Eagle Conservation Plan Guidance - April 2013

Eagle Conservation Plan Guidance April 2013.pdf

Eagle Take Permits and Fees, 50 CFR 22

Eagle Conservation Plan Guidance - April 2013

OMB: 1018-0167

Document [pdf]
Download: pdf | pdf
	U.S. Fish and Wildlife Service

	
	

Eagle Conservation Plan Guidance
Module 1 – Land-based Wind Energy
Version 2
	

	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	

Credit: Brian Millsap/USFWS

U.S. Fish and Wildlife Service
Division of Migratory Bird Management
April 2013

i

Disclaimer

This	Eagle	Conservation	Plan	Guidance	is	not	intended	to,	nor	
shall	it	be	construed	to,	limit	or	preclude	the	Service	from	
exercising	its	authority	under	any	law,	statute,	or	regulation,	
or	from	taking	enforcement	action	against	any	individual,	
company,	or	agency.		This	Guidance	is	not	meant	to	relieve	
any	individual,	company,	or	agency	of	its	obligations	to	
comply	with	any	applicable	Federal,	state,	tribal,	or	local	
laws,	statutes,	or	regulation.		This	Guidance	by	itself	does	not	
prevent	the	Service	from	referring	cases	for	prosecution,	
whether	a	company	has	followed	it	or	not.	

ii

EXECUTIVE SUMMARY
1. Overview
Of	all	America’s	wildlife,	eagles	hold	perhaps	the	most	revered	place	in	our	national	history	and	
culture.		The	United	States	has	long	imposed	special	protections	for	its	bald	and	golden	eagle	
populations.	Now,	as	the	nation	seeks	to	increase	its	production	of	domestic	energy,	wind	energy	
developers	and	wildlife	agencies	have	recognized	a	need	for	specific	guidance	to	help	make	wind	
energy	facilities	compatible	with	eagle	conservation	and	the	laws	and	regulations	that	protect	
eagles.	
	
To	meet	this	need,	the	U.S.	Fish	and	Wildlife	Service	(Service)	has	developed	the	Eagle	Conservation	
Plan	Guidance	(ECPG).		This	document	provides	specific	in‐depth	guidance	for	conserving	bald	and	
golden	eagles	in	the	course	of	siting,	constructing,	and	operating	wind	energy	facilities.		The	ECPG	
guidance	supplements	the	Service’s	Land‐Based	Wind	Energy	Guidelines	(WEG).		WEG	provides	a	
broad	overview	of	wildlife	considerations	for	siting	and	operating	wind	energy	facilities,	but	does	
not	address	the	in‐depth	guidance	needed	for	the	specific	legal	protections	afforded	to	bald	and	
golden	eagles.		The	ECPG	fills	this	gap.	
	
Like	the	WEG,	the	ECPG	calls	for	wind	project	developers	to	take	a	staged	approach	to	siting	new	
projects.		Both	call	for	preliminary	landscape‐level	assessments	to	assess	potential	wildlife	
interactions	and	proceed	to	site‐specific	surveys	and	risk	assessments	prior	to	construction.		They	
also	call	for	monitoring	project	operations	and	reporting	eagle	fatalities	to	the	Service	and	state	and	
tribal	wildlife	agencies.	
	
Compliance	with	the	ECPG	is	voluntary,	but	the	Service	believes	that	following	the	guidance	will	
help	project	operators	in	complying	with	regulatory	requirements	and	avoiding	the	unintentional	
“take”	of	eagles	at	wind	energy	facilities,	and	will	also	assist	the	wind	energy	industry	in	providing	
the	biological	data	needed	to	support	permit	applications	for	facilities	that	may	pose	a	risk	to	
eagles.	

	
2. The Bald and Golden Eagle Protection Act
The	Bald	and	Golden	Eagle	Protection	Act	(BGEPA)	is	the	primary	law	protecting	eagles.		BGEPA	
prohibits	“take”	of	eagles	without	a	permit	(16	USC	668‐668c).		BGEPA	defines	“take”	to	include	
“pursue,	shoot	at,	poison,	wound,	kill,	capture,	trap,	collect,	molest	or	disturb,”	and	prohibits	take	of	
individuals	and	their	parts,	nests,	or	eggs.		The	Service	expanded	this	definition	by	regulation	to	
include	the	term	“destroy”	to	ensure	that	“take”	includes	destruction	of	eagle	nests.		The	term	
“disturb”	is	further	defined	by	regulation	as	“to	agitate	or	bother	a	bald	or	golden	eagle	to	a	degree	
that	causes,	or	is	likely	to	cause,….injury	to	an	eagle,	a	decrease	in	productivity,	or	nest	
abandonment”	(50	CFR	22.3).	

	
3. Risks to Eagles from Wind Energy Facilities	
Wind	energy	development	can	affect	eagles	in	a	variety	of	ways.		First,	eagles	can	be	killed	by	
colliding	with	structures	such	as	wind	turbines.	This	is	the	primary	threat	to	eagles	from	wind	
facilities,	and	the	ECPG	guidance	is	primarily	aimed	at	this	threat.		Second,	disturbance	from	pre‐
construction,	construction,	or	operation	and	maintenance	activities	might	disturb	eagles	at	
concentration	sites	or	and	result	in	loss	of	productivity	at	nearby	nests.		Third,	serious	disturbance	
or	mortality	effects	could	result	in	the	permanent	or	long	term	loss	of	a	nesting	territory.		
Additionally,	disturbances	near	important	eagle	use	areas	or	migration	concentration	sites	might	
stress	eagles	so	much	that	they	suffer	reproductive	failure	or	mortality	elsewhere,	to	a	degree	that	

iii
could	amount	to	prohibited	take.		All	of	these	impacts,	unless	properly	permitted,	are	violations	of	
BGEPA.	

	
4. Eagle Take Permits
The	Service	recognizes	that	wind	energy	facilities,	even	those	developed	and	operated	with	the	
utmost	effort	to	conserve	wildlife,	may	under	some	circumstances	result	in	the	“take”	of	eagles	
under	BGEPA.		However,	in	2009,	the	Service	promulgated	new	permit	rules	for	eagles	that	address	
this	issue	(50	CFR	22.26	and	22.27).	
	
Under	these	new	rules	the	Service	can	issue	permits	that	authorize	individual	instances	of	take	of	
bald	and	golden	eagles	when	the	take	is	associated	with,	but	not	the	purpose	of,	an	otherwise	lawful	
activity,	and	cannot	practicably	be	avoided.		The	regulations	also	authorize	permits	for	
“programmatic”	take,	which	means	that	instances	of	“take”	may	not	be	isolated,	but	may	recur.		The	
programmatic	take	permits	are	the	most	germane	permits	for	wind	energy	facilities.		However,	
under	these	regulations,	any	ongoing	or	programmatic	take	must	be	unavoidable	even	after	the	
implementation	of	advanced	conservation	practices	(ACPs).	
	
The	ECPG	is	written	to	guide	wind‐facility	projects	starting	from	the	earliest	conceptual	planning	
phase.		For	projects	already	in	the	development	or	operational	phase,	implementation	of	all	stages	
of	the	recommended	approach	in	the	ECPG	may	not	be	applicable	or	possible.		Project	developers	or	
operators	with	operating	or	soon‐to‐be	operating	facilities	and	who	are	interested	in	obtaining	a	
programmatic	eagle	take	permit	should	contact	the	Service.		The	Service	will	work	with	project	
developers	or	operators	to	determine	if	the	project	might	be	able	to	meet	the	permit	requirements	
in	50	CFR	22.26.		The	Service	may	recommend	that	the	developer	monitor	eagle	fatalities	and	
disturbance,	adopt	reasonable	measures	to	reduce	eagle	fatalities	from	historic	levels,	and	
implement	compensatory	mitigation.		Sections	of	the	ECPG	that	address	these	topics	are	relevant	to	
both	planned	and	operating	wind	facilities	(Appendices	E	and	F	in	particular).	Operators	of	wind	
projects	(and	other	activities)	that	were	in	operation	prior	to	2009	that	pose	a	risk	to	golden	eagles	
may	qualify	for	programmatic	eagle	take	permits	that	do	not	automatically	require	compensatory	
mitigation.	This	is	because	the	requirements	for	obtaining	programmatic	take	authorization	are	
designed	to	reduce	take	from	historic,	baseline	levels,	and	the	preamble	to	the	Eagle	Permit	Rule	
specified	that	unavoidable	take	remaining	after	implementation	of	avoidance	and	minimization	
measures	at	such	projects	would	not	be	subtracted	from	regional	eagle	take	thresholds.	

	
5. Voluntary Nature of the ECPG
Wind	project	operators	are	not	legally	required	to	seek	or	obtain	an	eagle	take	permit.		However,	
the	take	of	an	eagle	without	a	permit	is	a	violation	of	BGEPA,	and	could	result	in	prosecution.		The	
methods	and	approaches	suggested	in	the	ECPG	are	not	mandatory	to	obtain	an	eagle	take	permit.	
The	Service	will	accept	other	approaches	that	provide	the	information	and	data	required	by	the	
regulations.		The	ECP	can	be	a	stand‐alone	document,	or	part	of	a	larger	bird	and	bat	strategy	as	
described	in	the	WEG,	so	long	as	it	adequately	meets	the	regulatory	requirements	at	50	CFR	22.26	
to	support	a	permit	decision.		However,	Service	employees	who	process	eagle	take	permit	
applications	are	trained	in	the	methods	and	approaches	covered	in	the	ECPG.	Using	other	
methodologies	may	result	in	longer	application	processing	times.	

	
	6. Eagle Take Thresholds
Eagle	take	permits	may	be	issued	only	in	compliance	with	the	conservation	standards	of	BGEPA.		
This	means	that	the	take	must	be	compatible	with	the	preservation	of	each	species,	defined	(in	
USFWS	2009a)	as	“consistent	with	the	goal	of	stable	or	increasing	breeding	populations.”	

iv
	
To	ensure	that	any	authorized	“take”	of	eagles	does	not	exceed	this	standard,	the	Service	has	set	
regional	take	thresholds	for	each	species,	using	methodology	contained	in	the	National	
Environmental	Policy	Act	(NEPA)	Final	Environmental	Assessment	(FEA)	developed	for	the	new	
eagle	permit	rules	(USFWS	2009b).		The	Service	looked	at	regional	populations	of	eagles	and	set	
take	thresholds	for	each	species	(upper	limits	on	the	number	of	eagle	mortalities	that	can	be	
allowed	under	permit	each	year	in	these	regional	management	areas).	
	
The	analysis	identified	take	thresholds	greater	than	zero	for	bald	eagles	in	most	regional	
management	areas.		However,	the	Service	determined	that	golden	eagle	populations	might	not	be	
able	to	sustain	any	additional	unmitigated	mortality	at	that	time,	and	set	the	thresholds	for	this	
species	at	zero	for	all	regional	populations.		This	means	that	any	new	authorized	“take”	of	golden	
eagles	must	be	at	least	equally	offset	by	compensatory	mitigation	(specific	conservation	actions	to	
replace	or	offset	project‐induced	losses).	
	
The	Service	also	put	in	place	measures	to	ensure	that	local	eagle	populations	are	not	depleted	by	
take	that	would	be	otherwise	regionally	acceptable.		The	Service	specified	that	take	rates	must	be	
carefully	assessed,	both	for	individual	projects	and	for	the	cumulative	effects	of	other	activities	
causing	take,	at	the	scale	of	the	local‐area	eagle	population	(a	population	within	a	distance	of	43	
miles	for	bald	eagles	and	140	miles	for	golden	eagles).	This	distance	is	based	on	the	median	
distance	to	which	eagles	disperse	from	the	nest	where	they	are	hatched	to	where	they	settle	to	
breed.	
	
The	Service	identified	take	rates	of	between	1	and	5	percent	of	the	total	estimated	local‐area	eagle	
population		as	significant,	with	5	percent	being	at	the	upper	end	of	what	might	be	appropriate	
under	the	BGEPA	preservation	standard,	whether	offset	by	compensatory	mitigation	or	not.		
Appendix	F	provides	a	full	description	of	take	thresholds	and	benchmarks,	and	provides	suggested	
tools	for	evaluating	how	these	apply	to	individual	projects.	

	
7. An Approach for Developing and Evaluating Eagle ACPs
Permits	for	eagle	take	at	wind‐energy	facilities	are	programmatic	in	nature	as	they	will	authorize	
recurring	take	rather	than	isolated	incidences	of	take.		For	programmatic	take	permits,	the	
regulations	require	that	any	authorized	take	must	be	unavoidable	after	the	implementation	of	
advanced	conservation	practices	(ACPs).		ACPs	are	defined	as	“scientifically	supportable	measures	
that	are	approved	by	the	Service	and	represent	the	best	available	techniques	to	reduce	eagle	
disturbance	and	ongoing	mortalities	to	a	level	where	remaining	take	is	unavoidable”	(50	CFR	22.3).	
	
Because	the	best	information	currently	available	indicates	there	are	no	conservation	measures	that	
have	been	scientifically	shown	to	reduce	eagle	disturbance	and	blade‐strike	mortality	at	wind	
projects,	the	Service	has	not	currently	approved	any	ACPs	for	wind	energy	projects.	
	
The	process	of	developing	ACPs	for	wind	energy	facilities	has	been	hampered	by	the	lack	of	
standardized	scientific	study	of	potential	ACPs.		The	Service	has	determined	that	the	best	way	to	
obtain	the	needed	scientific	information	is	to	work	with	industry	to	develop	ACPs	for	wind	projects	
as	part	of	an	adaptive‐management	regime	and	comprehensive	research	program	tied	to	the	
programmatic‐take‐permit	process.		In	this	scenario,	ACPs	will	be	implemented	at	operating	wind	
facilities	with	an	eagle	take	permit	on	an	“experimental”	basis	(the	ACPs	are	considered	
experimental	because	they	would	not	currently	meet	the	definition	of	an	ACP	in	the	eagle	permit	
regulation).	The	experimental	ACPs	would	be	scientifically	evaluated	for	their	effectiveness,	as	
described	in	detail	in	this	document,	and	based	on	the	results	of	these	studies,	could	be	modified	in	

v
an	adaptive	management	regime.		This	approach	will	provide	the	needed	scientific	information	for	
the	future	establishment	of	formal	ACPs,	while	enabling	wind	energy	facilities	to	move	forward	in	
the	interim.	
	
Despite	the	current	lack	of	formally	approved	ACPs,	there	may	be	other	conservation	measures	
based	on	the	best	available	scientific	information	that	should	be	applied	as	a	condition	on	
programmatic	eagle	take	permits	for	wind‐energy	facilities.		A	project	developer	or	operator	will	be	
expected	to	implement	any	reasonable	avoidance	and	minimization	measures	that	may	reduce	take	
of	eagles	at	a	project.		In	addition,	the	Service	and	the	project	developer	or	operator	will	identify	
other	site‐specific	and	possibly	turbine‐specific	factors	that	may	pose	risks	to	eagles,	and	agree	on	
the	experimental	ACPs	to	avoid	and	minimize	those	risks.	Unless	the	Service	determines	that	there	
is	a	reasonable	scientific	basis	to	implement	the	experimental	ACPs	up	front	(or	it	is	otherwise	
advantageous	to	the	developer	to	do	so),	we	recommend	that	such	measures	be	deferred	until	such	
time	as	there	is	eagle	take	at	the	facility	or	the	Service	determines	that	the	circumstances	and	
evidence	surrounding	the	take	or	risk	of	take	suggest	the	experimental	ACPs	might	be	warranted.		
The	programmatic	eagle	take	permit	would	specify	the	experimental	ACPs,	if	circumstances	
warrant,	and	the	permit	would	be	conditioned	on	the	project	operator’s	agreement	to	implement	
and	monitor	the	experimental	ACPs.	
	
Because	the	ACPs	would	be	experimental,	the	Service	recommends	that	they	be	subject	to	a	cost	cap	
that	the	Service	and	the	project	developer	or	operator	would	establish	as	part	of	the	initial	
agreement	before	issuance	of	an	eagle	permit.		This	would	provide	financial	certainty	as	to	what	
maximum	costs	of	such	measures	might	be.		The	amount	of	the	cap	should	be	proportional	to	
overall	risk.	
	
As	the	results	from	monitoring	experimental	ACPs	across	a	number	of	facilities	accumulate	and	are	
analyzed,	scientific	information	in	support	of	certain	experimental	ACPs	may	accrue,	whereas	other	
ACPs	may	show	little	value	in	reducing	take.		If	the	Service	determines	that	the	available	science	
demonstrates	an	experimental	ACP	is	effective	in	reducing	eagle	take,	the	Service	will	formally	
approve	that	ACP	and	require	its	implementation	up	front	on	new	projects	when	and	where	
warranted.	
	
As	the	ECPG	evolves,	the	Service	will	not	expect	project	developers	or	operators	to	retroactively	
redo	analyses	or	surveys	using	the	new	approaches.		The	adaptive	approach	to	the	ECPG	should	not	
deter	project	developers	or	operators	from	using	the	ECPG	immediately.	

	
8. Mitigation Actions to Reduce Effects on Eagle Populations
Where	wind	energy	facilities	cannot	avoid	taking	eagles	and	eagle	populations	are	not	healthy	
enough	to	sustain	additional	mortality,	applicants	must	reduce	the	unavoidable	mortality	to	a	no‐
net‐loss	standard	for	the	duration	of	the	permitted	activity.		No‐net‐loss	means	that	these	actions	
either	reduce	another	ongoing	form	of	mortality	to	a	level	equal	to	or	greater	than	the	unavoidable	
mortality,	or	lead	to	an	increase	in	carrying	capacity	that	allows	the	eagle	population	to	grow	by	an	
equal	or	greater	amount.		Actions	to	reduce	eagle	mortality	or	increase	carrying	capacity	to	this	no‐
net‐loss	standard	are	known	as	“compensatory	mitigation”	in	the	ECPG.		Examples	of	compensatory	
mitigation	activities	might	include	retrofitting	power	lines	to	reduce	eagle	electrocutions,	removing	
road‐killed	animals	along	roads	where	vehicles	hit	and	kill	scavenging	eagles,	or	increasing	prey	
availability.	
	
The	Service	and	the	project	developer	or	operator	seeking	a	programmatic	eagle	take	permit	
should	agree	on	the	number	of	eagle	fatalities	to	mitigate	and	what	actions	will	be	taken	if	actual	

vi
eagle	fatalities	differ	from	the	predicted	number.		The	compensatory	mitigation	requirement	and	
trigger	for	adjustment	should	be	specified	in	the	permit.		If	the	procedures	recommended	in	the	
ECPG	are	followed,	there	should	not	be	a	need	for	additional	compensatory	mitigation.		However,	if	
other,	less	risk‐averse	models	are	used	to	estimate	fatalities,	underestimates	might	be	expected	and	
the	permit	should	specify	the	threshold(s)	of	take	that	would	trigger	additional	actions	and	the	
specific	mitigation	activities	that	might	be	implemented.	
	
Additional	types	of	mitigation	such	as	preserving	habitat	–	actions	that	would	not	by	themselves	
lead	to	increased	numbers	of	eagles	but	would	assist	eagle	conservation	–	may	also	be	advised	to	
offset	other	detrimental	effects	of	permits	on	eagles.		Compensatory	mitigation	is	further	discussed	
below	(Stage	4	–	Avoidance	and	Minimization	of	Risk	and	Compensatory	Mitigation).	

	
9. Relationship of Eagle Guidelines (ECPG) to the Wind Energy Guidelines (WEG)
The	ECPG	is	intended	to	be	implemented	in	conjunction	with	other	actions	recommended	in	the	
WEG	that	assess	impacts	to	wildlife	species	and	their	habitats.		The	WEG	recommends	a	five‐tier	
process	for	such	assessments,	and	the	ECPG	fits	within	that	framework.	The	ECPG	focuses	on	just	
eagles	to	facilitate	collection	of	information	that	could	support	an	eagle	take	permit	decision.		The	
ECPG	uses	a	five‐stage	approach	like	the	WEG;	the	relationship	between	the	ECPG	stages	and	the	
WEG	tiers	is	shown	in	Fig.	1.	
	
Tiers	1	and	2	of	the	WEG	(Stage	1	of	the	ECPG)	could	provide	sufficient	evidence	to	demonstrate	
that	a	project	poses	very	low	risk	to	eagles.		Provided	this	assessment	is	robust,	eagles	may	not	
warrant	further	consideration	in	subsequent	WEG	tiers,	and	Stages	2	through	5	of	the	ECPG	and	
pursuit	of	an	eagle	take	permit	might	be	unnecessary.		A	similar	conclusion	could	be	reached	at	the	
end	of	Stage	2,	3,	or	4.		In	such	cases,	if	unpermitted	eagle	take	subsequently	occurs,	the	wind	
project	proponent	should	consult	with	the	U.S.	Fish	and	Wildlife	Service	to	determine	how	to	
proceed,	possibly	by	obtaining	an	eagle	take	permit.	
	
The	following	sections	describe	the	general	approach	envisioned	for	assessing	wind	project	impacts	
to	eagles	(also	see	the	Stage	Overview	Table	at	the	end	of	the	Executive	Summary).	

	
Tiers 1 and 2 of the WEG, Stage 1 of the ECPG
Tier	1	of	the	WEG	is	the	preliminary	site	evaluation	(landscape‐scale	screening	of	possible	
project	sites).	Tier	2	is	site	characterization	(broad	characterization	of	one	or	more	
potential	project	sites).		These	correspond	with	Stage	1	of	the	ECPG,	the	site‐assessment	
stage.		As	part	of	the	Tiers	1	and	2	process,	project	developers	should	carry	out	Stage	1	of	
the	ECPG	and	evaluate	broad	geographic	areas	to	assess	the	relative	importance	of	various	
areas	to	resident	breeding	and	non‐breeding	eagles,	and	to	migrant	and	wintering	eagles.	
During	Stage	1,	the	project	developer	or	operator	should	gather	existing	information	from	
publicly	available	literature,	databases,	and	other	sources,	and	use	those	data	to	judge	the	
appropriateness	of	various	potential	project	sites,	balancing	suitability	for	development	
with	potential	risk	to	eagles.	
	
To	increase	the	probability	of	meeting	the	regulatory	requirements	for	a	programmatic	take	
permit,	biological	advice	from	the	Service	and	other	jurisdictional	wildlife	agencies	should	
be	requested	as	early	as	possible	in	the	developer's	planning	process	and	should	be	as	
inclusive	as	possible	to	ensure	all	issues	are	being	addressed	at	the	same	time	and	in	a	
coordinated	manner.		Ideally,	consultation	with	the	Service,	and	state	and	tribal	wildlife	

vii
agencies	is	done	before	wind	developers	make	any	substantial	financial	commitment	or	
finalize	lease	agreements.		

	
Tier 3 of the WEG, Stages 2, 3, and 4 of the ECPG
During	Tier	3	of	the	WEG,	a	developer	conducts	field	studies	to	document	wildlife	use	and	
habitat	at	the	project	site	and	predict	project	impacts.		These	site‐specific	studies	are	critical	
to	evaluating	potential	impacts	to	all	wildlife	including	eagles.		The	developer	and	the	
Service	would	use	the	information	collected	to	support	an	eagle	take	permit	application,	
should	the	developer	seek	a	permit.		As	part	of	Tier	3,	the	ECPG	recommends	project	
developers	or	operators	implement	three	stages	of	assessment:	
 Stage	2	‐	site‐specific	surveys	and	assessments;	
 Stage	3	‐	predicting	eagle	fatalities;	and	
 Stage	4	‐	avoidance	and	minimization	of	risk	and	compensatory	mitigation.	

	
Stage 2 – Site Specific Surveys and Assessments
During	Stage	2	the	Service	recommends	the	project	developer	collect	quantitative	
data	through	scientifically	rigorous	surveys	designed	to	assess	the	potential	risk	of	
the	proposed	project	to	eagles.		The	Service	recommends	collecting	information	that	
will	allow	estimation	of	the	eagle	exposure	rate	(eagle‐minutes	flying	within	the	
project	footprint	per	hour	per	kilometer2),	as	well	as	surveys	sufficient	to	determine	
if	important	eagle	use	areas	or	migration	concentration	sites	are	within	or	in	close	
proximity	to	the	project	footprint	(see	Appendix	C).		In	the	case	of	small	wind	
projects	(one	utility‐scale	turbine	or	a	few	small	turbines),	the	project	developer	
should	consider	the	proximity	of	eagle	nesting	and	roosting	sites	to	a	proposed	
project	and	discuss	the	results	of	the	Stage	1	assessment	with	the	Service	to	
determine	if	Stage	2	surveys	are	necessary.		In	many	cases	the	hazardous	area	
associated	with	such	projects	will	be	small	enough	that	Stage	2	surveys	will	not	be	
necessary.	

	
Stage 3 – Predicting Eagle Fatalities
In	Stage	3,	the	Service	and	project	developers	or	operators	use	data	from	Stage	2	in	
models	to	predict	eagle	risk	expressed	as	the	average	number	of	fatalities	per	year	
extrapolated	to	the	tenure	of	the	permit.		These	models	can	compare	alternative	
siting,	construction,	and	operational	scenarios,	a	useful	feature	in	constructing	
hypotheses	regarding	predicted	effects	of	conservation	measures	and	experimental	
ACPs.		The	Service	encourages	project	developers	or	operators	to	use	the	
recommended	pre‐construction	survey	protocol	in	this	ECPG	in	Stage	2	to	help	
inform	our	predictive	models	in	Stage	3.		If	Service‐recommended	survey	protocols	
are	used,	this	risk	assessment	can	be	greatly	facilitated	using	model	tools	available	
from	the	Service.		If	project	developers	or	operators	use	other	forms	of	information	
for	the	Stage	2	assessment,	they	will	need	to	fully	describe	those	methods	and	the	
analysis	used	for	the	eagle	risk	assessment.		The	Service	will	require	more	time	to	
evaluate	and	review	the	data	because,	for	example,	the	Service	will	need	to	compare	
the	results	of	the	project	developer	or	operator’s	eagle	risk	assessment	with	
predictions	from	our	models.		If	the	results	differ,	we	will	work	with	the	project	
developers	or	operators	to	determine	which	model	results	are	most	appropriate	for	
the	Service’s	eventual	permitting	decisions.	
	

viii
The	Service	and	project	developers	or	operators	also	evaluate	Stage	2	data	to	
determine	whether	disturbance	take	is	likely,	and	if	so,	at	what	level.		Any	loss	of	
production	that	may	stem	from	disturbance	should	be	added	to	the	fatality	rate	
prediction	for	the	project.		The	risk	assessments	at	Stage	2	and	Stage	3	are	
consistent	with	developing	the	information	necessary	to	assess	the	efficacy	of	
conservation	measures,	and	to	develop	the	monitoring	required	by	the	permit	
regulations	at	50	CFR	22.26(c)(2).	

	
Stage 4 - Avoidance and Minimization of Risk and Compensatory Mitigation
In	Stage	4	the	information	gathered	should	be	used	by	the	project	developer	or	
operator	and	the	Service	to	determine	potential	conservation	measures	and	ACPs	(if	
available)	to	avoid	or	minimize	predicted	risks	at	a	given	site	(see	Appendix	E).		The	
Service	will	compare	the	initial	predictions	of	eagle	mortality	and	disturbance	for	
the	project	with	predictions	that	take	into	account	proposed	and	potential	
conservation	measures	and	ACPs,	once	developed	and	approved,	to	determine	if	the	
project	developer	or	operator	has	avoided	and	minimized	risks	to	the	maximum	
degree	achievable,	thereby	meeting	the	requirements	for	programmatic	permits	
that	remaining	take	is	unavoidable.		Additionally,	the	Service	will	use	the	
information	provided	along	with	other	data	to	conduct	a	cumulative	effects	analysis	
to	determine	if	the	project’s	impacts,	in	combination	with	other	permitted	take	and	
other	known	factors,	are	at	a	level	that	exceed	the	established	thresholds	or	
benchmarks	for	eagle	take	at	the	regional	and	local‐area	scales.		This	final	eagle	risk	
assessment	is	completed	at	the	end	of	Stage	4	after	application	of	conservation	
measures	and	ACPs	(if	available)	along	with	a	plan	for	compensatory	mitigation	if	
required.	
	
The	eagle	permit	process	requires	compensatory	mitigation	if	conservation	
measures	do	not	remove	the	potential	for	take,	and	the	projected	take	exceeds	
calculated	thresholds	for	the	eagle	management	unit	in	which	the	project	is	located.		
However,	there	may	also	be	other	situations	in	which	compensatory	mitigation	is	
necessary.	The	following	guidance	applies	to	those	situations	as	well.	
	
Compensatory	mitigation	can	address	pre‐existing	causes	of	eagle	mortality	(such	as	
eagle	electrocutions	from	power	poles)	or	it	can	address	increasing	the	carrying	
capacity	of	the	eagle	population	in	the	affected	eagle	management	unit.		However,	
there	needs	to	be	a	credible	analysis	that	supports	the	conclusion	that	implementing	
the	compensatory	mitigation	action	will	achieve	the	desired	beneficial	offset	in	
mortality	or	carrying	capacity.	
	
For	new	wind	development	projects,	if	compensatory	mitigation	is	necessary,	the	
compensatory	mitigation	action	(or	a	verifiable,	legal	commitment	to	such	
mitigation)	will	be	required	up	front	before	project	operations	begin	because	
projects	must	meet	the	statutory	eagle	preservation	standard	before	the	Service	
may	issue	a	permit.		For	operating	projects,	compensatory	mitigation	should	be	
applied	from	the	start	of	the	permit	period,	not	retroactively	from	the	time	the	
project	began.		The	initial	compensatory	mitigation	effort	should	be	sufficient	to	
offset	the	predicted	number	of	eagle	fatalities	per	year	for	five	years.	No	later	than	
at	the	end	of	the	five	year	period,	the	Service	and	the	project	operator	will	compare	
the	predicted	annual	take	estimate	to	the	realized	take	based	on	post‐construction	
monitoring.		If	the	triggers	identified	in	the	permit	for	adjustment	of	compensatory	

ix
mitigation	are	met,	those	adjustments	should	be	implemented.	In	the	case	where	the		
observed	take	was	less	than	estimated,	the	permittee	will	receive	a	credit	for	the	
excess	compensation	(the	difference	between	the	actual	mean	and	the	number	
compensated	for)	that	can	be	applied	to	other	take	(either	by	the	permittee	or	other	
permitted	individuals	at	his/her	discretion)	within	the	same	eagle	management	
unit.	The	Service,	in	consultation	with	the	permittee,	will	determine	compensatory	
mitigation	for	future	years	for	the	project	at	this	point,	taking	into	account	the	
observed	levels	of	mortality	and	any	reduction	in	that	mortality	that	is	expected	
based	on	implementation	of	additional	experimental	conservation	measures	and	
ACPs.		Monitoring	using	the	best	scientific	and	practicable	methods	available	should	
be	included	to	determine	the	effectiveness	of	the	resulting	compensatory	mitigation	
efforts.		The	Service	will	modify	the	compensatory	mitigation	process	to	adapt	to	
any	improvements	in	our	knowledge	base	as	new	data	become	available.	
	
At	the	end	of	Stage	4,	all	the	materials	necessary	to	satisfy	the	regulatory	
requirements	to	support	a	permit	application	should	be	available.		While	a	project	
operator	can	submit	a	permit	application	at	any	time,	the	Service	can	only	begin	the	
formal	process	to	determine	whether	a	programmatic	eagle	take	permit	can	be	
issued	after	completion	of	Stage	4.		Ideally,	National	Environmental	Policy	Act	
(NEPA)	and	National	Historic	Preservation	Act	(NHPA	)	analyses	and	assessments	
will	already	be	underway,	but	if	not,	Stage	4	should	include	necessary	NEPA	
analysis,	NHPA	compliance,	coordination	with	other	jurisdictional	agencies,	and	
tribal	consultation.	

Tier 4 and 5 of the WEG, Stage 5 of the ECPG
If	the	Service	issues	an	eagle	take	permit	and	the	project	goes	forward,	project	operators	
will	conduct	post‐construction	surveys	to	collect	data	that	can	be	compared	with	the	pre‐
construction	risk‐assessment	predictions	for	eagle	fatalities	and	disturbance.	The	
monitoring	protocol	should	include	validated	techniques	for	assessing	both	mortality	and	
disturbance	effects,	and	they	must	meet	the	permit‐condition	requirements	at	50	CFR	
22.26(c)(2).		In	most	cases,	intensive	monitoring	will	be	conducted	for	at	least	the	first	two	
years	after	permit	issuance,	followed	by	less	intense	monitoring	for	up	to	three	years	after	
the	expiration	date	of	the	permit.	Project	developers	or	operators	should	use	the	post‐
construction	survey	protocols	included	or	referenced	in	this	ECPG,	but	we	will	consider	
other	monitoring	protocols	provided	by	permit	applicants	though	the	process	will	likely	
take	longer	than	if	familiar	approaches	were	used.		The	Service	will	use	the	information	
from	post‐construction	monitoring	in	a	meta‐analysis	framework	to	weight	and	improve	
pre‐construction	predictive	models.	
	
Additionally	in	Stage	5,	the	Service	and	project	developers	or	operators	should	use	the	post‐
construction	monitoring	data	to	(1)	assess	whether	compensatory	mitigation	is	adequate,	
excessive,	or	deficient	to	offset	observed	mortality,	and	make	adjustments	accordingly;	and	
(2)	explore	operational	changes	that	might	be	warranted	at	a	project	after	permitting	to	
reduce	observed	mortality	and	meet	permit	requirements.	

	
10. Site Categorization Based on Mortality Risk to Eagles
Beginning	at	the	end	of	Stage	1,	and	continuing	at	the	end	of	Stages	2,	3,	and	4,	we	recommend	the	
approach	outlined	below	be	used	to	assess	the	likelihood	that	a	wind	project	will	take	eagles,	and	if	

x
so,	that	the	project	will	meet	standards	in	50	CFR	22.26	for	issuance	of	a	programmatic	eagle	take	
permit.	

	
Category 1 – High risk to eagles, potential to avoid or mitigate impacts is low
A	project	is	in	this	category	if	it:	
(1)	has	an	important	eagle‐use	area	or	migration	concentration	site	within	the	project	
footprint;	or	
(2)	has	an	annual	eagle	fatality	estimate	(average	number	of	eagles	predicted	to	be	
taken	annually)		>	5%	of	the	estimated	local‐area	population	size;	or	
(3)	causes	the	cumulative	annual	take	for	the	local‐area	population	to	exceed	5%	of	the	
estimated	local‐area	population	size.	
	
In	addition,	projects	that	have	eagle	nests	within	½	the	mean	project‐area	inter‐nest	
distance	of	the	project	footprint	should	be	carefully	evaluated.		If	it	is	likely	eagles	
occupying	these	territories	use	or	pass	through	the	project	footprint,	category	1	designation	
may	be	appropriate.	
	
Projects	or	alternatives	in	category	1	should	be	substantially	redesigned	to	at	least	meet	the	
category	2	criteria.		The	Service	recommends	that	project	developers	not	build	projects	at	
sites	in	category	1	because	the	project	would	likely	not	meet	the	regulatory	requirements.	
The	recommended	approach	for	assessing	the	percentage	of	the	local‐area	population	
predicted	to	be	taken	is	described	in	Appendix	F.	

	
Category 2 – High or moderate risk to eagles, opportunity to mitigate impacts
A	project	is	in	this	category	if	it:	
(1)	has	an	important	eagle‐use	area	or	migration	concentration	site	within	the	project	
area	but	not	in	the	project	footprint;	or	
(2)	has	an	annual	eagle	fatality	estimate	between	0.03	eagles	per	year	and	5%	of	the	
estimated	local‐area	population	size;	or	
(3)	causes	cumulative	annual	take	of	the	local‐area	population	of		less	than	5%	of	the	
estimated	local‐area	population	size.	
	
Projects	in	this	category	will	potentially	take	eagles	at	a	rate	greater	than	is	consistent	with	
maintaining	stable	or	increasing	populations,	but	the	risk	might	be	reduced	to	an	acceptable	
level	through	a	combination	of	conservation	measures	and	reasonable	compensatory	
mitigation.		These	projects	have	a	risk	of	ongoing	take	of	eagles,	but	this	risk	can	be	
minimized.		For	projects	in	this	category	the	project	developer	or	operator	should	prepare	
an	Eagle	Conservation	Plan	(ECP)	or	similar	plan	to	document	meeting	the	regulatory	
requirements	for	a	programmatic	permit.		The	ECP	or	similar	document	can	be	a	stand‐
alone	document,	or	part	of	a	larger	bird	and	bat	strategy	as	described	in	the	WEG,	so	long	as	
it	adequately	meets	the	regulatory	requirements	at	50	CFR	22.26	to	support	a	permit	
decision.		For	eagle	management	populations	where	take	thresholds	are	set	at	zero,	the	
conservation	measures	in	the	ECP	should	include	compensatory	mitigation	and	must	result	
in	no‐net‐loss	to	the	breeding	population	to	be	compatible	with	the	permit	regulations.		This	
does	not	apply	to	golden	eagles	east	of	the	100th	meridian,	for	which	no	non‐emergency	
take	can	presently	be	authorized	(USFWS	2009b).	

	
Category 3 – Minimal risk to eagles
A	project	is	in	this	category	if	it:	

xi
(1)	has	no	important	eagle	use	areas	or	migration	concentration	sites	within	the	project	
area;	and	
(2)	has	an	annual	eagle	fatality	rate	estimate	of	less	than	0.03;	and	
(3)	causes	cumulative	annual	take	of	the	local‐area	population	of	less	than	5%	of	the	
estimated	local‐area	population	size.	
	
Projects	in	category	3	pose	little	risk	to	eagles	and	may	not	require	or	warrant	eagle	take	
permits,	but	that	decision	should	be	made	in	coordination	with	the	Service.		Still,	a	project	
developer	or	operator	may	wish	to	create	an	ECP	or	similar	document	or	strategy	that	
documents	the	project’s	low	risk	to	eagles,	and	outlines	mortality	monitoring	for	eagles	and	
a	plan	of	action	if	eagles	are	taken	during	project	construction	or	operation.		This	would	
enable	the	Service	to	provide	a	permit	to	allow	a	de	minimis	amount	of	take	if	the	project	
developer	or	operator	wished	to	obtain	such	a	permit.	
	
The	risk	category	of	a	project	can	potentially	change	as	a	result	of	additional	site‐specific	analyses	
and	application	of	measures	to	reduce	the	risk.		For	example,	a	project	may	appear	to	be	in	category	
2	as	a	result	of	Stage	1	analyses,	but	after	collection	of	site‐specific	information	in	Stage	2	it	might	
become	clear	it	is	a	category	1	project.		If	a	project	cannot	practically	be	placed	in	one	of	these	
categories,	the	project	developer	or	operator	and	the	Service	should	work	together	to	determine	if	
the	project	can	meet	programmatic	eagle	take	permitting	requirements	in	50	CFR	22.26	and	22.27.		
Projects	should	be	placed	in	the	highest	category	(with	category	1	being	the	highest)	in	which	one	
or	more	of	the	criteria	are	met.	

	
11. Addressing Uncertainty

There	is	substantial	uncertainty	surrounding	the	risk	of	wind	projects	to	eagles,	and	of	ways	to	
minimize	that	risk.		For	this	reason,	the	Service	stresses	that	it	is	very	important	not	to	
underestimate	eagle	fatality	rates	at	wind	facilities.		Overestimates,	once	confirmed,	can	be	adjusted	
downward	based	on	post‐construction	monitoring	information	with	no	consequence	to	eagle	
populations.		Project	developers	or	operators	can	trade	or	be	credited	for	excess	compensatory	
mitigation,	and	debits	to	regional	and	local‐area	eagle‐take	thresholds	and	benchmarks	can	be	
adjusted	downwards	to	reflect	actual	fatality	rates.		However,	the	options	for	addressing	
underestimated	fatality	rates	are	extremely	limited,	and	pose	either	potential	hardships	for	wind	
developers	or	significant	risks	to	eagle	populations.	
	
Our	long‐term	approach	for	moving	forward	in	the	face	of	this	uncertainty	is	to	implement	eagle	
take	permitting	in	a	formal	adaptive	management	framework.	The	Service	anticipates	four	specific	
sets	of	adaptive	management	decisions:	(1)	adaptive	management	of	wind	project	siting	and	design	
recommendations;	(2)	adaptive	management	of	wind	project	operations;	(3)	adaptive	management	
of	compensatory	mitigation;	and	(4)	adaptive	management	of	population‐level	take	thresholds.		
These	are	discussed	in	more	detail	in	Appendix	A.		The	adaptive	management	process	will	depend	
heavily	on	pre‐	and	post‐construction	data	from	individual	projects,	but	analyses,	assessment,	and	
model	evaluation	will	rely	on	data	pooled	over	many	individual	wind	projects.		Learning	
accomplished	through	adaptive	management	will	be	rapidly	incorporated	into	the	permitting	
process	so	that	the	regulatory	process	adjusts	in	proportion	to	actual	risk.	

12. Interaction with the Service
The	Service	encourages	early,	frequent	and	thorough	coordination	between	project	developers	or	
operators	and	Service	and	other	jurisdictional‐agency	employees	as	they	implement	the	tiers	of	the	
WEG,	and	the	related	Stages	of	the	ECPG.		Close	coordination	will	aid	the	refinement	of	the	

xii
modeling	process	used	to	predict	fatalities,	as	well	as	the	post‐construction	monitoring	to	evaluate	
those	models.		We	anticipate	the	ECPG	and	the	recommended	methods	and	metrics	will	evolve	as	
the	Service	and	project	developers	or	operators	learn	together.		The	Service	has	created	a	cross‐
program,	cross‐regional	team	of	biologists	who	will	work	jointly	on	eagle‐programmatic‐take	
permit	applications	to	help	ensure	consistency	in	administration	and	application	of	the	Eagle	
Permit	Rule.		This	close	coordination	and	interaction	is	especially	important	as	the	Service	
processes	the	first	few	programmatic	eagle	take	permit	applications.	
	
The	Service	will	continue	to	refine	this	ECPG	with	input	from	all	stakeholders	with	the	objective	of	
maintaining	stable	or	increasing	breeding	populations	of	both	bald	and	golden	eagles	while	
simultaneously	developing	science‐based	eagle‐take	regulations	and	procedures	that	are	
appropriate	to	the	risk	associated	with	each	wind	energy	project.	
	

	
	
Stage Overview Table - Overview of staged approach to developing an Eagle Conservation Plan as
described in the ECPG. Stages are in chronological order. Stage 5 would only be applicable in cases where a
permit was issued at the end of Stage 4.

	
Stage	

Objective	

Actions	

Data	Sources	

1	

At	the	landscape	level,	identify	
Broad,	landscape‐scale	
potential	wind	facility	locations	
evaluation.	
with	manageable	risk	to	eagles.	

Technical	literature,	agency	files,	
on‐line	biological	databases,	data	
from	nearby	projects,	industry	
reports,	geodatabases,	experts.	

2	

Obtain	site‐specific	data	to	
predict	eagle	fatality	rates	and	
disturbance	take	at	wind‐facility	
sites	that	pass	Stage	1	
assessment.		Investigate	other	
aspects	of	eagle	use	to	consider	
assessing	distribution	of	
occupied	nests	in	the	project	
area,	migration,	areas	of	
seasonal	concentration,	and	
intensity	of	use	across	the	
project	footprint.	

Site‐specific	surveys	and	
intensive	observation	to	
determine	eagle	exposure	rate	
and	distribution	of	use	in	the	
project	footprint,	plus	locations	
of	occupied		eagle	nests,	
migration	corridors	and	
stopover	sites,	foraging	
concentration	areas,	and	
communal	roosts	in	the	project	
area.	

Project	footprint:	800‐m	radius	
point	count	surveys	and	
utilization	distribution	studies.		
Project	area:	nest	surveys,	
migration	counts	at	likely	
topographic	features,	
investigation	of	use	of	potential	
roost	sites	and	of	areas	of	high	
prey	availability.		Ideally	
conducted	for	no	less	than	2	
years	pre‐construction.	

3	

As	part	of	pre‐construction	
monitoring	and	assessment,	
estimate	the	fatality	rate	of	
eagles	for	the	facility	evaluated	
in	Stage	2,	excluding	possible	
additions	of	conservation	
measures	and	advanced	
conservation	practices	(ACPs).		
Consider	possible	disturbance	
effects.	

Use	the	exposure	rate	derived	
from	Stage	2	data	in	Service‐
provided	models	to	predict	the	 Point	count,	nest,	and	eagle	
annual	eagle	fatality	rate	for	the	 concentration	area	data	from	
Stage	2.	
project.		Determine	if	
disturbance	effects	are	likely	and	
what	they	might	be.	

xiii
Stage	

4	

Objective	

	

	

Data	Sources	

Re‐run	fatality	prediction	models	
with	risk	adjusted	to	reflect	
As	part	of	the	pre‐construction	
application	of	conservation	
assessment,	identify	and	
measures	and	ACPs	to	determine	
Fatality	estimates	before	and	
evaluate	conservation	measures	
fatality	estimate	(80%	upper	
after	application	of	conservation	
and	ACPs	that	might	avoid	or	
confidence	limit	or	equivalent).		
measures	and	ACPs,	using	point	
minimize	fatalities	and	
Calculate	required	
count	data	from	Stage	2.		
disturbance	effects	identified	in	
compensatory	mitigation	
Estimates	of	disturbance	effects	
Stage	3.		When	necessary,	
amount	where	necessary,	
from	Stage	3.	
identify	compensatory	
considering	disturbance	effects,	
mitigation	to	reduce	predicted	
if	any.		Identify	actions	needed	to	
take	to	a	no‐net‐loss	standard.	
accomplish	compensatory	
mitigation.	

Determine	if	regulatory	
Permit	
requirements	for	issuance	of	a	
Decision	
permit	have	been	met.	

5	

Actions	

Data	from	Stages	1,	2,	3	and	4;		
results	of	NEPA	analysis;	and	
The	Service	will	issue	or	deny	
considering	information	
the	permit	request	based	on	an	
obtained	during	tribal	
evaluation	of	the	ECP	or	other	
consultation	and	through	
form	of	application.	
coordination	with	the	states	and	
other	jurisdictional	agencies.	

During	post‐construction	
monitoring,	document	mean	
Conduct	fatality	monitoring	in	
annual	eagle	fatality	rate	and	
project	footprint.		Monitor	
effects	of	disturbance.		
Determine	if	initial	conservation	 activity	of	eagles	that	may	be	
disturbed	at	nest	sites,	
measures	are	working	and	
communal	roosts,	and/or	major	
should	be	continued,	and	if	
foraging	sites.		Ideally,	monitor	
additional	conservation	
measures	might	reduce	observed	eagle	use	of	project	footprint	via	
fatalities.	Monitor	effectiveness	 point	counts,	migration	counts,	
and/or	intensive	observation	of	
of	compensatory	mitigation.		
use	distribution.	
Ideally,	assess	use	of	area	by	
eagles	for	comparison	to	pre‐
construction	levels.	

Post‐construction	survey	
database	for	fatality	monitoring,	
Comparable	pre‐	and	post‐
construction	data	for	selected	
aspect	of	eagle	use	of	the	project	
footprint	and	adjoining	areas.	
	
All	post‐construction	surveys	
should	be	conducted	for	at	least	
2	years,	and	targeted	thereafter	
to	assess	effectiveness	of	any	
experimental	conservation	
measures	or	ACPs.	

