UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

How is a family of sedentary marine fishes shaped by its habitats, prey, and predators? Manning, Clayton Garin 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2017_november_manning_clayton.pdf [ 3.74MB ]
Metadata
JSON: 24-1.0355229.json
JSON-LD: 24-1.0355229-ld.json
RDF/XML (Pretty): 24-1.0355229-rdf.xml
RDF/JSON: 24-1.0355229-rdf.json
Turtle: 24-1.0355229-turtle.txt
N-Triples: 24-1.0355229-rdf-ntriples.txt
Original Record: 24-1.0355229-source.json
Full Text
24-1.0355229-fulltext.txt
Citation
24-1.0355229.ris

Full Text

	 i	HOW	IS	A	FAMILY	OF	SEDENTARY	MARINE	FISHES	SHAPED	BY	ITS	HABITATS,	PREY,	AND			PREDATORS?				by		Clayton	Garin	Manning				B.Sc.	(First-Class	Hons),	University	of	Calgary,	2012				A	THESIS	SUBMITTED	IN	PARTIAL	FULFILLMENT	OF			THE	REQUIREMENTS	FOR	THE	DEGREE	OF				MASTER	OF	SCIENCE		In		THE	FACULTY	OF	GRADUATE	AND	POSTDOCTORAL	STUDIES		(Zoology)				THE	UNIVERSITY	OF	BRITISH	COLUMBIA		(Vancouver)				August	2017				©Clayton	Garin	Manning,	2017			 		 ii	Abstract		 Overall,	this	thesis	expands	on	our	ecological	understanding	of	a	group	of	biologically	diverse	marine	fishes	by	investigating	how	they	are	shaped	by	their	habitats,	prey,	and	predators.	In	my	first	data	chapter,	I	used	the	seahorse	Hippocampus	whitei	as	a	case	study	for	investigating	the	ecological	correlates	of	syngnathid	abundance	and	distributions.	Expanding	on	research	that	had	looked	at	how	either	their	habitats,	prey,	or	predators	affected	their	populations,	I	considered	all	three	components	in	a	single	holistic	approach.	I	investigated	these	correlations	at	two	scales:	among	different	seagrass	beds	(200-6000	m	apart),	and	within	a	single	seagrass	bed	(<100	m	in	size).	I	found	that	habitat,	prey,	and	predator	variables	all	correlated	with	seahorse	density	or	height	distributions	to	varying	extents,	depending	on	the	scale	of	the	study.	Total	predators	was	negatively	associated	with	seahorse	density	across	seagrass	beds,	the	only	ecological	variable	that	was	correlated	with	seahorses	across	beds.	Within	seagrass	beds,	seahorse	locations	correlated	with	greater	depth,	denser	seagrass,	more	prey	types,	and	fewer	predators.	In	my	second	data	chapter,	I	reviewed	the	diets	and	feeding	behaviour	of	syngnathids,	bringing	together,	summarizing,	and	providing	new	insights	on	a	large	amount	of	fragmented	information	on	the	topic.	I	answered	three	central	questions.	1.	How	do	syngnathids	eat?	2.	How	does	feeding	and	diet	vary	across	a	morphogically	diverse	family	of	fishes?	3.	How	does	feeding	and	diet	vary	across	a	family	of	fishes	that	lives	in	a	three-dimensional	space?	I	answered	question	1	by	summarizing	a	number	of	different	studies	on	the	morphologies	and	kinematics	of	syngnathid	feeding	events.	I	answered	questions	2	and	3	using	a	meta-analysis	on	syngnathid	diets	found	the	literature.	Overall,	I	found	there	to	be	a	large	amount	of	variation	in	syngnathid	diets	that	I	hypothesize	is	caused	by	large		 iii	differences	in	prey	availability.	Of	the	explained	variation,	I	found	their	diets	were	most	strongly	correlated	with	their	relative	snout	lengths	and	gape	sizes.	These	feeding	morphologies	also	had	high	phylogenetic	signal,	suggesting	that	dietary	differences	across	genera	were	largely	explained	by	how	they	differed	with	respect	to	these	morphologies.																																					 iv	Lay	Summary		 Overall,	this	thesis	investigates	how	syngnathids—a	family	of	fish	that	includes	seahorses	and	pipefishes—are	affected	by	their	habitats,	prey,	and	predators.	First,	I	looked	at	how	seahorse	abundance	was	affected	by	habitat,	prey,	and	predators	in	eastern	Australia.	I	found	that	each	of	these	variables	affected	seahorse	density	or	distributions	to	varying	extents,	depending	on	the	scale	of	the	study.	Next,	I	reviewed	syngnathid	feeding,	and	provided	a	detailed	summary	on	how	syngnathids	eat,	and	also	analyzed	what	they	eat.	Overall,	I	found	a	lot	of	variation	in	what	syngnathids	eat,	and	suspect	this	is	because	there	is	a	lot	of	variation	in	what	is	available	to	them	in	their	environments.	I	also	found	their	diets	were	best	explained	by	the	size	and	shapes	of	their	snouts—highly	advanced	body	parts	that	have	evolved	to	help	these	fish	feed	on	fast	prey.													 v	Preface	The	research	questions	and	methodological	design	of	my	thesis	were	developed	in	collaboration	with	my	co-supervisors,	Drs.	Amanda	Vincent	and	Sarah	Foster.	I	collected	all	of	the	data	used	for	Chapter	2	with	the	help	of	my	host	supervisor	in	Australia	Dr.	Dave	Harasti,	and	my	research	assistants.	I	collected	the	data	and	information	used	in	Chapter	3	from	peer-reviewed	and	grey	literature.	I	carried	out	all	analyses,	and	prepared	all	manuscripts	in	this	thesis,	with	substantial	input	from	Amanda	Vincent	and	Sarah	Foster.	A	version	of	Chapter	2	is	in	the	final	stages	of	preparation	for	publication.	As	per	our	target	journal’s	requirements,	the	writing	is	in	passive	voice.	I	am	the	lead	author,	along	with	my	co-supervisors	Drs.	Amanda	Vincent	and	Sarah	Foster,	and	my	host-supervisor	Dr.	Dave	Harasti,	who	provided	guidance,	logistical	support,	and	assisted	with	fieldwork.	I	conducted	all	of	the	research	for	Chapter	2,	and	wrote	the	paper	in	collaboration	with	Drs.	Sarah	Foster	and	Amanda	Vincent.	Alistair	Poore	(University	of	New	South	Wales)	helped	train	my	assistants	and	me	in	crustacean	identification,	helped	mold	the	study	design,	and	provided	statistical	help.	Meagan	Abele	and	Natalie	Scadden,	my	research	assistants,	were	vital	to	the	data	collection	and	safety	management	of	my	project.	A	version	of	Chapter	3	is	in	the	final	stages	of	preparation	for	publication.	I	am	the	lead	author,	along	with	my	co-supervisors	Drs.	Amanda	Vincent	and	Sarah	Foster.	I	conducted	the	analysis	for	Chapter	3	and	wrote	the	paper	with	Drs.	Sarah	Foster	and	Amanda	Vincent.	I	collected	and	tabulated	the	data	used	for	Chapter	3	from	primary	and	grey	literature.		 vi	All	field	research	undertaken	in	Chapter	2	was	done	in	accordance	with	the	University	of	British	Columbia's	Animal	Care	Committee	permit	A12-0288	and	the	NSW	DPI	Animal	Care	and	Ethics	Committee	permit	15/01.																																				 vii	Table	of	Contents	Abstract	....................................................................................................................................................	ii	Lay	Summary	.........................................................................................................................................	iv	Preface	......................................................................................................................................................	v	Table	of	Contents................................................................................................................................vii	List	of	Tables	.........................................................................................................................................	xi	List	of	Figures	......................................................................................................................................	xiv	Acknowledgements	.........................................................................................................................	xvii	Chapter	1	Introduction	.......................................................................................................................	1	1.1	 Rationale	................................................................................................................................	1	1.2	 Background	...........................................................................................................................	1	1.3	 Research	objectives	and	thesis	outline	........................................................................	6	Chapter	2	Ecological	correlates	of	White's	seahorse	(Hippocampus	whitei)	abundance	and	size	distributions	at	different	spatial	scales	.......................................................................	8	2.1	 Introduction	..........................................................................................................................	8	2.2	 Materials	&	Methods	.......................................................................................................	13	2.2.1	 Study	species	...................................................................................................................................	13	2.2.2	 Study	locations	&	design	............................................................................................................	13	2.2.3	 Seahorse	surveys	...........................................................................................................................	17	2.2.4	 Predator	surveys	...........................................................................................................................	18	2.2.5	 Seagrass	surveys	............................................................................................................................	19		 viii	2.2.6	 Prey	surveys	....................................................................................................................................	19	2.2.7	 		Predicting	covariates	within	Little	Beach	.........................................................................	20	2.2.7.1		Seagrass	and	prey	...................................................................................................................	20	2.2.7.2		Predators	....................................................................................................................................	21	2.2.8	 Statistical	analyses	........................................................................................................................	22	2.3	 		Results	................................................................................................................................	25	2.3.1	 		Study	scale:	among	all	seagrass	beds	.................................................................................	25	2.3.1.1		Seahorse	survey	summary	statistics	..............................................................................	25	2.3.1.2		Correlates	of	seahorse	density	&	seahorse	height	....................................................	26	2.3.2	 		Study	scale:	within	Little	Beach	seagrass	bed	................................................................	26	2.3.3	 		Overall	associations	between	seahorses	and	ecological	correlates	......................	26	2.4	 Discussion	...........................................................................................................................	27	Chapter	3	Review	paper:	the	diet	and	feeding	behaviours	of	a	family	of	biologically	diverse	marine	fishes	(Family	Syngnathidae)	.........................................................................	48	3.1	 Introduction	.......................................................................................................................	48	3.2	 Methods	...............................................................................................................................	52	3.2.1	 Literature	review	...........................................................................................................................	52	3.2.2	 Diet	taxonomy	.................................................................................................................................	53	3.2.3	 Syngnathid	characteristics	........................................................................................................	54	3.2.4	 		Statistical	analyses	......................................................................................................................	55	3.2.4.1		Associations	between	syngnathid	characteristics	and	their	diets	.....................	55	3.2.4.2		Phylogenetic	signal	of	syngnathid	characteristics	....................................................	57	3.3	 		Results	................................................................................................................................	58		 ix	3.3.1	 		How	do	syngnathids	eat?	.........................................................................................................	58	3.3.1.1		Head	morphology	&	mechanics	of	a	feeding	event	..................................................	58	3.3.1.2		Stages	of	a	feeding	event	......................................................................................................	59	3.3.1.3		Energetics	of	feeding	.............................................................................................................	61	3.3.1.4		Diurnal	timing	of	feeding	.....................................................................................................	62	3.3.2	 How	does	feeding	and	diet	vary	across	a	speciose	marine	fish	family	that	is			morphologically	diverse?	.........................................................................................................................	63	3.3.2.1	Prey	items	by	genus	and	species	.......................................................................................	63	3.3.2.1.1	Entire	family	......................................................................................................................	63	3.3.2.1.2	Hippocampus	(seahorses)	...........................................................................................	64	3.3.2.1.3	Syngnathus	pipefishes	...................................................................................................	64	3.3.2.1.4	Seadragons	(Phyllopteryx)	and	other	pipefishes	..............................................	65	3.3.2.2	Morphology	.................................................................................................................................	65	3.3.2.2.1	Phylogenetic	signal	of	syngnathid	morphological	characteristics	.............	65	3.3.2.2.2	Body	form	&	orientation	...............................................................................................	66	3.3.2.2.3	Snout	shape	........................................................................................................................	66	3.3.2.2.4	Gape	size	..............................................................................................................................	68	3.3.2.2.5	Ontogenetics:	changes	in	snout	shape	&	gape	size	...........................................	69	3.3.2.3	Sex	&	reproductive	status	.....................................................................................................	71	3.3.3	 How	does	feeding	and	diet	vary	across	a	marine	fish	family	that	lives	in	a	three-dimensional	space?	.....................................................................................................................................	72	3.3.3.1	Variability	....................................................................................................................................	72	3.3.3.2	Tail	morphology	&	foraging	strategies	...........................................................................	73		 x	3.4	 Discussion	...........................................................................................................................	76	Chapter	4	Conclusions	...................................................................................................................	100	4.1		 Associations	with	habitats	..........................................................................................	101	4.2		 Associations	with	prey	.................................................................................................	101	4.3		 Associations	with	predators	.......................................................................................	103	4.4		 How	this	thesis	fits	in	to	the	literature	...................................................................	105	Bibliography	......................................................................................................................................	108	Appendices	........................................................................................................................................	141	Appendix	A:	The	determination	of	seahorse	sex,	maturity,	and	reproductive	status.	...........................................................................................................................................................	141	Appendix	B:	Tables	to	support	methods	and	results	in	Chapter	2	.............................	143	Appendix	C:	Tables	to	support	methods	and	results	in	Chapter	3	.............................	146																		 xi	List	of	Tables		Table	2.1	Summary	statistics	for	seahorse	(SH)	surveys	at	seven	plots	in	the	Port	Stephens	estuary,	NSW,	Australia,	during	the	November	(Nov)	and	February	(Feb)	sampling	campaigns.	SE	is	the	standard	error	of	the	mean.	Overall	density	represents	the	total	number	of	seahorses	divided	by	the	total	area	searched.	For	adult	height,	total	represents	the	mean	of	all	adults	pooled	across	plots.	February	Little	Beach	physical	maturity	ratios	do	not	add	to	total	number	of	seahorses	found	because	the	sex	of	one	individual	was	not	determined,	and	this	individual	was	not	included	in	any	other	calculations.	Plots	are	listed	from	smallest	to	largest	distance	from	the	estuary	mouth.	..............................................................................................................................................................................	35	Table	2.2	Summary	of	the	model-averaged	statistics	for	the	top	models	predicting:	(a)	seahorse	density	among	seagrass	beds,	(b)	adult	height	among	seagrass	beds,	(c)	and	resource	selection	function	within	the	Little	Beach	seagrass	bed.	LL	=	log-likelihood,	AICc	=	corrected	Akaike	information	criterion,	ΔAICc	=	difference	in	model	AICc	with	that	of	the	top	model,	wi	=	Akaike	weight,	df	=	number	of	model	parameters	including	intercepts	and	residuals.	The	following	abbreviations	have	been	made:	DPTH	=	depth,	DENS	=	seagrass	density,	HGHT	=	seagrass	height,	TPT	=	prey	types,	TPI	=	prey	density,	FLNG	=	fouling,	PRED	=	total	predators,	and	TPRED	=	types	of	predators.	......	36	Table	2.3	Model-averaged	parameter	estimates,	standard	error	(SE)	of	the	parameter,	correlate	relative	importance,	the	upper	and	lower	90%	parameter	confidence	intervals	(CI)	for	variables	predicting	resource	selection	function	within	the	Little	Beach	seagrass	bed.	....................................................................................................................................	37		 xii	Table	2.4	Summary	of	relationships	between	seagrass,	prey,	and	predator	covariates	with	(a)	seahorse	(SH)	density	and	(b)	adult	height	among	seagrass	beds,	(c)	resource	selection	function	(RSF)	within	the	Little	Beach	seagrass	bed	and	(d)	SH	height	within	Little	Beach,	among	SH	with	different	sexes	and	reproductive	statuses.	RA	=	reproductively	active.	................................................................................................................................	38	Table	3.1	Relative	importance	of	syngnathid	diets.	a	The	approximate	area	of	sampling.	For	comparative	purposes,	studies	in	close	proximity	(within	50	km	of	each	other)	have	the	same	location.	b	Bulk	dietary	studies	include	relative	values	that	each	food	item	contributes	to	the	total	volume	(%V),	weight	(%W),	or	area	(%A)	of	dietary	contents	collected,	numeric	dietary	studies	include	relative	values	that	each	food	item	contributes	to	the	total	number	of	food	items	(%N)	collected,	and	frequency	of	occurrence	(%FO)	studies	include	the	relative	number	of	stomach	samples	that	a	particular	food	item	occurs	in.	Each	row	represents	a	particular	species	in	a	particular	area	of	a	particular	study.	c	This	table	includes	additive	data,	so	if	a	taxon	is	centered	above	other	taxa	it	includes	those	numbers	in	its	total.	Numbers	are	added	to	columns	on	the	left	(e.g.	Crustacea	includes	Paracarida	and	Eucarida).	Sample	sizes	of	two	and	under	were	not	considered	in	statistical	analyses.	Blank	cells	represent	missing	data.	Shaded	cells	(for	%FO	data	only)	indicate	the	value	of	that	cell	was	not	provided	in	the	literature,	and	is	a	minimum	value	based	on	dietary	items	that	were	included	at	a	lower	taxonomic	resolution	(see	text)................................................................................................	84	Table	3.2	Measures	of	phylogenetic	signal	for	syngnathid	morphological	characteristics.	Statistical	significance	indicated	in	bold	(P	<	0.05).	N	indicates	the	number	of	syngnathid	species	that	were	used	to	assess	the	phylogenetic	signal.	.................................	94		 xiii	Table	3.3	Results	of	the	redundancy	analyses	that	measured	the	associations	between	multiple	independent	variables	(syngnathid	morphological	characteristics)	and	multiple	dependent	variables	(dietary	categories,	e.g.	amphipods).	Models	were	based	on	three	different	dietary	metrics:	bulk,	numeric,	and	frequency	of	occurrence	dietary	data.	Statistical	significance	indicated	in	bold	(P	<	0.05).	..........................................................	95																		 xiv	List	of	Figures	Figure	2.1	Location	of	the	seven	study	plots	along	the	southern	coast	of	the	Port	Stephens	estuary,	on	the	eastern	coast	of	New	South	Wales,	Australia.	Plot	names:	(1)	Shoal	Bay,	(2)	Little	Beach,	(3)	Fly	Point,	(4)	Seahorse	Gardens	1,	(5)	Seahorse	Gardens	2,	(6)	Pipeline,	(7)	Dutchies.	Shore	Data:	OpenStreetMap	(and)	contributors,	CC-BY-SA.	......	39	Figure	2.2	Map	of	the	Little	Beach	seagrass	beds	showing	an	example	of	the	stratified	random	sampling	design	used	at	each	of	the	seven	study	plots	(for	study	among	seagrass	beds).	Light	grey	represents	P.	australis.	White	background	includes	all	other	benthic	substrate	types,	predominantly	sand	and	Zostera	sp.	The	8	m	length	of	the	stratum	was	calculated	by	dividing	the	24	m	length	of	the	seagrass	bed	into	thirds.	Shore	Data:	OpenStreetMap	(and)	contributors,	CC-BY-SA.	.....................................................	41	Figure	2.3	Mean	(±	SE)	number	of	(a)	total	predators,	and	(b)	different	predator	species	observed	at	fixed	transects	at	Little	Beach,	measured	19-21	January	2016.	Hashed	polynomial	lines	of	best	fit	are	included:	(a)	y	=	0.0029x2	-	0.3473x	+	11.024	(r2	=	0.87),	(b)	y	=	0.0009x2	-	0.0849x	+	2.3039	(r2	=	0.96).	................................................................	42	Figure	2.4	(a)	Mean	seahorse	density	(±	SE),	(b)	mean	adult	seahorse	height,	(c)	proportion	of	seahorses	that	were	physically	mature	(adult),	and	(d)	proportion	of	adults	that	were	male,	by	plot	and	by	November	and	December	survey	campaigns.	Values	at	the	base	of	bars	indicate	the	number	of	adult	seahorses	(a,	d)	and	the	total	number	of	seahorses	(c).	Bars	sharing	a	common	letter	do	not	differ	significantly	(Tukey	HSD,	P	>	0.05)	................................................................................................................................	43	Figure	2.5	Model-averaged	effect	sizes	with	90%	confidence	intervals	(CI)	for	predictor	variables	of	seahorse	(a)	density	and	(b)	adult	height,	among	seagrass	beds.	Outputs		 xv	based	on	results	of	mixed-effects	models.	Parameter	estimates	are	indicated	to	the	right	of	the	CIs.	P-values	are	indicated	as	follows:	superscript	a	=	P	>	0.10,	*	=	0.10	<	P	<	0.05,	**	=	P	<	0.05.	....................................................................................................................................	45	Figure	2.6	Model-averaged	effect	sizes	with	90%	confidence	intervals	(CI)	for	predictor	variables	of	seahorse	height	within	the	Little	Beach	seagrass	bed,	among	(a)	all	seahorses,	(b)	all	reproductively	active	(RA)	seahorses,	(c)	females,	(d)	RA	females,	(e)	males,	(f)	RA	males.	Outputs	based	on	results	of	mixed-effects	models.	Parameter	estimates	are	indicated	to	the	right	of	the	CIs.	P-values	are	indicated	as	follows:	superscript	a	=	P	>	0.10,	*	=	0.10	<	P	<	0.05,	**	=	P	<	0.05.	Total	N	=	328;	includes	168	female,	148	male,	and	12	juvenile	(reproductively	inactive)	observations.	......................	47	Figure	3.1	Proportion	of	(a)	bulk	dietary	data	variance,	(b)	numeric	dietary	data	variance,	and	(c)	frequency	of	occurrence	dietary	variance	explained	by	syngnathid	body	traits	[component	a;	Relative	fin	size	+	Max.	standard	length	(StL)	+	Snout	depth	(SnD)	+	HL:SnL	+	SnL:SnD],	genus	[component	c],	covariance	between	syngnathid	body	traits	and	genus	[component	b],	and	unexplained	residuals	[component	d].	...............................	96	Figure	3.2	Figure	1	from	©	Van	Wassenbergh,	S.,	Strother,	J.	A.,	Flammang,	B.	E.,	Ferry-Graham,	L.	A.	&	Aerts,	P.,	Extremely	fast	prey	capture	in	pipefish	is	powered	by	elastic	recoil,	Journal	of	the	Royal	Society	Interface,	2008,	5,	20,	page	286,	by	permission	of	the	Royal	Society.	A	schematic	of	a	syngnathid	body	(Syngnathus	leptorhynchus)	during	a	feeding	strike.	Specialized	sternohyoideus	muscles	run	along	the	dorsal	and	ventral	sides	of	the	pipefish.	When	contracted,	they	pull	the	hyoid	arch	towards	the	body.	The	neurocranium	(including	the	snout)	then	rotates	away	from	the	body,	towards	the	prey.	.........................................................................................................................................	97		 xvi	Figure	3.3	Biplot	of	the	first	two	axes	of	the	redundancy	analyses	(RDA)	performed	on	bulk	dietary	data.	Points	represent	the	diet	of	a	particular	species	of	syngnathid	in	a	particular	area,	as	reported	by	a	particular	study.	Points	are	coloured	based	on	the	genus.	Environmental	vectors	for	syngnathid	body	characteristics	are	fit	onto	the	ordination,	and	the	direction	and	strength	of	the	gradient	is	represented	by	the	length	of	the	arrow.	...................................................................................................................................................	98	Figure	3.4	Biplot	of	the	first	two	axes	of	the	redundancy	analyses	(RDA)	performed	on	bulk	dietary	data.	Points	represent	the	diet	of	a	particular	species	of	syngnathid	in	a	particular	area,	as	reported	by	a	particular	study.	Points	are	coloured	based	on	the	genus.	Environmental	vectors	for	syngnathid	body	characteristics	are	fit	onto	the	ordination,	and	the	direction	and	strength	of	the	gradient	is	represented	by	the	length	of	the	arrow.	...................................................................................................................................................	99															 xvii	Acknowledgements	First	and	foremost,	I	would	like	to	thank	my	super-supportive	and	uber-helpful	supervisors,	Drs.	Amanda	Vincent	and	Sarah	Foster.	I	appreciate	the	belief	you	had	in	my	abilities,	and	for	giving	me	the	incredible	opportunity	to	operate	my	own	field	season	in	the	most	beautiful	field-site	possible,	Australia.	It	was	a	once	in	a	lifetime	experience	that	was	challenging,	rewarding,	and	a	lot	of	fun.	I	would	also	like	to	thank	Drs.	Diane	Srivastava	and	John	Richardson,	my	supervisory	committee,	for	the	constructive	discussions	that	led	to	my	project,	the	amazing	statistical	advice	that	turned	it	from	a	bunch	of	numbers	into	useful	data,	and	for	helping	me	refine	these	numbers	into	a	manuscript.			My	field	season	would	not	have	even	happened	were	it	not	for	Dr.	Dave	Harasti,	who	invited	me	to	his	field	station	in	Australia,	provided	me	with	everything	I	needed	to	succeed,	helped	with	fieldwork,	and	taught	me	everything	there	is	to	know	about	running	a	field	study.	Also,	a	huge	thanks	to	colleagues	at	the	NSW	Department	of	Primary	Industries,	who	provided	logistical	assistance	and	excellent	hospitality	in	Australia.	My	work	in	Australia	would	not	have	been	possible	without	my	fellow	Team	Sea	Stallion	members,	Meagan	Abele,	and	Natalie	Scadden.	Their	tremendous	assistance	in	the	water	kept	me	safe,	their	upbeat	attitudes	kept	me	sane,	and	their	tireless	work	ethic	allowed	me	to	collect	as	much	data	as	physically	possible.	I	would	like	to	acknowledge	Alistair	Poore	for	help	with	the	study	design	and	statistical	methods,	and	Tyler	Armitage	for	GIS	assistance.	This	thesis	is	a	contribution	from	Project	Seahorse.	The	fieldwork	was	financed	by	the	Natural	Sciences	and	Engineering	Research	Council	of	Canada	(NSERC),	the	Point	Defiance	Zoo	&	Aquarium,	and	Guylian	Chocolates	Belgium	through	its	partnership	for	marine	conservation	with	Project	Seahorse.		 xviii	The	past	three	years	have	been	a	crazy	journey,	and	made	better	by	the	people	who	supported	me,	and	laughed	with	me,	day	in	and	day	out.	Kyle	Gillespie,	I	can’t	believe	we	managed	to	pay	off	a	blimp	AND	build	a	world-famous	brewery.	Officelympics	champion	(disputed),	Ally	Stocks,	your	ability	to	brighten	a	day	is	unmatched.	Ravi	Maharaj,	thank	you	for	teaching	me	how	to	play	(football)	and	hanging	out	with	me	deep	into	nights	of	working	late.	To	the	rest	of	the	PS	family	during	my	tenure—Lindsay	Aylesworth,	Julia	Lawson,	Ting-Chun	Kuo,	Tanvi	Vaidyanathan,	Scott	Finestone,	Gina	Bestbier,	Tyler	Stiem,	Riley	Pollom,	Jenny	Selgrath,	Xong	Zhang,	Emilie	Stump,	Iwao	Fuji,	and	Lily	Stanton—thank	you	for	always	bringing	a	smile	to	my	face,	even	when	dealing	with	daunting	challenges.	Lastly	(but	definitely	not	leastly),	I	would	like	to	thank	my	friends	and	family	for	the	support	through	three	incredible	years,	and	one	of	the	most	difficult	of	my	life.	Mom,	Dad,	Jenay	and	Rylan—you	are	my	backbone,	and	my	rock.	I	don’t	know	what	I	did	to	deserve	you	guys,	but	I’m	damn	lucky,	and	a	far	better	person	because	of	it.	To	my	everyone	else	that	I	met	and	made	me	laugh—either	at	Zoology	Beers,	rehearsing	for	Huts,	or	somewhere	in-between—thanks.																 xix															 To	my	parents,	Kirk	and	Marilynne	 1	Chapter	1 Introduction		1.1 Rationale		 In	this	thesis,	I	explore	ecological	relationships	of	an	unusually	sedentary	family	of	marine	fishes—the	syngnathids.		All	animals	shape	and	are	shaped	by	their	interactions	with	their	habitats,	prey,	and	predators	they	encounter	during	their	lifespan,	as	they	try	to	survive,	grow,	and	reproduce.	The	specific	interactions	that	an	animal	experiences	can	vary	—but	as	a	whole	they	will	affect	the	animal’s	abundance	and	distribution	(Sih	et	al.,	1985;	Abrams,	2000;	Krebs,	2009).				1.2 Background		 An	animal	is	shaped	by	its	experiences	with	biological	and	physical	challenges,	and	the	costs	and	benefits	to	its	survival.	Perhaps	an	animal’s	most	important	interaction	is	with	its	habitat,	which	is	a	direct	and	indirect	source	of	resources,	and	can	mediate	the	interactions	it	has	with	other	members	of	community	(Orth	et	al.,	1984;	Heck	&	Crowder,	1991;	Canion	&	Heck,	2009).	Seagrasses,	for	example,	provide	a	direct	source	of	food	for	herbivorous	invertebrates	and	fish	(Jones	et	al.,	1994,	1997;	Bruno	&	Bertness,	2001).	