1

Table of Contents
	
Disclaimer .............................................................................................................................................................. i
EXECUTIVE SUMMARY ........................................................................................................................................ ii
1. Overview ..................................................................................................................................................... ii
2. The Bald and Golden Eagle Protection Act ................................................................................................ ii
3. Risks to Eagles from Wind Energy Facilities .............................................................................................. ii
4. Eagle Take Permits .................................................................................................................................... iii
5. Voluntary Nature of the ECPG ................................................................................................................... iii
6. Eagle Take Thresholds ............................................................................................................................... iii
7. An Approach for Developing and Evaluating Eagle ACPs......................................................................... iv
8. Mitigation Actions to Reduce Effects on Eagle Populations ..................................................................... v
9. Relationship of Eagle Guidelines (ECPG) to the Wind Energy Guidelines (WEG) .................................... vi
Tiers 1 and 2 of the WEG, Stage 1 of the ECPG ........................................................................................ vi
Tier 3 of the WEG, Stages 2, 3, and 4 of the ECPG .................................................................................. vii
Tier 4 and 5 of the WEG, Stage 5 of the ECPG .......................................................................................... ix
10. Site Categorization Based on Mortality Risk to Eagles .......................................................................... ix
Category 1 – High risk to eagles, potential to avoid or mitigate impacts is low ...................................... x
Category 2 – High or moderate risk to eagles, opportunity to mitigate impacts ...................................... x
Category 3 – Minimal risk to eagles........................................................................................................... x
11. Addressing Uncertainty ........................................................................................................................... xi
12. Interaction with the Service .................................................................................................................... xi
INTRODUCTION AND PURPOSE .......................................................................................................................... 4
1. Purpose........................................................................................................................................................ 4
2. Legal Authorities and Relationship to Other Statutes and Guidelines ..................................................... 6
3. Background and Overview of Process ........................................................................................................ 8
a. Risks to Eagles ........................................................................................................................................ 9
b. General Approach to Address Risk ......................................................................................................... 9
ASSESSING RISK AND EFFECTS ....................................................................................................................... 12
1. Considerations When Assessing Eagle Use Risk .................................................................................... 12
a. General Background and Rationale for Assessing Project Effects on Eagles ..................................... 12
b. Additional Considerations for Assessing Project Effects: Migration Corridors and Stopover Sites... 14
2. Eagle Risk Factors ..................................................................................................................................... 15
3. Overview of Process to Assess Risk......................................................................................................... 16
4. Site Categorization Based on Mortality Risk to Eagles ........................................................................... 25
a. Category 1 – High risk to eagles, potential to avoid or mitigate impacts is low ................................ 25
b. Category 2 – High or moderate risk to eagles, opportunity to mitigate impacts ................................ 25
c. Category 3 – Minimal risk to eagles ..................................................................................................... 26
5. Cumulative Effects Considerations .......................................................................................................... 26
a. Early Planning ........................................................................................................................................ 26
b. Analysis Associated with Permits ........................................................................................................ 27

2

ADAPTIVE MANAGEMENT ................................................................................................................................ 28
EAGLE CONSERVATION PLAN DEVELOPMENT PROCESS ................................................................................ 29
1. Contents of the Eagle Conservation Plan ................................................................................................. 30
a. Stage 1 .................................................................................................................................................. 31
b. Stage 2 .................................................................................................................................................. 31
c. Stage 3 ................................................................................................................................................... 31
d. Stage 4 .................................................................................................................................................. 31
e. Stage 5 – Post-construction Monitoring .............................................................................................. 31
INTERACTION WITH THE SERVICE .................................................................................................................... 32
INFORMATION COLLECTION.............................................................................................................................. 33
GLOSSARY .......................................................................................................................................................... 34
LITERATURE CITED ............................................................................................................................................. 40
APPENDIX A: ADAPTIVE MANAGEMENT ......................................................................................................... 44
1. Adaptive Management as a Tool ............................................................................................................. 45
2. Applying Adaptive Management to Eagle Take Permitting .................................................................... 46
a. Adaptive Management of Wind Project Operations ............................................................................ 46
b. Adaptive Management of Wind Project Siting and Design Recommendations .................................. 47
c. Adaptive Management of Compensatory Mitigation ........................................................................... 47
d. Adaptive Management of Population-Level Take Thresholds ............................................................. 47
Literature Cited .............................................................................................................................................. 48
APPENDIX B: STAGE 1 – SITE ASSESSMENT ................................................................................................... 50
Literature Cited .............................................................................................................................................. 52
APPENDIX C: STAGE 2 – SITE-SPECIFIC SURVEYS AND ASSESSMENT ......................................................... 53
1. Surveys of Eagle Use ................................................................................................................................ 53
a. Point Count Surveys .............................................................................................................................. 53
b. Migration Counts and Concentration Surveys...................................................................................... 60
c. Utilization Distribution (UD) Assessment ............................................................................................. 62
d. Summary ................................................................................................................................................ 63
2. Survey of the Project-area Nesting Population: Number and Locations of Occupied Nests of Eagles.. 64
Literature Cited .............................................................................................................................................. 66
APPENDIX D: STAGE 3 – PREDICTING EAGLE FATALITIES............................................................................... 68
1. Exposure .................................................................................................................................................... 69
2. Collision Probability .................................................................................................................................. 71
3. Expansion .................................................................................................................................................. 72

3

4. Fatalities.................................................................................................................................................... 72
5. Putting it all together: an example ........................................................................................................... 72
a. Patuxent Power Company Example ...................................................................................................... 73
b. Exposure ................................................................................................................................................ 74
b. Collision Probability............................................................................................................................... 75
c. Expansion ............................................................................................................................................... 75
d. Fatalities ................................................................................................................................................ 75
6. Additional Considerations ........................................................................................................................ 76
a. Small-scale projects .............................................................................................................................. 76
Literature Cited .............................................................................................................................................. 77
APPENDIX E: STAGE 4 – AVOIDANCE AND MINIMIZATION OF RISK USING ACPS AND OTHER
CONSERVATION MEASURES, AND COMPENSATORY MITIGATION............................................................... 78
Literature Cited .............................................................................................................................................. 79
APPENDIX F: ASSESSING PROJECT-LEVEL TAKE AND CUMULATIVE EFFECTS ANALYSES .......................... 80
Literature Cited .............................................................................................................................................. 85
APPENDIX G: EXAMPLES USING RESOURCE EQUIVALENCY ANALYSIS TO ESTIMATE THE
COMPENSATORY MITIGATION FOR THE TAKE OF GOLDEN AND BALD EAGLES FROM WIND ENERGY
DEVELOPMENT ................................................................................................................................................... 86
1. Introduction ............................................................................................................................................... 86
2. REA Inputs ................................................................................................................................................. 86
3. REA Example – WindCoA ......................................................................................................................... 88
a. REA Language and Methods ................................................................................................................. 89
b. REA Results for WindCoA ..................................................................................................................... 91
c. Summary of Bald Eagle REA Results .................................................................................................... 92
d. Discussion on Using REA ...................................................................................................................... 93
e. Additional Compensatory Mitigation Example..................................................................................... 93
f. Take from Disturbance ........................................................................................................................... 93
Literature Cited .............................................................................................................................................. 94
APPENDIX H: STAGE 5 – CALIBRATING AND UPDATING OF THE FATALITY PREDICTION AND CONTINUED
RISK-ASSESSMENT ........................................................................................................................................... 96
1. Fatality Monitoring ................................................................................................................................... 96
2. Disturbance Monitoring ............................................................................................................................ 98
3. Comparison of Post-Construction Eagle Use with Pre-Construction Use................................................ 99
Literature Cited .............................................................................................................................................. 99
	

4

INTRODUCTION AND PURPOSE
	

The	mission	of	the	Service	is	working	with	others	to	conserve,	protect	and	enhance	fish,	wildlife,	
plants	and	their	habitats	for	the	continuing	benefit	of	the	American	people.		As	part	of	this,	we	are	
charged	with	implementing	statutes	including	the	BGEPA,	MBTA	(Migratory	Bird	Treaty	Act),	and	
ESA	(Endangered	Species	Act).		BGEPA	prohibits	all	take	of	eagles	unless	otherwise	authorized	by	
the	Service.	A	goal	of	BGEPA	is	to	ensure	that	any	authorized	take	of	bald	and	golden	eagles	is	
compatible	with	their	preservation,	which	the	Service	has	interpreted	to	mean	allowing	take	that	is	
consistent	with	the	goal	of	stable	or	increasing	breeding	populations.		In	2009,	the	Service	
promulgated	regulations	authorizing	issuance	of	permits	for	non‐purposeful	take	of	eagles;	the	
ECPG	is	intended	to	promote	compliance	with	BGEPA	with	respect	to	such	permits	by	providing	
recommended	procedures	for:	
	
(1) 	conducting	early	pre‐construction	assessments	to	identify	important	eagle	use	areas;	
(2) 	analyzing	pre‐construction	information	to	estimate	potential	impacts	on	eagles;	
(3) 	avoiding,	minimizing,	and/or	compensating	for	potential	adverse	effects	to	eagles;	and	
(4) 	monitoring	for	impacts	to	eagles	during	construction	and	operation.	
	
The	ECPG	calls	for	scientifically	rigorous	surveys,	monitoring,	risk	assessment,	and	research	
designs	proportionate	to	the	risk	to	both	bald	and	golden	eagles.		The	ECPG	describes	a	process	by	
which	wind	energy	developers,	operators,	and	their	consultants	can	collect	and	analyze	information	
that	could	lead	to	a	programmatic	permit	to	authorize	unintentional	take	of	eagles	at	wind	energy	
facilities.		The	processes	described	here	is	not	required,	but	project	developers	or	operators	should	
coordinate	closely	with	the	Service	if	they	plan	to	use	an	alternative	approach	to	meet	the	
regulatory	requirements	for	a	permit.	

	
1. Purpose
The	Service	published	a	final	rule	(Eagle	Permit	Rule)	on	September	11,	2009		under	BGEPA	(50	
CFR	22.26)	authorizing	limited	issuance	of	permits	to	take	bald	eagles	(Haliaeetus	leucocephalus)	
and	golden	eagles	(Aquila	chrysaetos)	‘‘for	the	protection	of	...	other	interests	in	any	particular	
locality’’	where	the	take	is	compatible	with	the	preservation	of	the	bald	eagle	and	the	golden	eagle,	
is	associated	with	and	not	the	purpose	of	an	otherwise	lawful	activity,	and	cannot	practicably	be	
avoided	(USFWS	2009a).		The	ECPG	explains	the	Service’s	approach	to	issuing	programmatic	eagle	
take	permits	for	wind	energy	projects	under	this	authority,	and	provides	guidance	to	permit	
applicants	(project	developers	or	operators),	Service	biologists,	and	biologists	with	other	
jurisdictional	agencies	(state	and	tribal	fish	and	wildlife	agencies,	in	particular)	on	the	development	
of	Eagle	Conservation	Plans	(ECPs)	to	support	permit	issuance.	
	
Since	finalization	of	the	Eagle	Permit	Rule,	the	development	and	planned	development	of	wind	
facilities	(developments	for	the	generation	of	electricity	from	wind	turbines)	have	increased	in	the	
range	of	the	golden	eagle	in	the	western	United	States.		Golden	eagles	are	vulnerable	to	collisions	
with	wind	turbines	(Hunt	2002),	and	in	some	areas	such	collisions	could	be	a	major	source	of	
mortality	(Hunt	et	al.	1999,	2002;	USFWS	unpublished	data).		Although	significant	numbers	of	bald	
eagle	mortalities	have	not	yet	been	reported	at	North	American	wind	facilities,	deaths	have	
occurred	at	more	than	one	location	(USFWS,	unpublished	data),	and	the	closely	related	and	
behaviorally	similar	white‐tailed	eagle	(Haliaeetus	albicilla)	has	been	killed	regularly	at	wind	
facilities	in	Europe	(Krone	2003,	Cole	2009,	Nygård	et	al.	2010).		Because	of	this	risk	to	eagles,	
many	of	the	current	and	planned	wind	facilities	require	permits	under	the	Eagle	Permit	Rule	to	be	
in	compliance	with	the	law	if	and	when	an	eagle	is	taken	at	that	facility.		In	addition	to	being	legally	

5
necessary	to	comply	with	BGEPA	and	50	CFR	22.26,	the	conservation	practices	necessary	to	meet	
standards	required	for	issuance	of	these	permits	should	offset	the	short‐	and	long‐term	negative	
effects	of	wind	energy	facilities	on	eagle	populations.	Because	of	the	urgent	need	for	guidance	on	
permitting	eagle	take	at	wind	facilities,	this	initial	module	focuses	on	this	issue.		Many	of	the	
concepts	and	approaches	outlined	in	this	module	can	be	readily	exported	to	other	situations	(e.g.,	
solar	facilities,	electric	power	lines),	and	the	Service	expects	to	release	other	modules	in	the	future	
specifically	addressing	other	sources	of	eagle	take.	
	
The	ECPG	is	intended	to	provide	interpretive	guidance	to	Service	biologists	and	others	in	applying	
the	regulatory	permit	standards	as	specified	in	the	rule.		They	do	not	in‐and‐of	themselves	impose	
additional	regulatory	or	generally‐binding	requirements.		An	ECP	per	se	is	not	required,	even	to	
obtain	a	programmatic	eagle	take	permit.		As	long	as	the	permit	application	is	complete	and	
includes	the	information	necessary	to	evaluate	a	permit	application	under	50	CFR	22.26	or	22.27,	
the	Service	will	review	the	application	and	make	a	determination	if	a	permit	will	be	issued.		
However,	Service	personnel	will	be	trained	in	the	application	of	the	procedures	and	approaches	
outlined	in	the	ECPG,	and	developers	who	choose	to	use	other	approaches	should	expect	the	review	
time	on	the	part	of	the	Service	to	be	longer.		The	Service	recommends	that	the	basic	format	for	the	
ECP	be	followed	to	allow	for	expeditious	consideration	of	the	application	materials.	
	
Preparation	of	an	ECP	and	consultation	with	the	Service	are	voluntary	actions	on	the	part	of	the	
developer.		There	is	no	legal	requirement	that	wind	developers	apply	for	or	obtain	an	eagle	take	
permit,	so	long	as	the	project	does	not	result	in	take	of	eagles.		However,	take	of	an	eagle	without	an	
eagle	take	permit	is	a	violation	of	BGEPA,	so	the	developer	or	operator	must	weigh	the	risks	in	
his/her	decision.		The	Service	is	available	to	consult	with	the	developer	or	operator	as	he/she	
makes	that	decision.	
	
The	ECPG	is	written	to	guide	wind‐facility	projects	starting	from	the	earliest	conceptual	planning	
phase.		For	projects	already	in	the	development	or	operational	phase,	implementation	of	all	stages	
of	the	recommended	approach	in	the	ECPG	may	not	be	applicable	or	possible.		Project	developers	or	
operators	with	operating	or	soon‐to‐be	operating	facilities	and	who	are	interested	in	obtaining	a	
programmatic	eagle	take	permit	should	contact	the	Service.		The	Service	will	work	with	project	
developers	or	operators	to	determine	if	the	project	might	be	able	to	meet	the	permit	requirements	
in	50	CFR	22.26.		The	Service	may	recommend	that	the	developer	monitor	eagle	fatalities	and	
disturbance,	adopt	reasonable	measures	to	reduce	eagle	fatalities	from	historic	levels,	and	
implement	compensatory	mitigation.		Sections	of	the	ECPG	that	address	these	topics	are	relevant	to	
both	planned	and	operating	wind	facilities	(Appendices	E	and	F	in	particular).	Operators	of	wind	
projects	(and	other	activities)	that	were	in	operation	prior	to	2009	that	pose	a	risk	to	golden	eagles	
may	qualify	for	programmatic	eagle	take	permits	that	do	not	automatically	require	compensatory	
mitigation.	This	is	because	the	requirements	for	obtaining	programmatic	take	authorization	are	
designed	to	reduce	take	from	historic,	baseline	levels,	and	the	preamble	to	the	Eagle	Permit	Rule	
specified	that	unavoidable	take	remaining	after	implementation	of	avoidance	and	minimization	
measures	at	such	projects	would	not	be	subtracted	from	regional	eagle	take	thresholds	(U.	S.	Fish	
and	Wildlife	Service	2009a).	
	
The	ECPG	is	designed	to	be	compatible	with	the	more	general	guidelines	provided	in	the	U.S.	Fish	
and	Wildlife	Service	Land‐based	Wind	Energy	Guidelines	(WEG)	http://www.fws.gov/	
habitatconservation/windpower/wind_turbine_advisory_committee.html.		However,	because	the	
ECPG	describes	actions	which	help	to	comply	with		the	regulatory	requirements	in	BGEPA	for	an	
eagle	take	permit	as	described	in	50	CFR	22.26	and	22.27,	they	are	more	specific.		The	Service	will	
make	every	effort	to	ensure	the	work	and	timelines	for	both	processes	are	as	congruent	as	possible.	

6

	
2. Legal Authorities and Relationship to Other Statutes and Guidelines	
There	are	several	laws	that	must	be	considered	for	compliance	during	eagle	take	permit	application	
review	under	the	50	CFR	22.26	and	22.27	regulations:	BGEPA,	MBTA,	ESA,	the	National	
Environmental	Policy	Act	(NEPA)	(42	U.S.C.	4321	et.	seq.),	and	the	National	Historic	Preservation	
Act	(NHPA)	(16	U.S.C.	470	et	seq.).		BGEPA	is	the	primary	law	protecting	eagles.	BGEPA	defines	
“take”	to	include	“pursue,	shoot,	shoot	at,	poison,	wound,	kill,	capture,	trap,	collect,	molest	or	
disturb”	and	prohibits	take	of	individuals,	and	their	parts,	nests,	or	eggs	(16	USC	668	&	668c).	The	
Service	expanded	this	definition	by	regulation	to	include	the	term	“destroy”	to	ensure	that	“take”	
includes	destruction	of	eagle	nests	(50	CFR	22.3).		The	term	“disturb”	is	defined	by	regulation	at	50	
CFR	22.3	as	“to	agitate	or	bother	a	bald	or	golden	eagle	to	a	degree	that	causes,	or	is	likely	to	cause,	
…	injury	to	an	eagle,	a	decrease	in	productivity,	or	nest	abandonment…”	(USFWS	2007).		A	goal	of	
BGEPA	is	to	ensure	that	any	authorized	take	is	compatible	with	eagle	preservation,	which	the	
Service	has	interpreted	to	mean	it	can	authorize	take	that	is	consistent	with	the	goal	of	stable	or	
increasing	breeding	populations	of	bald	and	golden	eagles	(USFWS	2009b).	
	
In	2009,	two	new	permit	rules	were	created	for	eagles.	Under	50	CFR	22.26,	the	Service	can	issue	
permits	that	authorize	individual	instances	of	take	of	bald	and	golden	eagles	when	the	take	is	
associated	with,	but	not	the	purpose	of	an	otherwise	lawful	activity,	and	cannot	practicably	be	
avoided.		The	regulation	also	authorizes	ongoing	or	programmatic	take,	but	requires	that	any	
authorized	programmatic	take	be	unavoidable	after	implementation	of	advanced	conservation	
practices.		Under	50	CFR	22.27,	the	Service	can	issue	permits	that	allow	the	intentional	take	of	eagle	
nests	where	necessary	to	alleviate	a	safety	emergency	to	people	or	eagles,	to	ensure	public	health	
and	safety,	where	a	nest	prevents	use	of	a	human‐engineered	structure,	and	to	protect	an	interest	
in	a	particular	locality	where	the	activity	or	mitigation	for	the	activity	will	provide	a	net	benefit	to	
eagles.		Only	inactive	nests	are	allowed	to	be	taken	except	in	cases	of	safety	emergencies.	
	
The	new	Eagle	Permit	Rule	provides	a	mechanism	where	the	Service	may	legally	authorize	the	non‐
purposeful	take	of	eagles.	However,	BGEPA	provides	the	Secretary	of	the	Interior	with	the	authority	
to	issue	eagle	take	permits	only	when	the	take	is	compatible	with	the	preservation	of	each	species,	
defined	in	USFWS	(2009a)	as	“…consistent	with	the	goal	of	stable	or	increasing	breeding	
populations.”		The	Service	ensures	that	any	take	it	authorizes	under	50	CFR	22.26	does	not	exceed	
this	preservation	standard	by	setting	regional	take	thresholds	for	each	species	determined	using	
the	methodology	contained	in	the	NEPA	Final	Environmental	Assessment	(FEA)	developed	for	the	
new	permit	rules	(USFWS	2009b).		The	details	and	background	of	the	process	used	to	calculate	
these	take	thresholds	are	presented	in	the	FEA	(USFWS	2009b).		It	is	important	to	note	that	the	
take	thresholds	for	regional	eagle	management	populations	(eagle	management	units)	and	the	
process	by	which	they	are	determined	are	derived	independent	from	this	or	any	other	ECPG	
module.	
	
Many	states	and	tribes	have	regulations	that	protect	eagles,	and	may	require	permits	for	purposeful	
and	non‐purposeful	take.		Project	developers	or	operators	should	contact	all	pertinent	state	and	
tribal	fish	and	wildlife	agencies	at	the	earliest	possible	stage	of	project	development	to	ensure	
proper	coordination	and	permitting.		The	Service	will	coordinate	our	programmatic	take	permits	
with	all	such	jurisdictional	agencies.	
	
Wind	projects	that	are	expected	to	cause	take	of	endangered	or	threatened	wildlife	species	should	
still	receive	incidental	take	authorizations	under	sections	7	or	10	of	ESA	in	order	to	ensure	
compliance	with	Federal	law.		A	project	developer	or	operator	seeking	an	Incidental	Take	Permit	

7
(ITP)	through	the	ESA	section	10	Habitat	Conservation	Plan	(HCP)	process	may	be	issued	an	ITP	
only	if	the	permitted	activity	is	otherwise	lawful	(section	10(a)(1)(B)).	If	the	project	and	covered	
activities	in	the	HCP	are	likely	to	take	bald	or	golden	eagles,	the	project	proponent	should	obtain	a	
BGEPA	permit	or	include	the	bald	or	golden	eagle	as	a	covered	species	in	the	HCP	in	order	for	the	
activity	to	be	lawful	in	the	event	that	eagles	are	taken.		When	bald	or	golden	eagles	are	covered	in	
an	HCP	and	ITP,	the	take	is	authorized	under	BGEPA	even	if	the	eagle	species	is	not	listed	under	the	
ESA	(see	50	CFR	22.11(a)).	
	
If		bald	or	golden	eagles	are	included	as	covered	species	in	an	HCP,	the	avoidance,	minimization,	
and	other	mitigation	measures	in	the	HCP	must	meet	the	BGEPA	permit	issuance	criteria	of	50	CFR	
22.26,	and	include	flexibility	for	adaptive	management.	If	take	of	bald	or	golden	eagles	is	likely	but	
the	project	developer	or	operator	does	not	qualify	for	eagle	take	authorization	(or	chooses	not	to	
request	such	authorization),	an	ITP	may	be	issued	in	association	with	the	proposed	HCP.	The	
project	proponent	must	be	advised,	in	writing,	that	bald	or	golden	eagles	would	not	be	included	as	
covered	species	and	take	of	bald	eagles	or	golden	eagles	would	not,	therefore,	be	authorized	under	
the	incidental	take	permit.	The	project	developer	or	operator	must	also	be	advised	that	the	
incidental	take	permit	would	be	subject	to	suspension	or	revocation	if	take	of	bald	eagles	or	golden	
eagles	should	occur.	
	
In	addition	to	ESA,	wind	project	developers	or	operators	need	to	address	take	under	MBTA.		MBTA	
prohibits	the	taking,	hunting,	killing,	pursuit,	capture,	possession,	sale,	barter,	purchase,	transport,	
and	export	of	migratory	birds,	their	eggs,	parts,	and	nests,	except	when	authorized	by	the	
Department	of	the	Interior.		For	eagles,	the	BGEPA	take	authorization	serves	as	authorization	under	
MBTA	per	50	CFR	22.11(b).		For	other	MBTA‐protected	birds,	because	neither	the	MBTA	nor	its	
permit	regulations	at	50	CFR	Part	21	currently	provide	a	specific	mechanism	to	permit	
“unintentional”	take,	it	is	important	for	project	developers	or	operators	to	work	proactively	with	
the	Service	to	avoid	and	minimize	take	of	migratory	birds.	The	Service,	with	assistance	from	a	
Federal	Advisory	Committee,	developed	the	WEG	to	provide	a	structured	system	to	evaluate	and	
address	potential	negative	impacts	of	wind	energy	projects	on	species	of	concern.		Because	the	
Service	has	the	authority	to	issue	a	permit	for	non‐purposeful	take	of	eagles,	our	legal	and	
procedural	obligations	are	significantly	greater,	and	therefore	the	ECPG	is	more	focused	and	
detailed	than	the	WEG.		We	have	modeled	as	much	of	the	ECPG	as	possible	after	the	WEG,	but	there	
are	important	and	necessary	differences.	
	
NEPA	applies	to	issuance	of	eagle	take	permits	because	issuing	a	permit	is	a	federal	action.		While	
providing	technical	assistance	to	agencies	conducting	NEPA	analyses,	the	Service	will	participate	in	
the	other	agencies'	NEPA	to	the	extent	feasible	in	order	to	streamline	subsequent	NEPA	analyses	
related	to	a	project.		For	actions	that	may	result	in	applications	for	development	of	programmatic	
permits,	the	Service	may	participate	as	a	cooperating	agency	to	streamline	the	permitting	process.	
	
If	no	federal	nexus	exists,	other	than	an	eagle	permit,	or	if	the	existing	NEPA	of	another	agency	is	
not	adequate,	the	Service	must	complete	a	NEPA	analysis	before	it	can	issue	a	permit.		The	Service	
will	work	with	the	project	developer	or	operator	to	conduct	a	complete	NEPA	analysis,	including	
assisting	with	data	needs	and	determining	the	scope	of	analysis.		Project	developers	or	operators	
may	provide	assistance	that	can	expedite	the	NEPA	process	in	accordance	with	40	CFR	§1506.5.		
Additionally,	there	are	opportunities	to	“batch”	NEPA	analyses	for	proposed	projects	in	the	same	
geographic	area.		In	these	cases,	project	developers	or	operators	and	the	Service	could	pool	
resources	and	data,	likely	increasing	the	quality	of	the	product	and	the	efficiency	of	the	process.		
Developers	should	coordinate	closely	with	the	Service	for	projects	with	no	federal	nexus	other	than	

8
the	eagle	permit.		Close	coordination	between	project	developers	or	operators	and	the	Service	
regarding	the	data	needs	and	scope	of	the	analysis	required	for	a	permit	will	reduce	delays.	
	
Through	50	CFR	22.26	and	the	associated	FEA,	the	Service	defined	“mitigation”	as	per	the	Service	
Mitigation	Policy	(46	FR	7644,	Jan.	23,	1981),	and	the	President’s	Council	on	Environmental	Quality	
(40	CFR	1508.20	(a‐e)),	to	sequentially	include	the	following:	
	
(1)	Avoiding	the	impact	on	eagles	altogether	by	not	taking	a	certain	action	or	parts	of	an	action;	
(2)	Minimizing	impacts	by	limiting	the	degree	or	magnitude	of	the	action	and	its	
implementation;	
(3)	Rectifying	the	impact	by	repairing,	rehabilitating,	or	restoring	the	affected	environment;	
(4)	Reducing	or	eliminating	the	impact	over	time	by	implementing	preservation	and	
maintenance	operation	during	the	lifetime	of	the	action;	and	
(5)	Compensating	for	the	impact	by	replacing	or	providing	substitute	resources	or	
environments.	
	
Throughout	this	document	we	differentiate	between	mitigation,	which	covers	all	of	the	components	
listed	above,	and	compensatory	mitigation,	which	is	a	subset	of	(5)	above	and	directly	targets	
offsetting	permitted	disturbance	and	mortality	to	accomplish	a	no‐net‐loss	objective	at	the	scale	of	
the	eagle	management	unit.		The	Service	requires	compensatory	mitigation	(potentially	in	addition	
to	other	mitigation)	where	it	has	not	been	determined	that	eagle	populations	can	sustain	additional	
mortality.		The	NEPA	analysis	on	our	permits	and	the	discussion	of	mitigation	in	this	document	
follow	this	system,	and	in	this	ECPG	we	refer	to	(1)	–	(4)	as	conservation	measures	to	avoid	and	
minimize	take,	of	which	ACPs	are	a	subset,	and	to	(5)	as	compensatory	mitigation.	
	
Eagles	are	significant	species	in	Native	American	culture	and	religion	(Palmer	1988)	and	may	be	
considered	contributing	elements	to	a	“traditional	cultural	property”	under	Section	106	of	the	
NHPA.		Some	locations	where	eagles	would	be	taken	have	traditional	religious	and	cultural	
importance	to	Native	American	tribes	and	thus	have	the	potential	of	being	regarded	as	traditional	
cultural	properties	under	NHPA.		Permitted	take	of	one	or	more	eagles	from	these	areas,	for	any	
purpose,	could	be	considered	an	adverse	effect	to	the	traditional	cultural	property.		These	
considerations	will	be	incorporated	into	any	NEPA	analysis	associated	with	an	eagle	take	permit.	
	
Federally‐recognized	Indian	tribes	enjoy	a	unique	government‐to‐government	relationship	with	the	
United	States.		The	Service	recognizes	Indian	tribal	governments	as	the	authoritative	voice	
regarding	the	management	of	tribal	lands	and	resources	within	the	framework	of	applicable	laws.	It	
is	important	to	recall	that	many	tribal	traditional	lands	and	tribal	rights	extend	beyond	reservation	
lands.		The	Service	consults	with	Indian	tribal	governments	under	the	authorities	of	Executive	
Order	13175	“Consultation	and	Coordination	with	Indian	Tribal	Governments”	and	supporting	DOI	
and	Service	policies.		To	this	end,	when	it	is	determined	that	federal	actions	and	activities	may	
affect	a	tribe’s	resources	(including	cultural	resources),	lands,	rights,	or	ability	to	provide	services	
to	its	members,	the	Service	must,	to	the	extent	practicable,	seek	to	engage	the	affected	tribe(s)	in	
consultation	and	coordination.	

3. Background and Overview of Process
Increased	energy	demands	and	the	nationwide	goal	to	increase	energy	production	from	renewable	
sources	have	intensified	the	development	of	energy	facilities,	including	wind	energy.	The	Service	
supports	renewable	energy	development	that	is	compatible	with	fish	and	wildlife	conservation.		
The	Service	closely	coordinates	with	state,	tribal,	and	other	federal	agencies	in	the	review	and	

9
permitting	of	wind	energy	projects	to	address	potential	resource	effects,	including	effects	to	bald	
and	golden	eagles.	However,	our	knowledge	of	these	effects	and	how	to	address	them	at	this	time	is	
limited.		Given	this	and	the	Service’s	regulatory	mandate	to	only	authorize	actions	that	are	
“compatible	with	the	goal	of	stable	or	increasing	breeding	populations”	of	eagles	has	led	us	to	adopt	
an	adaptive	management	framework	predicated,	in	part,	on	the	precautionary	approach	for	
consideration	and	issuance	of	programmatic	eagle	take	permits.		This	framework	consists	of	case‐
specific	considerations	applied	within	a	national	framework,	and	with	the	outcomes	carefully	
monitored	so	that	we	maximize	learning	from	each	case.		The	knowledge	gained	through	
monitoring	can	then	be	used	to	update	and	refine	the	process	for	making	future	permitting	
decisions	such	that	our	ultimate	conservation	objectives	are	attained,	as	well	as	to	consider	
operational	adjustments	at	individual	projects	at	regular	intervals	where	deemed	necessary	and	
appropriate.		The	ECPG	provides	the	background	and	information	necessary	for	wind	project	
developers	or	operators	to	prepare	an	ECP	that	assesses	the	risk	of	a	prospective	or	operating	
project	to	eagles,	and	how	siting,	design,	and	operational	modifications	can	mitigate	that	risk.		
Implementation	of	the	final	ECP	must	reduce	predicted	eagle	take,	and	the	population	level	effect	of	
that	take,	to	a	degree	compatible	with	regulatory	standards	to	justify	issuance	of	a	programmatic	
take	permit	by	the	Service.	

	
a. Risks to Eagles
Energy	development	can	affect	eagles	in	a	variety	of	ways.		First,	structures	such	as	wind	
turbines	can	cause	direct	mortality	through	collision	(Hunt	2002,	Nygård	et	al.	2010).		This	
is	the	primary	threat	to	eagles	from	wind	facilities,	and	the	monitoring	and	avoidance	and	
minimization	measures	advocated	in	the	ECPG	primarily	are	aimed	at	this	threat.		Second,	
activities	associated	with	pre‐construction,	construction,	or	operation	and	maintenance	of	a	
project	might	cause	disturbance	and	result	in	loss	of	productivity	at	nearby	nests	or	
disturbance	to	nearby	concentrations	of	eagles.		Third,	if	disturbance	or	mortality	effects	
are	permanent,	they	could	result	in	the	permanent	or	long	term	loss	of	a	nesting	territory.		
All	of	these	impacts,	unless	properly	permitted,	are	violations	of	BGEPA	(USFWS	2009a).		
Additionally,	disturbances	near	important	eagle	use	areas	or	migration	concentration	sites	
might	stress	eagles	to	a	degree	that	leads	to	reproductive	failure	or	mortality	elsewhere;	
these	impacts	are	of	concern	as	well,	and	they	could	amount	to	prohibited	take,	though	such	
effects	are	difficult	to	predict	and	quantify.		Thus,	the	ECPG	addresses	both	direct	mortality	
and	disturbance.		Many	new	wind	projects	are	located	in	remote	areas	that	have	few,	if	any,	
transmission	lines.		The	Service	considers	new	transmission	lines	and	other	infrastructure	
associated	with	renewable	energy	projects	to	be	part	of	a	project.		Accordingly,	assessments	
of	project	impacts	should	include	transmission	lines	and	other	facilities,	not	merely	wind	
turbines.	

	
b. General Approach to Address Risk
Applicants	for	permits	under	50	CFR	22.26,	non‐purposeful	eagle	take,	are	required	to	avoid	
and	minimize	the	potential	for	take	of	eagles	to	the	extent	practicable.		Permits	for	wind‐
energy	development	are	programmatic	as	they	will	authorize	recurring	take,	rather	than	
isolated	incidences	of	take.		For	programmatic	take	permits,	the	regulations	at	50	CFR	22.26	
require	that	any	authorized	take	is	unavoidable	after	implementation	of	ACPs.		50	CFR	22.3	
defines	“advanced	conservation	practices”	as	“scientifically	supportable	measures	that	are	
approved	by	the	Service	and	represent	the	best	available	techniques	to	reduce	eagle	
disturbance	and	ongoing	mortalities	to	a	level	where	remaining	take	is	unavoidable.”	
	

10
Because	the	best	information	indicates	that	there	are	currently	no	available	scientifically	
supportable	measures	that	will	reduce	eagle	disturbance	and	blade‐strike	mortality	at	wind	
projects,	the	Service	has	not	currently	approved	any	ACPs	for	wind‐energy	projects.	
The	preamble	to	the	Eagle	Permit	Rule	envisioned	the	Service	and	industry	working	
together	to	identify	and	evaluate	possible	ACPs	(USFWS	2009a).		The	process	of	ACP	
development	for	wind‐energy	facilities	has	been	hampered	because	there	has	been	little	
standardized	scientific	study	of	potential	ACPs,	and	such	information	can	best	be	obtained	
through	experimental	application	of	ACPs	at	operating	facilities	with	eagle	take	permits.		
Given	this,	and	considering	the	pressing	need	to	develop	ACPs	for	wind‐energy	facilities,	the	
Service	believes	that	the	best	course	of	action	is	to	work	with	industry	to	develop	ACPs	for	
wind	projects	as	part	of	the	programmatic	take	permit	process.	
	
Under	this	scenario,	ACPs	would	be	implemented	at	operating	wind	facilities	with	an	eagle	
take	permit	on	an	“experimental”	basis	(the	ACPs	are	considered	experimental	because	they	
would	not	yet	meet	the	definition	of	an	ACP	in	the	eagle	permit	regulation).		The	
experimental	ACPs	would	be	scientifically	evaluated	for	their	effectiveness,	and	based	on	
the	results	of	these	studies,	could	be	modified	in	an	adaptive	management	regime.	
	
Despite	the	current	lack	of	available	ACPs,	the	best	available	scientific	information	may	
demonstrate	that	a	particular	avoidance,	minimization,	or	other	mitigation	action	should	be	
applied	as	a	condition	on	an	eagle	programmatic	take	permit	for	wind‐energy	facilities	(see	
50	C.F.R.	22.6(c)(1)).		A	project	developer	or	operator	will	still	be	expected	to	implement	
any	reasonable	avoidance	and	minimization	measures	that	may	reduce	take	of	eagles	at	a	
project.		However,	the	Service	and	the	project	developer	or	operator	will	discuss	and	agree	
on	other	site‐specific	and	possibly	turbine‐specific	factors	that	may	pose	risks	to	eagles	and	
experimental	ACPs	that	might	reduce	or	eliminate	those	risks	if	the	risks	are	substantiated	
by	the	best	available	science.		Unless	the	Service	determines	that	there	is	a	reasonable	
scientific	basis	to	implement	experimental	ACPs	up	front,	we	recommend	that	such	
measures	be	deferred	until	such	time	as	there	is	eagle	take	at	the	facility	or	the	Service	
determines	that	the	circumstances	and	evidence	surrounding	instances	of	take	or	risk	of	
take	suggest	the	experimental	ACPs	might	be	warranted.		This	agreement	would	be	
specified	as	a	condition	of	the	programmatic	eagle	take	permit.	
	
Because	ACPs	would	be	considered	experimental	in	these	situations,	we	recommend	that	
they	be	subject	to	a	cost	cap	that	the	Service	and	the	project	developer	or	operator	establish	
as	part	of	the	initial	agreement	before	issuance	of	a	permit,	thereby	providing	financial	
certainty	to	the	project	operator	or	developer	as	to	what	maximum	costs	of	such	measures	
might	be.		The	amount	of	the	cap	should	be	relevant	to	the	theorized	risk	factors	identified	
for	the	project,	and	proportional	to	overall	risk.	
	
If	eagle	take	is	confirmed	through	post‐construction	monitoring,	developers	or	operators	
would	be	expected	to	implement	the	experimental	ACP(s)	and	to	monitor	future	eagle	take	
relative	to	the	ACP(s)	as	part	of	the	adaptive	management	process	specified	in	Appendix	A,	
but	all	within	the	limits	of	the	pre‐determined	financial	cap.		As	the	results	from	monitoring	
experimental	ACPs	across	a	number	of	facilities	accumulates	and	is	analyzed	as	part	of	the	
adaptive	management	process,	scientific	information	in	support	of	certain	ACPs	may	accrue,	
whereas	other	ACPs	may	show	little	value	in	reducing	take.		If	the	Service	determines	that	
the	available	science	demonstrates	an	experimental	ACP	is	effective	in	reducing	eagle	take,	
the	Service	will	approve	that	ACP	and	require	its	implementation	up	front	on	new	projects	
when	and	where	warranted.	

11
	
Where	take	is	unavoidable	and	when	eagle	populations	at	the	scale	of	the	eagle	
management	unit	(as	defined	in	USFWS	2009b)	are	not	estimated	to	be	healthy	enough	to	
sustain	additional	mortality	over	existing	levels,	applicants	must	reduce	the	effect	of	
permitted	unavoidable	mortality	to	a	no‐net‐loss	standard	through	compensatory	
mitigation	for	the	duration	of	the	permitted	activity.		No‐net‐loss	means	that	unavoidable	
mortality	caused	by	the	permitted	activities	is	offset	by	compensatory	mitigation	that	
reduces	another,	ongoing	form	of	mortality	by	an	equal	or	greater	amount,	or	which	leads	to	
an	increase	in	carrying	capacity	that	allows	the	eagle	population	to	grow	by	an	equal	or	
greater	amount.		Compensatory	mitigation	may	also	be	necessary	to	offset	substantial	
effects	in	other	situations	(USFWS	2009a),	and	mitigation	designed	to	offset	other	
detrimental	effects	of	permits	on	eagles	may	be	advised	in	addition	to	compensatory	
mitigation	in	some	cases.		The	Service	and	the	project	developer	or	operator	seeking	a	
programmatic	eagle	take	permit	should	agree	on	the	number	of	eagle	fatalities	to	mitigate	
and	what	actions	will	be	taken	if	actual	eagle	fatalities	differ	from	the	predicted	number.		
The	compensatory	mitigation	requirement	and	trigger	for	adjustment	should	be	specified	in	
the	permit.		If	the	procedures	recommended	in	the	ECPG	are	followed,	there	should	not	be	a	
need	for	additional	compensatory	mitigation.		However,	if	other,	less	risk‐averse	models	are	
used	to	estimate	fatalities,	underestimates	might	be	expected	and	the	permit	should	specify	
the	threshold(s)	of	take	that	would	trigger	additional	actions	and	the	specific	mitigation	
activities	that	would	be	implemented	if	fatalities	are	underestimated.		The	approach	
described	in	the	ECPG	is	applicable	for	all	land‐based	wind	energy	projects	within	the	range	
of	the	bald	and	golden	eagle	where	interactions	with	wind	project	infrastructure	have	been	
documented	or	are	reasonably	expected	to	occur.		The	ECPG	is	intended	to	provide	a	
national	framework	for	assessing	and	mitigating	risk.	
	
As	part	of	the	application	process	for	a	programmatic	eagle	take	permit,	the	Service	
recommends	that	project	developers	or	operators	prepare	an	ECP	that	outlines	the	project	
development	process	and	includes	conservation	and	monitoring	plans	as	recommended	in	
this	ECPG.		The	ECPG	provides	examples	of	ways	that	applicants	can	meet	the	regulatory	
standards	in	the	rule,	and	while	other	approaches	may	be	acceptable,	the	Service	will	
determine	their	adequacy	on	a	case‐by‐case	basis.		As	noted	previously,	an	ECP	is	not	
required,	but	if	one	is	developed	following	the	approach	recommended	here,	it	will	expedite	
Service	review	of	the	project.	
	
There	is	substantial	uncertainty	surrounding	the	risk	of	wind	projects	to	eagles,	and	of	ways	to	
minimize	that	risk.		For	this	reason,	the	Service	strongly	recommends	that	care	be	taken	to	protect	
against	the	consequences	of	underestimating	eagle	fatality	rates	at	wind	facilities.		Overestimates,	
once	confirmed,	can	be	adjusted	downward	based	on	post‐construction	monitoring	information	
with	no	consequence	to	eagle	populations,	and	project	developers	or	operators	can	trade	or	be	
credited	for	excess	compensatory	mitigation.		However,	the	options	for	addressing	underestimated	
fatality	rates	are	extremely	limited,	and	pose	either	potential	hardships	for	wind	developers	or	
significant	risks	to	eagle	populations.	

12

ASSESSING RISK AND EFFECTS
1. Considerations When Assessing Eagle Use Risk
Bald	eagles	and	golden	eagles	associate	with	distinct	geographic	areas	and	landscape	features	
throughout	their	respective	ranges.		The	Service	defines	these	“important	eagle‐use	areas”	as	“an	
eagle	nest,	foraging	area,	or	communal	roost	site	that	eagles	rely	on	for	breeding,	sheltering,	or	
feeding,	and	the	landscape	features	surrounding	such	nest,	foraging	area,	or	roost	site	that	are	
essential	for	the	continued	viability	of	the	site	for	breeding,	feeding,	or	sheltering	eagles”	(USFWS	
2009a;	50	CFR	22.3).		Migration	corridors	and	migration	stopover	sites	also	provide	important	
foraging	areas	for	eagles	during	migration	(e.g.,	Restani	et	al.	2001,	Mojica	2008)	and	result	in	
seasonal	concentrations	of	eagles.		As	a	result,	the	presence	of	a	migration	corridor	or	stopover	site	
on	or	near	a	proposed	wind	development	project	could	increase	the	probability	of	encounters	
between	eagles	and	wind	turbines.		Although	these	sites	are	not	specifically	included	within	the	
regulatory	definition	of	an	important	eagle‐use	area	at	50	CFR	22.3,	the	presence	of	such	a	site	on	
or	near	a	proposed	wind	project	could	increase	the	likelihood	of	collisions.	
	
Wind	energy	projects	that	overlap,	or	are	proximate	to,	important	eagle	use	areas	or	migration	
concentration	sites	may	pose	risks	to	the	eagles	for	reasons	described	earlier.		Project	developers	
or	operators	should	identify	the	location	and	type	of	all	important	eagle	use	areas	or	migration	
concentration	sites	that	might	be	affected	by	a	proposed	wind	project	(e.g.,	within	the	project	area).		
If	recent	(within	the	previous	5	years)	local	data	are	available	on	the	spacing	of	eagle	nests	for	the	
project‐area	nesting	population,	those	data	can	be	used	to	determine	an	appropriate	boundary	for	
such	surveys	(as	described	in	Appendix	H).		Otherwise,	for	both	species	we	suggest	initial	surveys	
be	conducted	on	and	within	10	miles	of	a	project’s	footprint	to	establish	the	project‐area	mean	
inter‐nest	distance.		The	project	footprint	is	the	minimum	convex	polygon	(e.g.,	Mohr	1947)	that	
encompasses	the	wind	project	area	inclusive	of	the	hazardous	area	around	all	turbines	and	any	
associated	infrastructure,	including	utility	lines,	out‐buildings,	roads,	etc.		We	suggest	a	site‐specific	
approach	based	on	the	spacing	between	nearest,	simultaneously	occupied	nests	for	the	species	
present	in	the	area.		If	data	on	nest‐spacing	in	the	project	area	are	lacking,	project	proponents	or	
operators	may	wish	to	survey	up	to	10	miles,	as	this	is	½	the	largest	recorded	spacing	observed	for	
golden	eagles	in	the	Mojave/Sonoran	deserts	of	western	Arizona	(Millsap	1981).		.		For	subsequent	
monitoring	(e.g.,	post‐construction	monitoring	of	occupancy	and	productivity	of	pairs	potentially	
disturbed	by	the	project),	the	project‐area	mean	inter‐nest	distance	can	be	used	to	define	a	more	
relevant	project‐area	boundary.		The	10‐mile	perimeter	may	be	unnecessary	for	bald	eagles	in	
some	areas,	and	the	Service	acknowledges	there	needs	to	be	flexibility	in	the	application	of	this	
approach	to	accommodate	specific	situations.	
	
Evaluating	the	spatial	area	described	above	for	each	wind	project	is	a	key	part	of	the	programmatic	
take	permitting	process.		As	described	later,	surveys	should	be	conducted	initially	to	obtain	data	to	
predict	effects	of	wind	projects	on	eagles.		After	the	project	begins	operating,	studies	should	again	
be	conducted	to	determine	the	actual	effects.		The	following	sections	include	descriptions	and	
criteria	for	identifying	important	eagle	use	areas	or	migration	concentration	sites	in	these	
assessments.	
	

a. General Background and Rationale for Assessing Project Effects on Eagles
A	synthesis	of	publicly	available	databases	and	technical	literature	are	fundamental	to	the	
pre‐construction	assessment	component	of	an	ECP.		In	some	instances,	this	work	may	
reveal	information	on	use	of	a	proposed	project	area	by	eagles	that	is	strong	enough	to	
support	a	decision	on	whether	to	proceed	with	the	project.		In	most	cases,	if	available	

13
information	warrants	further	consideration	of	a	potential	wind	project	site,	on‐site	surveys	
should	be	implemented	to	further	document	use	of	the	project	area	by	eagles.		The	goal	of	
such	surveys	should	be	to	quantify	and	describe	use	of	the	project	area	by	breeding	
(territorial)	and	non‐breeding	eagles	across	seasons	and	years.		A	variety	of	survey	
approaches	may	be	needed	to	accomplish	this	goal.	
	
Although	potential	for	presence	of	all	types	of	important	eagle	use	areas	or	migration	
concentration	sites	should	be	considered	when	beginning	to	assess	a	potential	project	site,	
special	attention	is	typically	given	to	nests	and	nesting	pairs.		An	eagle	territory	is	defined	in	
50	CFR	22.3	as	an	area	that	contains,	or	historically	contained,	one	or	more	nests	within	the	
home	range	of	a	mated	pair	of	eagles.		We	recognize	that	usage	conflicts	with	the	true	
biological	meaning	of	the	term	territory,	but	we	use	it	herein	in	its	regulatory	context.		
Newton	(1979)	considered	the	nesting	territory	of	a	raptor	as	the	defended	area	around	a	
pair’s	nest	site	and	defined	the	home	range	as	“...the	area	traveled	by	the	individual	in	its	
normal	activities	of	food	gathering,	mating,	and	caring	for	the	young.”		For	golden	eagles	at	
least,	the	extent	of	the	home	range	and	territory	during	nesting	season	generally	are	
similar;	the	eagle	defends	its	territory	by	undulating	flight	displays	near	the	home	range	
boundaries	and	adjoining	territories	barely	overlap	(Harmata	1982,	Collopy	and	Edwards	
1989,	Marzluff	et	al.	1997).	
	
Avoidance	zones,	often	distinguished	by	specific	“buffer”	distances,	have	been	prescribed	to	
protect	nests	and	other	types	of	eagle	use	areas	from	disturbance.		Recommendations	for	
the	size	of	avoidance	zones	for	nests	of	bald	eagles	and	golden	eagles	have	sometimes	been	
based	on	documented	distances	between	nests	and	territory	boundaries.		For	example,	
McGrady	et	al.	(2002)	and	Watson	and	Davies	(2009)	indicated	nesting	territories	of	golden	
eagles	extend	to	at	least	4	miles	from	their	nests.		Garrett	et	al.	(1993)	found	that	bald	eagle	
territories	extend	at	least	2	miles	from	nests,	though	studies	in	areas	of	densely	packed	
breeding	territories	of	bald	eagles	suggest	much	smaller	distances	(Sherrod	et	al.	1976,	
Hodges	and	Robards	1982,	Anthony	2001).		A	recommendation	for	a	spatial	buffer	to	avoid	
disturbance	of	eagle	nests	can	hardly	be	applied	throughout	the	entire	range	of	either	
species	due	to	marked	variation	in	the	size	and	configuration	of	nesting	territories.		As	such,	
these	avoidance	prescriptions	have	been	conservative	because	there	are	few	site‐specific	
data	on	spatial	extent	of	territories	in	the	published	and	unpublished	literature.		For	bald	
eagles,	minimum‐distance	buffers	are	prescribed	by	the	Service	to	protect	nests,	foraging	
areas,	and	communal	roosts	against	disturbance	from	a	variety	of	activities	(USFWS	2007b).	
	
The	approach	we	recommend	in	the	ECPG	for	evaluating	siting	options	and	assessing	
potential	mortality	and	disturbance	effects	of	wind	facilities	on	eagles	is	to	conduct	
standardized	surveys	(e.g.,	point	counts)	to	estimate	eagle	exposure	within	the	project	
footprint.		We	further	suggest	augmenting	these	with	surveys	to	determine	locations	of	
important	eagle	use	areas	or	migration	concentration	sites	for	the	project‐area	eagle	
population.		The	project‐area	eagle	population	is	the	population	of	breeding,	resident	non‐
breeding,	migrating,	and	wintering	eagles	within	the	project	area.		As	described	previously	
and	in	Appendix	H,	if	recent	data	on	the	spacing	of	eagle	nests	in	the	project	area	are	
available,	it	may	be	appropriate	to	use	the	mean	species‐specific	inter‐nest	distance	
(assuming	there	is	no	reason	to	suspect	eagle	territories	in	the	project	area	are	configured	
such	that	the	mean	inter‐nest	distance	would	be	misleading)	as	the	outer	boundary	of	the	
project	area.		Such	a	choice,	however,	also	increases	the	importance	of	having	adequate	
eagle	exposure	information	from	the	project	footprint	for	all	seasons.		For	example,	a	winter	
communal	night	roost	of	eagles	further	than	one	mean	inter‐nest	distance	from	the	project	

14
boundary	could	produce	a	large	influx	of	eagles	into	the	footprint	in	winter.		Inadequate	
winter	eagle	exposure	sampling	(or	sampling	in	only	one	year,	if	the	night	roost	is	not	used	
annually)	in	combination	with	selection	of	a	project	area	based	on	nest	spacing	alone,	could	
result	in	a	failure	to	detect	this	increased	risk	to	eagles	in	winter.		Unpredicted	fatalities	that	
result	from	such	an	oversight	will	have	to	be	addressed	by	the	project	developers	or	
operators	eventually	through	increased	compensatory	mitigation,	operational	adjustments,	
or	both	to	continue	operating	under	the	authority	of	a	valid	eagle	permit.		Thus,	it	is	
important	that	the	combination	of	exposure	and	project‐area	surveys	adequately	capture	all	
risks	to	eagles.	
	
One‐half	the	mean	inter‐nest	distance	has	been	used	as	a	coarse	approximation	for	the	
territory	boundary	in	a	number	of	raptor	studies	(e.g.,	Thorstrom	2001,	Wichmann	et	al.	
2003,	Soutullo	et	al.	2006).		Eagle	pairs	at	nests	within	½	the	mean	project‐area	inter‐nest	
distance	of	the	project	footprint	are	potentially	susceptible	to	disturbance	take	and	blade‐
strike	mortality,	as	these	pairs	and	offspring	may	use	the	project	footprint.		We	recommend	
using	this	distance	to	delineate	territories	and	associated	breeding	eagles	at	risk	of	
mortality	or	disturbance.		Exposure	surveys	should	adequately	sample	the	parts	of	the	
project	footprint	potentially	used	by	these	eagle	pairs	so	they	are	captured	in	the	fatality	
estimates,	and	these	nests	should	be	included	in	post‐construction	occupancy	and	
productivity	monitoring	(see	Appendix	H).		This	information	is	useful	in	decisions	on	
whether	a	wind	project	might	meet	permit	requirements	at	50	CFR	22.26	considering	both	
predicted	take	through	fatalities	and	likely	take	from	disturbance;	for	evaluating	various	
siting	and	project‐configuration	alternatives;	and	in	monitoring	for	disturbance	effects	
during	the	post‐construction	period.		In	some	situations,	as	where	nests	are	concentrated	
on	linear	features	(such	as	cliffs	for	golden	eagles	or	along	rivers	for	bald	eagles),	½	the	
mean	inter‐nest	distance	may	not	encompass	all	important	parts	of	the	territory.		In	these	
situations	inferences	based	on	nest	spacing	should	be	used	cautiously.		The	overall	
effectiveness	of	this	approach	will	be	evaluated	through	post‐construction	monitoring	and	
the	adaptive	management	framework	described	later	in	this	ECPG.	

	
b. Additional Considerations for Assessing Project Effects: Migration Corridors and
Stopover Sites
Bald	eagles	and	golden	eagles	tend	to	migrate	along	north‐south	oriented	cliff	lines,	ridges,	
and	escarpments,	where	they	are	buoyed	by	uplift	from	deflected	winds	(Kerlinger	1989,	
Mojica	et	al.	2008).		Bald	eagles	typically	migrate	during	midday	by	soaring	on	thermal	
uplift	or	on	winds	aloft,	the	onset	of	dally	movements	migration	being	influenced	by	rising	
temperatures	and	favorable	winds	(Harmata	2002).	Both	species	will	forage	during	
migration	flights,	though	for	bald	eagles	foraging	often	is	limited	to	lakes,	rivers,	streams,	
and	other	wetland	systems	(Mojica	et	al.	2008).		Both	species	use	lift	from	heated	air	from	
open	landscapes	to	move	efficiently	during	migration	and	seasonal	movements,	gliding	
from	one	thermal	to	the	next	and	sometimes	moving	in	groups	with	other	raptor	species.	
	
Passage	rates	and	altitude	of	migrant	eagles	can	be	influenced	by	temperature,	barometric	
pressure,	winds	aloft,	storm	systems,	weather	patterns	at	the	site	of	origin,	and	wind	speed	
(Yates	et	al.	2001).		Both	species	avoid	large	water	bodies	during	migration	and	funnel	
along	the	shoreline,	often	becoming	concentrated	at	the	tips	of	peninsulas	or	in	other	
situations	where	movement	requires	water	crossings	(Newton	1979).		Eagles	annually	use	
stopover	sites	with	predictably	ample	food	supplies	(e.g.,	Restani	et	al.	2000,	Mojica	et	al.	
2008),	although	some	stopovers	may	be	brief	and	infrequent,	such	as	when	optimal	

15
migration	conditions	suddenly	become	unfavorable	and	eagles	are	forced	to	land	and	seek	
roosts.		Presence	of	a	migration	corridor	or	stopover	site	in	the	project	area	is	best	
documented	and	delineated	by	using	a	standard	“hawk	watch”	migration	count	as	
recommended	in	this	ECPG	as	part	of	site‐specific	surveys	or,	in	some	cases,	by	simply	
expanding	point	count	surveys	to	account	for	migration	incidence	during	what	normally	
would	be	the	peak	migration	period	(Appendix	C).	
	
Much	eagle	mortality	could	occur	if	communal	night	roosts	or	communal	foraging	areas	of	
eagles	are	separated	by	strings	of	wind	turbines	from	other	areas	used	by	eagles.		Outside	
the	breeding	season,	both	bald	eagles	and	golden	eagles	can	roost	communally.		Such	roosts	
can	include	individuals	of	all	ages	and	residency	status	(Platt	1976,	Craig	and	Craig	1984,	
Mojica	et	al.	2008).		During	the	breeding	season,	non‐breeding	bald	eagles	also	may	roost	
communally.		Large	roosts	of	eagles	tend	to	be	associated	with	nearby	foraging	areas.		
Conversely,	eagles	also	may	congregate	to	forage	at	sites	of	unusually	high	prey	or	carcass	
availability;	such	concentrations	of	bald	eagles	may	number	in	the	hundreds	(Buehler	
2000).		Methods	for	documenting	concentrations	of	eagles,	and	movements	to	and	from	
such	areas	in	relation	to	the	project	footprint	are	provided	in	Appendix	C.	

	
2. Eagle Risk Factors
Factors	that	influence	vulnerability	of	eagles	to	collisions	with	wind	turbines	are	poorly	known.		
Theoretically,	two	major	elements	are	likely	involved:	(1)	eagle	abundance,	and	(2)	the	presence	of	
features	or	circumstances	that	decrease	an	eagle’s	ability	to	perceive	and	avoid	collision.		However,	
the	relative	importance	of	these	factors,	and	how	they	interrelate,	remains	poorly	understood	for	
eagles	and	birds	in	general	(Strickland	et	al.	2011).		Table	1	lists	some	of	the	factors	known	or	
postulated	to	be	associated	with	turbineblade‐strike	risk	in	raptors,	but	evidence	for	or	against	
these	is	equivocal,	and	may	well	vary	between	sites.		While	some	of	these	factors	are	not	known	to	
affect	eagles,	because	of	the	similarity	of	flight	behavior	between	eagles	and	some	other	soaring	
raptors,	we	include	them	here	because	they	may	apply	to	eagles.		Evidence	across	multiple	studies	
suggests	that	in	addition	to	eagle	abundance,	two	main	factors	contribute	to	increased	risk	of	
collision	by	eagles:	(1)	the	interaction	of	topographic	features,	season,	and	wind	currents	that	
create	conditions	for	high‐risk	flight	behavior	near	turbines;	and	(2)	behavior	that	distracts	eagles	
and	presumably	makes	them	less	vigilant	(e.g.,	active	foraging	or	inter‐	and	intra‐specific	
interactions).	

	
	
	
Table 1. Factors potentially associated with wind turbine collision risk in raptors.		Not all factors apply to
eagles, and the influence of these factors may vary in association with other covariates on a case-by-case basis.
Risk	Factor	
Bird	Density		

Bird	Age	

Status	of	Knowledge	from	Literature
Mixed	findings;	likely	some	
relationship	but	other	factors	have	
overriding	influence	across	a	range	of	
species.	
Mixed	findings.		Higher	number	of	
fatalities	among	subadult	and	adult	
golden	eagles	in	one	area.		Higher	
fatalities	among	adult	white‐tailed	
eagles	in	another.	
	

Citations	
Barrios	and	Rodriguez	(2004),	De	
Lucas	et	al.	(2008),	Hunt	(2002),	
Smallwood	et	al.	(2009),	Ferrer	et	al.	
(2011)	

Hunt	(2002),	Nygård	et	al	(2010)	

16
Risk	Factor	

Interaction	with	
Other	Birds	

Status	of	Knowledge	from	Literature
White‐tailed	eagle	nesting	areas	close	
to	turbines	have	been	observed	to	have	
low	nest	success	and	be	abandoned	
over	time.	
Mixed	findings.	Higher	risk	to	resident	
adults	in	Egyptian	vultures	(Neophron	
percnopterus).		High	number	of	
mortalities	among	subadults	and	
floating	adults	in	golden	eagles	in	one	
other	study.	
Mixed	findings.		In	some	cases	for	some	
species,	risk	appears	higher	in	seasons	
with	greater	propensity	to	use	slope	
soaring	(fewer	thermals)	or	kiting	
flight	(windy	weather)	while	hunting.	
Species	most	at	risk	perform	more	
frequent	flights	that	can	be	described	
as	kiting,	hovering,	and	diving	for	prey.	
Higher	risk	when	interactive	behavior	
is	occurring.	

Active	Hunting/	
Prey	Availability	

High	risk	when	hunting	close	to	
turbines,	across	a	range	of	species.	

Proximity	to	
Nests	

Bird	Residency	
Status	

Season	

Flight	Style	

Turbine	Height	
Rotor	Speed	
Rotor‐swept	
Area	

Topography	

Wind	Speed	

Mixed,	contradictory	findings	across	a	
range	of	species.	
Higher	risk	associated	with	higher	
blade‐tip	speed	for	golden	eagles	in	one	
study,	but	this	finding	may	not	be	
generally	applicable.	
Meta‐analysis	found	no	effect,	but	
variation	among	studies	clouds	
interpretation.		
Several	studies	show	higher	risk	of	
collisions	with	turbines	on	ridge	lines	
and	on	slopes.		Also	a	higher	risk	in	
saddles	that	present	low‐energy	ridge	
crossing	points.	
Mixed	findings,	probably	locality	
dependent.	

Citations	
Nygård	et	al	(2010)	

Barrios	and	Rodriguez	(2004),	Hunt	
(2002)	

Barrios	and	Rodriguez	(2004),	De	
Lucas	et	al.	(2008),		Hoover	and	
Morrision	(2005),	Smallwood	et	al.	
(2009)	
Smallwood	et	al.	(2009)	
Smallwood	et	al.	(2009)	
Barrios	and	Rodriguez	(2004),	De	
Lucas	et	al.	(2008),	Hoover	and	
Morrision	(2005),	Hunt	(2002),	
Smallwood	et	al.	(2009)	
Barclay	et	al. (2007),	De	Lucas	et	al.	
(2008)	
Chamberlain	et	al.	(2006)	

Barclay	et	al.	(2007)	
Barrios	and	Rodriguez	(2004),	De	
Lucas	et	al.	(2008),		Hoover	and	
Morrission	(2005),	Smallwood	and	
Thelander	(2004)	
Barrios	and	Rodriguez	(2004),	
Hoover	and	Morrision	(2005),	
Smallwood	et	al.	(2009)	

	

3. Overview of Process to Assess Risk
This	ECPG,	and	in	particular	the	eagle	fatality	prediction	model	described	in	Appendix	D,	relies	on	
the	assumption	that	there	is	predictable	relationship	between	pre‐construction	eagle	occurrence	
and	abundance	in	the	project	footprint	and	subsequent	fatalities.		Assessing	the	veracity	of	this	
operating	hypothesis	is	a	key	element	of	the	adaptive	management	component	of	the	ECPG.		The	
ECPG	outlines	a	decision‐making	process	that	gathers	information	at	each	stage	of	project	
development,	with	an	increasing	level	of	detail.		This	approach	provides	a	framework	for	making	

17
decisions	sequentially	at	three	critical	phases	in	project	development:	(1)	siting,	(2)	construction,	
and	(3)	operations.		The	greatest	potential	to	avoid	and	minimize	impacts	to	eagles	occurs	if	eagle	
risk	factors	are	taken	into	account	at	the	earliest	phase	of	project	development.		If	siting	and	
construction	have	proceeded	without	consideration	of	risks	to	eagles,	significant	opportunities	to	
avoid	and	minimize	risk	may	have	been	lost.		This	can	potentially	result	in	greater	compensatory	
mitigation	requirements	or,	in	the	worst	case,	an	unacceptable	level	of	mortality	for	eagles.	
	
The	related,	but	more	general,	WEG	advocates	using	a	five‐tiered	approach	for	iterative	decision	
making	relative	to	assessing	and	addressing	wildlife	effects	from	wind	facilities.		Elements	of	all	of	
those	tiers	apply	here,	but	the	process	for	eagles	is	more	specifically	defined	and	falls	into	five	
broadly	overlapping,	iterative	stages	that	largely	do	not	parallel	the	WEG’s	five	tiers	(Figures	1	and	
2).	
	
Stage	1	for	eagles	(Appendix	B)	combines	Tiers	1	and	2	from	the	WEG,	and	consists	of	an	initial	site	
assessment.		In	this	stage	project	developers	or	operators	evaluate	broad	geographic	areas	to	
assess	the	relative	importance	of	various	areas	to	resident	breeding	and	non‐breeding	eagles,	and	
to	migrant	and	wintering	eagles.		The	Service	is	available	to	assist	project	developers	or	operators	
in	beginning	to	identify	important	eagle	use	areas	or	migration	concentration	sites	and	potential	
eagle	habitat	at	this	stage.		To	increase	the	probability	of	meeting	the	regulatory	requirements	for	a	
programmatic	take	permit,	biological	advice	from	the	Service	and	other	jurisdictional	wildlife	
agencies	should	be	requested	as	early	as	possible	in	the	developer's	planning	process	and	should	be	
as	inclusive	as	possible	to	ensure	all	issues	are	being	address	at	the	same	time	and	in	a	coordinated	
manner.		Ideally,	consultation	with	the	Service,	and	state	and	tribal	wildlife	agencies	is	done	prior	to	
any	substantial	financial	commitment	or	finalization	of	lease	agreements.		During	Stage	1	the	
project	developer	or	operator	should	gather	existing	information	from	publicly	available	literature,	
databases,	and	other	sources,	and	use	those	data	to	judge	the	appropriateness	of	various	potential	
project	sites,	balancing	suitability	for	development	with	potential	risk	to	eagles.	
	
Once	a	site	has	been	selected,	the	next	stage,	Stage	2,	is	site‐specific	surveys	and	assessments	
(this	is	the	first	component	of	Tier	3	in	the	WEG;	Appendix	C).		During	Stage	2	the	project	developer	
or	operator	should	collect	quantitative	data	through	scientifically	rigorous	surveys	designed	to	
assess	the	potential	risk	of	the	proposed	project	to	eagles.		In	the	case	of	small	wind	projects	(one	
or	a	few	small	turbines),	the	project	developer	or	operator	should	apply	the	predictive	model	
described	in	Stage	3	(below)	to	determine	if	stage	2	surveys	are	necessary.		In	many	cases,	the	
hazardous	area	associated	with	such	projects	will	be	small	enough	that	Stage	2	surveys	will	not	be	
necessary	to	demonstrate	that	the	project	will	likely	not	take	eagles.	
	
In	Stage	3,	the	predicting	eagle	fatalities	stage,	the	Service	and	project	developers	or	operators	
use	data	from	Stage	2	in	standardized	models	linked	to	the	Service’s	adaptive	management	process	
to	generate	predictions	of	eagle	risk	in	the	form	of	average	number	of	fatalities	per	year	
extrapolated	to	the	tenure	of	the	permit	(see	Appendix	D).		These	models	can	be	used	to	
comparatively	evaluate	alternative	siting,	construction,	and	operational	scenarios,	a	useful	feature	
in	constructing	hypotheses	regarding	predicted	effects	of	conservation	measures	and	ACPs.		We	
encourage	project	developers	or	operators	to	use	the	recommended	pre‐construction	survey	
protocol	in	this	ECPG	in	Stage	2	to	help	inform	our	predictive	models	in	Stage	3.		If	Service‐
recommended	survey	protocols	are	used,	this	risk	assessment	can	be	greatly	facilitated	using	model	
tools	available	from	the	Service.		If	project	developers	or	operators	use	other	forms	of	information	
for	the	Stage	2	assessment,	they	will	need	to	fully	describe	those	methods	and	the	analysis	used	for	
the	eagle	risk	assessment,	and	more	time	will	be	required	for	Service	biologists	to	evaluate	and	

18
review	the	data.		For	example,	the	Service	will	compare	the	results	of	the	project	developer	or	
operator’s	eagle	risk	assessment	with	predictions	from	our	models,	and	if	the	results	differ,	we	will		

	
Figure 1. Chart comparing Land-based Wind Energy Guideline tiers with Eagle Conservation Plan Guidance stages.

	
	

work	with	the	project	developers	or	operators	to	determine	which	model	results	are	most	
appropriate	for	the	Service’s	eventual	permitting	decisions.		The	Service	and	project	developers	or	
operators	also	evaluate	Stage	2	data	to	determine	whether	disturbance	take	is	likely,	and	if	so,	at	
what	level.		Any	loss	of	production	that	may	stem	from	disturbance	should	be	added	to	the	fatality	
rate	prediction	for	the	project.		The	risk	assessments	at	Stage	2	and	Stage	3	are	consistent	with	
developing	the	information	necessary	to	assess	the	efficacy	of	conservation	measures,	and	to	
develop	the	monitoring	required	by	the	permit	regulations	at	50	CFR	22.26(c)(2).	

			

Stage	4	is	the	avoidance	and	minimization	of	risk	using	conservation	measures	and	ACPs	and	
compensatory	mitigation	(if	required).	
	
Conservation	measures	and	ACPs.		Regardless	of	which	approach	is	employed	in	the	Stage	
3	assessment,	in	Stage	4	the	information	gathered	should	be	used	by	the	project	developer	
or	operator	and	the	Service	to	determine	potential	conservation	measures	and	ACPs	(if	
available)	that	can	be	employed	to	avoid	and/or	minimize	the	predicted	risks	at	a	given	site	
(see	Appendix	E).		The	Service	will	compare	the	initial	predictions	of	eagle	mortality	and	
disturbance	for	the	project	with	predictions	that	take	into	account	proposed	and	potential	
conservation	measures	and	ACPs	to	determine	if	the	project	developer	or	operator	has	
avoided	and	minimized	risks	to	the	maximum	degree	achievable,	thereby	meeting	the	
requirements	for	programmatic	permits	in	50	CFR	22.26	that	remaining	take	is	
unavoidable.		Additionally,	the	Service	will	use	the	information	provided	along	with	other	

19
data	to	conduct	a	cumulative	effects	analysis	to	determine	if	the	project’s	impacts,	in	
combination	with	other	permitted	take	and	other	known	factors	affecting	the	local‐area	and		

20

	
	

Figure 2. Figure 1 from WEG, adapted to show where and how eagles are considered in that process and which Stage and section of the ECPG
are applicable at each Tier of the WEG. Note that existing, operational wind energy projects enter the process between Tiers 3 and 4.

21
eagle	management	unit	population(s),	are	at	a	level	that	exceed	established	thresholds	or	
benchmarks	(see	Appendix	F).		This	final	eagle	risk	assessment	is	completed	at	the	end	of	
Stage	4	after	application	of	conservation	measures	and	ACPs	along	with	a	plan	for	
compensatory	mitigation	if	required.	
	
Compensatory	Mitigation.		Compensatory	mitigation	occurs	in	the	eagle	permitting	
process	if	conservation	measures	and	ACPs	do	not	remove	the	potential	for	take,	and	the	
projected	take	exceeds	calculated	thresholds	for	the	species‐specific	eagle	management	unit	
in	which	the	project	is	located.		Compensatory	mitigation	may	also	be	necessary	in	other	
situations	as	described	in	the	preamble	to	50	CFR	22.26	(USFWS	2009a),	and	the	following	
guidance	applies	to	those	situations	as	well.	
	
Compensatory	mitigation	can	address	any	pre‐existing	mortality	source	affecting	the	
species‐specific	eagle	management	unit	impacted	by	the	project	(e.g.	environmental	lead	
abatement,	addressing	eagle	electrocutions	due	to	high	risk	power	poles,	etc.)	that	was	in	
effect	at	the	time	of	the	FEA	in	2009	(USFWS	2009b),	or	it	can	address	increasing	the	
carrying	capacity	of	the	eagle	population	in	the	affected	eagle	management	unit.		However,	
there	needs	to	be	a	credible	analysis	that	supports	the	conclusion	that	implementing	the	
compensatory	mitigation	action	will	achieve	the	desired	beneficial	offset	in	mortality	or	
carrying	capacity.		All	compensatory	mitigation	projects	will	be	subjected	to	random	
inspections	by	the	Service	or	appointed	subcontractors	to	examine	efficacy,	accuracy,	and	
reporting	rigor.	
	
For	new	wind	development	projects,	if	compensatory	mitigation	is	necessary,	the	
compensatory	mitigation	action	(or	a	verifiable,	legal	commitment	to	such	mitigation)	will	
be	required	up	front	before	project	operations	commence	because	projects	must	meet	the	
statutory	and	regulatory	eagle	preservation	standard	before	the	Service	may	issue	a	permit.		
For	operating	projects	that	may	meet	permitting	requirements,	compensatory	mitigation	
should	be	applied	from	the	start	of	the	permit	period,	not	retroactively	from	the	initiation	of	
project	operations.		The	initial	compensatory	mitigation	contribution	effort	should	be	
sufficient	to	offset	take	at	the	upper	80%	confidence	limit	(or	equivalent)	of	the	predicted	
number	of	eagle	fatalities	per	year	for	a	five‐year	period	starting	with	the	date	the	project	
becomes	operational	(or,	for	operating	projects,	the	date	the	permit	is	signed).		No	later	
than	at	the	end	of	the	five	year	period,	the	predicted	annual	take	estimate	will	be	compared	
to	the	realized	take	as	estimated	by	post‐construction	monitoring.		If	the	triggers	identified	
in	the	permit	for	adjustment	of	compensatory	mitigation	are	met,	those	adjustments	should	
be	implemented.	In	the	case	where	the		realized	take	is	less	than	predicted,	the	permittee	
will	receive	a	credit	for	the	excess	compensation	(the	difference	between	the	actual	mean	
and	the	number	compensated	for)	that	can	be	applied	to	other	take	(either	by	the	permittee	
or	other	permitted	individuals	at	his/her	discretion)	within	the	same	eagle	management	
unit.		Compensatory	mitigation	for	future	years	for	the	project	will	be	determined	at	this	
point,	taking	into	account	the	observed	levels	of	mortality	and	any	reduction	in	that	
mortality	that	is	expected	based	on	implementation	of	additional	experimental	
conservation	measures	and	ACPs	that	might	reduce	fatalities.	
	
To	illustrate	an	acceptable	process	for	calculating	compensatory	mitigation,	the	Service	has	
prepared	an	example	of	a	strategy	using	Resource	Equivalency	Analysis	(REA)	to	quantify	
the	number	of	power	pole	retrofits	needed	to	offset	the	take	of	golden	eagles	at	a	wind	
project	(see	Appendix	G).		The	Service	used	the	example	of	eliminating	electrocutions	
because:	(1)	high‐risk	power	poles	cause	quantifiable	adverse	impacts	to	eagles;	(2)	the	‘per	

22
eagle’	effects	of	high‐risk	power	pole	retrofitting	are	quantifiable	and	verifiable	through	
accepted	practices;	(3)	success	of	and	subsequent	maintenance	of	retrofitting	can	be	
monitored;	and	(4)	electrocution	from	high‐risk	power	poles	is	known	to	cause	eagle	
mortality	and	this	can	be	corrected.		The	potential	for	take	of	eagles	is	estimated	using	
informed	modeling,	as	described	in	Stage	3	of	the	ECPG	(Appendix	D).		This	fatality	
prediction	is	one	of	several	fundamental	variables	that	are	used	to	populate	the	REA	(see	
REA	Inputs,	Appendix	G).		The	REA	generates	a	project‐area	eagle	impact	calculation	
(debit),	expressed	in	bird‐years,	and	an	estimate	of	the	quantity	of	compensatory	mitigation	
(credit)	(e.g.,	power	pole	retrofits)	necessary	to	offset	this	impact.		Compensatory	
mitigation	would	then	be	implemented	either	directly	by	the	project	developer	or	operator	
or	through	a	formal,	binding	agreement	with	a	third	party	to	implement	the	required	
actions.	
	
Effectiveness	monitoring	of	the	resulting	compensatory	mitigation	projects	should	be	
included	within	the	above	options	using	the	best	scientific	and	practicable	method	
available.		The	Service	will	modify	the	compensatory	mitigation	process	to	adapt	to	any	
improvements	in	our	knowledge	base	as	new	data	become	available.	
	
At	the	end	of	Stage	4,	all	the	materials	necessary	to	satisfy	the	regulatory	requirements	to	support	a	
permit	application	should	be	available.		While	the	application	can	be	submitted	at	any	time,	it	is	
only	after	completion	of	Stage	4	that	the	Service	can	begin	the	formal	process	to	determine	whether	
a	programmatic	eagle	take	permit	can	be	issued	or	not.		Ideally,	NEPA	and	NHPA	analyses	and	
assessments	will	already	be	underway,	but	if	not,	Stage	4	should	include	necessary	NEPA	analysis,	
NHPA	compliance,	coordination	with	other	jurisdictional	agencies,	and	tribal	consultation.	
	
If	a	permit	is	issued	and	the	project	goes	forward,	Stage	5	of	the	process	is	calibration	and	updating	
of	the	fatality	prediction	and	continued	risk	assessment,	equivalent	to	Tier	4	and,	in	part,	Tier	5	in	
the	WEG.		During	this	stage,	post‐construction	surveys	are	conducted	to	generate	empirical	data	for	
comparison	with	the	pre‐construction	risk‐assessment	fatality	and	disturbance	predictions.		The	
monitoring	protocol	should	include	both	validated	techniques	for	assessing	mortality,	and	for	
estimating	effects	of	disturbance	to	eagles,	and	they	must	meet	the	permit‐condition	requirements	
at	50	CFR	22.26(c)(2).		We	anticipate	that	in	most	cases,	intensive	monitoring	to	estimate	the	true	
annual	fatality	rate	and	to	assess	possible	disturbance	effects	will	be	conducted	for	at	least	the	first	
two	years	after	permit	issuance,	followed	by	less	intense	monitoring	for	up	to	three	years	after	the	
expiration	date	of	the	permit,	in	accordance	with	monitoring	requirements	at	50	CFR	22.26(c)(2).		
We	recommend	project	developers	or	operators	use	the	post‐construction	survey	protocols	
included	or	referenced	in	this	ECPG,	but	we	will	consider	other	monitoring	protocols	provided	by	
permit	applicants.		We	will	use	the	information	from	post‐construction	monitoring	in	a	meta‐
analysis	framework	to	weight	and	improve	pre‐construction	predictive	models.		Additionally	in	
Stage	5	the	Service	and	project	developers	or	operators	should	use	the	post‐construction	
monitoring	data	to	(1)	assess	whether	compensatory	mitigation	is	adequate,	excessive,	or	deficient	
to	offset	observed	mortality,	and	make	adjustments	accordingly;	and	(2)	explore	operational	
changes	that	might	be	warranted	at	a	project	after	permitting	to	reduce	observed	mortality	and	
ensure	that	permit	condition	requirements	at	50	CFR	22.26(c)(7)	are	met.	
	
Table	2	provides	a	summary	of	the	roles	of	the	project	developer	or	operator	and	the	Service,	
responsibilities,	and	decision	points	at	each	stage.	

	

23

Table 2. Roles, responsibilities of the project developers and operators and the Service, and decision
points at each stage of the ECP process.
	
Stage	

Project	developer/operator	role


1	






2	






3	








4	


Conduct	a	desktop	landscape‐level	
assessment	for	known	or	likely	
occurrence	of	eagles,	including	
reconnaissance	visits	to	prospective	
sites.	
Consult	with	the	Service	on	potential	for	
any	obvious	negative	impacts	on	eagles	
in	at	least	general	locale	of	prospective	
sites.	
Decision	point:	select	site(s)	for	Stage	2	
study,	if	appropriate.	
Conduct	detailed,	site‐specific	field	
studies	in	the	project	area	to	inform	
eagle	fatality	prediction	model,	
document	important	eagle	use	areas	or	
migration	concentration	sites,	and	
identify	possible	eagle	disturbance	
issues.	
Coordinate	in	advance	with	the	Service	
and	other	jurisdictional	agencies	to	
ensure	studies	will	satisfy	regulatory	
requirements	for	permitting.	
Decision	point:	choose	whether	to	move	
to	Stage	3.	

Service	role	










Optionally	generate	an	estimated	annual	
eagle	fatality	prediction	for	the	site(s)	
and	an	assessment	of	eagle	disturbance	 
risk	using	data	from	Stage	2	and	
model(s)	of	choice.	
Report	on	all	other	germane	aspects	of	

eagle	use	such	as	communal	roosts	and	
nest	or	territory	locations.	

Decision	point:	choose	whether	to	move	
to	Stage	4.	

Identify	conservation	measures	and	ACP	 
s	that	can	be	used	to	avoid	and	minimize	
take	identified	in	Stage	3.	

Optionally	generate	revised	fatality	and	
disturbance	estimates,	taking	into	

account	conservation	measures	and	
ACPs.	

Identify	and	develop	necessary	
agreements	for	compensatory	

mitigation	to	offset	take,	if	required.	

Recommend	and	help	provide	existing	data	and	
input	if	requested.	
Provide	preliminary	consultation	on	
appropriateness	of	application	for	eagle	take	
permits	for	sites	considered	and	the	likelihood	
permits	could	be	issued.	
Review	available	Stage	1	data	and	advise	what	
Stage	2	data	are	recommended.	
Decision	point:	none.	

Consult	on	field	study	design	and	approach	in	
coordination	with	other	jurisdictional	agencies.	
Decision	point:	None.	

Generate	an	initial	eagle	fatality	estimate	for	
site(s),	using	the	Service	model	and	survey	data	
from	Stage	2.	
Assess	likelihood	of	disturbance	to	eagles;	
quantify	extent	and	impact	of	disturbance,	if	any	
likely.	
Make	preliminary	recommendation	on	risk	
category.	
Consult	with	developer/operator	to	interpret	and	
resolve	discrepancies	in	conclusions	and	risk	
category	recommendation.	
Decision	point:	None.	
Re‐run	Service	fatality	model	to	predict	fatalities	
with	conservation	measures	and	ACPs.	
Re‐assess	potential	for	disturbance	take	with	
conservation	measures	and	ACPs.	
Coordinate	with	developer/operator	to	reach	
agreement	on	predicted	take	and	risk	category.	
Coordinate	with	developer/operator	on	
compensatory	mitigation,	if	requested.	
Provide	revised	preliminary	assessment	of	
likelihood	site(s)	will	be	permittable	if	requested.	

24
Stage	

Project	developer/operator	role


Decision	point:	choose	whether	to	
submit	eagle	take	permit	application.	

Service	role	


Decision	point:	None.	



Coordinate	and	consult	on	writing	of	ECP	or	
equivalent,	including	proposed	plan	for	post‐
construction.	
Convey	adequacy	of	ECP	or	equivalent	to	
developer/operator.	
Evaluate	permit	application	for	regulatory	
sufficiency.	
Draft	permit	conditions	drawing	on	relevant	
components	of	ECP	or	equivalent.	
Conduct	cumulative	effects	analysis.	
Conduct	NEPA	review.	
Conduct	NHPA	evaluation.	
Coordinate	with	other	jurisdictional	agencies.	
Consult	with	Tribes.	
Establish	limits	on	future	operational	adjustments	
proportionate	to	risk,	in	coordination	with	
applicant.	
Decision	point:	whether	permit	can	be	issued.	
Monitor	compliance	with	permit	conditions.	
Review	post‐construction	monitoring	data,	
including	comparison	of	predicted	and	observed	
annual	fatality	rate	and	disturbance.	
At	no	more	than	5‐year	intervals,	determine	
whether	revision	of	the	estimated	fatality	rate,	
adjustments	to	monitoring,	implementation	of	
additional	experimental	conservation	measures	
and	ACPs,	and	compensatory	mitigation	are	
warranted.	
Effect	any	necessary	adjustments	by	crediting	
back	excess	compensatory	mitigation,	or	by	
assessing	additional	compensatory	mitigation	for	
fatalities	in	excess	of	predictions.	
Combine	monitoring	data	with	that	from	other	
projects	for	meta‐analysis	within	adaptive	
management	framework.	
Decision	point:	determine	what	adjustments	need	
to	be	made	to	compensatory	mitigation	level,	and	
whether	additional	conservation	measures	and	
ACPs	are	warranted	or	not.	





Permit	
Decision	



Draft	ECP	or	equivalent,	including	a	plan	
for	post‐construction	monitoring	of	
eagle	fatality	and	disturbance.	
Submit	a	permit	application	that	meets	
requirements	at	50	CFR	22.26	or	22.27,	
including	ECP	or	equivalent	information	
as	part	of	application	package.	
Choose	whether	to	assist	Service	in	
conducting	NEPA.	
Decision	point:	None.	

















5	




Implement	post‐construction	
monitoring	in	accordance	with	permit	
conditions,	including	immediate	
reporting	of	any	eagle	take.	
Participate	in	scheduled	reviews	of	
post‐construction	monitoring	results.	
Effect	additional	compensatory	
mitigation	if	necessary.	
Implement	and	monitor	additional	
conservation	measures	and	ACPS,	if	
warranted,	within	scope	of	permit	
sideboards.	
Decision	point:	choose	whether	to	apply	
for	permit	renewal	near	the	end	of	
permit	term.	









25

4. Site Categorization Based on Mortality Risk to Eagles
We	recommend	the	approach	outlined	below	be	used	to	categorize	the	likelihood	that	a	site	or	
operational	alternative	will	meet	standards	in	50	CFR	22.26	for	issuance	of	a	programmatic	eagle	
take	permit.	

	
a. Category 1 – High risk to eagles, potential to avoid or mitigate impacts is low
A	project	is	in	this	category	if	it:	
(1)	has	an	important	eagle‐use	area	or	migration	concentration	site	within	the	project	
footprint;	or	
(2)	has	a	species‐specific	uncertainty‐adjusted	annual	fatality	estimate	(average	number	
of	eagles	predicted	to	be	taken	annually)		>	5%	of	the	estimated	species‐specific	
local‐area	population	size;	or	
(3)	causes	the	cumulative	annual	take	for	the	local‐area	population	to	exceed	5%	of	the	
estimated	species‐specific	local‐area	population	size.	
	
In	addition,	projects	that	have	eagle	nests	within	½	the	mean	project‐area	inter‐nest	
distance	of	the	project	footprint	should	be	carefully	evaluated	(see	Appendix	H).		If	it	is	
likely	eagles	occupying	these	territories	use	or	pass	through	the	project	footprint,	category	
1	designation	may	be	appropriate.	
	
Projects	or	alternatives	in	category	1	should	be	substantially	redesigned	if	they	are	to	at	
least	meet	the	category	2	criteria.		Construction	of	projects	at	sites	in	category	1	is	not	
recommended	because	the	project	would	likely	not	meet	the	regulatory	requirements	for	
permit	issuance	and	may	place	the	project	developer	or	operator	at	risk	of	violating	the	
BGEPA.		The	recommended	approach	for	assessing	the	percentage	of	the	local‐area	
population	predicted	to	be	taken	is	described	in	Appendix	F.	

	
b. Category 2 – High or moderate risk to eagles, opportunity to mitigate impacts
A	project	is	in	this	category	if	it:	
(1)	has	an	important	eagle‐use	area	or	migration	concentration	site	within	the	project	
area	but	not	in	the	project	footprint;	or	
(2)	has	a	species‐specific	uncertainty‐adjusted	fatality	estimate	between	0.03	eagles	per	
year	and	5%	of	the	estimated	species‐specific	local‐area	population	size;	or	
(3)	causes	cumulative	annual	take	of	the	species‐specific	local‐area	population	of		less	
than	5%	of	the	estimated	local‐area	population	size.	
	
Projects	in	this	category	will	potentially	take	eagles	at	a	rate	greater	than	is	consistent	with	
maintaining	stable	or	increasing	populations,	but	the	risk	might	be	reduced	to	an	acceptable	
level	through	a	combination	of	conservation	measures	and	reasonable	compensatory	
mitigation.		These	projects	have	a	risk	of	ongoing	take	of	eagles,	but	this	risk	can	be	
minimized.		For	projects	in	this	category	the	project	developer	or	operator	should	prepare	
an	ECP	or	similar	plan	to	document	meeting	the	regulatory	requirements	for	a	
programmatic	permit.		For	eagle	management	populations	where	take	thresholds	are	set	at	
zero,	the	conservation	measures	in	the	ECP	should	include	compensatory	mitigation	and	
must	result	in	no‐net‐loss	to	the	breeding	population	to	be	compatible	with	the	permit	
regulations.		This	does	not	apply	to	golden	eagles	east	of	the	100th	meridian,	for	which	no	
non‐emergency	take	can	presently	be	authorized	(USFWS	2009b).	

	

26

c. Category 3 – Minimal risk to eagles
A	project	is	in	this	category	if	it:	
(1)	has	no	important	eagle	use	areas	or	migration	concentration	sites	within	the	project	
area;	and	
(2)	has	a	species‐specific	uncertainty‐adjusted	annual	fatality	rate	estimate	of	less	than	
0.03	for	both	species	of	eagle;	and	
(3)	causes	cumulative	annual	take	of	the	local‐area	population	of	less	than	5%	of	the	
estimated	species‐specific	local‐area	population	size.	
	
Projects	in	category	3	pose	little	risk	to	eagles	and	may	not	require	or	warrant	eagle	take	
permits,	but	that	decision	should	be	made	in	coordination	with	the	Service.		Still,	a	project	
developer	or	operator	may	wish	to	create	an	ECP	that	documents	the	project’s	low	risk	to	
eagles,	and	outlines	mortality	monitoring	for	eagles	and	a	plan	of	action	if	eagles	are	taken	
during	project	construction	or	operation.		If	take	should	occur,	the	developer	or	operator	
should	contact	the	Service	to	discuss	ways	to	avoid	take	in	the	future.		Such	an	ECP	would	
enable	the	Service	to	provide	a	permit	to	allow	a	de	minimis	amount	of	take	if	the	project	
developer	or	operator	wished	to	obtain	such	a	permit.	
	
The	risk	category	of	a	project	has	the	potential	to	change	from	one	of	higher	risk	to	one	of	lower	
risk	or	one	of	lower	risk	to	one	of	higher	risk	through	additional	site‐specific	analyses	and	
application	of	measures	to	reduce	the	risk.		For	example,	a	project	may	appear	to	be	in	category	2	as	
a	result	of	Stage	1	analyses,	but	after	collection	of	site‐specific	information	in	Stage	2	it	might	
become	clear	it	is	a	category	1	project.		If	a	project	cannot	practically	be	placed	in	one	of	these	
categories,	the	project	developer	or	operator	and	the	Service	should	work	together	to	determine	if	
the	project	can	meet	programmatic	eagle	take	permitting	requirements	in	50	CFR	22.26	and	22.27.		
Projects	should	be	placed	in	the	highest	category	(with	category	1	being	the	highest)	in	which	one	
or	more	of	the	criteria	are	met.	

	
5. Cumulative Effects Considerations
a. Early Planning
Regulations	at	50	CFR	22.26	require	the	Service	to	consider	the	cumulative	effects	of	
programmatic	eagle	take	permits.		Cumulative	effects	are	defined	as:	“the	incremental	
environmental	impact	or	effect	of	the	proposed	action,	together	with	impacts	of	past,	
present,	and	reasonably	foreseeable	future	actions”	(50	CFR	22.3).		Thorough	cumulative	
effects	analysis	will	depend	on	effective	analysis	during	the	NEPA	process	associated	with	
an	eagle	permit.		Scoping	and	other	types	of	preliminary	analyses	can	help	identify	
important	cumulative‐effects	factors	and	identify	applicable	past,	present,	and	future	
actions.	Comprehensive	evaluation	during	early	planning	may	identify	measures	that	would	
avoid	and	minimize	the	effects	to	the	degree	that	take	of	eagles	is	not	likely	to	occur.		In	that	
case,	there	may	be	no	permit,	and	thus	no	need	for	NEPA	associated	with	an	eagle	take	
permit.		When	a	wind	project	developer	or	operator	seeks	an	eagle	take	permit,	a	
comprehensive	cumulative	effects	analysis	at	the	early	planning	stage	will	serve	to	
streamline	subsequent	steps,	including	the	NEPA	process.	
	
The	Service	recommends	that	cumulative	effects	analyses	be	consistent	with	the	principles	
of	cumulative	effects	outlined	in	the	Council	on	Environmental	Quality	(CEQ)	handbook,	
"Considering	Cumulative	Effects	under	the	National	Environmental	Policy	Act	(1997)	(CEQ	
handbook).	The	Service	recommends	consideration	of	the	following	examples	from	the	CEQ	

27
handbook	that	may	apply	to	cumulative	effects	to	eagles	and	the	ecosystems	they	depend	
upon:	
	
(1)	Time	crowding	‐	frequent	and	repetitive	effects	on	an	environmental	system;	
(2)	Time	lags	‐	delayed	effects;	
(3)	Space	crowding	‐	High	spatial	density	of	effects	on	an	environmental	system;	
(4)	Cross‐	boundary	‐	Effects	occur	away	from	the	source;	
(5)	Fragmentation	‐	change	in	landscape	pattern;	
(6)	Compounding	effects	‐	Effects	arising	from	multiple	sources	or	pathways;	
(7)	Indirect	effects	‐	secondary	effects;	and	
(8)	Triggers	and	thresholds	‐	fundamental	changes	in	system	behavior	or	structure.	

	
b. Analysis Associated with Permits
The	cumulative	effects	analysis	for	a	wind	project	and	a	permit	authorization	should	include	
whether	the	anticipated	take	of	eagles	is	compatible	with	eagle	preservation	as	required	at	
50	CFR	22.26,	including	indirect	impacts	associated	with	the	take	that	may	affect	eagle	
populations.		It	should	also	include	consideration	of	the	cumulative	effects	of	other	
permitted	take	and	additional	factors	affecting	eagle	populations.	
	
Whether	or	not	a	permit	authorization	is	compatible	with	eagle	preservation	was	analyzed	
in	the	FEA	that	established	the	thresholds	for	take	(USFWS	2009b).		The	scale	of	that	
analysis	was	based	upon	eagle	management	units	as	defined	in	USFWS	(2009b).		However,	
the	scale	for	cumulative	effects	analysis	of	wind	projects	and	associated	permits	should	
include	consideration	of	the	effects	at	the	local‐population	scale	as	well.	
	
The	cumulative	effects	analyses	for	programmatic	permits	should	cover	the	time	period	
over	which	the	take	will	occur,	not	just	the	period	the	permit	will	cover,	including	the	effect	
of	the	proposed	action,	other	actions	affecting	eagles,	predicted	climate	change	impacts,	and	
predicted	changes	in	number	and	distribution	of	affected	eagle	populations.		Effects	
analyses	should	note	whether	the	project	is	located	in	areas	where	eagle	populations	are	
increasing	or	predicted	to	increase	based	on	available	data,	over	the	lifetime	of	the	project,	
even	if	take	is	not	anticipated	in	the	immediate	future.		In	addition,	conditions	where	
populations	are	saturated	should	be	considered	in	cumulative	effects	analyses.		Numerous	
relatively	minor	disruptions	to	eagle	behavior	from	multiple	activities,	even	if	spatially	or	
temporally	distributed,	may	lead	to	disturbance	that	would	not	have	resulted	from	fewer	or	
more	carefully	sited	activities	(e.g.,	Whitfield	et	al.	2007).		Additional	detailed	guidance	for	
cumulative	impacts	analyses	can	be	found	on	the	Council	on	Environmental	Quality	website	
at	http://ceq.hss.doe.gov/nepa/	ccenepa/	ccenepa.htm.	
	
Specific	recommendations	for	conducting	cumulative	effects	analysis	of	the	authorized	take	
under	eagle	programmatic	take	permits	is	provided	in	Appendix	F.	

	
	

28

ADAPTIVE MANAGEMENT
	

Management	of	wind	facilities	to	minimize	eagle	take,	through	decisions	about	siting,	design,	
operation,	and	compensatory	mitigation,	is	a	set	of	recurrent	decisions	made	in	the	face	of	
uncertainty.		The	Department	of	the	Interior	has	a	long	history	of	approaching	such	decisions	
through	a	process	of	adaptive	management	(Williams	et	al.	2007).		The	purpose	of	adaptive	
management	is	to	improve	long‐term	management	outcomes,	by	recognizing	where	key	
uncertainties	impede	decision	making,	seeking	to	reduce	those	uncertainties	over	time,	and	
applying	that	learning	to	subsequent	decisions	(Walters	1986).	
	