In	response,	herbivores	modify	their	distributions	according	to	the	distribution	of	seagrasses	(Fretwell	&	Lucas,	1970;	Kennedy	et	al.,	1993),	and	in	doing	so,	can	become	a	concentrated	source	of	food	for	predatory	fish	(Stoner,	1980;	Bell	&	Westoby,	1986;	Jenkins	et	al.,	2002).	In	this	way	habitat	complexity	brings	different	species	together,	thereby	facilitating	interactions	(i.e.	predator-prey,	competition	and	parasitism)	among	the	organisms	it	supports	(Briand	&	Cohen,	1987;	Cohen	et	al.,	1990;	Cohen,	1994;	Vander	Zanden	&	Fetzer,		 2	2007).	In	addition	to	facilitating	interactions,	habitats	can	also	mediate	interactions	among	community	members—such	as	providing	some	members	with	refuge	from	predation,	and	others	with	a	physical	structure	from	which	to	ambush	prey	(Heck	&	Crowder,	1991;	Jones	et	al.,	1994,	1997;	Bruno	&	Bertness,	2001).	Overall,	habitats	create	and	maintain	environmental	conditions	that	are	more	favorable	to	life,	increasing	the	fitness	of	associated	species	and	allowing	for	greater	species	diversity	(Jones	et	al.,	1997;	Bruno	&	Bertness,	2001).	In	doing	so,	habitats	affect	the	distribution	and	population	size	of	both	prey	and	predators	alike	(Heck	&	Crowder,	1991;	Jones	et	al.,	1994,	1997;	Bruno	&	Bertness,	2001).				In	their	efforts	to	survive,	grow,	and	reproduce,	animals	can	affect	the	abundance	and	distribution	of	other	species	(Sih	et	al.,	1985;	Abrams,	2000).	All	predators	depend	on	their	prey	as	an	energy	source,	with	prey	physically	shaping	the	morphology	and	physiology	of	predators	(Schoener,	1971;	West	et	al.,	1991;	Norton,	1991).	In	return,	of	course,	prey	are	shaped	by	their	predators	through	selection	on	traits	that	reduce	capture.	In	exploiting	a	food	source	for	their	own	needs,	predators	are	a	direct	source	of	mortality	on	the	prey	population,	and	reduce	prey	abundance	in	the	process	(Sih	et	al.,	1985).	In	addition,	predators	may	indirectly	shape	the	distribution	of	their	prey	by	inducing	behavioural	changes,	as	the	prey	try	to	find	ways	of	reducing	predation	pressure	(Sih,	1984;	Lima	&	Dill,	1990;	Lima,	1998;	Werner	&	Peacor,	2003).	Predators	and	prey	are	in	an	ever-evolving	arms	race	with	consequences	for	both.	Predation	can	bring	about	the	evolution	of	defensive	traits	among	the	prey	population,	such	as	the	development	of	defensive	toxins,	morphology	that	aids	in	rapid	predator	evasion,	or	behavioural	changes	(Sih,	1984;	Lima	&	Dill,	1990;	Lima,	1998;	Werner	&	Peacor,	2003).	Each	of	these	changes		 3	has	consequences	for	the	predators,	who	must	find	ways	to	counter	them	by	changing	their	search,	capture,	and	consumption	of	prey	(Schoener,	1971;	West	et	al.,	1991;	Norton,	1991).	Predators	might	be	expected	to	select	their	prey	according	to	the	energetic	costs	and	benefits	associated	with	consuming	them.			Sedentary	predators	could	be	expected	to	have	particularly	strong	relationships	with	their	habitats,	prey,	and	predators	because	they	are	long-term	residents	in	one	geographic	area.	The	more	sedentary	an	animal,	the	less	likely	it	is	to	disperse	to	locations	with	more	favourable	conditions,	and	the	more	dependent	it	therefore	becomes	on	local	environmental	conditions	(Steffan-Dewenter	et	al.,	2002;	Öckinger	et	al.,	2009).	For	example,	while	more	mobile	predators	are	able	to	seek	out	their	preferred	prey	availability,	sedentary	ambush	predators	are	dependent	on	the	local	suite	of	prey	available	to	them.	If	local	prey	conditions	are	more	stable,	predators	may	become	reliant	on	a	certain	group	of	prey,	an	effect	that	may	be	reinforced	by	the	evolution	of	specialized	traits	that	aid	in	their	exploit	(Schoener,	1971;	Dawkins	&	Krebs,	1979;	Abrams,	2000).	Alternatively,	conditions	where	the	number	and	suite	of	prey	available	is	under	flux	may	favour	the	development	of	a	more	generalist	palate	(Bommarco	et	al.,	2010).	Sedentary	species	have	similar	limitations	when	it	comes	to	poor	habitat	or	predator	conditions,	and	it	may	be	that	these	species	have	to	make	the	best	of	the	local	situation.			The	marine	environment	is	so	highly	connected—with	so	much	movement	of	organisms	and	so	many	fluid	influences	on	habitats,	along	with	lower	costs	of	movement—that	sedentary	marine	species	might	have	looser	relationships	with	their	environments	than	terrestrial	species.	Aquatic	environments	are	defined	by	the	liquid	medium,	which	provides	organisms	with	physical	structure	to	survive,	grow,	and	reproduce	(Thorne-	 4	Miller,	1999).	As	a	result,	aquatic	environments	are	able	to	support	more	trophic	levels	and	longer	food	chains	than	two-dimensional	terrestrial	habitats	(Briand	&	Cohen,	1987;	Vander	Zanden	&	Fetzer,	2007).	Overall,	marine	animals	interact	with	a	greater	number	of	species—both	within	and	among	trophic	levels—than	those	in	terrestrial	or	freshwater	habitats	(Cohen	et	al.,	1990;	Cohen,	1994).	For	sedentary	marine	predators,	these	interactions	may	not	necessarily	be	with	other	resident	species.	Marine	currents	are	responsible	for	high	rates	of	flux,	dispersing	organisms,	energy,	and	nutrients	from	habitat	to	habitat	(Hixon	&	Menge,	1991;	Palmer	et	al.,	1996;	Cowen	&	Sponaugle,	2009;	McManus	&	Woodson,	2012).	Planktonic	crustaceans,	for	example—an	important	food	source	for	numerous	marine	fish—move	largely	by	means	of	marine	currents	(Palmer	et	al.,	1996;	McManus	&	Woodson,	2012).	Such	processes	can	result	in	large	prey	or	nutrient	differences	at	small	spatial	or	temporal	scales	(Menge	&	Olson,	1990;	Palmer	et	al.,	1996).	The	result	is	that	sedentary	marine	predators	may	be	fairly	generalist,	being	better	suited	to	handle	a	large	number	of	conditions	than	more	mobile	species.		Syngnathids	are	a	diverse	family	of	sedentary	marine	predators	whose	relationships	with	predators,	prey	and	habitats	are	poorly	understood.	Overall,	no	work	has	explored	the	simultaneous	influences	of	habitats,	prey,	and	predators	on	their	abundance	and	distribution.	Research	on	associations	between	syngnathids	and	their	habitats	has	primarily	focused	on	habitat	preferences	(e.g.	Howard	&	Koehn,	1985;	Rosa	et	al.,	2007;	Harasti	et	al.,	2014a),	or	has	not	considered	habitat	variables	together	with	other	potential	factors,	such	as	their	communities	(Curtis	&	Vincent,	2005;	Caldwell	&	Vincent,	2012;	Aylesworth	et	al.,	2015).	While	much	is	known	about	where,	when,	and	on	what	syngnathids	feed	(e.g.	Howard	&	Koehn,	1985;	Bergert	&	Wainwright,	1997;	Kendrick	&		 5	Hyndes,	2005),	the	role	that	prey	have	on	where	we	find	syngnathids,	and	in	what	abundance,	is	not.	Likewise,	we	know	very	little	about	the	effects	that	predators	actually	have	on	populations	of	syngnathids.	Despite	two	recent	studies	that	have	identified	some	marine	animals	that	consume	syngnathids	(Kleiber	et	al.,	2011;	Harasti	et	al.,	2014b),	no	study	has	looked	at	how	predators	matter	to	syngnathids	in	the	context	of	a	complex	marine	environment,	with	numerous	confounding	variables	that	need	to	be	considered.	Given	that	syngnathids	show	strong	site	fidelity	(habitat),	and	are	also	at	the	nexus	of	bottom-up	resource	(prey)	effects	and	top-down	predatory	effects	(predators)	–	we	must	consider	all	sides	of	the	story	to	fully	understand	what	matters	to	syngnathid	populations.		To	date,	there	has	been	no	work	on	general	feeding	patterns	among	syngnathids	across	genera,	locations,	and	studies.	This	is	despite	a	number	of	studies	looking	at	syngnathid	diets	in	the	wild	(e.g.	Steffe	et	al.,	1989;	Franzoi	et	al.,	2004;	Kendrick	&	Hyndes,	2005;	Gurkan	et	al.,	2011a,	2011b),	and	many	others	that	have	contributed	to	an	understanding	of	their	feeding	kinematics	(e.g.	Bergert	&	Wainwright,	1997;	Van	Wassenbergh	et	al.,	2009,	2011)	and	foraging	tactics	(e.g.	Howard	&	Koehn,	1985;	Ryer,	1988;	Tipton	&	Bell,	1988).	Syngnathids	exhibit	large	variations	in	feeding-related	morphology	and	behaviour	across	genera	(Howard	&	Koehn,	1985;	Kendrick	&	Hyndes,	2005),	which	better	equip	them	to	capture	the	prey	that	they	are	more	likely	to	encounter	(Van	Wassenbergh	et	al.,	2008,	2009;	Roos	et	al.,	2009b).	These	morphological	and	behavioural	adaptations	offer	a	distinct	opportunity	to	investigate	the	function	that	syngnathids’	characteristics	have	on	their	foraging	and	diets.	In	my	thesis,	I	seek	to	draw	an	unusually	broad	picture	of	syngnathid	ecology.	I	was	motivated	to	understand	what	components	of	a	syngnathid’s	world—including	their		 6	habitats,	prey,	and	predators—affect	their	abundance	and	distributions.	My	aim	was	to	build	on	previous	seahorse	habitat	work	that	has	investigated	the	associations	between	seahorses	and	either	habitat,	prey,	or	predator	variables—and	approach	it	from	a	holistic	perspective—considering	elements	of	all	three	groups	of	variables	at	once.	I	also	wanted	to	compare	and	contrast	these	associations	at	different	scales,	to	see	if	the	variables	that	correlate	with	syngnathids	within	a	site	also	correlate	among	different	plots.	While	conducting	a	literature	review	to	prepare	myself	for	this	work,	I	noticed	that	there	were	a	large	number	of	interesting,	yet	very	different	studies	related	to	syngnathid	feeding—and	that	this	information	had	not	been	summarized,	or	reviewed	in	any	way.	I	became	keen	to	seek	generalities	(through	comparative	analyses)	about	syngnathid	feeding	across	genera,	locations,	and	studies.	Specifically,	I	was	interested	in	understanding	the	associations	between	syngnathid	characteristics	(in	particular,	morphology)	and	their	foraging	strategies,	and	diets.	My	interest	was	heightened	by	the	opportunity	to	compare	across	300	species	and	57	genera,	in	a	family	with	diverse	morphologies	and	habitats.			1.3 Research	objectives	and	thesis	outline		My	research	focused	on	answering	five	questions:	1. What	habitat,	prey,	and	predator	variables	are	associated	with	seahorse	abundance	and	size	distributions	among	different	seagrass	beds	in	an	Australian	estuary?	(Chapter	2)	2. What	habitat,	prey,	and	predator	variables	are	correlated	with	seahorse	distribution	in	a	single	seagrass	bed?	(Chapter	2)	3. How	do	syngnathids	eat?	(Chapter	3)		 7	4. How	does	feeding	and	diet	vary	across	a	speciose	marine	fish	family	that	is	morphologically	diverse?	(Chapter	3)	5. How	does	feeding	and	diet	vary	across	a	marine	fish	family	that	lives	in	a	three-dimensional	space?	(Chapter	3)		 Chapter	2	expands	on	previous	seahorse	habitat	studies	by	quantifying	seahorse	densities	and	size	distributions	and	looking	at	correlations	between	these	response	variables	and	habitat,	prey,	and	predator	variables.	I	look	at	these	relationships	both	at	a	medium	scale	(0.2-6	km;	among	different	seagrass	beds)	and	a	small	scale	(>	100	m;	within	a	single	seagrass	bed).	My	data	came	from	my	own	field	work	on	the	east	coast	of	Australia.	Chapter	3	presents	the	first	synthetic	overview	of	syngnathid	diet	and	feeding.	Although	much	has	been	published	on	the	diets,	kinematics	and	foraging	of	syngnathids,	there	has	been	no	effort	to	synthesize	this	information	into	a	useful	resource,	or	identify	potential	patterns.	First,	I	synthesize	information	on	the	mechanisms	and	morphologies	involved	in	syngnathid	feeding	strikes.	Next,	I	scoured	peer-reviewed	and	grey	literature	and	performed	a	meta-analysis	that	investigates	the	associations	between	syngnathid	diets	and	their	feeding	morphologies,	and	habitat	associations.	I	present	this	meta-analyses	in	the	context	of	what	is	known	about	syngnathid	diets	and	feeding	behaviours	to	date.	Finally,	I	end	with	a	general	discussion	on	my	findings,	and	how	they	help	answer	my	research	questions	(Chapter	4).			 8	Chapter	2 Ecological	correlates	of	White's	seahorse	(Hippocampus	whitei)	abundance	and	size	distributions	at	different	spatial	scales		2.1	 Introduction	Numerous	living	and	non-living	variables	can	shape	where	organisms	can	be	found,	and	in	what	abundance.	For	example,	extreme	abiotic	conditions	(e.g.	current	velocity,	water	chemistry,	turbidity)	or	poor	resource	availability	in	an	area	can	exclude	organisms	(Whittaker	et	al.,	1973;	Bruno	&	Bertness,	2001).	Likewise,	physical	components	of	an	organism's	habitat,	such	as	the	size	of	refuges	or	degree	of	habitat	complexity	can	determine	if	a	prey	species	is	able	to	persist	under	the	threat	of	predation	(Orth	et	al.,	1984;	Heck	&	Crowder,	1991;	Canion	&	Heck,	2009).	Interactions	among	members	of	a	community	(i.e.	predation,	competition	and	parasitism)	also	have	direct	and	indirect	impacts	on	the	survival	of	multiple	species	(Sih	et	al.,	1985;	Pace	et	al.,	1999).	Beyond	this,	organisms	can	indirectly	affect	the	distribution	of	each	other	by	inducing	behavioural	changes	(e.g.	to	escape	detection	by	predators;	Lima	&	Dill,	1990;	Werner	&	Peacor,	2003;	Preisser	et	al.,	2005).	The	way	that	these	relationships	play	out	is	determined	by	characteristics	of	the	organisms	involved,	the	mediating	effect	of	the	habitat	in	which	they	live,	and	the	interactions	among	these	biotic	and	abiotic	variables.	Research	on	animal	abundances	and	distributions	to	date,	however,	has	often	focused	on	habitat,	prey,	or	predator	variables	individually,	rather	than	components	of	all	three.	Habitats	play	a	direct	role	in	the	life	of	an	organism	by	providing	the	physical	structure,	and	indirectly	supplying	resources	that	are	needed	for	survival,	growth	and	reproduction.	Many	habitats	and	their	associated	communities	are	structured	by	the	presence	of	one	or	more	foundation	species—organisms	that	often	form,	and	have	strong		 9	effects	on	the	functioning	of	a	community	(Dayton,	1972;	Bruno	&	Bertness,	2001).	Foundation	species	can	indirectly	provide	necessary	resources,	can	be	a	direct	source	of	food	themselves,	can	provide	community	members	with	refuge	from	predation,	and	can	reduce	the	physiological	stress	for	other	species	(Jones	et	al.,	1994,	1997;	Bruno	&	Bertness,	2001).	Habitats	with	greater	structural	complexity	have	been	shown	to	reduce	predator	foraging	success	(Orth	et	al.,	1984;	Heck	&	Crowder,	1991;	Canion	&	Heck,	2009)	and	increase	prey	populations	(Heck	&	Orth,	2006),	such	that	many	species	use	complex	habitats	for	their	reproduction	(Kroon	et	al.,	2000;	Sancho	et	al.,	2000;	Gladstone,	2007)	and	nursery	grounds	(Heck	et	al.,	2003).	In	a	global	meta-analysis,	juvenile	fish	and	aquatic	invertebrates	experienced	greater	survival	and	growth	rates	in	complex	habitats,	such	as	seagrass	beds,	than	in	those	with	less	structural	complexity	(Heck	et	al.,	2003).	While	much	research	has	looked	at	how	habitat	variables	affect	species	abundances	and	distributions,	those	species’	relationships	with	their	communities	are	more	complex,	and	less	understood.	The	relationships	among	species	in	a	community	greatly	affect	their	abundances	and	distributions	(Paine,	1966).	The	relative	abundance	of	predators	and	prey	are	intimately	linked,	as	predators	kill	prey	and	convert	a	portion	of	their	biomass	into	new	predators	(Arditi	&	Ginzburg,	1989;	McCauley	et	al.,	1993;	Abrams	&	Ginzburg,	2000).	Without	prey	biomass	to	consume,	however,	predator	numbers	will	decrease	(Arditi	&	Ginzburg,	1989;	Abrams	&	Ginzburg,	2000).	Predator-prey	relationships	affect	the	distributions	of	both	species,	as	each	adjusts	its	behaviours	to	improve	its	own	fitness	(Sih	et	al.,	1985;	Lima	&	Dill,	1990;	Abrams,	2000).	For	example,	as	a	result	of	favourable	abiotic	conditions,	habitat	edges	frequently	have	higher	concentrations	of	herbivores	than	habitat		 10	interiors	(Ries	et	al.,	2004).	This	can	lead	to	higher	predator	abundances	as	they	adjust	their	distribution	to	match	their	prey	(Fretwell	&	Lucas,	1970;	Kennedy	et	al.,	1993;	Ries	et	al.,	2004).	In	response,	prey	make	life-history	trade-offs	and	may	sacrifice	the	improved	resources	and	relocate	to	refuges	or	alternative	locations	to	avoid	predation	(Sih,	1984;	Lima	&	Dill,	1990;	Lima,	1998;	Werner	&	Peacor,	2003).	When	investigating	organisms	such	as	sedentary	species,	which	both	consume	others	and	are	consumed	themselves,	it	is	particularly	vital	to	consider	both	their	prey	and	their	predators.	Sedentary	species	might	be	expected	to	have	especially	strong	relationships	with	their	particular	habitat	and	community,	being	shaped	and	adapted	to	them.	Species	that	are	sedentary	and	rely	on	crypsis	are	largely	limited	to	habitats	that	provide	suitable	refuge	or	substrate	for	cryptic	opportunities	(Ruxton	et	al.,	2004),	as	leaving	a	safe	space	to	forage	or	mate	greatly	increases	the	risk	of	being	preyed	upon	(Sih,	1992).	Sedentary	species	that	are	ambush	predators	are	also	dependent	on	local	prey	characteristics	as	they	feed	on	nearby	prey	(Gerritsen,	1984;	Sih	&	Moore,	1990).	When	animals	engage	in	reproductive	activities	such	as	building	nests	for	their	young,	or	parental	care,	they	are	more	vulnerable	to	predation	(Magnhagen,	1991).	The	effect	of	these	habitat	characteristics	is	exacerbated	when	animals	have	small	home	ranges	and	are	unlikely	to	disperse	to	more	favourable	conditions	(Roberts	&	Hawkins,	1999).	The	distribution	of	these	animals	might	therefore	be	expected	to	be	shaped	by	their	habitat	microstructure	and	by	fine-scale	differences	in	prey	and	predator	availability.	Yet	few,	if	any,	studies	have	evaluated	the	relative	effects	of	habitat,	prey,	and	predator	variables	on	the	abundance	or	distribution	of	a	sedentary	animal	species.		 11	The	relationship	between	a	community	and	any	particular	species	is	particularly	complex	in	the	three-dimensional	structure	of	the	ocean.	Three-dimensional	habitats	provide	organisms	with	more	physical	opportunities	to	survive,	grow,	and	reproduce	(Crowder	&	Norse,	2008).	In	aquatic	environments,	this	structure	is	provided	by	the	liquid	medium	itself,	and	by	foundation	species	such	as	seagrasses,	coral,	macrophytes,	and	mangroves	(Jones	et	al.,	1994;	Thorne-Miller,	1999).	These	habitats	are	able	to	support	more	trophic	levels	and	longer	food	chains	than	two-dimensional	habitats	(Briand	&	Cohen,	1987;	Vander	Zanden	&	Fetzer,	2007).	Food	webs	in	marine	habitats	are	particularly	intricate,	as	animals	in	these	environments	interact	with	a	greater	number	of	species	both	within	and	among	trophic	levels	as	compared	to	terrestrial	or	freshwater	habitats	(Cohen	et	al.,	1990;	Cohen,	1994).	Marine	animals	are	therefore	more	intimately	connected	with	their	communities,	which	have	the	ability	to	affect	species’	abundance	and	distribution	(Werner	&	Peacor,	2003).		Seahorses	(Hippocampus	spp.)	are	a	genus	of	marine	fishes	with	complex	life	histories	whose	relationships	with	their	habitats	and	communities	are	poorly	understood.	Life-history	traits	indicate	that	seahorse	populations	should	be	greatly	influenced	by	the	characteristics	of	their	habitats	and	communities.	These	are	sedentary,	ambush	predators	with	small	home	ranges	and	intensive	male	parental	care	(Vincent	&	Sadler,	1995;	Foster	&	Vincent,	2004;	Vincent	et	al.,	2005).	They	are	poor	swimmers,	and	spend	most	of	their	time	waiting	among	benthic	substrates	(e.g.	seagrass,	coral)	for	prey	to	enter	their	strike	zone	(Howard	&	Koehn,	1985;	James	&	Heck,	1994;	Kendrick	&	Hyndes,	2005).	To	anchor	themselves	for	feeding	strikes,	and	to	remain	cryptic	to	predators,	seahorses	grasp	marine	substrates	with	their	prehensile	tail	(Foster	&	Vincent,	2004).	Reproduction	is	also	tied	to		 12	the	habitat,	as	pairs	within	many	species	maintain	long-term	bonds	that	are	reinforced	by	daily	dances	across	their	home	ranges	(Vincent	et	al.,	1992;	Vincent,	1995;	Vincent	&	Sadler,	1995).	These	characteristics	suggest	intimate	relationships	with	both	habitats	and	communities,	but	these	have	not	been	well	identified	in	the	literature.	Most	in-situ	habitat	studies	on	seahorses	to	date	have	examined	aspects	of	their	habitat	such	as	holdfast	preferences	(Bell	et	al.,	2003;	Dias	&	Rosa,	2003;	Martin-Smith	&	Vincent,	2005;	Morgan	&	Vincent,	2007;	Rosa	et	al.,	2007;	Harasti	et	al.,	2014a),	and	habitat	correlates	such	as	depth,	temperature,	and	water	velocity	(Curtis	&	Vincent,	2005;	Caldwell	&	Vincent,	2012;	Aylesworth	et	al.,	2015).	Those	studies	that	examined	biotic	variables	either	did	not	consider	abiotic	variables	(Harasti	et	al.,	2014b),	or	did	not	differentiate	between	seahorse	prey	and	predators	(Curtis	&	Vincent,	2005).	To	improve	our	ecological	understanding	of	seahorses,	a	study	that	considers	habitat,	prey,	and	predator	variables	together	is	needed.	This	study	is	the	first	to	explicitly	investigate	the	relative	importance	of	habitat	characteristics	and	community	correlates	(prey	and	predators)	to	seahorse	populations	in	the	wild,	as	a	model	sedentary	marine	species.	This	study	investigates	the	associations	between	ecological	parameters	and	the	densities	and	size	distributions	of	White's	seahorse	(Hippocampus	whitei	Bleeker	1895)	in	a	New	South	Wales	estuary.	The	goal	of	this	study	was	to	better	understand	how	these	characteristics	are	correlated	with	seahorse	abundances	and	size	distributions	at	different	scales.	To	do	this,	seagrass,	prey	and	predator	correlates	were	compared	to	seahorse	abundance	and	size	data,	both	among	different	seagrass	beds	in	an	estuary	(denoted	‘medium-scale	study’	in	the	text),	and	within	a	single	seagrass	bed	(denoted	‘small-scale	study’	in	the	text).				 13	2.2	 Materials	&	Methods		2.2.1	 Study	species		Hippocampus	whitei	is	a	medium-sized	seahorse	species	(mean	height	±	SD	was	107	±	27	mm;	this	study)	found	in	coastal	waters	along	the	east	coast	of	Australia	(Lourie	et	al.,	2016).	They	live	among	a	variety	of	marine	substrates,	including	soft	coral	(Dendronephthya	australis),	various	branching	macroalgae	(e.g.	Sargassum	sp.,	Codium	sp.;	Harasti	et	al.,	2014a),	seagrasses	(e.g.	Zostera	muelleri	subsp.	capricorni,	Posidonia	australis.;	Vincent	et	al.,	2005;	Harasti	et	al.,	2014a)	and	artificial	structures	such	as	protective	swimming	net	enclosures	(Clynick,	2008;	Harasti	et	al.,	2010).	Juvenile	H.	whitei	are	more	cryptic	and	prefer	more	complex	habitats	than	adults	(Harasti	et	al.,	2014a).	Hippocampus	whitei	adults	have	small	home-ranges	(Vincent	et	al.,	2005),	and	form	faithful	pair-bonds	during	the	breeding	season,	which	extends	from	September	to	February	(austral	summer;	Vincent	&	Sadler,	1995;	Harasti	et	al.,	2012).		2.2.2	 Study	locations	&	design		Study	scale:	among	seagrass	beds		 The	among	seagrass	beds	study	was	carried	out	along	a	6-km	stretch	of	the	southern	coastline	of	the	Port	Stephens	marine	estuary	in	New	South	Wales	(NSW),	Australia	(Figure	2.1).	The	study	was	limited	to	the	Eastern	Port	of	the	estuary,	where	H.	whitei	were	known	to	occur	(Harasti	et	al.,	2012).	The	benthic	substrate	of	this	region	is	a	mixture	of	Posidonia	australis	(Hooker	1858)	seagrass,	Zostera	muelleri	subsp.	capricorni	seagrass,	sand,	soft	coral	(Dendronephthya	australis),	various	branching	macroalgae	(e.g.	Sargassum	sp.,	Calerpa	sp.),	and	sponges	(Davis	et	al.,	2015).	Posidonia	australis	is	the	most		 14	abundant	holdfast	type	in	shallow	coastal	areas	of	the	estuary	(to	a	depth	of	approximately	5	m;	Davis	et	al.,	2015).	Seven	P.	australis-dominated	seagrass	beds,	located	between	200	m	and	6000	m	apart	from	one	another,	were	the	focus	of	this	study	(Figure	2.1).	In	all	beds,	P.	australis	occurred	at	high-densities,	usually	150-200	shoots	m-2,	with	leaves	reaching	30-50	cm	(Table	B1).	The	other	benthic	substrates	listed	above	were	occasionally	found	within	the	beds.	Any	generic	mention	of	'seagrass'	in	this	paper	refers	specifically	to	P.	australis.		The	aim	of	the	among	seagrass	beds	study	was	to	determine	how	seahorse	density	and	body	size	correlated	with	(i)	seagrass	characteristics,	(ii)	availability	of	prey	and	(iii)	presence	of	predators.	To	meet	this	aim,	two	sampling	campaigns	were	conducted	at	each	of	the	seven	seagrass	beds	during	the	2015-2016	austral	summer:	(1)	10	November	-	23	December	2015;	and	(2)	3-25	February	2016.	Surveys	to	characterize	seahorses,	seagrass,	prey,	and	predators	are	described	in	separate	sections	below.	The	locations	of	the	plots	were	determined	by	randomly	selecting	a	specific	point	along	the	seaward	edge	of	each	seagrass	bed.	From	each	of	these	seven	points,	a	square	sampling	area	was	established	that	measured	30	m	parallel	to	(width),	and	30	m	perpendicular	(length)	from	the	shoreline	(Figure	2.2).	However,	the	shape	of	the	seagrass	beds,	and	sometimes	discontinuous	nature	of	the	seagrass,	meant	the	sample	areas	never	reached	900	m2,	and	instead	ranged	from	111	to	752	m2,	with	a	mean	of	551	m2	across	the	seven	plots	(Table	2.1).			 The	seven	plots	were	sampled	using	a	random-stratified	design.	The	area	of	seagrass	at	each	plot	was	divided	into	three	strata	by	distance	from	the	seaward	edge	of	the	seagrass	patch	(shallow,	medium,	deep).	Each	stratum	was	of	equal	length,	measured	perpendicular	to	shore	(Figure	2.2).	If	the	length	of	the	seagrass	bed	did	not	reach	30	m,	it		 15	was	still	divided	into	thirds.	The	outline	of	each	stratum	was	traced	by	slowly	swimming	its	perimeter	while	towing	a	GPS	device	attached	to	a	safety	float	(as	described	in	Poulos	et	al.,	2013).	The	data	were	then	uploaded	to	Google	Earth	Pro	as	point	data,	from	which	the	area	of	each	stratum	was	calculated.		Study	scale:	within	Little	Beach	seagrass	bed		 One	seagrass	bed,	Little	Beach,	had	a	much	higher	abundance	of	seahorses	than	the	other	six	beds.	Little	Beach	was	therefore	used	as	the	area	to	study	the	ecological	correlates	of	seahorse	abundance	and	size	distributions	within	a	single	bed.	The	Little	Beach	seagrass	bed	is	bisected	by	a	3	m	wide	jetty	(Figure	2.2).	To	the	west	of	the	jetty,	is	the	smaller	350	m2	'West	Patch',	and	to	the	east	of	the	jetty,	the	larger	1248	m2	'Main	Patch'.	Taken	together,	the	Little	Beach	seagrass	bed	is	90	m	wide	and	25	m	long	(Figure	2.2).	The	seaward	edge	of	the	bed	was	approximately	3.5	m	deep.	Despite	having	a	relatively	higher	number	of	seahorses,	the	P.	australis	density	and	substrate	characteristics	were	similar	to	the	other	plots	across	the	estuary	(Table	B1).		The	aim	of	the	within	Little	Beach	study	was	to	determine	how	(1)	seahorse	distribution	and	(2)	body	size	correlated	with	seagrass	characteristics,	availability	of	prey	and	presence	of	predators	within	a	single	seagrass	bed.	(1)	The	goal	of	this	part	of	the	study	was	to	determine	if	seahorses	were	found	at	locations	(0.25	m2	areas	of	seagrass)	within	Little	Beach	that	had	seagrass,	prey,	and	predator	characteristics	proportionate	to	their	availability,	or	if	seahorses	were	preferentially	selecting	locations	with	particular	characteristics.	A	use-availability	study	(as	defined	by	Johnson	et	al.,	2006)	was	carried	out	at	the	Little	Beach	site	in	January	2016,		 16	during	which	seagrass,	prey,	and	predator	characteristics	were	measured	where	seahorses	were	found	('used'	locations,	those	adjacent	to	every	fifth	H.	whitei	found	during	a	seahorse	survey,	see	below),	and	at	random	locations	available	to	seahorses	but	not	necessarily	being	used	by	them	('available'	locations,	sampled	at	haphazard	locations	during	free	swims	through	Little	Beach).	Using	a	resource	selection	function	model	(RSF),	the	seagrass,	prey,	and	predator	characteristics	were	then	compared	between	‘used’	and	‘available’	locations	to	determine	if	seahorse	were	selecting	for	particular	characteristics	(Boyce	et	al.,	2002;	Manly	et	al.,	2002;	Lele	&	Keim,	2006;	McDonald,	2013;	Northrup	et	al.,	2013).	Surveys	to	characterise	seahorses,	seagrass	and	prey	are	described	below.	Predator	data	for	'used'	and	'available'	locations	within	Little	Beach	were	predicted,	as	described	below.		To	determine	the	number	of	‘used’	and	‘available’	samples	required	to	encompass	a	representative	proportion	of	the	population,	the	seahorse	population	of	Little	Beach	was	estimated	using	a	mark-resight	model	as	has	been	done	for	H.	whitei	populations	in	Port	Stephens	in	the	past	(Harasti	et	al.,	2012).	This	mark-resight	model	consisted	of	eight	x	60-minute	free	swims	across	the	entire	site	over	a	5-day	period	(15-19	January	2016).	Each	seahorse	was	tagged	using	unique	combinations	of	three	small	visual	implant	fluorescent	elastomer	tags	(VIFE;	Northwest	Marine	Technologies,	USA).	VIFE	has	been	shown	to	have	minimal	impact	on	the	mortality	and	behaviour	of	H.	whitei	and	other	seahorse	species	(Woods	&	Martin-Smith,	2004;	Woods,	2005;	Harasti	et	al.,	2012).	The	population	was	assumed	closed	during	this	short	time	period	(i.e.	no	immigration,	emigration,	births	or	deaths;	Seber,	1986).	Using	a	closed	capture	estimate	in	the	program	NOREMARK	(White,	1996),	the	population	was	estimated	to	be	312	individuals	(0.195	seahorses	m-2;	95%	CI	=	222-495).	After	factoring	in	the	time	taken	to	process	the	samples	taken	at	a	single		 17	location,	it	was	determined	that	roughly	20%	of	the	population	(‘used’	locations),	and	the	same	number	of	‘available’	locations,	could	be	sampled	for	ecological	correlates.	Therefore,	the	ecological	correlates	associated	with	every	fifth	seahorse	were	sampled.	A	total	of	66	‘used’	and	71	‘available’	samples	were	collected	from	11-30	January	2016.		(2)	The	associations	between	ecological	correlates	(seagrass,	prey,	and	predator	characteristics)	and	the	body	size	of	seahorses	within	Little	Beach	were	also	investigated.	Seahorses	observed	in	the	(i)	November	and	(ii)	February	sampling	campaigns	at	Little	Beach	as	a	part	of	the	study	among	seagrass	beds,	as	well	as	(iii)	in	the	use-availability	study	within	Little	Beach	were	used	to	investigate	seahorse	size	distributions	within	a	single	seagrass	bed.	Seagrass,	prey,	and	predator	data	were	predicted,	as	described	below.		2.2.3	 Seahorse	surveys		 For	the	medium-scale	study,	during	the	November	and	February	sampling	campaigns,	the	entirety	of	each	plot	was	searched	for	H.	whitei.	A	transect	tape	was	laid	1	m	in	from	the	seaward	edge	of	the	shallow	stratum,	parallel	to	shore.	The	length	of	the	transect	was	searched	for	seahorses	by	two	divers	using	SCUBA;	a	1	m	strip	on	either	side	of	the	transect	tape	was	slowly	searched	by	each	diver	with	the	assistance	of	a	30	cm	metal	stick.	The	transect	was	then	moved	2	m	closer	to	shore	and	the	survey	was	repeated.	This	continued	until	the	entirety	of	each	stratum	in	each	of	the	seven	survey	plots	was	searched.	All	seahorses	were	measured	underwater	(see	below),	and	then	returned	to	the	holdfast	from	which	they	were	taken.	To	prevent	duplicate	counting,	each	seahorse	was	marked	using	VIFEs	as	described	above.	The	location	of	each	tagged	seahorse	was	recorded	using	a	towed	GPS.	The	body	size,	sex	and	reproductive	status	of	each	seahorse	was	recorded.		 18	Body	size	was	measured	as	height—the	length	from	the	tip	of	the	coronet	to	the	end	of	the	outstretched	tail	(Lourie	et	al.,	2004).	The	sex,	maturity,	and	reproductive	status	of	each	individual	was	then	determined	using	a	combination	of	height	and	brood	pouch	cues,	in	combination	with	height	at	maturity	models	(Appendix	A).		For	the	small-scale	study,	seahorse	surveys	consisted	of	2	m	wide	transects	along	the	length	of	the	site,	perpendicular	to	shore,	starting	on	the	most	western	edge.	Transects	were	searched	by	two	divers,	and	seahorses	processed	as	described	above—but	in	this	case	divers	only	sampled	every	fifth	seahorse	they	encountered.			