In	the	case	of	managing	eagle	populations	in	the	face	of	energy	development	there	is	considerable	
uncertainty	to	be	reduced.		For	example,	evidence	shows	that	in	some	areas	or	specific	situations,	
large	soaring	birds,	specifically	raptors,	are	vulnerable	to	colliding	with	wind	turbines	(Barrios	and	
Rodriguez	2004,	Kuvlesky	et	al.	2007).		However,	we	are	uncertain	about	the	relative	importance	of	
factors	that	influence	that	risk.		We	are	also	uncertain	about	the	best	way	to	mitigate	the	effects	of	
wind	turbine	developments	on	raptors;	we	suspect	some	strategies	might	be	effective,	others	are	
worth	trying.		We	also	suspect	that	a	few	species,	including	golden	eagles	(USFWS	2009b),	may	be	
susceptible	enough	to	collisions	with	wind	turbines	that	populations	may	be	negatively	affected.		
Thus,	there	are	uncertainties	at	several	levels	that	challenge	our	attempts	to	manage	eagle	
populations:	(1)	at	the	level	of	understanding	factors	that	affect	collision	risk,	(2)	at	the	level	that	
influences	population	trends,	and	(3)	about	the	efficacy	of	various	mitigation	options.		The	Service,	
our	conservation	partners,	and	industry	will	never	have	the	luxury	of	perfect	information	before	
needing	to	act	to	manage	eagles.		Our	goal	is	to	reduce	that	uncertainty	through	use	of	formal	
adaptive	management,	thereby	improving	our	predictive	capability	over	time.		Applying	a	
systematic,	cohesive,	nationally‐consistent	strategy	of	management	and	monitoring	is	necessary	to	
accomplish	this	goal.	
	
In	the	context	of	wind	energy	development	and	eagle	management	under	the	ECPG,	there	are	four	
specific	sets	of	decisions	that	will	be	approached	through	adaptive	management:	(1)	adaptive	
management	of	wind	project	operations;	(2)	adaptive	management	of	wind	project	siting	and	
design	recommendations;	(3)	adaptive	management	of	compensatory	mitigation;	and	(4)	adaptive	
management	of	population‐level	take	thresholds.		These	are	discussed	in	more	detail	in	Appendix	A.		
The	adaptive	management	process	will	depend	heavily	on	pre‐	and	post‐construction	data	from	
individual	projects,	but	analyses,	assessment,	and	model	evaluation	will	rely	on	data	pooled	over	
many	individual	wind	projects.		Therefore,	individual	project	developers	or	operators	will	have	
limited	direct	responsibilities	for	conducting	adaptive	management	analyses,	other	than	to	provide	
data	through	post‐construction	monitoring.	

	
	

	

29

EAGLE CONSERVATION PLAN DEVELOPMENT PROCESS
	

The	following	sections	of	the	ECPG,	including	attached	appendices,	provide	a	descriptive	
instructional	template	for	developing	an	ECP.		Throughout	this	section,	we	use	the	term	ECP	to	
include	any	other	document	or	collection	of	documents	that	could	be	considered	equivalent	to	an	
ECP.		The	ECP	is	an	integral	part	of	the	permit	process,	and	the	following	chronological	step‐by‐step	
outline	shows	how	the	pieces	fit	together:	
	
The	ECPG	provides	guidance	and	serves	as	a	reference	for	project	developers	or	operators,	the	
Service,	and	other	jurisdictional	agency	biologists	when	developing	and	evaluating	ECPs.		Using	the	
ECPG	as	a	non‐binding	reference,	the	Service	will	work	with	project	developers	or	operators	to	
develop	an	ECP.		The	ECP	documents	how	the	project	developer	or	operator	intends	to	comply	with	
the	regulatory	requirements	for	programmatic	permits	and	the	associated	NEPA	process	by	
avoiding	and	minimizing	the	risk	of	taking	eagles	up‐front,	and	formally	evaluating	possible	
alternatives	in	(ideally)	siting,	configuration,	and	operation	of	wind	projects.		The	Service’s	ability	
to	influence	siting	and	configuration	factors	depends	on	the	stage	of	development	of	the	project	at	
the	time	the	project	developer	or	operator	comes	to	us.	
	
The	Service	recommends	that	project	developers	or	operators	develop	an	ECP	following	the	five‐
staged	approach	described	earlier.		During	Stages	1	through	4,	projects	or	alternatives	should	be	
placed	in	one	of	the	three	risk	categories,	with	increasing	certainty	by	Stage	4.		The	ECP	should	
provide	detailed	information	on	siting,	configuration,	and	operational	alternatives	that	avoid	and	
minimize	eagle	take	to	the	point	any	remaining	take	is	unavoidable	and,	if	required,	mitigates	that	
remaining	take	to	meet	the	statutory	preservation	standard.	The	Service	will	use	the	ECP	and	other	
application	materials	to	either	develop	an	eagle	take	permit	for	the	project,	or	to	determine	that	the	
project	cannot	be	permitted	because	risk	to	eagles	is	too	high	to	meet	the	regulatory	permit	
requirements.	
	
For	permitted	projects,	the	Service	will	use	the	80%	upper	confidence	limit	or	similar	risk‐averse	
estimate	(e.g.,	the	upper	limit	of	the	80%	credible	interval	is	used	in	the	Service’s	predictive	model	
described	in	Appendix	D)	of	the	mean	annual	predicted	unavoidable	eagle	take	to	determine	likely	
population‐level	effects	of	the	permit	and	compensatory	mitigation	levels,	if	required.		For	
predicted	recurring	eagle	take	that	is	in	excess	of	calculated	eagle	management	unit	take	
thresholds,	the	Service	will	either	(a)	approve	a	compensatory	mitigation	proposal	from	the	project	
developer	or	operator;	or	(b)	accept,	if	sufficient,	a	commitment	of	funds	to	an	appropriate	
independent	third	party	that	is	formally	obligated	(via	contract	or	other	agreement	with	the	project	
developer	or	operator)	to	perform	the	approved	mitigation	work.		Under	either	(a)	or	(b),	the	
compensatory	mitigation	cost	and	actions	will	be	calibrated	so	as	to	offset	the	predicted	
unavoidable	take,	such	that	we	bring	the	individual	permit’s	(and	cumulatively	over	all	such	
permits’)	predicted	mortality	effect	to	a	no‐net‐loss	standard.		Compensatory	mitigation	will	
initially	be	based	on	the	upper	80%	confidence	limit	of	the	predicted	mean	annual	fatality	rate	(or	
similar	risk‐averse	estimate)	over	a	five	year	period,	and	it	will	be	adjusted	for	future	years	based	
on	the	observed	fatality	rate	over	the	initial	period	of	intensive	post‐construction	monitoring	(no	
less	than	2	years).		Compensatory	mitigation,	as	well	as	other	forms	of	mitigation	aimed	at	reducing	
other	detrimental	effects	of	permits	on	eagles,	may	also	be	necessary	in	other	situations	where	
predicted	effects	to	eagle	populations	are	substantial	and	not	consistent	with	stable	or	increasing	
breeding	populations	of	eagles.	
	
Post‐construction	monitoring	may	be	required	as	a	condition	of	an	eagle	programmatic	take	permit	
and	will	be	required	for	wind‐energy	projects	that	may	potentially	take	eagles.		This	monitoring	

30
should	be	systematic	and	standardized	to	be	suitable	for	use	in	a	formal	adaptive	management	
framework	to	evaluate	and	improve	the	predictive	accuracy	of	our	models.		In	addition,	the	
information	will	be	used	by	the	Service	and	the	project	developer	or	operator	to	determine	if,	after	
no	more	than	five	years	of	post‐construction	monitoring,	the	80%	upper	confidence	limit	on	the	
predicted	mean	number	of	annual	fatalities	adequately	captured	the	observed	estimated	mean	
number	of	fatalities	annually.		If	the	observed	and	predicted	estimates	of	annual	fatalities	are	
different,	either	additional	compensatory	mitigation	will	be	required	retroactively	to	offset	higher‐
than‐predicted	levels	of	take	(assuming	the	actual	number	of	eagles	taken	was	greater	than	the	
number	actually	compensated	for),	or	the	permittee	will	receive	a	credit	for	the	excess	
compensation	(the	difference	between	the	actual	mean	and	the	number	compensated	for)	that	can	
be	applied	to	other	take	(either	by	the	permittee	or	other	permitted	individuals	at	his/her	
discretion)	within	the	same	eagle	management	unit	at	any	time	in	the	future.	
	
At	no	more	than	five‐years	from	the	date	a	permit	is	issued,	the	permittee	will	compile	and	the	
Service	and	the	permittee	will	review	fatality	information	for	the	project	to	determine	if	
experimental	ACPs	should	be	implemented	to	potentially	reduce	eagle	mortalities	based	on	the	
observed,	specific	situation	at	each	site.		As	discussed	previously,	at	the	time	of	permit	issuance	the	
Service	and	the	project	developer	or	operator	will	agree	to	an	upper	limit	on	the	cost	of	such	future	
experimental	ACPs,	which	will	only	be	implemented	if	warranted	by	eagle	disturbance	or	mortality	
data.		If	these	experimental	ACPs	are	likely	to	reduce	mortalities	at	the	project	in	the	future,	the	
amount	of	future	compensatory	mitigation	will	be	decreased	accordingly	(e.g.	if	ACPs	are	predicted	
to	reduce	the	fatality	rate	from	three	to	two	eagles	annually,	compensatory	mitigation	would	only	
be	required	to	offset	the	future	predicted	take	of	two	eagles	per	year).		In	such	cases,	additional	
post‐implementation	monitoring	should	be	conducted	to	determine	the	effectiveness	of	the	
experimental	ACPs.		In	cases	where	observed	fatalities	exceed	predicted	to	the	degree	category	1	
fatality‐rate	criteria	are	confirmed	to	have	been	met	or	exceeded	by	a	permitted	project,	and	for	
whatever	reason	experimental	ACPs	or	additional	conservation	measures	cannot	be	implemented	
to	reduce	fatalities	to	category	2	levels	or	below,	the	Service	may	have	to	rescind	the	permit	for	that	
project	to	remain	in	compliance	with	regulatory	criteria.	
	
Programmatic	eagle	take	permits	will	be	conditioned	to	require	access	to	the	areas	where	take	is	
possible	and	where	compensatory	mitigation	is	being	implemented	by	Service	personnel,	or	other	
qualified	persons	designated	by	the	Service,	within	reasonable	hours	and	with	reasonable	notice	
from	the	Service,	for	purposes	of	monitoring	the	site(s).		The	regulations	provide,	and	a	condition	of	
any	permit	issued	will	require,	that	the	Service	may	conduct	such	monitoring	while	the	permit	is	
valid,	and	for	up	to	three	years	after	it	expires	(50	CFR	22.26(c)(4)).		In	general,	verifying	
compliance	with	permit	conditions	is	a	secondary	purpose	of	site	visits;	the	primary	purpose	is	to	
monitor	the	effects	and	effectiveness	of	the	permitted	action	and	mitigation	measures.		This	may	be	
done	if	a	project	developer	or	operator	is	unable	to	observe	or	report	to	the	Service	the	information	
required	by	the	annual	report—or	it	may	serve	as	a	“quality	control”	measure	the	Service	can	use	to	
verify	the	accuracy	of	reported	information	and/or	adjust	monitoring	and	reporting	requirements	
to	provide	better	information	for	purposes	of	adaptive	management.	

	
1. Contents of the Eagle Conservation Plan
This	section	provides	a	recommended	outline	for	an	ECP,	with	a	short	description	of	what	should	
be	contained	in	each	section.		See	previous	sections	and	referenced	appendices	for	details	on	the	
stages	and	categories.	

	

31

a. Stage 1
Data	from	Stage	1	should	be	presented	and	summarized	in	this	section	of	the	ECP.		The	
project	developer	or	operator	should	work	with	the	Service	to	place	potential	wind–facility	
site	in	a	category	based	on	the	Stage	1	information.		For	detailed	recommendations	on	the	
Stage	1	process,	see	Appendix	B.	

b. Stage 2
Data	from	Stage	2	should	be	presented	and	summarized	in	this	section	of	the	ECP.		For	
detailed	recommendations	on	the	Stage	2	methods	and	metrics,	see	Appendix	C.		The	risk	
categorization	should	be	re‐assessed	in	this	section,	taking	into	account	Stage	2	results.	

c. Stage 3
In	this	section	of	the	ECP,	project	developers	or	operators	should	work	in	coordination	with	
the	Service	to	calculate	a	prediction	of	the	annual	eagle	fatality	rate	and	confidence	interval	
for	the	project	using	data	generated	from	the	Stage	2	assessment.		The	initial	estimate	of	the	
fatality	rate	should	not	take	into	account	possible	conservation	measures	and	ACPs;	these	
will	be	factored	in	as	part	of	Stage	4.		For	detailed	recommendations	on	Stage	3	methods	
and	metrics,	see	Appendix	D.		The	risk	categorization	should	be	re‐assessed	in	this	section,	
taking	into	account	Stage	3	results.	

d. Stage 4
This	section	of	the	ECP	should	describe	how	proposed	conservation	measures	and	ACPs	
should	reduce	the	fatality	rate	generated	in	stage	3,	and	what	compensatory	mitigation	
measures	will	be	employed	to	offset	unavoidable	take,	if	required.		This	section	facilitates	
demonstrating	how	conservation	measures	and	ACPs	have	reduced	the	raw	predicted	
fatality	rate	to	the	unavoidable	standard.		For	detailed	recommendations	on	considerations	
for	the	development	of	conservation	measures	and	ACPs	see	Appendix	E.		The	risk	
categorization	should	be	re‐assessed	in	this	section,	taking	into	account	Stage	4	results.	This	
should	be	the	final	pre‐construction	risk	categorization	for	the	proposed	project.		This	
section	should	also	fully	describe	the	proposed	compensatory	mitigation	approach	(if	
required).		For	detailed	recommendations	regarding	compensatory	mitigation,	see	
Appendix	G.	

e. Stage 5 – Post-construction Monitoring
In	this	section	of	the	ECP,	the	project	developer	or	operator	should	describe	the	proposed	
post‐construction	survey	methodology	for	the	project.		Detailed	recommendations	for	post‐
construction	monitoring	are	in	Appendix	H.		The	Stage	5	post‐construction	monitoring	plan	
is	the	final	section	of	the	ECP.	

	
	
	

	

32

INTERACTION WITH THE SERVICE
	

As	noted	throughout	this	ECPG,	frequent	and	thorough	coordination	between	project	developers	or	
operators	and	Service	and	other	jurisdictional‐agency	employees	is	crucial	to	the	development	of	
an	effective	and	successful	ECP.		Close	coordination	will	also	be	necessary	in	the	refinement	of	the	
modeling	process	used	to	predict	fatalities,	as	well	as	in	post‐construction	monitoring	to	evaluate	
those	models.		We	anticipate	the	ECPG	and	the	recommended	methods	and	metrics	will	evolve	
rapidly	as	the	Service	and	project	developers	or	operators	learn	together.		The	Service	has	created	a	
cross‐program,	cross‐regional	team	of	biologists	who	will	work	jointly	on	eagle‐programmatic‐take	
permit	applications	to	help	ensure	consistency	in	administration	and	application	of	the	Eagle	
Permit	Rule.		This	close	coordination	and	interaction	is	especially	important	as	the	Service	
processes	the	first	few	programmatic	eagle	take	permit	applications.	
	
The	Service	will	continue	to	refine	this	ECPG	with	input	from	all	stakeholders	with	the	objective	of	
maintaining	stable	or	increasing	breeding	populations	of	both	bald	and	golden	eagles	while	
simultaneously	developing	science‐based	eagle‐take	regulations	and	procedures	that	are	
appropriate	to	the	risk	associated	with	each	wind	energy	project.		As	the	ECPG	evolves,	the	Service	
will	not	expect	project	developers	or	operators	to	retroactively	redo	analyses	or	surveys	using	the	
new	approaches.		The	adaptive	approach	to	the	ECPG	should	not	deter	project	developers	or	
operators	from	using	it	immediately.	

	
	
	

	

33

INFORMATION COLLECTION
	

The	Bald	and	Golden	Eagle	Protection	Act	authorizes	us	to	collect	information	in	order	to	issue	
permits	for	eagle	take.		The	Eagle	Conservation	Plan	Guidance	defines	and	clarifies	the	information	
required	for	a	permit	application	(FWS	Form	3‐200‐71)	and	the	associated	annual	report	(FWS	
Form	3‐202‐15).	We	use	the	collected	information	to	evaluate	whether	the	take	is	compatible	with	
the	preservation	of	the	eagle;	to	determine	if	take	is	likely	and	how	it	can	be	avoided	and	
minimized;	to	determine	if	the	applicant	will	take	reasonable	measures	to	minimize	the	take;	and	to	
assess	how	the	activity	actually	affects	eagles	in	order	to	adjust	mitigation	measures	for	that	project	
and	for	future	permits.	
	
We	may	not	conduct	or	sponsor,	nor	are	you	required	to	respond,	to	a	collection	of	information	
unless	it	displays	a	currently	valid	Office	of	Management	and	Budget	control	number.		The	burden	
for	the	information	collection	associated	with	eagle	permits	and	reports	is	approved	under	OMB	
Control	No.	1018‐0022	(Federal	Fish	and	Wildlife	Permit	Applications	and	Reports‐‐Migratory	
Birds	and	Eagles)	and	OMB	Control	No.	1018‐0148	(Land‐Based	Wind	Energy	Guidelines).	

	
	
	
	

	
	

34

GLOSSARY
	
Active	nest	–	see	occupied	nest.	
Adaptive	resource	management	–	an	iterative	decision	process	that	promotes	flexible	decision‐
making	that	can	be	adjusted	in	the	face	of	uncertainties	as	outcomes	from	management	actions	
and	other	events	become	better	understood.	
Advanced	conservation	practices	(ACP)	–	means	scientifically	supportable	measures	that	are	
approved	by	the	Service	and	represent	the	best	available	techniques	to	reduce	eagle	
disturbance	and	ongoing	mortalities	to	a	level	where	remaining	take	is	unavoidable.		ACPs	are	a	
special	subset	of	conservation	measures	that	must	be	implemented	where	they	are	applicable.	
Adult	–	an	eagle	five	or	more	years	of	age.	
Alternate	nests	–	additional	sites	within	a	nesting	territory	that	are	available	to	be	used.	
Avoidance	and	minimization	measures	–	conservation	actions	targeted	to	remove	or	reduce	
specific	risk	factors	(e.g.,	avoiding	important	eagle	use	areas	and	migration	concentration	sites,	
placing	turbines	away	from	ridgelines).		A	subset	of	conservation	measures.	
Benchmark	–	an	eagle	harvest	rate	at	the	local‐area	population	scale	that	should	trigger	
heightened	scrutiny.	
Breeding	territory	–	equivalent	to	eagle	territory.	
Calculated	take	thresholds	–	annual	allowable	eagle	take	limits	established	in	USFWS	(2009b).	
Collision	probability	(risk)	–	the	probability	that	an	eagle	will	collide	with	a	turbine	given	
exposure.	
Compensatory	mitigation	–	replacement	of	project‐induced	losses	to	fish	and	wildlife	resources.	
Substitution	or	offsetting	of	fish	and	wildlife	resource	losses	with	resources	considered	to	be	of	
equivalent	biological	value.		In	the	case	of	an	the	ECPG,	an	action	in	the	eagle	permitting	process	
that	offsets	the	predicted	take	of	eagles	if	ACPs	and	other	conservation	measures	do	not	
completely	remove	the	potential	for	take,	and	projected	take	exceeds	calculated	take	thresholds	
for	the	species	or	the	eagle	management	unit	affected	(or	in	some	cases,	under	other	
circumstances	as	described	in	USFWS	2009a).	
Conservation	measures	–	actions	that	avoid	(this	is	best	achieved	at	the	siting	stage),	minimize,	
rectify,	reduce,	eliminate,	or	mitigate	an	effect	over	time.		ACPs	are	conservation	measures	that	
have	scientific	support	and	which	must	be	implemented	where	they	are	applicable.	
Discount	rate	–	the	interest	rate	used	in	calculating	the	present	value	of	expected	yearly	benefits	
and	costs.	
Disturb	‐	means	to	agitate	or	bother	a	bald	or	golden	eagle	to	a	degree	that	causes,	or	is	likely	to	
cause,	based	on	the	best	scientific	information	available,	(1)	injury	to	an	eagle,	(2)	a	decrease	in	
its	productivity,	by	substantially	interfering	with	normal	breeding,	feeding,	or	sheltering	
behavior,	or	(3)	nest	abandonment,	by	substantially	interfering	with	normal	breeding,	feeding,	
or	sheltering	behavior.	
Eagle	Conservation	Plans	(ECP)	–	a	document	produced	by	the	project	developer	or	operator	in	
coordination	with	the	Service	that	supports	issuance	of	an	eagle	take	permit	under	50	CFR	
22.26	and	potentially	22.27	(or	demonstrates	that	such	a	permit	is	unnecessary).	
Eagle	Management	Unit	–	regional	eagle	populations	defined	in	the	FEA	(USFWS	2009b).		For	
golden	eagles,	eagle	management	units	follow	Bird	Conservation	Regions	(Figure	2),	whereas	
bald	eagle	management	units	largely	follow	Service	regional	boundaries	(Figure	3).	
Eagle	exposure	rate	–	Eagle‐minutes	flying	within	the	project	footprint	(in	proximity	to	turbine	
hazards)	per	hour	(hr)	per	kilometer2	(km2).	
Eagle	nest	(or	nest)	–	any	readily	identifiable	structure	built,	maintained	or	used	by	bald	eagles	or	
golden	eagles	for	the	purposes	of	reproduction	(as	defined	in	50	CFR	22.3).	

35
Eagle	territory	–	an	area	that	contains,	or	historically	contained,	one	or	more	nests	within	the	
home	range	of	a	mated	pair	of	eagles	(from	the	regulatory	definition	of	“territory”	at	50	CFR	
22.3).		“Historical”	is	defined	here	as	at	least	the	previous	5	years.	
Experimental	ACPs	–	prospective	conservation	measures	identified	at	the	start	of	a	programmatic	
eagle	take	permit	that	are	not	implemented	immediately,	but	are	deferred	pending	the	results	
of	post‐construction	monitoring.		If	such	monitoring	indicates	the	measures	might	reduce	
observed	eagle	fatalities,	they	should	be	implemented	and	monitored	for	a	sufficient	period	of	
time	to	determine	their	effectiveness.	
Fatality	monitoring	–	searching	for	eagle	carcasses	beneath	turbines	and	other	facilities	to	
estimate	the	number	of	fatalities.	
Fatality	rate	–	(1)	in	fatality	prediction	models,	the	fatality	rate	is	the	number	of	eagle	fatalities	per	
hr		per	km	2	;	(2)	elsewhere	in	the	ECPG	it	is	the	number	of	eagles	taken	or	predicted	to	be	taken	
per	year.	
Floater	(floating	adult)	–	an	adult	eagle	that	has	not	settled	on	a	breeding	territory.	
Hazardous	area	–	Rotor‐swept	area	around	a	turbine	or	proposed	turbine	(km2).	
Home	range	–	the	area	traveled	by	and	eagle	in	its	normal	activities	of	food	gathering,	mating,	and	
caring	for	young.		Breeding	home	range	is	the	home	range	during	the	breeding	season,	and	the	
non‐breeding	home	range	is	the	home	range	outside	the	breeding	season.	
Important	eagle‐use	area	–	an	eagle	nest,	foraging	area,	or	communal	roost	site	that	eagles	rely	on	
for	breeding,	sheltering,	or	feeding,	and	the	landscape	features	surrounding	such	nest,	foraging	
area,	or	roost	site	that	are	essential	for	the	continued	viability	of	the	site	for	breeding,	feeding,	
or	sheltering	eagles	(as	defined	at	50	CFR	22.26).	
Inactive	nest	–	a	bald	eagle	or	golden	eagle	nest	that	is	not	currently	being	used	by	eagles	as	
determined	by	the	continuing	absence	of	any	adult,	egg,	or	dependent	young	at	the	nest	for	at	
least	10	consecutive	days	immediately	prior	to,	and	including,	at	present.	An	inactive	nest	may	
become	active	again	and	remains	protected	under	the	Eagle	Act.	
Inventory	–	systematic	observations	of	the	numbers,	locations,	and	distribution	of	eagles	and	eagle	
resources	such	as	suitable	habitat	and	prey	in	an	area.	
Jurisdictional	agency	–	a	government	agency	with	jurisdictional	authority	to	regulate	an	activity	
(e.g.,	a	state	or	tribal	fish	and	wildlife	agency,	a	state	or	federal	natural	resource	agency,	etc.).	
Juvenile	–	an	eagle	less	than	one	year	old.	
Kiting	–	stationary	or	near‐stationary	hovering	by	a	raptor,	usually	while	searching	for	prey.	
Local‐area	population	–	is	as	defined	in	USFWS	(2009b),	and	refers	to	the	eagle	population	within	
a	distance	from	the	project	footprint	equal	to	the	species	median	natal‐dispersal	distance	(43	
miles	for	bald	eagles	and	140	miles	for	golden	eagles).	
Mean	inter‐nest	distance	–	the	mean	nearest‐neighbor	distance	between	simultaneously	occupied	
eagle	nests.	
Meteorological	towers	(met	towers)	–	towers	erected	to	measure	meteorological	events	such	as	
wind	speed,	direction,	air	temperature,	etc.	
Migration	concentration	sites	–	places	where	geographic	features	(e.g.,	north‐south	oriented	
ridgelines,	peninsulas)	funnel	migrating	eagles,	resulting	in	concentrated	use	during	migration	
periods.	
Migration	corridors	–	the	routes	or	areas	where	eagles	may	concentrate	during	migration	(e.g.,	
funneling	areas	along	ridgetops,	at	tips	of	peninsulas)	as	a	result	of	the	interplay	between	
weather	variables	and	topography.	
Migration	counts	–	standardized	counts	that	can	be	used	to	determine	relative	numbers	of	diurnal	
raptors	passing	over	an	established	point	during	fall	or	spring	migration.	
Mitigation	–	avoidance,	minimization,	rectification,	reduction	over	time,	and	compensation	for	
negative	impacts	to	bald	eagles	and	golden	eagles	from	the	permitted	actions.		In	the	ECPG,	we	

36
use	the	term	compensatory	mitigation	to	describe	the	subset	of	mitigation	actions	designed	to	
offset	take	to	achieve	the	no‐net‐loss	standard.	
Monitoring	–	(1)	a	process	of	project	oversight	such	as	checking	to	see	if	activities	were	conducted	
as	agreed	or	required;	(2)	making	measurements	of	uncontrolled	events	at	one	or	more	points	
in	space	or	time	with	space	and	time	being	the	only	experimental	variable	or	treatment;	(3)		
making	measurements	and	evaluations	through	time	that	are	done	for	a	specific	purpose,	such	
as		to	check	status	and/or	trends	or	the	progress	towards	a	management	objective.	
No‐net‐loss	–	no	net	change	in	the	overall	eagle	population	mortality	or	natality	rate	after	issuance	
of	a	permit	that	authorizes	take,	because	compensatory	mitigation	reduces	another	form	of	
mortality,	or	increases	natality,	by	a	comparable	amount.	
Occupied	nest	–	a	nest	used	for	breeding	in	the	current	year	by	a	pair	of	eagles.		Presence	of	an	
adult,	eggs,	or	young,	freshly	molted	feathers	or	plucked	down,	or	current	year’s	mutes	
(whitewash)	suggest	site	occupancy.		In	years	when	food	resources	are	scarce,	it	is	not	
uncommon	for	a	pair	of	eagles	to	occupy	a	nest	yet	never	lay	eggs;	such	nests	are	considered	
occupied.	
Occupied	territory	–	an	area	that	encompasses	a	nest	or	nests	or	potential	nest	sites	and	is	
defended	by	a	mated	pair	of	eagles.	
Operational	adjustments	–	modifications	made	to	an	existing	wind	project	that	changes	how	that	
project	operates	(e.g.,	increasing	turbine	cut	in	speeds,	implementing	curtailment	of	turbines	
during	periods	of	high	eagle	use).	
Posterior	distribution	(Bayesian)	–	a	distribution	that	quantifies	the	uncertainty	in	the	model	
parameters	after	incorporating	the	observed	data.		The	distributions	are	usually	summarized	by	
intervals	around	the	median.	
Present	value	–	within	the	context	of	a	Resource	Equivalency	Analysis	(REA),	refers	to	the	value	of	
debits	and	credits	based	on	an	assumed	annual	discount	rate	(3%).		This	term	is	commonly	
used	in	economics	and	implies	that	resources	lost	or	gained	in	the	future	are	of	less	value	to	us	
today.	
Prior	distribution	(Bayesian)	–	a	distribution	that	quantifies	the	uncertainty	in	the	model	
parameters	from	previous	data	or	past	knowledge.		A	non‐informative	prior	can	be	used	to	
imply	that	little	or	nothing	is	known	about	the	parameters.	
Programmatic	take	–	take	that	is	recurring,	is	not	caused	solely	by	indirect	effects,	and	that	occurs	
over	the	long	term	or	in	a	location	or	locations	that	cannot	be	specifically	identified	(as	defined	
in	50	CFR	22.3).	
Project	area	–	the	area	that	includes	the	project	footprint	as	well	as	contiguous	land	that	shares	
relevant	characteristics.		For	eagle‐take	considerations,	the	Service	recommends	the	project	
area	be	either	project	footprint	and	a	surrounding	perimeter	equal	to	the	mean	species‐specific	
inter‐nest	distance	for	eagles	locally,	or	the	project	footprint	and	a	10‐mile	perimeter.	
Project‐area	inter‐nest	distance	–	the	mean	nearest‐neighbor	distance	between	simultaneously	
occupied	eagle	nests	of	a	species	(including	occupied	nests	in	years	where	no	eggs	are	laid).		We	
recommend	calculating	this	metric	from	the	nesting	territory	survey	in	Stage	2,	using	all	nesting	
territories	within	the	project	area,	ideally	over	multiple	years.	
Project‐area	nesting	population	–	number	of	pairs	of	eagles	nesting	within	the	project	area.	
Project‐area	eagle	population	–	the	population	of	eagles,	considering	breeding,	migrating,	and	
wintering	eagles,	within	the	project	area.	
Project	footprint	–	the	minimum‐convex	polygon	that	encompasses	the	wind‐project	area	
inclusive	of	the	hazardous	area	around	all	turbines	and	any	associated	utility	infrastructure,	
roads,	etc.	
Project	developer	or	operator	–	any	developer	or	operator	that	proposes	to	construct	a	wind	
project.	

37
Productivity	─	the	number	of	juveniles	 ledged	from	an	occupied	nest,	often	reported	as	a	mean	
over	a	sample	of	nests.	
Renewable	energy	–	energy	produced	by	solar,	wind,	geothermal	or	any	other	methods	that	do	not	
require	fossil	fuels.	
Resource	Equivalency	Analysis	(REA)	–	in	the	context	of	the	ECPG,	a	methodology	used	to	
compare	the	injury	to	or	loss	of	eagles	caused	by	wind	facilities	(debit)	to	the	benefits	from	
projects	designed	to	improve	eagle	survival	or	increase	productivity	(credits).		Compensation	is	
evaluated	in	terms	of	eagles	and	their	associated	services	instead	of	by	monetary	valuation	
methods.	
Retrofit	–	any	activity	that	results	in	the	modification	of	an	existing	power	line	structure	to	make	it	
bird	safe.	
Risk‐averse	–	a	conservative	estimate	in	the	face	of	considerable	uncertainty.		For	example,	the	
Service	typically	will	use	the	upper	80%	credible	interval	of	the	median	estimated	number	of	
annual	eagle	fatalities	for	permit	decisions	in	an	effort	to	avoid	underestimating	fatality	rates	at	
wind	projects.	
Risk	validation	–	as	part	of	Stage	5	assessment,	where	post‐construction	surveys	are	conducted	to	
generate	empirical	data	for	comparison	with	the	pre‐construction	risk	assessment	predictions	
to	validate	if	the	initial	assumptions	were	correct.	
Roosting	–	activity	where	eagles	seek	cover,	usually	during	night	or	periods	of	severe	weather	(e.g.,	
cold,	wind,	snow).		Roosts	are	usually	found	in	protected	areas,	typically	tree	rows	or	trees	
along	a	river	corridor.	
Seasonal	concentration	areas	–	areas	used	by	concentrations	of	eagles	seasonally,	usually	
proximate	to	a	rich	prey	source.	
Site	categorization	–	a	standardized	approach	to	categorize	the	likelihood	that	a	site	or	
operational	alternative	will	meet	standards	in	50	CFR	22.26	for	issuance	of	a	programmatic	
eagle	take	permit.	
Stopover	sites	–	areas	temporarily	used	by	eagles	to	rest,	seek	forage,	or	cover	on	their	migration	
routes.	
Subadult	–	an	eagle	between	1	and	4	years	old,	typically	not	of	reproductive	age.	
Survey	–combined	inventory	and	monitoring.	
Take	threshold	–	an	upper	limit	on	the	annual	eagle	harvest	rate	for	each	species‐specific	eagle	
management	unit.		Thresholds	were	set	in	the	Final	Environmental	Assessment	on	the	Eagle	
Permit	Rule	(USFWS	2009b).	
Territory	–	area	that	contains,	or	historically	contained,	one	or	more	nests	within	the	home	range	
of	a	mated	pair	of	eagles	(from	50	CFR	22.3).	
Unoccupied	nest	–	those	nests	not	selected	by	raptors	for	use	in	the	current	nesting	season.		See	
also	inactive	nest.	
U.S.	Fish	and	Wildlife	Service	Draft	Land‐based	Wind	Energy	Guidelines	(WEG)	–	a	document	
that	describes	a	multi‐tiered	process	to	site,	construct,	operate	and	monitor	wind	facilities	in	
ways	that	avoid,	minimize,	and	mitigate	impacts	to	wildlife.	
Wind	facilities	–	developments	for	the	generation	of	electricity	from	wind	turbines.	
Wind	project	–	developments	for	the	generation	of	electricity	from	wind	turbines.	
Wind	turbine	–	a	machine	for	converting	the	kinetic	energy	in	wind	into	mechanical	energy,	which	
is	then	converted	to	electricity.	

	
	

	

38

	

	
Figure 2. Map of golden eagle management units, from USFWS (2009b).

	
	

	

39

Figure 3. Map of bald eagle management units, from USFWS (2009b).	

	

	

40

LITERATURE CITED
	
Anthony,	R.	G.		2001.		Low	productivity	of	bald	eagles	on	Prince	of	Wales	Island,	Southeast	Alaska.	
Raptor	Research	35:1‐8.	
Arnett,	E.	B.		2006.		A	preliminary	evaluation	on	the	use	of	dogs	to	recover	bat	fatalities	at	wind	
energy	facilities.	Wildlife	Society	Bulletin	34(5):1440–1445.	
Arnett,	E.	B.,	D.	B.	Inkley,	D.	H.	Johnson,	R.	P.	Larkin,	S.	Manes,	A.	M.	Manville,	R.	Mason,	M.	Morrison,	
M.	D.	Strickland,	and	R.	Thresher.		2007.		Impacts	of	wind	energy	facilities	on	wildlife	and	
wildlife	habitat.	Technical	Review	07‐2,	The	Wildlife	Society,	Bethesda,	Maryland,	USA.	
Barclay,	R.	M.	R.,	E.	F.	Baerwald,	and	J.	C.	Gruver.		2007.		Variation	in	bat	and	bird	fatalities	at	wind	
energy	facilities:	assessing	the	effects	of	rotor	size	and	tower	height.	Canadian	Journal	of	
Zoology	85:	381–387.	
Barrios,	L.	and	A.	Rodriguez.		2004.		Behavioural	and	environmental	correlates	of	soaring‐bird	
mortality	at	on‐shore	wind	turbines.	Journal	of	Applied	Ecology	41:72‐81.	
Buckland,	S.	T.,	D.	R.	Anderson,	K.	P.	Burnham,	J.	L.	Laake,	D.	L.	Borchers,	and	L.	Thomas.		2001.		
Introduction	to	distance	sampling.		Oxford	University	Press,	New	York,	New	York,	USA.	
Buehler,	D.	A.		2000.		Bald	eagle	(Haliaeetus	leucocephalus).	The	Birds	of	North	America	no.	506	(A.	
Poole,	ed.).		The	Birds	of	North	America	Online,	Cornell	Lab	of	Ornithology,	Ithaca,	New	York,	
USA.	http://bna.birds.cornell.edu/bna/species/506.	
Chamberlain,	D.		E.,	M.	R.	Rehfisch,	A.	D.	Fox,	M.	Desholm,	and	S.	J.	Anthony.		2006.		The	effect	of	
avoidance	rates	on	bird	mortality	predictions	made	by	wind	turbine	collision	risk	models.	Ibis	
148:198‐202.	
Cole,	S.		2009.		How	much	is	enough?		Determining	adequate	levels	of	environmental	compensation	
for	wind	power	impacts	using	equivalency	analysis.	In	European	Offshore	Wind	Conference	
2009,	14‐16	September	2009,	Stockholm,	Sweden.	
Collopy,	M.	W.,	and	T.	C.	Edwards,	Jr.		1989.		Territory	size,	activity	budget,	and	role	of	undulating	
flight	in	nesting	golden	eagles.	Journal	of	Field	Ornithology	60:43‐51.	
Craig,	T.	H.,	and	E.	H.	Craig.		1984.		A	large	concentration	of	roosting	golden	eagles	in	southeastern	
Idaho.	Auk	101:610‐613.	
Craig,	T.	H.,	E.	H.	Craig,	and	L.	R.	Powers.		1984.		Recent	changes	in	eagle	and	buteo	densities	in	
southeastern	Idaho.	Murrelet	65:91‐93.	
De	Lucas,	M.,	G.	F.	E.	Janss,	D.	P.	Whitfield	and	M.	Ferrer.		2008.		Collision	fatality	of	raptors	in	wind	
farms	does	not	depend	on	raptor	abundance.	Journal	of	Applied	Ecology	45:1695–1703.	
Ferrer,	M.	de	Lucas,	G.	F.	E.	Janss,	E.	Casado,	A.	R.	Munõz,	M.	J.	Bechard,	and	C.	P.	Calabuig.		2011.		
Weak	relationship	between	risk	assessment	studies	and	recorded	mortality	in	wind	farms.	
Journal	of	Applied	Ecology	doi:	10.1111/j.1365‐2664.2011.02054.x.	
Fuller,	M.	R.,	J.	J.	Millspaugh,	K.	Church,	and	R.	Kenward.		2005,		Wildlife	radiotelemetry.	Pages	377‐
417	in	C.	E.	Braun	(ed.),	Techniques	for	wildlife	investigations	and	management,	6th	edition.	The	
Wildlife	Society,	Bethesda,	Maryland,	USA.	
Garrett,	M.	G.,	J.	W.	Watson,	and	R.	G.	Anthony.		1993.		Bald	eagle	home	range	and	habitat	use	in	the	
Columbia	River	estuary.	Journal	of	Wildlife	Management	57:19‐27.	
Gregory,	M.	J.	P,	A.	G.	Gordon,	and	R.	Moss.		2002.		Impact	of	nest‐trapping	and	radio‐tagging	on	
breeding	golden	eagles	Aquila	chrysaetos	in	Argyll,	Scotland.	Ibis	145:113‐119.	
Harmata,	A.	R.		1982.		What	is	the	function	of	undulating	flight	display	in	golden	eagles?	Raptor	
Research	16:103‐109.	
Harmata,	A.R.		2002.		Vernal	migration	of	bald	eagles	from	a	southern	Colorado	wintering	area.	
Journal	of	Raptor	Research	36:256‐264.	

41
Hodges,	J.	I.	and	F.	C.	Robards.		1982.		Observations	of	3,850	bald	eagle	nests	in	southeast	Alaska.	
Pages	37‐46	in	A	symposium	and	workshop	on	raptor	management	and	biology	in	Alaska	and	
western	Canada.	(W.	N.	Ladd	and	P.	F.	Schempf,	eds.)	U.S.	Fish	and	Wildlife	Service,	Anchorage,	
Alaska,	USA.	
Hoechlin,	D.	R.		1976.		Development	of	golden	eagles	in	southern	California.		Western	Birds	7:137‐
152.	
Hoover,	S.	L.,	and	M.	L.	Morrison.		2005.		Behavior	of	red‐tailed	hawks	in	a	wind	turbine	
development.	Journal	of	Wildlife	Management	69:150‐159.	
Hunt,	W.	G.		1998.		Raptor	floaters	at	Moffat’s	equilibrium.	Oikos	82:191‐197.	
Hunt,	W.	G.,	R.	E.	Jackman,	T.	L.	Brown,	and	L.	Culp.		1999.		A	population	study	of	golden	eagles	in	
the	Altamont	Pass	Wind	Resource	Area:	population	trend	analysis	1994‐1997.		Report	to	
National	Renewable	Energy	Laboratory,	Subcontracts	XAT‐5‐15174‐01,	XAT‐6‐16459‐01.		
Predatory	Bird	Research	Group,	University	of	California,	Santa	Cruz,	California,	USA.	
Hunt,	G.		2002.		Golden	eagles	in	a	perilous	landscape:	predicting	the	effects	of	mitigation	for	wind	
turbine	bladestrike	mortality.	California	Energy	Commission	Report	P500‐02‐043F.	
Sacramento,	California,	USA.	
Huso,	M.	M.	P.		2010.		An	estimator	of	wildlife	fatality	from	observed	carcasses.	Environmetrics	DOI:	
10.1002/env.1052.	
Kenward,	R.	E.		2001.		A	manual	for	wildlife	tagging.	Academic	Press,	London,	UK.	
Kerlinger,	P.		1989.		Flight	strategies	of	migrating	hawks.	University	of	Chicago	Press,	Chicago,	
Illinois,	USA.	
Kochert,	M.	N.		1972.		Population	status	and	chemical	contamination	in	golden	eagles	in	
southwestern	Idaho.	M.S.	Thesis.	University	of	Idaho,	Moscow,	Idaho,	USA.	
Kochert,	M.	N.,	K.	Steenhof,	C.	L.	Mcintyre,	and	E.	H.	Craig.		2002.		Golden	eagle	(Aquila	chrysaetos).	
The	Birds	of	North	America	No.	684	(A.	Poole,	Ed.).	The	Birds	of	North	America	Online.	Cornell	
Lab	of	Ornithology,	Ithaca,	New	York,	USA.	http://bna.birds.cornell.edu/bna/species/684.	
Krone,	O.		2003.		Two	white‐tailed	sea	eagles	(Haliaeetus	albicilla)	collide	with	wind	generators	in	
northern	Germany.		Journal	of	Raptor	Research	37:174‐176.	
Kunz,	T.	H.,	E.	B.	Arnett,	B.	M.	Cooper,	W.	P.	Erickson,	R.	P.	Larkin,	T.	Mabee,	M.	L.	Morrison,	M.	D.	
Strickland,	and	J.	M.	Szewczak.		2007.		Assessing	impacts	of	wind‐energy	development	on	
nocturnally	active	birds	and	bats:	a	guidance	document.	Journal	of	Wildlife	Management	71:	
2449‐2486.	
Kuvlesky,	W.	P.	Jr.,	L.	A.	Brennan,	M.	L.	Morrison,	K.	K.	Boydston,	B.	M.	Ballard,	and	F.	C.	Bryant.		
2007.		Wind	energy	development	and	wildlife	conservation:	challenges	and	opportunities.	
Journal	of	Wildlife	Management	71:2487‐2498.	
Marzluff,	J.	M.,	M.	S.	Vekasy,	M.	N.	Kochert,	and	K.	Steenhof.		1997.		Productivity	of	golden	eagles	
wearing	backpack	radiotransmitters.	Journal	of	Raptor	Research	31:223‐227.	
McGrady,	M.	J.,	J.	R.	Grant,	I.	P.	Bainbridge,	and	D.	R.	A.	McLeod.		2002.		A	model	of	golden	eagle	
(Aquila	chrysaetos)	ranging	behavior.	Journal	of	Raptor	Research	36(1)	supplement:62‐69.	
McLeod,	D.	R.	A,	D.	P.	Whitfield,	and	M.	J.	McGrady.		2002.		Improving	prediction	of	golden	eagle	
(Aquila	chrysaetos)	ranging	in	western	Scotland	using	GIS	and	terrain	modeling.	Journal	of	
Raptor	Research	36(1)	Supplement:70‐77.	
Millsap,	B.	A.		1981.		Distributional	status	and	ecology	of	Falconiformes	in	west	central	Arizona,	
with	notes	on	ecology,	reproductive	success,	and	management.	U.	S.	Bureau	of	Land	
Management	Technical	Note	355.	
Mohr,	C.	O.	1947.	Table	of	equivalent	populations	of	North	American	small	mammals.	American	
Midland	Naturalist	37:223–249.	
Mojica,	E.	K.,	J.	M.	Meyers,	B.	A.	Millsap,	and	K.	T.	Haley.		2008.		Migration	of	sub‐adult	Florida	bald	
eagles.		Wilson	Journal	of	Ornithology	120:304‐310.	

42
Moorcroft,	P.	R.,	M.	A.	Lewis,	and	R.	L.	Crabtree.		1999.		Home	range	analysis	using	a	mechanistic	
home	range	model.	Ecology	80:1656‐1665.	
Newton,	I.		1979.		Population	ecology	of	raptors.		Buteo	Books,	Vermillion,	South	Dakota,	USA.	
Nygård,	T,	K.,	E.	L.	Bevanger,	Ø.	Dahl,	A.	Flagstad,	P.	L.	Follestad,	Hoel,	R.	May,	and	O.	Reitan.		2010.		
A	study	of	white‐tailed	eagle	Haliaeetus	albicilla	movements	and	mortality	at	a	wind	farm	in	
Norway.	in	D.	Senapathi	(ed.),	Climate	change	and	birds:	adaptation,	mitigation	and	impacts	on	
avian	populations.	British	Ornithologists’	Union	Proceedings,	Peterborough,	United	Kingdom.	
http://www.bou.org.uk/	bouproc‐net/ccb/nygard‐etal.pdf	(last	visited	9	October	2011).	
Palmer,	R.	S.		1988.		Golden	eagle	Aquila	chrysaetos.		Pages	180‐231	in	R.	S.	Palmer	(ed.),	Handbook	
of	North	American	birds,	volume	5,	diurnal	raptors	(part	2).	Yale	University	Press,	New	Haven,	
Connecticut,	USA.	
Platt,	J.	B.		1976.		Bald	eagles	wintering	in	a	Utah	desert.	American	Birds	30:783‐788.	
Phillips,	R.	L.	and	A.	E.	Beske.		1984.		Resolving	conflicts	between	energy	development	and	nesting	
golden	eagles.	Pages	214‐219	in	R.	D.	Comer,	J.	M.	Merino,	J.	W.	Monarch,	C.	Pustmueller,	M.	
Stalmaster,	R.	Stoecker,	J.	Todd,	and	W.	Wright	(eds.).	Proceedings	of	asymposium:	Issues	and	
technology	in	the	management	of	impacted	western	wildlife.	Steamboat	Springs,	Colorado,	
November	15‐17,	1982.	Thorne	Ecological	Institute,	Boulder,	Colorado,	USA.	
Restani,	M.,	A.	R.	Harmata,	and	E.	M.	Madden.		2000.		Numerical	and	functional	responses	of	migrant	
bald	eagles	exploiting	a	seasonally	concentrated	food	source.		Condor	102:561‐568.	
Seaman,	D.	E.,	and	R.	A.	Powell.	1996.	An	evaluation	of	the	accuracy	of	kernel	density	estimators	for	
home	range	analysis.	Ecology	77:2075–2085.	
Sherrod,	S.	K.,	C.	M.	White,	and	F.	S.	L.	Williamson.	1976.		Biology	of	the	bald	eagle	on	Amchitka	
Island,	Alaska.	Living	Bird	15:145‐182.	
Smallwood,	K.	S.	and	C.	G.	Thelander.		2004.		Developing	methods	to	reduce	bird	fatalities	in	the	
Altamont	Wind	Resource	Area.	Final	report	prepared	by	BioResource	Consultants	to	the	
California	Energy	Commission,	Public	Interest	Energy	Research‐Environmental	Area,	Contract	
No.	500‐01‐019.	
Smallwood,	K.	S.,	L.	Rugge,	M.	L.	Morrison.	2009.		Influence	of	behavior	on	bird	mortality	in	wind	
energy	developments.	Journal	of	Wildlife	Management	73:1082–1098.	
Soutullo,	A.,	V.	Urios,	M.	Ferrer,	and	S.	G.	Penarrubia.		2006.		Post‐fledging	behaviour	in	golden	
eagles	Aquila	chrysaetos:	onset	of	juvenile	dispersal	and	progressive	distancing	from	the	nest.	
Ibis	148:307‐312.	
Strickland,	M.D.,	E.B.	Arnett,	W.P.	Erickson,	D.H.	Johnson,	G.D.	Johnson,	M.L.,	Morrison,	J.A.	Shaffer,	
and	W.	Warren‐Hicks.	2011.	Comprehensive	Guide	to	Studying	Wind	Energy/Wildlife	
Interactions.	Prepared	for	the	National	Wind	Coordinating	Collaborative,	Washington,	D.C.,	USA.	
Thorstrom,	R.		2001.		Nest‐site	characteristics	and	breeding	density	of	two	sympatric	forest‐falcons	
in	Guatemala.	Ornitologia	Neotropical	12:337‐343.	
USFWS.		1983.		Northern	states	bald	eagle	recovery	plan.	U.S.	Fish	and	Wildlife	Service,	Division	of	
Migratory	Bird	Management,	Washington	D.C.,	USA.	
USFWS.		2007.		Protection	of	eagles;	definition	of	“Disturb.”	Federal	Register	72(107):31132‐31140.	
USFWS.		2009a.		Eagle	permits;	take	necessary	to	protect	interests	in	particular	localities.	Federal	
Register	74(175):46836‐46879.	
USFWS.		2009b.		Final	environmental	assessment.		Proposal	to	permit	take	provided	under	the	Bald	
and	Golden	Eagle	Protection	Act.	U.S.	Fish	and	Wildlife	Service,	Division	of	Migratory	Bird	
Management,	Washington	D.C.,	USA.	
Walker,	D.,	M.	McGrady,	A.	McCluskie,	M.	Madders,	and	D.	R.	A.	McLeod.		2005.		Resident	golden	
eagle	ranging	behaviour	before	and	after	construction	of	a	windfarm	in	Argyll.	Scottish	Birds	
25:24‐40.	
Watson,	J.	W.,	and	R.	W.	Davies.		2009.		Range	use	and	contaminants	of	golden	eagles	in	Washington.		
Progress	Report.	Washington	Department	of	Fish	and	Wildlife,	Olympia,	Washington.	

43
Whitfield,	D.	P.,	A.	H.	Fielding,	M.	J.	P.	Gregory,	A.	G.	Gordon,	D.	R.	A.McLeod,	and	P.	F.	Haworth.		
2007.		Complex	effects	of	habitat	loss	on	golden	eagles	Aquila	chrysaetos.	Ibis	149:26–36.	
Wichmann,	M.	C.,	F.	Jeltsch,	W.	R.	J.	Dean,	K.	A.	Moloney	and	C.	Wissel.		2003.		Implication	of	climate	
change	for	the	persistence	of	raptors	in	arid	savanna.	Oikos	102:186–202.	
Williams,	B.	K.,	R.	C.	Szaro,	and	C.	D.	Shapiro.		2007.		Adaptive	Management:	The	U.S.	Department	of	
the	Interior	Technical	Guide.	Adaptive	Management	Working	Group,	U.S.	Department	of	the	
Interior,	Washington,	DC,	USA.	
Worton,	B.	J.		1989.		Kernel	methods	for	estimating	the	utilization	distribution	in	home‐range	
studies.	Ecology	70:164–168.	
Yates,	R.	E.,	B.	R.	McClelland,	P.	T.	McClelland,	C.	H.	Key,	and	R.	E.	Bennetts.		2001.		The	influence	of	
weather	on	golden	eagle	migration	in	northwestern	Montana.	Journal	of	Raptor	Research	
35:81‐90.	

	
	
	

	

44

APPENDIX A: ADAPTIVE MANAGEMENT
	

Management	of	wind	facilities	to	minimize	eagle	take	through	decisions	about	siting,	design,	
operation,	and	compensatory	mitigation,	is	a	set	of	recurrent	decisions	made	in	the	face	of	
uncertainty.		The	Department	of	the	Interior	has	a	long	history	of	approaching	such	decisions	
through	a	process	of	adaptive	management	(Williams	et	al.	2007).		The	purpose	of	adaptive	
management	is	to	improve	long‐term	management	outcomes,	by	recognizing	where	key	
uncertainties	impede	decision	making,	seeking	to	reduce	those	uncertainties	over	time,	and	
applying	that	learning	to	subsequent	decisions	(Walters	1986).	
	
Adaptive	management	is	a	special	case	of	decision	analysis	applied	to	recurrent	decisions	(Lyons	et	
al.	2008).		Like	all	formal	decision	analysis,	it	begins	with	the	identification	of	fundamental	
objectives—the	long‐term	ends	sought	through	the	decision	(step	2,	Fig.	A‐1).		These	objectives	are	
the	primary	concern,	and	all	the	other	elements	are	designed	around	them.		With	these	objectives	in	
mind,	alternative	actions	are	considered,	and	the	consequences	of	these	alternatives	are	evaluated	
with	regard	to	how	well	they	might	achieve	the	objectives.		But	in	many	decisions,	there	is	critical	
uncertainty	that	impedes	the	decision	(step	6,	Fig.	A‐1),	that	is,	the	decision‐maker	is	missing	
knowledge	that	affects	which	alternative	might	be	best.		In	recurrent	decisions,	there	exists	the	
opportunity	to	reduce	that	uncertainty,	by	monitoring	the	outcomes	of	early	actions,	and	apply	that	
learning	to	later	actions.		It	is	valuable	to	note	that	learning	is	not	pursued	for	its	own	sake,	but	only	
insofar	as	it	helps	improve	long‐term	management	by	reducing	these	uncertainties.	
	
There	are	two	hallmarks	of	a	formal	interpretation	of	adaptive	management,	like	that	described	
above.		The	first	hallmark	is	the	a	priori	identification	of	the	critical	uncertainty.		In	this	way,	
adaptive	management	is	not	a	blind	search	for	some	unspecified	new	insights,	but	a	focused	effort	
to	reduce	the	uncertainty	that	stands	in	the	way	of	better	decision‐making.		The	second	hallmark	is	
that	the	means	of	adaptation	is	clear,	that	is,	the	way	in	which	new	information	will	be	applied	to	
subsequent	decisions	is	articulated.	
	
There	is,	however,	recognition	that	unanticipated	learning	does	occur	in	any	real	system,	and	this	
learning	can	sometimes	lead	to	valuable	insights.		In	so‐called	“double‐loop	learning”	(Argyris	and	
Shon	1978),	the	learning	might	even	lead	to	a	re‐framing	of	the	decision,	a	re‐examination	of	the	
objectives,	or	consideration	of	new	alternatives	(this	could	be	represented	by	a	loop	from	step	7	to	
step	1	in	Fig.	A‐1).		In	the	context	of	eagle	management	at	wind	facilities,	the	Service’s	focus	is	on	
the	inner‐loop	learning	(represented	by	the	feedback	from	step	7	to	8	to	4	in	Fig.	A‐1),	but	
unanticipated	learning	will	not	be	ignored.	
	
In	the	case	of	managing	eagle	populations	in	the	face	of	energy	development,	there	is	considerable	
uncertainty	to	be	reduced.		For	example,	we	believe	that	in	some	areas	or	specific	situations,	large	
soaring	birds,	specifically	raptors,	might	be	especially	vulnerable	to	colliding	with	wind	turbines	
(Barrios	and	Rodriguez	2004,	Kuvlesky	et	al.	2007),	but	we	are	uncertain	about	the	relative	
importance	of	factors	that	influence	that	risk.		We	are	also	uncertain	about	the	best	way	to	mitigate	
the	effects	of	wind	turbine	developments	on	raptors;	we	suspect	some	strategies	might	be	effective,	
others	are	worth	trying.		We	also	suspect	that	a	few	species,	including	golden	eagles	(USFWS	2009),	
may	be	susceptible	enough	to	collisions	with	wind	turbines	that	populations	may	be	negatively	
affected.		Thus,	there	are	uncertainties	at	several	levels	that	challenge	our	attempts	to	manage	eagle	
populations:	(1)	at	the	level	of	understanding	factors	that	affect	collision	risk,	(2)	at	the	level	that	
influences	population	trends,	and	(3)	about	the	efficacy	of	various	mitigation	options.		The	Service,	
our	conservation	partners,	and	industry	will	never	have	the	luxury	of	perfect	information	before	
needing	to	act	to	manage	eagles.		We	are	therefore	left	to	make	management	decisions	based	on	the	

45
best	available	information	with	some	inherent	degree	of	uncertainty	about	the	outcomes	of	those	
decisions.		Our	goal	is	to	reduce	that	uncertainty	through	use	of	formal	adaptive	management,	
thereby	improving	our	predictive	capability	over	time.		Applying	a	systematic,	cohesive,	nationally‐
consistent	strategy	of	management	and	monitoring	is	necessary	to	accomplish	this	goal.	

	
	
	

1

2

Problem
Framing

Elicit
Objectives

4

Update
Predictive
Models (Learn)

Develop
Alternatives

5,9 Identify

Evaluate
Consequences

8

3

Preferred
Alternative

6 Evaluate
Critical
Uncertainty

Implement
Action

7
Monitor

	
Figure A-1: A framework for adaptive resource management (ARM). At the core of adaptive
management is critical uncertainty that impedes the identification of a preferred alternative. When
decisions are recurrent, implementation coupled with monitoring can resolve uncertainty, and allow
future decisions to reflect that learning. (Figure from Runge 2011).

	

	

	

	

1. Adaptive Management as a Tool
Using	adaptive	management	as	a	tool	to	manage	wildlife	populations	is	not	new	to	the	Service.		We	
and	other	agencies	are	increasingly	using	the	principles	of	adaptive	management	across	a	range	of	
programs,	including	waterfowl	harvest	management	(Johnson	et	al.	1997),	endangered	species	
(Runge	2011),	and	habitat	management	at	local	and	landscape	scales	(Lyons	et	al.	2008).		Applying	
adaptive	management	to	complex	resource	management	issues	is	promoted	throughout	the	
Department	of	the	Interior	(Williams	et	al.	2007).	

46
	
Waterfowl	harvest	management	is	the	classic	example	of	adaptive	resource	management.		Hunting	
regulations	are	reset	each	year	in	the	United	States	and	Canada	through	the	application	of	adaptive	
management	principles	(Johnson	et	al.	1997).		A	key	uncertainty	in	waterfowl	management	is	the	
extent	to	which	harvest	mortality	is	compensated	by	reductions	in	non‐harvest	mortality	or	by	
increases	in	productivity	(Williams	et	al.	1996).		Various	population	models	have	been	built	based	
on	competing	hypotheses	to	answer	this	question;	these	competing	models	make	different	
predictions	about	how	the	population	will	respond	to	hunting.		Every	year	the	Service	and	the	
Canadian	Wildlife	Service	monitor	waterfowl	and	environmental	conditions	to	estimate	population	
size,	survival	rates,	productivity,	and	hunting	rates.		These	data	feed	into	the	various	competing	
models,	and	the	models	are	evaluated	annually	based	on	how	well	they	predict	changes	in	
waterfowl	populations.		Models	that	perform	best	year‐after‐year	accrue	increasing	weight	(i.e.,	
evidence	in	support	of	the	underlying	hypothesis).		Weighted	model	outputs	directly	lead	to	
recommended	sets	of	hunting	regulations	(e.g.,	bag	limits	and	season	lengths)	for	the	subsequent	
year.		Over	time,	by	monitoring	the	population	effects	of	various	harvest	rates	on	survivorship,	and	
environmental	conditions	on	productivity,	our	uncertainty	about	the	degree	to	which	harvest	is	
compensated	by	other	factors	has	been	reduced,	allowing	for	the	setting	of	harvest	rates	with	
greater	confidence	every	year.		The	application	of	adaptive	management	principles	to	waterfowl	
harvest	regulation	has	helped	the	Service	and	its	partners	achieve	or	exceed	population	goals	for	
most	species	of	waterfowl	(NAWMP	2004).	
	
Adaptive	management	is	a	central	component	of	the	Service’s	approach	to	collaborative	
management	at	the	landscape	scale,	through	strategic	habitat	conservation	(NEAT	2006).		The	
principles	of	adaptive	management	are	also	embedded	in	endangered	species	management	(Ruhl	
2004,	Runge	2011),	including	in	recovery	planning	(Smith	2011)	and	habitat	conservation	planning	
(Wilhere	2002).		Indeed,	the	Service	recognizes	that	adaptive	management	is	a	normative	concept	
in	modern	ecological	decision‐making	(Callicott	et	al.	1999),	and	embraces	it	as	a	fundamental	tool.	
	

2. Applying Adaptive Management to Eagle Take Permitting
In	the	context	of	wind	energy	development	and	eagle	management	under	the	ECPG,	there	are	four	
specific	sets	of	decisions	that	are	suitable	for	an	adaptive	management	approach.	

	
	
a. Adaptive Management of Wind Project Operations
The	most	immediate	and	direct	opportunity	for	adaptive	management	is	at	the	site‐level	for	
wind	facilities	after	construction.		The	relevant	uncertainty	is	in	the	predictions	of	eagle	
take	at	the	project,	and	the	operational	factors	that	influence	the	level	of	take.		The	role	of	
adaptive	management	at	this	scale	will	be	analyzed	and	evaluated	in	the	NEPA	associated	
with	each	permit.		Under	the	ECPG,	a	wind	project	would	initially	work	with	the	Service	to	
generate	predictions	of	take,	given	the	siting,	design,	and	operational	parameters	of	the	
project.		These	predictions	are	made	under	uncertainty,	and	the	risk	to	eagles	associated	
with	this	uncertainty	is	factored	into	the	compensatory	mitigation	terms	of	the	permit	
under	BGEPA.		After	a	site	becomes	operational,	ongoing	surveys	of	realized	take	can	be	
compared	to	the	predictions	of	take.		At	the	review	points	of	the	permit	(typically,	every	five	
years),	the	Service	and	the	operator	will	review	the	observed	take.		If	the	observed	take	
exceeds	the	predicted	and	permitted	take,	the	Service	will	work	with	the	operator	to	
identify	measures	that	could	be	taken	to	reduce	the	take	below	the	permitted	threshold	
(within	the	limits	jointly	agreed	to	at	the	outset	of	the	permit	period).		The	monitoring	data	
may	provide	clues	about	how	this	could	be	done,	for	example,	by	identifying	where	and	
when	most	of	the	take	is	occurring.		On	the	other	hand,	if	the	observed	take	is	significantly	

47
less	than	the	predicted	take,	the	Service	can	work	with	the	operator	to	update	the	
predictions	of	take	for	the	next	review	period,	adjust	the	conditions	for	compensatory	
mitigation,	and	return	credits	to	the	operator	for	any	excess	compensatory	mitigation.	
	
In	a	related	manner,	for	both	new	and	existing	facilities,	ongoing	monitoring	can	provide	
information	to	reduce	uncertainty	about	the	effectiveness	of	conservation	measures	and	
ACPs.		In	particular,	experimental	conservation	measures	and	ACPs	are	actions	taken	by	the	
operator	that	are	thought	to	reduce	mortality	risk,	but	there	is	uncertainty	about	how	
effective	some	of	these	measures	can	be.		In	the	end,	the	purpose	of	adaptive	management	
of	operations	is	to	reduce	mortality	of	eagles	while	also	reducing	the	impact	of	conservation	
measures	and	ACPs	on	power	generation	at	wind	facilities.	

	
b. Adaptive Management of Wind Project Siting and Design Recommendations
Through	the	ECPG	and	the	permit	review	process,	the	Service	makes	recommendations	to	
operators	about	how	to	site	and	design	wind	facilities	to	reduce	eagle	disturbance	and	
mortality.		These	recommendations	are	based	on	the	best	available	science,	but	
acknowledge	that	our	understanding	of	the	interaction	between	eagles	and	wind	facilities	is	
incomplete.		Adaptive	management	provides	the	opportunity	to	respond	to	increasing	
understanding	about	this	interaction.	
	
The	particular	focus	of	this	layer	of	adaptive	management	is	the	predictions	of	take	that	are	
made	by	considering	pre‐construction	surveys	and	risk	factors	(see	APPENDIX	D).		The	
proposed	models	are	initially	quite	coarse	in	their	ability	to	make	predictions,	but	the	
Service,	in	partnership	with	the	U.S.	Geological	Survey	(USGS),	plans	to	refine	these	models.		
The	key	uncertainties	concern	the	risk	factors	that	are	important	in	predicting	eagle	take.		
For	example,	how	important	is	the	proximity	to	nesting	sites,	prey	concentrations,	or	
ridgelines	in	determining	the	risk	posed	by	any	wind	turbine?		Multiple	models	will	be	
developed	to	express	uncertainty	in	these	risk	factors,	and	the	predictions	from	these	
multiple	models	will	be	compared	to	the	patterns	of	observed	take	at	existing	facilities.		
Using	multiple	models	to	express	uncertainty	allows	inclusion	and	evaluation	of	alternative	
models	from	different	sources.		The	learning	that	emerges	will	be	used	to	improve	the	
predictions	from	the	models,	which	in	turn,	will	allow	future	recommendations	about	siting	
and	design	to	be	enhanced.		In	this	case,	the	benefit	of	the	monitoring	at	individual	sites	
accrues	to	the	wind	industry	as	a	whole.	

	
c. Adaptive Management of Compensatory Mitigation
The	determination	of	appropriate	levels	of	compensatory	mitigation,	such	as	through	a	
resource	equivalency	analysis	(REA,	see	APPENDIX	F),	is	based	on	two	predictions:	the	level	
of	take	expected	at	a	project;	and	the	amount	of	mitigation	required	to	offset	that	take.		As	
noted	above,	site‐level	learning,	through	observation	of	realized	take,	can	be	used	to	update	
predictions	of	take,	and	compensatory	mitigation	can	be	adjusted	accordingly.		In	addition,	
the	accrued	experience	across	sites,	through	monitoring	of	the	effectiveness	of	
compensatory	mitigation	projects	and	eagle	population	responses,	can	be	used	to	update	
the	methods	and	parameters	in	the	REA	methods	used	to	determine	the	appropriate	level	of	
compensatory	mitigation.	

	
d. Adaptive Management of Population-Level Take Thresholds
Healthy,	robust	populations	of	animals	can	sustain	some	degree	of	incidental	take,	without	
long‐term	adverse	impacts	to	the	population	or	the	ecosystem.		The	amount	of	take	that	is	

48
sustainable	and	that	can	be	authorized	is	a	function	of	both	scientific	factors	(e.g.,	the	
intrinsic	growth	rate	and	carrying	capacity	of	the	population)	and	policy	interpretation	(e.g.,	
the	amount	of	potential	growth	that	can	be	allocated	to	take,	and	the	risk	tolerance	for	
excessive	take)	(Runge	et	al.	2009).		The	capacity	to	sustain	incidental	take	arises	from	the	
resilience	in	populations	due	to	the	ability	to	compensate	for	that	take	by	increasing	
survival	or	reproductive	rates.	
	
At	the	scale	of	regional	populations	(e.g.,	bird	conservation	regions	for	golden	eagles),	the	
central	question	for	eagles	is	not	altogether	different	than	it	is	for	waterfowl:	to	what	extent	
is	mortality	from	energy	development,	or	any	other	anthropogenic	source,	compensated	by	
reductions	in	mortality	from	other	sources,	or	by	increases	in	productivity?		These	
questions	are	best	answered	by	building	population	models	founded	on	competing	
hypotheses	that	incorporate	estimates	of	mortality,	productivity,	and	the	variation	around	
those	vital	rates.		What	is	needed	is	a	systematic	effort	to	collect	information	on	mortality,	
breeding,	and	population	status	to	feed	those	models.		Similar	to	waterfowl	management,	
reducing	uncertainty	in	population‐level	models	for	eagle	management	will	require	rolling	
up	the	results	of	local	monitoring	and	research	across	the	distribution	of	eagles.		The	results	
will	allow	the	Service	to	make	more	informed	management	recommendations	to	reach	the	
Service’s	population	goal	of	stable	or	increasing	breeding	populations	for	both	eagle	
species.	
	
At	present,	the	Service’s	regulations	call	for	no	increase	in	net	take	of	golden	eagles,	under	a	
protective	concern	that	the	current	level	of	take	exceeds	a	sustainable	threshold.		As	our	
understanding	of	golden	eagle	population	size	and	status	increases,	and	our	knowledge	of	
vital	rates	and	potential	resilience	improves,	the	Service	and	USGS	will	reanalyze	the	
potential	for	instituting	take	thresholds	for	golden	eagles.		Take	thresholds	for	bald	eagles	
will	also	be	re‐assed	no	less	frequently	than	every	five	years	(USFWS	2009).		If	thresholds	
for	either	species	are	increased	and	additional	take	is	authorized,	continued	population	
monitoring	will	be	critical	in	providing	feedback	on	population	response	(i.e.,	step	4	to	8	in	
Fig.	A‐1).	

	
	
Literature Cited
Argyris,	C.,	and	D.	Shon.	1978.	Organizational	Learning:	a	Theory	of	Action	Learning.	Addison‐
Wesley,	Reading,	Massachusetts.	
Barrios,	L.,	and	A.	Rodriguez.	2004.	Behavioural	and	environmental	correlates	of	soaring‐bird	
mortality	at	on‐shore	wind	turbines.	Journal	of	Applied	Ecology	41:72‐81.	
Callicott,	J.	B.,	L.	B.	Crowder,	and	K.	Mumford.	1999.	Current	normative	concepts	in	conservation.	
Conservation	Biology	13:22‐35.	
Johnson,	F.	A.,	C.	T.	Moore,	W.	L.	Kendall,	J.	A.	Dubovsky,	D.	F.	Caithamer,	J.	R.	Kelley,	Jr.,	and	B.	K.	
Williams.	1997.	Uncertainty	and	the	management	of	mallard	harvests.	Journal	of	Wildlife	
Management	61:202‐216.	
Kuvlesky,	W.	P.,	Jr,	L.	A.	Brennan,	M.	L.	Morrison,	K.	K.	Boydston,	B.	M.	Ballard,	and	F.	C.	Bryant.	
2007.	Wind	energy	development	and	wildlife	conservation:	challenges	and	opportunities.	The	
Journal	of	wildlife	management	71:2487‐2498.	
Lyons,	J.	E.,	M.	C.	Runge,	H.	P.	Laskowski,	and	W.	L.	Kendall.	2008.	Monitoring	in	the	context	of	
structured	decision‐making	and	adaptive	management.	Journal	of	Wildlife	Management	
72:1683‐1692.	

49
National	Environmental	Assessment	Team	[NEAT].	2006.	Strategic	Habitat	Conservation.	U.S.	Fish	
and	Wildlife	Service,	Arlington,	Virginia,	USA.	
North	American	Waterfowl	Management	Plan,	Plan	Committee	[NAWMP].	2004.	North	American	
Waterfowl	Management	Plan	2004.		Strategic	Guidance:	Strengthening	the	Biological	
Foundation.	Canadian	Wildlife	Service,	U.S.	Fish	and	Wildlife	Service,	Secretaria	de	Medio	
Ambiente	y	Recursos	Naturales.	
Ruhl,	J.	2004.	Taking	adaptive	management	seriously:	A	case	study	of	the	Endangered	Species	Act.	
University	of	Kansas	Law	Review	52:1249‐1284.	
Runge,	M.	C.	2011.	Adaptive	management	for	threatened	and	endangered	species.	Journal	of	Fish	
and	Wildlife	Management	2.	
Runge,	M.	C.,	J.	R.	Sauer,	M.	L.	Avery,	B.	F.	Blackwell,	and	M.	D.	Koneff.	2009.	Assessing	allowable	take	
of	migratory	birds.	Journal	of	Wildlife	Management	73:556‐565.	
Smith,	C.	B.	2011.	Adaptive	management	on	the	central	Platte	River	‐	Science,	engineering,	and	
decision	analysis	to	assist	in	the	recovery	of	four	species.	Journal	of	Environmental	
Management	92:1414‐1419.	
USFWS.		2009.		Final	environmental	assessment.		Proposal	to	permit	take	provided	under	the	Bald	
and	Golden	Eagle	Protection	Act.		U.S.	Fish	and	Wildlife	Service,	Division	of	Migratory	Bird	
Management,	Washington	D.C.,	USA.	
Walters,	C.	J.	1986.	Adaptive	management	of	renewable	resources.	Macmillan,	New	York,	New	York,	
USA.	
Wilhere,	G.	F.	2002.	Adaptive	management	in	habitat	conservation	plans.	Conservation	Biology	
16:20‐29.	
Williams,	B.	K.,	F.	A.	Johnson,	and	K.	Wilkins.	1996.	Uncertainty	and	the	adaptive	management	of	
waterfowl	harvests.	The	Journal	of	wildlife	management	60:223‐232.	
Williams,	B.	K.,	R.	C.	Szaro,	and	C.	D.	Shapiro.	2007.	Adaptive	Management:	The	U.S.	Department	of	
the	Interior	Technical	Guide.	Adaptive	Management	Working	Group,	U.S.	Department	of	the	
Interior,	Washington,	DC,	USA.	

	
	

	

50

APPENDIX B: STAGE 1 – SITE ASSESSMENT
	

Occurrence	of	eagles	and	their	use	of	landscapes	vary	across	broad	spatial	scales.		The	first	step	in	
project	development	is	to	conduct	a	landscape‐scale	assessment,	based	mainly	on	publicly	available	
information,	to	identify	sites	within	a	large	geographic	area	that	have	both	high	potential	for	wind	
energy	and	low	potential	for	negative	impacts	on	eagles	if	a	project	is	developed.		Stage	1	
corresponds	to	Tiers	1	and	2	of	the	WEG	and,	along	with	Stage	2	herein	and	Tier	3	in	the	WEG,	
comprise	the	pre‐construction	evaluation	of	wind	energy	projects.		Depending	on	the	outcome	of	
Stage	1,	developers	decide	whether	to	proceed	to	the	next	stage,	“...	requiring	a	greater	investment	
in	data	collection	to	answer	certain	questions”	(referring	to	Tier	3,	in	the	WEG;	see	also	Table	B‐1).		
The	WEG	should	be	examined	for	general	considerations	relevant	to	Stage	1;	this	appendix	and	the	
following	APPENDIX	C	focus	on	considerations	specific	to	eagles.	
	
The	Stage	1	assessment	should	evaluate	wind	energy	potential	within	the	ecological	context	of	
eagles,	including	considerations	for	the	eagle’s	annual	life‐cycle,	i.e.,	breeding,	dispersal,	migration,	
and	wintering.		The	goal	at	this	stage	is	to	determine	whether	prospective	wind	project	sites	are	
within	areas	known	or	likely	to	be	used	by	eagles	and,	if	so,	begin	to	determine	the	relative	
spatiotemporal	extent	and	type	of	eagle	use	of	the	sites.		Areas	used	heavily	by	eagles	are	likely	to	
fall	into	category	1;	development	in	these	areas	should	be	avoided	because	the	Service	probably	
could	not	issue	project	developers	or	operators	a	programmatic	permit	for	take	that	complies	with	
all	regulatory	requirements.		Stage	1	assessment	is	a	relatively	straightforward	“desktop”	process	
that	probably	should	conduct	before	significant	financial	resources	have	been	committed	to	
developing	a	particular	project.	
	
Multiple	data	sources	can	be	consulted	when	evaluating	a	prospective	site’s	value	to	eagles.		
Wildlife	biologists	and	other	natural	resource	professionals	from	federal	agencies	including	the	
Service,	and	tribal,	state,	and	county	agencies	should	be	consulted	early	in	the	Stage	1	process	to	
help	ensure	all	relevant	information	is	being	considered.		Information	mainly	encompasses	
physiographic	and	biological	factors	that	could	affect	eagle	risk	associated	with	wind	energy	
development.		Questions	generally	focus	on:	(1)	recent	or	historical	nesting	and	seasonal	
occurrence	data	for	eagles	at	the	prospective	area;	(2)	migration	or	other	regular	movement	by	
eagles	through	the	area	or	surrounding	landscape;	(3)	seasonal	concentration	areas	such	as	a	
communal	roost	site	in	a	mature	riparian	woodland	or	a	prairie	dog	(Cynomys	spp.)	town	serving	as	
a	major	forage	base;	and	(4)	physical	features	of	the	landscape,	especially	topography,	that	may	
attract	or	concentrate	eagles.		“Historical”	is	defined	here	as	5	or	more	years;	a	search	for	historical	
data	should	encompass	at	least	the	previous	5	years.		Data	from	far	longer	time	periods	may	be	
available	but	should	be	cautiously	scrutinized	for	confounding	factors	such	as	land	use	change	that	
diminish	the	data’s	relevance.	
	
Preliminary	site	evaluation	could	begin	with	a	review	of	publically	available	information,	including	
resource	databases	such	as	NatureServe	(http://www.natureserve.org/)	and		the	American	Wind	
Wildlife	Institute’s	Landscape	Assessment	Tool	(LAT;	http://www.awwi.org/initiatives/	
landscape.aspx);	information	from	relevant	tribal,	state,	and	federal	agencies,	including	the	Service;	
state	natural	heritage	databases;	state	Wildlife	Action	Plans;	raptor	migration	databases	such	as	
those	available	through	Hawk	Migration	Association	of	North	America	(http://www.hmana.org)	or	
HawkWatch	International	(http://www.hawkwatch.org);	peer‐reviewed	literature	and	published	
technical	reports;	and	geodatabases	of	land	cover,	land	use,	and	topography	(e.g.,	the	LAT	
integrates	several	key	geodatabases).		Additional	information	on	a	site’s	known	or	potential	value	
to	eagles	can	be	garnered	by	directly	contacting	persons	with	eagle	expertise	from	universities,	
conservation	organizations,	and	professional	or	state	ornithological	or	natural	history	societies.		

51
Some	of	this	wide	assortment	of	desktop	information	and	certain	knowledge	gaps	identified	
probably	will	necessitate	validation	through	site‐level	reconnaissance,	as	suggested	in	the	WEG.	
	
Using	these	and	other	data	sources,	a	series	of	questions	should	be	considered	to	help	place	the	
prospective	project	site	or	alternate	sites	into	an	appropriate	risk	category.		Relevant	questions	
include	(modified	from	the	WEG):	
	
1. Does	existing	or	historical	information	indicate	that	eagles	or	eagle	habitat	(including	
breeding,	migration,	dispersal,	and	wintering	habitats)	may	be	present	within	the	
geographic	region	under	development	consideration?	
2. Within	a	prospective	project	site,	are	there	areas	of	habitat	known	to	be	or	potentially	
valuable	to	eagles	that	would	be	destroyed	or	degraded	due	to	the	project?	
3. Are	there	important	eagle	use	areas	or	migration	concentration	sites	documented	or	
thought	to	occur	in	the	project	area?	
4. Does	existing	or	historical	information	indicate	that	habitat	supporting	abundant	prey	for	
eagles	may	be	present	within	the	geographic	region	under	development	consideration	
(acknowledging,	wherever	appropriate,	that	population	levels	of	some	prey	species	such	as	
black‐tailed	jackrabbits	(Lepus	californicus)	cycle	dramatically	[Gross	et	al.	1974]	such	that	
they	are	abundant	and	attract	eagles	only	in	certain	years	[e.g.,	Craig	et	al.	1984])?	
5. For	a	given	prospective	site,	is	there	potential	for	significant	adverse	impacts	to	eagles	
based	on	answers	to	above	questions	and	considering	the	design	of	the	proposed	project?	
	
We	recommend	development	of	a	map	that,	based	on	answers	to	the	above	questions,	indicates	
areas	that	fall	under	site	category	1,	i.e.,	areas	where	wind	energy	development	would	pose	
obvious,	substantially	high	risks	to	eagle	populations.		Remaining	areas	could	be	tentatively	
categorized	as	either	moderate	to	high	but	mitigable	risk	or	minimal	risk	to	eagle	populations	
(category	2	or	category	3).		Prospective	sites	that	fall	into	category	1	at	this	point	are	unlikely	
candidates	for	a	programmatic	permit	for	take	of	eagles,	although	classification	of	a	site	at	Stage	1	
might	be	regarded	as	tentative	(see	“Assessing	Risk	and	Effects;	4.	Site	Categorization	Based	on	
Mortality	Risk	to	Eagles”	in	the	ECPG.		If	a	site	appears	to	be	a	category	1	site	based	on	the	outcome	
of	Stage	1,	the	developer	can	decide	whether	information	at	that	stage	adequately	supports	a	
category	decision	or	whether	to	invest	in	Stage	2	assessment	to	clarify	preliminary	indications	of	
Stage	1	(Table	B‐1).		Sites	that	tentatively	fall	into	categories	2	or	3	at	Stage	1	can	move	on	to	Stage	
2	assessment,	but	could	ultimately	be	excluded	as	permit	candidates	after	more	site‐specific	data	
are	collected	in	Stage	2.	
	
Again,	the	goal	of	Stage	1	site	assessment	in	this	ECPG	is	to	determine	whether	prospective	wind	
project	sites	are	within	areas	known	or	likely	to	be	used	by	eagles	and,	if	so,	begin	to	assess	the	
spatiotemporal	extent	and	type	of	eagle	use	the	sites	receive	or	are	likely	to	receive.		Thus,	the	
ultimate	goal	of	Stage	1	is	to	determine	whether	sites	exhibit	any	obvious	substantial	risk	for	eagles.		
For	those	that	do	not,	the	Stage	1	site	assessment	will	provide	fundamental	support	for	the	design	
of	detailed	surveys	in	Stage	2,	decisions	which	influence	optimal	allocation	of	the	financial	
investment	in	surveys	and	quality	of	data	collected.		In	some	situations,	the	Stage	1	site	assessment	
may	provide	enough	information	to	adequately	estimate	impacts	and	support	decisions	on	site	
categorization	(and,	where	relevant,	potential	conservation	measures	and	appropriate	levels	of	
compensatory	mitigation),	rendering	Stage	2	assessment	unnecessary	(Table	B‐1).	

	
	

52

Literature Cited
Craig,	T.	H.,	E.	H.	Craig,	and	L.	R.	Powers.		1984.		Recent	changes	in	eagle	and	buteo	abundance	in	
southeastern	Idaho.	Murrelet	65:91‐93.	
Gross,	J.	E.,	L.	C.	Stoddart,	and	F.	H.	Wagner.		1974.		Demographic	analysis	of	a	northern	Utah	
jackrabbit	population.	Wildlife	Monograph	40.	

	
	
Table B-1. Framework for decisions on investment at Stage 2 level to address chief information needs.
A bidirectional arrow represents a continuum of conditions.

	
	

Area	of	
Information	
Need	
	
Seasonal	
abundance	
	
Nesting	records	
	
Migration	
corridors	
	
Communal	
roosts	
	
Prey	availability	
or	foraging	
hotspots	

Strength	of	Stage	1	Information	Base	for	Assessing	Risk	to	
Eagles	
Robust:	
well	investigated	and	supported,	at	
least	semi‐quantitative	
documentation	from	most	recent	2‐5	
years,	encompassing	potential	site(s)	
or	adjoining	areas	from	which	reliable	
inferences	can	be	made	

↔

Weak:
characterized	by	little	
supportive	information	and	
marginal	certainty	overall,	at	
best	only	general	descriptions,	
conjecture,	or	limited	
inferences	from	other	areas	or	
regions	

	

↔

	

	

	

↔
↔

	

	

	

↔

	

	

↔

	

↔

Uncertain	risk	level	–	strong	
survey	effort	at	Stage	2	level	
advised	

Relevant	areas	of	information	need	
are	well‐addressed	and	risk	level	is	
clearly	low	–	Stage	2	may	not	be	
Outcome	and	
implications	for	 warranted	or	else	modest	or	limited‐
focus	survey	effort	at	Stage	2	level	
additional	
recommended	
assessment	
	
needs	at	Stage	2	 Relevant	areas	of	information	need	
are	well‐addressed	and	risk	level	is	
level:	
moderate	or	high	–		strong	effort	at	
Stage	2	level	advised	

53

APPENDIX C: STAGE 2 – SITE-SPECIFIC SURVEYS AND ASSESSMENT
1. Surveys of Eagle Use
Information	collected	in	Stage	2	is	used	mainly	to	generate	predictions	of	the	mean	annual	number	
of	eagle	fatalities	for	a	prospective	wind	energy	project	and	to	identify	important	eagle	use	areas	or	
migration	concentration	sites	that	could	be	affected	by	the	project.		Information	from	Stage	2	is	also	
used	to	assess	the	likelihood	of	disturbance	take	of	eagles.		An	array	of	survey	types	could	be	used	
to	quantify	use	by	eagles	of	a	proposed	project	area.		This	section	focuses	on	four	types	of	surveys	
recommended	for	assessing	risk	to	eagles	at	proposed	wind	projects.		The	first	three	are	surveys	of	
eagle	use	within	the	proposed	project	footprint.		These	include:	(1)	point	count	surveys,	which	
mainly	generate	occurrence	data	that	form	underpinnings	of	the	risk	assessment	model	
recommended	herein;	(2)	migration	(“hawk	watch”)	counts,	documenting	hourly	passage	rates	of	
eagles;	and	(3)	utilization	distribution	(UD)	assessment,	an	accounting	of	the	intensity	of	use	of	
various	parts	of	the	home	range	within	the	project	footprint;	and	(4)	surveys	of	nesting	territory	
occupancy	in	the	project	area.		Where	uncertainties	exist	regarding	survey	methods,	our	
recommendations	tend	to	be	conservative	such	that	biases	in	survey	data,	if	any,	are	more	likely	to	
favor	greater	rather	than	lower	estimates	of	use	and	ultimately	more	rather	than	less	protection	for	
eagles.		This	approach	is	consistent	with	the	Service’s	policy	of	taking	a	risk‐averse	stance	in	the	
face	of	existing	uncertainty	with	respect	to	eagle	programmatic	take	permits.	
	
In	addition	to	fatality	estimation	and	informing	a	site	categorization	decision,	Stage	2	studies	of	
eagles	should	help	answer	the	following	questions	(modified	from	the	WEG):	
1. What	is	the	distribution,	relative	abundance,	behavior,	and	site	use	of	eagles	and	to	what	
extent	do	these	factors	expose	eagles	to	risk	from	the	proposed	wind	energy	project?	
2. What	are	the	potential	risks	of	adverse	impacts	of	the	proposed	wind	energy	project	to	
individual	and	local	populations	of	eagles	and	their	habitats?	
3. How	can	developers	avoid,	minimize,	and	mitigate	identified	adverse	impacts?	
4. Are	there	studies	that	should	be	initiated	at	this	stage	that	would	be	continued	in	post‐
construction?	

	
a. Point Count Surveys
Point	counts	(i.e.,	circular‐plot	surveys)	often	are	used	to	assess	relative	abundance,	
population	trends,	and	habitat	preferences	of	birds	(Johnson	1995).		The	Service	advocates	
use	of	point	count	surveys	as	the	means	of	providing	primary	input	for	models	predicting	
fatality	rate	of	eagles	associated	with	wind	turbines.		However,	we	acknowledge	the	term	
point	count	survey	does	not	accurately	describe	the	approach	we	advocate	for	collecting	
data	to	support	fatality	rate	estimation	at	wind	energy	projects.		The	Service’s	approach	in	
this	regard	is	point‐based	recording	of	activity	duration	(minutes	of	flight)	within	a	three‐
dimensional	plot.		In	contrast,	point	count	surveys,	as	typically	conducted,	yield	indices	of	
relative	abundance	or	frequency	of	occurrence	(in	addition	to	trend,	density	estimation,	and	
habitat	association,	depending	on	how	data	are	collected;	Ralph	et	al.	1993).		With	that	said,	
most	records	of	eagle	flight	duration	are	likely	to	be	classified	as	1	minute,	per	the	approach	
recommended	in	this	section,	and	as	such	resemble	records	of	occurrence	for	data	from	
point	count	surveys.		Although	a	bit	of	a	misnomer	in	this	regard,	“point	count	survey”	is	
applied	broadly	herein	to	include	both	point‐based	records	of	flight	time	and	traditional	
point	count	surveys	because	sampling	frameworks	for	each	so	closely	overlap	and	both	data	
types	can	be	gathered	simultaneously,	along	with	other	information	described	in	this	
appendix.		There	may	be	other	means	of	generating	count	data	to	support	the	fatality	model	

54
described	in	this	document.		Consideration	of	alternative	approaches	for	predicting	fatality	
at	such	projects	may	require	greater	time	and	additional	reviews.	
	
The	general	approach	for	conducting	a	fixed‐radius	point	count	survey	is	to	travel	to	a	pre‐
determined	point	on	the	landscape	and	record	individual	birds	detected	–	whether	
observed,	only	heard,	or	both	observed	and	heard	–	within	a	circular	plot,	the	boundary	of	
which	is	at	a	fixed	distance	from	the	point	and	is	marked	in	the	field	in	several	places	(Hutto	
et	al.	1986,	Ralph	et	al.	1993).		In	addition	to	plot	radius,	the	survey	is	standardized	by	count	
duration.		Sometimes	a	variable‐radius	plot	method	(Reynolds	et	al.	1980)	is	used,	yielding	
species‐by‐species	detectability	coefficients	to	appropriately	bound	the	plot	radius	(i.e.,	
sampling	area)	for	each	species.		A	variety	of	point	count	survey	methods	have	been	used	
specifically	for	raptors	(reviewed	in	Anderson	[2007];	the	North	American	Breeding	Bird	
Survey	[Sauer	et	al.	2009]	is	a	random‐systematic,	continent‐wide	point	count	survey	of	
bird	population	trends,	including	those	of	many	raptor	species).	However,	a	fixed‐radius	
approach	with	circular	plots	of	800‐m	radius	typically	is	used	for	surveying	eagles	and	
other	large	(greater	than	crow	[Corvus	spp.]‐size)	diurnal	species	of	raptors	at	proposed	
wind	energy	projects	in	the	United	States	(Strickland	et	al.	2011).	
	
The	optimal	duration	of	point	count	survey	for	eagles	is	a	focus	of	current	research.		For	
now,	for	point	count	surveys	of	eagles	at	proposed	wind	energy	projects,	the	Service	
recommends	counts	of	1,	2,	or	more	hours	duration	instead	of	20‐	to	40‐minute	counts	
typically	used	(Strickland	et	al.	2011).	Longer	counts	also	facilitate	integration	of	other	
survey	types	(e.g.,	development	of	utilization	distribution	profiles).		Many	raptor	biologists	
have	suggested	that	the	likelihood	of	detecting	an	eagle	during	a	20‐	to	40‐minute	point	
count	survey	is	extremely	low	in	all	but	locales	of	greatest	eagle	activity	and	datasets	
generated	by	pre‐construction	point	count	surveys	of	this	duration	typically	are	replete	
with	counts	of	zero	eagles,	resulting	in	unwieldy	confidence	intervals	and	much	uncertainty.		
Moreover,	time	spent	traveling	to	and	accessing	points	for	20‐minute	surveys	may	exceed	
time	spent	conducting	the	observations.		For	example,	250	1‐hour	surveys	conducted	
annually	at	a	project	of	average	size		(e.g.,	15	sampling	points,	1	to	3	km	apart)	and	travel	
conditions	require	roughly	the	same	total	field	time	as	needed	for	500	20‐minute	surveys,	
yet	yield	50%	more	observation	hours	(250	versus	167),	with	correspondingly	greater	
probability	of	detecting	eagles.		Another	advantage	of	longer	counts	is	that	they	reduce	
biases	created	if	some	eagles	avoid	conspicuous	observers	as	they	approach	their	points	
and	begin	surveys,	although	some	observers	may	become	fatigued	and	overlook	eagles	
during	longer	counts.		A	potential	trade	off	of	fewer	visits,	of	course,	is	diminished	
accounting	of	temporal	variation	(e.g.,	variable	weather	conditions	or	an	abrupt	migration	
event).		While	counting	at	fewer	points	for	longer	periods	might	also	reduce	the	ability	to	
sample	more	area,	we	advocate	maintain	the	minimum	spatial	coverage	of	at	least	30%	of	
the	project	footprint.		Until	there	is	more	evidence	that	shorter	count	intervals	are	adequate	
to	estimate	eagle	exposure,	we	believe	that	a	sampling	strategy	including	counts	of	longer	
duration,	albeit	fewer	total	counts,	may	in	the	end	improve	sampling	efficiency	and	data	
quality.	
	
A	key	assumption	of	fatality	prediction	models	based	on	data	from	point	count	surveys	is	
that	occurrence	of	eagles	at	a	proposed	project	footprint	before	construction	bears	a	
positive	relationship	with	turbine‐collision	mortality	after	the	project	becomes	operational	
(Strickland	et	al.	2011).		Support	for	this	assumption	from	published	literature	is	limited	for	
eagles	and	other	diurnal	raptors	at	this	time,	however.		In	a	recent	study	of	raptors	at	20	
projects	in	Europe,	no	overall	relationship	was	evident	between	either	of	two	pre‐

55
construction	risk	indices	and	post‐construction	mortality	(Ferrer	et	al.	2011).		However,	the	
authors	based	risk	indices	only	in	part	on	data	from	pre‐construction	point	counts;	factors	
incorporated	into	risk	indices	included	a	somewhat	subjective	decision	on	species‐specific	
sensitivity	to	collision	and	conservation	status.		Despite	this,	a	weak	relationship	between	
pre‐construction	flight	activity	and	post‐construction	mortality	was	suggested	for	the	most	
common	species,	griffon	vulture	(Gyps	fulvus)	and	kestrels	(Falco	spp.).		Neither	Aquila	nor	
Haliaeetus	eagles	occurred	in	the	study.		On	coastal	Norway,	however,	a	high	density,	local	
population	of	the	white‐tailed	eagle,	a	species	closely	related	and	ecologically	similar	to	the	
bald	eagle,	experienced	substantial	turbine‐collision	fatality	and	loss	of	nesting	territories	
after	development	of	a	wind	energy	project	(Nygård	et	al.	2010).		The	relationship	between	
pre‐construction	occurrence	and	post‐construction	mortality	might	be	less	clear	if	eagles	
and	other	raptor	species	avoided	areas	after	wind	energy	projects	were	constructed	(e.g.,	
Garvin	et	al.	2011),	but	in	general	such	displacement	seems	negligible	(Madders	and	
Whitfield	2006).	
	
Precision,	consistency,	and	utility	of	data	derived	from	point	count	surveys	depend	greatly	
on	the	sampling	framework	and	field	approach	for	conducting	the	counts,	which	in	turn	
depend	somewhat	on	study	objectives	and	the	array	of	species	under	consideration.		
Precision	and	reliability	of	data	from	point	count	surveys	for	eagles	can	be	much	improved	
upon	–	and	need	for	a	risk‐averse	approach	lessened	–	by	incorporating	some	basic,	
common‐sense	sideboards	into	the	survey	design.		One	of	these,	longer	count	duration,	is	
discussed	above.		Below	are	examples	of	ideal	design	features	for	point	count	surveys	of	
eagle	use	of	proposed	wind	energy	projects,	particularly	when	fatality	rate	prediction	is	a	
primary	objective.		Some	of	these	extend	from	Strickland	et	al.	(2011)	and	references	
therein,	although	the	first	is	not	in	accord	with	corresponding	guidance	in	that	document.	
	