2.2.4	 Predator	surveys		 For	the	medium-scale	study,	underwater	visual	censuses	(UVC)	were	used	to	survey	potential	H.	whitei	predators.	Within	each	stratum,	one	diver	swam	along	a	single	haphazardly	located	transect	(max	30	m	in	length)	laid	parallel	to	shore,	at	a	constant	speed	of	roughly	4	m	min-1.	As	the	transect	was	being	laid,	all	fish	observed	within	2	m	of	the	tape	(both	sides)	and	to	a	height	of	4	m	above	the	seafloor	were	recorded	(to	the	lowest	taxonomic	level	possible).	As	the	transect	was	being	retrieved,	the	seagrass	along	its	length	was	searched	for	sedentary	fish	and	crabs	by	two	divers;	a	1	m	width	on	either	side	of	the	transect	tape	was	searched	by	each	diver.	Predators	in	this	paper	includes	fish	and	octopus	species	known	to	eat	H.	whitei—including	dusky	flathead	Platycephalus	fuscus,	red	rock	cod	Scorpaena	jacksoninsis,	striped	anglerfish	Antennarius	striatus,	Sydney	octopus	Octopus	tetricus,	and	blue-lined	octopus	Hapalochlaena	fasciata	(Harasti	et	al.,	2014b).	It	also	includes	species	that	might	eat	fish	similar	to	seahorses	(e.g.	yellowfin	pike	Dinolestes	lewini;	based	on	Froese	&	Pauly,	2017).	Predator	variables	included	'total	predators'		 19	(number	of	individual	predators	along	a	transect)	and	'predator	types'	(number	of	predator	species	observed	along	a	transect).		For	the	small-scale	study,	predators	were	surveyed	at	Little	Beach	on	three	consecutive	days,	19-21	January	2016,	along	25	m	transects	laid	perpendicular	to	shore,	at	eight	fixed	locations—0,	20,	35,	50,	65,	and	80	m	from	the	most	western	part	of	'West	Patch'.	Predators	were	surveyed	using	UVC,	as	described	above.		2.2.5	 Seagrass	surveys		 In	the	medium-scale	study,	seagrass	was	quantified	in	three	quadrats	placed	at	random	along	each	predator	transect	(see	above).	The	seagrass	variables	used	in	analyses	were	the	mean	seagrass	density	and	seagrass	height	within	a	stratum.	For	the	small-scale	study	focused	on	Little	Beach,	quadrats	were	either	centred	around	a	seahorse	(‘used’	locations)	or	placed	haphazardly	(‘available’	locations).	Quadrats	of	50	cm	x	50	cm	(0.25	m2)	were	used	to	characterize	seagrass	(West,	1990).	The	number	of	P.	australis	shoots	in	each	quadrat	was	counted	('seagrass	density')	and	the	longest	blade	in	each	of	five	haphazardly	selected	shoots	in	the	quadrat	was	measured	('seagrass	height';	Orth	et	al.,	2002;	Ceccherelli	et	al.,	2007).			2.2.6	 Prey	surveys		 To	determine	prey	availability	in	both	the	among	and	within	seagrass	bed	studies,	2-3	blades	of	a	haphazardly	selected	P.	australis	shoot	were	collected	within	each	0.25	m2	seagrass	quadrat.	Seagrass	blades	were	sealed	in	plastic	bags,	taken	to	the	surface,	and	immediately	preserved	in	5%	formalin	for	later	processing.		 20	In	the	lab,	seagrass	samples	were	rinsed	through	a	50	μm	sieve	to	collect	all	small	(<1	cm)	epifauna	on	the	blade.	All	potential	invertebrate	prey	were	counted	and	identified	to	broad	functional	categories	under	a	microscope	(e.g.	amphipods,	copepods,	polychaete	worms;	Horinouchi	&	Sano,	2000;	Yip	et	al.,	2015;	Valladares	et	al.,	2016).	The	determination	of	potential	prey	was	based	on	the	diets	of	other	seahorse	species	as	very	limited	information	was	available	on	the	diet	of	H.	whitei	specifically	(Burchmore	et	al.,	1984).	Epiphytic	growth	was	scraped	off	the	seagrass	blades,	and	both	the	epiphytic	growth	and	seagrass	leaves	were	dabbed	dry	and	weighed	to	the	nearest	0.01	g.	Prey	variables	were	quantified	as	'Prey	Types’	(number	of	functional	prey	categories	in	a	sample)	and	'Prey	Density'	(number	of	individual	prey	items	divided	by	the	mass	of	the	P.	australis	sample).	'Fouling	density'	was	calculated	as	the	mass	of	epiphytic	growth	divided	by	the	mass	of	the	P.	australis	on	which	it	grew.		Prey	types,	prey	density	and	fouling	density	values	were	averaged	(as	a	mean)	among	the	three	quadrats	for	each	stratum	in	the	among	seagrass	bed	study.	For	the	Little	Beach	study,	these	values	were	enumerated	for	all	used	and	available	samples.				2.2.7	 		Predicting	covariates	within	Little	Beach			2.2.7.1		Seagrass	and	prey		 Ideally,	all	seahorse	observations	at	Little	Beach	would	have	been	used	in	the	analysis	exploring	the	association	between	ecological	variables	and	seahorse	height	(N	=	328).	However,	seagrass	and	prey	information	was	only	available	for	seahorses	observed	during	the	use-availability	study	(N	=	66).	In	order	to	generate	associated	seagrass	and		 21	prey	information	for	the	other	262	seahorse	observations,	the	ordinary	kriging	function	in	ArcMap's	Geostatistical	Analyst	10.2	was	used	to	generate	a	series	of	interpolated	spatial	maps	for	seagrass	and	prey	characteristics.	These	maps	were	subsequently	used	to	predict	the	seagrass	and	prey	characteristics	for	every	seahorse	observation	based	on	their	location	within	the	interpolated	spatial	maps.	The	following	parameters	were	predicted:	depth,	seagrass	density,	seagrass	height,	fouling	density	and	prey	types.	The	variability	in	prey	density	prevented	reliable	predictions,	and	was	therefore	excluded	from	this	analysis.	These	interpolations	were	used	in	analyses	of	seahorse	height	within	Little	Beach.		2.2.7.2		Predators			 Because	the	only	predator	information	for	Little	Beach	was	from	fixed	transects,	predator	data	had	to	be	predicted	for	all	‘used’	and	‘available’	locations	in	the	RSF,	and	all	seahorse	observations	in	the	height	analysis.	The	total	numbers	of	predators,	and	predator	types,	were	regressed	against	the	distance	from	the	most	western	part	of	'West	Patch'	using	polynomial	equations	(Figure	2.3;	see	caption	for	model	outputs).	The	equations	were	subsequently	used	to	predict	the	total	number	of	predators,	and	predator	types,	for	all	‘used’	and	‘available’	locations	in	the	RSF,	and	for	all	seahorse	observations	in	the	analysis	of	seahorse	height	within	Little	Beach.							 22	2.2.8	 Statistical	analyses		Study	scale:	among	seagrass	beds		 Generalized	linear	mixed	effects	models	were	used	to	determine	the	extent	to	which	seagrass,	prey,	and	predator	characteristics	explained	differences	in	H.	whitei	density	and	height	among	plots.	Using	Cleveland	dotplots,	all	seagrass,	prey,	and	predator	covariates	were	evaluated	for	the	presence	of	outliers	(Zuur	et	al.,	2010).	In	both	the	seahorse	density	and	height	models,	the	effect	of	outliers	in	fouling	density	data	was	removed	by	a	log-transformation.	Variance	Inflation	Factor	(VIF)	analysis	was	used	to	determine	and	deal	with	multicollinearity	in	two	steps,	following	the	protocol	described	in	(Zuur	et	al.,	2010).	First,	VIF	was	applied	to	independent	variables	within	the	same	group	(seagrass,	prey,	and	predator	covariates),	such	that	the	variables	contributing	the	most	variance	(i.e.	the	highest	VIF)	were	sequentially	removed	until	the	VIF	scores	of	all	covariates	within	each	group	was	less	than	3.	Second,	the	process	was	repeated	for	the	remaining	variables	across	groups.	Although	a	VIF	of	3	is	generally	a	strict	cut-off	for	covariate	inclusion	(Zuur	et	al.,	2010),	the	global	models	were	constrained	by	degrees	of	freedom	and	required	greater	selectivity.	Resulting	independent	variables	used	in	the	global	model	to	predict	H.	whitei	density	were:	depth,	seagrass	density,	prey	types,	prey	density,	and	total	predators;	and	for	seahorse	height	were:	seagrass	density,	seagrass	height,	prey	types,	prey	density,	total	predators,	and	predator	types.	Seahorse	density	fit	a	log-normal	distribution,	and	the	untransformed	density	values	were	fit	using	a	generalized	linear	mixed	effects	model	(GLMM)	with	a	Gaussian	distribution	and	a	log-link	function.	Mean	seahorse	height	was	fit	using	a	normally	distributed	linear	mixed	effects	model.	To	account	for	spatial	autocorrelation	and	site-level		 23	effects	in	both	models,	stratum	nested	within	plot	was	included	as	a	random	effect	(Bolker	et	al.,	2009).	To	account	for	temporal	autocorrelation,	a	time-level	random	effect	was	used	(Bolker	et	al.,	2009).	The	models	were	run	using	the	lme4	package,	version	1.1-12	in	R	3.3.1	(www.r-project.org).	An	information-criterion	approach	was	used	to	select	a	group	of	models	that	best	describe	the	data,	but	individually	may	not	be	distinguishable	in	their	ability	to	model	the	data	(Grueber	et	al.,	2011).	All	potential	models	were	first	ranked	by	their	AICc	score,	and	then	parameter	estimates	were	averaged	(with	weight-adjustment)	over	models	that	were	within	4	AICc	units	of	the	best	model	(Table	2.2;	Grueber	et	al.,	2011).		Study	scale:	within	Little	Beach	seagrass	bed		 An	exponential	logistic	RSF	was	applied	to	the	use-availability	design	to	compare	characteristics	measured	at	locations	adjacent	to	seahorses	to	those	measured	at	available	locations.	The	RSF	of	a	particular	location	is	proportional	to	the	probability	of	H.	whitei	selecting	that	particular	location,	given	the	set	of	seagrass,	prey,	and	predator	characteristics	measured	there.	By	comparing	the	RSFs	of	used	and	available	locations,	it	can	be	determined	if	seahorses	were	selecting	locations	with	particular	ecological	characteristics.	The	model	was	run	using	the	ResourceSelection	package	version	0.3-0	in	R	(www.r-project.org).	Prior	to	running	the	model,	outliers	within,	and	multicollinearity	among	covariates	were	evaluated	as	in	the	among	seagrass	bed	study.	The	global	model	had	the	following	independent	variables:	depth,	seagrass	density,	seagrass	height,	fouling	density,	prey	types,	prey	density,	and	total	predators	(Table	2.2).	An	information-criterion,	model-averaging	approach	was	used	to	estimate	the	parameters	as	outlined	in	the	among		 24	seagrass	bed	study	(Grueber	et	al.,	2011).	Seahorse	height	data	were	modelled	using	a	series	of	linear	mixed	effects	models	for	the	following	groups	of	H.	whitei:	all	seahorses,	all	reproductively	active	(RA)	seahorses,	females,	RA	females,	males,	and	RA	males.	Outliers	were	evaluated	as	in	the	among	seagrass	bed	study.	Multicollinearity	was	evaluated	as	in	the	among	seagrass	bed	study;	however,	a	VIF	cut-off	of	4	was	used	because	there	were	a	high	number	of	observations	and	lower	number	of	candidate	variables,	so	the	models	were	not	constrained	by	degrees	of	freedom	(Zuur	et	al.,	2010).	Predator	and	interpolated	seagrass	and	prey	data	were	included	as	independent	variables.	Global	models	had	the	following	independent	variables:	depth,	seagrass	height,	prey	types,	total	predators,	and	predator	types.	Because	91	of	the	328	sightings	represented	the	same	individuals	(sighted	in	both	November	and	February),	autocorrelation	that	would	result	from	multiple	sightings	of	the	same	individual	was	accounted	for	by	including	an	individual-level	random	effect	(Bolker	et	al.,	2009).	To	account	for	temporal	autocorrelation,	a	time-level	random	effect	was	included	in	the	model.	The	model	was	run	using	the	lme4	package,	version	1.1-12	in	R	(www.r-project.org).	Information-criterion,	model-averaging	approaches	were	used	to	estimate	parameters	for	each	seahorse	reproductive	group,	as	outlined	in	the	among	seagrass	bed	study	(Table	B2;	Grueber	et	al.,	2011).		All	research	undertaken	in	this	project	was	done	in	accordance	with	the	University	of	British	Columbia's	Animal	Care	Committee	permit	A12-0288	and	the	NSW	DPI	Animal	Care	and	Ethics	Committee	permit	15/01.				 25	2.3	 		Results		2.3.1	 		Study	scale:	among	all	seagrass	beds			2.3.1.1		Seahorse	survey	summary	statistics			 Among	the	plots	sampled	along	the	Port	Stephens	estuary,	the	number	of	seahorses	found	varied	across	both	space	and	time	(Figure	2.4;	Table	2.1).	During	both	sampling	campaigns,	Little	Beach	and	Pipeline	had	the	highest	overall	seahorse	densities	(Figure	2.4a).	Seahorses	found	in	plots	other	than	Little	Beach	accounted	for	just	19%	of	individuals	found	in	November,	but	nearly	half	of	those	found	in	February	(48%).	More	seahorses	were	found	in	February	than	November	in	all	plots	except	Little	Beach	and	Pipeline.	A	total	of	nine	seahorses	were	found	in	November	across	Seahorse	Gardens	1,	Seahorse	Gardens	2	and	Dutchies,	but	this	increased	to	73	in	February;	the	increase	in	these	three	plots	was	largely	due	to	an	increased	number	of	encountered	females	(2	to	27),	and	juveniles	(6	to	34).	In	contrast,	the	number	of	seahorses	encountered	at	Little	Beach	and	Pipeline	decreased	only	slightly	from	November	to	February	(Table	2.1).		The	height	of	adult	seahorses	did	not	differ	significantly	between	months	(ANOVA,	F1,257	=	0.83,	P	>	0.05),	but	did	differ	among	plots	(Figure	2.4b;	ANOVA,	F5,257=	15.20,	P	<	0.001).	Nearly	half	(48%)	of	all	adults	were	male	in	November,	and	just	under	half	(42%)	were	male	in	February	(Table	2.1).	A	higher	percentage	of	seahorses	were	physically	mature	in	November	(95%)	than	February	(74%;	Table	2.1).	The	total	number	of	observed	adults	was	the	same	in	November	and	February,	but	the	number	of	juveniles	was	more	than	six	times	greater	in	February	(Table	2.1).	Little	Beach	and	Pipeline	also	had	the	highest	proportion	of	individuals	that	were	physically	mature	(with	the	exception	of	Fly	Point	in	November,	which	had	only	three	seahorses;	Figure	2.4c).		 26	2.3.1.2		Correlates	of	seahorse	density	&	seahorse	height			 Among	seagrass	beds,	seahorse	density	was	significantly	negatively	associated	with	total	predators	(Figure	2.5a).	None	of	the	seagrass,	prey,	or	predator	variables	was	significantly	associated	with	seahorse	height	among	seagrass	beds	(Figure	2.5b).		2.3.2	 		Study	scale:	within	Little	Beach	seagrass	bed			 Within	Little	Beach,	seahorses	were	more	likely	to	be	found	in	deeper	locations	with	denser	seagrass,	a	greater	number	of	prey	types,	and	fewer	total	predators	(Table	2.3).	Seagrass	height,	fouling	density	and	prey	density	were	not	significant	determinants	of	where	seahorses	were	found	(Table	2.3).	Within	the	Little	Beach	seagrass	bed,	seahorse	height	increased	significantly	with	depth	for	all	seahorses,	as	well	as	subsets	of	the	population:	all	RA	seahorses	(Figure	2.6a),	females	(Figure	2.6c),	RA	females	(Figure	2.6d)	and	RA	males	(Figure	2.6f).	Seahorse	height	also	increased	significantly	with	total	predators	for	all	RA	seahorses	(Figure	2.6b),	and	females	(Figure	2.6c;	Table	2.3).	Seagrass	height,	prey	types,	and	predator	types	did	not	correlate	significantly	with	seahorse	height	in	any	subset	of	the	population	(Figure	2.6),	and	none	of	the	independent	variables	correlated	with	male	height	(Figure	2.6e).		2.3.3	 		Overall	associations	between	seahorses	and	ecological	correlates		 Although	seahorses	selected	deeper	locations,	such	locations	were	also	associated	with	smaller	seahorses	(except	males;	Table	2.4).	Seahorses	selected	locations	with	denser	seagrass	(Table	2.4).	Seagrass	height	and	fouling	density	were	not	correlated	with	any	seahorse	parameter	within	or	among	seagrass	beds	(Table	2.4).	Seahorses	selected		 27	locations	with	more	prey	types	(Table	2.4).	Prey	density	was	not	correlated	with	any	seahorse	parameters	both	within	or	among	seagrass	beds	(Table	2.4).	Seahorses	selected	locations	with	fewer	predators	within	Little	Beach,	and	predators	were	negatively	associated	with	seahorses	among	different	seagrass	beds	(Table	2.4).	Additionally,	there	was	a	positive	association	between	predators	and	seahorse	height	(among	RA	seahorses	and	females;	Table	2.4).	Predator	types	were	not	associated	with	any	seahorse	parameters	both	in	or	among	seagrass	beds	(Table	2.4).			 	2.4	 Discussion			 Habitat,	prey,	and	predator	variables	all	correlated	with	seahorse	density	or	height	distributions	to	varying	extents,	depending	on	the	scale	of	the	study.	In	general,	habitat	variables	predicted	seahorse	distribution	better	within	seagrass	beds	than	across	them,	consistent	with	passive	habitat	selection	at	large	scales	and	active	habitat	selection	at	the	level	of	the	seagrass	bed,	perhaps	hinting	that	seahorses	take	what	opportunities	they	can	in	the	bed	where	they	find	themselves.	Seahorses	also	show	a	negative	association	with	predators	once	all	other	variables	are	accounted	for,	either	because	of	direct	mortality	or	behavioural	changes	among	seahorses.	Within	a	seagrass	bed,	this	sedentary	seahorse	species	was	associated	with	denser	seagrass	and	deeper	habitat.	It	may	be	that	such	habitat	offered	more	food	and	less	detection,	especially	of	their	relatively	visible	courtship	and	mating.	That	said	there	was	no	clear	association	between	the	distribution	of	seahorses	and	the	abundance	of	their	epiphytic	prey.	All	of	the	results	of	this	study	must,	however,	be	considered	in	the	context	of	a	population	in	transition.	Human	activities	have	led	to	tremendous	habitat	loss,	which	may	explain	why	seahorses	are	now	found	in	seagrass	beds		 28	and	not	in	their	previous	coral	and	sponge	habitats.	It	is	thus	possible	that	relationships	between	seahorses	and	their	abiotic	environment	and	biotic	community	are	still	in	flux.		 Seahorses	showed	a	stronger	association	with	environmental	variables	within	a	single	seagrass	bed	than	among	seagrass	beds,	indicating	that	they	may	have	to	make	the	best	of	the	habitat	in	which	they	settle.	Although	most	species	are	planktonic	after	birth—likely	dispersed	by	currents—they	settle	at	about	2–8	weeks,	after	which	they	are	relatively	sedentary	(Foster	&	Vincent,	2004).	While	adults	of	some	seahorse	species	make	seasonal	migrations	to	deeper	waters,	most	species	studied	to	date	maintain	small	home-ranges	for	prolonged	periods,	likely	because	of	their	poor	swimming	ability,	sit-and-wait	predatory	style,	and	stable	social	structures	(Foster	&	Vincent,	2004).	In	fact,	H.	whitei	shows	particularly	strong	site	associations	across	seasons	and	years	(Vincent	et	al.,	2005;	Harasti	et	al.,	2014a).	For	example,	none	of	the	1100	H.	whitei	tagged	by	Harasti	(2016)	between	2007	and	2009	was	ever	found	at	a	different	site	along	the	estuary.	Movement	among	different	seagrass	beds	would	likely	be	costly	in	terms	of	risk	of	exposure	to	predators,	strong	currents,	and	the	risk	of	not	finding	a	mate.	In	contrast,	seagrass	beds	reduce	flow	within	their	canopies	(Hendriks	et	al.,	2008),	and	provide	refuge	from	predators	(Orth	et	al.,	1984;	Heck	&	Crowder,	1991;	Canion	&	Heck,	2009).	Seahorses	are	therefore	able	to	move	within	seagrass	beds	without	being	exposed	to	the	strong	tidal	currents	of	the	Port	Stephens	estuary	(Davis	et	al.,	2015),	while	also	maintaining	crypsis.	Given	that	seahorses	show	strong	fidelity	with	small	areas,	it	is	not	surprising	that	this	study	found	H.	whitei	to	be	associated	with	many	ecological	correlates	within	a	seagrass	bed,	and	very	few	among	seagrass	beds.			 29	The	negative	association	between	seahorse	and	predator	abundance	may	arise	from	changes	in	seahorse	behaviour	rather	than	direct	mortality.	Observation	and	gut-content	studies	suggest	relatively	low	predation	on	seahorses	(Kleiber	et	al.,	2011).	Indeed	the	most	intense	evaluation	of	predation	on	seahorses	to	date,	involving	hundreds	of	hours	of	observation	of	more	than	2000	individual	seahorses,	recorded	only	13	predation	events	(Harasti	et	al.,	2014b).	This	should	not	be	that	surprising	as	seahorses	and	their	syngnathid	relatives	are	highly	cryptic	fish	with	the	ability	to	adjust	coloration	and	to	grow	body	filaments	to	camouflage	with	their	seagrass	habitats	(Foster	&	Vincent,	2004).	For	predators	that	are	able	to	find	syngnathids,	they	are	a	poor	meal;	they	are	bony,	difficult	to	digest,	and	a	low	energy	food	source	(Harris	et	al.,	2008).	In	most	cases,	predators	that	feed	on	syngnathids	are	opportunistic,	generalist	feeders	(Kleiber	et	al.,	2011).	Taken	together,	this	suggests	that	the	negative	association	of	seahorses	and	predators	may	instead	reflect	a	predator-induced	change	in	behaviour	(Werner	&	Peacor,	2003;	Preisser	et	al.,	2005),	with	seahorses	redistributing	themselves	to	avoid	predation,	a	common	strategy	among	fish	(Milinski,	1993).	Predator-avoidance	behaviours	may	be	particularly	pronounced	during	H.	whitei's	breeding	season,	which	is	when	this	study	was	conducted.	Seahorses	are	more	vulnerable	to	predation	during	breeding	as	they	engage	in	risky	courtship	behaviours	and	have	lengthy	parental	care	(Magnhagen,	1991;	Foster	&	Vincent,	2004).	While	the	present	study	is	not	the	first	to	show	a	negative	association	between	seahorses	and	their	predators	(Harasti	et	al.,	2014b),	it	is	the	first	to	show	this	association	is	independent	of	other	ecological	variables.	Hippocampus	whitei	may	have	selected	denser	seagrass	locations	because	the	increased	habitat	complexity	improved	its	success	as	an	ambush	predator.	Previous		 30	research	has	suggested	preference	for	certain	habitat	characteristics	is	not	consistent	among	seahorse	species,	perhaps	because	of	differences	in	foraging	strategies.	Relatively	sedentary	seahorse	species	may	favour	dense	seagrass	where	the	increased	habitat	complexity	has	been	shown	to	either	increase	or	have	no	effect	on	the	success	of	ambush	feeding	(James	&	Heck,	1994;	Flynn	&	Ritz,	1999).	In	contrast,	since	habitat	complexity	impedes	the	foraging	success	for	more	active	fishes,	more	active	seahorse	species	may	prefer	more	open	landscapes	(Orth	et	al.,	1984;	Heck	&	Crowder,	1991;	Canion	&	Heck,	2009).	The	sedentary	seahorse	Hippocampus	guttulatus	was	found	in	more	densely	vegetated	areas	than	its	more	active	congener,	Hippocampus	hippocampus,	probably	because	it	preferred	greater	complexity	for	feeding	purposes	(Curtis	&	Vincent,	2005).	Likewise,	the	ambush	predator	Hippocampus	erectus	was	shown	to	change	from	an	ambush	(sit-and-wait)	foraging	strategy	to	one	that	was	more	active	when	moved	to	less	dense	seagrass	habitats,	presumably	since	they	lost	their	cryptic	advantage	in	areas	with	less	seagrass	(James	&	Heck,	1994).	The	finding	that	H.	whitei	selected	denser	seagrass	is	therefore	consistent	with	its	extremely	sedentary	nature	(C.	Manning,	personal	observation;	Vincent	et	al.,	2005;	Harasti	et	al.,	2014a),	and	may	reflect	a	preference	for	better	foraging	prospects	in	these	areas.		Since	seahorses	change	feeding	strategies	as	they	age,	it	is	surprising	that	seahorse	size	was	not	correlated	with	prey	variables,	whether	with	or	among	seagrass	beds.	As	seahorses	grow,	they	are	able	to	feed	on	a	greater	variety	of	prey	types	and	eat	a	greater	quantity	of	food	(Flynn	&	Ritz,	1999;	Castro	et	al.,	2008).	This	type	of	ontogenetic	shift	in	diet	is	also	seen	in	other	syngnathids	(Bennett,	1989;	Rosa	et	al.,	2007),	and	teleost	fishes	in	general	(Wootton,	2012).	Even	if	seahorses	are	unlikely	to	move	among	seagrass	beds		 31	after	they	settle,	it	might	be	expected	that	larger	seahorses	would	aggregate	within	seagrass	beds	in	areas	with	more	prey	types	and	higher	prey	densities.	It	is	possible	that	the	method	of	prey	sampling	used	in	this	study,	in	which	prey	items	were	collected	from	the	surface	of	seagrass	blades,	did	not	represent	the	prey	used	by	different	age	classes	of	seahorses.	Although	prey	associated	with	the	surface	of	plants	are	a	major	source	of	food	for	seahorses	(Horinouchi	&	Sano,	2000;	Teixeira	&	Musick,	2001;	Kendrick	&	Hyndes,	2005;	Kitsos	et	al.,	2008;	Storero	&	Gonzalez,	2008;	Gurkan	et	al.,	2011b),	planktonic	and	epibenthic	prey	-	not	considered	in	this	study	-	are	important	dietary	items	for	smaller	seahorses	(Teixeira	&	Musick,	1995;	Kanou	&	Kohno,	2001;	Castro	et	al.,	2008).	For	the	scale	of	prey	sampling	in	the	present	study	(based	on	0.25m2	quadrats),	however,	the	collection	of	planktonic	prey	was	impractical.	A	goal	of	future	studies	that	aim	to	better	understand	the	full-scope	of	relationships	between	seahorses	and	their	prey	should	be	to	include	planktonic	and	epibenthic	species.		This	is	the	first	study	to	empirically	show	that	larger	seahorses	aggregate	at	shallower	depths	within	a	habitat	patch	possibly	because	of	spatial	differences	in	feeding	and	reproductive	opportunities.	A	different	result	was	reported	by	Harasti	et	al.	(2014a),	who	found	no	difference	in	the	depths	at	which	adults	and	juveniles	were	found.	Within	Little	Beach,	greater	depth	is	associated	with	the	seaward	edge	of	the	P.	australis	bed.	Seagrass	edges	are	a	physical	barrier	to	nutrient	flow	and	have	the	ability	to	intercept	zooplankton	carried	to	seagrass	beds	in	the	water	column	(Jenkins	&	Sutherland,	1997;	Hendriks	et	al.,	2008),	concentrating	these	resources	along	the	seaward	edge.	Since	planktonic	crustaceans	(not	considered	in	this	study)	are	important	dietary	items	for	smaller	seahorses,	smaller	individuals	may	benefit	from	foraging	in	these	areas	(Teixeira	&		 32	Musick,	1995;	Castro	et	al.,	2008).	Habitat	edges	could	also	benefit	smaller	seahorses	by	allowing	them	to	occasionally	forage	for	planktonic	prey	over	open	sand,	while	remaining	close	to	their	protective	vegetative	habitat	(Orth	et	al.,	1984).	In	addition,	adult	seahorses	may	have	preferred	shallower	depths	because	the	seagrass	provides	better	refuge	from	predation	while	engaging	in	risky	reproductive	activities	such	as	courtship	and	parental	care	(Magnhagen,	1991).		The	seahorse	densities	found	in	this	study	among	all	seagrass	beds	in	the	Port	Stephens	seagrass	beds	may	reflect	a	mass	habitat	shift	of	the	population	in	response	to	habitat	loss	at	greater	depths.	Notable	seahorse	population	declines	have	been	reported	in	Port	Stephens	in	the	last	decade,	linked	to	a	loss	of	the	soft	coral	and	sponge	habitats	preferred	by	seahorses,	apparently	as	a	result	of	mooring	installation,	anchor	damage,	and	sand	inundation	(Harasti,	2016).	A	multi-year	survey	of	Port	Stephens	(2006-2009)	found	seahorses	concentrated	in	deeper	soft	coral	and	sponges,	with	very	few	in	the	seagrass	patches	that	were	surveyed	in	the	present	study	(Harasti	et	al.,	2014a;	Harasti,	2016).	Nearly	ten	years	later,	this	study	estimated	a	seahorse	population	at	one	seagrass	bed	(Little	Beach)	at	a	density	almost	twice	that	previously	recorded	in	in	the	deeper	habitats	(Harasti,	2016).	The	corollary	is	that	seahorses	were	very	difficult	to	find	in	the	deeper	soft	coral	and	sponge	habitats,	unlike	Harasti	(2016).	It	is	unclear	from	this	study	whether	seagrass	is	a	preferred	habitat	type	for	this	species,	or	is	being	used	because	preferred	types	have	been	lost	locally	(Harasti	et	al.,	2014a).	While	seahorse	surveys	conducted	during	this	study	suggest	that	the	population	decline	was	not	as	bad	as	initially	suspected	(Harasti,	2016),	the	loss	of	preferred	habitat	at	greater	depths	places	pressure	on	H.	whitei	(Foster	&	Vincent,	2004).	In	the	context	of	substantial	declines	in	soft	coral	and	sponge		 33	habitat	at	depth	(Harasti,	2016),	seagrass	may	now	be	playing	an	important	role	in	H.	whitei	ecology.	This	study	should	be	considered	in	decisions	about	closing	areas	to	fishing.	Some	research	indicates	that	spatial	bans	on	recreational	fishing	can	result	in	more	predatory	fish	(Jouvenel	&	Pollard,	2001;	Schroeder	&	Love,	2002),	leading	to	declines	in	populations	of	smaller	fish,	including	cryptic	species	(Willis	&	Anderson,	2003).	This	mechanism	has	also	been	argued	to	apply	to	seahorses	in	Port	Stephens.	In	two	separate	comparisons,	Harasti	et	al.	(2014b)	found	significantly	more	predators	and	significantly	fewer	seahorses	within	two	sites	that	banned	recreational	fishing	(sanctuary	zone)	compared	to	two	sites	that	allowed	it	(habitat	protection	zone).	In	this	study,	however,	the	highest	densities	of	seahorses	were	found	inside	the	sanctuary	zone	(Little	Beach).	This	is	compatible	with	a	single	study	in	the	Philippines	that	showed	no	reserve	effect	for	seahorse	abundance,	although	seahorses	inside	the	MPAs	were	somewhat	larger	than	outside	(Yasué	et	al.,	2012).	More	research	is	necessary	to	understand	how	closing	areas	to	fishing	might	affect	seahorses	in	areas	with	and	without	predators.		In	summary,	this	study	demonstrates	the	importance	of	a	holistic	approach	that	considers	an	integration	of	habitat,	prey,	and	predator	variables	in	animal	studies.	Animals	do	not	live	in	isolation;	they	are	involved	in	trophic	interactions	with	other	organisms	(Paine,	1966;	Polis	&	Strong,	1995),	and	often	have	intimate	relationships	with	their	habitats	(Jones	et	al.,	1994,	1997;	Bruno	&	Bertness,	2001).	This	makes	it	crucial	that	both	their	communities	and	their	habitats	are	considered,	especially	when	they	are	sedentary,	or	involved	in	both	sides	of	the	predator-prey	dynamic.	Using	seahorses	as	an	example,	research	to	date	has	generally	looked	at	the	relationship	between	animals	and	either	their		 34	habitat	(Bell	et	al.,	2003;	Dias	&	Rosa,	2003;	Curtis	&	Vincent,	2005;	Martin-Smith	&	Vincent,	2005;	Rosa	et	al.,	2007;	Caldwell	&	Vincent,	2012;	Harasti	et	al.,	2014a;	Aylesworth	et	al.,	2015),	or	a	component	of	their	community	(Harasti	et	al.,	2014b).	This	study	is	therefore	unique	among	seahorse	studies	in	evaluating	seahorse	abundance	and	distribution	in	models	that	consider	habitat,	prey,	and	predator	correlates	together.	This	study’s	results,	showing	that	habitat,	prey,	and	predator	variables	were	all	important	correlates	of	seahorse	abundance	at	various	scales,	suggests	that	more	such	studies	would	be	valuable.																 35	Table	2.1	Summary	statistics	for	seahorse	(SH)	surveys	at	seven	plots	in	the	Port	Stephens	estuary,	NSW,	Australia,	during	the	November	(Nov)	and	February	(Feb)	sampling	campaigns.	SE	is	the	standard	error	of	the	mean.	Overall	density	represents	the	total	number	of	seahorses	divided	by	the	total	area	searched.	For	adult	height,	total	represents	the	mean	of	all	adults	pooled	across	plots.	February	Little	Beach	physical	maturity	ratios	do	not	add	to	total	number	of	seahorses	found	because	the	sex	of	one	individual	was	not	determined,	and	this	individual	was	not	included	in	any	other	calculations.	Plots	are	listed	from	smallest	to	largest	distance	from	the	estuary	mouth.		 		 		 Abundances	 		 Adult	Height	(mm)	 	 Sex	Ratios	 		 Physical	Maturity	Ratios		 	 Total	SH	 Density	(SH	100m-2)	 	 November	 February	 	 November	 February	 	 November	 February	Plot	 Area	(m2)	 Nov	 Feb	 Nov	 Feb	 	 Mean	 SE	 Mean	 SE	 	 Male	 Female	 Male	 Female	 	 Adult	 Juvenile	 Adult	 Juvenile	Shoal	Bay	 625	 0	 0	 0.0	 0.0	 	 -	 -	 -	 -	 	 0	 0	 0	 0	 	 0	 0	 0	 0	Little	Beach	 600	 112	 92	 18.7	 15.3	 	 119	 1	 124	 2	 	 55	 56	 38	 46	 	 111	 1	 84	 7	Fly	Point	 604	 3	 3	 0.5	 0.5	 	 99	 2	 -	 -	 	 2	 1	 0	 0	 	 3	 0	 0	 3	SHG	1	 561	 5	 39	 0.9	 7.0	 	 125	 13	 95	 3	 	 1	 1	 8	 15	 	 2	 3	 23	 16	SHG	2	 605	 1	 21	 0.2	 3.5	 	 -	 -	 103	 6	 	 0	 0	 3	 9	 	 0	 1	 12	 9	Pipeline	 111	 15	 10	 13.5	 9.0	 	 120	 3	 116	 8	 	 6	 9	 5	 4	 	 15	 0	 9	 1	Dutchies	 752	 3	 13	 0.4	 1.7	 	 131	 -	 94	 8	 	 0	 1	 1	 3	 	 1	 2	 4	 9	Overall	 3858	 139	 178	 3.6	 4.6	 		 119	 1	 115	 2		64	 68	 55	 77	 		 132	 7	 132	 45		 36	Table	2.2	Summary	of	the	model-averaged	statistics	for	the	top	models	predicting:	(a)	seahorse	density	among	seagrass	beds,	(b)	adult	height	among	seagrass	beds,	(c)	and	resource	selection	function	within	the	Little	Beach	seagrass	bed.	LL	=	log-likelihood,	AICc	=	corrected	Akaike	information	criterion,	ΔAICc	=	difference	in	model	AICc	with	that	of	the	top	model,	wi	=	Akaike	weight,	df	=	number	of	model	parameters	including	intercepts	and	residuals.	