 Surveys	of	eagles	and	other	large	birds	are	exclusive	of	those	for	small	birds,	to	
avoid	overlooking	large	birds	while	searching	at	a	much	smaller	scale	for	a	much	
different	suite	of	birds.		The	relatively	brief	(e.g.,	10‐minute)	point	counts	for	small	
birds	could	be	conducted	during	the	same	visit,	but	before	or	after	the	count	of	large	
birds.	
 In	open	areas	where	observers	may	be	conspicuous,	counts	are	conducted	from	a	
portable	blind	or	from	a	blind	incorporated	into	a	vehicle	to	reduce	the	possibility	
that	some	individual	eagles	avoid	observers,	,thus		reducing	likelihood	of	detection.		
Blinds	are	designed	to	mask	conspicuous	observer	movement	while	not	impeding	
views	of	surroundings.	
 Point	locations	may	be	shifted	slightly	to	capitalize	on	whatever	vantage	points	may	
be	available	to	enhance	the	observer’s	view	of	surroundings.	
 Elevated	platforms	(e.g.,	blinds	on	scaffolding	or	high	in	trees,	truck‐mounted	lifts)	
are	used	to	facilitate	observation	in	vistas	obstructed	by	tall	vegetation,	topographic	
features,	or	anthropogenic	structures.	
 The	observer’s	visual	field	at	a	point	count	plot,	if	less	than	800	m	(e.g.,	due	to	
obstruction	by	forest	cover),	is	mapped.		The	percentage	of	the	plot	area	that	is	
visible	is	factored	into	the	calculation	of	area	surveyed.	
 Observers	use	the	most	efficient,	logical	route	to	move	among	points,	changing	the	
starting	point	with	the	beginning	of	each	survey	cycle	such	that	each	point	is	
surveyed	during	a	range	of	daylight	hours.	
 Systematic	scans	of	the	point	count	plot	using	binoculars	alternating	with	scans	via	
the	unaided	eye	to	detect	close	and	distant	eagles,	and	with	overhead	checks	for	

56











eagles	that	may	have	been	overlooked	during	peripheral	scanning	(Bildstein	et	al.	
2007).	
Observers	are	trained	and	their	skills	are	tested,	including	accurate	identification	
and	distance	estimation	(both	horizontal	and	vertical;		e.g.,	eagles	greater	than	600	
m	horizontal	distance	may	not	be	detected	by	some	observers	and	correction	for	
differences	among	individual	observers	may	be	warranted).	
The	boundary	of	each	point	count	plot	is	identified	via	distinct	natural	or	
anthropogenic	features	or	marked	conspicuously	(e.g.,	flagging	on	poles)	at	several	
points	for	distance	reference.		Distance	intervals	within	the	plot	also	are	marked	if	
observations	are	to	be	categorized	accordingly;	rangefinder	instruments	are	useful	
in	this	regard.	
Surveys	are	distributed	across	daylight	hours	(e.g.,	morning	–	sunrise	to	1100	hours;	
midday	–	1101‐1600;	evening	1601	to	sunset).		In	areas	or	during	seasons	where	
eagle	flight	is	more	likely	during	midday	than	in	early	morning	or	evening	(e.g.,	
migration	[Heintzelman	1986]),	sampling	efficiency	could	be	increased	by	
temporally	stratifying	surveys	to	more	intensively	cover	the	midday	period.	
A	map	(e.g.,	1:24,000	scale	topographic	quadrangle)	or	aerial	photographs	
indicating	topographic	and	other	reference	features	plus	locations	of	point	count	
plots	is	used	as	the	primary	recording	instrument	in	the	field.		A	GPS	with	GIS	
interface	may	serve	in	this	regard.	
Time	and	position	of	each	individual	eagle	is	recorded	on	the	map,	e.g.,	at	the	
beginning	of	each	minute	of	observation,	if	not	more	frequently.	

	
The	following	examples	of	suggested	sideboards	pertain	especially	to	point	count	surveys	
supplying	data	for	the	fatality	prediction	method	recommended	in	this	document:	
	
 Following	a	point	count	survey,	the	duration	of	observation	of	each	eagle	flying	
within	the	plot	is	summarized	in	number	of	minutes,	rounded	to	the	next	highest	
integer	(e.g.,	an	eagle	observed	flying	within	the	plot	for	about	15	seconds	is	1	eagle‐
minute,	another	observed	within	for	about	1	minute	10	seconds	is	2	eagle‐minutes,	
and	so	on;	most	observations	likely	will	equal	1	eagle‐minute).	
 Eagles	are	mapped	when	perched	or	when	otherwise	not	flying,	but	the	summary	of	
eagle‐minutes	for	a	count	excludes	these	observations	and	includes	only	eagles	in	
flight.	
 Horizontal	distance	of	each	eagle‐minute	is	estimated	and	recorded	as	≤	800	m	or	>	
800	m.		Vertical	distance	of	each	eagle‐minute	is	estimated	and	recorded	as	≤	200	m	
(at	or	below	conservative	approximation	of	maximum	height	of	blade	tip	of	tallest	
turbine)	or	>	200	m.		Thus,	the	point	count	“plot”	is	a	200‐m	high	cylinder	with	a	
radius	of	800	m.	
 Surveys	are	done	under	all	weather	conditions	except	that	surveys	are	not	
conducted	when	visibility	is	less	than	800	m	horizontally	and	200	m	vertically.	
 Data	from	point	count	surveys	are	archived	in	their	rawest	form	to	be	available	
when	fatality	is	estimated	as	detailed	in	this	document	(APPENDIX	D).	
	
Other	information	recorded	during	point	counts	may	prove	useful	in	project	assessment	
and	planning,	or	in	additional	data	analyses	(some	requiring	data	pooled	from	many	
projects),		e.g.:	
	

57








Flight	paths	of	eagles,	including	those	outside	the	plot,	are	recorded	on	reference	
maps,	using	topographic	features	or	markers	placed	in	the	field	as	location	
references.		Eagle	flight	paths	are	recorded	also	before	and	after	point	count	surveys	
and	incidental	to	other	field	work.		Flight	paths	are	summarized	on	a	final	map,	with	
those	recorded	during	point	count	surveys	distinguished	from	others	to	roughly	
account	for	spatial	coverage	bias.		Documentation	of	flight	paths	can	aid	planning	to	
avoid	areas	of	high	use	(Strickland	et	al.	2011).	
Behavior	and	activity	prevalent	during	each	1‐minute	interval	is	recorded	as	(e.g.)	
soaring	flight	(circling	broadly	with	wings	outstretched);	unidirectional	flapping‐
gliding;	kiting‐hovering;	stooping	or	diving	at	prey;	stooping	or	diving	in	an	
agonistic	context	with	other	eagles	or	other	bird	species;	undulating/territorial	
flight;	perched;	or	other	(specified).	
Age	class	of	individual	eagles	is	recorded,	e.g.,	juvenile	(first	year),	immature	or	
subadult	(second	to	fourth	year),	adult	(fifth	year	or	greater),	or	unknown.	
Weather	data	are	recorded,	including	wind	direction	and	speed,	extent	of	cloud	
cover,	precipitation	(if	any),	and	temperature	(Strickland	et	al.	2011).	
Distance	measures	are	used	to	estimate	detectability	for	improving	estimates	from	
counts	(Buckland	et	al.	2001)	and	could	be	used	to	assess	whether	eagles	avoid	
observers.		Horizontal	distance	of	each	eagle‐minute	is	estimated	and	categorized,		
e.g.,	in	100‐m	intervals	to	>	800	m.	

	
The	key	consideration	for	planning	point	count	surveys	at	proposed	wind	energy	projects	is	
sampling	effort.		We	advise	that	project	developers	or	operators	coordinate	closely	with	the	
Service	regarding	the	appropriate	seasonal	sampling	effort,	as	sampling	considerations	are	
complex	and	depend	in	part	on	case‐specific	objectives.		We	also	reiterate	that	these	(and	
most	other)	surveys	should	be	conducted	for	at	least	2	years	before	project	construction	
and,	in	most	cases,	across	all	seasons.		In	general,	sampling	effort	should	be	commensurate	
with	the	relative	level	of	risk	at	a	proposed	project	footprint	if	this	can	be	surmised	reliably	
from	the	Stage	1	assessment.		If	Stage	1	information	cannot	support	reasonably	certain	risk	
categorization,	Stage	2	surveys	should	be	conducted	as	described	here	to	clearly	ascertain	
whether	eagles	are	known	or	likely	to	use	the	area.		If	a	project	is	determined	to	be	category	
2,	products	of	point	count	surveys	should	include	data	for	the	fatality	model	detailed	in	this	
document	(APPENDIX	D).		If	there	is	compelling	Stage	1	evidence	indicating	no	use	in	a	
given	season,	zero	use	could	be	assumed	and	point	count	surveys	in	that	season	might	be	
unnecessary.	
	
In	general,	goals	for	the	Stage	2	surveys	are	either	to:	(1)	confirm	category‐3	status	for	a	
project,	or	(2)	to	generate	a	fatality	rate	estimate.		Regardless	of	which	of	these	survey	goals	
apply	to	a	particular	project,	we	recommend	first	identifying	potential	sites	for	wind	
turbines,	including	alternate	sites,	then	calculating	the	total	area	(km	2	)	encompassing	a	1‐
km	buffer	around	all	the	sites.		We	suggest	1	km	because	this	approximates	optimal	spacing	
of	a	generic	2.5‐MW	turbine	(Denholm	et	al.		2009),	and	the	area	outside	this	may	not	be	
representative	of	topographic	features	and	vegetation	types	that	characterize	turbine	
strings	within	the	project	footprint.	This	approach	assures	close	association	between	
sampling	sites	and	likely	turbine	locations,	as	recommended	by	Strickland	et	
al.		(2011).		Next,	we	recommend	that	at	least	30%	of	the	area	within	1	km	of	turbines	be	
considered	as	the	total	km	2		area	to	be	covered	by	800‐m	radius	point	count	plots	(with	a	
sample	area	for	each	plot	of	2	km2).		Our	recommended	30%	minimum	is	based	on	the	
actual	minimum	coverage	at	eight	wind	facilities	under	review	by	the	Service	at	the	time	
version	2	of	the	ECPG	was	being	developed.	

58
	
The	first	case	(i.e.,	(1)	above)	is	the	use	of	point	count	data	to	validate	whether	a	proposed	
project	meets	category	3	criteria	when	Stage	1	information	is	inadequate.		Based	on	
experience	with	current	parameters	of	the	“prior	term”	in	our	predictive	model	(see	
APPENDIX	D),	we	calculate	an	average	of	20	hours	per	turbine	as	an	optimal	level	of	annual	
sampling	via	point	count	survey	(e.g.,	equivalent	of	ten	4‐hour	point	count	surveys	at	each	
of	20	sample	points	for	a	40‐turbine	project;	our	20‐hour	recommendation	considers	the	
hazardous	area	created	by	a	generic	2.5‐MW	turbine	with	a	rotor	diameter	of	about100	m;	
sample	effort	for	turbines	with	smaller	rotor	diameters	would	be	less).		As	sampling	effort	
falls	from	this	level,	uncertainty	regarding	fatality	risk	rises	sharply,	calling	for	an	
increasingly	risk	averse	basis	for	risk	categorization.		Although	20	sample	hours	per	turbine	
may	be	necessary	initially	for	validating	category	3	determination	where	little	Stage	1	
information	exists,	we	expect	this	will	decrease	as	more	projects	are	incorporated	into	the	
adaptive	management	meta‐analyses	that	will	refine	the	prior	term.	
	
The	second	case	(i.e.,	(2)	above)	is	where	Stage	1	evidence	is	strong	enough	to	support	the	
decision	that	a	project	is	category	2	(or	category	3	with	potential	for	re‐evaluation	as	
category	2).		Fatality	rate	estimation	becomes	the	main	objective	of	point	count	surveys	and	
demands	for	sampling	effort	can	be	reduced.		We	recommend	a	minimum	of	1	hour	of	
observation	per	point	count	plot	per	month	but	at	least	2	hours	of	observation	per	point	
count	is	warranted	for	a	season	for	which	Stage	1	evidence	is	ambiguous	or	suggests	high	
use.	
	
These	ideas	on	minimum	observation	hours	stem	from	the	Service’s	initial	experience	in	
fatality	estimation	(see	APPENDIX	D:	Stage	3	–	Predicting	Eagle	Fatalities).		However,	as	
noted	above,	with	more	field	applications	of	our	fatality	prediction	model	we	should	be	able	
to	refine	our	ability	to	characterize	uncertainty	based	in	part	on	site‐specific	characteristics,	
something	the	Service’s	current	model	does	not	do.		Again,	to	develop	a	reasonable,	
informed	sampling	approach,	we	urge	project	developers	to	engage	early	with	the	Service	in	
discussions	about	sampling	design	and	strategies.	
	
The	example	below	includes	determination	of	the	number	of	point	count	plots	for	a	project.	
Example
The	site	for	a	100‐MW,	40‐turbine	project	proposed	in	open	foothills	of	central	New	
Mexico	encompasses	40	km2	(16	mi2).		During	the	Stage	1	assessment,	data	from	a	
hawk	watch	organization	indicates	the	area	is	25	miles	east	of	a	north‐south	
mountain	ridge	that	sustains	a	moderate	level	of	migration	by	golden	eagles	each	
fall	but	receives	little	use	in	spring.		According	to	the	state	ornithological	society,	the	
region	also	is	thought	to	attract	golden	eagles	during	winter,	but	this	is	based	on	
sparse	anecdotal	accounts.		Aerial	nesting	surveys	by	the	Service	5	years	ago	yielded	
no	evidence	of	eagle	nests	within	10	miles	of	the	proposed	project,	although	use	of	
the	area	by	non‐breeding	resident	eagles	during	spring	and	summer	cannot	be	ruled	
out.		Reconnaissance	visits	and	review	of	land	cover	and	other	habitat	layers	in	
geodatabases	support	the	general	indication	that	the	area	is	important	to	golden	
eagles	during	at	least	part	of	the	year.	
	
Stage	1	Summary:	Of	primary	concern	at	the	prospective	project	site	is	potential	for	
risk	to	golden	eagles	during	fall	migration.		Evidence	of	this	at	the	Stage	1	level	is	
somewhat	equivocal,	however,	because	the	known	migration	pathway	is	outside	the	

59
project	area.		Further	examination	of	use	in	spring,	summer,	and	especially	winter	
also	seems	warranted.		Questions	include	temporal	(seasonal)	and	spatial	
(distribution	within	project)	use.		The	overarching	goal	is	to	quantify	risk	to	eagles	
posed	by	the	proposed	project,	mainly	by	estimating	fatality	rate.		If	fatality	is	
anticipated,	a	secondary	goal	is	to	determine	whether	the	predicted	level	is	
acceptable	and,	if	not,	whether	fatality	can	be	avoided	and	minimized	through	
specified	project	design	and	operation	features.	
	
The	primary	tool	for	predicting	fatality	is	the	point	count	survey.		However,	if	the	
pre‐construction	assessment	is	robust	and	optimally	designed,	point	count	surveys	
will	provide	insight	on	distribution	of	use	within	the	project	footprint	especially	
near	proposed	turbine	sites,	and	on	migration	timing	and	movement	pathways.	
	
Sampling Effort
A.		Number	of	points,	i.e.,	point	count	plots,	and	spatial	allocation:	
1. 40	turbines	are	proposed	for	project	
2. potential	sites	for	turbines	have	been	selected	
3. area	within	1	km	of	turbines	covers	total	of	100	km2	
4. 30%	of	total	area	=	30	km2	
5. number	of	800‐m	radius	(area	of	each,	2‐km2)	point	count	plots	
recommended	=	30/2	=	15	plots	
6. survey	points	are	distributed	among	turbine	strings	via	random‐systematic	
allocation,	with	each	point	no	more	than	1	km	from	a	prospective	turbine	
site	
	
B.		Number	of	counts	per	point	per	season	and	duration	of	each	point	count	survey:	
1. Based	on	some	Stage	1	evidence	of	low	use	in	this	example,	1	hour	of	
observation	per	point	count	plot	per	month	seems	appropriate	during	each	
of	winter	(e.g.,	mid‐December	through	mid‐March),	spring	(mid‐March	
through	mid‐June),	and	summer	(mid‐June	through	mid‐September)	
seasons.		A	count	duration	of	1	hour	is	selected	to	maximize	efficiency	in	the	
field	
2. Survey	effort	is	doubled	during	the	mid‐September	through	mid‐December	
fall	migration	season	for	golden	eagles,	based	on	Stage	1	evidence	of	fall	
migration	nearby	and	need	for	more	definitive	data	on	eagle	occurrence,	
timing,	and	distribution	within	the	footprint.		This	could	be	done	by	using	
either	two	1‐hour	counts	or	a	2‐hour	count	per	point	per	month;	the	latter	is	
chosen	to	maximize	field	efficiency	and	better	emulate	migration	count	
methods.		The	1‐hour	counts	may	lend	better	insight	on	temporal	variation,	
but	in	this	example	each	monthly	session	of	15	2‐hour	counts	requires	an	
observer	3‐4	days	to	complete,	affording	some	accounting	of	day‐to‐day	
variation.	
3. The	total	yearly	effort	in	this	example	is	nine	1‐hour	counts	and	three	2‐
hour	counts	at	each	of	15	points,	yielding	225	total	observation	hours.	
	
The	raw	data,	in	number	of	eagle‐minutes,	appear	as	follows	(e.g.,	for	the	first	fall	
season	sampled,	with	one	2‐hour	count	per	point	per	month):	

	

	

60

	
Point	no.	

	

Point	count	visit	number	–	Fall	Season,	Year	1	
1	(early	fall)	
2	(mid‐fall)	
3	(late	fall)	

1	

0	

0	

0	

2	

0	

0	

0	

3	

0	

0	

0	

4	

0	

0	

0	

5	

0	

0	

0	

6	

0	

0	

0	

7	

1	

1	

0	

8	

0	

0	

0	

9	

0	

0	

0	

10	

0	

2	

1	

11	

0	

0	

0	

12	

0	

2	

0	

13	

0	

0	

0	

14	

0	

1	

0	

15	

0	

0	

0	

The	first	year’s	fall	point	count	survey	totals	90	observation	hours,	the	equivalent	of	
nine	10‐hour	migration	counts.		Thus,	the	fall	point	count	surveys	could	yield	much	
insight	on	eagle	migration	–	perhaps	even	substituting	for	focused	migration	counts	
–	especially	if	the	sample	is	stratified	so	point	count	surveys	mainly	cover	the	
midday	period	when	eagles	are	most	likely	to	be	moving.		(see	b.	Migration	Counts	
and	Concentration	Surveys,	below).		Observations	made	during	point	count	surveys	
in	all	seasons	also	could	support	a	map	of	flight	paths	to	roughly	indicate	the	
distribution	of	use	of	the	area	by	eagles	relative	to	turbine	sites	(see	c.	Utilization	
Distribution	(UD)	Assessment,	below).	

	

Fatality	estimation	should	be	adequately	supported	by	the	data,	although	multiple	survey	
years	are	likely	needed	to	account	for	annual	variation.		Data	for	fatality	estimation	should	
be	made	available	to	the	Service	in	the	rawest	form,	as	in	the	above	example.	

	
b. Migration Counts and Concentration Surveys
Wherever	potential	for	eagle	migration	exists,	migration	counts	should	be	conducted	unless	
the	Stage	1	assessment	presents	compelling	evidence	that	the	project	area	does	not	include	
or	is	not	part	of	a	migration	corridor	or	a	migration	stopover	site.		Migration	counts	convey	
relative	numbers	of	diurnal	raptors	passing	over	an	established	point	per	unit	time	
(Bildstein	et	al.	2007,	Dunn	et	al.	2008),	usually	a	migration	concentration	site.		Examples	of	
sites	include	north‐south	oriented	ridges,	cliff	lines,	or	deeply	incised	river	valleys;	terminal	
points	or	coast	lines	of	large	water	bodies;	or	peninsulas	extending	into	large	water	bodies	
(Kerlinger	1989,	Bildstein	2006,	Mojica	et	al.	2008).		Migration	counts	could	be	considered	a	
specialized	type	of	point	count,	one	for	which	the	plot	radius	is	unlimited	(Reynolds	et	al.	
1980)	and	the	count	period	is	quite	long,	from	6	hours	to	a	full	day.	
	

61
In	contrast	to	the	allocation	of	sample	points	for	point	count	surveys	at	proposed	wind	
energy	projects,	migration	counts	typically	are	conducted	from	one	to	a	few	points	within	or	
adjacent	to	a	proposed	project	footprint.		Points	are	widely	spaced,	located	primarily	at	
places	that	collectively	provide	greatest	visual	coverage	especially	of	topographic	features	
likely	to	attract	or	funnel	migrating	raptors.		At	many	proposed	projects,	however,	survey	
points	for	migration	counts	could	be	the	same	as	or	a	subset	of	those	used	for	point	count	
surveys,		e.g.,	per	the	above	example	(under	1a.	Point	Count	Surveys),	such	that	migration	
counts	at	a	given	point	simultaneously	contribute	point	count	data.		Consideration	should	
be	given	to	restructuring	point	count	surveys	to	this	end,	including	temporal	stratification	
to	more	effectively	account	for	potential	eagle	migration	and	improve	precision	of	exposure	
estimates.		As	another	example,	during	an	anticipated	6‐week	peak	of	eagle	migration	in	fall,	
point	count	duration	could	be	extended	to	6	hours.		If	the	surveys	were	to	cover	either	the	
first	6	hours	or	the	last	6	hours	of	the	day,	the	two	survey	periods	would	overlap	by	several	
hours	in	midday,	better	covering	the	time	of	day	when	eagles	are	most	likely	moving	
(Heintzelman	1986).		The	data	may	have	to	be	adjusted	slightly	when	used	for	fatality	
estimation,	however.	
	
Strickland	et	al.	(2011)	summarize	some	important	details	for	conducting	raptor	migration	
counts	at	proposed	wind	energy	sites.		Counts	should	be	conducted	using	standard	
techniques	(Bildstein	et	al.	2007,	Dunn	et	al.	2008)	during	at	least	peak	periods	of	passage	
(see	the	Hawk	Migration	Association	of	North	America’s	[HMANA]	website	for	information	
on	seasonal	passage	periods	for	eagles	at	various	migration	survey	sites:	
http://www.hmana.org).		Migration	counts	may	involve	staffing	survey	points	up	to	75%	of	
days	during	peak	passage	(Dunn	et	al.	2008).		If	at	least	a	modest	eagle	migration	is	
evidenced	(i.e.,	multiple	individuals	observed	passing	unidirectionally	during	each	of	
multiple	days),	surveys	should	be	continued	for	at	least	2	years	and	into	the	operational	
phase	to	validate	initial	observations	and	help	assess	evidence	of	collision	and	influence	of	
turbines	on	migration	behavior.		Migration	count	data	should	be	provided	to	the	Service	as	
an	appendix	to	the	ECP,	using	a	reporting	format	similar	to	that	used	by	HMANA.		As	with	
point	count	surveys,	training	of	migration	survey	staff	should	include	assessment	of	raptor	
identification	skills	and	of	ability	of	individuals	to	detect	eagles	in	flight	under	a	broad	range	
of	distances	and	weather	conditions.	
	
Potential	for	non‐breeding	(either	winter	or	summer)	season	concentrations	of	eagles	in	or	
near	the	project	footprint	should	begin	to	be	evaluated	in	Stage	1,	including	close	scrutiny	of	
potential	habitat	via	geospatial	imagery	and	follow	up	reconnaissance	visits	(see	APPENDIX	
B).		Non‐breeding	bald	eagles	often	use	communal	roosts	and	forage	communally	(Platt	
1976,	Mojica	et	al.	2008).		Golden	eagles	may	do	so	on	occasion,	with	other	golden	eagles	
and/or	with	bald	eagles	(Craig	and	Craig	1984).		Both	species	can	become	concentrated	on	
spring	and	fall	migration	under	particular	combinations	of	weather	and	topographic	
conditions,	or	may	annually	use	traditional	stopover	sites	during	migration.		The	Stage	1	
assessment	may	suggests	that	seasonal	concentrations	of	eagles	regularly	occur	within	the	
project	area,	either	because	of	favorable	conditions	(e.g.,	clusters	of	large	trees	along	rivers	
offering	potential	roost	sites,	stopover	concentrations	of	migrating	waterfowl)	or	because	of	
indications	from	prior	anecdotal	or	systematically	collected	records.		The	Stage	2	
assessment	should	include	surveys	designed	to	further	explore	evidence	of	any	such	
occurrences.		If,	based	on	the	outcome	of	Stage	1,	there	is	no	compelling	reason	to	believe	
concentration	areas	are	lacking,	an	efficient	way	to	begin	to	probe	for	concentration	areas	is	
simply	to	extend	the	duration	of	point	count	surveys	and	perhaps	conduct	them	more	
frequently.		Expanded	point	count	surveys,	distributed	evenly	across	the	day	during	the	first	

62
year	of	Stage	2,	should	provide	at	least	a	preliminary	indication	of	regular	movements	to	
and	from	what	may	be	roosts	or	prey	hotspots	within	or	outside	the	project	footprint.		
Moreover,	expanded	point	count	surveys	conducted	near	potential	turbine	sites	(see	design	
recommendations	in	a.	Point	Count	Surveys,	above)	can	better	inform	turbine	siting	
decisions	in	relation	to	eagle	use	of	concentration	areas,	if	such	areas	exist.		The	increased	
survey	effort	also	could	contribute	towards	a	more	precise	indication	of	eagle	exposure	in	a	
fatality	estimate	for	the	proposed	project	(APPENDIX	D).	
	
Early	in	Stage	2,	evidence	from	Stage	1	of	concentration	areas	in	the	project	area	may	be	
corroborated	or	new	evidence	of	concentrations	may	surface.		In	either	case,	focused	
surveys	(e.g.,	via	direct	observation	or	by	aircraft)	can	be	implemented	to	document	their	
locations	and	daily	timing	and	spatial	patterns	of	their	use	by	eagles	in	relation	to	the	
proposed	project	footprint	throughout	the	season(s).		For	example,	surveys	for	wintering	
concentrations	of	bald	eagles	could	be	conducted,	following	USFWS	(1983)	guidance.		
Direct,	systematic	observation	from	vantage	points	in	early	morning	and	evening	is	the	
most	practical	means	of	documenting	roost	locations	and	movements	of	eagles	to	and	from	
roosts	on	a	local	scale	(Steenhof	et	al.	1980,	Crenshaw	and	McClelland	1989).		Aerial	
surveys	may	be	needed	for	repeated	surveys	of	eagles	at	extensive	roosts	(Chandler	et	al.	
1995).		Direct	observation	can	be	used	to	compare	occurrence	and	activity	of	eagles	before	
and	after	construction	and	operation	of	a	project	(Becker	2002)	and	may	be	a	valid	means	
to	identify	disturbance	effects	on	roosting	concentrations.	

	
c. Utilization Distribution (UD) Assessment
UD	can	be	thought	of	as	animal’s	spatial	distribution	or	intensity	of	use	of	various	parts	of	a	
given	area,	such	as	its	home	range.		A	basic	though	perhaps	labor‐intensive	approach	for	
documenting	spatial	distribution	of	use	across	all	or	part	of	a	proposed	project	footprint	by	
eagles	is	to	systematically	observe	and	record	eagle	movements	and	activities	(e.g.,	
territorial	display,	prey	delivery	flight)	on	maps	in	the	field	then	convert	the	data	into	GIS	
formats	for	standard	analyses	(e.g.,	Walker	et	al.	2005).		For	example,	a	grid	of	square	cells,	
each	0.5	x	0.5	km,	can	be	framed	by	the	Universal	Transverse	Mercator	(UTM)	system	
across	a	map	of	the	area	of	interest	to	record	eagle	observations	in	each	0.25	km2	cell.		The	
area	of	interest	is	divided	into	non‐overlapping	observation	sectors,	each	with	a	vantage	
point	that	affords	unobstructed	viewing	of	grid	cells	to	more	than	1	km	in	all	directions.		
Observation	periods	last	at	least	4	hours	and	include	all	daylight	hours	and	account	for	
roost	sites.		If	necessary,	two	(or	more)	observers	working	from	separate	vantage	points	
can	pinpoint	locations	of	eagles	through	triangulation.	
	
The	data	can	be	analyzed	by	simply	counting	the	number	of	flights	intersecting	each	cell.		An	
eagle’s	distribution	of	use	can	then	be	estimated	by	using	standard	kernel	analyses	(Worton	
1989,	1995,	Seaman	and	Powell	1996,	Kenward	2001)	or	other	probabilistic	approaches,	
comparable	to	Moorcroft	et	al.	(1999),	McGrady	et	al.	(2002),	and	McLeod	et	al.	(2002).		
Having	concern	over	potential	autocorrelation,	Walker	et	al.	(2005)	randomly	selected	
independent	locations	of	golden	eagles	along	flight	paths	to	establish	a	point	database	for	
standard	UD	analyses.		They	determined	that	locations	would	be	independent	if	separated	
by	at	least	45	minutes.		McGrady	et	al.	(2002)	conservatively	used	a	1‐hour	minimum	to	
separate	points,	even	though	their	data	indicated	a	20‐minute	interval	would	suffice.		
Concerns	with	autocorrelation	in	UD	analyses	have	recently	diminished,	however	(Feiberg	
et	al.	2010).		Most	study	of	eagle	UD	has	focused	on	resident	birds	especially	breeding	
adults	on	their	nesting	territories.		Size	and	shape	of	use	areas	can	vary	seasonally	(Newton	

63
1979),	so	documentation	of	spatial	use	by	resident	eagles	should		encompass	all	seasons	in	
addition	to	accounting	for	annual	variation.	
	
A	substantial	advantage	of	a	direct	observation	approach	compared	to	telemetry	
techniques,	which	typically	target	only	one	or	two	resident	eagles	at	a	proposed	project,	is	
that	it	disregards	age	and	breeding	and	residency	status.		Included	are	overwintering	
individuals;	dispersing	juveniles;	post‐fledging	young	from	nearby	territories	and	juveniles	
dispersing	from	other	areas	or	regions;	and	adults	from	adjoining	territories	plus	non‐
breeding	adults	(i.e.,	“floaters,”	Hunt	1998)	and	subadults	that	may	occur	along	boundaries	
of	breeding	territories.		In	many	instances,	identification	of	individual	eagles	may	not	be	
important	and	final	results	of	a	generalized	UD	analysis	may	be	based	on	data	pooled	from	
multiple	birds,	some	of	which	were	indistinguishable	from	each	other	in	the	field.		A	
disadvantage	of	this	approach	is	that	position	accuracy	based	on	direct	observation	across	
expansive	landscapes	is	coarse	compared	to	using	telemetry	with	GPS	capability,	and	
generally	declines	with	distance,	increasing	topographic	and	forest	cover,	and	during	early	
morning	and	late	evening	hours.		This	can	be	resolved	to	some	extent	by	limiting	the	size	
and	increasing	the	number	of	observation	sectors	(in	addition	to	using	multiple	observers),	
but	for	most	pre‐construction	information	needs,	a	high	degree	of	accuracy	is	unessential	
for	UD	data.		Last,	it	is	unlikely	that	UD	needs	to	be	assessed	across	entire	project	footprints.		
Instead,	it	is	more	likely	used	to	target	specific	areas	of	concern,	such	as	areas	where	eagles	
nest	or	frequently	forage,	and	to	refine	knowledge	of	use	of	particular	areas	to	better	inform	
turbine	siting	decisions.		The	method	obviously	has	little	utility	in	areas	of	low	eagle	
occurrence.	
	
Although	we	acknowledge	telemetry	offers	some	distinct	benefits	for	assessing	risks	and	
impacts	of	wind	projects,	use	of	the	method	for	eagles	has	other	drawbacks.		Specific	
individual	eagles	must	be	targeted	for	capture	and	not	all	eagles	using	a	given	project	
footprint	are	equally	likely	to	be	captured	or	provide	useful	data	(e.g.,	migrants	may	be	
readily	captured	but	leave	the	area	before	providing	much	data).		More	importantly,	
capturing	and	radio‐marking	eagles	can	have	negative	effects	on	behavior,	productivity,	and	
re‐use	of	nest	sites	(e.g.,	Marzluff	et	al.	1997,	Gregory	et	al.	2002),	and	recent	information	
suggests	a	negative	effect	in	some	cases	on	survival,	especially	of	golden	eagles	captured	as	
adults	and	released	with	large	(70‐	to	100‐g),	solar‐charged	transmitters	(USFWS,	
unpublished	information).		These	effects	must	be	better	understood	before	routine	use	of	
telemetry	techniques	can	be	recommended	as	components	of	wind‐facility	assessments.		
Until	then,	the	Service	discourages	the	use	of	telemetry	in	assessments	of	eagle	use	
associated	with	wind	energy	projects;	survey	approaches	suggested	herein	do	not	require	
telemetry.	

	
d. Summary
The	Service	encourages	development	of	cost‐effective	sampling	designs	that	simultaneously	
address	multiple	aspects	of	use	of	proposed	wind	energy	projects	by	eagles,	though	
emphasizes	that	high‐quality	point	count	data	to	support	fatality	rate	estimation	should	be	
considered	the	highest	priority.		In	many	cases,	the	sampling	framework	for	point	count	
surveys	likely	can	be	extended	to	reasonably	assess	migration	incidence,	UD,	and	other	
objectives.		Although	field‐based	data	that	directly	support	fatality	estimation	are	most	
important,	development	of	methods	for	addressing	other	objectives	is	encouraged,	such	as	
the	use	of	digital	trail	cameras	to	document	eagle	occurrence	at	carcass	stations.		
Regardless,	we	recommend	that	pre‐construction	surveys	at	proposed	wind	energy	sites	

64
encompass	a	minimum	of	2	years,	including	at	least	1	year	characterized	by	robust	
sampling	that	integrates	multiple	survey	types.	

	
2. Survey of the Project-area Nesting Population: Number and Locations of Occupied Nests of
Eagles
To	evaluate	project	siting	options	and	help	assess	potential	effects	of	wind	energy	projects	on	
breeding	eagles,	we	recommend	determining	locations	of	occupied	nests	of	eagles	within	the	
project	area	for	no	less	than	two	breeding	seasons	prior	to	construction.		The	primary	objective	of	a	
survey	of	the	project‐area	nesting	population	is	to	determine	the	number	and	locations	of	occupied	
nests	and	the	approximate	centers	of	occupied	nesting	territories	of	eagles	within	the	project	area.		
If	recent	(i.e.,	within	the	past	5	years)	data	are	available	on	spacing	of	occupied	eagle	nests	for	the	
project‐area	nesting	population,	the	data	can	be	used	to	delineate	an	appropriate	boundary	for	the	
project	area	as	described	in	APPENDIX	H.		Otherwise,	we	suggest	that	project	area	be	defined	as	the	
project	footprint	and	all	area	within	10	miles.	
	
In	this	ECPG	document	we	use	raptor	breeding	terminology	originally	proposed	by	Postupalsky	
(1974)	and	largely	followed	today	(Steenhof	and	Newton	2007).		An	occupied	nest	is	a	nest	
structure	at	which	any	of	the	following	is	observed:	(1)	an	adult	eagle	in	an	incubating	position,	(2)	
eggs,	(3)	nestlings	or	fledglings,	(4)	occurrence	of	a	pair	of	adult	eagles	(or,	sometimes	subadults,		
e.g.,	Steenhof	et	al.	[1983])	at	or	near	a	nest	through	at	least	the	time	incubation	normally	occurs,	
(5)	a	newly	constructed	or	refurbished	stick	nest	in	the	area	where	territorial	behavior	of	a	raptor	
had	been	observed	early	in	the	breeding	season,	or	(6)	“A	recently	repaired	nest	with	fresh	sticks	
(clean	breaks)	or	fresh	boughs	on	top,	and/or	droppings	and/or	molted	feathers	on	its	rim	or	
underneath”	(Postupalsky	1974).	
	
A	nest	that	is	not	occupied	is	termed	unoccupied.		An	occupied	nesting	territory	includes	one	
occupied	nest	and	may	include	alternate	nests,		i.e.,	any	of	several	other	nest	structures	within	the	
nesting	territory.		Sometimes	“active	nest”	is	used	to	encompass	occupied	nests	in	which	eggs	were	
laid	plus	those	at	which	no	eggs	were	laid.		Here,	as	elsewhere	in	the	ECPG	and	in	Postupalsky	
(1974),	an	active	nest	is	considered	one	in	which	an	egg	or	eggs	have	been	laid.		A	nest	that	is	active	
is	also,	by	default,	occupied.		A	nest	that	is	not	active	is	inactive,	and	there	is	a	regulatory	definition	
for	the	term	inactive	nest	(50	CFR	22.3.		Not	all	pairs	of	bald	eagles	and	golden	eagles	attempt	to	
nest	or	nest	successfully	every	year	(Buehler	2000,	Kochert	et	al.	2002),	and	nesting	territories	
where	pairs	are	present	but	do	not	attempt	to	nest	could	in	some	cases	be	misclassified	as	
unoccupied.		Accurate	comprehension	of	territory	distribution	and	determination	of	occupancy	
status	is	the	crux	of	determining	the	project‐area	nesting	population.	
	
The	project‐area	nesting	population	survey	should	include	all	potential	eagle	nesting	habitat	within	
the	project	area.		At	least	two	checks	via	aircraft	or	two	ground‐based	observations	are	
recommended	to	designate	a	nest	or	territory	as	unoccupied,	as	long	as	all	potential	nest	sites	and	
alternate	nests	are	visible	and	monitored	(i.e.,	alternate	nests	may	be	widely	separated	such	that	a	
full‐length,	ground‐based	observation	should	be	devoted	to	each).		Ground‐based	observations	
should	be	conducted	for	at	least	4	hours	each		(occupancy	may	be	verified	in	less	time),	aided	by	
spotting	scopes,	from	at	least	0.8	km	from	the	nest(s),	during	weather	conducive	to	eagle	activity	
and	good	visibility.		Surveys	of	occupancy	should	be	conducted	at	least	30	days	apart,	ideally	during	
the	normal	courtship	and	mid‐incubation	periods,	respectively.		Surveys	later	in	the	breeding	
season	are	likely	to	overlook	some	territorial	pairs	that	that	did	not	lay	eggs	or	failed	early	in	the	
nesting	season.		Timing	of	surveys	should	be	based	on	local	nesting	chronologies;	Service	staff	can	
provide	recommendations.		If	an	occupied	nest	or	a	pair	of	eagles	is	located,	the	territory	should	

65
continue	to	be	searched	for	alternate	nest	sites.		This	information	can	help	determine	the	relative	
value	of	individual	nests	to	a	territory	if	ever	there	are	applications	for	permits	to	take	inactive	
nests,	and	when	determining	whether	abandonment	of	a	particular	nest	may	result	in	loss	of	a	
territory.	
	
Use	of	aerial	surveys	followed	by	ground‐based	surveys	at	targeted	sites	can	be	an	ideal	approach	
to	determine	nest	and	territory	occupancy.		Helicopters	are	an	accepted	and	efficient	means	for	
inventory	of	extensive	areas	of	potential	nesting	habitat	for	eagles,	although	fixed‐wing	aircraft	can	
be	used	where	potential	nest	sites	are	widely	scattered	and	conspicuous.		Aerial	surveys	for	eagle	
nests	in	woodland	habitat	may	require	two	to	three	times	as	much	time	as	aerial	surveys	for	nests	
on	cliffs.		When	surveying	rugged	terrain	by	helicopter,	cliffs	should	be	approached	from	the	front,	
rather	than	flying	over	from	behind	or	suddenly	appearing	from	around	corners	or	buttresses.		
Inventories	by	helicopter	should	be	flown	at	slow	speeds,	about	30	to	40	knots.		All	potentially	
suitable	nest	sites	should	be	scrutinized;	multiple	passes	at	several	elevation	bands	may	be	
necessary	to	provide	complete	coverage	of	nest	site	habitat	on	large	cliff	complexes.		Hovering	for	
up	to	15	seconds	no	closer	than	50	m	from	a	nest	may	be	necessary	to	verify	the	nesting	species,	
photograph	the	nest	site,	and,	if	late	in	the	nesting	season,	allow	the	observer	to	count	and	estimate	
age	of	young	in	the	nest.		Aerial	surveys	may	not	be	appropriate	in	some	areas	such	as	bighorn	
sheep	lambing	areas;	to	avoid	such	sensitive	areas,	state	resource	agencies	should	be	consulted	
when	planning	surveys.		Additional	guidelines	for	aerial	surveys	for	eagles	and	other	raptors	are	
reviewed	in	Anderson	(2007).	
	
Surveys	should	be	conducted	only	by	biologists	with	extensive	experience	in	surveys	of	raptors	and	
appropriate	training	in	aerial	surveys	(see	review	in	Anderson	2007).		Whether	inventories	are	
conducted	on	the	ground	or	aerially,	metrics	of	primary	interest	to	the	Service	for	the	project‐area	
nesting	population	include:	
	
1. number	and	locations	of	nest	structures	that	are	verified	or	likely	to	be	eagle	nests	
2. number	and	locations	of	eagle	nests	currently	or	recently	occupied	based	on	criteria	
outlined	herein	
3. estimated	number	and	approximate	boundaries	and	centers	of	eagle	breeding	territories,	
based	on	records	of	nest	site	occupancy	and	clustering	of	nests.	
	
Additionally,	productivity	(i.e.,	reproductive	success,	defined	here	as	the	mean	number	of	nestlings	
surviving	to	>	56	and	≥	67	days	of	age	per	occupied	nest	for	golden	eagles	and	bald	eagles,	
respectively)	may	be	of	interest	for	assessing	disturbance	effects,	although	utility	of	productivity	
data	at	a	given	project	likely	will	be	limited	due	to	small	sample	size	and	factors	confounding	the	
interpretation	of	results.		A	meta‐analysis	approach	based	on	productivity	data	from	many	projects	
is	contemplated	as	part	of	the	adaptive	management	process	accompanying	the	ECPG,	and	may	
contribute	to	understanding	of	disturbance	effects	on	this	aspect	of	eagle	breeding	biology.		
Moreover,	abandonment	of	territories	–	the	gravest	manifestation	and	clearest	evidence	of	
disturbance	effects	–	could	be	documented	through	the	occupancy	surveys	recommended	herein,	if	
these	surveys	are	repeated	after	project	construction.		We	reiterate	that	accurate	comprehension	of	
territory	distribution	and	determination	of	occupancy	status	should	be	the	primary	goal	of	nesting	
surveys.	

	
	

66

Literature Cited
Anderson,	D.	E.	2007.	Survey	techniques.	Pages	89‐100	in	D.	M.	Bird	and	K.	L.	Bildstein	(eds.),	
Raptor	research	and	management	techniques.	Hancock	House,	Blaine,	Washington.	
Becker,	J.	M.		2002.		Response	of	wintering	bald	eagles	to	industrial	construction	in	southeastern	
Washington.	Wildlife	Society	Bulletin	30:875‐878.	
Bildstein,	K.	L.	2006.	Migrating	raptors	of	the	world,	their	ecology	and	conservation.	Cornell	
University	Press,	Ithaca,	New	York.	
Bildstein,	K.	L.,	J.	P.	Smith,	and	R.	Yosef.	2007.	Migration	counts	and	monitoring.	Pages	102‐115	in	D.	
M.	Bird	and	K.	L.	Bildstein	(eds.),	Raptor	research	and	management	techniques.	Hancock	House,	
Blaine,	Washington.	
Buckland,	S.	T.,	D.	R.	Anderson,	K.	P.	Burnham,	J.	L.	Laake,	D.	L.	Borchers,	and	L.	J.	Thomas.	2001.	An	
introduction	to	distance	sampling:	estimating	abundance	of	biological	populations.	Oxford	
University	Press,	Oxford,	United	Kingdom.	
Buehler,	D.	A.	2000.	Bald	Eagle	Haliaeetus	leucocephalus.	No.	506	in	A.	Poole	and	F.	Gill	(eds.),	The	
Birds	of	North	America.	The	Birds	of	North	America,	Inc.,	Philadelphia,	Pennsylvania.	
Chandler,	S.	K.,	J.	D.	Fraser,	D.	A.	Buehler,	and	J.	K.	D.	Seegar.		1995.		Perch	trees	and	shoreline	
development	as	predictors	of	bald	eagle	distribution	on	Chesapeake	Bay.	Journal	of	Wildlife	
Management	59:325‐332.	
Craig,	T.	H.,	and	E.	H.	Craig.		1984.		A	large	concentration	of	roosting	golden	eagles	in	southeastern	
Idaho.		Auk	101:610‐613.	
Crenshaw,	J.	G.,	and	B.	R.	McClelland.		1989.		Bald	eagle	use	of	a	communal	roost.	Wilson	Bulletin	
101:626‐633.	
Denholm,	P.,	M.	Hand,	M.	Jackson,	and	S.	Ong.	2009.	Land‐use	requirements	of	modern	wind	power	
plants	in	the	United	States.	National	Renewable	Energy	Laboratory	Technical	Report	NREL/TP‐
6A2‐45834.	www.nrel.gov/docs/fy09osti/45834.pdf	(last	visited	October	22,	2011)	
Dunn,	E.	H.,	D.	J.	T.	Hussell,	and	E.	R.	Inzunza.		2008.		Recommended	methods	for	population	
monitoring	at	raptor‐migration	watch	sites.	Pages	447‐459	in	K.	L.	Bildstein,	J.	P.	Smith,	E.	R.	
Inzunza	and	R.	R.	Veit	(eds.).		State	of	North	America’s	birds	of	prey.		Series	in	Ornithology	No.	3.	
Nuttall	Ornithological	Club.	Cambridge,	Massachusetts	and	American	Ornithologists'	Union,	
Washington,	D.C.	
Fieberg,	J.,	J.	Matthiopoulos,	M.	Hebblewhite,	M.	S.	Boyce,	and	J.	L.	Frair.		2010.		Correlation	and	
studies	of	habitat	selection:	problem,	red	herring,	or	opportunity?		Philosophical	Transactions	
of	the	Royal	Society	Biological	Sciences	365:	2233‐2244.		doi:	10.1098/rstb.2010.0079.	
Garvin,	J.	C.,	C.	S.	Jennelle,	D.	Drake,	and	S.	M.	Grodsky.	2011.	Response	of	raptors	to	a	windfarm.	
Journal	of	Applied	Ecology	48:199‐209.	
Heintzelman,	D.	S.	1986.	The	migrations	of	hawks.	Indiana	University	Press,	Bloomington	and	
Indianapolis,	Indiana.	
Hutto,	R.	L.,	S.	M.	Pletschet,	and	P.	Henricks.	1986.	A	fixed‐radius	point	count	for	nonbreeding	and	
breeding	season	use.	Auk	103:593‐602.	
Johnson,	D.	H.		1995.	Point	counts	of	birds:	what	are	we	estimating?	Pages	117‐123	in	C.	J.	Ralph	and	
J.	R.	Sauer	(eds.),	Monitoring	bird	populations	by	point	counts.	General	Technical	Report	PSW‐
GTR‐149,	U.S.	Forest	Service,	Pacific	Southwest	Research	Station,	Albany,	California.	
Kerlinger,	P.	1989.	Flight	strategies	of	migrating	hawks.	University	of	Chicago	Press,	Chicago,	
Illinois.	
Kochert,	M.	N.,	K.	Steenhof,	C.	L.	McIntyre,	and	E.	H.	Craig.	2002.	Golden	Eagle	Aquila	chrysaetos.	No.	
684	in	A.	Poole	and	F.	Gill	(eds.),	The	Birds	of	North	America.	The	Birds	of	North	America,	Inc.,	
Philadelphia,	Pennsylvania.	
Madders,	M.,	and	D.	P.	Whitfield.	2006.	Upland	raptors	and	the	assessment	of	wind	farm	impacts.	
Ibis	148:43‐56.	