The	following	abbreviations	have	been	made:	DPTH	=	depth,	DENS	=	seagrass	density,	HGHT	=	seagrass	height,	TPT	=	prey	types,	TPI	=	prey	density,	FLNG	=	fouling,	PRED	=	total	predators,	and	TPRED	=	types	of	predators.		Model	and	parameters	included	 LL	 AICc	 ΔAICc	 wi	 df	Among	seagrass	beds	 	 	 	 	 		(a)	Seahorse	density	 	 	 	 	 		 	 PRED	 -104.80	 224.3	 0.00	 0.47	 6		 	 TPI,	PRED	 -104.56	 226.86	 2.56	 0.13	 7		 	 TPT,	PRED	 -104.62	 226.97	 2.67	 0.12	 7		 	 DPTH,	PRED	 -104.75	 227.22	 2.92	 0.11	 7		 	 DENS,	PRED	 -104.79	 227.31	 3.01	 0.1	 7		 	 TPI	 -106.76	 228.23	 3.93	 0.07	 6		(b)	Adult	seahorse	height	 	 	 	 	 		 	 DENS,	HGHT,	TPT,	TPI,	PRED,	TPRED	(global	model)	 -94.10	 229.05	 0.00	 0.25	 11		 	 HGHT,	TPT,	TPI,	PRED,	TPRED	 -97.55	 229.77	 0.72	 0.17	 10		 	 DENS,	TPT,	TPI,	PRED,	TPRED	 -97.60	 229.87	 0.81	 0.17	 10		 	 DENS,	HGHT,	TPT,	PRED,	TPRED	 -97.76	 230.18	 1.12	 0.14	 10		 	 DENS,	HGHT,	TPT,	TPI,	TPRED	 -97.76	 230.19	 1.14	 0.14	 10		 	 DENS,	HGHT,	TPT,	TPI,	PRED	 -97.83	 230.32	 1.27	 0.13	 10	Within	Little	Beach	seagrass	bed	 	 	 	 	 		(c)	Resource	Selection	Function		 	 	 	 	 		 	 DPTH,	DENS,	TPT,	TPI,	PRED	 -225.47	 462.04	 0.00	 0.32	 5		 	 DPTH,	DENS,	FLNG,	TPT,	TPI,	PRED	 -224.97	 463.53	 1.48	 0.15	 6		 	 DPTH,	DENS,	HGHT,	TPT,	TPI,	PRED	 -225.26	 464.1	 2.05	 0.12	 6		 	 DPTH,	TPT,	TPI,	PRED	 -227.89	 464.52	 2.47	 0.09	 4		 	 DPTH,	DENS,	PRED	 -229.16	 464.74	 2.70	 0.08	 3		 	 DPTH,	DENS,	FLNG,	TPT,	PRED	 -226.88	 464.87	 2.83	 0.08	 5		 	 DPTH,	DENS,	TPT,	PRED	 -228.40	 465.53	 3.48	 0.06	 4		 	 DPTH,	DENS,	HGHT,	FLNG,	TPT,	TPI,	PRED	(global	model)	 -224.85	 465.86	 3.82	 0.05	 7			 		 DPTH,	DENS,	HGHT,	PRED	 -228.57	 465.88	 3.83	 0.05	 4		 37		Table	2.3	Model-averaged	parameter	estimates,	standard	error	(SE)	of	the	parameter,	correlate	relative	importance,	the	upper	and	lower	90%	parameter	confidence	intervals	(CI)	for	variables	predicting	resource	selection	function	within	the	Little	Beach	seagrass	bed.			Variable	 Parameter	estimate	 SE	 Relative	importance	 5%	CI	 95%	CI	Depth	 0.856**	 0.255	 1.00	 0.437	 1.276	Seagrass	density	 0.031**	 0.015	 0.91	 0.006	 0.056	Seagrass	height	 0.026a	 0.045	 0.21	 -0.048	 0.101	Fouling	(log)	 -3.095a	 3.121	 0.28	 -8.229	 2.039	Prey	types	 0.207*	 0.113	 0.87	 0.021	 0.393	Prey	density	(log)	 -1.099a	 0.682	 0.73	 -2.220	 0.022	Total	predators	 -0.805**	 0.179	 1.00	 -1.100	 -0.511		a,	P	>	0.10	*	0.10	>	P	>	0.05	**	P	<	0.05	 38	Table	2.4	Summary	of	relationships	between	seagrass,	prey,	and	predator	covariates	with	(a)	seahorse	(SH)	density	and	(b)	adult	height	among	seagrass	beds,	(c)	resource	selection	function	(RSF)	within	the	Little	Beach	seagrass	bed	and	(d)	SH	height	within	Little	Beach,	among	SH	with	different	sexes	and	reproductive	statuses.	RA	=	reproductively	active.				 		 Among	seagrass	beds	 		 Within	Little	Beach	seagrass	bed		 	 (a)	SH	Density	 (b)	Adult	SH	Height	 	 (c)	RSF	 (d)	SH	Height	Parameter	 	 		 		 		 All	SH	 RA	SH	 Females	 RA	Females	 Males	 RA	Males	Seagrass	covariates	 	 	 	 	 	 	 	 	 	 		 Depth	 0	 	 	 +	 -	-	 -	-	 -	-	 -	-	 0	 -		 Seagrass	density	 0	 0	 	 +	 	 	 	 	 	 		 Seagrass	height	 	 0	 	 0	 0	 0	 0	 0	 0	 0		 Fouling	 	 0	 	 0	 	 	 	 	 	 	Prey	covariates	 	 	 	 	 	 	 	 	 	 		 Prey	types	 0	 0	 	 +	 0	 0	 0	 	 0	 0		 Prey	density	 0	 0	 	 0	 	 	 	 	 	 	Predator	covariates	 	 	 	 	 	 	 	 	 	 		 Total	predators	 -	-	 0	 	 -	-	 0	 +	 +	+	 0	 0	 0		 Predators	types	 		 0	 		 		 0	 0	 0	 0	 0	 0		+	+,	positive	relationship,	P	<	0.05	+,	positive	relationship,	0.10	>	P	>	0.05	0,	no	significant	relationship,	P	>	0.10	-	-,	negative	relationship,	P	<	0.05	-,	negative	relationship,	0.10	>	P	>	0.05	blank	cell,	not	included	in	final	model					 39					Figure	2.1	Location	of	the	seven	study	plots	along	the	southern	coast	of	the	Port	Stephens	estuary,	on	the	eastern	coast	of	New	South	Wales,	Australia.	Plot	names:	(1)	Shoal	Bay,	(2)	Little	Beach,	(3)	Fly	Point,	(4)	Seahorse	Gardens	1,	(5)	Seahorse	Gardens	2,	(6)	Pipeline,	(7)	Dutchies.	Shore	Data:	OpenStreetMap	(and)	contributors,	CC-BY-SA.	 40			 41	Figure	2.2	Map	of	the	Little	Beach	seagrass	beds	showing	an	example	of	the	stratified	random	sampling	design	used	at	each	of	the	seven	study	plots	(for	study	among	seagrass	beds).	Light	grey	represents	P.	australis.	White	background	includes	all	other	benthic	substrate	types,	predominantly	sand	and	Zostera	sp.	The	8	m	length	of	the	stratum	was	calculated	by	dividing	the	24	m	length	of	the	seagrass	bed	into	thirds.	Shore	Data:	OpenStreetMap	(and)	contributors,	CC-BY-SA.		 42		Figure	2.3	Mean	(±	SE)	number	of	(a)	total	predators,	and	(b)	different	predator	species	observed	at	fixed	transects	at	Little	Beach,	measured	19-21	January	2016.	Hashed	polynomial	lines	of	best	fit	are	included:	(a)	y	=	0.0029x2	-	0.3473x	+	11.024	(r2	=	0.87),	(b)	y	=	0.0009x2	-	0.0849x	+	2.3039	(r2	=	0.96).		 43		Figure	2.4	(a)	Mean	seahorse	density	(±	SE),	(b)	mean	adult	seahorse	height,	(c)	proportion	of	seahorses	that	were	physically	mature	(adult),	and	(d)	proportion	of	adults	that	were	male,	by	plot	and	by	November	and	December	survey	campaigns.	Values	at	the	base	of	bars	indicate	the	number	of	adult	seahorses	(a,	d)	and	the	total	number	of	seahorses	(c).	Bars	sharing	a	common	letter	do	not	differ	significantly	(Tukey	HSD,	P	>	0.05)		 44			 45	Figure	2.5	Model-averaged	effect	sizes	with	90%	confidence	intervals	(CI)	for	predictor	variables	of	seahorse	(a)	density	and	(b)	adult	height,	among	seagrass	beds.	Outputs	based	on	results	of	mixed-effects	models.	Parameter	estimates	are	indicated	to	the	right	of	the	CIs.	P-values	are	indicated	as	follows:	superscript	a	=	P	>	0.10,	*	=	0.10	<	P	<	0.05,	**	=	P	<	0.05.	 46			 47		Figure	2.6	Model-averaged	effect	sizes	with	90%	confidence	intervals	(CI)	for	predictor	variables	of	seahorse	height	within	the	Little	Beach	seagrass	bed,	among	(a)	all	seahorses,	(b)	all	reproductively	active	(RA)	seahorses,	(c)	females,	(d)	RA	females,	(e)	males,	(f)	RA	males.	Outputs	based	on	results	of	mixed-effects	models.	Parameter	estimates	are	indicated	to	the	right	of	the	CIs.	P-values	are	indicated	as	follows:	superscript	a	=	P	>	0.10,	*	=	0.10	<	P	<	0.05,	**	=	P	<	0.05.	Total	N	=	328;	includes	168	female,	148	male,	and	12	juvenile	(reproductively	inactive)	observations.		 48	Chapter	3 Review	paper:	the	diet	and	feeding	behaviours	of	a	family	of	biologically	diverse	marine	fishes	(Family	Syngnathidae)			3.1	 Introduction			 An	animal’s	ability	to	meet	the	energetic	demands	of	growth	and	reproduction	are	directly	related	to	its	ability	to	capture	and	consume	prey.	For	predators,	the	energy	and	nutrients	required	for	biological	functions	comes	exclusively	from	their	prey	(Paine,	1966;	Pimm,	1982).	Those	that	fail	to	capture	and	consume	enough	food	to	meet	their	energetic	and	nutritional	demands	will	either	die,	or	will	be	unable	to	allocate	energy	towards	growth	and	reproduction	(Schoener,	1971;	Hislop	et	al.,	1978;	Chambers	&	Trippel,	1997;	Lester	et	al.,	2004;	Kooijman,	2010).	However,	predators	must	first	find,	catch,	and	handle	prey	before	they	are	able	to	consume	them	(Schoener,	1971;	Pyke	et	al.,	1977;	Gendron	&	Staddon,	1983).	Each	of	these	steps	expends	energy	and	increases	the	chances	of	a	failed	attack,	forcing	predators	to	optimize	their	foraging	and	make	difficult	prey	selection	decisions	that	net	them	the	greatest	energy	benefit	(Pyke	et	al.,	1977).	Many	animals,	for	example,	will	only	eat	larger	prey	because	they	are	more	profitable,	and	do	not	require	any	more	handling	time	than	smaller	prey	(Richardson	&	Verbeek,	1986;	Costa,	2009).		Because	of	the	high	fitness	consequences	of	failing	to	meet	energy	demands,	predators	evolve	morphologies	and	behaviours	that	enable	them	to	exploit	their	prey	resources	better	(Schoener,	1971;	Dawkins	&	Krebs,	1979).	Predation	is	a	huge	selective	pressure	for	both	predators	and	their	prey,	and	can	result	in	an	“arms	race”:	the	rapid	co-evolution	of	traits	selected	to	either	aid	in	the	capture	of	prey,	or	in	the	avoidance	of	predators	(Dawkins	&	Krebs,	1979;	Abrams,	2000).	Predators	evolve	energetically	expensive	offensive	tactics	that	can	influence	development,	morphology,	or	behavioural		 49	traits	(Dawkins	&	Krebs,	1979;	Abrams,	2000).	Examples	include	the	evolution	of	morphologies	that	help	predators	seek	(Warrant	&	Locket,	2004),	capture	(Norton,	1991)	and	consume	(West	et	al.,	1991)	prey,	as	well	as	specialized	search	behaviours	(Schoener,	1971),	and	physiological	resistance	to	defensive	neurotoxins	excreted	by	prey	(Brodie	&	Brodie,	1999).	In	marine	ecosystems	particularly,	the	intimate	coupling	of	predators	and	their	prey	is	mediated	by	their	three-dimensional	habitats	(Rose,	2000).	The	vast	majority	of	marine	species	can	move	in	three	dimensions,	allowing	organisms	more	physical	space	in	which	to	live	(Crowder	&	Norse,	2008).	In	marine	environments,	this	structure	is	provided	by	the	liquid	medium	itself,	and	by	foundation	species	(e.g.	seagrasses,	coral,	macrophytes)	that	dramatically	increase	the	structural	complexity	of	a	landscape	(Jones	et	al.,	1994;	Thorne-Miller,	1999;	Bruno	&	Bertness,	2001).	As	in	terrestrial	environments,	habitat	complexity	provides	prey	with	refuge	from	predation	(Jones	et	al.,	1994,	1997;	Bruno	&	Bertness,	2001)	while	providing	predators	with	a	backdrop	from	which	to	ambush	prey	that	cannot	see	them	(Orth	&	Heck,	1980;	James	&	Heck,	1994).	Unlike	in	most	terrestrial	environments,	however,	marine	predators	must	pursue	prey	in	three	dimensions.	Even	some	of	the	most	sedentary,	benthic	marine	crustaceans	are	capable	of	swimming	away	from	predators	(Main,	1987;	Orav-Kotta	&	Kotta,	2004;	Zamzow	et	al.,	2010).	To	capture	prey	that	are	able	to	both	hide	and	move	in	three	dimensions,	some	marine	predators	have	evolved	unique	prey	capture	techniques.		Syngnathid	fishes	(Family	Syngnathidae)	are	a	family	of	marine	predators	that	ambush	small	prey	items	in	complex	habitats.	The	300	described	species—within	57	genera—live	in	seagrass,	coral	reefs,	mangroves,	macrophytes,	and	artificial	structures	in		 50	tropical	and	temperate	coastal	waters	around	the	world	(Froese	&	Pauly,	2017).	They	are	notable	for	their	extensive	male	parental	care,	which	varies	by	genus	from	simple	ventral	gluing	of	the	eggs	through	a	series	of	pouch	enhancements	to	the	fully	sealed	brood	pouch	of	the	seahorses.	Syngnathids	have	highly	advanced	prey-capture	techniques	that	aid	them	in	the	capture	of	prey	(Howard	&	Koehn,	1985;	Kendrick	&	Hyndes,	2005).	All	species	use	pivot	feeding,	the	rapid	propulsion	of	the	mouth	towards	a	prey	item	(de	Lussanet	&	Muller,	2007;	Van	Wassenbergh	et	al.,	2011,	2014).	This	technique	is	faster	than	the	reaction	time	of	even	their	fastest	prey	items	(Gemmell	et	al.,	2013),	and	is	a	large	reason	they	are	successful	marine	predators.	Syngnathids	also	differ	greatly	in	their	feeding-related	morphologies,	as	adult	body	sizes	range	30	to	600	mm	(standard	length)	and	they	have	snouts	that	vary	dramatically	in	shape	and	size	(Dawson	1982,	1985).		We	have	considerable	amounts	of	fragmented	information	about	syngnathid	diets	and	feeding	behaviours.	Syngnathid	feeding	mechanics	have	been	the	focus	of	numerous	studies	that	investigated	their	structure,	kinematics	and	evolutionary	development	(e.g.	Bergert	&	Wainwright	1997;	Van	Wassenbergh	et	al.,	2009,	2011,	2014,	Roos	et	al.,	2010,	2011).	Other	studies	have	investigated	the	role	that	different	variables	have	on	syngnathid	feeding,	including,	but	not	limited	to,	morphology	(e.g.	Howard	&	Koehn	1985;	Kendrick	&	Hyndes	2005),	sex	and	reproductive	status	(e.g.	D’Entremont,	2002;	Berglund	et	al.,	2006;	Kitsos	et	al.,	2008),	ontogenetics	(e.g.	Brown,	1972;	Castro	et	al.,	2008),	habitat	structure	(e.g.	Howard	&	Koehn,	1985;	Curtis	&	Vincent,	2005;	Kendrick	&	Hyndes,	2005),	and	diurnal/seasonal	variability	(e.g.	Ryer	&	Orth,	1987;	Woods,	2002;	Uncumusaoglu	et	al.,	2017).	In	addition,	dozens	of	studies	have	tabulated	syngnathid	diets	in	the	wild	(e.g.	Steffe		 51	et	al.,	1989;	Gaughan	&	Potter	1997;	Woods,	2002;	Kendrick	&	Hyndes,	2005;	Kitsos	et	al.,	2008;	Smith	et	al.,	2011a).		While	numerous	studies	have	looked	at	the	diets	and	foraging	behaviours	of	syngnathids,	no	work	has	summarized	general	patterns	in	the	mechanics	and	morphologies	involved	in	syngnathid	feeding	events.	Also,	no	study	has	explored	how	syngnathid	diets	and	foraging	behaviours	vary	across	genera,	locations,	and	studies	within	this	morphologically	diverse	family	of	fishes,	which	lives	in	a	variable	three-dimensional	space.	Our	first	goal	is	to	provide	a	resource	that	summarizes	important	ecological	knowledge	about	syngnathid	feeding	in	the	wild.	Our	second	goal	is	to	identify	general	patterns	in	syngnathid	diets	and	discern	how	they	are	associated	with	body	characteristics.	Here	we	test	the	hypothesis	that	syngnathid	feeding	morphologies—which	differ	considerably	within	the	family—are	correlated	with	their	diets,	across	genera,	locations,	and	studies.	Specifically,	we	look	at	syngnathid	morphologies	that	are	directly	related	to	the	size	and	speed	of	prey	they	can	capture	(i.e.	snouts,	gapes),	and	those	that	can	indirectly	affect	where	syngnathids	feed	(i.e.	overall	body	size,	and	relative	fin	sizes).	We	expect	there	to	be	large	variation	in	the	diets	of	syngnathids	in	these	analyses	due	to	differences	in	prey	availability.	Additionally,	we	expect	that	snout	morphologies—which	have	been	shown	to	affect	syngnathid	diets	in	the	past	(Kendrick	&	Hyndes,	2005)—will	be	correlated	with	what	they	eat.	In	this	review	we	synthesize	information	about	the	diets	and	feeding	behaviours	of	syngnathids	in	the	wild,	with	a	focus	on	exploring	the	associations	between	syngnathid	morphologies	and	their	diets.				 52	3.2	 Methods		 We	conducted	a	meta-analysis	of	peer-reviewed	literature,	unpublished	university	dissertations,	and	government	reports	on	the	diet	and	feeding	of	syngnathids	to	summarize	the	known	information.	The	emphasis	of	this	study	was	to	evaluate	how,	what,	where	and	when	syngnathids	eat,	and	what	affects	their	diets	and	feeding	behaviours.	We	provide	a	review	of	the	literature,	and	perform	comparative	analyses	to	see	if	there	are	patterns	in	syngnathid	diets	across	genera,	locations,	and	studies.		3.2.1	 Literature	review		 We	used	information	from	and	sources	cited	in	FishBase	(Froese	&	Pauly,	2017),	and	materials	found	during	searches	on	Google	Scholar	up	until	June	2017.	We	divided	search	terms	into	two	groups.	(i)	taxa:	syngnathid*,	pipefish,	seahorse,	seadragon,	and	all	57	syngnathid	genera	names	(e.g.	Hippocampus,	Syngnathus),	and	(ii)	diet:	diet,	feed*,	forag*,	prey-capture.	We	then	searched	all	pairwise	combinations	using	one	term	from	each	group.	Our	goal	was	to	obtain	the	broadest	possible	understanding	of	syngnathid	feeding	in	the	wild.	Our	narrative	preferentially	included	in-situ	studies,	and	supplemented	them	with	ex-situ	studies	if	they	included	novel	information	not	already	found	on	syngnathids	in	the	wild.	Our	statistical	analyses	were	based	on	a	subset	of	the	diet	studies,	including	studies	conducted	on	syngnathids	in	the	wild	which	involved	more	than	two	individuals.				 53	3.2.2	 Diet	taxonomy		 We	categorized	all	prey	items	reported	quantitatively	in	in-situ	diet	studies	using	the	World	Register	of	Marine	Species	(www.marinespecies.org).	Converting	all	cited	species	to	current	taxonomic	classifications	made	the	diet	studies	consistent	and	comparable.	Dietary	data	were	reported	as	either	bulk	data	(the	relative	contribution	that	a	food	item	made	to	volume,	weight,	or	area	of	the	stomach	contents),	numeric	data	(the	relative	contribution	that	a	food	item	made	to	the	total	number	of	food	items	in	stomach	contents)	or	frequency	of	occurrence	(FO;	the	number	of	stomachs	that	contained	a	certain	food	item).	To	compare	the	greatest	number	of	studies	at	similar	taxonomic	resolutions,	we	summed	all	additive	dietary	information	(bulk	and	numeric	data)	from	lower	into	higher	taxonomic	categories	when	not	explicitly	calculated	by	the	authors.	For	example,	although	most	studies	did	not	report	the	volume	of	Crustacea	(subphylum)	explicitly,	they	did	report	volumes	of	lower	taxonomic	levels	(i.e.	Peracarida	and	Eucarida)	that	encompass	Crustacea,	such	that	the	volume	of	crustaceans	could	be	inferred.	We	also	evaluated	non-additive	information	(FO),	but	these	data	could	not	be	summed	to	higher	taxonomic	levels.	If	the	FO	of	a	taxonomic	level	was	not	explicitly	reported,	we	estimated	its	value	by	recording	the	maximum	value	of	taxonomic	levels	below,	and	within	it.	For	example,	if	the	frequency	of	Crustacea	was	not	reported,	we	estimated	it	by	recording	the	maximum	value	of	the	taxonomic	groups	that	were	within	it.	If	Peracarida	FO	was	55%,	Eucarida	45%,	and	Copepoda	15%,	Crustacea	would	be	estimated	as	55%.	Since	we	do	not	know	which	stomachs	had	which	items	(the	basis	for	FO	data),	we	can	only	say	“at	least	55%	of	stomachs	had	Crustacea".			 54	3.2.3	 Syngnathid	characteristics		 For	all	syngnathid	species	with	in-situ	dietary	data,	we	recorded	maximum	standard	length	(StL),	and	ratios	of	standard	length	to	head	length	(StL:HL),	head	length	to	snout	length	(HL:SnL),	and	snout	length	to	snout	depth	(SnL:SnD).	Standard	length	was	recorded	from	FishBase	(Froese	&	Pauly,	2017).	Ratio	information	was	recorded	from	species	descriptions	reported	in	field	guides,	measured	from	specimens	at	the	Australian	Museum	in	Sydney,	Australia,	and	measured	from	photos	(Table	C1).	For	consistency,	we	also	preferentially	selected	information	in	that	order.	We	selected	mean	ratios	when	available,	but	took	the	median	value	if	reported	as	a	range.	We	then	used	these	ratios	to	estimate	head	length,	snout	length,	and	snout	depth	from	StL	for	each	species.	We	were	forced	to	standardize	syngnathid	sizes	in	this	way	because	studies	rarely	reported	the	specific	sizes	of	syngnathids	used	in	their	studies	(and	almost	never	reported	HL,	SnL,	or	SnD).	We	measured	the	relative	size	of	caudal	fins	(relative	fin	size)	by	measuring	the	area	of	the	caudal	fin	and	dividing	it	by	the	area	of	the	syngnathid’s	body,	calculated	from	their	profile	view.	For	consistency,	we	preferentially	selected	adult	females	and	did	not	include	pectoral	or	anal	fins	in	our	calculations.	We	preferentially	analyzed	photos	over	sketches	(Table	C1),	and	because	of	the	time	required	to	process	each	image	and	the	lack	of	quality	profile	images	available	for	most	species,	we	calculated	all	areas	on	one	image	per	species	in	ImageJ	version	1.59m9m	(National	Institutes	of	Health,	Bethesda,	MD).					 55	3.2.4	 		Statistical	analyses		3.2.4.1		Associations	between	syngnathid	characteristics	and	their	diets			 To	determine	what	body	characteristics	are	correlated	with	the	diets	of	syngnathids,	we	executed	redundancy	analyses	(RDA;	Van	Den	Wollenberg,	1977;	McArdle	&	Anderson,	2001)	on	bulk	(volume,	weight,	and	area	data),	numeric,	and	frequency	of	occurrence	(FO)	dietary	data.	In	total,	we	included	48	diets	in	our	analysis	of	bulk	dietary	data	(across	14	genera,	34	species,	and	28	studies),	33	diets	for	numeric	data	(across	5	genera,	19	species,	and	16	studies)	and	40	diets	for	FO	data	(across	12	genera,	30	species,	and	18	studies).	The	justification	for	the	(i)	dependent	and	(ii)	independent,	and	(iii)	possible	conditioning	variables	included	in	this	model	is	explained	below.	First,	based	on	our	additive	calculations	(see	above),	we	considered	dietary	response	variables	using	a	blend	of	taxonomy	and	biological	knowledge.	Dietary	data	are	often	reported	within	functional	groups	that	do	not	align	perfectly	with	taxonomic	groupings.	For	example,	many	studies	report	the	volume	of	amphipods	and	copepods	in	the	same	list,	although	Amphipoda	is	an	order	and	Copepoda	is	a	subclass.	We	therefore	considered	the	lowest	level	of	taxonomic	resolution	that	could	be	calculated	for	all	studies,	within	these	functional	groups.	As	another	example,	because	we	cannot	calculate	the	values	for	individual	orders	of	Copepoda	(i.e.	calanoid,	cyclopoid	and	harpacticoids)	for	some	studies	(those	that	only	report	to	the	level	of	Copepoda),	the	lowest	taxonomic	resolution	across	all	studies	within	this	functional	group	is	to	the	level	of	Copepoda.	The	following	ten	response	variables	(prey	items)	were	therefore	considered:	amphipods,	copepods,	decapods,	gastropods,	isopods,	mysids,	ostracods,	polychaetes,	tanaids,	total	eggs	(sum	of	eggs	from	all	prey	taxa),	and	total	larvae	(sum	of	larvae	from	all	prey	taxa).	To	conserve		 56	degrees	of	freedom	in	our	model,	we	then	removed	food	items	that	had	a	mean	of	less	than	5%	across	all	samples.	For	bulk	data,	we	retained	the	following	five	prey	items:	amphipods,	copepods,	mysids,	decapods,	and	total	eggs.	For	numeric	data,	we	retained	five	prey	items:	amphipods,	copepods,	mysids,	decapods,	and	ostracods.	For	FO	data,	we	retained	seven	prey	items:	amphipods,	copepods,	mysids,	decapods,	ostracods,	total	larvae,	and	‘other’	(which	included	various	items	such	as	algae,	sediment,	insects,	etc.).	Second,	as	candidate	independent	variables,	we	included	relative	fin	size,	maximum	standard	length	(StL),	head	length	(HL),	snout	length	(SnL),	snout	depth	(SnD),	the	ratio	of	head	length	to	snout	length	(HL:SnL),	and	the	ratio	of	snout	length	to	snout	depth	(SnL:SnD)	of	syngnathid	species.	We	used	Variance	Inflation	Factor	(VIF)	to	determine	and	deal	with	multicollinearity.	We	applied	VIF	to	all	predictor	variables	and	removed	the	variables	contributing	the	most	variance	(i.e.	the	highest	VIF)	until	the	VIF	scores	of	all	predictors	was	less	than	5	(Zuur	et	al.,	2010).	HL	and	SnL	were	removed	from	all	three	models.	StL:HL	was	not	considered	as	a	candidate	variable	because	it	inflated	the	VIF	of	all	models.	Third,	we	wanted	to	determine	if	it	was	necessary	to	control	for	a	certain	variable	in	our	model,	by	partialing	out	its	variance.	Therefore,	to	determine	the	relative	ability	different	components	of	the	model	had	in	explaining	the	variation	in	syngnathid	diets,	we	used	an	RDA	to	partial	out	the	variance	into	syngnathid	body	characteristics	and	genus	(see	Figure	3.1).	If	syngnathid	body	characteristics	were	to	explain	a	large	proportion	of	the	variance	in	their	diets,	the	relative	variation	explained	purely	by	the	body	characteristics,	independent	of	any	covariance	with	genus,	would	be	high	[component	a].	A	similar	trend	would	be	seen	if	a	large	contribution	of	the	variation	was	explained	by	the	pure	effects	of		 57	their	genus	[component	c].	However,	diet	may	also	depend	on	the	covariance	between	body	characteristics	and	genus	[component	b].	Some	variance	in	diet	will	remain	unexplained	[component	d].	As	is	evident,	very	little	of	the	variance	in	bulk	diet	is	explained	by	the	pure	effects	of	genus	(component	c;	Figure	3.1a).	A	much	higher	proportion	of	variance	is	explained	by	the	pure	effects	of	body	characteristics	[component	a;	Figure	3.1a]	and	the	covariance	between	syngnathid	body	characteristics	and	genus	(component	b;	Figure	3.1a).	For	numeric	data,	no	variance	was	explained	by	genus	alone	(component	c;	Figure	3.1b),	and	little	variance	was	explained	by	the	covariance	between	syngnathid	body	traits	and	genus	(component	b;	Figure	3.1b).	For	FO	data,	only	7%	of	the	variance	was	explained	by	genus	alone	(component	c;	Figure	3.1c).	This	information,	in	conjunction	with	the	high	phylogenetic	signals	of	these	characteristics	(Table	3.1),	provides	justification	for	removing	the	variance	explained	by	genus	alone	in	our	three	models.	As	a	result,	no	conditional	terms	were	considered	in	either	RDA.		To	test	the	significance	that	syngnathid	body	characteristics	had	on	bulk	and	numeric	syngnathid	diets,	we	executed	a	permuational	multivariate	analysis	of	variance	(PERMANOVA)	with	999	permutations.	We	did	all	variance	partitioning	and	RDA	analyses	using	the	vegan	package,	version	2.4-3	in	R	3.4.1	(www.r-project.org).		3.2.4.2		Phylogenetic	signal	of	syngnathid	characteristics		 We	estimated	the	phylogenetic	signal	of	syngnathid	body	characteristics	using	Bloomberg’s	K	statistic.	Phylogenetic	signal	measures	the	similarity	of	related	species	(with	respect	to	a	trait),	as	compared	to	species	drawn	randomly	from	a	phylogenetic	tree		 58	(Blomberg	&	Garland,	2002).	A	K	statistic	of	zero	indicates	that	the	trait	values	are	randomly	distributed	within	the	phylogeny,	and	a	value	approaching	1	indicates	that	the	trait	values	are	evolving	as	expected	by	Brownian	motion	models	(Blomberg	et	al.,	2003).	A	value	over	1	means	the	trait	has	more	phylogenetic	signal	than	expected	from	evolutionary	models.	We	used	the	most	recent	and	comprehensive	phylogenetic	tree	for	syngnathids	(Hamilton	et	al.,	2017),	and	estimated	the	phylogenetic	signals	using	the	phytools	package,	version	0.6-00	in	R	3.4.1	(www.r-project.org).		3.3	 		Results		3.3.1	 		How	do	syngnathids	eat?		 	3.3.1.1		Head	morphology	&	mechanics	of	a	feeding	event			 Although	syngnathid	feeding	mechanisms	vary	across	the	family	(Van	Wassenbergh,	et	al.,	2011),	species	share	a	number	of	general	characteristics.	Syngnathids	have	jaws	that	are	tilted	slightly	upwards	at	the	end	of	an	elongated	snout	and	lack	teeth	(Figure	3.2;	Bergert	&	Wainwright	1997;	Flammang	et	al.,	2009).	Syngnathid	snouts	are	rigid,	tubular	structures	made	of	specialized	neurocranial	and	suspensorial	bones	(Bergert	&	Wainwright,	1997;	Flammang	et	al.,	2009).	Syngnathid	feeding	involves	a	coupling	between	hyoid	rotation	and	neurocranial	elevation	and	is	powered	by	elastic	energy	generated	by	specialized	muscle	groups	(Figure	3.2;	Muller	&	Osse,	1984;	Bergert	&	Wainwright	,1997;	Van	Wassenbergh	et	al.,	2008;	Flammang	et	al.,	2009;	Roos	et	al.,	2009a).	The	hyoid	is	a	V-shaped	bone	complex	on	the	bottom	side	of	the	snout,	running	parallel	to	the	snout	when	in	a	resting	position.	The	expansive	phase	(see	next	paragraph)	of	a	feeding	event	is		 59	initiated	by	hyoid	rotation,	and	occurs	about	half	a	second	before	the	start	of	neurocranial	elevation	(Table	C2;	Roos	et	al.,	2009b).	By	contracting	the	sternohyoideus	muscles,	which	run	along	the	top	and	bottom	side	of	syngnathids	(Figure	3.2;	Van	Wassenbergh	et	al.,	2008,	2014),	the	hyoid	arch	is	rotated	away	from	the	snout	(Muller	&	Osse,	1984;	Bergert	&	Wainwright,	1997).	The	bones	involved	in	hyoid	rotation	cause	suspensorial	bones	of	the	snout	to	be	driven	away	from	the	body,	towards	their	prey	(Flammang	et	al.,	2009;	Van	Wassenbergh	et	al.,	2013).	As	the	skull	expands	laterally	during	head	rotation,	it	causes	a	build-up	of	negative	pressure	within	the	head	that	is	used	to	inhale	the	prey	item	(Bergert	&	Wainwright,	1997;	Van	Wassenbergh	et	al.,	2013).		3.3.1.2		Stages	of	a	feeding	event			 Syngnathids	are	among	the	fastest	feeders	of	all	fish,	capturing	prey	between	3	and	8	ms	after	initiating	an	attack	(Table	C2;	Bergert	&	Wainwright,	1997;	Van	Wassenbergh	et	al.,	2011a).	Syngnathid	feeding	events	share	similarities	with	other	suction	feeders,	involving	preparation,	expansion	and	recovery	phases	(Lauder,	1985;	Bergert	&	Wainwright,	1997).	During	the	preparatory	phase,	syngnathids	visually	scan	their	environment	in	search	of	prey.	For	a	syngnathid	attached	to	a	holdfast,	this	involves	using	its	cryptic	ability	and	waiting	for	a	prey	item	to	move	into	its	area	(James	&	Heck,	1994).	If	unattached,	the	syngnathid	may	actively	swim	through	the	water	column	in	search	of	prey	(James	&	Heck,	1994).	Syngnathids	preferentially	select	oddly	positioned	or	coloured	prey	in	swarms	because	they	are	more	easily	tracked,	and	overall	syngnathids	are	more	successful	at	capturing	prey	from	smaller	groups	(Ocken	&	Ritz,	1994).	Once	prey	are		 60	located,	syngnathids	fix	their	independently	moving	eyes	on	it	(James	&	Heck,	1994;	Ocken	&	Ritz,	1994).	A	syngnathid	must	then	successfully	orient	its	mouth	close	enough	to	a	prey	that	the	mouth	can	reach	the	prey	with	a	pivot	of	its	head	(Muller	&	Osse,	1984).	This	approach	may	involve	extending	their	body	closer	to	the	prey	if	the	syngnathid	is	attached	to	holdfast,	or	swimming	towards	it	if	unattached	(Ryer,	1988;	Felicio	et	al.,	2006).	Slower	approaches	are	more	successful	than	faster	ones,	as	they	are	less	likely	to	elicit	a	prey	escape	response	(Hippocampus	zosterae;	Gemmell	et	al.,	2013).	Syngnathids	have	evolved	a	head	morphology	that	limits	fluid	disturbance	during	a	feeding	strike,	reducing	the	chances	of	being	detected	by	prey	(Gemmell	et	al.,	2013).	Feeding	events	are	accompanied	by	a	'clicking'	sound	(Bergert	&	Wainwright,	1997;	Colson	et	al.,	1998;	Ripley	&	Foran,	2007;	Haris	et	al.,	2014).		 After	positioning	its	body	so	the	snout	is	within	a	few	millimeters	of	their	prey	(Table	C2),	a	syngnathid	snaps	its	mouth	towards	its	prey	by	very	rapidly	rotating	its	head	away	from	its	body	(see	detailed	description	in	previous	paragraph;	Bergert	&	Wainwright,	1997;	Flammang	et	al.,	2009;	Roos	et	al.,	2009a).	Its	mouth	is	propelled	along	an	arc	towards	its	prey	in	less	than	10	ms	(Table	C2;	Van	Wassenbergh	et	al.,	2011a),	much	faster	than	the	reaction	latency	of	planktonic	copepods,	themselves	among	the	fastest	syngnathid	prey	items	(Gemmell	et	al.,	2013).	The	explosive	head	rotation	allows	the	syngnathid	to	maintain	a	constant	body	position,	unlike	most	teleosts	which	lunge	when	feeding	(Bergert	&	Wainwright,	1997).	Following	a	feeding	strike,	during	the	recovery	phase,	the	syngnathid	head	and	hyoid	return	to	resting	positions,	and	the	snout	volume	returns	to	normal	(Muller	&	Osse,	1984;	Bergert	&	Wainwright,	1997).	Recovery	generally	takes	about	one	second	(Bergert	&	Wainwright,	1997).		 61	3.3.1.3		Energetics	of	feeding			 Syngnathids	have	low	metabolic	rates,	and	eat	numerous	low-energy	prey	items	in	the	attempt	to	meet	their	energetic	demands	(Dunham,	2010).	As	a	result	of	having	short,	rudimentary	guts,	and	no	differentiated	stomachs,	syngnathids	are	relatively	inefficient	at	extracting	energy	from	their	food	(Dunham,	2010).	Syngnathids	also	have	relatively	inefficient	gills	which	reduce	their	ability	to	metabolize	the	food	they	do	eat	(Prein	&	Kunzmann,	1987).	This	effect	is	exacerbated	because	syngnathids	feed	on	small	crustaceans	with	high	surface	area-volume	ratios	(i.e.	relatively	low	energy	prey),	meaning	that	they	must	eat	proportionally	more	frequently	in	order	to	acquire	the	energy	that	could	be	extracted	from	larger	prey	(Kooijman,	2000).	At	times	when	prey	availability	is	not	limiting,	gut	volume	and	rates	of	gut	evacuation	may	limit	feeding	rates	of	seahorses	(Sheng	et	al.,	2006)	and	pipefishes	(Ryer	&	Boehlert,	1983).	In	contrast,	at	times	of	low	prey	availability,	syngnathids	may	reduce	the	gut	evacuation	rate,	and	increase	assimilation	efficiency	to	maximize	energy	gains	(Ryer	&	Boehlert,	1983).		Research	on	captive	feeding	studies	suggests	seahorses	eat	proportionally	more	than	pipefishes.	