67
McGrady,	M.	J.,	J.	R.	Grant,	I.	P.	Bainbridge,	and	D.	R.	A.	McLeod.		2002.		A	model	of	golden	eagle	
(Aquila	chrysaetos)	ranging	behavior.	Journal	of	Raptor	Research	36(1)	supplement:62‐69.	
Mojica,	E.	K.,	J.	M.	Meyers,	B.	A.	Millsap,	and	K.	T.	Haley.		2008.		Migration	of	sub‐adult	Florida	bald	
eagles.		Wilson	Journal	of	Ornithology	120:304‐310.	
Nygård,	T,	K.,	E.	L.	Bevanger,	Ø.	Dahl,	A.	Flagstad,	P.	L.	Follestad,	Hoel,	R.	May,	and	O.	Reitan.	2010.	A	
study	of	white‐tailed	eagle	Haliaeetus	albicilla	movements	and	mortality	at	a	wind	farm	in	
Norway.	in	D.	Senapathi	(ed.),	Climate	change	and	birds:	adaptation,	mitigation	and	impacts	on	
avian	populations.	British	Ornithologists’	Union	Proceedings,	Peterborough,	United	Kingdom.		
http://www.bou.org.uk/bouproc‐net/ccb/nygard‐etal.pdf	(last	visited	October	9,	2011).	
Platt,	J.	B.	1976.	Bald	eagles	wintering	in	a	Utah	desert.	American	Birds	30:783‐788.	
Potupalsky,	S.	1974.	Raptor	reproductive	success:	some	problems	with	methods,	criteria,	and	
terminology.	Raptor	Research	Report	2:21‐31.	
Ralph,	C.	J.,	G.	R.	Geupel,	P.	Pyle,	T.	E.	Martin,	and	D.	F.	DeSante.	1993.	Handbook	of	field	methods	for	
monitoring	landbirds.	General	Technical	Report	PSW‐GTR‐144,	U.S.	Forest	Service,	Pacific	
Southwest	Research	Station,	Albany,	California.	
Reynolds,	R.	T.,	J.	M.	Scott,	and	R.	A.	Nussbaum.	1980.	A	variable	circular‐plot	method	for	estimating	
bird	numbers.	Condor	82:309‐313.	
Sauer,	J.	R.,	J.	E.	Hines,	J.	E.	Fallon,	K.	L.	Pardieck,	D.	J.	Ziolkowski,	Jr.,	and	W.	A.	Link.	2011.	The	North	
American	Breeding	Bird	Survey,	results	and	analysis	1966	‐	2009.	Version	3.23.2011.	U.S.	
Geological	Survey,	Patuxent	Wildlife	Research	Center,	Laurel,	Maryland,	USA.	
Steenhof,	K.,	S.	S.	Berlinger,	and	L.	H.	Fredrickson.		1980.		Habitat	use	by	wintering	bald	eagles	in	
South	Dakota.	Journal	of	Wildlife	Management	44:798‐805.	
Steenhof,	K.,	M.	N.	Kochert,	and	J.	H.	Doremus.	1983.	Nesting	of	subadult	golden	eagles	in	
southwestern	Idaho.	Auk	100:743‐747.	
Steenhof,	K.,	and	I.	Newton.		2007.		Assessing	nesting	success	and	productivity.	Pages	181‐191	in	D.	
M.	Bird	and	K.	Bildstein	(eds.),	Raptor	Research	and	Management	Techniques.	Hancock	House,	
Blaine,	Washington,	USA.	
Strickland,	M.	D.,	E.	B.	Arnett,	W.	P.	Erickson,	D.	H.	Johnson,	G.	D.	Johnson,	M.	L.	Morrison,	J.	A.	
Shaffer,	and	W.	Warren‐Hicks.	2011.	Comprehensive	guide	to	studying	wind	energy/wildlife	
interactions.	Prepared	for	the	National	Wind	Coordinating	Collaborative,	Washington,	D.C.	
USFWS.		1983.		Northern	states	bald	eagle	recovery	plan.	U.S.	Fish	and	Wildlife	Service,	Division	of	
Migratory	Bird	Management,	Washington	D.C.	http://www.fws.gov/midwest/eagle/recovery/	
be_n_recplan.pdf	(last	visited	October	9,	2011).	
Walker,	D.,	M.	McGrady,	A.	McCluskie,	M.	Madders,	and	D.	R.	A.	McLeod.		2005.		Resident	golden	
eagle	ranging	behaviour	before	and	after	construction	of	a	windfarm	in	Argyll		Scottish	Birds	
25:24‐40.

68

APPENDIX D: STAGE 3 – PREDICTING EAGLE FATALITIES
	
The	Service	uses	a	Bayesian	method	(see	Gelman	et	al.	2003)	to	predict	the	annual	fatality	rate	for	a	
wind‐energy	facility,	using	explicit	models	to	define	the	relationship	between	eagle	exposure	
(resulting	from	the	Stage	2	assessment,	APPENDIX	C),	collision	probability,	and	fatalities	(verified	
during	post‐construction	monitoring	in	Stage	5,	APPENDIX	H),	and	to	account	for	uncertainty.	The	
relationships	between	eagle	abundance,	fatalities,	and	their	interactions	with	factors	influencing	
collision	probability	are	still	poorly	understood	and	appear	to	vary	widely	depending	on	multiple	
site‐specific	factors	(see	Assessing	Risk	and	Effects;	2.		Eagle	Risk	Factors	in	the	ECPG).	The	baseline	
model	presented	below	is	a	foundation	for	modeling	fatality	predictions	from	eagle	exposure	to	
wind	turbine	hazards.	In	addition	to	generating	the	fatality	estimate	that	will	be	a	component	of	the	
Service's	analysis	of	the	permit	application,	the	model	also	serves	as	a	basis	for	learning	and	the	
exploration	of	other	candidate	models	that	attempt	to	better	incorporate	specific	factors	and	
complexity.	The	Service	encourages	project	developers	or	operators	to	develop	additional	
candidate	models	(both	a	priori	and	post	hoc)	for	direct	comparison	with,	and	evaluation	of,	the	
baseline	model	and	modeling	approach.	Our	ability	to	learn	over	time	and	reduce	uncertainty	by	
incorporating	new	information	into	our	modeling	approach	through	an	adaptive	management	
framework	(see	APPENDIX	A)	enables	us	to	improve	site‐specific	estimation	of	eagle	fatalities,	
reduce	uncertainty	in	predictions,	and,	ultimately,	improve	management	decisions	relating	to	
eagles	and	wind	energy	in	a	responsible	and	informed	way.	Rigorous	post‐construction	monitoring	
is	a	critical	component	of	evaluating	model	performance	over	time	(see	APPENDIX	H).	
	
Variables	used	in	the	formulas	below	are	summarized	in	Table	D‐1	for	ease	of	reference.		The	total	
annual	eagle	fatalities	(F)	as	the	result	of	collisions	with	wind	turbines	can	be	represented	as	the	
product	of	the	rate	of	eagle	exposure	(λ)	to	turbine	hazards,	the	probability	that	eagle	exposure	will	
result	in	a	collision	with	a	turbine	(C),	and	an	expansion	factor	(ε)	that	scales	the	resulting	fatality	
rate	to	the	parameter	of	interest,	the	annual	predicted	fatalities	for	the	project:	
	
.	
	
Using	the	Bayesian	estimation	framework,	we	define	prior	distributions	for	exposure	rate	and	
collision	probability;	the	expansion	factor	is	a	constant	and	therefore	does	not	require	a	prior	
distribution.	Next,	we	calculate	the	exposure	posterior	distribution	from	its	prior	distribution	and	
observed	data.	The	expanded	product	of	the	posterior	exposure	distribution	and	collision	
probability	prior	yields	the	predicted	annual	fatalities.	

	
	

	

69

Table D-1. Abbreviations and descriptions of variables used in the Service method for predicting annual eagle
fatalities.
Abbreviation	

Variable	

Description	

F	

Annual	fatalities	

Annual	eagle	fatalities	from	turbine	collisions	

λ	

Exposure	rate	

Eagle‐minutes	flying	below	200	m	in	height	within	the	project	
footprint	(in	proximity	to	turbine	hazards)	per	hr	per	km2	

C	

Collision	
probability	

The	probability	of	an	eagle	colliding	with	a	turbine	given	exposure	

ε	

Expansion	factor	

Product	of	daylight	hours	and	total	hazardous	area	(hr∙km2)	

k	

Eagle‐minutes	

Number	of	minutes	that	eagles	were	observed	flying	below	200	m	
during	survey	counts	

δ	

Turbine	
hazardous	area	

Rotor‐swept	area	around	a	turbine	or	proposed	turbine	from	0	to	200	
m	(km2)	

n	

Trials	

Number	of	trials	for	which	events	could	have	been	observed	(the	
number	of	hr∙km2	observed)	

τ	

Daylight	hours	

Total	daylight	hours	(e.g.	4383	hr	per	year)	

nt	

Number	of	
turbines	

Number	of	turbines	(or	proposed	turbines)	for	the	project	

	

	
	
	
1. Exposure	
The	exposure	rate	λ	is	the	expected	number	of	exposure	events	(eagle‐minutes)	per	daylight	hour	
per	square	kilometer	(hr∙	km2).		We	defined	the	prior	distribution	for	exposure	rate	based	on	
information	from	a	range	of	projects	under	Service	review	and	others	described	with	sufficient	
detail	in	Whitfield	(2009).	The	exposure	prior	predicts	an	exposure	rate	from	a	mixture	distribution	
of	project‐specific	Gamma	distributions	(Figure	D‐1).	We	used	the	Gamma	distribution	because	all	
values	are	positive	and	real	(see	Gelman	et	al.,	1995,	p.	474–475).	The	mixture	distribution	is	
summarized	by	a	new	Gamma	distribution	(our	prior	distribution	for	exposure)	with	a	mean	
(0.352)	and	standard	deviation	(0.357)	derived	from	the	conditional	distributions	(Gelman	et	al,	
1995,	equation	1.7	p.	20).	The	resulting	prior	distribution	for	exposure	rate	is:
	
	 	~	

∝,

,	with	shape	and	rate	parameters	of	α	=	0.97	and	β	=	2.76.	

	
Simulation	trials	produced	consistent	results.	The	prior	distribution	is	meant	to	include	the	range	of	
possible	exposure	rates	for	any	project	considered.	
	
	

	

70

	

	

Figure D-1. The prior probability distribution Gamma (0.97, 2.76), for exposure rate, λ, with a mean of 0.352
(indicated by the reference line) and standard deviation of 0.357. The distribution is positively skewed such
that exposure is generally at or near 0 with fewer higher values shown by the black curve. The project-specific
distributions (gray curves) were used to determine the mixture distribution (dashed curve) which determined the prior
distribution parameters.

	
	
	

Eagle	exposure	data	collected	during	the	pre‐construction	phase	surveys	(see	APPENDIX	C)	can	be	
used	to	update	this	prior	and	determine	the	posterior	distribution	that	will	be	used	to	estimate	the	
predicted	fatalities.		The	Service	may	also	be	able	to	work	with	a	project	developer	or	operator	on	a	
case‐by‐case	basis	to	use	the	prior	λ	distribution	to	generate	a	risk‐averse	fatality	prediction	for	
projects	where	no	pre‐construction	survey	data	are	available.		Assuming	the	observed	exposure	
minutes	follow	a	Poisson	distribution	with	rate	λ,	the	resulting	posterior	λ	distribution	is:	
	
,
.	
	 	~	
∝ ∑
	
The	new	posterior	λ	parameters	are	the	sum	of	α	from	the	prior	and	the	events	observed	(eagle	
minutes,	ki),	and	the	sum	of	β	from	the	prior	and	the	number	of	trials,	n,	for	which	events	could	
have	been	observed	(the	number	of	“trials”	is	the	number	of	hr∙km2	that	were	observed).	Note	that	
by	including	realistic	time	and	area	data	from	the	pre‐construction	surveys,	the	relative	influence	of	
the	prior	λ	distribution	on	the	resulting	posterior	λ	distribution	for	exposure	rate	becomes	
negligible.	In	other	words,	with	adequate	sampling,	the	data	will	determine	the	posterior	
distribution,	not	the	prior.	The	posterior	λ	distribution	can	then	be	used	to	estimate	the	annual	
fatality	distribution.	
	

71
In	addition,	this	posterior	λ	distribution	can	now	serve	as	a	prior	distribution	for	the	next	iteration	
of	the	predictive	model	in	an	adaptive	framework	(see	APPENDIX	A),	at	least	for	the	project	under	
consideration	and	potentially	in	a	more	general	way	as	the	posteriors	from	multiple	sites	are	
considered;	in	this	way,	we	build	ongoing	information	directly	into	the	predictive	process.	

	
2. Collision Probability
Collision	probability	C	is	the	probability,	given	exposure	(1	minute	of	flight	in	the	hazardous	area,	
),	of	an	eagle	colliding	with	a	turbine;	for	the	purposes	of	the	model,	all	collisions	are	considered	
fatal.		We	based	the	prior	distribution	on	a	Whitfield	(2009)	study	of	avoidance	rates	from	four	
independent	sites.		Averaging	avoidance	from	those	sites	yielded	a	mean	and	standard	deviation	for	
collision	probability	of	0.0058,	0.0038,	respectively	(note	this	is	consistent	with	eagle	avoidance	
rates	in	other	risk	assessment	approaches,	e.g.	99%).		This	in	turn	defined	the	prior	C	distribution	
as:	
	
	 	~	
, ´ ,	with	parameters	ν	and	ν´	of	2.31	and	396.69	(Figure	D‐2).	
	
The	Beta	distribution	is	used	to	describe	values	between	0	and	1	(Gelman	et	al.,1995,	p.	476–477).		
The	prior	C	distribution	attempts	to	include	the	range	of	possible	collision	probabilities	across	the	
set	of	potential	sites	to	be	considered.	

	
	

60
0

20

40

Density

80

100

120

Collision Probability Prior

0.000

0.005

0.010

0.015

Pr(Collision|Exposure Minute)

0.020

	

Figure D-2. The probability distribution for the collision probability prior, a Beta(2.31, 396.69) distribution
with a mean of 0.0058 (indicated by the reference line) and a standard deviation of 0.0038. The distribution
is positively skewed such that most collision probabilities will be small.

	
	

At	the	time	of	pre‐construction	permitting,	the	prior	C	distribution	will	be	used	to	estimate	the	
annual	predicted	fatalities.		After	construction,	post‐construction	monitoring	can	be	used	to	
determine	the	posterior	C	distribution	by	updating	the	prior	C	distribution.	

72
	
Assuming	the	observations	of	fatalities	follow	a	binomial	distribution	with	rate	C,	the	posterior	
distribution	of	the	rate	C	will	be	a	beta	distribution	(the	beta	distribution	and	the	binomial	
distribution	are	a	conjugate	pair):	
	
	 	~	
, ´
,	
	
where	f	is	the	number	of	fatalities	estimated	from	the	Stage	5	post‐construction	monitoring,	and	g	is	
the	estimated	number	of	exposure	events	that	did	not	result	in	a	fatality.		The	posterior	distribution	
for	C	cannot	be	calculated	until	a	project	has	been	built,	has	started	operations,	and	at	least	one	
season	of	post‐construction	monitoring	has	been	completed.		Once	determined,	the	posterior	C	
distribution	can	then	be	used	to	generate	a	prediction	for	annual	fatalities	and	can	serve	as	a	prior	C	
for	the	next	iteration	of	the	predictive	model	(see	APPENDIX	A).	

	
3. Expansion
The	expansion	factor	(ε)	scales	the	resulting	per	unit	fatality	rate	(fatalities	per	hr	per	km2)	to	the	
daylight	hours,	τ,	in	1	year	(or	other	time	period	if	calculating	and	combining	fatalities	for	seasons	
or	stratified	areas)	and	total	hazardous	area	(km2)	within	the	project	footprint:	
	
∑
,	
	
where	nt	is	the	number	of	turbines,	and	δ	is	the	circular	area	centered	at	the	base	of	a	turbine	with	a	
radius	equal	to	the	rotor‐swept	radius	of	the	turbine;	we	define	this	as	the	hazardous	area	
surrounding	a	turbine.	In	this	model,	to	simplify	data	requirements	and	assumptions,	we	consider	
both	eagle	use	and	hazardous	area	as	2‐dimensional	areas,	since	the	height	of	the	sampled	and	
hazardous	areas	are	the	same	(200	m)	and	will	cancel	out	in	the	calculations.	Alternative	models	
that	consider	3‐dimensional	space	could	also	be	considered,	though	the	expansion	factor	should	be	
adjusted	accordingly.		The	units	for	ε	are	hr∙	km2	per	year	(or	time	period	of	interest).	

	
4. Fatalities
Now	we	can	generate	the	distribution	of	predicted	annual	fatalities	as	the	expanded	product	of	the	
posterior	exposure	rate	and	the	prior	collision	probability	(once	post‐construction	data	is	available,	
the	posterior	collision	probability	would	be	used	to	update	our	fatality	distribution):	
	
∙
	 ∙
	 .	
	
We	can	then	determine	the	mean,	median,	standard	deviation,	and	80%	quantile	(this	will	be	the	
upper	credible	limit)	directly	from	the	distribution	of	predicted	fatalities.	

	
5. Putting it all together: an example
The	Patuxent	Power	Company	example	below	illustrates	the	calculation	of	predicted	fatalities	from	
exposure	data	from	a	hypothetical	project	site.		This	data	will	normally	come	from	the	field	surveys	
in	Stage	2,	but	for	the	purposes	of	this	example,	we	have	generated	fabricated	observation	data.		
The	advantage	of	simulating	data	in	such	an	exercise	is	that	we	can	manipulate	model	inputs	to	
critically	evaluate	the	performance	of	the	model.	Additional	examples	are	provided	at	the	end	of	
this	document	to	illustrate	the	general	approach	and	clarify	specific	considerations	that	may	apply	
to	certain	projects.	

	

73

a. Patuxent Power Company Example		
Patuxent	Power	Company	conducted	surveys	for	eagles	at	a	proposed	location	for	a	small‐	
to	medium‐sized	wind	facility	(18	turbines,	each	with	a	50	meter	rotor	diameter)	following	
the	recommended	methods	in	the	ECPG	(see	Table	D‐2).		They	conducted	168	counts	at	7	
points	and	60	eagle‐min	of	exposure	were	observed.		Each	count	was	2‐hr	in	duration,	and	
covered	a	circular	area	of	radius	0.8	km.		Thus,	675.6	km2∙hr	were	observed	in	total.	

	
	
	
Table D-2. Exposure data for Patuxent Power Company example. In this hypothetical example, 168 counts
were performed. Each count was 2-hr in duration and covered a 0.8 km radius circle. Thus, the total time and area
2
sampled was 675.6 km ·hr. In that time, 60 exposure events (eagle-min) were observed.
Visit	
1	
2	
3	
4	
5	
6	
7	
8	
9	
10	
11	
12	
13	
14	
15	
16	
17	
18	
19	
20	
21	
22	
23	
24	
Total	

	

P1	
0	
0	
0	
0	
0	
0	
0	
0	
0	
0	
1	
0	
0	
2	
0	
0	
0	
1	
0	
0	
0	
1	
1	
0	
6	

P2
0	
0	
1	
1	
1	
0	
1	
0	
0	
0	
0	
1	
0	
0	
0	
0	
0	
0	
0	
0	
0	
0	
0	
0	
5	

P3
2	
1	
2	
0	
0	
1	
0	
0	
0	
0	
1	
0	
1	
0	
0	
0	
0	
1	
0	
2	
0	
0	
0	
0	
11	

P4
0	
0	
0	
0	
1	
1	
0	
0	
0	
0	
1	
0	
0	
0	
2	
1	
2	
1	
1	
0	
0	
0	
3	
0	
13	

P5
2	
0	
0	
0	
0	
0	
0	
0	
0	
0	
0	
1	
0	
0	
2	
0	
0	
0	
0	
1	
1	
0	
0	
0	
7	

P6
0	
0	
0	
1	
1	
0	
1	
1	
0	
0	
0	
0	
0	
0	
0	
0	
0	
0	
2	
0	
0	
0	
0	
0	
6	

P7
1	
1	
1	
1	
1	
1	
1	
0	
0	
0	
0	
0	
1	
2	
1	
0	
0	
0	
0	
0	
0	
1	
0	
0	
12	

Total	
5	
2	
4	
3	
4	
3	
3	
1	
0	
0	
3	
2	
2	
4	
5	
1	
2	
3	
3	
3	
1	
2	
4	
0	
60	

74

b. Exposure	
The	posterior	distribution	for	the	exposure	rate	is:	
	
	 	~	
∝, ,	remember,	
	 	~	
0.97, 2.76 ,	Figure	D1;	where,	
	
0.97

	

60	

	

60.97	

	

	

678.31	

∙

	
2.76

168	

2	

0.8	

	

	
Thus,	
	
	 	~	

60.97, 678.31 ;	the	units	for	λ	are	per	hr	per	km2.	

	
The	posterior	distribution	is	shown	in	Figure	D‐3.		The	mean	and	standard	deviation	of	
exposure	rate	are	0.09	and	0.01,	respectively.		Note	that	there	is	little	influence	of	the	prior	
on	this	posterior,	because	the	sampling	effort	was	substantial.	

	
	

	
Figure D-3. The posterior distribution for exposure rate for the example project, “Patuxent Power
Company.” This gamma distribution has a mean (indicated by the reference line) of 0.09 and a standard deviation of
0.01.

75

b. Collision Probability
We	do	not	have	any	additional	information	about	collision	probability,	C,	so	we	will	use	the	
prior	distribution,	which	has	a	mean	of	0.0058	and	a	standard	deviation	of	0.0038:	
	
	 	~	
2.31,396.69 ;	see	Figure	D‐2.	

	
c. Expansion
The	expansion	rate,	ε,	is	the	number	of	daylight	hours	in	a	year	(τ)	multiplied	by	the	
hazardous	area	(δ)	around	the	18	turbines	proposed	for	the	project:	
	
∙ 18	 154.9	 ∙
.	
4,383	 ∙ 0.025	

	
d. Fatalities
To	determine	the	distribution	for	the	predicted	annual	fatalities,	the	exposure	and	collision	
risk	distributions	need	to	be	multiplied	by	each	other	and	expanded.		The	resulting	
distribution	cannot	be	calculated	in	closed	form;	it	is	easiest	to	generate	it	through	
simulations.		In	this	example,	after	running	100,000	simulations,	the	predicted	distribution	
for	annual	fatalities	(Figure	D‐4)	has	a	mean	of	0.082	and	a	standard	deviation	of	0.055.		
The	80%	quantile	is	0.12	eagle	fatalities	per	year.	
	

	

0

2

4

Density

6

8

10

Predicted Annual Fatalities

0.00

0.05

0.10

0.15
Fatality Rate

0.20

0.25

0.30

	

Figure D-4. The probability distribution for predicted annual fatalities. The histogram shows the simulation
results. The mean (0.082) and 80% quantile (0.12) are represented by the reference lines (black and gray,
respectively). The standard deviation is 0.055.

	
	

76

	

The	Service’s	baseline	model	for	the	proposed	Patuxent	wind	facility	predicts	that	80%	of	
the	time	that	annual	fatalities	would	be	0.12	eagles	or	fewer,	suggesting	that	an	eagle	
collision	fatality	would	be	predicted	to	occur	at	the	project	site	every	8‐9	years	on	average.	
The	facility	had	a	medium	amount	of	eagle	activity	at	the	site,	but	the	small	size	of	the	
project	kept	the	predicted	fatality	numbers	lower	than	they	would	have	been	for	a	larger	
project	in	the	same	location.	Ideally,	we	would	consider	other	candidate	models	alongside	
the	baseline	model	presented	here	and	compare	their	relative	performance	using	data	
collected	in	Stage	5.	

	
6. Additional Considerations
This	initial	estimate	of	fatality	rate	should	not	take	into	account	possible	conservation	measures	
and	ACPs	(e.g.	changes	in	turbine	siting	or	seasonal	curtailments);	these	will	be	factored	in	as	part	
of	Stage	4	(APPENDIX	E).		Additionally,	any	loss	of	production	that	may	stem	from	disturbance	is	
not	considered	in	these	calculations,	but	should	be	added	to	these	estimates	and	later	adjusted	
based	on	post‐construction	monitoring	as	described	in	Stage	5.		This	stage	and	Stage	5	of	the	ECP	
will	require	close	coordination	between	the	project	developer	or	operator	and	the	Service.	

	
a. Small-scale Projects
Small‐scale	projects	(generally	these	will	be	residential	or	small‐business	projects)	may	
pose	a	low	enough	risk	that	Stage	2	surveys	are	unnecessary	to	demonstrate	that	the	
project	is	not	likely	not	take	eagles.	This	presumes	that	Stage	1	surveys	are	conducted	and	
show	no	important	eagle	use	areas	or	migration	concentration	sites	in	the	project	area.		In	
such	cases,	the	fatalities	predicted	by	the	collision	fatality	model	are	the	expanded	product	
of	the	exposure	prior	and	the	collision	probability	prior;	the	exposure	prior	is	not	updated	
to	create	a	posterior	as	it	would	be	for	projects	with	survey	data	(Figure	D‐5).	With	the	
prior	distributions	currently	used	for	exposure	rate	and	collision	probability	(note	that	the	
parameters	for	the	priors	distributions	are	part	of	the	adaptive	management	framework	
and	will	change	as	new	information	becomes	available),	the	80	percent	quantile	of	the	
predicted	fatality	distribution	for	projects	with	less	than	approximately	2.4x10‐3	km2	of	
hazardous	area	predicts	fatalities	at	a	rate	less	than	1	eagle	in	30	years	(not	likely	to	take	
eagles).	This	is	equivalent	to	a	single	turbine	with	a	rotor	diameter	of	approximately	55	m,	
or	more	than	45	turbines	with	8	m	rotor	diameter	(each	of	which	has	the	capacity	to	exceed	
typical	home	energy	needs).		The	calculation	of	hazardous	area	is	presented	in	this	
Appendix	under	‘Expansion’.	If	the	collision	model	prediction	based	on	the	exposure	prior	
predicts	that	take	of	eagles	will	occur	(e.g.,	if	the	hazardous	area	is	greater	than	2.4x10‐3	
km2),	Stage	2	preconstruction	sampling	for	eagle	use	of	the	project	area	is	recommended	
(see	APPENDIX	C).	The	data	from	Stage	2	surveys	will	be	used	to	update	the	exposure	prior	
distribution	and	produce	a	project‐specific	fatality	prediction.	Projects	are	encouraged	to	
consult	with	the	Service	early	in	the	planning	process	as	components	of	the	fatality	
prediction	model	will	continue	to	evolve	and	may	change	over	time.	

	

	

77

	

	

	

	

Figure D-5. Predicted fatalities for projects with small hazardous areas based on the prior-only collision fatality
-3
2
model; projects with less than 2.4x10 km hazardous area are predicted to take less than 1 eagle in 30 years.	

	

	
The	Service	is	working	on	the	development	of	additional	tools	to	assist	project	developers	or	
operators	with	estimating	predicted	fatalities	given	different	inputs	and	allowing	for	the	flexibility	
to	incorporate	other	factors	into	additional	candidate	models.	We	encourage	project	developers	or	
operators	to	begin	coordinating	with	the	Service	early	in	the	process	(Stage	1	or	Stage	2)	so	that	we	
can	collaboratively	develop	a	suite	of	candidate	models	to	consider.	

	
	
Literature Cited
Gelman,	A.,	Carlin,	J.	B.,	Stern,	H.	S.,	and	D.	B.	Rubin.	2003.	Bayesian	Data	Analysis,	2nd	ed.	London,	
Chapman	&	Hall.	
Whitfield,	D.	P.		2009.		Collision	avoidance	of	golden	eagles	at	wind	farms	under	the	‘Band’	collision	
risk	model.	Report	from	Natural	Research	to	Scottish	Natural	Heritage,	Banchory,	UK.	

	
	
	

78

APPENDIX E: STAGE 4 – AVOIDANCE AND MINIMIZATION OF RISK USING ACPS AND OTHER
CONSERVATION MEASURES, AND COMPENSATORY MITIGATION
	
The	most	important	factor	when	considering	potential	effects	to	eagles	is	the	siting	of	a	wind	
project.		Based	on	information	gathered	in	Stage	2	and	analyzed	in	Stage	3,	the	project	developer	or	
operator	should	revisit	the	site	categorization	from	the	Stage	1	assessment	to	determine	if	the	
site(s)	still	falls	into	an	acceptable	category	of	risk	(at	this	stage,	acceptable	categories	are	2	and	3,	
and	very	rarely	1).		When	information	suggests	that	a	proposed	wind	project	has	a	high	eagle	
exposure	rate	and	presents	multiple	risk	factors	(e.g.,	is	proximate	to	an	important	eagle‐use	area	
or	migration	concentration	site		and	Stage	2	data	suggest	eagles	frequently	use	the	proposed	wind‐
project	footprint),	it	should	be	considered	a	category	1	site;	we	recommend	relocating	the	project	
to	another	area	because	a	location	at	that	site	would	be	unlikely	to	meet	the	regulatory	
requirements	for	a	programmatic	permit.		If	the	site	falls	into	categories	2	or	3,	or	rarely	some	
category	1	sites	where	there	is	potential	to	adequately	abate	risk,	the	ECP	should	next	address	
conservation	measures	and	ACPs	that	might	be	employed	to	minimize	or,	ideally,	avoid	eagle	
mortality	and	disturbance.		To	meet	regulatory	requirements,	ACPs,	if	available,	must	be	employed	
such	that	any	remaining	eagle	take	is	unavoidable.	
	
In	this	section	of	the	ECP,	we	recommend	project	developers	or	operators	re‐run	models	predicting	
eagle	fatality	rates	after	implementing	conservation	measures	and	available	ACPs	for	all	the	
plausible	alternatives.		This	re‐analysis	serves	two	purposes:	(1)	it	demonstrates	the	degree	to	
which	minimization	and	avoidance	measures	might	reduce	effects	to	eagle	populations	compared	
to	the	baseline	project	configuration,	and	(2)	it	provides	a	prediction	of	unavoidable	eagle	
mortality.	Conservation	measures	and	ACPs	should	be	tailored	to	specifically	address	the	risk	
factors	identified	in	Stage	3	of	the	ECP.		This	section	of	the	ECP	should	describe	in	detail	the	
measures	proposed	to	be	implemented	and	their	expected	results.	
	
The	Service	does	not	advocate	the	use	of	any	particular	conservation	measures	and	merely	
provides	the	below	list	as	examples.		Moreover,	at	this	time	none	of	these	measures	have	been	
approved	as	ACPs	for	wind	projects.		Ultimately,	project	developers	or	operators	will	propose	and	
implement	site	specific	conservation	measures	and	ACPs	(as	they	become	available)	in	cooperation	
with	local	Service	representatives	in	order	to	meet	the	regulatory	standard	of	reducing	any	
remaining	take	to	a	level	that	is	unavoidable.	
	
Examples	of	conservation	measures	that	could	be	considered	before	and	during	project	
construction,	depending	on	the	specific	risk	factors	involved,	include:	
1. Minimize	the	area	and	intensity	of	disturbances	during	pre‐construction	and	construction	
periods.	
2. Prioritize	locating	development	on	lands	that	provide	minimal	eagle	use	potential	including	
highly	developed	and	degraded	sites.	
3. Utilize	existing	transmission	corridors	and	roads.	
4. Set	turbines	back	from	ridge	edges.	
5. Site	structures	away	from	high	eagle	use	areas	and	the	flight	zones	between	them.	
6. Dismantle	nonoperational	meteorological	towers.	
7. Bury	power	lines	to	reduce	avian	collision	and	electrocution.	
8. Follow	the	Avian	Power	Line	Interaction	Committee	(APLIC)	guidance	on	power	line	
construction	and	design	(APLIC	2006).	
9. Minimize	the	extent	of	the	road	network.	

79
10. Avoid	the	use	of	structures,	or	remove	existing	structures,	that	are	attractive	to	eagles	for	
perching.	
11. Avoid	construction	designs	(including	structures	such	as	meteorological	towers)	that	
increase	the	risk	of	collision,	such	as	guy	wires.		If	guy	wires	are	used,	mark	them	with	bird	
flight	diverters	(according	to	the	manufacturer’s	recommendation).	
12. Avoid	siting	turbines	in	areas	where	eagle	prey	are	abundant.	
13. Avoid	areas	with	high	concentrations	of	ponds,	streams,	or	wetlands.	
	
Examples	of	avoidance	and	minimization	measures	that	could	be	considered	during	project	
operation,	depending	on	the	specific	risk	factors	involved,	include:	
	
1. Maintain	facilities	and	grounds	in	a	manner	that	minimizes	any	potential	impacts	to	eagles	
(e.g.	minimize	storage	of	equipment	near	turbines	that	may	attract	prey,	avoid	seeding	forbs	
below	turbines	that	may	attract	prey,	etc.).	
2. Avoid	practices	that	attract/enhance	prey	populations	and	opportunities	for	scavenging	
within	the	project	area.	
3. Take	actions	to	reduce	vehicle	collision	risk	to	wildlife	and	remove	carcasses	from	the	
project	area	(e.g.	deer,	elk,	livestock,	etc.).	
4. Instruct	project	personnel	and	visitors	to	drive	at	low	speeds	(<	25	mph)	and	be	alert	for	
wildlife,	especially	in	low	visibility	conditions.	
	
When	post‐construction	fatality	information	becomes	available,	the	project	developer	or	operator	
and	the	Service	should	consider	implementing	all	or	a	subset	of	the	additional	conservation	
measures	and	experimental	ACPs	that	were	considered	at	the	time	the	permit	was	issued	(see	
ASSESSING	RISK	AND	EFFECTS,	3b.	General	Approach	to	Address	Risks	in	the	ECPG).	
	
Examples	of	experimental	ACPs	that	could	be	identified	initially	or	after	evaluation	of	post‐
construction	fatality	monitoring	data,	depending	on	the	specific	risk	factors	involved,	include:	
	
1. Seasonal,	daily,	or	mid‐day	shut‐downs	(particularly	relevant	in	situations	where	eagle	
strikes	are	seasonal	in	nature	and	limited	to	a	few	turbines,	or	occur	at	a	particular	time	of	
day).	
2. Turbine	removal	or	relocation.	
3. Adjusting	turbine	cut‐in	speeds.	
4. Use	of	automated	detection	devices	(e.g.	radar,	etc.)	to	control	the	operation	of	turbines.	

	
	
Literature Cited
Avian	Power	Line	Interaction	Committee	(APLIC).		2006.		Suggested	practices	for	avian	protection	
on	power	lines:	the	state	of	the	art	in	2006.	Edison	Electric	Institute,	APLIC,	and	the	California	
Energy	Commission.	Washington	D.C.	and	Sacramento,	CA,	USA.	http://www.aplic.org/	
SuggestedPractices2006(LR‐2watermark).pdf.	

	

	

80

APPENDIX F: ASSESSING PROJECT-LEVEL TAKE AND CUMULATIVE EFFECTS ANALYSES
	

The	Service	is	required	to	evaluate	and	consider	the	effects	of	programmatic	take	permits	on	eagles	
at	the	eagle	management	unit,	local‐area,	and	project‐area	population	scales,	including	cumulative	
effects,	as	part	of	its	permit	application	review	process	(50	CFR	22.26	(f)(1)	and	USFWS	2009).		The	
Service	will	rely	on	information	a	developer	provides	from	the	Stage	1	and	Stage	2	assessments,	as	
well	as	all	other	available	information	on	mortality	and	other	population‐limiting	effects	at	the	
various	population	scales,	when	preparing	its	cumulative	impact	assessment.	The	Service’s	NEPA	
on	the	Eagle	Permit	Rule	evaluated	and	set	sustainable	take	levels	at	the	eagle	management	unit	
scale	(USFWS	2009).		However,	that	NEPA	analysis	did	not	assess	impacts	at	other	population	
scales.		A	significant	part	of	the	cumulative	effects	evaluation	is	assessing	the	effect	of	the	proposed	
take	in	combination	with	take	caused	by	previously	authorized	actions	and	reasonably	foreseeable	
future	actions	on	the	local‐area	eagle	population(s),	and	it	is	this	analysis	that	is	the	focus	of	this	
appendix.	
	
The	purpose	of	this	part	of	the	cumulative	effects	evaluation	is	to	identify	situations	where	take,	
either	at	the	individual	project	level	or	in	combination	with	other	authorized	or	foreseeable	future	
actions	and	other	limiting	factors	at	the	local‐area	population	scale,	may	be	approaching	levels	that	
are	biologically	problematic	or	which	cannot	reasonably	be	offset	through	compensatory	
mitigation.		In	previous	assessments	of	the	effect	of	falconry	take	on	raptor	populations	(Millsap	
and	Allen	2006),	the	Service	identified	annual	take	levels	of	5%	of	annual	production	to	be	
sustainable	for	a	range	of	healthy	raptor	populations,	and	annual	take	levels	of	1%	of	annual	
production	as	a	relatively	benign	harvest	rate	over	at	least	short	intervals	when	population	status	
was	uncertain.		This	approach	was	used	to	establish	take	thresholds	at	the	eagle	management	unit	
scale	(USFWS	2009).		The	Service	considered	several	alternatives	for	benchmark	harvest	rates	at	
the	local‐area	population	scale,	and	after	comparative	evaluation	identified	take	rates	of	between	
1%	and	5%	of	the	estimated	total	eagle	population	size	at	this	scale	as	significant,	with	5%	being	at	
the	upper	end	of	what	might	be	appropriate	under	the	BGEPA	preservation	standard,	whether	
offset	by	compensatory	mitigation	or	not.		These	local‐area	harvest	rate	benchmarks	are	overlain	
by	the	more	conservative	take	thresholds	for	the	eagle	management	units,	so	the	overall	harvest	
rate	at	the	eagle	management	unit	scale	should	not	exceed	levels	established	in	the	Final	
Environmental	Assessment	(USFWS	2009).	
	
The	Service	recommends	a	top‐down	approach	for	this	assessment:	(1)	identify	numbers	of	eagles	
that	may	be	taken	safely	at	the	national	level	(i.e.,	a	national‐level	benchmarks);	(2)	allocate	take	
opportunities	among	regional	eagle	management	units	(USFWS	2009)	as	a	function	of	the	
proportion	of	eagles	in	each	unit	(i.e.,	regional‐level	benchmarks);	(3)	further	allocate	take	
opportunities	to	the	local‐area	population	scale	as	a	function	of	inferred	eagle	population	size	at	
that	scale	(assuming,	in	the	absence	of	better	data	on	eagle	distribution	at	the	scale	of	the	eagle	
management	unit,	a	uniform	distribution	of	that	population);	and	(4)	incorporating	benchmarks	
that	can	be	used	to	assess	the	likely	sustainability	of	predicted	levels	of	take	at	the	local‐area	scale.		
Through	a	spatial	accounting	system,	permitted	take	is	managed	to	ensure	that	the	benchmarks	
also	consider	cumulative	effects	at	the	local‐area	eagle	population	scale	as	a	guard	against	
authorizing	excessive	take	at	this	scale.	
	
In	Table	F‐1,	we	work	through	this	approach	using	the	hypothetical	example	of	eight	individual	yet	
identical	projects,	one	in	each	bald	eagle	management	unit.		Each	of	these	projects	has	a	314	mi2	
footprint,	and	affects	a	local‐area	bald	eagle	population	over	8824	square	mile	(mi2)	area.		For	this	
example,	we	use	a	take	rate	of	5%	of	the	local‐area	bald	eagle	population	per	year	as	the	maximum	
acceptable	take	rate.		In	this	example,	the	5%	benchmark	take	rate	over	the	eight	projects	is	150	

81
individual	bald	eagles	per	year,	and	the	range	of	allowable	take	rates	at	this	scale	varies	across	
management	units	from	<1	bald	eagle	per	year	in	the	southwest	to	67	per	year	in	Alaska.		Table	F‐2	
provides	population	and	eagle	management	unit	area	statistics	for	golden	eagles	to	aid	in	
performing	these	calculations	for	that	species.	
	
As	noted	above,	in	cases	where	the	local‐area	eagle	populations	of	proximate	projects	overlap,	the	
overlap	should	be	taken	into	account	in	a	cumulative	effects	analysis	so	that	the	cumulative	take	on	
the	local‐area	population	scale	can	be	considered	against	population	benchmarks.		Figure	F‐1	
illustrates	one	method	to	do	this,	and	Table	F‐3	provides	the	calculations	for	this	example.		These	
examples	use	bald	eagles,	but	the	same	concept	and	approach	can	be	used	for	golden	eagles,	with	
Bird	Conservation	Regions	(BCRs)	defining	the	eagle	management	units.		The	example	in	Figure	F‐1	
involves	bald	eagles	in	Region	3.		Project	1	(in	green)	has	a	footprint	of	41	miles2	(mi2),	and	affects	a	
local‐area	bald	eagle	population	over	6854	mi2	(light	green	buffer	around	the	project	footprint).		
Following	the	approach	in	Table	F‐1,	project	1	was	issued	a	programmatic	take	permit	with	a	
maximum	annual	project‐level	take	of	21	bald	eagles	per	year	(see	Table	F‐3).		Project	2	(in	red,	the	
same	size	as	project	1)	applied	for	a	programmatic	eagle	take	permit	5	years	later.		The	calculated	
project‐level	bald	eagle	take	for	project	2	is	20	bald	eagles	per	year,	but	under	the	5%	benchmark,	
maximum	take	for	1563	mi2	of	project	2’s	local‐area	bald	eagle	population	(totaling	5	bald	eagles	
per	year)	was	already	allocated	to	project	1	(the	hatched‐marked	area	of	overlap	between	the	local	
areas	of	project	1	and	project	2).		Therefore,	the	calculated	local‐area	bald	eagle	take	for	project	2	
exceeds	the	5%	benchmark.		Thus,	the	decision‐maker	for	the	permit	for	project	2	should	carefully	
consider	whether	this	project	can	be	permitted	as	designed	under	the	requirements	of	our	
regulations	at	50	CFR	22.26.	
	
The	examples	assume	acceptable	compensatory	mitigation	opportunities,	when	they	are	required,	
are	limitless.		They	are	not,	and	where	compensatory	mitigation	is	necessary	to	offset	the	permitted	
take,	the	availability	of	compensatory	mitigation	can	become	the	proximate	factor	limiting	take	
opportunities.	
	
A	critical	assumption	of	this	approach	is	that	eagle	density	is	uniform	across	eagle	regions.		The	
potential	consequence	of	this	assumption	is	to	over	protect	eagles	in	areas	of	high	density	and	
under	protect	them	in	areas	of	low	density.		As	the	Service	and	others	develop	more	reliable	models	
for	predicting	the	distribution	of	eagles	within	regional	management	populations	at	finer	scales,	
these	approaches	should	be	used	in	place	of	an	assumption	of	uniform	distribution	in	the	analyses	
suggested	here.	

82

Table F-1. Example of the proposed method to calculate local-area annual eagle take benchmarks. The
example uses bald eagles (BAEA), and is based on a hypothetical scenario where a single project with a circular
footprint of 10-mile radius is proposed in each BAEA region. See Figure F-1 for an example of how to assess the
cumulative effects of such permitted take over the local-area population.
	
BAEA	
Management	
Unit	

Estimated	
Population	
Sizea	

Region	
Size	(mi2)	

R1	

7105	

245336	

Maximum	
Take	Rate	
(%	local‐
area	
population	
per	year)b	
5.0	

R2	

797	

565600	

R3	

27617	

R4	

Management	
Unit	Eagle	
Density	
(BAEA/	mi2)c	

Local	
Area	
(mi2)d	

Local‐area	
5%	
Benchmark	
(eagles	per	
year)e	

0.029	

8824	

13	

5.0	

0.001	

8824	

>1	

447929	

5.0	

0.062	

8824	

27	

13111	

464981	

5.0	

0.028	

8824	

12	

R5	

14021	

237687	

5.0	

0.059	

8824	

26	

R6	

5385	

732395	

5.0	

0.007	

8824	

3	

R7	

86550	

570374	

5.0	

0.152	

8824	

67	

R8	

889	

265779	

5.0	

0.003	

8824	

1	

Sum	

155474	

150	

	

a	Taken	directly	from	USFWS	(2009).	

b	A	take	rate	of	5%	is	the	Service’s	upper	benchmark	for	take	at	the	local‐area	population	scale.	
c	Management	unit	eagle	density	=	population	size	/	management	unit	size.	

d	The	local‐area	for	this	example	is	the	project	footprint	(in	this	case,	a	circle	with	radius	of	10	miles)	plus	a	

buffer	of	43	additional	miles	(43	miles	is	the	average	natal	dispersal	distance	for	the	BAEA)		=	3.142	*	532	.	
e	The	local‐area	5%	benchmark	=	(Local‐area*Regional	Eagle	Density)*0.05.	

83

Table F-2. Background information necessary to estimate the local-area take benchmarks for golden
eagles (GOEA). Columns are as in Table F-1. The local-area for golden eagles, which is not used in this table, is
calculated using the median natal dispersal distance of 140 miles (USFWS 2009).

GOEA	Management	Unit	
Alaska	
Northern	Pacific	Rainforest	
Prairie	Potholes	
Sierra	Nevada	
Shortgrass	Prairie	
Coastal	California	
Sonoran	and	Mojave	Desert	
Sierra	Madre	Occidental	
Chihuahuan	Desert	
Great	Basin	
Northern	Rockies	
Southern	Rockies	and	
Colorado	Plateau	
Badlands	and	Prairies	

Estimated	
Population	
Sizea	

BCR	Size	
(mi2)b	

Management	Unit	
Eagle	Density	
(GOEA	per	mi2)	

5
11
15
18
32
33
34
35
9
10

2400
108
1680
84
1080
960
600
360
720
6859
6172

557007
68777
160794
20414
148540
63919
95593
47905
72455
269281
199666

0.0043	
0.0016	
0.0104	
0.0041	
0.0073	
0.0150	
0.0063	
0.0075	
0.0099	
0.0255	
0.0309	

16	

3770	

199522	

0.0189	

17

7800

141960

0.0549	

BCR	
Number	

Sum

32593	

	
a	Taken	directly	from	USFWS	2009.	

b	BCR	area	values	are	from	the	North	American	Bird	Conservation	Region	website	at:	http://www.bsc‐
eoc.org/international/bcrmain.html	(last	visited	8	December	2011).	

	
	

	

84

	
			

Project 1

Project 2

60

0

60

120 Miles

	

	
	
Figure F-1. Example of the proposed method for ensuring local-area take benchmarks are not exceeded
through the cumulative take authorized over multiple projects. Project 1 is in green, project 2 is in red, and
the overlap in their local-area eagle bald eagle populations is the hatched-marked area (see text). This same
approach could be used to assess the cumulative effects of other forms of take and anthropomorphic impacts for
which data on population effects are available.

	

	

85

Table F-3. Calculations used to determine local-area bald eagle take for the example in Fig. F-1, where
project 1 is first-in-time, and the local-area bald eagle (BAEA) populations for the two projects overlap.
Calculations are as described in the footnotes to table F-1.

	
Project	

Project	1	(first‐
in‐time)	
Project	2,	
unadjusted	
Overlap	Area	
Project	2,	
adjusted	

Region	3	
Region	
BAEA	
Size	
Population	
(mi2)	
Size	

Maximum	
Take		Rate	
(%	local‐
area	
population	
per	year)b	

Regional	
Eagle	
Density	
(BAEA	
per	mi2)	

Local‐
area	
(mi2)	

Local‐area	
5%	
Benchmark	
(eagles	per	
year)e	

27617	

447929

5.0	

0.062	

6854	

21	

27617	

447929

5.0	

0.062	

6550	

20	

27617	

447929

5.0	

0.062	

1562	

5	

27617	

447929

5.0	

0.062	

13404	

15	

Literature Cited
USFWS.	2007.		Final	environmental	assessment,	take	of	raptors	from	the	wild	under	the	falconry	
regulations	and	the	raptor	propagation	regulations.		U.S.	Fish	and	Wildlife	Service,	Division	of	
Migratory	Bird	Management,	Washington,	D.C.	
USFWS.		2008.		Final	environmental	assessment	and	management	plan,	take	of	migrant	peregrine	
falcons	from	the	wild	for	use	in	falconry,	and	reallocation	of	nestling/fledgling	take.		U.S.	Fish	
and	Wildlife	Service,	Division	of	Migratory	Bird	Management,	Washington,	D.C.	
USFWS.		2009.		Final	environmental	assessment,	proposal	to	permit	take	as	provided	under	the	
Bald	and	Golden	Eagle	Protection	Act.		U.S.	Fish	and	Wildlife	Service,	Division	of	Migratory	Bird	
Management,	Washington,	D.C.	
USFWS.		2011.		Draft	eagle	conservation	plan	guidance.		U.S.	Fish	and	Wildlife	Service,	Division	of	
Migratory	Bird	Management,	Washington,	D.C.	