Pipefish	have	been	reported	to	feed	at	a	rate	of	between	3.8	and	4.5%	of	their	body	mass	per	day	(Ryer	&	Boehlert,	1983;	Bennett	&	Branch,	1990),	while	seahorses	consume	as	much	as	23.4%	of	their	body	mass	per	day	(Do	et	al.,	1996;	Lin	et	al.,	2007).	Juvenile	seahorses	have	been	reported	to	consume	214-360	copepod	larvae	per	hour	(Herald	&	Rakowicz,	1951;	Payne	&	Rippingale,	2000).	Temperatures	deviating	from	birth-temperature,	and	high	water	velocities	have	been	shown	to	decrease	seahorse	feeding	rates	(Sheng	et	al.,	2006;	Qin	et	al.,	2014).	Low	temperatures	restrict	seahorse	eating	physiologically,	as	gut	evacuation	rates	increase	with	temperature	(Ryer	&	Boehlert,	1983).		 62	3.3.1.4		Diurnal	timing	of	feeding		 Syngnathids	are	visual	feeders	with	diurnal	feeding	cycles	that	are	dictated	by	light	availability.	A	comparison	by	Kendrick	(2002)	between	daytime	and	nighttime	stomach	contents	among	twelve	syngnathid	species	suggested	most	fed	predominantly	during	the	day.	This	study	looked	at	Stigmatopora	argus	and	Stigmatopora	nigra	feeding	in	greater	detail,	showing	a	similar	pattern	between	gut	fullness	and	the	percentage	of	undigested	prey	in	pipefish	stomachs	(Kendrick,	2002).	For	both	species,	gut	fullness	and	undigested	prey	increased	sharply	after	sunrise,	continued	to	increase	until	shortly	after	noon,	decreased	from	after	noon	until	shortly	before	midnight,	and	remained	low	until	the	next	sunrise	(Kendrick,	2002).	Other	pipefishes	have	shown	similar	patterns,	eating	exclusively	during	daylight	hours	(Syngnathus	acus;	Bennett	&	Branch,	1990).	At	night,	at	least	some	seahorses	attached	themselves	to	holdfasts	where	they	remained	mostly	inactive	(Hippocampus	reidi;	Felicio	et	al.,	2006),	with	empty	guts	or	guts	that	contained	only	highly	digested	prey	(Hippocampus	histrix	and	Hippocampus	trimaculatus;	Do	et	al.,	1996).	To	date	only	two	species	of	seahorses	have	been	reported	feeding	at	night;	Hippocampus	comes	were	found	foraging	early	in	the	morning,	possibly	as	a	result	of	fishing	pressure	(Perante	et	al.,	2002),	and	Hippocampus	erectus	fed	more	actively	at	night	in	Sweetings	Pond,	Bahamas	(S.	Foster,	personal	observation).	Likewise,	the	only	evidence	of	nocturnal	pipefish	feeding	was	among	Syngnathus	folletti—however	even	they	consumed	more	prey	and	on	more	prey	types	during	the	day	(Garcia	et	al.,	2005).	Captive	seahorses	are	able	to	feed	in	low-light	conditions,	but	will	rarely	feed	in	complete	darkness	(James	&	Heck,	1994;	Perante	et	al.,	2002;	Felicio	et	al.,	2006;	Garcia	et	al.,	2012).	Juvenile	seahorses	were	found	to	move	and	feed	continuously	if	provided		 63	constant	light	(Hippocampus	trimaculatus;	Sheng	et	al.,	2006).	When	provided	with	a	simulated	24-hour	day,	peak	seahorse	feeding	occurred	in	the	morning	between	06:00	and	08:00	(Hippocampus	kuda;	Do	et	al.,	1996),	their	greatest	gut	fullness	was	at	08:00	(Truong	&	Nga,	1995),	and	they	did	not	feed	at	all	between	20:00	and	06:00	(Truong	&	Nga,	1995).		3.3.2	 How	does	feeding	and	diet	vary	across	a	speciose	marine	fish	family	that	is			morphologically	diverse?			3.3.2.1	Prey	items	by	genus	and	species		 The	sources	and	citations	for	this	entire	section	are	found	in	Table	3.1.	We	explore	all	syngnathids	before	singling	out	the	Hippocampus	(all	seahorses)	and	the	Syngnathus	pipefishes,	as	they	are	the	most	frequently	cited	genera.	In	this	section	only,	the	species	names	in	brackets	relate	to	the	syngnathids	and	not	the	prey.		3.3.2.1.1	Entire	family		 Overall,	syngnathids	primarily	eat	small	crustaceans.	Crustaceans	are	responsible	for	more	than	75%	of	the	bulk	and	75%	of	the	numeric	prey	items	of	most	syngnathid	diets.	Across	studies	the	types	of	crustaceans	in	diets	varies	quite	considerably.	However,	amphipods	and	copepods	are	most	often	the	dominant	prey	types	in	both	bulk	and	numeric	data.	Although	they	usually	account	for	a	relatively	low	percentage	of	bulk	and	numeric	diets,	in	some	studies	mysids	and	decapods	are	important	to	syngnathid	diets.	Amphipods	and	copepods	are	also	found	in	most	syngnathid	diets	(FO).	Although	to	a	lesser	extent,	mysids	and	decapods	are	also	frequently	(FO)	found	in	many	syngnathid	stomachs.		 64	3.3.2.1.2	Hippocampus	(seahorses)		 Amphipods	are	the	main	bulk	dietary	item	for	a	number	of	seahorse	species	including	Hippocampus	breviceps,	Hippocampus	coronatus,	Hippocampus	erectus	and	Hippocampus	whitei.	Other	primary	bulk	items	include	copepods	(Hippocampus	zosterae)	and	total	eggs	(Hippocampus	hippocampus).	The	primary	numeric	prey	items	include	amphipods	(Hippocampus	guttulatus	and	Hippocampus	hippocampus),	copepods	(Hippocampus	mohnikei,	and	Hippocampus	zosterae),	and	decapods	(Hippocampus	patagonicus).	Amphipods	also	occur	in	more	stomachs	(FO)	than	any	other	prey	item	for	Hippocampus	breviceps,	Hippocampus	erectus,	Hippocampus	guttulatus,	Hippocampus	patagonicus,	and	Hippocampus	subelongatus.	Decapods	and	copepods	occur	in	a	large	number	of	stomachs	(FO)	for	many	seahorse	species,	mysids	are	eaten	by	many	Hippocampus	guttulatus,	and	tanaids	by	Hippocampus	histrix.		3.3.2.1.3	Syngnathus	pipefishes		 Copepods	are	the	major	bulk	prey	item	for	Syngnathus	abaster	and	Syngnathus	acus.	Other	primary	bulk	prey	items	include	decapods	(Syngnathus	floridae,	Syngnathus	typhle),	isopods	(Syngnathus	folletti),	and	ostracods	(Syngnathus	louisianae).	Copepods	are	the	most	numerous	dietary	item	for	Syngnathus	abaster,	Syngnathus	taenionotus,	and	Syngnathus	typhle.	Amphipods,	decapods,	copepods	and	mysids	occur	in	a	large	number	of	stomachs	(FO).				 65	3.3.2.1.4	Seadragons	(Phyllopteryx)	and	other	pipefishes		 Amphipods	are	the	dominant	bulk	prey	item	for	Filicampus	tigris,	Histiogamphelus	cristatus,	Pugnaso	curtirostris,	and	Vanacampus	phillipi,	and	are	a	major	contributor	to	diets	of	Lissocampus	caudalis,	Mitotichys	semistriatus,	and	Nerophis	ophidion.	Other	primary	dietary	items	include	copepods	(Stigmatopora	spp.,	Lissocampus	caudalis,	Mitotichthys	meraculus,	Urocampus	carinirostris),	mysids	(Mitotichthys	meraculus,	Phyllopteryx	taeniolatus,	Vanacampus	poecilolaemus),	and	decapods	(Syngnathoides	biaculeatus).	For	numeric	data,	primary	diet	contributors	are	copepods	(Anarchopterus	criniger,	Nerophis	lumbriciformis,	Urocampus	carinirostris).	For	frequency	data,	primary	contributors	are	amphipods	(Filicampus	tigris,	Histiogamphelus	cristatus,	Vanacampus	phillipi),	mysids	(Mitotichthys	meraculus,	Phyllopteryx	taeniolatus,	Vanacampus	poecilolaemus),	and	copepods	(Stigmatopora	spp.).			3.3.2.2	Morphology		3.3.2.2.1	Phylogenetic	signal	of	syngnathid	morphological	characteristics		 Across	syngnathids,	feeding	morphologies	were	more	similar	among	closely-related	species	that	they	were	among	randomly-selected	species	from	the	clade.	Relative	fin	size,	max	standard	length	(StL),	head	length	(HL),	snout	length	(SnL),	snout	depth	(SnD),	StL:HL,	HL:SnL,	and	SnL:SnD	all	showed	significant	phylogenetic	signal	(Table	3.2).	All	measured	traits	had	K	values	over	0.60,	and	all	but	two	traits	had	K	values	over	0.80.	As	well,	HL	and	SnL	had	K	values	slightly	over	1,	indicating	that	related	species	were	more	similar	to	each	other—with	respect	to	these	traits—than	would	be	expected	from	evolution.		 66	3.3.2.2.2	Body	form	&	orientation			 During	feeding	events,	syngnathid	body-orientation	is	optimized	to	maximize	either	strike	distance,	or	mouth	velocity	(Van	Wassenbergh	et	al.,	2011a).	When	compared	proportionally	to	head	length,	the	curved	body-form	of	seahorses	increases	the	distance	from	which	it	can	capture	prey,	and	the	straight	body	form	of	pipefishes	allows	for	greater	mouth	velocity	for	some	genera	(including	Doryrhamphus	spp.	and	Syngnathus	spp.;	Van	Wassenbergh	et	al.,	2011b).	In	addition	to	the	vertical	mouth	movement	associated	with	all	syngnathid	feeding	strikes,	seahorses	also	move	their	mouths	forward,	towards	prey	(Van	Wassenbergh	et	al.,	2011b).	In	contrast,	many	syngnathids	that	feed	from	a	straight-body	orientation	move	their	mouths	vertically,	and	even	backwards	in	some	cases	(Van	Wassenbergh	et	al.,	2011b).	However,	the	angled,	seahorse-like	head	orientation	and	strike	kinematics	of	the	pipefish	Corythoichthys	intestinalis	displayed	an	intermediate	feeding	body	orientation	(Van	Wassenbergh	et	al.,	2011b).	Like	all	pipefish,	Corythoichthys	intestinalis	has	a	straight-body	form	with	no	angle	between	its	head	and	trunk	(Leysen	et	al.,	2011),	yet	this	syngnathid	bends	its	head	and	moves	its	mouth	forward	during	a	feeding	event,	much	like	a	seahorse.		3.3.2.2.3	Snout	shape		 Across	genera,	locations,	and	studies,	the	relative	length	of	syngnathid	snouts	was	significantly	correlated	with	the	bulk	diets	of	syngnathids,	and	the	frequency	that	prey	items	occurred	in	syngnathid	diets	(Table	3.3).	Syngnathids	with	relatively	longer	snouts	(low	HL:SnL)	tended	to	feed	on	decapods	(e.g.	Syngnathoides	biaculates),	mysids	(e.g.		 67	Phyllopteryx	taeniolatus),	total	eggs	(e.g.	Syngnathoides	biaculates),	copepods	(e.g.	Stigmatopora	spp.)	or	decapods	(e.g.	Syngnathoides	biaculates)	in	bulk	(Figure	3.3),	and	on	‘other’	dietary	items	(e.g.	Mitotichthys	meraculus),	mysids	(e.g.	Vanacampus	poecilolaemus),	decapods	(e.g.	Syngnathus	floridae)	and	total	larvae	(e.g.	Syngnathus	acus)	in	frequency	(FO	data;	Figure	3.4).	In	contrast,	syngnathids	with	relatively	shorter	snouts	tended	to	feed	on	amphipods	(e.g.	Hippocampus	breviceps)	in	bulk	(Figure	3.3),	and	on	ostracods	(e.g.	Lissocampus	caudalis),	amphipods	(e.g.	Pugnaso	curtirostris),	and	copepods	(e.g.	Lissocampus	caudalis)	in	frequency	(Figure	3.4).	Snout	length	and	relative	snout	length	did	not	correlate	with	numeric	diet	data	(Table	3.3).	Snout	shape	appears	to	dictate	from	how	far	away	a	syngnathid	can	attack	a	prey	item,	and	the	speed	with	which	it	can	capture	the	item	(Roos	et	al.,	2010,	2011;	Van	Wassenbergh	et	al.,	2011a).	A	major	difference	between	pipefish	and	seahorse	head	shape	can	be	attributed	to	differences	in	snout	shape,	as	many	pipefish	species	have	snouts	that	are	relatively	longer	and	narrower	than	seahorses	(Leysen	et	al.,	2011).	Longer-snouted	syngnathids	can	attack	prey	from	further	away	as	the	angle	to	reach	prey	is	reduced	compared	to	shorter-snouted	relatives	(de	Lussanet	&	Muller,	2007).	If	snout	width	increased	proportionally	with	snout	length,	the	head	rotation	velocity	of	longer-snouted	syngnathids	would	be	compromised	as	a	result	of	the	increased	drag	(de	Lussanet	&	Muller,	2007;	Leysen	et	al.,	2011).	To	compensate,	species	of	pipefishes	that	have	relatively	longer	snouts	tend	to	have	relatively	narrower	snouts	too,	allowing	for	faster	head	rotation	(de	Lussanet	&	Muller,	2007;	Leysen	et	al.,	2011).	Snout	shape	has	therefore	been	linked	to	the	mobility	of	prey	consumed,	with	longer	snouts	being	better	suited	to	capturing	more	mobile	prey	types	(Kendrick	&	Hyndes,	2005).	In	a	lagoon	off	western	Sicily,	sympatric		 68	pipefishes	maintained	different	food	niches,	purportedly	as	a	result	of	snout	length	differences	(Campolmi	et	al.,	1996).	The	longer-snouted	Syngnathus	typhle	was	better	suited	at	catching	fast	pelagic	organisms	while	the	broader-snouted	Syngnathus	abaster	caught	sessile,	benthic	prey	(Campolmi	et	al.,	1996).	Syngnathids	with	longer	snouts	are	more	specialized	than	those	with	shorter	snouts,	with	shorter	snouted	individuals	eating	a	wider	range	of	prey	(Kendrick	&	Hyndes,	2005).		 The	effect	that	snout	shape	has	on	the	suction	ability	of	syngnathids	is	unclear	from	the	literature.	According	to	Roos	et	al.,	(2011),	smaller	snouts	(both	in	length	and	width)	can	inhale	prey	more	quickly	(Roos	et	al.,	2011).	Muller	and	Osse	(1984)	disagree,	suggesting	that	the	velocity	and	acceleration	of	suction	increases	with	greater	snout	length.			3.3.2.2.4	Gape	size		 Across	genera,	locations,	and	studies,	the	relative	size	of	syngnathid	snout	depths	was	significantly	correlated	with	the	bulk	diets	of	syngnathids,	and	the	frequency	that	prey	items	occurred	in	syngnathid	diets	(Table	3.3).	Syngnathids	with	relatively	deeper	snouts	(low	SnL:SnD)	tended	to	feed	on	amphipods	(e.g.	Hippocampus	erectus)	in	bulk	(Figure	3.3),	and	on	mysids,	‘other’	items,	amphipods,	decapods,	and	total	larvae	in	frequency	(Figure	3.4).	On	the	other	hand,	those	with	relatively	smaller	snout	depths	tended	to	feed	on	copepods	(e.g.	Stigmatopora	spp.)	and	mysids	(e.g.	Phyllopteryx	taeniolatus)	in	bulk	(Figure	3.3),	and	copepods	(e.g.	Stigmatopora	spp.)	in	frequency	(Figure	3.4).	Snout	depth	and	relative	snout	depth	did	not	correlate	with	numeric	diet	data	(Table	3.3).	Gape	size	limits	the	size	of	prey	that	syngnathids	can	consume	(Mercer,	1973;	Ryer,		 69	1988).	Capture	success	decreases	if	a	syngnathid	attempts	to	consume	prey	that	is	proportionately	too	large	for	its	gape	(Mercer,	1973;	Ryer,	1988).	Among	syngnathids	with	similar	snout	lengths,	gape	size	dictates	which	types	of	mobile	prey	items	they	consume	—	with	smaller-gaped	syngnathids	eating	smaller	prey	(Kendrick	&	Hyndes,	2005).	The	size	of	prey	items	consumed	by	syngnathids	often	match	their	gape	size	quite	well	Garcia	et	al.,	2005),	being	between	half	and	three	quarters	the	width	of	the	mouth	(Gaughan	&	Potter,	1997;	Celino	et	al.,	2012).	The	size	of	prey	items	consumed	by	Urocampus	carinirostris,	for	example,	was	limited	by	the	disproportionately	small	gape	size	to	body	size	(Gaughan	&	Potter,	1997).			3.3.2.2.5	Ontogenetics:	changes	in	snout	shape	&	gape	size		 Syngnathids	produce	juveniles	that	are	not	only	equipped	with	fully	developed	prey-capture	abilities—itself	a	rarity	among	fish—but	in	some	cases	also	possess	record-breaking	abilities.	Juvenile	Hippocampus	reidi	have	been	recorded	rotating	their	hyoids	over	90°	at	speed	of	80	000°	s-1,	among	the	fastest	velocities	ever	recorded	among	fish	prey-capture	systems	(Table	C2;	Van	Wassenbergh	et	al.,	2009).	Likewise,	juvenile	head	rotation	was	recorded	at	30	000°	s-1,	more	than	three	times	faster	than	adults	(Table	C2;	Roos	et	al.,	2009b;	Van	Wassenbergh	et	al.,	2009).	It	is	likely	that	the	exceptionally	fast	juvenile	cranial	rotation	abilities	allow	them	to	reach	their	prey	more	quickly	than	adults.	This	is	particularly	impressive	given	the	greater	angle	over	which	they	must	attack.	Juveniles	are	also	able	to	inhale	proportionally	more	water	and	at	a	greater	speeds	than		 70	adults	(Roos	et	al.,	2011).	Together,	it	is	likely	that	the	lightning-quick	feeding	events	that	syngnathids	are	famous	for	is	even	faster	in	juveniles.		 As	individuals	grow,	so	do	their	gapes	and	snouts.	Syngnathids	are	born	with	relatively	shorter	and	broader	snouts	that	lengthen	and	narrow	as	they	age	(Roos	et	al.,	2010,	2011).	Across	all	ontogenetic	stages,	seahorse	snout	length	is	optimized	to	reach	prey	as	quickly	as	possible	(Roos	et	al.,	2010).	These	animals	are	susceptible	to	the	varying	kinematic	constrains	associated	with	differing	snout	morphologies	throughout	their	lives.	Their	snout	morphology	therefore	dictates	the	size	of	prey	that	they	are	capable	of	inhaling.	Changes	in	snout	shape	and	gape	size	are	largely	responsible	for	ontogenetic	prey	choice	differences	(Oliveira	et	al.,	2007;	Sakurai	et	al.,	2009).	As	a	result,	syngnathids	start	as	specialist	feeders	as	juveniles,	feeding	on	a	limited	suite	of	small	prey	and	then	become	more	successful,	generalist	feeders	as	adults,	feeding	on	a	wider	range	of	prey	(Ryer	&	Orth,	1987;	Gaughan	&	Potter,	1997;	Flynn	&	Ritz,	1999;	Oliveira	et	al.,	2007;	Castro	et	al.,	2008;	Taskavak	et	al.,	2010).		While	we	did	not	find	any	across-genera	correlations	between	maximum	standard	length	and	the	bulk,	numeric,	or	FO	diets	of	syngnathids	(Table	3.3),	a	number	of	studies	have	shown	this	trend	within	species.	For	example,	mean	prey	item	weight	increased	with	seahorse	size	for	Hippocampus	guttulatus	and	Hippocampus	hippocampus,	with	small-to-intermediate	sized	Hippocampus	guttulatus	feeding	primarily	on	decapod	larvae	and	larger	individuals	eating	more	mysids	(Gurkan	et	al.,	2011b).	Among	Hippocampus	abdominalis,	larger	individuals	consumed	proportionally	more	caridean	shrimp	than	smaller	individuals,	which	preferred	smaller	prey	such	as	amphipods	(Woods,	2002).	Other	seahorse	species	were	found	to	prey	primarily	on	copepods	as	juveniles	and	then	transition		 71	to	amphipods	(Hippocampus	erectus:	Teixeira	&	Musick,	2001;	Hippocampus	kuda:	Truong	and	Nga	1995)	and	decapods	(Hippocampus	reidi:	Castro	et	al.,	2008;	Hippocampus	mohnikei:	Kanou	and	Kohno	2001)	as	adults.	Syngnathus	spp.	experience	similar	ontogenetic	changes	with	individuals	shifting	towards	larger	prey	types	at	larger	sizes	(Livingston,	1982,	1984;	Ryer	&	Orth,	1987;	Tipton	&	Bell,	1988;	Franzoi	et	al.,	1993;	Teixeira	&	Musick,	1995).	Pipefish	diets	shifted	from	copepods	in	smaller	individuals	to	larger	prey	such	as	amphipods	(Brown,	1972;	Bennett,	1989),	mysids,	and	caridean	shrimp	(Brown,	1972;	Oliveira	et	al.,	2007)	at	larger	sizes.	Smaller	Syngnathus	acus	ate	copepods	and	larger	individuals	ate	decapod	eggs	and	larvae	at	larger	sizes	(Taskavak	et	al.,	2010).	Likewise,	the	large	prey	item,	caridean	shrimp,	were	found	to	be	the	only	prey	in	the	largest	specimens	of	Syngnathus	floridae	(Brook,	1977).			 Contrary	to	the	above	studies,	a	few	reports	indicate	no	ontogenetic	changes	in	syngnathid	diets	(Nerophis	ophidion:	Lyons	&	Dunne,	2004;	Hippocampus	patagonicus:	Storero	&	Gonzalez,	2008;	Syngnathoides	biaculeatus:	Horinouchi	et	al.,	2012).			3.3.2.3	Sex	&	reproductive	status			 Among	seahorses,	the	relationship	between	sex	and	reproductive	status	and	feeding	has	been	inconsistent,	with	some	studies	showing	that	females	eat	more	than	males	(Kitsos	et	al.,	2008),	and	others	showing	no	difference	(Woods,	2002;	Felicio	et	al.,	2006;	Storero	&	Gonzalez,	2008;	Gurkan	et	al.,	2011b).	Among	Hippocampus	reidi,	adult	males	changed	their	diets	and	fed	on	smaller	prey	when	they	became	reproductively	active	(Castro	et	al.,	2008).	However,	among	Hippocampus	guttulatus,	neither	sex	nor	reproductive	status	affected		 72	seahorse	diets	of	(D’Entremont,	2002).		The	effect	of	sex	and	reproductive	status	on	feeding	has	also	been	variable	among	pipefishes,	although	a	number	of	studies	have	suggested	females	eat	more	than	males.	Compared	to	males,	female	Syngnathus	fuscus	and	Syngnathus	floridae	ate	more	(Teixeira	&	Musick,	1995),	and	female	Stigmatopora	nigra,	Stigmatopora	argus	(Steffe	et	al.,	1989)	and	Syngnathus	typhle	(Oliveira	et	al.,	2007)	had	fuller	guts.	Similarly,	Nerophis	lumbriciformis	females	consumed	a	greater	quantity	and	diversity	of	prey	than	non-reproductive	males	(Lyons	&	Dunne,	2004).	Female	Syngnathus	folletti	fed	on	a	larger	range	of	prey	sizes	than	males	who	preyed	primarily	on	small	isopods	and	copepods	(Garcia	et	al.,	2005).	The	trend	also	translates	to	ex-situ	studies	on	Syngnathus	typhle,	where	reproductively	active	females	ate	more	and	larger	prey	than	reproductively	active	males	(Svensson,	1988).	In	contrast	to	these	studies,	Berglund	et	al.	(2006)	found	that	Syngnathus	typhle	males	invested	more	time	into	feeding	than	females,	and	Lyons	and	Dunne	(2004)	found	that	among	Nerophis	lumbriciformis	males,	egg-bearing	individuals	ate	more	prey	than	non-reproductive	males.	Diets	of	males	and	female	did	not	differ	for	Syngnathus	typhle,	Stigmatopora	nigra,	Urocampus	carinirostris,	Vanacapus	phillpi	and	Mitotichthys	semistriatus	(Howard	&	Koehn,	1985).		3.3.3	 How	does	feeding	and	diet	vary	across	a	marine	fish	family	that	lives	in	a	three-dimensional	space?		3.3.3.1	Variability			 Syngnathid	diets	have	been	shown	to	vary	seasonally	and	geographically,	largely	as	a	result	of	differences	in	the	abundance	and	composition	of	prey.	For	example,	the	diet	of		 73	an	Indo-Pacific	seahorse	was	comprised	of	37%	and	7%	(by	volume)	of	eucarids	and	peracarids,	respectively,	at	one	site,	and	0%	and	67%	on	the	opposite	side	of	the	narrow	Malaysian	Peninsula	(Hippocampus	trimaculatus;	Yip	et	al.,	2015).	A	similar	effect	was	found	among	Hippocampus	patagonicus	diets	in	Argentina,	which	showed	large	variation	between	two	sites	located	less	than	5	km	from	each	other	(Storero	&	Gonzalez,	2008).	Likewise,	despite	some	studies	that	found	no	seasonal	effect	(Huh	&	Kitting,	1985;	Tipton	&	Bell,	1988;	Teixeira	&	Musick,	1995),	the	majority	of	work	has	shown	that	syngnathid	diets	change	seasonally	(Brown,	1972;	Livingston,	1982;	Ryer	&	Orth,	1987;	Franzoi	et	al.,	1993;	Motta	et	al.,	1995;	Woods,	2002;	Oliveira	et	al.,	2007;	Taskavak	et	al.,	2010;	Gurkan	et	al.,	2011a;	Horinouchi	et	al.,	2012).	 			3.3.3.2	Tail	morphology	&	foraging	strategies			 Across	genera,	locations,	and	studies,	there	was	no	correlation	between	the	relative	size	of	syngnathid	caudal	fins	and	the	bulk	diets	of	syngnathids,	numeric	diets,	or	the	frequency	that	prey	items	occurred	in	syngnathid	diets	(Table	3.3).		Syngnathid	tail	morphology—including	its	grasping	ability	and	the	relative	size	of	a	potential	caudal	fin—is	thought	to	affect	their	foraging	strategy,	and	in	turn,	what	types	of	prey	they	eat.	Syngnathid	tail	morphology	can	broadly	be	broken	down	into	three	groups	(based	on	even	broader	categories	in	Neutens	et	al.	(2014):	(1)	syngnathids	that	lack	caudal	fins,	but	have	prehensile	tails	that	can	grasp	holdfasts	(including	seahorses),	(2)	pipefishes	that	have	caudal	fins	but	lack	prehensile	abilities	(including	most	pipefish	genera),	and	(3)	seadragons,	which	lack	both	caudal	fins	and	prehensile	abilities.		 74	Syngnathids	belonging	to	group	1,	including	seahorses	and	certain	genera	of	pipefishes	(e.g.	Nerophis	spp.,	Stigmatopora	spp.,	Urocampus	spp.)	rely	on	rapid	dorsal	and	pectoral	fin	oscillations	for	movement	(Consi	et	al.,	2001).	Without	caudal	fin	propulsion,	this	group	of	syngnathids	swims	comparatively	slowly,	but	with	good	maneuverability	(Consi	et	al.,	2001).	Instead,	these	syngnathids	prefer	to	remain	attached	to	a	holdfast,	adopting	a	sit-and-wait	feeding	strategy	(Howard	&	Koehn,	1985;	James	&	Heck,	1994;	Ocken	&	Ritz,	1994;	Kendrick	&	Hyndes,	2005).	In	this	position,	syngnathids	are	difficult	for	prey	to	detect.	Seahorses	are	able	to	adjust	their	colouration	to	match	their	background,	and	some	grow	long	skin	filaments	to	improve	crypsis	(Foster	&	Vincent,	2004).	Likewise,	some	pipefishes	belonging	to	this	group—with	a	greenish-brown	body	colour,	and	a	slow,	rhythmic	body	movement—resemble	eelgrass	leaves	(Howard	&	Koehn,	1985).	While	attached	to	a	holdfast,	these	sedentary	ambush	predators	feed	from	the	water	column	or	from	the	surfaces	of	plants	(Howard	&	Koehn,	1985).	In	contrast,	pipefishes	belonging	to	group	2	are	typically	more	mobile,	active	feeders	(Howard	&	Koehn,	1985;	Kendrick	&	Hyndes,	2005).	Among	these	pipefishes	there	are	genera	that	are	associated	with	the	vegetation,	often	resting	horizontally	within	the	canopy	(e.g.	Mitotichthys	spp.),	and	others	that	lay	near	or	on	the	substrate	(e.g.	Vanacampus	spp.;	Howard	&	Koehn,	1985,	Kendrick	&	Hyndes,	2005).	The	latter	group	are	often	camouflaged	brown	and	resemble	mud	(Howard	&	Koehn,	1985;	Kendrick	&	Hyndes,	2005).	Although	syngnathids	with	caudal	fins	will	occasionally	feed	from	their	sitting	positions,	they	routinely	swim	clear	of	the	vegetation	to	capture	prey	(Howard	&	Koehn,	1985;	Kendrick	&	Hyndes,	2005).	These	pipefish	genera	typically	feed	on	a	wider	range	of	epibenthic	and	planktonic	prey	types	(Howard	&	Koehn,	1985;	Kendrick	&	Hyndes,	2005).	Seadragons	(Group	3;	Phyllopteryx	spp.	and	Phycodurus		 75	sp.)	—which	resemble	drifting	algae	and	lack	both	prehensile	tails	and	caudal	fins—swim	above	the	vegetation	and	feed	on	prey	in	the	open	water	(Kendrick	&	Hyndes,	2005).	 			 Syngnathids	have	been	shown	to	switch	to	a	more	sedentary	strategy	in	areas	of	greater	habitat	complexity,	leading	to	diets	with	fewer	mobile	prey.	Niche	partitioning	may	drive	some	syngnathids	to	use	more	active	feeding	strategies.	For	example,	Hippocampus	hippocampus	was	found	to	swim	and	hunt	prey	in	more	open,	unvegetated	areas	than	its	congener,	Hippocampus	guttulatus,	which	conformed	to	the	traditional	seahorse	sit-and-wait	strategy	in	more	vegetated	areas	(Curtis	&	Vincent,	2005).	Additionally,	some	have	found	that	higher	habitat	complexity	promoted	seahorses	to	adopt	a	more	sedentary	feeding	strategy,	presumably	because	of	greater	ambush	opportunities	(James	&	Heck,	1994;	Curtis	&	Vincent,	2005;	Felicio	et	al.,	2006).	In	aquaria	with	limited	vegetation,	seahorses	abandoned	the	typical	sit-and-wait	strategy	and	would	instead	search	for	and	consume	prey	while	swimming	(Felicio	et	al.,	2006).	A	similar	change	in	strategy	was	found	with	Syngnathus	fuscus,	which	abandoned	their	post	to	chase	prey	in	low	vegetation	aquaria,	but	rarely	did	so	in	those	with	greater	vegetation	densities	(Ryer,	1988).	Greater	habitat	complexity	may	also	promote	a	switch	from	planktonic	to	epibenthic	prey	types	among	some	syngnathids.	For	example,	with	increased	seagrass	density,	mid-sized	Syngnathus	scovelli	shifted	their	diets	from	more	mobile	to	less	mobile	copepods	(Krejci,	2012),	and	Stigmatopora	nigra	shifted	from	planktonic	copepods	to	benthic	copepods	(Smith	et	al.,	2011a).	Overall,	research	on	the	impacts	of	habitat	complexity	on	syngnathid	feeding	and	success	have	had	mixed	results.	For	example,	some	research	on	seahorses	has	found	increased	capture	success	in	more	complex	habitats	(Flynn	&	Ritz,	1999),	and	others	have	found	no	effect	(James	&	Heck,	1994).	Likewise,	increased	habitat	complexity	was		 76	associated	with	greater	gut	fullness	among	Stigmatopora	argus,	but	not	among	its	congener,	Stigmatopora	nigra	(Steffe	et	al.,	1989;	Smith	et	al.,	2011a).			 In	addition,	syngnathids	may	change	their	foraging	strategy	based	on	their	satiation	level,	switching	to	more	active	foraging	when	hungry	(Ocken	&	Ritz,	1994).	Satiated	seahorses	are	more	likely	to	conserve	energy,	attach	to	a	holdfast	and	attack	easier	prey	whereas	hungrier	seahorses	will	actively	swim	and	become	less	selective	(Ocken	&	Ritz,	1994).	Unattached	seahorses	have	been	shown	to	have	greater	capture	success	than	those	associated	with	a	holdfast	(Ocken	&	Ritz,	1994).	In	this	way,	the	need	for	a	rapid	acquisition	of	energy	makes	active	swimming	more	important	than	the	higher	costs	associated	with	it	(Ocken	&	Ritz,	1994).		3.4	 Discussion		 While	all	syngnathids	feed	predominantly	on	combinations	of	epibenthic	and	planktonic	crustaceans,	we	deduce	that	the	relative	amounts	and	specific	types	of	prey	are	dictated	by	a	combination	of	prey	availability	and	syngnathid	morphology.	First,	research	on	syngnathid	diets	has	shown	that	studies	in	different	areas	and	season	often	yield	very	different	results—likely	because	of	prey	availability.	We	found	large	unexplained	diet	variation,	perhaps	because	of	vast	differences	in	prey	availability	across	the	global	locations	and	many	seasons	included	in	our	analyses.	Second,	we	were	surprised	to	find	that	no	metric	of	body	size	was	related	to	what	they	ate—despite	that	the	literature	repeatedly	showing	that	larger	syngnathids	ate	larger	prey.	Instead,	we	showed	that	across	genera,	locations,	and	studies,	the	relative	snout	morphologies	of	syngnathids	correlated		 77	with	their	diet.	These	snout	morphologies	may	also	explain	the	relatively	generalist	feeding	strategies	among	certain	species,	and	may	better	explain	the	capture	of	prey	with	differing	mobility	than	their	own	swimming	abilities.	Our	phylogenetic	results	show	that	among	syngnathids,	more	closely	related	species	are	more	similar	in	feeding	morphologies	than	less-related	species	(Blomberg	et	al.,	2003;	Revell	et	al.,	2008).	Differences	in	diet	observed	among	different	genera	probably	arises,	therefore,	because	of	differences	in	important	feeding	morphologies.		Our	results	show	considerable	variation	in	syngnathid	diets	suggesting	that	adult	syngnathids	are	relatively	generalist	feeders	and	that	their	diet	depends	largely	on	prey	availability.	In	all	three	dietary	RDA	models	(bulk,	numeric,	FO)	the	majority	of	the	variation	was	neither	explained	by	genus,	nor	morphological	characteristics.	Our	need	to	standardize	the	size	of	syngnathids	in	our	across-genera	comparisons—to	maximum	adult	sizes—may	have	obscured	some	variation	in	syngnathid	diets.	Adults	of	many	syngnathids	have	relatively	generalist	diets	compared	to	juveniles	(Ryer	&	Orth,	1987;	Gaughan	&	Potter,	1997;	Flynn	&	Ritz,	1999;	Oliveira	et	al.,	2007;	Castro	et	al.,	2008;	Taskavak	et	al.,	2010).	Adults	are	physically	able	to	feed	on	a	wider	range	of	prey	because	they	can	find,	capture,	and	consume	more	types	of	prey	(Flynn	&	Ritz,	1999),	they	extract	energy	and	nutrients	from	their	prey	more	efficiently	than	juveniles	(Brown	&	Maurer,	1989),	and	their	feeding	morphologies	(e.g.	mouth)	are	larger,	and	less	size	restrictive	(Scharf	et	al.,	2000).	Many	studies	have	suggested	that	differing	prey	availability	among	locations	and	seasons,	in	combination	with	this	generalist	ability	of	adult	syngnathids,	is	responsible	for	the	large	diet	differences	recorded	among	locations	and	seasons	(e.g.	Storero	&	Gonzalez,	2008;	Yip	et	al.,	2015).	It	is	therefore	quite	likely	that	much	of	our	unexplained	diet		 78	variation	was	due	to	differences	in	prey	availability	among	the	various	study	locations	included	in	our	analyses,	as	is	the	case	with	other	fishes	(Griffiths,	1973,	1975).	To	date,	no	syngnathid	study	has	explicitly	compared	diet	to	the	prey	that	are	available	to	the	fish.	It	could	be	that	adults	are	actually	selecting	for	certain	prey	when	their	distributions	overlap,	but	its	effect	may	be	masked	by	other	samples	in	areas	where	the	overlap	does	not	occur	(Griffiths,	1973,	1975).		It	was	surprising	that	we	were	not	able	to	detect	a	correlation	between	syngnathid	body	size	and	diet	across	genera,	locations,	and	studies.	The	ecological	principle	that	larger	predators	feed	on	larger	prey	has	been	shown	before	within	syngnathid	species	(Livingston,	1982,	1984;	Ryer	&	Orth,	1987;	Tipton	&	Bell,	1988;	Franzoi	et	al.,	1993;	Truong	&	Nga,	1995;	Teixeira	&	Musick,	1995,	2001;	Kanou	&	Kohno,	2001;	Woods,	2005;	Castro	et	al.,	2008;	Gurkan	et	al.,	2011a;	2011b),	and	among	various	clades	of	marine	fishes	(Scharf	et	al.,	2000;	Costa,	2009;	Barnes	et	al.,	2010).	It	is	generally	understood	that	this	may	be	the	result	of	two	factors.	First,	larger	predators	have	larger	feeding	morphologies	(e.g.	mouths),	and	are	therefore	better	suited,	physically,	to	handling	larger	prey	(Scharf	et	al.,	2000).	Second,	larger	predators	have	greater	energy	demands	which	are	more	easily	met	if	they	optimize	their	foraging,	and	feed	on	larger,	more	energetically	beneficial	prey	(Costa,	2009).	Therefore,	we	expect	we	would	have	found	correlations	between	syngnathid	diets	and	either	their	body	size	or	non-relative	values	of	feeding	morphologies	including	snout	length	or	snout	depths.	Instead,	however,	we	found	correlations	between	their	diets	and	relative	snout	lengths	and	depths.	Our	results,	which	complement	other	studies,	explicitly	show—for	the	first	time—	 79	that	differences	in	syngnathid	diets	across	genera,	locations,	and	studies	can	largely	be	attributed	to	the	relative	shape	and	sizes	of	their	snouts.	In	general,	pipefishes	have	relatively	longer	snouts	than	seahorses	(Table	C1).	Our	results	expand	on	those	of	Kendrick	&	Hyndes	(2005),	as	we	show	here	that	relative	snout	length	has	an	effect	on	the	diets	of	syngnathids	across	genera,	locations,	and	studies.	The	relatively	long	snouts	typical	of	pipefishes	are	often	used	to	capture	more	mobile	prey,	such	as	copepods	and	mysids.	To	make	up	for	increased	drag,	longer	snouts	are	usually	quite	narrow,	too	(Table	C1;	de	Lussanet	&	Muller	2007;	Leysen	et	al.,	2011).	Syngnathids	with	long,	narrow	snouts	have	less	angle	to	cover	during	head	rotation	and	less	drag	to	overcome,	and	some	have	hypothesized	that	longer,	narrower	snouts	would	result	in	the	fastest	prey-capture	times	(de	Lussanet	&	Muller,	2007;	Van	Wassenbergh	et	al.,	2011a).	While	this	snout	shape	allows	for	speed,	it	comes	at	the	expense	of	reduced	gape	size	(Table	C1;	de	Lussanet	&	Muller	2007;	Leysen	et	al.,	2011).	This	is	important,	and	our	study	was	the	first	to	show	that	it	is	relative	gape	size	that	matters	to	syngnathid	diets.	We	found	that	syngnathids	with	relatively	small	gapes	had	limited	diets,	and	were	presumably	forced	to	feed	on	very	small	copepods.	Stigmatopora	spp.	—a	genus	of	pipefishes	with	the	longest	snouts	and	smallest	gapes	(relative	measurements;	Table	C1)	of	all	syngnathids	in	this	study—had	diets	that	were	equally	extreme,	composed	almost	entirely	of	copepods	(in	bulk	and	FO).	This	genus	may	represent	one	end	of	the	spectrum	in	syngnathid	snout	morphology,	in	fact,	as	this	pattern	was	consistent	among	three	studies	(Steffe	et	al.,	1989;	Kendrick	&	Hyndes,	2005;	Smith	et	al.,	2011a).	In	contrast	to	most	pipefishes	are	seahorses,	having	notably	short	and	broad	snouts.	Relatively	broad	snouts	allow	syngnathids	to	fit	more	types	of	prey	in	their	gape,	so	it	is	not	surprising	that	most	prey	types	ended	up	in	the	stomachs	of		 80	species	with	relatively	larger	gapes	more	frequently	(FO)	than	those	with	smaller	gapes.	While	broad	snouts	increased	the	breadth	of	prey	diversity	they	were	able	to	eat,	the	bulk	of	their	diets	was	primarily	taken	up	by	larger,	slower	prey	like	amphipods.		Our	findings	that	the	relative	amounts	of	mobile	and	sedentary	prey	in	the	diets	of	syngnathids	is	influenced	more	by	snout	morphology	than	syngnathid	mobility	are	at	odds	with	studies	of	other	fish	predators.	It	has	been	argued	that	mobile	predators	feed	on	sedentary	prey	more	frequently	than	sedentary	predators	do	because	they	encounter	these	prey	types	more	often	(Cooper	et	al.,	1985).	Such	a	pattern	had	been	suggested	for	syngnathids,	too,	as	a	reason	for	the	difference	in	diets	between	syngnathids	that	are	free-swimming	versus	those	attached	to	a	holdfast	(Howard	&	Koehn,	1985).	Syngnathid	foraging	mobility	has	been	partially	explained	by	tail	morphology,	which	ranges	from	the	grasping,	prehensile	tails	of	sedentary	seahorses	to	the	massive	caudal	fins	of	mobile	flagtail	pipefishes	(Neutens	et	al.,	2014).	In	contrast	to	this	hypothesis,	however,	studies	on	pipefishes	suggest	that	less	active	syngnathids	may	instead	feed	on	less	active	prey	(Smith	et	al.,	2011b;	Krejci,	2012).	It	is	possible	that	there	is	a	two-fold	reason	for	greater	numbers	of	less-mobile	prey	in	the	diets	of	syngnathids	feeding	in	more	complex	habitats.	First,	the	increasing	seagrass	density	promotes	a	sit	and	wait	strategy	where	syngnathids	feed	from	among	the	surfaces	of	seagrass	blades,	where	less	mobile	prey	typically	reside	(James	&	Heck,	1994;	Curtis	&	Vincent,	2005;	Felicio	et	al.,	2006).	Second,	densities	of	less	mobile	prey	increase	with	greater	seagrass	density	(Stoner,	1980;	Bell	&	Westoby,	1986;	Jenkins	et	al.,	2002).	Therefore,	if	seagrass	density	is	great	enough,	sedentary	syngnathids	may	actually	encounter	more	sedentary	prey	than	mobile	syngnathids,	despite	the	latter’s	active	search	behaviours.	In	addition,	the	propensity	of	seahorses	to	feed	on	relatively	sedentary		 81	prey—while	being	a	sedentary	predator	themselves—seems	to	suggest	an	alternative	explanation.	Perhaps	general	trends	in	the	mobility	of	syngnathid	prey	are	more	a	consequence	of	their	snout	morphology	than	their	mobility.	In	support	of	this	view,	we	were	not	able	to	detect	an	association	between	diet	and	relative	caudal	fin	size,	a	proxy	of	mobility,	in	this	this	study.	Overall,	our	results	show	that	syngnathid	diet	was	better	explained	by	body	characteristics	than	by	genus	alone.	We	showed	that	seahorse	and	pipefish	diets	can	overlap	considerably	because	of	overlaps	in	feeding-dependent	morphologies.	For	example,	Pugnaso	curtirostris	is	a	pipefish	with	a	relatively	short	and	narrow	snout	characteristic	similar	to	seahorses.	And,	like	seahorses,	this	pipefish	feeds	primarily	on	amphipods	and	decapods.	The	opposite	trend	can	also	be	true;	despite	feeding	while	attached	to	a	holdfast,	like	seahorses,	Stigmatopora	spp.	have	very	little	diet	overlap	with	most	seahorses.	Instead,	they	have	very	long	and	narrow	snouts	which	are	likely	responsible	for	their	copepod-dominated	diets.	Syngnathid	diet	may	be	a	function	of	the	species’	body	characteristics,	with	any	differences	among	genera	arising	because	of	differences	in	those	morphologies	among	genera.	Our	phylogenetic	results	support	this,	demonstrating	that	among	syngnathids,	more	closely	related	species	are	more	similar	in	feeding	morphologies	than	are	less-related	species	(Blomberg	et	al.,	2003;	Revell	et	al.,	2008).	It	is	therefore	reasonable	that	we	see	differences	in	diet	among	different	genera,	as	they	are	statistically	less	similar	to	each	other	for	important	feeding	morphologies	than	species	within	a	genus.	Our	study	shows	that	meta-analyses	can	cut	across	inevitable	variability	across	genera,	locations,	and	methods	to	identify	new	broad	taxonomic	patterns	and	to	expand		 82	previous	generalities	to	a	broader	scale.	Review	studies	also	serve	to	expose	gaps	in	the	literature	and	prompt	future	studies.	For	example,	to	fully	understand	prey	selectivity	among	syngnathids,	we	suggest	the	need	for	a	quantitative	comparison	between	their	prey	consumption	and	the	availability	of	these	prey.	The	results	of	our	study	suggest	that	across	genera,	locations,	and	studies,	syngnathids	forage	in	a	way	that	has	the	greatest	overall	energy	benefit,	but	is	constrained	by	prey	availability	and	their	feeding	morphologies.	Syngnathids	which	live	in	denser	habitats	more	often	adopt	a	sedentary	foraging	strategy,	presumably	because	of	greater	ambush	opportunities	(James	&	Heck,	1994;	Curtis	&	Vincent,	2005;	Felicio	et	al.,	2006).	Since	all	syngnathids	must	eat	constantly,	on	account	of	their	rudimentary	digestive	tract—and	sedentary	syngnathids	must	feed	on	what	is	around	them—it	makes	energetic	sense	for	more	sedentary	types	to	be	less	picky	(Dunham,	2010).	More	sedentary	syngnathids	may	have	therefore	evolved	relatively	larger	gapes	that	allow	them	to	feed	on	a	larger	suite	of	prey,	including	larger,	more	energetically	beneficial	types,	as	we	found	in	this	study.	Syngnathids	with	relatively	larger	gapes	also	have	relatively	shorter	snouts	(de	Lussanet	&	Muller,	2007;	Leysen	et	al.,	2011).	This	may	be	an	evolutionary	adaptation	to	the	greater	densities	of	sedentary	prey	types	found	in	highly	complex	habitats	(Stoner,	1980;	Bell	&	Westoby,	1986;	Jenkins	et	al.,	2002),	or	the	result	of	a	morphological	trade-off	with	snout	length	(de	Lussanet	&	Muller,	2007;	Leysen	et	al.,	2011).	Either	way,	shorter,	broader	snouts	are	likely	beneficial	to	more	sedentary	syngnathids	which	benefit	from	feeding	on	a	wider	range	of	prey—including	the	large,	sedentary	types	most	often	found	in	their	complex	habitats.	In	the	same	way,	longer,	narrower	snouts	are	more	beneficial	to	those	feeding	on	smaller,	faster	prey	that	are	common	in	the	water	column.	This	may	be	why	we	often	see	more	mobile	pipefish	with		 83	these	types	of	snouts.	Unfortunately,	very	few	studies	have	compared	the	home	range	or	mobility	of	species	with	their	dietary	breadth.	In	one	study,	however,	more	specialized	monkeys	were	found	to	have	larger	home	ranges,	purportedly	because	they	needed	to	travel	further	to	find	their	preferred	food	types	(Clutton-Brock,	1975).	Our	understanding	of	syngnathid	prey	choices	and	their	ability	to	meet	their	energy	demands	would	benefit	from	a	study	that	looked	at	the	energy	benefits	of	different	prey	types,	and	the	energy	costs	of	different	foraging	strategy.	 84	Table	3.1	Relative	importance	of	syngnathid	diets.	a	The	approximate	area	of	sampling.	For	comparative	purposes,	studies	in	close	proximity	(within	50	km	of	each	other)	have	the	same	location.	b	Bulk	dietary	studies	include	relative	values	that	each	food	item	contributes	to	the	total	volume	(%V),	weight	(%W),	or	area	(%A)	of	dietary	contents	collected,	numeric	dietary	studies	include	relative	values	that	each	food	item	contributes	to	the	total	number	of	food	items	(%N)	collected,	and	frequency	of	occurrence	(%FO)	studies	include	the	relative	number	of	stomach	samples	that	a	particular	food	item	occurs	in.	Each	row	represents	a	particular	species	in	a	particular	area	of	a	particular	study.	c	This	table	includes	additive	data,	so	if	a	taxon	is	centered	above	other	taxa	it	includes	those	numbers	in	its	total.	Numbers	are	added	to	columns	on	the	left	(e.g.	Crustacea	includes	Paracarida	and	Eucarida).	Sample	sizes	of	two	and	under	were	not	considered	in	statistical	analyses.	Blank	cells	represent	missing	data.	Shaded	cells	(for	%FO	data	only)	indicate	the	value	of	that	cell	was	not	provided	in	the	literature,	and	is	a	minimum	value	based	on	dietary	items	that	were	included	at	a	lower	taxonomic	resolution	(see	text).	Abbreviated	locations;	PoF	=	Port	of	Frematle.																					 85			 		 		 		 Taxon	 		 		 		 		 		 		 		 		 		 		 		 		 			 	 	 	 Subphylum	 Crustacea	 		 	 	 	 Subclass	 		 	 	 	 	 	 	 	 Copepoda	 		 	 	 	 Superorder	 		 Peracarida	 Eucarida	 		 	 	 	 		 	 	 	 Order	 		 		Amphipoda	 Mysida	 Isopoda	 Tanaidacea	 		 Decapoda	 		 Calanoida	 Cyclopoida	 Harpacticoida	 	Genus	 Species	 Location	of	Study	 N	 Diet	Metric	 		 		 		 		 		 		 		 		 		 		 		 		 Reference	Filicampus	 tigris	PoF,	Australia	 10	 %V	 ++++	 ++++	 ++++	 +	 -	 -	 +	 +	 +	 -	 -	 +	 Kendrick	&	Hydnes	(2005)	Hippichthys	 cyanospilus	Pulau	Tinggi,	Malaysia	 25	 %A	 ++	 +	 +	 +	 -	 -	 +	 +	 +	 	 	 	Lim,	unpublished	data	(2015)	Hippichthys	 cyanospilus	Trang,	Thailand	 16	 %V	 ++++	 ++	 ++	 -	 -	 -	 ++	 ++	 +	 	 	 	 Horinouchi	et	al.	(2012)	Hippocampus	 abdominalis	Wellington	Harbour,	NZ	 89	 %V	 +++	 ++	 +	 +	 +	 +	 ++	 ++	 +	 -	 +	 +	 Woods	(2002)	Hippocampus	 breviceps	PoF,	Australia	 67	 %V	 ++++	 ++++	 +++	 +	 +	 +	 -	 -	 +	 +	 +	 +	 Kendrick	&	Hydnes	(2005)	Hippocampus	 coronatus	Tokyo,	Japan	 60	 %V	 ++++	 ++++	 +++	 +	 -	 -	 -	 -	 +	 	 	 -	Horinouchi	&	Sano	(2000)	Hippocampus	 erectus	Tampa	Bay,	FL	 NA	 %V	 ++++	 +++	 +++	 -	 +	 -	 +	 +	 +	 	 	 	 Dunham	(2010)	Hippocampus	 guttulatus	Aegean	Sea,	Turkey	 16	 %W	 ++	 ++	 +	 ++	 +	 -	 +	 -	 +	 +	 -	 -	 Gurkan	et	al.	(2011b)	Hippocampus	 hippocampus	Aegean	Sea,	Turkey	 21	 %W	 +	 +	 +	 +	 -	 -	 -	 -	 +	 -	 -	 +	 Gurkan	et	al.	(2011b)	Hippocampus	 reidi	Mamanguape	estuary,	Brazil	 280	 %A	 +++	 +	 +	 -	 +	 -	 ++	 ++	 ++	 +	 +	 +	 Castro	et	al.	(2008)	Hippocampus	 spiniosissimus	 East	 29	 %V	 ++	 +	 +	 +	 -	 -	 ++	 ++	 +	 -	 -	 +	 Yip	et	al.		 86	Coast,	Malaysia	 (2015)	Hippocampus	 spiniosissimus	West	Coast,	Malaysia	 4	 %V	 +++	 +++	 ++	 ++	 -	 -	 -	 -	 +	 	 	 -	 Yip	et	al.	(2015)	Hippocampus	 subelongatus	PoF,	Australia	 22	 %V	 ++++	 +++	 ++	 +	 -	 -	 ++	 ++	 -	 -	 -	 -	 Kendrick	&	Hydnes	(2005)	Hippocampus	 trimaculatus	East	Coast,	Malaysia	 36	 %V	 +++	 +	 +	 +	 -	 -	 ++	 ++	 -	 -	 -	 -	 Yip	et	al.	(2015)	Hippocampus	 trimaculatus	West	Coast,	Malaysia	 16	 %V	 ++++	 +	 +	 -	 -	 -	 +	 +	 +++	 	 	 +	 Yip	et	al.	(2015)	Hippocampus	 whitei	Sydney,	Australia	 8	 %V	 ++++	 ++++	 +++	 ++	 -	 -	 +	 +	 +	 	 	 	 Burchmore	et	al.	(1984)	Hippocampus	 zosterae	Tampa	Bay,	FL	 87	 %W	 ++++	 +	 +	 -	 -	 -	 +	 +	 ++++	 +	 +	 ++++	 Tipton	&	Bell	(1988)	Histiogamphelus	 cristatus	PoF,	Australia	 58	 %V	 ++++	 ++++	 +++	 +	 -	 -	 +	 +	 +	 +	 +	 +	 Kendrick	&	Hydnes	(2005)	Lissocampus	 caudalis	PoF,	Australia	 8	 %V	 ++++	 ++	 ++	 -	 -	 -	 -	 -	 +++	 +	 -	 ++	 Kendrick	&	Hydnes	(2005)	Mitotichthys	 meraculus	PoF,	Australia	 16	 %V	 ++++	 ++++	 +	 ++++	 -	 -	 -	 -	 -	 -	 -	 -	 Kendrick	&	Hydnes	(2005)	Mitotichthys	 semistriatus	Melbourne,	Australia	 8	 %V	 ++++	 ++	 ++	 -	 -	 -	 -	 -	 +++	 ++	 -	 -	 Howard	&	Koehn	(1985)	Nerophis	 ophidion	Aegean	Sea,	Turkey	 43	 %W	 ++++	 ++	 ++	 -	 -	 -	 -	 -	 ++	 -	 +	 +	 Gurkan	et	al.	(2011a)	Phyllopteryx	 taeniolatus	PoF,	Australia	 29	 %V	 ++++	 ++++	 +	 ++++	 -	 -	 +	 +	 +	 +	 -	 -	 Kendrick	&	Hydnes	(2005)	Pugnaso	 curtirostris	PoF,	Australia	 39	 %V	 ++++	 +++	 +++	 +	 +	 +	 +	 +	 ++	 +	 +	 ++	 Kendrick	&	Hydnes	(2005)	Stigmatopora	 argus	PoF,	Australia	 165	 %V	 ++++	 +	 +	 -	 -	 -	 +	 +	 ++++	 +++	 +	 +	 Kendrick	&	Hydnes	(2005)	Stigmatopora	 argus	Sydney,	Australia	 40	 %V	 ++++	 +	 +	 +	 +	 -	 -	 -	 ++++	 	 	 	 Steffe	et	al.	(1989)	Stigmatopora	 argus	 Sydney,	 1	 %V	 +++	 -	 -	 -	 -	 -	 -	 -	 +	 	 	 	 Burchmore		 87	Australia	 et	al.	(1984)	Stigmatopora	 nigra	Melbourne,	Australia	 350	 %W	 ++++	 +	 +	 -	 +	 +	 -	 -	 ++++	 	 	 +	 Smith	et	al.	(2011a)	Stigmatopora	 nigra	PoF,	Australia	 144	 %V	 ++++	 +	 +	 +	 -	 -	 +	 +	 ++++	 +++	 +	 +	 Kendrick	&	Hydnes	(2005)	Stigmatopora	 nigra	Sydney,	Australia	 40	 %V	 ++++	 +	 +	 +	 -	 -	 -	 -	 ++++	 	 	 	 Steffe	et	al.	(1989)	Syngnathiodes	 biaculeatus	Trang,	Thailand	 88	 %V	 +++	 +	 +	 +	 -	 -	 ++	 ++	 -	 -	 -	 -	 Horinouchi	et	al.	(2012)	Syngnathoides	 biaculeatus	Amitori	Bay,	Japan	 15	 %V	 ++++	 +	 -	 +	 -	 -	 +++	 +++	 +	 	 	 -	 Nakamura	et	al.	(2003)	Syngnathus	 acus	Bot	estuary,	SA	 77	 %W	 ++++	 ++	 +	 -	 +	 +	 -	 -	 +++	 ++	 	 +	Bennett	&	Branch	(1990)	Syngnathus	 floridae	Chesapeake	Bay,	USA	 1295	 %W	 ++++	 ++	 +	 +	 +	 -	 ++	 ++	 +	 	 	 	Teixeira	&	Musick	(1995)	Syngnathus	 folletti	Patos	lagoon,	Brazil	 108	 %A	 ++++	 +++	 +	 -	 +++	 -	 -	 -	 +	 	 	 	 Garcia	et	al.	(2005)	Syngnathus	 fuscus	Chesapeake	Bay,	USA	 3311	 %W	 ++++	 ++++	 ++++	 +	 +	 -	 +	 +	 +	 	 	 	Teixeira	&	Musick	(1995)	Syngnathus	 fuscus	Chesapeake	Bay,	USA	 136	 %W	 ++++	 +++	 ++	 +	 ++	 -	 -	 -	 +	 +	 -	 -	 Ryer	&	Orth	(1987)	Syngnathus	 fuscus	Cape	Hatteras,	NC	 13	 %W	 ++++	 ++++	 +	 ++++	 -	 -	 -	 -	 -	 -	 -	 -	 Bowman	et	al.	(2000)	Syngnathus	 schlegali	Kwangyang	Bay,	Korea	 1347	 %W	 ++++	 +++	 +++	 +	 -	 +	 +	 +	 ++	 ++	 -	 -	 Huh	&	Kwak,	(1997)	Syngnathus	 schlegali	Ishikari,	Japan	 51	 %W	 ++++	 ++++	 ++	 ++	 +	 -	 -	 -	 +	 +	 +	 +	 Sakurai	et	al.	(2009)	Syngnathus	 schlegali	Tokyo,	Japan	 41	 %V	 ++++	 +	 +	 +	 -	 -	 -	 -	 +++	 	 	 -	Horinouchi	&	Sano	(2000)	Syngnathus	 scovelli	Tampa	Bay,	FL	 178	 %W	 +++	 +	 +	 -	 -	 -	 +	 +	 ++	 +	 +	 ++	 Tipton	&	Bell	(1988)	Syngnathus	 scovelli	Indian	River	 101	 %V	 +++	 +	 +	 -	 -	 -	 +	 +	 +	 +	 -	 +	 Krejci	(2012)		 88	Lagoon,	FL	Syngnathus	 scovelli	Tampa	Bay,	FL	 30	 %W	 ++++	 ++	 ++	 -	 +	 -	 ++	 ++	 +	 	 	 	 Motta	et	al.	(1995)	Syngnathus	 typhle	Ria	Formosa,	Portugal	 411	 %W	 +++	 +	 +	 +	 +	 -	 +++	 +++	 +	 	 	 	 Oliveira	et	al.	(2007)	Urocampus	 carinirostris	Melbourne,	Australia	 17	 %V	 ++++	 +	 +	 -	 -	 -	 -	 -	 ++++	 +	 +++	 +	 Howard	&	Koehn	(1985)	Vanacampus	 phillipi	PoF,	Australia	 26	 %V	 ++++	 +++	 ++	 +	 +	 +	 +	 +	 +	 +	 +	 +	 Kendrick	&	Hydnes	(2005)	Vanacampus	 phillipi	Melbourne,	Australia	 13	 %V	 +++	 ++	 ++	 -	 +	 -	 -	 -	 ++	 ++	 -	 +	 Howard	&	Koehn	(1985)	Vanacampus	 poecilolaemus	PoF,	Australia	 67	 %V	 ++++	 +++	 +	 +++	 +	 -	 +	 +	 +	 +	 -	 +	 Kendrick	&	Hydnes	(2005)		 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	Anarchopterus	 criniger	Cedar	Key,	FL	 27	 %N	 ++++	 ++	 +	 +	 +	 -	 +	 +	 +++	 	 	 	 Brown	(1972)	Hippocampus	 erectus	Chesapeake	Bay,	USA	 136	 %N	 ++++	 ++++	 ++++	 -	 +	 -	 +	 +	 +	 	 	 	Teixeira	&	Musick	(2001)	Hippocampus	 guttulatus	Rhodes	Island,	Greece	 279	 %N	 ++++	 +++	 ++	 +	 +	 +	 ++	 ++	 +	 	 	 	 Kitsos	et	al.	(2008)	Hippocampus	 guttulatus	Aegean	Sea,	Turkey	 16	 %N	 +++	 +++	 +	 ++	 +	 -	 +	 -	 +	 +	 -	 -	 Gurkan	et	al.	(2011b)	Hippocampus	 hippocampus	Aegean	Sea,	Turkey	 21	 %N	 +++	 +++	 +	 ++	 -	 -	 -	 -	 +	 -	 -	 +	 Gurkan	et	al.	(2011b)	Hippocampus	 hippocampus	Rhodes	Island,	Greece	 19	 %N	 ++++	 +++	 ++	 +	 -	 -	 +	 +	 -	 	 	 	 Kitsos	et	al.	(2008)	Hippocampus	 mohnikei	Tokyo,	Japan	 64	 %N	 ++++	 -	 -	 -	 -	 -	 -	 -	 ++++	 +	 +++	 -	 Kanou	&	Kohno	(2001)	Hippocampus	 patagonicus	San	Antonio	Bay,	Argentin 32	 %N	 ++++	 ++	 ++	 -	 -	 -	 +++	 +++	 -	 -	 -	 -	 Storero	&	Gonzalez	(2008)		 89	a	Hippocampus	 zosterae	Tampa	Bay,	FL	 87	 %N	 ++++	 +	 +	 -	 -	 -	 +	 +	 ++++	 +	 +	 +++	 Tipton	&	Bell	(1988)	Nerophis	lumbriciformis	Galway	Bay,	Ireland	 220	 %N	 ++++	 +	 +	 -	 +	 -	 -	 -	 ++++	 +	 +	 +++	 Lyons	&	Dunne	(2004)	Syngnathus	 abaster	Sacca	di	Scardovari,	Italy	 180	 %N	 ++++	 +	 +	 +	 +	 -	 +	 +	 ++++	 +	 +	 ++++	 Franzoi	et	al.	(1993)	Syngnathus	 abaster	Stagnone	di	Marsala	lagoon,	Italy	 87	 %N	 ++++	 ++	 +	 +	 +	 +	 -	 -	 +++	 -	 -	 +++	 Campolmi	et	al.	(1996)	Syngnathus	 acus	Aegean	Sea,	Turkey	 95	 %N	 ++++	 +	 +	 +	 +	 -	 +	 +	 ++	 +	 +	 ++	 Taskavak	et	al.	(2010)	Syngnathus	 acus	Kromme	estuary,	SA	 2	 %N	 ++++	 -	 -	 -	 -	 -	 ++++	 ++++	 -	 -	 -	 -	 Hanekom	&	Baird,	(1984)	Syngnathus	 floridae	Chesapeake	Bay,	USA	 1295	 %N	 ++++	 ++++	 ++	 ++	 +	 -	 -	 -	 +	 +	 -	 -	 Mercer	(1973)	Syngnathus	 floridae	Chesapeake	Bay,	USA	 1295	 %N	 ++++	 +	 +	 +	 +	 -	 +	 +	 +++	 	 	 	Teixeira	&	Musick	(1995)	Syngnathus	 floridae	Ecofina	River,	FL	 875	 %N	 ++++	 +	 +	 +	 -	 -	 ++	 ++	 +	 -	 -	 -	 Livingston	(1984)	Syngnathus	 floridae	Cedar	Key,	FL	 140	 %N	 ++++	 +	 +	 +	 +	 -	 +	 +	 +	 	 	 	 Brown	(1972)	Syngnathus	 folletti	Patos	lagoon,	Brazil	 108	 %N	 ++++	 +++	 +	 -	 +++	 +	 +	 +	 ++	 	 	 	 Garcia	et	al.	(2005)	Syngnathus	 fuscus	Chesapeake	Bay,	USA	 3311	 %N	 ++++	 ++	 ++	 +	 +	 -	 +	 +	 +++	 	 	 	Teixeira	&	Musick	(1995)	Syngnathus	 fuscus	Chesapeake	Bay,	USA	 1295	 %N	 ++++	 ++++	 +++	 +	 +	 -	 -	 -	 +	 +	 -	 -	 Mercer	(1973)	Syngnathus	 louisianae	Cedar	Key,	FL	 18	 %N	 ++++	 +	 +	 +	 +	 -	 +	 +	 +	 	 	 	 Brown	(1972)	Syngnathus	 schlegali	Kwangyang	Bay,	Korea	 1347	 %N	 ++++	 +++	 ++	 +	 -	 +	 +	 +	 ++	 ++	 -	 -	 Huh	&	Kwak	(1997)		 90	Syngnathus	 scovelli	Ecofina	River,	FL	 291	 %N	 ++++	 ++	 ++	 +	 +	 -	 +	 +	 +	 -	 -	 +	 Livingston	(1984)	Syngnathus	 scovelli	Tampa	Bay,	FL	 178	 %N	 ++++	 +	 +	 -	 -	 -	 +	 +	 ++++	 +	 +	 ++	 Tipton	&	Bell	(1988)	Syngnathus	 scovelli	Cedar	Key,	FL	 150	 %N	 ++++	 +	 +	 +	 +	 -	 +	 +	 ++	 	 	 	 Brown	(1972)	Syngnathus	 scovelli	Indian	River	Lagoon,	FL	 101	 %N	 ++++	 +	 +	 -	 -	 -	 +	 +	 +++	 +	 -	 +++	 Krejci	(2012)	Syngnathus	 scovelli	Tampa	Bay,	FL	 30	 %N	 +++	 ++	 ++	 -	 +	 -	 +	 +	 +	 	 	 	 Motta	et	al.	(1995)	Syngnathus	 taenionotus	Sacca	di	Scardovari,	Italy	 137	 %N	 +++	 +	 +	 +	 +	 -	 +	 +	 +++	 ++	 +	 ++	 Franzoi	et	al.	(1993)	Syngnathus	 typhle	Ria	Formosa,	Portugal	 411	 %N	 ++++	 +	 +	 +	 +	 -	 +	 +	 ++++	 	 	 	 Oliveira	et	al.	(2007)	Syngnathus	 typhle	Aegean	Sea,	Turkey	 95	 %N	 ++	 -	 -	 -	 -	 -	 +	 +	 ++	 +	 -	 ++	 Uncumusaoglu	et	al.	(2017)	Syngnathus	 typhle	Stagnone	di	Marsala	lagoon,	Italy	 94	 %N	 ++++	 ++++	 +	 ++++	 -	 -	 -	 -	 -	 -	 -	 -	 Campolmi	et	al.	(1996)	Urocampus	 carinirostris	Wilson	Inlet,	Australia	 268	 %N	 +++	 -	 -	 -	 -	 -	 -	 -	 +++	 +	 ++	 -	 Gaughan	&	Potter	(1997)		 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	 	Filicampus	 tigris	PoF,	Australia	 10	 %FO	 ++++	 ++++	 ++++	 +++	 -	 -	 +	 +	 +	 -	 -	 +	 Kendrick	&	Hydnes	(2005)	Hippichthys	 cyanospilus	Pulau	Tinggi,	Malaysia	 25	 %FO	 ++	 ++	 ++	 +	 -	 -	 +	 +	 ++	 	 	 	Lim,	unpublished	data	(2015)	Hippocampus	 abdominalis	Wellington	Harbour,	NZ	 59	 %FO	 ++++	 ++	 ++	 +	 +	 +	 +++	 +++	 +	 	 +	 +	 Woods	(2002)	Hippocampus	 breviceps	PoF,	Australia	 67	 %FO	 ++++	 ++++	 ++++	 +	 ++	 +	 -	 -	 +++	 +	 +	 +++	 Kendrick	&	Hydnes	(2005)		 91	Hippocampus	 erectus	Chesapeake	Bay,	USA	 136	 %FO	 ++	 ++	 ++	 +	 +	 -	 +	 +	 +	 	 	 	Teixeira	&	Musick	(2001)	Hippocampus	 guttulatus	Rhodes	Island,	Greece	 279	 %FO	 ++++	 ++++	 ++++	 ++	 +	 +	 +++	 +++	 +	 	 	 	 Kitsos	et	al.	(2008)	Hippocampus	 guttulatus	Aegean	Sea,	Turkey	 16	 %FO	 ++++	 ++++	 ++	 ++++	 +	 -	 +	 -	 +	 +	 -	 -	 Gurkan	et	al.	(2011b)	Hippocampus	 hippocampus	Aegean	Sea,	Turkey	 21	 %FO	 +	 +	 +	 +	 -	 -	 -	 -	 +	 -	 -	 +	 Gurkan	et	al.	(2011b)	Hippocampus	 hippocampus	Rhodes	Island,	Greece	 19	 %FO	 +++	 +++	 +++	 ++	 -	 -	 +	 +	 -	 	 	 	 Kitsos	et	al.	(2008)	Hippocampus	 histrix	Khanh	Hoa,	Vietnam	 15	 %FO	 +++	 +++	 +++	 +	 -	 +++	 ++	 ++	 -	 -	 -	 -	 Do	et	al.	(1996)	Hippocampus	 patagonicus	San	Antonio	Bay,	Argentina	 23	 %FO	 ++++	 ++++	 ++++	 -	 -	 -	 +++	 +++	 -	 -	 -	 -	 Storero	&	Gonzalez	(2008)	Hippocampus	 patagonicus	San	Antonio	Bay,	Argentina	 9	 %FO	 ++++	 ++++	 ++++	 -	 -	 -	 -	 -	 -	 -	 -	 -	 Storero	&	Gonzalez	(2008)	Hippocampus	 reidi	Mamanguape	estuary,	Brazil	 280	 %FO	 ++++	 +	 +	 -	 +	 -	 +	 +	 ++++	 ++	 ++	 ++++	 Castro	et	al.	(2008)	Hippocampus	 spiniosissimus	East	Coast,	Malaysia	 29	 %FO	 +	 +	 +	 +	 -	 -	 +	 +	 +	 -	 -	 +	 Yip	et	al.	(2015)	Hippocampus	 spiniosissimus	West	Coast,	Malaysia	 4	 %FO	 +	 +	 +	 +	 -	 -	 -	 -	 +	 	 	 -	 Yip	et	al.	(2015)	Hippocampus	 subelongatus	PoF,	Australia	 22	 %FO	 ++++	 +++	 +++	 ++	 -	 -	 ++	 ++	 +	 -	 -	 +	 Kendrick	&	Hydnes	(2005)	Hippocampus	 trimaculatus	East	Coast,	Malaysia	 36	 %FO	 ++	 +	 +	 +	 -	 -	 ++	 ++	 +	 	 	 -	 Yip	et	al.	(2015)	Hippocampus	 trimaculatus	 Khanh	 19	 %FO	 +++	 +++	 +++	 +	 -	 ++	 +	 +	 -	 -	 -	 -	 Do	et	al.		 92	Hoa,	Vietnam	 (1996)	Hippocampus	 trimaculatus	West	Coast,	Malaysia	 16	 %FO	 +	 +	 +	 -	 -	 -	 +	 +	 +	 	 	 +	 Yip	et	al.	(2015)	Hippocampus	 zosterae	Tampa	Bay,	FL	 87	 %FO	 ++++	 ++	 ++	 -	 -	 -	 +	 +	 ++++	 ++++	 ++	 ++++	 Tipton	&	Bell	(1988)	Histiogamphelus	 cristatus	PoF,	Australia	 58	 %FO	 ++++	 ++++	 ++++	 ++	 -	 -	 +	 +	 +	 +	 +	 +	 Kendrick	&	Hydnes	(2005)	Lissocampus	 caudalis	PoF,	Australia	 8	 %FO	 ++++	 ++++	 ++++	 -	 -	 -	 -	 -	 ++++	 ++++	 -	 ++++	 Kendrick	&	Hydnes	(2005)	Anarchopterus	 criniger	Cedar	Key,	FL	 27	 %FO	 ++++	 +++	 +++	 +	 +	 -	 +	 +	 ++++	 	 	 	 Brown	(1972)	Mitotichthys	 meraculus	PoF,	Australia	 16	 %FO	 ++++	 ++++	 ++	 ++++	 -	 -	 -	 -	 -	 -	 -	 -	 Kendrick	&	Hydnes	(2005)	Phyllopteryx	 taeniolatus	PoF,	Australia	 29	 %FO	 ++++	 ++++	 +	 ++++	 -	 -	 ++	 ++	 +	 +	 -	 +	 Kendrick	&	Hydnes	(2005)	Pugnaso	 curtirostris	PoF,	Australia	 39	 %FO	 ++++	 ++++	 ++++	 +	 +	 +	 +	 +	 ++++	 ++	 +	 ++++	 Kendrick	&	Hydnes	(2005)	Stigmatopora	 argus	PoF,	Australia	 165	 %FO	 ++++	 +	 +	 +	 -	 +	 +	 +	 ++++	 ++++	 ++++	 ++++	 Kendrick	&	Hydnes	(2005)	Stigmatopora	 nigra	Melbourne,	Australia	 350	 %FO	 ++++	 +++	 +++	 -	 +	 +	 -	 -	 ++++	 	 	 ++++	 Smith	et	al.	(2011a)	Stigmatopora	 nigra	PoF,	Australia	 144	 %FO	 ++++	 +	 +	 +	 -	 -	 +	 +	 ++++	 ++++	 ++++	 ++++	 Kendrick	&	Hydnes	(2005)	Syngnathus	 acus	Aegean	Sea,	Turkey	 95	 %FO	 ++++	 ++++	 ++++	 +	 +	 -	 ++	 ++	 ++++	 +	 ++	 ++++	 Tackavak	et	al.	(2010)	Syngnathus	 floridae	Cedar	Key,	FL	 140	 %FO	 ++++	 ++++	 ++	 ++++	 +	 -	 ++++	 ++++	 ++	 	 	 	 Brown	(1972)	Syngnathus	 louisianae	Cedar	Key,	FL	 18	 %FO	 ++++	 ++++	 ++++	 ++++	 +	 -	 ++++	 ++++	 +++	 	 	 	 Brown	(1972)	Syngnathus	 schlegali	Kwangyang	Bay,	Korea	 1347	 %FO	 ++	 ++	 ++	 +	 -	 +	 +	 +	 ++	 ++	 -	 -	 Huh	&	Kwak	(1997)	Syngnathus	 scovelli	Tampa	Bay,	FL	 178	 %FO	 ++++	 +++	 +++	 -	 -	 -	 +	 +	 ++++	 ++++	 +++	 ++++	 Tipton	&	Bell	(1988)		 93	Syngnathus	 scovelli	Cedar	Key,	FL	 150	 %FO	 ++++	 ++++	 ++++	 ++	 +	 -	 ++	 ++	 +++	 	 	 	 Brown	(1972)	Syngnathus	 scovelli	Tampa	Bay,	FL	 30	 %FO	 ++++	 ++++	 ++++	 -	 +	 -	 +++	 +++	 +	 	 	 	 Motta	et	al.	(1995)	Syngnathus	 typhle	Ria	Formosa,	Portugal	 411	 %FO	 ++	 +	 +	 +	 +	 -	 ++	 ++	 +	 	 	 	 Oliveira	et	al.	(2007)	Syngnathus	 typhle	Aegean	Sea,	Turkey	 95	 %FO	 +	 -	 -	 -	 -	 -	 +	 +	 +	 +	 -	 +	 Uncumusaoglu	et	al.	(2017)	Vanacampus	 phillipi	PoF,	Australia	 26	 %FO	 ++++	 ++++	 ++++	 ++	 ++	 +	 +	 +	 +++	 +	 +	 +++	 Kendrick	&	Hydnes	(2005)	Vanacampus	 poecilolaemus	PoF,	Australia	 67	 %FO	 ++++	 ++++	 +	 ++++	 +	 -	 ++	 ++	 +	 +	 -	 +	 Kendrick	&	Hydnes	(2005)			–								=0%		+								>0	to	25%	++						>25	to	50%	+++				>50	to	75%	++++		>75%		 94	Table	3.2	Measures	of	phylogenetic	signal	for	syngnathid	morphological	characteristics.	Statistical	significance	indicated	in	bold	(P	<	0.05).	N	indicates	the	number	of	syngnathid	species	that	were	used	to	assess	the	phylogenetic	signal.				 		 		 Bloomberg's	K	Characteristic	 N	 	 K	 P-value	Relative	fin	size	 27	 	 0.93	 0.006	Max.	Standard	length	(StL)	 67	 	 0.96	 0.001	Head	length	(HL)	 67	 	 1.02	 0.001	Snout	length	(SnL)	 67	 	 1.02	 0.001	Snout	depth	(SnD)	 67	 	 0.84	 0.001	StL:HL	 67	 	 0.95	 0.001	HL:SnL	 67	 	 0.67	 0.001	SnL:SnD	 67	 		 0.72	 0.001																	 95	Table	3.3	Results	of	the	redundancy	analyses	that	measured	the	associations	between	multiple	independent	variables	(syngnathid	morphological	characteristics)	and	multiple	dependent	variables	(dietary	categories,	e.g.	amphipods).	Models	were	based	on	three	different	dietary	metrics:	bulk,	numeric,	and	frequency	of	occurrence	dietary	data.	Statistical	significance	indicated	in	bold	(P	<	0.05).		Model	 Variable	 F	 P-value	Bulk	dietary	data	 	 		 Relative	fin	size	 0.29	 0.870		 Max	Standard	length	(StL)	 1.46	 0.223		 Snout	length	(SnL)	 2.42	 0.068		 Snout	depth	(SnD)	 1.14	 0.329		 HL:SnL	 4.44	 0.010		SnL:SnD	 5.93	 0.003	Numeric	dietary	data	 	 		 Relative	fin	size	 0.10	 0.985		 Max	Standard	length	(StL)	 0.41	 0.748		 Snout	depth	(SnD)	 1.46	 0.221		 HL:SnL	 0.47	 0.702		 SnL:SnD	 0.27	 0.850	Frequency	of	occurrence	data	 	 		 Relative	fin	size	 1.57	 0.160		 Max	Standard	length	(StL)	 1.96	 0.081		 Snout	length	(SnL)	 1.21	 0.301		 Snout	depth	(SnD)	 1.12	 0.347		 HL:SnL	 2.29	 0.044			 SnL:SnD	 1.58	 0.154									 96		Figure	3.1	Proportion	of	(a)	bulk	dietary	data	variance,	(b)	numeric	dietary	data	variance,	and	(c)	frequency	of	occurrence	dietary	variance	explained	by	syngnathid	body	traits	[component	a;	Relative	fin	size	+	Max.	standard	length	(StL)	+	Snout	depth	(SnD)	+	HL:SnL	+	SnL:SnD],	genus	[component	c],	covariance	between	syngnathid	body	traits	and	genus	[component	b],	and	unexplained	residuals	[component	d].		 97			Figure	3.2	Figure	1	from	©	Van	Wassenbergh,	S.,	Strother,	J.	A.,	Flammang,	B.	E.,	Ferry-Graham,	L.	A.	&	Aerts,	P.,	Extremely	fast	prey	capture	in	pipefish	is	powered	by	elastic	recoil,	Journal	of	the	Royal	Society	Interface,	2008,	5,	20,	page	286,	by	permission	of	the	Royal	Society.	A	schematic	of	a	syngnathid	body	(Syngnathus	leptorhynchus)	during	a	feeding	strike.	Specialized	sternohyoideus	muscles	run	along	the	dorsal	and	ventral	sides	of	the	pipefish.	When	contracted,	they	pull	the	hyoid	arch	towards	the	body.	The	neurocranium	(including	the	snout)	then	rotates	away	from	the	body,	towards	the	prey.			 98		Figure	3.3	Triplot	of	the	first	two	axes	of	the	redundancy	analyses	(RDA)	performed	on	bulk	dietary	data.	Points	represent	the	diet	of	a	particular	species	of	syngnathid	in	a	particular	area,	as	reported	by	a	particular	study.	Points	are	coloured	based	on	the	genus.	Environmental	vectors	for	syngnathid	body	characteristics	are	fit	onto	the	ordination,	and	the	direction	and	strength	of	the	gradient	is	represented	by	the	length	of	the	arrow.	0 5 10−50510RDA1RDA2−101FinratioMaxSLSnL.SnDHL.SnLSnDAmphipodaMysidaDecapodaCopepodaTotalEggs−101FilicampusHippichthysHippocampusHistiogamphelusLissocampusMitotichthysNerophisPhyllopteryxPugnasoStigmatoporaSyngnathiodesSyngnathusUrocampusVanacampus	 99		Figure	3.4	Triplot	of	the	first	two	axes	of	the	redundancy	analyses	(RDA)	performed	on	bulk	dietary	data.	Points	represent	the	diet	of	a	particular	species	of	syngnathid	in	a	particular	area,	as	reported	by	a	particular	study.	Points	are	coloured	based	on	the	genus.	Environmental	vectors	for	syngnathid	body	characteristics	are	fit	onto	the	ordination,	and	the	direction	and	strength	of	the	gradient	is	represented	by	the	length	of	the	arrow.				−10 −5 0 5−6−4−20246RDA1RDA2−101Finratio MaxSLSnL.SnDHL.SnLSnDAmphipodaMysidaDecapodaCopepodaOstracoda TotalLarvaeOther−101AnarchopterusFilicampusHippichthysHippocampusHistiogamphelusLissocampusMitotichthysPhyllopteryxPugnasoStigmatoporaSyngnathusVanacampus	 100	Chapter	4 Conclusions	My	thesis	helps	reveal,	for	the	first	time,	the	interplay	between	habitats,	prey,	and	predators	in	shaping	the	ecology	of	fishes	in	the	family	Syngnathidae.	In	Chapter	2,	I	used	the	seahorse	Hippocampus	whitei	as	a	case	study,	exploring	ecological	correlates	of	abundance	and	distributions.	Expanding	on	previous	work	that	had	investigated	how	either	their	habitats	or	their	prey	or	their	predators	affected	their	populations,	I	considered	all	three	components	of	their	environment	together,	using	a	holistic	approach.	I	also	investigated	how	the	correlations	differed	at	two	scales:	among	different	seagrass	beds	(200-6000	m	apart),	and	within	a	single	seagrass	bed	(<100	m	in	size).	I	found	that	habitat,	prey,	and	predator	variables	all	correlated	with	seahorse	density	or	their	size	distributions	to	varying	extents,	depending	on	the	scale	of	the	study.	In	Chapter	3,	I	then	probed	syngnathid	foraging,	synthesizing	a	large	amount	of	fragmented	information	on	their	feeding	and	diets.	My	focus	was	to	relate	their	diets	to	their	relatively	diverse	feeding	morphologies	and	habitat	use	across	the	family.	In	this—the	first	study	to	compare	syngnathid	diets	across	genera,	locations,	and	studies—I	found	their	diets	were	best	explained	by	their	snout	morphologies.	These	morphologies	also	had	high	phylogenetic	signal,	suggesting	that	dietary	differences	across	genera	were	largely	explained	by	how	these	diverse	fishes	differed	with	respect	to	these	morphologies.	Ecological	studies	of	this	family	are	of	considerable	interest	because	the	unusually	intimate	and	proximate	relationship	between	these	mostly	sedentary	species	and	their	environment.			 101	4.1		 Associations	with	habitats	The	results	of	Chapter	2	suggest	that	Hippocampus	whitei	tolerate	the	habitats	in	which	they	settle,	but	then	make	the	best	of	the	hand	they	have	been	dealt.	I	found	that	seahorse	distribution	was	correlated	with	more	ecological	variables	within	a	single	bed	than	among	different	beds.	This	is	reasonable,	as	these	seahorses	are	slow	swimmers	(Consi	et	al.,	2001),	and	travel	to	new	patches	(200	to	6000	m	away	in	my	study)	would	require	that	they	cross	patches	of	open	sand	where	they	would	be	more	exposed	to	predators	and	strong	currents	(Brown,	1999;	Hendriks	et	al.,	2008;	Canion	&	Heck,	2009).	While	there	is	evidence	that	some	seahorses	migrate	to	deeper	waters	after	the	breeding	season	(Vincent	&	Sadler,	1995),	doing	so	during	the	breeding	season	could	be	risky.	