	

	

86

APPENDIX G: EXAMPLES USING RESOURCE EQUIVALENCY ANALYSIS TO ESTIMATE THE
COMPENSATORY MITIGATION FOR THE TAKE OF GOLDEN AND BALD EAGLES FROM WIND
ENERGY DEVELOPMENT
1. Introduction	
This	appendix	provides	Resource	Equivalency	Analysis	(REA)	examples	developed	by	the	Service	to	
illustrate	the	calculation	of	compensatory	mitigation	for	the	annual	loss	of	golden	(GOEA)	eagles	
and	bald	(BAEA)	eagles	caused	by	wind	power	if	conservation	measures	and	ACPs	do	not	remove	
the	potential	for	take,	and	the	projected	take	exceeds	calculated	thresholds	for	the	species	or	
management	population	affected.		These	examples	result	in	estimates	of	the	number	of	high‐risk	
electric	power	poles	that	would	need	to	be	retrofitted	per	eagle	taken	based	on	the	inputs	provided	
below.		Detailed	explanatory	documentation,	literature,	and	supporting	REA	spreadsheets	are	now	
located	at:	www.fws.gov/windenergy/index.html	
	
As	a	framework	for	compensatory	mitigation,	it	needs	to	be	clear	that	the	results	provided	below	
are	an	illustration	of	how	REA	works	given	the	current	understanding	of	GOEA	and	BAEA	life	
history	inputs,	effectiveness	of	retrofitting	high‐risk	electric	power	poles,	the	expected	annual	take,	
and	the	timing	of	both	the	eagle	take	permit	and	implementation	of	compensatory	mitigation.		As	
would	be	expected,	the	estimated	number	of	eagle	fatalities	and	the	permit	renewal	period	affect	
the	number	of	poles	to	be	retrofitted.		Delays	in	retrofitting	would	lead	to	more	retrofitted	poles	
owed.		New	information	on	changes	in	the	level	of	take,	understanding	of	the	eagle	life	history,	or	
effectiveness	of	retrofitting	could	be	used	to	change	the	number	of	retrofitted	poles	needed	for	
compensation.		Finally,	while	only	electric	pole	retrofitting	is	presented	here	in	detail,	the	REA	
metric	of	bird‐years	lends	itself	to	consideration	of	other	compensatory	mitigation	options	to	
achieve	the	no‐net‐loss	standard	in	the	future.		With	enough	reliable	information,	any	
compensatory	mitigation	that	directly	leads	to	an	increased	number	of	GOEA	and	BAEA	(e.g.,	
habitat	restoration)	or	the	avoided	loss	of	these	eagles	(e.g.,	reducing	vehicle/eagle	collisions,	
making	livestock	water	tanks	‘eagle‐safe’,	lead	ammunition	abatement,	etc.)	could	be	considered	for	
compensation	within	the	context	of	the	REA.	

	
2. REA Inputs	
The	best	available	peer‐reviewed,	published	data	are	provided	in	Tables	G‐1	and	G‐2.		It	should	be	
noted	that	additional	modeling	work	within	the	REA	may	be	needed,	particularly	on	issues	related	
to	migration,	adult	female	survivorship,	natal	dispersal,	age	at	first	breeding,	and	population	sex	
ratio.	

	
	

	

87

	
Table G-1. EXAMPLE INPUTS. REA Inputs to Develop a Framework of Compensatory Mitigation for Potential Take
of GOEA from Wind Energy Development

	
Parameter	

REA	Input	

Reference	

Start	year	of	permit	
Length	of	permit	renewal	
period	
Estimated	take	

2012	

Example.	

5	years	

Example.	

1	eagle/year	

Example.	
28	years,	3	months,	USGS	Bird	Banding	
Lab.	
Consistent	with	Cole	(2010)	approach.	
 20%	juveniles	(age	class	(0‐1))	
 35%	sub‐adults	(11.67%	for	each	age	
class	from	age	class	(1‐2)	through	age	
class	(3‐4))	
 45%	adults	(1.73%	for	each	age	class	
from	age	class	(4‐5)	through	age	class	
(29‐30))	
Assume	age	class	is	distributed	evenly	
over	time.		Age	distribution	derived	
from	models	presented	in	USFWS	2009.	

Average	maximum	
lifespan	

Age	distribution	of	birds	
killed	at	wind	facilities	
(based	on	age	
distribution	of	GOEA	
population)	

30	years	

(0‐1)	
(1‐4)	
(4‐30)	

Age	start	reproducing	
Expected	years	of	
reproduction	
%	of	adult	females	that	
reproduce	annually	
Productivity	(mean	
number	of	individuals	
fledged	per	occupied	nest	
annually)	
year	0‐1	survival	
year	1‐2	survival	
year	2‐3	survival	
year	3‐4	survival	
year	4+	survival	
Relative	productivity	of	
mitigation	option	

Discount	rate	

	

20%	
35%	
45%	

Age	5
[age	class	(5‐6)]	

Steenhof	et	al.	1984;	Kochert	et	al.	2002	

25	years	

=	(Maximum		Lifespan)	–	(Age	Start	
Reproducing)		(Harmata	2002)	

80%	

Steenhof	et	al.	1997	

0.61	

USFWS	2009	

61%
79%
79%
79%
90.9%

USFWS	2009	

Example.	 Compensatory	mitigation	
involves	retrofitting	high‐risk	electric	
0.0036	eagle	
power	poles,	thus	avoiding	the	loss	of	
electrocutions/pole/year
GOEA	from	electrocution	(Lehman	et	al.	
2010).	
A	3%	discount	rate	is	commonly	used	
for	valuing	lost	natural	resource	
3%	
services	(Freeman	1993,		Lind	1982,		
NOAA	1999;	and	court	decisions	on	
damage	assessment	cases)	

	

88

	
Table G-2. EXAMPLE INPUTS. REA Inputs to Develop a Framework of Compensatory Mitigation for Potential Take
of BAEA from Wind Energy Development
	
Parameter	

REA	Input	

Start	year	of	permit	
Length	of	permit	
renewal	period	
Estimated	take	

Relative	productivity	of	
mitigation	option	

Discount	rate	

	

Example.	

5	years	

Example.	

30	years	

Age	distribution	of	
birds	killed	at	wind	
(0‐1)	
facilities	(based	on	age	 (1‐4)	
distribution	of	BAEA	
(4‐30)	
population)	

Expected	years	of	
reproduction	
%	of	adult	females	that	
reproduce	annually	
Productivity	
year	0‐1	survival	
year	1‐2	survival		
year	2‐3	survival		
year	3‐4	survival	
year	4+	survival			

2011	
1	eagle/year	

Average	maximum	
lifespan	

Age	start	reproducing	

Reference	

15.4%	
30%	
54.6%	

Age	5
[age	class	(5‐6)]	
25	years	

Example.	
32	years	10	months;	Longevity	record	
from	USGS	Bird	Banding	Lab.		Consistent	
with	Cole	(2010)	approach.	
 15.4%	juveniles	(age	class	(0‐1))	
 30%	sub‐adults	(10%	for	each	age	
class	from	age	class	(1‐2)	through	age	
class	(3‐4))	
 54.6%	adults	(2.1%	for	each	age	class	
from	age	class	(4‐5)	through	age	class	
(29‐30))	
Assume	age	class	is	distributed	evenly	
over	time.		Age	distribution	derived	
from	models	presented	in	USFWS	2009.	
Buehler	2000	
=	(Maximum	Lifespan)	–	(Age	Start	
Reproducing)	

42%	

Hunt	1998,	per.	comm.	Millsap	

1.3
77%
88%
88%
88%
83%

Millsap	et	al. 2004
Millsap	et	al.	2004	

Example. 	Mitigation	involves	
0.0036	eagle	
retrofitting	high‐risk	electric	power	
electrocutions/pole/year	 poles,	thus	avoiding	the	loss	of	BAEA	
from	electrocution	(Lehman	et.	al	2010).
A	3%	discount	rate	is	commonly	used	
for	valuing	lost	natural	resource	
3%	
services	(Freeman	1993;	Lind	1982;	
NOAA	1999;	and	court	decisions	on	
damage	assessment	cases).	

89

3. REA Example – WindCoA
The	Service	developed	the	following	hypothetical	scenario	for	permitting	and	compensatory	
mitigation	to	be	applied	to	the	take	of	GOEA1		from	wind	power	operations:	
	
WindCoA	conducted	three	years	of	pre‐construction	surveys	to	determine	relative	abundance	of	
GOEA	at	their	proposed	wind	project	in	Texas.		The	survey	data	was	then	used	to	populate	a	risk	
assessment	model	to	generate	an	eagle	fatality	estimate.		The	initial	fatality	estimate	of	two	eagles	
per	year	was	further	reduced	after	WindCoA	implemented	a	few	mutually	agreed	upon	ACPs.		The	
final	fatality	estimate	generated	from	the	risk	assessment	model,	after	consideration	of	the	
advanced	conservation	practices,	was	an	annual	take	of	one	GOEA	per	year	over	the	life	of	the	
permit	starting	in	2012.	
	
WindCoA	decided	to	conduct	an	REA	to	determine	the	number	of	high‐risk	power	poles	that	would	
need	to	be	retrofitted	to	get	to	no‐net‐loss.		The	company	used	the	Service’s	GOEA	REA	inputs	and	
assumed	the	power	pole	retrofit	would	occur	in	calendar	year	2012,	thus	offsetting	the	potential	
loss	of	eagles	at	the	newly	operating	wind	project	with	avoidance	of	electrocution	of	an	equal	
number	of	GOEA.		Through	proper	operation	and	maintenance	(O&M),	the	retrofitted	poles	are	
assumed	to	be	effective	in	avoiding	the	loss	of	eagles	for	10	years.		The	results	of	the	model	are	
expressed	in	the	total	number	of	electric	power	poles	to	be	retrofitted	to	equate	to	no‐net‐loss	of	5	
eagles	for	the	5‐year	permit	renewal	period	(1	eagle	annually	over	five	years).		These	results	are	
extrapolated	over	the	expected	operating	life	of	the	wind	project,	which	is	assumed	to	be	30	years,	
for	a	total	take	of	30	eagles.	
	
The	results	of	the	REA	indicated	that	WindCoA	needed	to	retrofit	approximately	149	power	poles	
for	the	first	5‐year	permit	period	(see	Table	G‐3).		Using	an	estimated	cost	of	$7500/pole,	the	
Service	estimated	that	WindCoA	could	contribute	$1,117,500	to	a	third‐party	mitigation	account	or	
contract	the	retrofits	directly.		After	determining	that	they	could	fund	the	retrofits	directly	at	a	
lower	cost,	WindCoA	decided	to	partner	with	UtilityCoB	to	get	the	required	number	of	poles	
retrofitted.		UtilityCoB	had	previously	conducted	a	risk	assessment	of	their	equipment	and	had	
identified	high‐risk	poles	that	were	likely	to	take	golden	eagles.		Through	a	written	agreement,	
WindCoA	provided	funding	to	UtilityCoB	to	retrofit	the	required	number	of	power	poles	and	
maintain	the	retrofits	for	10	years.		In	addition,	WindCoA	contracted	with	ConsultCoC	to	perform	
effectiveness	monitoring	of	the	retrofitted	power	poles	for	2	years.		The	contract	required	that	
ConsultCoC	visit	each	retrofitted	power	pole	every	4	months	(quarterly)	to	perform	fatality	
searches	and	check	for	proper	operation	and	maintenance	of	the	equipment.		The	Service	reviewed	
the	compensatory	mitigation	project	proposed	by	WindCoA	and	found	it	to	be	consistent	with	
requirements	at	50	CFR	22.26.		After	reviewing	the	signed	contract	between	WindCoA,	UtilityCoB,	
and	ConsultCoC,	the	Service	issued	a	programmatic	eagle	take	permit	to	WindCoA.	

	
a. REA Language and Methods
As	discussed	in	greater	detail	in	documents	on	the	supporting	website,	this	REA	includes:	
	
 The	direct	loss	of	GOEA/BAEA	eagles	from	the	take	(debit	in	bird‐years);	
 The	relative	productivity	of	retrofitting	high‐risk	power	poles,	which	is	the	
effectiveness	in	avoiding	the	loss	of	GOEA/BAEA	by	electrocution	as	a	mitigation	
offset	(measured	in	total	bird‐years	per	pole);	and	

1

Using the inputs provided in Table G-2, this scenario may also be applied to BAEA.

90


	The	mitigation	owed,	which	is	the	total	debit	divided	by	the	relative	productivity	
(scaling)	to	identify	the	number	of	high‐risk	power	poles	that	need	retrofitting	to	
completely	offset	the	take	of	GOEA/BAEA	eagles	(credit).	

	
There	are	up	to	16	steps	when	conducting	a	REA.		Depending	on	whether	foregone	future	
reproduction	(part	of	the	debit)	is	included,	there	are	up	to	13	total	steps	involved	in	
calculating	the	injury	side	(debit)	of	a	REA,	and	three	additional	steps	involved	in	estimating	
compensatory	mitigation	owed	(credit).		Please	refer	to	the	technical	note	“Scaling	Directly	
Proportional	Avoided	Loss	Mitigation/Restoration	Projects”	on	the	supporting	website	
(www.fws.gov/windenergy)	for	further	information	on	the	development	of	REA	inputs	and	
the	inclusion	of	lost	reproduction.		Notably,	in	the	case	of	an	avoided	loss	project	where	the	
estimated	prevented	loss	of	bird‐years	(e.g.,	through	mitigation)	is	directly	proportional	to	
the	loss	of	bird‐years	(e.g.,	from	“take”),	the	life	history	inputs	(e.g.,	longevity,	age	
distribution,	survival	rates,	reproduction)	do	not	affect	the	final	results	of	the	credit	owed.		
That	is,	the	retrofitting	of	high‐risk	power	poles	is	a	directly	proportional	avoided	loss,	so	
only	the	level	of	take	(number	of	eagles	annually),	the	avoided	loss	of	eagles	per	mitigated	
electric	pole,	the	number	of	years	the	mitigated	pole	is	effective	in	avoiding	the	loss	of	
eagles,	and	the	timing	of	the	mitigation	relative	to	the	take	affect	the	final	credit	owed.		It	
should	also	be	noted	that	the	annual	take	of	one	eagle	is	used	in	the	example	because	the	
lost	bird‐years	associated	with	one	eagle	can	be	easily	multiplied	by	the	actual	take	to	
estimate	the	total	debit	in	bird‐years.	
	
The	following	is	a	brief	discussion	of	REA	variables	used	in	the	Service’s	WindCoA	example	
that	affect	the	outcome	of	the	compensatory	mitigation	calculation:	
 Relative	Productivity	of	Mitigation	(0.0036	electrocutions/pole/year)	–	This	
rate	is	taken	directly	from	published	literature	on	eagle	electrocution	rates	in	
northeastern	Utah	and	northwestern	Colorado	and	is	specific	to	eagles	(Lehman	et	
al.	2010).		Although	the	referenced	study	also	lists	a	higher	rate	(0.0066)	that	
includes	all	known	eagle	mortalities,	this	rate	included	eagles	that	may	have	died	
from	causes	unrelated	to	electrocution.	
 Years	of	Avoided	Loss	Per	Retrofitted	Pole	(10	Years)	–	The	Service	uses	a	
period	of	10	years	for	crediting	the	project	developer	or	operator	for	the	avoided	
loss	of	eagles	from	power	pole	retrofits.		This	is	a	reasonable	amount	of	time	to	
assume	that	power	pole	retrofits	will	remain	effective.		However,	project	developers	
or	operators	should	consider	entering	into	agreements	with	utility	companies	or	
contractors	for	the	long‐term	maintenance	of	retrofits.		Evidence	of	this	type	of	
agreement	could	increase	the	amount	of	credit	received	by	the	project	developer	or	
operator	and,	as	a	result,	decrease	the	amount	of	compensatory	mitigation	required.	
 Permit	Renewal	Period	(5	Years)	–	This	will	be	the	review	period	that	is	used	by	
the	Service	for	adaptive	management	purposes	and	re‐calculation	of	compensatory	
mitigation.		The	Service	believes	that	this	length	of	time	will	enable	the	project	
developer	or	operator	to	continue	to	meet	the	statutory	and	regulatory	eagle	
preservation	standard.		This	permit	review	tenure	will	remain	the	same	regardless	
of	the	overall	tenure	of	the	permit.	
 Retrofit	Cost/Payment	($7,500/pole)	–	The	Service	received	input	directly	from	
the	industry	regarding	the	actual	costs	to	retrofit	power	poles.		Estimates	ranged	
from	a	low	of	approximately	$400	to	over	$11,000	given	that	costs	vary	according	to	
many	factors.		The	Service	believes	that	$7,500	represents	a	reasonable	estimate	for	
the	current	cost	to	retrofit	power	poles	in	the	United	States.		Project	developers	or	

91
operators	are	encouraged	to	contract	directly	for	retrofits	as	this	will	likely	not	be	as	
costly	as	contributing	$7,500/pole	to	an	eagle	compensatory	mitigation	account.	

	
b. REA Results for WindCoA
Using	the	WindCoA	example	described	above,	along	with	the	REA	inputs	provided	in	Table	
G‐1,	Table	G‐3	provides	a	summary	of	the	results:	

	
	
Table G-3. WindCoA Example: Compensatory Mitigation Owed for a 5-Year Permitted Take of 5 GOEA
Extrapolated to the 30-Year Expected Operating Life of the Wind Project (30 GOEA in Total).
Total	Debit	for	Take	of	1	GOEA	

28.485	

	÷Relative	Productivity	of	High‐
Risk	Electric	Pole	Retrofitting	

÷0.191	

=	Mitigation	Owed	for	5‐Year	
Permitted	Take		
x	#	Cycles	of	5‐Year	Permit	
Reviews	
=Total	Mitigation	Owed		
*PV=Present	Value	

	
	
	

=149.136	
x	6	=	894.818	

PV*	bird‐years	for	5	years	of	GOEA	take	
Avoided	loss	of	PV	bird‐years	per	
retrofitted	pole	
(assumes	10	years	of	avoided	loss	per	pole	
based	on	the	commitment	from	
UtilityCoB)	
Poles	to	be	retrofitted	to	achieve	no‐net‐
loss		
Poles	to	be	retrofitted	to	achieve	no‐net‐
loss	for	the	30‐year	expected	operating	life	
of	the	wind	project	

If	all	of	the	REA	inputs	remain	the	same	after	the	initial	five	years,	then	the	estimated	
149.14	poles	may	be	multiplied	by	the	expected	number	of	permit	reviews	to	provide	an	
estimate	of	the	total	number	of	poles	that	would	eventually	be	retrofitted.		For	example,	for	
the	30‐year	life	cycle	of	the	WindCoA	wind	project,	149.14	poles	would	be	multiplied	by	6	
permit	renewals	to	equal	approximately	895	high‐risk	power	poles	in	total	to	be	retrofitted	
as	compensatory	mitigation	for	the	take	of	30	GOEA	over	30	years	(1	eagle	annually).		While	
this	example	shows	the	effectiveness	of	the	mitigation	method	as	lasting	for	10	years,	it	may	
be	the	case	that	the	method	selected	is	more	or	less	effective	at	avoiding	the	loss	of	eagles	
(e.g.,	5	years,	more	than	10	years).		The	REA	can	be	adjusted	for	the	expected	effectiveness	
of	mitigation,	and	more	or	fewer	high‐risk	power	poles	would	need	to	be	mitigated.		All	
estimates	of	compensatory	mitigation	are	contingent	on	proper	operation	and	maintenance	
being	conducted	by	UtilityCoB	or	a	contractor	to	ensure	that	the	expected	effectiveness	is	
achieved.	
	
For	purposes	of	illustration,	should	WindCoA	choose	to	use	the	GOEA	inputs	provided	in	
Table	G‐1	and	their	fatality	estimate	is	that	5	GOEA	will	be	taken	annually,	the	results	may	
be	easily	adjusted	as	shown	in	Table	G‐4:	

	

	

92

	

	

Table G-4. WindCoA Example: Compensatory Mitigation Owed for a 5-Year Permitted Take of 25 GOEA
Extrapolated to the 30-Year Expected Operating Life of the Wind Project (150 GOEA in Total).
Total	Debit	for	Take	of	1	GOEA	

28.485	

PV	bird‐years	for	5	years	of	GOEA	take	
from	Table	F‐3		

x	Actual	Annual	Take	of	GOEA		

x	5	=142.425	

PV	bird‐years	for	5	years	of	GOEA	take	

÷	Relative	Productivity	of	High‐
Risk	Electric	Pole	Retrofitting	
=	Mitigation	Owed	for	5‐Year	
Permitted	Take	
x	#	Cycles	of	5‐Year	Permit	
Reviews	=	Total	Mitigation	
Owed		
PV=Present	Value	

÷0.191	
=745.681	
x	6	=4474.086	

Avoided	loss	of	PV	bird‐years	per	
retrofitted	pole	(assumes	10	years	of	
avoided	loss	per	pole	based	on	the	
commitment	from	UtilityCoB)	
Poles	to	be	retrofitted	to	achieve	no‐net‐
loss		
Poles	to	be	retrofitted	to	achieve	no‐net‐
loss	for	the	30‐year	expected	operating	life	
of	the	wind	project	

	

c. Summary of Bald Eagle REA Results
Following	the	same	process	described	above	for	GOEA	(i.e.,	using	the	WindCoA	example	and	
the	BAEA	REA	inputs	provided	in	Table	G‐2),	Table	G‐5	provides	a	summary	of	the	results	
for	bald	eagles:	

	
	
	
Table G-5. Example of Compensatory Mitigation Owed for a 5-Year Permitted Take of 5 BAEA Extrapolated to the
30-Year Expected Operating Life of the Wind Project (30 BAEA in Total).	
Total	Debit	for	Take	of	1	BAEA	
÷	Relative	Productivity	of	High‐
Risk	Electric	Pole	Retrofitting	
	=	Mitigation	Owed	for	5‐Year	
Permitted	Take	
x	#	Cycles	of	5‐Year	Permit	
Reviews	=	Total	Mitigation	
Owed	
PV=Present	Value	

20.229
÷0.136	
=149.136	
x	6	=894.818	

PV	bird‐years	for	5	years	of	BAEA	take
Avoided	loss	of	PV	bird‐years	per	
retrofitted	pole	
Poles	to	be	retrofitted	to	achieve	no‐net‐
loss	
Poles	to	be	retrofitted	to	achieve	no‐net‐
loss	for	the	30‐year	expected	operating	life	
of	the	wind	project	

	

	
	

Although	there	are	differences	between	GOEA	and	BAEA	life	history	inputs	(e.g.,	longevity,	
age	distribution,	survival	rates,	reproduction),	the	estimated	avoided	loss	of	bird‐years	
through	mitigation	is	directly	proportional	to	the	loss	of	bird‐years	from	the	take,	so	the	life	
history	inputs	do	not	affect	the	final	results	of	the	credit	owed.		Because	there	was	no	
change	in	the	level	of	take	(number	of	eagles	annually),	the	avoided	loss	of	eagles	per	

93
mitigated	electric	pole,	the	number	of	years	the	mitigated	pole	is	effective	in	avoiding	the	
loss	of	eagles,	or	the	timing	of	the	mitigation	relative	to	the	take,	there	is	no	change	in	the	
credit	owed.		To	help	illustrate,	when	comparing	the	results	of	BAEA	to	GOEA,	both	the	
debit	(20.23÷28.49)	and	the	relative	productivity	of	electric	pole	retrofitting	(0.14÷0.19)	for	
BAEA	are	approximately	70%	of	GOEA,	so	the	amount	of	retrofitting	owed	is	the	same.		That	
is,	both	the	numerator	of	the	scaling	equation	(total	debit)	and	the	denominator	(relative	
productivity	of	mitigation)	were	changed	proportionally	(approximately	70%),	so	there	is	
no	change	in	the	mitigation	owed.	

	
d. Discussion on Using REA
The	ECPG	does	not	mandate	the	use	of	REA.		Rather,	the	Service	recognized	the	need	for	a	
reliable,	transparent,	reproducible,	and	cost‐effective	tool	to	expedite	wind	power	permits,	
while	ensuring	sufficient	compensatory	mitigation	for	the	take	of	golden	eagles	and	bald	
eagles	from	operations	to	meet	regulatory	permitting	requirements.		Although	there	is	a	
learning	curve,	REA	meets	these	basic	needs.		This	appendix	and	materials	on	the	
supporting	website	explain	the	methods,	share	the	tools	to	run	REAs,	and	discuss	how	
changes	in	the	different	inputs	can	affect	the	results.		Should	project	developers	or	
operators/applicants	choose	to	use	the	provided	inputs,	methods,	and	tools,	the	Service	will	
be	able	to	appropriately	focus	on	the	expected	take	of	eagles.		Project	developers	or	
operators/applicants	have	the	discretion	to	offer	alternative	REA	inputs	or	use	different	
compensatory	mitigation	modeling	methods.		However,	they	will	need	to	provide	sufficient	
evidence	and	tools	(if	necessary)	to	ensure	that	the	Service	can	provide	appropriate	review	
of	the	results,	and	should	expect	that	such	an	effort	will	likely	take	additional	time.	

	
e. Additional Compensatory Mitigation Example
In	the	United	States,	another	known	cause	of	mortality	to	eagles,	both	bald	and	golden,	is	
vehicle	collisions.		Eagles	are	susceptible	to	being	struck	by	vehicles	as	they	feed	on	
carcasses	along	roadsides,	particularly	in	areas	of	the	United	States	where	large	numbers	of	
ungulates	concentrate	seasonally	(e.g.	winter,	breeding	season,	etc.).		As	a	compensatory	
mitigation	strategy,	a	project	developer	or	operator	may	decide	to	collect	data	(or	use	
existing	data	if	it	is	available)	on	the	annual	number	of	eagle	mortalities	that	result	from	
vehicle	collisions	in	a	specified	geographic	area	or	along	a	specific	stretch	of	roadway.		This	
data	could	then	be	used	to	generate	an	estimate	of	the	number	of	eagle	mortalities	that	
could	be	prevented	in	the	same	area	by	removing	carcasses	from	roadsides.		If	there	was	
sufficient	evidence	that	this	was	a	valid	project	(e.g.	quantifiable	and	verifiable),	the	project	
developer	or	operator	could	contract	to	have	these	roadsides	‘cleaned’	of	carcasses	during	
the	time	of	year	that	ungulates	concentrate	and	eagles	are	known	to	be	struck.		The	credible	
estimate	of	eagle	mortalities	that	would	be	avoided	through	carcass	removal	would	be	the	
value	of	the	compensatory	mitigation	achieved.	

	
f. Take from Disturbance
Project	developers	or	operators	should	work	with	the	Service	to	determine	if	take	from	
disturbance	is	likely	to	occur.		This	should	be	predicted	in	advance	based	on	Stage	3	data,	
and	verified		through	post‐construction	monitoring	in	Stage	5.		The	following	are	
recommended	take	calculations	based	on	information	contained	within	the	FEA	(USFWS	
2009):	
	
For	the	standard	bald	eagle	population:	

94



Take	resulting	from	disturbance	at	one	nest	on	only	one	occasion	=	take	of	1.3	
individuals	
One	nest	take	resulting	in	the	permanent	abandonment	of	a	territory	=	take	of	1.3	
individuals	for	the	first	year,	then	take	of	8	individuals	annually	until	data	show	the	
number	of	breeding	pairs	has	returned	to	or	exceeded	the	original	estimated	
number	for	the	eagle	management	unit.	

	
For	the	standard	golden	eagle	population:	
 Take	resulting	from	disturbance	at	one	nest	on	only	one	occasion	=	take	of	0.8	
individuals	
 One	nest	take	resulting	in	the	permanent	abandonment	of	a	territory	=	take	of	0.8	
individuals	for	the	first	year,	then	take	of	4	individuals	annually	until	data	show	the	
number	of	breeding	pairs	has	returned	to	or	exceeded	the	original	estimated	
number	for	the	eagle	management	unit.	
	
Using	the	data	presented	in	the	above	WindCoA	example,	the	compensatory	mitigation	
required	for	disturbance	resulting	in	the	loss	of	productivity	from	one	GOEA	nest	for	one	
year	would	result	in	the	following:	
	
1. Disturbance	take	of	one	GOEA	nest	on	one	occasion	=	0.8	GOEA,	
2. From	the	REA,	the	take	of	one	GOEA	for	one	year	=	6	PV	bird‐years,	
3. Six	PV	bird‐years/GOEA	*	0.8	GOEA	=	4.8	PV	bird‐years,	and	
4. From	the	REA,	4.8	PV	bird‐years	÷	0.191	PV	bird‐years/pole	retrofitted	(for	10	year	
maintenance	of	poles)	=	25.1	poles	retrofitted.	
	
WindCoA	would	be	required	to	retrofit	a	total	of	174.24	poles	(149.14	poles	for	the	lethal	
take	of	5	GOEA	(see	Table	G‐3)	+	24.5	poles	for	the	disturbance	take	of	one	GOEA	nest)	to	
cover	the	initial	five	year	permitted	take.	

	
	
Literature Cited
Buehler,	D.	A.	2000.	Bald	Eagle	(Haliaeetus	leucocephalus),	The	Birds	of	North	America	Online	(A.	
Poole,	Ed.).	Ithaca:	Cornell	Lab	of	Ornithology;	Retrieved	from	the	Birds	of	North	America	
Online:	http://bna.birds.cornell.edu/bna/species/506.	
Cole,	S.	2010.	How	much	is	enough?	Determining	adequate	levels	of	environmental	compensation	
for	wind	power	impacts	using	resource	equivalency	analysis:	An	illustrative	and	hypothetical	
case	study	of	sea	eagle	impacts	at	the	Smola	Wind	Farm,	Norway.	Epsilon	Open	Archive	
Publishing,	Swedish	Agricultural	University.	
Freeman,	A.M.	III.		1993.	The	Measurement	of	Environmental	and	Resource	Values:	Theory	and	
Methods.		(Resources	for	the	Future,	Washington,	DC).	
Harmata,	A.	R.	2002.	Encounters	of	Golden	Eagles	banded	in	the	Rocky	Mountain	West.		J.	Field	
Ornithol.	73:23‐32.	
Hunt,	W.G.	1998.	Raptor	floaters	at	Moffat’s	equilibrium.	Oikos	81:1‐7.	
Kochert,	M.	N.,	K.	Steenhof,	C.	L.	Mcintyre	and	E.	H.	Craig.	2002.	Golden	Eagle	(Aquila	chrysaetos),	
The	Birds	of	North	America	Online	(A.	Poole,	Ed.).	Ithaca:	Cornell	Lab	of	Ornithology;	Retrieved	
from	the	Bird	of	North	America	Online:	http://bna.birds.cornell.edu/bna/species/684.	
Lehman,	R.	N.,	Savidge,	J.	A.,	Kennedy,	P.	L.	and	Harness,	R.	E.	(2010),	Raptor	Electrocution	Rates	for	
a	Utility	in	the	Intermountain	Western	United	States.	Journal	of	Wildlife	Management,	74:	459‐
470.	

95
Lind,	R.		1982.	A	Primer	on	the	Major	Issues	Relating	to	the	Discount	Rate	for	Evaluating	National	
Energy	Options	in	Discounting	for	Time	and	Risk	in	Energy	Policy,	edited	by	R.	Lind.		
Washington:	Resources	for	the	Future.	
Millsap,	B.A.,	T.	Breen,	E.	McConnell,	T.	Steffer,	L.	Phillips,	N.	Douglass,	S.	Taylor.	2004.	Comparative	
fecundity	and	survival	of	bald	eagles	fledged	from	suburban	and	rural	natal	areas	in	Florida.	
Journal	of	Wildlife	Management	68:1018‐1031.	
NOAA.		1999.	Discounting	and	the	Treatment	of	Uncertainty	in	Natural	Resource	Damage	Assessment.	
Technical	Paper	99‐1	(Silver	Spring,	MD:	NOAA).	
Steenhof,	K.,	M.N.	Kochert,	and	M.	Q.	Moritsch.	1984.	Dispersal	and	migration	of	southwestern	Idaho	
raptors.	J.	Field	Ornithol.	55:	357‐368.	
Steenhof,	K.,	M.	N.	Kochert,	and	T.	L.	McDonald.	1997.	Interactive	effects	of	prey	and	weather	on	
Golden	Eagle	reproduction.	J.	Anim.	Ecol.	66:	350‐362.	
USFWS.	2009.	Final	environmental	assessment.	Proposal	to	permit	take	provided	under	the	Bald	
and	Golden	Eagle	Protection	Act.	U.S.	Fish	and	Wildlife	Service,	Division	of	Migratory	Bird	
Management,	Washington	D.C.,	USA.	

96

APPENDIX H: STAGE 5 – CALIBRATING AND UPDATING OF THE FATALITY PREDICTION AND
CONTINUED RISK-ASSESSMENT
	

Given	the	degree	of	uncertainty	that	currently	exists	surrounding	the	risk	of	wind	facilities	to	eagles	
and	the	factors	that	contribute	to	that	risk,	post‐construction	monitoring	is	one	of	the	most	
significant	activities	that	will	be	undertaken	by	eagle	programmatic	take	permit	holders.		Post‐
construction	monitoring	has	two	basic	components	when	applied	to	eagle	take:	(1)	estimating	the	
mean	annual	fatality	rate,	and	(2)	assessing	possible	disturbance	effects	on	neighboring	nests	and	
communal	roosts.		Provided	that	assessments	conducted	during	Stages	1‐4	are	consistent,	robust,	
and	reliably	performed	as	suggested	in	this	ECPG,	the	pre‐construction	data	should	provide	a	solid	
platform	for	development	of	the	Stage	5	monitoring	and	assessment	studies.	

	
1. Fatality Monitoring
All	wind	facilities	that	are	permitted	to	take	eagles	will	need	to	conduct	fatality	monitoring	to	
ensure	compliance	with	regulatory	requirements.		Fatality	monitoring	must	be	conducted	at	all	
wind	facilities	that	are	permitted	to	take	eagles.		We	anticipate	that	in	most	cases,	intensive	
monitoring	to	estimate	the	true	annual	fatality	rate	and	to	assess	possible	disturbance	effects	will	
be	conducted	for	at	least	the	first	two	years	after	permit	issuance,	followed	by	less	intense	
monitoring	for	up	to	three	years	after	the	expiration	date	of	the	permit,	in	accordance	with	
monitoring	requirements	at	50	CFR	22.26(c)(2).		However,	additional	intensive,	targeted	
monitoring	may	be	necessary	to	determine	the	effectiveness	of	additional	conservation	measures	
and	ACPs	implemented	to	reduce	observed	fatalities.		Such	monitoring	should	be	rigorous	and	
sufficient	to	yield	a	reasonable	estimate	of	the	mean	annual	eagle	fatality	rate	for	the	project.		
General	considerations	for	designing	fatality	monitoring	programs	can	be	found	in	Strickland	et	al.	
(2011)	and	the	WEG,	and	these	sources	should	be	consulted	in	the	development	of	a	post‐
construction	study	design.		Because	the	post‐construction	monitoring	protocol	will	be	included	as	a	
condition	of	the	programmatic	take	permit,	the	design	of	such	monitoring	will	be	determined	
jointly	by	the	permittee	and	the	Service.		Additionally,	the	Service	and	USGS	are	investing	significant	
resources	into	research	to	test	and	assess	post‐construction	monitoring	approaches	for	eagles,	thus	
we	expect	to	be	able	to	offer	useful	input	in	the	design	of	such	monitoring	programs.		Fatality	
monitoring	for	eagles	can	be	combined	with	monitoring	mortality	of	other	wildlife	so	long	as	
sampling	intensity	takes	into	account	the	relative	infrequency	of	eagle	mortality	events.	
	
Fatality‐monitoring	efforts	involve	searching	for	eagle	carcasses	beneath	turbines	and	other	
facilities	to	estimate	the	number	of	fatalities.		The	primary	objectives	of	these	efforts	are	to:	(1)	
estimate	eagle	fatality	rates	for	comparison	with	the	model‐based	predictions	prior	to	construction,	
and	(2)	to	determine	whether	individual	turbines	or	strings	of	turbines	are	responsible	for	the	
majority	of	eagle	fatalities,	and	if	so,	the	factors	associated	with	those	turbines	that	might	account	
for	the	fatalities	and	which	might	be	addressed	via	conservation	measures	and	ACPs.	
	
Fatality	monitoring	results	should	be	of	sufficient	statistical	validity	to	provide	a	reasonably	precise	
estimate	of	the	eagle	mortality	rate	at	a	project	to	allow	meaningful	comparisons	with	pre‐
construction	predictions,	and	to	provide	a	sound	basis	for	determining	if,	and	if	so	which,	
conservation	measures	and	ACPs	might	be	appropriate.		The	basic	method	of	measuring	fatality	
rates	is	the	carcass	search.	All	fatality	monitoring	should	include	estimates	of	carcass	removal	and	
carcass	detection	bias	(scavenger	removal	and	searcher	efficiency)	likely	to	influence	those	rates,	
using	the	currently	accepted	methods.		Fatality	and	bias	correction	efforts	should	occur	across	all	
seasons	to	assess	potential	temporal	variation.		Where	seasonal	eagle	concentrations	were	

97
identified	in	the	Stage	2	assessment,	sampling	protocols	should	take	these	periodic	pulses	in	
abundance	into	account	in	the	sample	design.	
	
Carcass	searches	underestimate	actual	mortalities	at	wind	turbines,	but	with	appropriate	sampling,		
carcass	counts	can	be	adjusted	to	account	for	biases	in	detection	(Kunz	et	al.	2007,	Arnett	et	al.	
2007,	NRC	2007,	Huso	2010).		Important	sources	of	bias	and	error	include:	(1)	low	or	highly	
variable	fatality	rates;	(2)	carcass	removal	by	scavengers;	(3)	differences	in	searcher	efficiency;	(4)	
failure	to	account	for	the	influence	of	site	(e.g.,	vegetative)	conditions	in	relation	to	carcass	removal	
and	searcher	efficiency;	and	(5)	fatalities	or	injured	birds	that	may	land	or	move	outside	search	
plots.		Strickland	et	al	(2011)	provide	a	concise	overview	of	fatality	prediction	models	and	
considerations	in	the	selection	of	a	model.		In	the	case	of	eagles,	a	primary	consideration	in	the	
selection	of	a	model	and	in	the	sampling	design	is	the	relative	rarity	of	collisions,	even	at	sites	
where	fatality	rates	are	comparatively	high.	
	
Regardless	of	the	approach	selected,	we	recommend	the	following	data	be	collected	for	each	search:	
1. Date.	
2. Start	time.	
3. End	time.	
4. Interval	since	last	search.	
5. Observer.	
6. Which	turbine	area	was	searched	(including	decimal‐degree	latitude	longitude	or	UTM	
coordinates	and	datum).	
7. Weather	data	for	each	search,	including	the	weather	for	the	interval	since	the	last	search.	
8. GPS	track	of	the	search	path.	
	
When	a	dead	eagle	is	found,	the	following	information	should	be	recorded	on	a	fatality	data	sheet:	
1. Date.	
2. Species.	
3. Age	and	sex	(following	criteria	in	Pyle	2008)	when	possible.	
4. Band	number	and	notation	if	wearing	a	radio‐transmitter	or	auxiliary	marker.	
5. Observer	name.	
6. Turbine	or	pole	number	or	other	identifying	character.	
7. Distance	of	the	carcass	from	the	turbine	or	pole.	
8. Azimuth	of	the	carcass	from	the	turbine	or	pole.	
9. Decimal‐degree	latitude	longitude	or	UTM	coordinates	of	the	turbine	or	pole	and	carcass.	
10. Habitat	surrounding	the	carcass.	
11. Condition	of	the	carcass	(entire,	partial,	scavenged).	
12. Description	of	the	carcass	(e.g.,	intact,	wing	sheared,	in	multiple	pieces).	
13. A	rough	estimate	of	the	time	since	death	(e.g.,	<1	day,	>	a	week),	and	how	estimated.	
14. A	digital	photograph	of	the	carcass.	
15. Information	on	carcass	disposition.	
	
In	some	cases,	eagle	take	permits	may	specify	other	biological	materials	or	data	that	should	be	
collected	from	eagle	carcasses	(e.g.,	feathers,	tissue	samples).		Rubber	gloves	should	be	used	to	
handle	all	carcasses	to	eliminate	possible	disease	transmission.		All	eagle	fatalities	(not	just	those	
found	on	post‐construction	surveys)	and	associated	information	should	be	immediately	reported	to	
the	Service’s	Office	of	Law	Enforcement	and	to	the	Service’s	migratory	bird	permit	issuing	office	if	
the	facility	is	operating	under	an	eagle	take	permit.		Eagle	carcasses	should	not	be	moved	until	such	
notification	occurs,	after	which	carcass	disposition	should	be	in	accordance	with	permit	conditions	
or	Service	direction.	

98

	
2. Disturbance Monitoring
Project	developers	or	operators	may		also	be	required	to	monitor	many	of	the	eagle	nesting	
territories	and	communal	roost	sites	identified	in	the	Stage	2	assessments	as	stated	in	the	permit	
regulations	at	50	CFR	22.26(c)(2)for	at	least	two	years	after	project	construction	and	for	up	to	
three	years	after	the	cessation	of	the	activity.		The	objective	of	such	monitoring	will	be	to	determine	
post‐construction	(1)	territory	or	roost	occupancy	rates,	(2)	nest	success	rates,	and	(3)	
productivity.		On	a	project‐by‐project	basis,	changes	in	any	of	these	reproductive	measures	may	not	
be	indicative	of	disturbance.		However,	patterns	may	become	apparent	when	the	Service	and	USGS	
pool	data	appropriately	and	analyze	findings	from	many	projects	in	the	context	of	a	meta‐analysis	
within	the	adaptive	management	framework.	
	
Eagle	nesting	territories	most	likely	to	be	affected	by	disturbance	from	a	wind	project	are	those	that	
have	use	areas	within	or	adjacent	to	the	project	footprint.		The	Service	will	accept	an	assumption	
that	all	eagle	pairs	at	or	within	the	mean	project‐area	inter‐nest	distance	(as	determined	from	the	
Stage	2	assessment)	of	the	project	boundary	are	territories	that	may	be	at	risk	of	disturbance	(e.g.,	
if	the	mean	nearest‐neighbor	distance	between	simultaneously	occupied	eagle	territories	in	the	
Stage	2	assessment	is	2	miles,	we	would	expect	disturbance	to	most	likely	affect	eagles	within	2	
miles	of	the	project	boundary;	Figures	H‐1	though	H‐4).		Eagle	pairs	nesting	within	½	the	project‐
area	mean	intern‐nest	distance	are	the	highest	candidates	for	disturbance	effects,	and	should	
receive	special	attention	and	consideration.	
	
Where	nesting	habitat	is	patchy	or	eagle	nesting	density	is	low	such	that	nearest‐neighbors	are	
outside	a	10‐mile	wide	perimeter	of	the	project	footprint,	we	recommend	either:	(1)	extending	the	
project‐area	survey	outward	to	include	the	nearest‐neighbors	for	the	purposes	of	estimating	the	
mean	inter‐nest	distance	value,	or	(2)	undertaking	detailed	observational	studies	of	the	eagles	
occupying	territories	within	the	typical	project‐area	to	assess	use	patterns	and	ranging	behavior	
relative	to	the	project	footprint.		We	recognize	that	selecting	option	(1)	for	golden	eagles	would	
extend	the	project	area	beyond	the	maximum	of	10	miles	advocated	in	the	ECPG,	but	in	some	areas	
it	is	possible	golden	eagles	using	nests	further	than	10	miles	from	the	project	footprint	may	occur	
there.		Regardless	of	which	approach	is	used,	territories	that	meet	this	distance	criterion	should	be	
re‐sampled	annually	for	no	less	than	two	years	after	the	project	is	operational	following	identical	
survey	and	reporting	procedures	as	were	used	in	the	Stage	2	assessment.	
	
If	such	monitoring	shows	strong	evidence	of	direct	disturbance	from	a	project,	project	developers	
or	operators	and	the	Service	will	consider	additional	conservation	measures	and	ACPs	that	might	
be	effective	in	reducing	the	effect.		Such	measures	would	be	within	the	sideboards	established	at	
the	time	of	permit	issuance.	Alternatively,	the	project	developer	or	operator	may	be	required	to	
provide	compensatory	mitigation	to	offset	the	estimated	decreases	in	productivity	to	the	extent	
necessary	to	meet	the	statutory	requirement	to	preserve	eagles.	
	
The	Service	and	the	project	developer	or	operator	should	agree	on	a	site‐specific,	post‐construction	
survey	protocol	for	eagle	concentration	areas	identified	in	Stage	2	and	make	an	a	priori	decision	on	
how	to	interpret	and	act	on	potential	outcomes.		Mortalities	of	eagles	using	proximate	communal	
roosts	will	be	accounted	for	through	the	protocol	for	monitoring	post‐construction	fatalities.		
However,	if	communal	roosts	are	no	longer	used	by	eagles	because	of	disturbance,	that	effect	
should	be	determined,	evaluated,	and	where	population‐level	effects	are	indicated,	mitigated.	

	

99

3. Comparison of Post-Construction Eagle Use with Pre-Construction Use
As	noted	elsewhere,	Service	fatality	models	assume	eagle	use	of	the	project	footprint	does	not	
change	as	a	result	of	project	development.		However,	there	is	little	information	to	support	this	
assumption,	and	the	ability	to	accurately	predict	fatality	rates	could	be	greatly	improved	by	
comparative	information	on	post‐construction	eagle	use.		The	Service	encourages	project	
developers	or	operators	to	consider	conducting	exposure	surveys	similar	in	design	and	intensity	to	
pre‐construction	survey	work	to	test	this	assumption	where	and	when	feasible.	

	
	
Literature Cited
Arnett,	E.	B.		2006.		A	preliminary	evaluation	on	the	use	of	dogs	to	recover	bat	fatalities	at	wind	
energy	facilities.	Wildlife	Society	Bulletin	34(5):1440–1445.	
Huso,	M.	M.	P.		2010.		An	estimator	of	wildlife	fatality	from	observed	carcasses.	Environmetrics	DOI:	
10.1002/env.1052.	
Kochert,	M.	N.,	K.	Steenhof,	C.	L.	Mcintyre,	and	E.	H.	Craig.		2002.		Golden	eagle	(Aquila	chrysaetos).	
The	Birds	of	North	America	No.	684	(A.	Poole,	Ed.).	The	Birds	of	North	America	Online.	Cornell	
Lab	of	Ornithology,	Ithaca,	New	York,	USA.	http://bna.birds.cornell.edu/bna/species/684.	
Kunz,	T.	H.,	E.	B.	Arnett,	B.	M.	Cooper,	W.	P.	Erickson,	R.	P.	Larkin,	T.	Mabee,	M.	L.	Morrison,	M.	D.	
Strickland,	and	J.	M.	Szewczak.		2007.		Assessing	impacts	of	wind‐energy	development	on	
nocturnally	active	birds	and	bats:	a	guidance	document.	Journal	of	Wildlife	Management	71:	
2449‐2486.	
National	Research	Council	(NRC).		2007.		Environmental	impacts	of	wind‐energy	projects.	National	
Academies	Press.	Washington,	D.C.,	USA.		www.nap.edu.	
Strickland,	M.D.,	E.B.	Arnett,	W.P.	Erickson,	D.H.	Johnson,	G.D.	Johnson,	M.L.,	Morrison,	J.A.	Shaffer,	
and	W.	Warren‐Hicks.		2011.		Comprehensive	Guide	to	Studying	Wind	Energy/Wildlife	
Interactions.	Prepared	for	the	National	Wind	Coordinating	Collaborative,	Washington,	D.C.,	USA.	

	
	
Figures H-1 to H-4 (following pages). Suggested approach for determining project-area and identifying eagle
nesting territories to monitor for disturbance effects during Stage 5.

100

	

101

	

102

103

	
	
	

	

	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	
	

U.S. Fish and Wildlife Service
Division of Migratory Bird Management
April 2013	


File Typeapplication/pdf
File Modified0000-00-00
File Created0000-00-00

© 2024 OMB.report | Privacy Policy