After	all,	seahorses	often	live	in	low	densities	within	patchy	distributions	(Foster	&	Vincent,	2004).	Instead,	seahorses	are	well	suited	to	living	stationary	lives	in	complex	habitats,	thanks	to	excellent	maneuverability	and	camouflage.	Within	those	habitats,	the	data	from	Little	Beach	suggests	they	choose	areas	with	fewer	predators,	more	prey	options,	and	denser	canopies—locations	that	presumably	improved	their	prospects	of	surviving,	growing,	and	reproducing.		4.2		 Associations	with	prey	My	results	indicate	that	many	sedentary	syngnathids	are	generalist	feeders,	and	select	locations	that	have	a	diverse	range	of	food	options.	Presumably,	this	is	because	it	is	energetically	beneficial	for	a	slow-moving	animal	to	be	less	selective,	consuming	what	they	encounter	(Gerritsen,	1984;	Sih	&	Moore,	1990).	Our	work	revealed	no	relationship		 102	between	prey	metrics	and	abundance	and	distribution	of	the	sedentary	seahorse,	H.	whitei,	among	seagrass	beds	(Chapter	2).	Syngnathids	that	have	very	small	home	ranges	(e.g.	seahorses)	are	unlikely	to	make	risky	long-distance	movements	for	food	options	(Vincent	et	al.,	2005;	Dunham,	2010).	If	they	do	move	to	improve	their	feeding	prospects,	it	is	probably	at	small	scales.	For	example—although	I	did	not	monitor	their	actual	movement	within	a	seagrass	bed—I	found	that	seahorses	selected	locations	within	a	seagrass	bed	with	greater	numbers	of	prey	types.	It	may	be	that	sedentary	syngnathids	relocate	short	distances,	within	the	habitat	patch	in	which	they	are	already	settled,	to	locations	that	have	a	greater	number	of	prey	types	to	select	from.	These	options	could	be	energetically	beneficial,	allowing	syngnathids	to	change	the	food	they	eat	depending	on	their	current	energy	needs,	such	as	during	the	energetically	expensive	breeding	season	(Dunham,	2010).		Although	this	thesis	suggests	that	syngnathid	diet	is	driven	largely	by	prey	availability,	a	quantitative	comparison	between	syngnathid	diets	and	what	is	available	to	them	is	badly	needed	to	assess	selectivity.	As	evident	in	our	review,	diet	is	a	function	of	many	factors	including	the	relative	sizes	of	predators	and	their	prey,	predator	capture	success—and	the	relative	distributions	of	predators	and	their	prey.	In	the	absence	of	any	work	that	has	empirically	compared	syngnathid	feeding	to	the	number	and	types	of	prey	available	to	them,	however,	I	was	left	to	look	for	patterns	in	diet	across	genera	and	speculate	on	the	feeding	selectivity	of	different	syngnathids.	A	study	that	compared	syngnathid	diets	to	prey	availability	would	disentangle	the	potential	variables	that	could	lead	to	erroneous	conclusions	that	can	occur	in	both	directions.	For	example,	one	might	wrongfully	conclude	that	syngnathids	were	selecting	for	a	particular	prey	type—if	they	were	concentrated	in	their	stomachs—when	they	were	simply	feeding	at	random	in	an		 103	area	with	a	notably	high	concentration	of	that	prey	type	(Griffiths,	1973).	Alternatively,	a	stomach	with	many	prey	types	of	equal	volumes	could	mislead	some	to	believe	that	a	syngnathid	fed	randomly,	when	in	reality	their	preferred	prey	types	were	simply	not	present	in	the	area	(Griffiths,	1973).	A	study	that	looks	at	syngnathid	feeding	in	relation	to	availability	would	greatly	add	to	our	understanding—and	one	that	did	it	an	area	with	many	different	syngnathids	would	allow	for	analyses	that	could	compare	diet	selectivity	to	morphology,	behaviour	and	phylogeny.	As	an	example	of	one	such	location,	Kendrick	&	Hyndes	(2005)	found	12	species	of	syngnathid	within	9	genera	in	a	network	of	seagrasses	in	southwestern	Australia.		4.3		 Associations	with	predators	While	it	would	be	valuable	to	know	the	relative	effects	of	direct	mortality	and	predator	evasion	on	seahorse	distributions,	SCUBA-assisted	observational	studies	are	likely	not	the	answer.	In	Chapter	2,	I	concluded	that	the	negative	association	between	seahorses	and	predators	was	likely	the	result	of	predator-evasion	tactics	among	seahorses,	and	not	the	result	of	direct	mortality.	This	hypothesis	was	largely	born	from	anecdotal	evidence,	and	work	that	showed	predation	on	seahorses	to	be	rare.	As	an	example,	only	13	predation	events	were	recorded	in	the	most	thorough	evaluation	of	predation	on	seahorses	to	date,	involving	thousands	of	hours	of	observation	and	2000	individuals	(Harasti	et	al.,	2014b).	It	may	very	well	be	that	syngnathids	are	only	eaten	on	rare	occasion	by	opportunistic,	generalist	feeders	(Kleiber	et	al.,	2011)	because	of	their	unpalatable	and	energy-poor	morphologies	(Harris	et	al.,	2008).	Or,	perhaps	they	are	depredated,	but	we	are	not	looking	in	the	correct	ways.	Using	SCUBA	can	strongly	influence	what	fish	we	see		 104	underwater,	and	their	behaviours	when	we	do	see	them	(Dickens	et	al.,	2011).	To	avoid	these	diver	effects,	tethering	experiment	have	been	used	in	observational	studies	to	monitor	predation	events	(Aronson	&	Heck,	1995).	Since	many	syngnathids	show	strong	site	fidelity,	researchers	can	monitor	syngnathids	without	having	to	sacrifice	them.	Recent	work,	for	example,	has	shown	that	GoPro	cameras	can	be	set	up	facing	seahorses	in	the	wild	and	effectively	monitor	their	behaviours—including	when	they	are	attacked	by	predators	(Claasens,	2017).	Indeed	the	use	of	GoPro	cameras	in	remote	observation	studies	is	becoming	more	and	more	common	in	marine	sciences	(Struthers	et	al.,	2015).	If	adapted	correctly,	we	may	be	able	to	measure	relative	frequency	with	which	syngnathids	are	consumed	by	their	predators,	and	better	understand	how	syngnathid	distributions	are	shaped	by	their	predators.	Future	work	on	syngnathid	predators	should	focus	on	discerning	the	impact	of	predation	on	mortality	of	newborn	juveniles.	While	conducting	my	fieldwork,	one	of	the	most	memorable	moments	was	witnessing	a	seahorse	birth	in	the	wild.	I	was	struck	by	how	helpless	the	newborn	seahorses	seemed	as	they	drifted	towards	the	surface—at	the	mercy	of	the	ocean	currents.	This	was	not	an	uncommon	occurrence	as	many	seahorse	species	have	a	planktonic	phase	in	their	first	few	weeks	of	existence	(Foster	&	Vincent,	2004).	That	said,	we	know	very	little	about	the	fate	of	newborn	syngnathids	(Vincent	&	Giles,	2003),	and	a	large	percentage	may	end	up	as	food	for	piscivorous	fish	(Foster	&	Vincent,	2004).	Newborn	syngnathids	are	extremely	small	(2-20	mm;	Foster	&	Vincent,	2004)	and	could	conceivably	fit	into	the	gapes	of	most	piscivorous	fish.	So	despite	the	poor	nutritional	quality	of	syngnathids	(Harris	et	al.,	2008)—which	has	been	credited	as	a	reason	for	the	low	rates	of	predation	on	adults	(Kleiber	et	al.,	2011)—opportunistic	feeders		 105	would	likely	not	hesitate	to	eat	newborns	(Gerking,	1994).	In	fact,	spikes	in	the	newborn	syngnathid	populations	(i.e.	during	breeding	seasons)	can	lead	to	higher	rates	of	predation	on	syngnathids	(Kleiber	et	al.,	2011).	It	may	be	that	opportunistic	predators—which	rarely	feed	on	adults	syngnathids—have	a	large	role	in	the	very	high	mortality	rate	among	newborn	syngnathids.	As	such,	a	study	that	tracked	the	fate	of	newborn	syngnathids	could	shed	light	on	the	effect	of	predators	in	juvenile	recruitment.		4.4		 How	this	thesis	fits	in	to	the	literature	My	discoveries	in	this	thesis	can	make	a	contribution	to	applied	questions	of	seahorse	conservation—many	species	are	threatened—and	management.	Prior	to	this	thesis,	we	often	knew	where	seahorses	lived	and	what	kinds	of	habitats	they	occupied,	but	were	less	aware	of	how	this	mattered	with	respect	to	their	populations.	This	thesis	identified	that	at	the	scale	of	200	m	to	6000	m,	that	total	predators	was	the	only	variable	that	was	correlated	with	seahorse	abundance.	The	association	between	seahorses	and	predators	continued	within	a	single	habitat	patch	(<100	m),	showing	that	seahorses	occurred	in	areas	where	predator	numbers	were	lower.	Such	knowledge	matters	in	designing	marine	management	and	particularly	in	placing	MPAs.	One	study	has	argued	that	that	MPAs,	by	fostering	an	increase	in	the	number	of	predatory	fishes,	could	actually	add	to	the	pressures	on	seahorses	(Harasti	et	al.,	2014b).	The	results	of	this	thesis	do	not	align	with	that	hypothesis,	however,	as	I	found	our	highest	seahorse	densities	in	a	sanctuary	zone	that	banned	recreational	fishing.	While	fish	known	to	eat	seahorses	are	highly	desirable	fishing	targets	(Harasti	et	al.,	2014b),	it	could	be	that	recreational	fishing	does		 106	not	remove	the	predatory	pressure	to	a	large	enough	extent	to	have	a	tangible	effect	on	seahorse	numbers.	MPAs	are	most	likely	to	protect	seahorses	where	seahorse	extraction	is	prohibited	or	where	destructive	fishing	that	damages	habitats	is	banned	(Vincent	et	al.,	2011).	Overall	our	understanding	of	how	MPAs	affect	these	flagship	fishes	is	still	at	an	elementary	stage	(Yasué	et	al.,	2012).	To	provide	conservationists	with	the	tools	to	best	select	the	locations	for	MPAs,	we	need	to	better	understand	how	they	affect	seahorses,	both	directly	and	indirectly.	Using	syngnathids	as	a	case	study,	my	thesis	improves	on	our	understanding	of	how	animals	are	shaped	by	their	world—including	their	habitats,	their	prey,	and	their	predators.	I	used	syngnathids	as	a	case	study—a	group	of	morphologically	diverse	marine	fishes	which	have	evolved	to	live	in	certain	habitats,	eat	certain	prey,	and	hide	from	certain	predators.	To	best	understand	how	syngnathids	interact	with	their	environments—my	thesis	shows	that	it	is	important	to	consider	all	three	components.	First,	syngnathids	are	shaped	by	their	habitats.	Syngnathids	are	found	in	all	of	the	world’s	oceans,	living	amongst	many	different	types	of	coastal	habitats—including	coral	reefs,	macroalgae,	mangroves,	seagrasses,	sponges	and	artificial	structures	(Dawson,	1982,	1985;	Foster	&	Vincent,	2004).	While	some	species	have	evolved	to	live	in	specific	habitat	types,	many	syngnathids	are	found	living	in	more	than	one	type.	In	this	thesis,	I	found	H.	whitei	distributions	were	shaped	by	specific	components	of	their	habitats,	including	depth	and	density.	This	was	potentially	an	example	of	the	mediating	effect	that	habitats	can	have	on	predator-prey	interactions,	as	more	complex	locations	may	improve	the	feeding	prospects	of	sedentary	syngnathids	such	as	H.	whitei.	Second,	syngnathids	are	also	shaped	by	their	prey.	In	my	thesis,	I	showed	that	seahorses	selected	locations	with	more	prey	types,	presumably		 107	because	more	prey	options	would	be	energetically	beneficial.	I	also	reviewed	how	syngnathids	have	been	physically	shaped	by	their	prey.	Syngnathids	have	evolved	highly-advanced	prey-capture	morphologies	that	aid	in	the	capture	of	elusive	prey	(Van	Wassenbergh	et	al.,	2008,	2009;	Roos	et	al.,	2009b;	Gemmell	et	al.,	2013).	These	morphologies	vary	across	genera,	and	I	showed	that	across	genera,	locations,	and	studies,	feeding	morphologies—and	in	particular,	their	snout	shape—are	correlated	with	what	they	eat.	Third,	syngnathids	are	also	shaped	by	their	predators.	In	response	to	predation,	syngnathids	demonstrate	various	forms	of	behavioural	and	physiological	crypsis.	Some	pipefish	species	mimic	the	rhythmic	movements	of	seagrass,	while	seahorses	are	able	to	change	colour	to	match	their	backgrounds,	and	grow	long	skin	filaments	(Howard	&	Koehn,	1985;	Foster	&	Vincent,	2004).	This	was	the	first	study	to	demonstrate	that	syngnathid	abundance	and	distributions	correlates	with	presence	of	their	predators,	after	controlling	for	other	facets	of	their	environment.	Overall,	I	show	that	syngnathid	abundances	and	distributions	are	affected	by	their	habitats,	prey,	and	predators,	to	some	degree.	To	best	understand	how	an	animal	interacts	with—and	is	shaped	by—its	environment,	my	thesis	demonstrates	that	animal	studies	should	focus	on	investigating	these	relationships	from	multiple	angles.							 108	Bibliography		Abrams,	P.	A.	(2000).	The	evolution	of	predator-prey	interactions:	theory	and	evidence.	Annual	Review	of	Ecological	Systems	31,	79–105.	doi:	10.1146/annurev.ecolsys.31.1.79	Abrams,	P.	A.	&	Ginzburg,	L.	R.	(2000).	The	nature	of	predation:	prey	dependent,	ratio	dependent	or	neither?	Trends	in	Ecology	and	Evolution	15,	337–341.	doi:	10.1016/S0169-5347(00)01908-X	Arditi,	R.	&	Ginzburg,	L.	R.	(1989).	Coupling	in	predator-prey	dynamics:	ratio-dependence.	Journal	of	Theoretical	Biology	139,	311–326.	doi:	10.1016/S0022-5193(89)80211-5	Aronson,	R.	B.	&	Heck,	K.	L.	(1995).	Tethering	experiments	and	hypothesis	testing	in	ecology.	Marine	Ecology	Progress	Series	121,	307–309.	Aylesworth,	L.	A.,	Xavier,	J.	H.	A.,	Oliveira,	T.	P.	R.,	Tenorio,	G.	D.,	Diniz,	A.	F.	&	Rosa,	I.	L.	(2015).	Regional-scale	patterns	of	habitat	preference	for	the	seahorse	Hippocampus	reidi	in	the	tropical	estuarine	environment.	Aquatic	Ecology	49,	499–512.	doi:	10.1007/s10452-015-9542-3	Barnes,	C.,	Maxwell,	D.,	Reuman,	D.	C.,	Jennings,	S.,	Barnes,	C.,	Maxwell,	D.,	Reuman,	D.	C.	&	Jennings,	S.	(2010).	Global	patterns	in	predator	—	prey	size	relationships	reveal	size	dependency	of	trophic	transfer	efficiency.	Ecology	91,	222–232.	Bell,	E.	M.,	Lockyear,	J.	F.,	McPherson,	J.	M.,	Marsden,	A.	D.	&	Vincent,	A.	C.	J.	(2003).	First	field	studies	of	an	endangered	South	African	seahorse,	Hippocampus	capensis.	Environmental	Biology	of	Fishes	67,	35–46.	doi:	10.1023/A:1024440717162	Bell,	J.	D.	&	Westoby,	M.	(1986).	Abundance	of	macrofauna	in	dense	seagrass	is	due	to		 109	habitat	preference,	not	predation.	Oecologia	68,	205–209.	Bennett,	B.	A.	(1989).	The	diets	of	fish	in	three	south-western	Cape	estuarine	systems.	South	African	Journal	of	Zoology	24,	163–177.	doi:	10.1080/02541858.1989.11448149	Bennett,	B.	A.	&	Branch,	G.	M.	(1990).	Relationships	between	production	and	consumption	of	prey	species	by	resident	fish	in	the	Bot,	a	cool	temperate	South	African	estuary.	Estuarine,	Coastal	and	Shelf	Science	31,	139–155.	doi:	10.1016/0272-7714(90)90043-Q	Bergert,	B.	A.	&	Wainwright,	P.	C.	(1997).	Morphology	and	kinematics	of	prey	capture	in	the	sygnathid	fishes	Hippocampus	erectus	and	Sygnathus	floridae.	Marine	Biology	127,	563–570.	Berglund,	A.,	Rosenqvist,	G.	&	Robinson-Wolrath,	S.	(2006).	Food	or	sex—males	and	females	in	a	sex	role	reversed	pipefish	have	different	interests.	Behavioral	Ecology	and	Sociobiology	60,	281–287.	doi:	10.1007/s00265-006-0166-4	Blomberg,	S.	P.	&	Garland,	T.	(2002).	Tempo	and	mode	in	evolution:	phylogenetic	inertia,	adaptation	and	comparative	methods.	Journal	of	Evolutionary	Biology	15,	899–910.	doi:	10.1046/j.1420-9101.2002.00472.x	Blomberg,	S.	P.,	Garland,	T.	&	Ives,	A.	R.	(2003).	Testing	for	phylogenetic	signal	in	comparative	data:	behavioral	traits	are	more	labile.	Evolution	57,	717.	doi:	10.1554/0014-3820(2003)057[0717:TFPSIC]2.0.CO;2	Bolker,	B.	M.,	Brooks,	M.	E.,	Clark,	C.	J.,	Geange,	S.	W.,	Poulsen,	J.	R.,	Stevens,	M.	H.	H.	&	White,	J.-S.	S.	(2009).	Generalized	linear	mixed	models:	a	practical	guide	for	ecology	and		 110	evolution.	Trends	in	Ecology	and	Evolution	24,	127–135.	doi:	10.1016/j.tree.2008.10.008	Bommarco,	R.,	Biesmeijer,	J.	C.,	Meyer,	B.,	Stuart,	P.,	Potts,	S.	G.,	Po,	J.,	Roberts,	M.,	Steffan-Dewenter,	I.	&	Ockinger,	E.	(2010).	Dispersal	capacity	and	diet	breadth	modify	the	response	of	wild	bees	to	habitat	loss.	Proceedings	of	the	Royal	Society	of	London	B.	doi:	10.1098/rspb.2009.2221	Bowman,	R.	E.,	Stillwell,	C.	E.,	Michaels,	W.	L.	&	Grosslein,	M.	D.	(2000).	Food	of	Northwest	Atlantic	Fishes	and	Two	Common	Species	of	Squid.	U.S.	Department	of	Commerce	National	Oceanic	and	Atmospheric	Administration	Nation	Marine	Fisheries	Service	Northeast	Region.	Retrieved	from	http://nefsc.noaa.gov/nefsc/publications/tm/tm155/tm155.pdf	Boyce,	M.	S.,	Vernier,	P.	R.,	Nielsen,	S.	E.	&	Schmiegelow,	F.	K.	A.	(2002).	Evaluating	resource	selection	functions.	Ecological	Modelling	157,	281–300.	doi:	10.1016/S0304-3800(02)00200-4	Briand,	F.	&	Cohen,	J.	E.	(1987).	Environmental	correlates	of	food	chain	length.	Science	238,	956–960.	doi:	10.1126/science.3672136	Brodie,	E.	D.	&	Brodie,	E.	D.	(1999).	Costs	of	exploiting	poisonous	prey:	evolutionary	trade-offs	in	a	predator-prey	arms	race.	Evolution	53,	626–631.	Brook,	I.	M.	(1977).	Trophic	relationships	in	a	seagrass	community	(Thalassia	testudinum),	in	Card	Sound,	Florida.	Fish	dies	in	relation	to	macrobenthic	and	cryptic	faunal	abundance.	Transactions	of	the	American	Fisheries	Society	106,	219–229.		 111	Brown,	J.	D.	(1972).	A	comparative	life	history	study	of	four	species	of	pipefishes	(family	Syngnathidae)	in	Florida.	University	of	Florida.	Brown,	J.	H.	&	Maurer,	B.	A.	(1989).	Macroecology:	the	division	of	food	and	space	among	species	on	continents.	Science,	New	Series	243,	1145–1150.	Brown,	J.	S.	(1999).	Vigilance,	patch	use	and	habitat	selection:	foraging	under	predation	risk.	Evolutionary	Ecology	Research	1,	49–71.	Bruno,	J.	F.	&	Bertness,	M.	D.	(2001).	Habitat	modification	and	facilitation	in	benthic	marine	communities.	In:	Marine	Community	Ecology	(Bertness,	M.	D.,	Gaines,	S.	D.	&	Hay,	M.	E.,	eds),	pp.	201–218.	Sunderland,	MA:	Sinauer	Associates.	Burchmore,	J.	J.,	Pollard,	D.	A.	&	Bell,	J.	D.	(1984).	Community	structure	and	trophic	relationships	of	the	fish	fauna	of	an	estuarine	Posidonia	australis	seagrass	habitat	in	Port	Hacking,	New	South	Wales.	Aquatic	Botany	18,	71–87.	doi:	10.1016/0304-3770(84)90081-0	Caldwell,	I.	R.	&	Vincent,	A.	C.	J.	(2012).	Revisiting	two	sympatric	European	seahorse	species:	apparent	decline	in	the	absence	of	exploitation.	Aquatic	Conservation:	Marine	and	Freshwater	Ecosystems	22,	427–435.	doi:	10.1002/aqc.2238	Campolmi,	M.,	Franzoi,	P.	&	Mazzola,	A.	(1996).	Observations	on	pipefish	(Syngnathidae)	biology	in	the	Stagnone	lagoon	(west	Sicily).	Publications	Especial	de	I’	Institut	Espana	de	Oceanografie	21,	205–209.	Canion,	C.	R.	&	Heck,	K.	L.	(2009).	Effect	of	habitat	complexity	on	predation	success:	re-evaluating	the	current	paradigm	in	seagrass	beds.	Marine	Ecology	Progress	Series	393,		 112	37–46.	doi:	10.3354/meps08272	Castro,	A.	L.	C.,	Diniz,	A.	F.,	Martins,	I.	Z.,	Vendel,	A.	L.,	Oliveira,	T.	P.	R.	&	Rosa,	I.	L.	(2008).	Assessing	diet	composition	of	seahorses	in	the	wild	using	a	nondestructive	method:	Hippocampus	reidi	(Teleostei:	Syngnathidae)	as	a	study-case.	Neotropical	Ichthyology	6,	637–644.	doi:	10.1590/S1679-62252008000400012	Ceccherelli,	G.,	Campo,	D.	&	Milazzo,	M.	(2007).	Short-term	response	of	the	slow	growing	seagrass	Posidonia	oceanica	to	simulated	anchor	impact.	Marine	Environmental	Research	63,	341–349.	doi:	10.1016/j.marenvres.2006.10.004	Celino,	F.	T.,	Hilomen-Garcia,	G.	V.	&	del	Norte-Campos,	A.	G.	C.	(2012).	Feeding	selectivity	of	the	seahorse,	Hippocampus	kuda	(Bleeker),	juveniles	under	laboratory	conditions.	Aquaculture	Research	43,	1804–1815.	doi:	10.1111/j.1365-2109.2011.02988.x	Chambers,	R.	C.	&	Trippel,	E.	A.	(1997).	Early	Life	History	and	Recuritment	in	Fish	Populations.	London:	Chapman	&	Hall.	Claasens,	L.	(2017).	Big	brother	is	watching:	gaining	insights	into	seahorse	behaviour	using	GoPro	cameras.	In:	SyngBio	2017	-	Fantastical	Fishes:	Seahorses,	Pipefishes	and	Seadragons	into	the	Future.	Tampa,	FL.	Clutton-Brock,	T.	H.	(1975).	Feeding	behaviour	of	red	colobus	and	black	and	white	colobus	in	East	Africa.	Folia	Primatologica	23,	165–207.	Clynick,	B.	G.	(2008).	Harbour	swimming	nets:	a	novel	habitat	for	seahorses.	Aquatic	Conservation:	Marine	and	Freshwater	Ecosystems	18,	483–492.	doi:	10.1002/aqc	Cohen,	J.	E.	(1994).	Marine	and	continental	food	webs:	three	paradoxes?	Philosophical		 113	Transactions	of	the	Royal	Society	B	343,	57–69.	doi:	10.1098/rstb.1994.0008	Cohen,	J.	E.,	Braind,	F.	&	Newman,	C.	M.	(1990).	Community	Food	Webs:	Data	and	Theory.	New	York:	Springer-Verlag.	Colson,	D.	J.,	Patek,	S.	N.,	Brainerd,	E.	L.	&	Lewis,	S.	M.	(1998).	Sound	production	during	feeding	in	Hippocampus	seahorses	(Syngnathidae).	Environmental	Biology	of	Fishes	221–229.	Consi,	T.	R.,	Seifert,	P.	A.,	Triantafyllou,	M.	S.	&	Edelman,	E.	R.	(2001).	The	dorsal	fin	engine	of	the	seahorse	(Hippocampus	sp.).	Journal	of	Morphology	97,	80–97.	Cooper,	S.	D.,	Smith,	D.	W.	&	Bence,	J.	R.	(1985).	Prey	selection	by	freshwater	predators	with	different	foraging	strategies.	Canadian	Journal	of	Fish	and	Aquatic	Sciences	42,	1720–1732.	Costa,	G.	C.	(2009).	Predator	size,	prey	size,	and	dietary	niche	breadth	relationships	in	marine	predators.	Ecology	90,	2014–2019.	Cowen,	R.	K.	&	Sponaugle,	S.	(2009).	Larval	dispersal	and	marine	population	connectivity.	Annual	Review	of	Marine	Sciences	1,	443–466.	doi:	10.1146/annurev.marine.010908.163757	Crowder,	L.	B.	&	Norse,	E.	(2008).	Essential	ecological	insights	for	marine	ecosystem-based	management	and	marine	spatial	planning.	Marine	Policy	32,	772–778.	doi:	10.1016/j.marpol.2008.03.012	Curtis,	J.	M.	R.	&	Vincent,	A.	C.	J.	(2005).	Distribution	of	sympatric	seahorse	species	along	a	gradient	of	habitat	complexity	in	a	seagrass-dominated	community.	Marine	Ecology		 114	Progress	Series	291,	81–91.	doi:	10.3354/meps291081	D’Entremont,	J.	(2002).	Sex-differences	in	feeding	behaviour	and	diet	in	Hippocampus	guttulatus.	In:	Life	History	and	Ecology	of	Seahorses:	Implications	for	Conservation	and	Management	(Foster,	S.	J.	&	Vincent,	A.	C.	J.,	eds).	Davis,	T.	R.,	Gallen,	C.,	Laird,	R.	&	Harasti,	D.	(2015).	Mapping	sub-tidal	habitats	in	the	Eastern	Port	of	Port	Stephens.	Retrieved	from	http://hunter.lls.nsw.gov.au/__data/assets/pdf_file/0008/572291/portstephenshabitatmappingreport.pdf	Dawkins,	R.	&	Krebs,	J.	R.	(1979).	Arms	races	between	and	within	species.	Proceedings	of	the	Royal	Society	of	London	B	205,	489–511.	Dawson,	C.	E.	(1982).	Fishes	of	the	Western	North	Atlantic,	Part	8.	(eds	Cohen,	D.	M.,	Marshall,	N.	B.,	Ebeling,	A.	W.,	et	al.).	New	Haven,	CT:	Yale	University.	Dawson,	C.	E.	(1985).	Indo-Pacific	Pipefishes,	Red	Sea	to	the	Americas.	Ocean	Springs,	MS:	The	Gulf	Coast	Research	Laboratory.	Dayton,	P.	(1972).	Toward	an	understanding	of	resilience	and	the	potential	effects	of	enrichments	to	the	benthos	at	McMurdo	Sound,	Antarctica.	In:	Proceedings	of	the	Colloquium	of	Conservation	Problems	in	Antarctica,	pp.	81–96.	Lawrence,	KS:	Allen	Press.	de	Lussanet,	M.	H.	E.	&	Muller,	M.	(2007).	The	smaller	your	mouth,	the	longer	your	snout:	predicting	the	snout	length	of	Syngnathus	acus,	Centriscus	scutatus	and	other	pipette	feeders.	Journal	of	the	Royal	Society,	Interface	/	the	Royal	Society	4,	561–573.	doi:		 115	10.1098/rsif.2006.0201	Dias,	T.	L.	P.	&	Rosa,	I.	L.	(2003).	Habitat	preferences	of	a	seahorse	species,	Hippocampus	reidi	(Teleostei:	Syngnathidae)	in	Brazil.	Aqua	Journal	of	Ichthyology	and	Aquatic	Biology	6,	165–176.	Dickens,	L.	C.,	Goatley,	C.	H.	R.,	Tanner,	J.	K.	&	Bellwood,	D.	R.	(2011).	Quantifying	relative	diver	effects	in	underwater	visual	censuses.	PLoS	ONE	6,	e18965.	doi:	10.1371/journal.pone.0018965	Do,	H.	H.,	Truong,	S.	K.	&	Ho,	T.	H.	(1996).	Feeding	behaviour	and	food	of	seahorses	in	Vietnam.	In:	Proceedings	of	the	Third	International	Conference	on	the	Marine	Biology	of	the	South	China	Sea.	Hong	Kong:	Hong	Kong	University	Press.	Dunham,	N.	M.	(2010).	The	life	history	and	energy	budget	of	Hippocampus	erectus	in	Tampa	Bay,	Florida.	University	of	South	Florida.	Retrieved	from	http://scholarcommons.sc.edu/cgi/viewcontent.cgi?article=1345&context=sph_epidemiology_biostatistics_facpub	Felicio,	A.	K.	C.,	Rosa,	I.	L.,	Souto,	A.	&	Freitas,	R.	H.	A.	(2006).	Feeding	behavior	of	the	longsnout	seahorse	Hippocampus	reidi	Ginsburg,	1933.	Journal	of	Ethology	24,	219–225.	doi:	10.1007/s10164-005-0189-8	Flammang,	B.	E.,	Ferry-Graham,	L.	A.,	Rinewalt,	C.,	Ardizzone,	D.,	Davis,	C.	&	Trejo,	T.	(2009).	Prey	capture	kinematics	and	four-bar	linkages	in	the	bay	pipefish,	Syngnathus	leptorhynchus.	Zoology	112,	86–96.	doi:	10.1016/j.zool.2008.04.003	Flynn,	A.	J.	&	Ritz,	D.	A.	(1999).	Effect	of	habitat	complexity	and	predatory	style	on	the		 116	capture	success	of	fish	feeding	on	aggregated	prey.	Journal	of	the	Marine	Biological	Association	of	the	UK	79,	487–494.	doi:	10.1017/S0025315498000617	Foster,	S.	J.	&	Vincent,	A.	C.	J.	(2004).	Life	history	and	ecology	of	seahorses:	implications	for	conservation	and	management.	Journal	of	Fish	Biology	65,	1–61.	doi:	10.1111/j.1095-8649.2004.00429.x	Franzoi,	P.,	Maccagnani,	R.,	Rossi,	R.	&	Ceccherelli,	V.	U.	(1993).	Life	cycles	and	feeding	habits	of	Syngnathus	taenionotus	and	S.	abaster	(Pisces,	Syngnathidae)	in	a	brackish	bay	of	the	Po	River	Delta	(Adriatic	Sea).	Marine	Ecology-Progress	Series	97,	71–81.	Franzoi,	P.,	Riccato,	F.,	Franco,	A.	&	Torricelli,	P.	(2004).	Dietary	differences	in	three	pipefish	species	(Osteichthyes,	Syngnathidae)	related	to	snout	morphology.	Biologia	Marina	Mediterranea	11,	592–594.	Fretwell,	S.	D.	&	Lucas,	H.	L.	(1970).	On	territorial	behaviour	and	other	factors	influencing	habitat	distribution	in	birds.	Acta	Biotheoretica	19,	16–36.	doi:	10.1007/BF01601953	Froese,	R.	&	Pauly,	D.	(2017).	FishBase.	Retrieved	November	1,	2016,	from	www.fishbase.org	Garcia,	A.	M.,	Geraldi,	R.	M.	&	Vieira,	J.	P.	(2005).	Diet	composition	and	feeding	strategy	of	the	southern	pipefish.	Neotropical	Ichthyology	3,	427–432.	Garcia,	L.	M.	B.,	Hilomen-Garcia,	G.	V.,	Celino,	F.	T.,	Gonzales,	T.	T.	&	Maliao,	R.	J.	(2012).	Diet	composition	and	feeding	periodicity	of	the	seahorse	Hippocampus	barbouri	reared	in	illuminated	sea	cages.	Aquaculture	358–359,	1–5.	doi:	10.1016/j.aquaculture.2012.06.013		 117	Gaughan,	D.	J.	&	Potter,	I.	C.	(1997).	Analysis	of	diet	and	feeding	strategies	within	an	assemblage	of	estuarine	larval	fish	and	an	objective	assessment	of	dietary	niche	overlap.	Fishery	Bulletin	95,	722–731.	Gemmell,	B.	J.,	Sheng,	J.	&	Buskey,	E.	J.	(2013).	Morphology	of	seahorse	head	hydrodynamically	aids	in	capture	of	evasive	prey.	Nature	Communications	4,	2840.	doi:	10.1038/ncomms3840	Gendron,	R.	P.	&	Staddon,	J.	E.	R.	(1983).	Searching	for	cryptic	prey:	the	effect	of	search	rate.	The	American	Naturalist	121,	172–186.	Gerking,	S.	D.	(1994).	The	Feeding	Ecology	of	Fish.	San	Diego:	Academic	Press,	Inc.	Gerritsen,	J.	(1984).	Size	efficiency	reconsidered:	a	general	foraging	model	for	free-swimming	aquatic	animals.	The	American	Naturalist	123,	450–467.	doi:	10.1086/284216	Gladstone,	W.	(2007).	Selection	of	a	spawning	aggregation	site	by	Chromis	hypsilepis	(Pisces:	Pomacentridae):	habitat	structure,	transport	potential,	and	food	availability.	Marine	Ecology	Progress	Series	351,	235–247.	doi:	10.3354/meps07108	Griffiths,	D.	(1973).	The	food	of	animals	in	an	acid	moorland	pond.	Journal	of	Animal	Ecology	42,	285–293.	Griffiths,	D.	(1975).	Prey	availability	and	the	food	of	predators.	Ecology	56,	1209–1214.	Grueber,	C.	E.,	Nakagawa,	S.,	Laws,	R.	J.	&	Jamieson,	I.	G.	(2011).	Multimodel	inference	in	ecology	and	evolution:	challenges	and	solutions.	Journal	of	Evolutionary	Biology	24,	699–711.	doi:	10.1111/j.1420-9101.2010.02210.x		 118	Gurkan,	S.,	Sever,	T.	M.	&	Taskavak,	E.	(2011a).	Seasonal	food	composition	and	prey-length	relationships	of	pipefish	Nerophis	ophidion	(Linnaeus,	1758)	inhabiting	the	Aegean	Sea.	Acta	Adriatica	52,	5–14.	Gurkan,	S.,	Taskavak,	E.,	Sever,	T.	M.	&	Akalin,	S.	(2011b).	Gut	contents	of	two	European	seahorses	Hippocampus	hippocampus	and	Hippocampus	guttulatus	in	the	Aegean	Sea,	coasts	of	Turkey.	Pakistan	Journal	of	Zoology	43,	1197–1201.	doi:	0030-9923/2011/0006-1197	Hamilton,	H.,	Saarman,	N.,	Short,	G.,	Sellas,	A.	B.,	Moore,	B.,	Hoang,	T.,	Grace,	C.	L.,	Gomon,	M.,	Crow,	K.	&	Brian	Simison,	W.	(2017).	Molecular	phylogeny	and	patterns	of	diversification	in	syngnathid	fishes.	Molecular	Phylogenetics	and	Evolution	107,	388–403.	doi:	10.1016/j.ympev.2016.10.003	Hanekom,	N.	&	Baird,	D.	(1984).	Fish	community	structures	in	Zostera	and	non-Zostera	regions	of	the	Kromme	estuary,	St	Francis	Bay.	South	African	Journal	of	Zoology	19,	295–301.	Harasti,	D.	(2016).	Declining	seahorse	populations	linked	to	loss	of	essential	marine	habitats.	Marine	Ecology	Progress	Series	546,	173–181.	doi:	10.3354/meps11619	Harasti,	D.,	Glasby,	T.	M.	&	Martin-Smith,	K.	M.	(2010).	Striking	a	balance	between	retaining	populations	of	protected	seahorses	and	maintaining	swimming	nets.	Aquatic	Conservation:	Marine	and	Freshwater	Ecosystems	20,	159–166.	doi:	10.1002/aqc	Harasti,	D.,	Martin-Smith,	K.	M.	&	Gladstone,	W.	(2012).	Population	dynamics	and	life	history	of	a	geographically	restricted	seahorse,	Hippocampus	whitei.	Journal	of	Fish	Biology	81,	1297–1314.	doi:	10.1111/j.1095-8649.2012.03406.x		 119	Harasti,	D.,	Martin-Smith,	K.	M.	&	Gladstone,	W.	(2014a).	Ontogenetic	and	sex-based	differences	in	habitat	preferences	and	site	fidelity	of	White’s	seahorse	Hippocampus	whitei.	Journal	of	Fish	Biology	85,	1413–1428.	doi:	10.1111/jfb.12492	Harasti,	D.,	Martin-Smith,	K.	M.	&	Gladstone,	W.	(2014b).	Does	a	no-take	marine	protected	area	benefit	seahorses?	PLoS	ONE	9.	doi:	10.1371/journal.pone.0105462	Haris,	K.,	Chakraborty,	B.,	Menezes,	A.,	Sreepada,	R.	A.	&	Fernandes,	W.	A.	(2014).	Multifractal	detrended	fluctuation	analysis	to	characterize	phase	couplings	in	seahorse	(Hippocampus	kuda)	feeding	clicks.	Journal	of	the	Acoustical	Society	of	America	136,	1972–1981.	doi:	10.1121/1.4895713	Harris,	M.	P.,	Newell,	M.,	Daunt,	F.,	Speakman,	J.	R.	&	Wanless,	S.	(2008).	Snake	Pipefish	Entelurus	aequoreus	are	poor	food	for	seabirds.	Ibis	150,	413–415.	doi:	10.1111/j.1474-919X.2007.00780.x	Heck,	K.	L.	&	Crowder,	L.	B.	(1991).	Habitat	structure	and	predator-prey	interactions	in	vegetated	aquatic	systems.	In:	Habitat	Structure	(Bell,	S.	S.,	McCoy,	E.	D.	&	Mushinsky,	H.	R.,	eds),	pp.	281–299.	Dordrecht:	Springer	Netherlands.	doi:	10.1007/978-94-011-3076-9_14	Heck,	K.	L.,	Hays,	G.	&	Orth,	R.	J.	(2003).	Critical	evaluation	of	nursery	hypothesis	for	seagrasses.	Marine	Ecology	Progress	Series	253,	123–136.	doi:	10.3354/meps253123	Heck,	K.	L.	&	Orth,	R.	J.	(2006).	Predation	in	Seagrass	Beds.	In:	Seagrasses:	Biology,	Ecology	and	Conservation	(Larkum,	A.,	Orth,	R.	J.	&	Duarte,	C.,	eds),	pp.	537–550.	Dordrecht:	Springer	Netherlands.	doi:	10.1007/1-4020-2983-7_22		 120	Hendriks,	I.	E.,	Sintes,	T.,	Bouma,	T.	J.	&	Duarte,	C.	M.	(2008).	Experimental	assessment	and	modeling	evaluation	of	the	effects	of	the	seagrass	Posidonia	oceanica	on	flow	and	particle	trapping.	Marine	Ecology	Progress	Series	356,	163–173.	doi:	10.3354/meps07316	Herald,	E.	S.	&	Rakowicz,	M.	(1951).	Stable	requirements	for	raising	seahorses.	Aquarium	Journal	22,	234–242.	Hislop,	J.	R.	G.,	Robb,	A.	P.	&	Gauld,	J.	A.	(1978).	Observations	on	effects	of	feeding	level	on	growth	and	reproduction	in	haddock,	Melanogrammus	aeglefinue	(L.)	in	captivity.	Journal	of	Fish	Biology	13,	85–98.	Hixon,	M.	A.	&	Menge,	B.	A.	(1991).	Species	interactive	diversity:	prey	refuges	modify	the	effects	of	predation	and	competition.	Theoretical	Population	Biology	39,	178–200.	Horinouchi,	M.	&	Sano,	M.	(2000).	Food	habits	of	fishes	in	a	Zostera	marina	bed	at	Aburatsubo,	central	Japan.	Ichthyological	Research	47,	163–173.	doi:	10.1007/BF02684237	Horinouchi,	M.,	Tongnunui,	P.,	Furumitsu,	K.,	Nakamura,	Y.,	Kanou,	K.,	Yamaguchi,	A.,	Okamoto,	K.	&	Sano,	M.	(2012).	Food	habits	of	small	fishes	in	seagrass	habitats	in	Trang,	southern	Thailand.	Fisheries	Science	78,	577–587.	doi:	10.1007/s12562-012-0485-5	Howard,	R.	K.	&	Koehn,	J.	D.	(1985).	Population	dynamics	and	feeding	ecology	of	pipefish	(Syngnathidae)	associated	with	eelgrass	beds	of	Western	Port,	Victoria.	Marine	and	Freshwater	Research	36,	361–370.	doi:	10.1071/MF9850361		 121	Huh,	S.-H.	&	Kitting,	C.	L.	(1985).	Trophic	relationships	among	concentrated	populations	of	small	fishes	in	segrass	meadows.	Journal	of	Experimental	Marine	Biology	and	Ecology	92,	29–43.	Huh,	S.-H.	&	Kwak,	S.	N.	(1997).	Feeding	habits	of	Syngnathus	schlegeli	in	eelgrass	(Zostera	marina)	bed	in	Kwangyang	Bay.	Journal	of	the	Korean	Fisheries	Socieity	30,	896–902.	James,	P.	L.	&	Heck,	K.	L.	(1994).	The	effects	of	habitat	complexity	and	light	intensity	on	ambush	predation	within	a	smulated	seagrass	habitat.	Journal	of	Experimental	Marine	Biology	and	Ecology	176,	187–200.	doi:	10.1016/0022-0981(94)90184-8	Jenkins,	G.	P.	&	Sutherland,	C.	R.	(1997).	The	influence	of	habitat	structure	on	nearshore	fish	assemblages	in	a	southern	Australian	embayment:	colonisation	and	turnover	rate	of	fishes	associated	with	artificial	macrophyte	beds	of	varying	physical	structure.	Journal	of	Experimental	Marine	Biology	and	Ecology	218,	103–125.	doi:	10.1016/S0022-0981(97)00071-3	Jenkins,	G.	P.,	Walker-Smith,	G.	K.	&	Hamer,	P.	A.	(2002).	Elements	of	habitat	complexity	that	influence	harpacticoid	copepod	associated	with	seagrass	beds	in	a	temperate	bay.	Oecologia	131,	598–605.	Johnson,	C.	J.,	Nielsen,	S.	E.,	Merrill,	E.	H.,	Trent,	L.	&	Boyce,	M.	S.	(2006).	Resource	selection	functions	based	on	use–availability	data:	theoretical	motivation	and	evaluation	methods.	Journal	of	Wildlife	Management	70,	347–357.	doi:	10.2193/0022-541X(2006)70	Jones,	C.	G.,	Lawton,	J.	H.	&	Shachak,	M.	(1994).	Organisms	as	ecosystem	engineers.	Oikos	69,	373–386.	doi:	10.2307/3545850		 122	Jones,	C.	G.,	Lawton,	J.	H.	&	Shachak,	M.	(1997).	Positive	and	negative	effects	of	organisms	as	physical	ecosystem	engineers.	Ecology	78,	1946–1957.	doi:	10.1890/0012-9658(1997)078[1946:PANEOO]2.0.CO;2	Jouvenel,	J.-Y.	&	Pollard,	D.	A.	(2001).	Some	effects	of	marine	reserve	protection	on	the	population	structure	of	two	spearfishing	target-fish	species,	Dicentrarchus	labrax	(Moronidae)	and	Sparus	aurata	(Sparidae),	in	shallow	inshore	waters,	along	a	rocky	coast	in	the	northwestern	Mediterranean	.	Aquatic	Conservation:	Marine	and	Freshwater	Ecosystems	11,	1–9.	doi:	10.1002/aqc.424	Kanou,	K.	&	Kohno,	H.	(2001).	Early	life	history	of	a	seahorse,	Hippocampus	mohnikei,	in	Tokyo	Bay,	Japan.	Ichthyological	Research	48,	361–368.	doi:	10.1007/s10228-001-8159-9	Kendrick,	A.	J.	(2002).	Resource	utilisation	and	reproductive	biology	of	syngnathid	fishes	in	a	seagrass-dominated	marine	environment	in	south-western	Australia.	Murdoch	University.	Kendrick,	A.	J.	&	Hyndes,	G.	A.	(2005).	Variations	in	the	dietary	compositions	of	morphologically	diverse	syngnathid	fishes.	Environmental	Biology	of	Fishes	72,	415–427.	doi:	10.1007/s10641-004-2597-y	Kennedy,	M.,	Gray,	R.	D.,	Kennedy,	M.	&	Gray,	R.	D.	(1993).	Can	ecological	theory	predict	the	distribution	of	foraging	animals?	A	critical	analysis	of	experiments	on	the	ideal	free	distribution.	Oikos	68,	158–166.	doi:	10.2307/3545322	King,	J.	(2007).	Fisheries	Biology,	Assessment	and	Management.	Oxford,	UK:	John	Wiley	&	Sons.	doi:	10.1002/9781118688038		 123	Kitsos,	M.	S.,	Tzomos,	T.,	Anagnostopoulou,	L.	&	Koukouras,	A.	(2008).	Diet	composition	of	the	seahorses,	Hippocampus	guttulatus	Cuvier,	1829	and	Hippocampus	hippocampus	(L.,	1758)	(Teleostei,	Syngnathidae)	in	the	Aegean	Sea.	Journal	of	Fish	Biology	72,	1259–1267.	doi:	10.1111/j.1095-8649.2007.01789.x	Kleiber,	D.,	Blight,	L.	K.,	Caldwell,	I.	R.	&	Vincent,	A.	C.	J.	(2011).	The	importance	of	seahorses	and	pipefishes	in	the	diet	of	marine	animals.	Reviews	in	Fish	Biology	and	Fisheries	21,	205–223.	doi:	10.1007/s11160-010-9167-5	Kooijman,	S.	A.	L.	M.	(2010).	Dynamic	Energy	Budget	Theory:	For	Metabolic	Organization.	New	York:	Cambridge	University	Press.	Krebs,	C.	J.	(2009).	Ecology:	The	Experimental	Analysis	of	Distribution	and	Abundance,	6th	ed.	San	Francisco,	CA:	Benjamin	Cummings.	Krejci,	S.	E.	(2012).	Habitat	preferences	and	the	effects	of	seagrass	density	on	population	demographics	and	feeding	ecology	of	pipefish	in	the	Indian	River	Lagoon,	FL.	Florida	Institute	of	Technology.	Kroon,	F.	J.,	De	Graaf,	M.	&	Liley,	N.	R.	(2000).	Social	organisation	and	competition	for	refuges	and	nest	sites	in	Coryphopterus	nicholsii	(Gobiidae),	a	temperate	protogynous	reef	fish.	Environmental	Biology	of	Fishes	57,	401–411.	doi:	10.1023/A:1007600728267	Kuiter,	R.	H.	(2000).	Seahorses,	Pipefishes	and	Their	Relatives:	A	Comprehensive	Guide	to	Syngnathiformes.	Chorleywood,	UK:	TMC	Publishing.	Lauder,	G.	V.	(1985).	Aquatic	feeding	in	lower	vertebrates.	In:	Functional	Vertebrate		 124	Morphology	(Hildebrand,	M.,	Bramble,	D.	M.,	Liem,	K.	F.,	et	al.,	eds),	pp.	210–229.	Cambridge,	MA.	Lele,	S.	R.	&	Keim,	J.	L.	(2006).	Weighted	distributions	and	estimation	of	resource	selection	probability	functions.	Management	87,	3021–3028.	doi:	10.1890/0012-9658(2006)87[3021:WDAEOR]2.0.CO;2	Lester,	N.	P.,	Shuter,	B.	J.	&	Abrams,	P.	A.	(2004).	Interpreting	the	von	Bertalanffy	model	of	somatic	growth	in	fishes:	the	cost	of	reproduction	1625–1631.	doi:	10.1098/rspb.2004.2778	Leysen,	H.,	Roos,	G.	&	Adriaens,	D.	(2011).	Morphological	variation	in	head	shape	of	pipefishes	and	seahorses	in	relation	to	snout	length	and	developmental	growth.	Journal	of	Morphology	272,	1259–1270.	doi:	10.1002/jmor.10982	Lima,	S.	L.	(1998).	Nonlethal	effects	in	the	ecology	of	predator-prey	interactions.	BioScience	48,	25–34.	doi:	10.2307/1313225	Lima,	S.	L.	&	Dill,	L.	M.	(1990).	Behavioral	decisions	made	under	the	risk	of	predation:	a	review	and	prospectus.	Canadian	Journal	of	Zoology	68,	619–640.	doi:	10.1139/z90-092	Lin,	Q.,	Gao,	Y.,	Sheng,	J.,	Chen,	Q.,	Zhang,	B.	&	Lu,	J.	(2007).	The	effects	of	food	and	the	sum	of	effective	temperature	on	the	embryonic	development	of	the	seahorse,	Hippocampus	kuda	Bleeker.	Aquaculture	262,	481–492.	doi:	10.1016/j.aquaculture.2006.11.011	Livingston,	R.	J.	(1982).	Trophic	organization	of	fishes	in	a	coastal	seagrass	system.	Marine	Ecology	Progress	Series	7,	1–12.	doi:	10.3354/meps007001		 125	Livingston,	R.	J.	(1984).	Trophic	responses	of	fishes	to	habitat	variability	in	coastal	seagrass	systems.	Ecology	65,	1258–1275.	Lourie,	S.	A.,	Foster,	S.	J.,	Cooper,	E.	W.	T.	&	Vincent,	A.	C.	J.	(2004).	A	Guide	to	the	Identification	of	Seahorses.	Washington	D.C.:	University	of	British	Columbia	and	World	Wildlife	Fund.	Retrieved	from	http://www.traffic.org/species-reports/traffic_species_fish29.pdf	Lourie,	S.	A.,	Pollom,	R.	A.	&	Foster,	S.	J.	(2016).	A	global	revision	of	the	Seahorses	Hippocampus	Rafinesque	1810	(Actinopterygii:	Syngnathiformes):	Taxonomy	and	biogeography	with	recommendations	for	further	research.	Zootaxa	4146,	1–66.	doi:	10.11646/zootaxa.4146.1.1	Lourie,	S.	A.,	Pritchard,	J.	C.,	Casey,	S.	P.,	Truong,	S.	K.,	Hall,	H.	J.	&	Vincent,	A.	C.	J.	(1999).	The	taxonomy	of	Vietnam’s	exploited	seahorses	(family	Syngnathidae).	Biological	Journal	of	the	Linnean	Society	66,	231–256.	doi:	10.1111/j.1095-8312.1999.tb01886.x	Lyons,	D.	O.	&	Dunne,	J.	J.	(2004).	Inter-	and	intra-gender	analyses	of	feeding	ecology	of	the	worm	pipefish	(Nerophis	lumbriciformis).	Journal	of	the	Marine	Biological	Association	of	the	UK	84,	461–464.	doi:	10.1017/S0025315404009452h	Magnhagen,	C.	(1991).	Predation	risk	as	a	cost	of	reproduction.	Trends	in	Ecology	and	Evolution	6,	183–186.	doi:	10.1016/0169-5347(91)90210-O	Main,	K.	L.	(1987).	Predator	avoidance	in	seagrass	meadows:	prey	behavior,	microhabitat	selection,	and	cryptic	coloration.	Ecology	68,	170–180.	Manly,	B.	F.	J.,	McDonald,	L.	L.,	Thomas,	D.	L.,	McDonald,	T.	L.	&	Erickson,	W.	P.	(2002).		 126	Resource	Selection	by	Animals:	Statistical	Design	and	Analysis	for	Field	Studies.	doi:	10.1007/0-306-48151-0	Martin-Smith,	K.	M.	&	Vincent,	A.	C.	J.	(2005).	Seahorse	declines	in	the	Derwent	estuary,	Tasmania	in	the	absence	of	fishing	pressure.	Biological	Conservation	123,	533–545.	doi:	10.1016/j.biocon.2005.01.003	McArdle,	B.	H.	&	Anderson,	M.	J.	(2001).	Fitting	multivariate	models	to	community	data:	a	comment	on	distance-based	redundancy	models.	Ecology	82,	290–297.	McCauley,	E.,	Wilson,	W.	G.	&	de	Roos,	A.	M.	(1993).	Dynamics	of	age-structured	and	spatially	structured	predator-prey	interactions:	individual-based	models	and	population-level	formulations.	The	American	Naturalist	142,	412–442.	doi:	10.1086/285547	McDonald,	T.	L.	(2013).	The	point	process	use-availability	or	presence-only	likelihood	and	comments	on	analysis.	Journal	of	Animal	Ecology	82,	1174–1182.	doi:	10.1111/1365-2656.12132	McManus,	M.	A.	&	Woodson,	C.	B.	(2012).	Plankton	distribution	and	ocean	dispersal.	The	Journal	of	Experimental	Biology	215,	1008–1016.	doi:	10.1242/jeb.059014	Menge,	B.	A.	&	Olson,	A.	M.	(1990).	Role	of	scale	and	environmental	factors	in	regulation	of	community	structure.	Trends	in	Ecology	and	Evolution	5.	Mercer,	L.	P.	(1973).	The	comparative	ecology	of	two	species	of	pipefish	(Syngnathidae)	in	the	York	River,	Virginia.	The	College	of	William	and	Mary	in	Virginia.	Milinski,	M.	(1993).	Predation	risk	and	feeding	behaviour.	In:	Behaviour	of	Teleost	Fishes		 127	(Pitcher,	T.	J.,	ed),	pp.	285–306.	Dordrecht:	Springer	Netherlands.	Morgan,	S.	K.	&	Vincent,	A.	C.	J.	(2007).	The	ontogeny	of	habitat	associations	in	the	tropical	tiger	tail	seahorse	Hippocampus	comes	Cantor,	1850.	Journal	of	Fish	Biology	71,	701–724.	doi:	10.1111/j.1095-8649.2007.01535.x	Morgan,	S.	K.	&	Vincent,	A.	C.	J.	(2013).	Life-history	reference	points	for	management	of	an	exploited	tropical	seahorse.	Marine	and	Freshwater	Research	64,	185–200.	doi:	10.1071/MF12171	Motta,	P.	J.,	Clifton,	K.	B.,	Hernandez,	P.,	Eggold,	B.	T.,	Giordano,	S.	D.,	Wilcox,	R.	&	Jenny,	S.	(1995).	Feeding	relationships	among	nine	species	of	seagrass	fishes	of	Tampa	Bay,	Florida.	Bulletin	of	Marine	Science	56,	185–200.	Muller,	M.	&	Osse,	J.	W.	M.	(1984).	Hydrodynamics	of	suction	feeding	in	fish.	Transactions	of	the	Zoological	Society	of	London	37,	51–135.	doi:	10.1111/j.1096-3642.1984.tb00068.x	Nakamura,	Y.,	Horinouchi,	M.,	Nakai,	T.	&	Sano,	M.	(2002).	Food	habits	of	fishes	in	a	seagrass	bed	on	a	fringing	coral	reef	at	Iriomote	Island,	southern	Japan.	Ichthyological	Research	50,	15–22.	doi:	10.1007/s102280300002	Neutens,	C.,	Adriaens,	D.,	Christiaens,	J.,	De	Kegel,	B.,	Dierick,	M.,	Boistel,	R.	&	Van	Hoorebeke,	L.	(2014).	Grasping	convergent	evolution	in	syngnathids:	A	unique	tale	of	tails.	Journal	of	Anatomy	224,	710–723.	doi:	10.1111/joa.12181	Northrup,	J.	M.,	Hooten,	M.	B.,	Anderson,	C.	R.	&	Wittemyer,	G.	(2013).	Practical	guidance	on	characterizing	availability	in	resource	selection	functions	under	a	use-availability		 128	design.	Ecology	94,	1456–1463.	doi:	10.1890/12-1688.1	Norton,	S.	F.	(1991).	Capture	success	and	diet	of	cottid	fishes:	the	role	of	predator	morphology	and	attack	kinematics.	Ecology	72,	1807–1819.	Ocken,	A.	E.	J.	&	Ritz,	D.	A.	(1994).	Prey	capture	techniques	of	the	seahorse	Hippocampus	abdominalis	feeding	on	swarming	prey.	Zoology	Department.	University	of	Tasmania.	Öckinger,	E.,	Franzén,	M.,	Rundlöf,	M.	&	Smith,	H.	G.	(2009).	Mobility-dependent	effects	on	species	richness	in	fragmented	landscapes.	Basic	and	Applied	Ecology	10,	573–578.	doi:	10.1016/j.baae.2008.12.002	Oliveira,	F.,	Erzini,	K.	&	Gonçalves,	J.	M.	S.	(2007).	Feeding	habits	of	the	deep-snouted	pipefish	Syngnathus	typhle	in	a	temperate	coastal	lagoon.	Estuarine,	Coastal	and	Shelf	Science	72,	337–347.	doi:	10.1016/j.ecss.2006.11.003	Orav-Kotta,	H.	&	Kotta,	J.	(2004).	Food	and	habitat	choice	of	the	isopod	Idotea	baltica	in	the	northeastern	Baltic	Sea.	Hydrobiologia	514,	79–85.	Orth,	R.	J.	&	Heck,	K.	L.	(1980).	Structural	Components	of	eelgrass	(Zostera	marina)	meadows	in	the	lower	Chesapeake	Bay	-	Fishes.	Estuaries	3,	278–288.	Orth,	R.	J.,	Heck,	K.	L.	&	Tunbridge,	D.	J.	(2002).	Predation	on	seeds	of	the	seagrass	Posidonia	australis	in	Western	Australia.	Marine	Ecology	Progress	Series	244,	81–88.	doi:	10.3354/meps244081	Orth,	R.	J.,	Heck,	K.	L.	&	van	Montfrans,	J.	(1984).	Faunal	communities	in	seagrass	beds:	a	review	of	the	influence	of	plant	structure	and	prey	characteristics	on	predator-prey	relationships.	Ecosystems	7,	339–350.	doi:	10.2307/1351618		 129	Pace,	M.	L.,	Cole,	J.	J.,	Carpenter,	S.	R.	&	Kitchell,	J.	F.	(1999).	Trophic	cascades	revealed	in	diverse	ecosystems.	Trends	in	Ecology	and	Evolution	14,	483–488.	doi:	10.1016/S0169-5347(99)01723-1	Paine,	R.	T.	(1966).	Food	web	complexity	and	species	diversity.	The	American	Naturalist	103,	91–93.	Palmer,	M.	A.,	Allan,	D.	&	Butman,	C.	A.	(1996).	Dispersal	as	a	regional	process	affecting	the	local	dynamics	of	marine	and	stream	benthic	invertebrates.	Trends	in	Ecology	and	Evolution	11,	322–326.	doi:	10.1016/0169-5347(96)10038-0	Payne,	M.	F.	&	Rippingale,	R.	J.	(2000).	Rearing	West	Australian	seahorse,	Hippocampus	subelongatus,	juveniles	on	copepod	nauplii	and	enriched	Artemia.	Aquaculture	188,	353–361.	doi:	10.1016/S0044-8486(00)00349-5	Perante,	N.	C.,	Pajaro,	M.	G.,	Meeuwig,	J.	J.	&	Vincent,	A.	C.	J.	(2002).	Biology	of	a	seahorse	species,	Hippocampus	comes	in	the	central	Philippines.	Journal	of	Fish	Biology	60,	821–837.	doi:	10.1111/j.1095-8649.2002.tb02412.x	Pimm,	S.	L.	(1982).	Food	Webs.	Dordrecht:	Springer	Netherlands.	Polis,	G.	A.	&	Strong,	D.	R.	(1995).	Food	web	complexity	and	community	dynamics	147,	813–846.	doi:	10.1086/285880	Poulos,	D.	E.,	Harasti,	D.,	Gallen,	C.	&	Booth,	D.	J.	(2013).	Biodiversity	value	of	a	geographically	restricted	soft	coral	species	within	a	temperate	estuary.	Aquatic	Conservation:	Marine	and	Freshwater	Ecosystems	23,	838–849.	doi:	10.1002/aqc.2362	Prein,	M.	&	Kunzmann,	A.	(1987).	Structural	organization	of	the	gills	of	pipefish	(Teleostei,		 130	Syngnathidae).	Zoomorphology	107,	161–168.	Preisser,	E.	L.,	Bolnick,	D.	I.	&	Benard,	M.	F.	(2005).	Scared	to	death?	The	effects	of	intimidation	and	consumption	in	predator–prey	interactions.	Ecology	86,	501–509.	doi:	10.1890/04-0719	Pyke,	G.	H.,	Pulliam,	H.	R.	&	Charnov,	E.	L.	(1977).	Optimal	foraging:	a	selective	review	of	theory	and	tests.	The	Quarterly	Review	of	Biology	52,	137–154.	Qin,	G.,	Zhang,	Y.,	Huang,	L.	&	Lin,	Q.	(2014).	Effects	of	water	current	on	swimming	performance,	ventilation	frequency,	and	feeding	behavior	of	young	seahorses	(Hippocampus	erectus).	Journal	of	Experimental	Marine	Biology	and	Ecology	461,	337–343.	doi:	10.1016/j.jembe.2014.09.001	Revell,	L.	J.,	Harmon,	L.	J.	&	Collar,	D.	C.	(2008).	Phylogenetic	signal,	evolutionary	process,	and	rate.	Systematic	Biology	57,	591–601.	doi:	10.1080/10635150802302427	Richardson,	H.	&	Verbeek,	N.	A.	M.	(1986).	Diet	selection	and	optimization	by	northwestern	crows	feeding	on	Japanese	littleneck	clams.	Ecology	67,	1219–1226.	Ries,	L.,	Fletcher,	R.	J.,	Battin,	J.	&	Sisk,	T.	D.	(2004).	Ecological	responses	to	habitat	edges:	mechanisms,	models,	and	variability	explained.	Annual	Review	of	Ecology,	Evolution	and	Systematics	35,	491–522.	doi:	10.1146/annurev.ecolsys.35.112202.130148	Ripley,	J.	L.	&	Foran,	C.	M.	(2007).	Influence	of	estuarine	hypoxia	on	feeding	and	sound	production	by	two	sympatric	pipefish	species	(Syngnathidae).	Marine	Environmental	Research	63,	350–367.	doi:	10.1016/j.marenvres.2006.10.003	Roberts,	C.	M.	&	Hawkins,	J.	P.	(1999).	Extinction	risk	in	the	sea.	Trends	in	Ecology	and		 131	Evolution	14,	241–246.	doi:	10.1016/S0169-5347(98)01584-5	Roos,	G.,	Leysen,	H.,	Van	Wassenbergh,	S.,	Herrel,	A.,	Jacobs,	P.,	Dierick,	M.,	Aerts,	P.	&	Adriaens,	D.	(2009a).	Linking	morphology	and	motion:	a	test	of	a	four-bar	mechanism	in	seahorses.	Physiological	and	Biochemical	Zoology:	PBZ	82,	7–19.	doi:	10.1086/589838	Roos,	G.,	Van	Wassenbergh,	S.,	Herrel,	A.	&	Aerts,	P.	(2009b).	Kinematics	of	suction	feeding	in	the	seahorse	Hippocampus	reidi.	The	Journal	of	Experimental	Biology	212,	3490–8.	doi:	10.1242/jeb.033050	Roos,	G.,	Van	Wassenbergh,	S.,	Aerts,	P.,	Herrel,	A.	&	Adriaens,	D.	(2011).	Effects	of	snout	dimensions	on	the	hydrodynamics	of	suction	feeding	in	juvenile	and	adult	seahorses.	Journal	of	Theoretical	Biology	269,	307–317.	doi:	10.1016/j.jtbi.2010.10.023	Roos,	G.,	Van	Wassenbergh,	S.,	Herrel,	A.,	Adriaens,	D.	&	Aerts,	P.	(2010).	Snout	allometry	in	seahorses:	insights	on	optimisation	of	pivot	feeding	performance	during	ontogeny.	The	Journal	of	Experimental	Biology	213,	2184–93.	doi:	10.1242/jeb.040972	Rosa,	I.	L.,	Oliveira,	T.	P.	R.,	Castro,	A.	L.	C.,	Moraes,	L.	E.,	Xavier,	J.	H.	A.,	Nottingham,	M.	C.,	Dias,	T.	L.	P.,	Bruto-Costa,	L.	V.,	Araújo,	M.	E.,	Birolo,	A.	B.,	Mai,	A.	C.	G.	&	Monteiro-Neto,	C.	(2007).	Population	characteristics,	space	use	and	habitat	associations	of	the	seahorse	Hippocampus	reidi	(Teleostei:	Syngnathidade).	Neotropical	Ichthyology	5,	405–414.	doi:	10.1590/S1679-62252007000300020	Rose,	K.	A.	(2000).	Why	are	quantitative	relationships	between	environmental	quality	and	fish	populations	so	elusive.	Ecological	Applications	10,	367–385.		 132	Ruxton,	G.	D.,	Sherratt,	T.	N.	&	Speed,	M.	P.	(2004).	Avoiding	Attack:	The	Evolutionary	Ecology	of	Crypsis,	Warning	Signals	&	Mimicry.	Oxford,	UK:	Oxford	University	Press.	doi:	10.1093/acprof:oso/9780198528609.001.0001	Ryer,	C.	H.	(1988).	Pipefish	foraging:	effects	of	fish	size,	prey	size	and	altered	habitat	complexity.	Marine	Ecology	Progress	Series	48,	37–45.	doi:	10.3354/meps048037	Ryer,	C.	H.	&	Boehlert,	G.	W.	(1983).	Feeding	chronology,	daily	ration,	and	the	effects	of	temperature	upon	gastric	evacuation	in	the	pipefish,	Syngnathus	fuscus.	Environmental	Biology	of	Fishes	9,	301–306.	doi:	10.1007/BF00692379	Ryer,	C.	H.	&	Orth,	R.	J.	(1987).	Feeding	ecology	of	the	northern	pipefish,	Syngnathus	fuscus,	in	a	seagrass	community	of	the	Lower	Chesapeake	Bay.	Estuaries	10,	330–336.	doi:	10.1007/BF02689864	Sakurai,	I.,	Kaneta,	T.,	Nakayama,	T.,	Fukuda,	H.	&	Kaneko,	T.	(2009).	Food	habits	of	fish	communities	in	a	Sargassum	confusum	bed	off	the	coast	of	Ishikari,	Hokkaido,	Japan.	Nippon	Suisan	Gakkaishi	75,	365–375.	Sancho,	G.,	Petersen,	C.	W.	&	Lobel,	P.	S.	(2000).	Predator-prey	relations	at	a	spawning	aggregation	site	of	coral	reef	fishes.	Marine	Ecology	Progress	Series	203,	275–288.	doi:	10.3354/meps203275	Scharf,	F.	S.,	Juanes,	F.	&	Rountree,	R.	A.	(2000).	Predator	size	-	prey	size	relationships	of	marine	fish	predators:	interspecific	variation	and	effects	of	ontogeny	and	body	size	on	trophic-niche	breadth.	Marine	Ecology	Progress	Series	208,	229–248.	doi:	10.3354/meps208229		 133	Schoener,	T.	W.	(1971).	Theory	of	feeding	strategies.	Annual	Review	of	Ecology	and	Systematics	2,	369–404.	Schroeder,	D.	M.	&	Love,	M.	S.	(2002).	Recreational	fishing	and	marine	fish	populations	in	California.	California	Cooperative	Oceanic	Fisheries	Investigations	43,	182–190.	Seber,	G.	A.	F.	(1986).	A	review	of	estimating	animal	abundance.	Biometrics	42,	267–292.	doi:	10.2307/2531049	Sheng,	J.,	Lin,	Q.,	Chen,	Q.,	Gao,	Y.,	Shen,	L.	&	Lu,	J.	(2006).	Effects	of	food,	temperature	and	light	intensity	on	the	feeding	behavior	of	three-spot	juvenile	seahorses,	Hippocampus	trimaculatus	Leach.	Aquaculture	256,	596–607.	doi:	10.1016/j.aquaculture.2006.02.026	Sih,	A.	(1984).	The	behavioral	response	race	between	predator	and	prey.	The	American	Naturalist	123,	143–150.	doi:	10.1086/284193	Sih,	A.	(1992).	Prey	uncertainty	and	the	balancing	of	antipredator	and	feeding	needs.	The	American	Naturalist	139,	1052–1069.	doi:	10.1086/285372	Sih,	A.,	Crowley,	P.,	McPeek,	M.,	Petranka,	J.	&	Strohmeier,	K.	(1985).	Predation,	competition,	and	prey	communities:	a	review	of	field	experiments.	Annual	Review	of	Ecological	Systems	16,	269–311.	doi:	10.1146/annurev.es.16.110185.001413	Sih,	A.	&	Moore,	R.	D.	(1990).	Interacting	effects	of	predator	and	prey	behavior	in	determining	diets.	In:	Behavioural	Mechanisms	of	Food	Selection	(Huges,	R.	N.,	ed),	pp.	771–795.	Berlin:	Springer.	doi:	10.1007/978-3-642-75118-9_37	Smith,	T.	M.,	Hindell,	J.	S.,	Jenkins,	G.	P.,	Connolly,	R.	M.	&	Keough,	M.	J.	(2011a).	Fine-scale		 134	spatial	and	temporal	variations	in	diets	of	the	pipefish	Stigmatopora	nigra	within	seagrass	patches.	Journal	of	Fish	Biology	78,	1824–1832.	doi:	10.1111/j.1095-8649.2011.02977.x	Smith,	T.	M.,	Hindell,	J.	S.,	Jenkins,	G.	P.,	Connolly,	R.	M.	&	Keough,	M.	J.	(2011b).	Edge	effects	in	patchy	seagrass	landscapes:	the	role	of	predation	in	determining	fish	distribution.	Journal	of	Experimental	Marine	Biology	and	Ecology	399,	8–16.	doi:	10.1016/j.jembe.2011.01.010	Steffan-Dewenter,	I.,	Munzenberg,	U.,	Burger,	C.,	Thies,	C.	&	Tscharntke,	T.	(2002).	Scale-dependent	effects	of	landscape	context	on	three	polinator	guilds.	Ecology	83,	1421–1432.	Steffe,	A.	S.,	Westoby,	M.	&	Bell,	J.	D.	(1989).	Habitat	selection	and	diet	in	two	species	of	pipefish	from	seagrass:	sex	differences.	Marine	Ecology	Progress	Series	55,	23–30.	doi:	10.3354/meps055023	Stoner,	A.	W.	(1980).	The	role	of	seagrass	biomass	in	the	organization	of	benthic	macrofaunal	assemblages.	Bulletin	of	Marine	Science	30,	537–551.	Storero,	L.	P.	&	Gonzalez,	R.	A.	(2008).	Feeding	habits	of	the	seahorse	Hippocampus	patagonicus	in	San	Antonio	Bay	(Patagonia,	Argentina).	Journal	of	the	Marine	Biological	Association	of	the	United	Kingdom	88,	1503–1508.	doi:	10.1017/S0025315408002506	Struthers,	D.	P.,	Danylchuk,	A.	J.,	Wilson,	A.	D.	M.	&	Steven,	J.	(2015).	Action	cameras:	bringing	aquatic	and	fisheries	research	into	view.	Fisheries	40,	502–512.	doi:	10.1080/03632415.2015.1082472		 135	Svensson,	I.	(1988).	Reproductive	costs	in	two	sex-role	reversed	pipefish	species	(Syngnathidae).	The	Journal	of	Animal	Ecology	57,	929–942.	doi:	10.2307/5102	Taskavak,	E.,	Gurkan,	S.,	Sever,	T.	M.,	Akalin,	S.	&	Ozaydin,	O.	(2010).	Gut	contents	and	feeding	habits	of	the	great	pipefish,	Syngnathus	acus	Linnaeus,	1758,	in	Izmir	Bay	(Aegean	Sea,	Turkey).	Zoology	in	the	Middle	East	50,	75–82.	doi:	10.1080/09397140.2010.10638414	Teixeira,	R.	L.	&	Musick,	J.	A.	(1995).	Trophic	ecology	of	two	congeneric	pipefishes	(Syngnathidae)	of	the	lower	York	River,	Virginia.	Environmental	Biology	of	Fishes	43,	295–309.	doi:	10.1007/BF00005862	Teixeira,	R.	L.	&	Musick,	J.	A.	(2001).	Reproduction	and	food	habits	of	the	lined	seahorse,	Hippocampus	erectus	(Teleostei:	Syngnathidae)	of	Chesapeake	Bay,	Virginia.	Revista	Brasileira	de	Biologia	61,	79–90.	doi:	10.1590/S0034-71082001000100011	Thorne-Miller,	B.	(1999).	The	Living	Ocean:	Understanding	and	Protecting	Marine	Biodiversity.	Washington	D.C.:	Island	Press.	Tipton,	K.	&	Bell,	S.	S.	(1988).	Foraging	patterns	of	two	syngnathid	fishes:	importance	of	harpacticoid	copepods.	Marine	Ecology	Progress	Series	47,	31–13.	doi:	10.3354/meps047031	Truong,	S.	K.	&	Nga,	T.	N.	M.	(1995).	Reproduction	of	two	species	of	seahorses,	Hippocampus	histrix	and	H.	trimaculatus,	in	Binh	Thuan	waters.	Bao	Cao	Khoa	Hoc	27,	68.	Uncumusaoglu,	A.	A.,	Gurkan,	S.	&	Taskavak,	E.	(2017).	Seasonally	prey	composition	of		 136	Broad-nosed	pipefish,	Syngnathus	typhle,	distributed	in	the	coasts	of	Aegean	Sea,	Turkey	26,	2673–2677.	Valladares,	S.,	Soto,	D.	X.	&	Planas,	M.	(2017).	Dietary	composition	of	endangered	seahorses	determined	by	stable	isotope	analysis.	Marine	and	Freshwater	Research	68,	831–839.	doi:	http://dx.doi.org/10.1071/MF16013	Van	Den	Wollenberg,	A.	N.	(1977).	Redundancy	analysis	an	alternative	for	canonical	correlation	analysis.	Psychometrika	42,	207–219.	Van	Wassenbergh,	S.,	Dries,	B.	&	Herrel,	A.	(2014).	New	insights	into	muscle	function	during	pivot	feeding	in	seahorses.	PLoS	ONE	9.	doi:	10.1371/journal.pone.0109068	Van	Wassenbergh,	S.,	Leysen,	H.,	Adriaens,	D.	&	Aerts,	P.	(2013).	Mechanics	of	snout	expansion	in	suction-feeding	seahorses:	musculoskeletal	force	transmission.	The	Journal	of	Experimental	Biology	216,	407–17.	doi:	10.1242/jeb.074658	Van	Wassenbergh,	S.,	Roos,	G.,	Aerts,	P.,	Herrel,	A.	&	Adriaens,	D.	(2011a).	Why	the	long	face?	A	comparative	study	of	feeding	kinematics	of	two	pipefishes	with	different	snout	lengths.	Journal	of	Fish	Biology	78,	1786–1798.	doi:	10.1111/j.1095-8649.2011.02991.x	Van	Wassenbergh,	S.,	Roos,	G.	&	Ferry,	L.	(2011b).	An	adaptive	explanation	for	the	horse-like	shape	of	seahorses.	Nature	2,	164.	doi:	10.1038/ncomms1168	Van	Wassenbergh,	S.,	Roos,	G.,	Genbrugge,	A.,	Leysen,	H.,	Aerts,	P.	&	Herrel,	A.	(2009).	Suction	is	kid’s	play:	extremely	fast	suction	in	newborn	seahorses.	Biology	Letters	5.	doi:	10.1098/rsbl.2008.0765		 137	Van	Wassenbergh,	S.,	Strother,	J.	A.,	Flammang,	B.	E.,	Ferry-Graham,	L.	A.	&	Aerts,	P.	(2008).	Extremely	fast	prey	capture	in	pipefish	is	powered	by	elastic	recoil.	Journal	of	the	Royal	Society	Interface	5,	285–296.	doi:	10.1098/rsif.2007.1124	Vander	Zanden,	M.	J.	&	Fetzer,	W.	W.	(2007).	Global	patterns	of	aquatic	food	chain	length.	Oikos	116,	1378–1388.	doi:	10.1111/j.2007.0030-1299.16036.x	Vincent,	A.	C.	J.	(1995).	A	role	for	daily	greetings	in	maintaining	seahorse	pair	bonds.	Animal	Behaviour	49,	258–260.	doi:	10.1016/0003-3472(95)80178-2	Vincent,	A.	C.	J.,	Ahnesjö,	I.,	Berglund,	A.	&	Rosenqvist,	G.	(1992).	Pipefishes	and	seahorses:	are	they	all	sex	role	reversed?	Trends	in	Ecology	and	Evolution	7,	237–241.	doi:	10.1016/0169-5347(92)90052-D	Vincent,	A.	C.	J.,	Evans,	K.	L.	&	Marsden,	A.	D.	(2005).	Home	range	behaviour	of	the	monogamous	Australian	seahorse	Hippocampus	whitei.	Environmental	Biology	of	Fishes	72,	1–12.	doi:	DOI	10.1007/s10641-004-4192-7	Vincent,	A.	C.	J.,	Foster,	S.	J.	&	Koldewey,	H.	J.	(2011).	Conservation	and	management	of	seahorses	and	other	Syngnathidae.	Journal	of	Fish	Biology	78,	1681–1724.	doi:	10.1111/j.1095-8649.2011.03003.x	Vincent,	A.	C.	J.	&	Giles,	B.	G.	(2003).	Correlates	of	reproductive	success	in	a	wild	population	of	Hippocampus	whitei.	Journal	of	Fish	Biology	63,	344–355.	doi:	10.1046/j.1095-8649.2003.00154.x	Vincent,	A.	C.	J.	&	Sadler,	L.	M.	(1995).	Faithful	pair	bonds	in	wild	seahorses,	Hippocampus	whitei.	Animal	Behaviour	50,	1557–1569.	doi:	10.1016/0003-3472(95)80011-5		 138	Warrant,	E.	J.	&	Locket,	N.	A.	(2004).	Vision	in	the	deep	sea.	Biological	Reviews	79,	671–712.	doi:	10.1017/S1464793103006420	Werner,	E.	E.	&	Peacor,	S.	D.	(2003).	A	review	of	trait-mediated	indirect	interactions	in	ecological	communities.	Ecology	84,	1083–1100.	doi:	10.1890/0012-9658(2003)084[1083:AROTII]2.0.CO;2	West,	K.,	Cohen,	A.	&	Baron,	M.	(1991).	Morphology	and	behavior	of	crabs	and	gastropods	from	Lake	Tanganyika,	Africa:	implications	for	lacustrine	predator-prey	evolution.	Evolution	45,	589–607.	West,	R.	J.	(1990).	Depth-related	structural	and	morphological	variations	in	an	Australian	Posidonia	seagrass	bed.	Aquatic	Botany	36,	153–166.	doi:	10.1016/0304-3770(90)90079-Z	White,	G.	C.	(1996).	NOREMARK:	population	estimate	from	mark-resighting	surveys.	Wildlife	Society	Bulletin	24,	50–52.	Whitehead,	P.,	Bauchot,	M.,	Hureau,	J.,	Nielsen,	J.	&	Tortonese,	E.	(1984).	Fishes	of	the	North-Eastern	Atlantic	and	the	Mediterranean	(Vol.	2).	Paris:	UNESCO.	Whittaker,	R.	H.,	Levin,	S.	A.	&	Root,	R.	B.	(1973).	Niche,	habitat,	and	ecotope.	The	American	Naturalist	107,	321–338.	doi:	10.1086/282837	Willis,	T.	J.	&	Anderson,	M.	J.	(2003).	Structure	of	cryptic	reef	fish	assemblages:	relationships	with	habitat	characteristics	and	predator	density.	Marine	Ecology	Progress	Series	257,	209–221.	doi:	10.3354/meps257209	Woods,	C.	M.	C.	(2002).	Natural	diet	of	the	seahorse	Hippocampus	abdominalis.	New	Zealand		 139	Journal	of	Marine	and	Freshwater	Research	36,	655–660.	doi:	10.1080/00288330.2002.9517121	Woods,	C.	M.	C.	(2005).	Evaluation	of	VI-alpha	and	PIT-tagging	of	the	seahorse	Hippocampus	abdominalis.	Aquaculture	International	13,	175–186.	doi:	10.1007/s10499-004-1087-8	Woods,	C.	M.	C.	&	Martin-Smith,	K.	M.	(2004).	Visible	implant	fluorescent	elastomer	tagging	of	the	big-bellied	seahorse,	Hippocampus	abdominalis.	Fisheries	Research	66,	363–371.	doi:	10.1016/S0165-7836(03)00183-8	Wootton,	R.	(2012).	Ecology	of	Teleost	Fishes.	Dordrecht:	Springer	Netherlands.	doi:	10.1007/978-94-009-0829-1	Yasué,	M.,	Nellas,	A.	&	Vincent,	A.	C.	J.	(2012).	Seahorses	helped	drive	creation	of	marine	protected	areas,	so	what	did	these	protected	areas	do	for	the	seahorses?	Environmental	Conservation	39,	183–193.	doi:	10.1017/S0376892911000622	Yip,	M.	Y.,	Lim,	A.	C.	O.,	Chong,	V.	C.,	Lawson,	J.	M.	&	Foster,	S.	J.	(2015).	Food	and	feeding	habits	of	the	seahorses	Hippocampus	spinosissimus	and	Hippocampus	trimaculatus	(Malaysia).	Journal	of	the	Marine	Biological	Association	of	the	United	Kingdom	95,	1033–1040.	doi:	10.1017/S0025315414001660	Zamzow,	J.	P.,	Amsler,	C.	D.,	Mcclintock,	J.	B.	&	Baker,	B.	J.	(2010).	Habitat	choice	and	predator	avoidance	by	Antarctic	amphipods:	the	roles	of	algal	chemistry	and	morphology.	Marine	Ecology	Progress	Series	400,	155–163.	doi:	10.3354/meps08399	Zuur,	A.	F.,	Ieno,	E.	N.	&	Elphick,	C.	S.	(2010).	A	protocol	for	data	exploration	to	avoid		 140	common	statistical	problems.	Methods	in	Ecology	and	Evolution	1,	3–14.	doi:	10.1111/j.2041-210X.2009.00001.x																																									 141	Appendices		Appendix	A:	The	determination	of	seahorse	sex,	maturity,	and	reproductive	status.		 Adult	males	were	identified	by	the	presence	of	a	brood	pouch,	and	older	juvenile	males	based	on	a	dark	oval	region	where	a	brood	pouch	was	developing.	Individuals	without	a	brood	pouch	or	a	dark	oval	region	were	considered	to	have	'unknown	sex'	and	'unknown	reproductive	status'	until	height	at	physical	maturity	(HTM)	was	determined	(see	below).	The	pregnancy	states	of	adult	males	were	assigned	as	follows:	(1)	pouch	tight	and	empty	-	no	recent	pregnancy,	(2)	pouch	round	-	currently	pregnant,	(3)	pouch	rotund	–	male	about	to	release	young,	and	(4)	pouch	flaccid	–	male	released	young	recently	(	Vincent	&	Sadler,	1995;	Perante	et	al.,	2002;	Harasti	et	al.,	2012).	Juvenile	males	were	considered	to	be	physically	immature,	with	a	pregnancy	status	of	0.	Although	having	a	fully	developed	brood	pouch	suggests	physical	maturity	(Morgan	&	Vincent,	2013),	only	males	that	displayed	evidence	of	engaging	in	reproduction	(pregnancy	states	2-4)	were	considered	reproductively	active	(Harasti	et	al.,	2012;	Morgan	&	Vincent,	2013).	Since	female	maturity	can	only	be	determined	by	killing	individuals	and	dissecting	their	ovaries	(Foster	&	Vincent,	2004),	females	were	necessarily	assumed	to	mature	(both	physically	and	reproductively)	at	the	same	height	as	males	(see	next	paragraph).			 The	height	of	physical	and	reproductive	maturity	for	H.	whitei	across	Port	Stephens	was	determined	using	linear	models	fit	to	the	height	of	males,	with	binary	response	variables.	To	model	HTM,	males	were	considered	either	(0)	physically	immature	juveniles	(state	0),	or	(1)	physically	mature	adults	(states	1-4).	To	model	the	height	of	reproductive	maturity	(HTR),	males	were	considered	either	(0)	reproductively	inactive	(states	0-1),	or	(1)	reproductively	active	(states	2-4).	HTM	and	HTR	were	determined	by	the	50%	inflection		 142	points	in	these	models	(King,	2007).	Mean	±	SE	HTM	was	72.3	±	3.7	mm,	and	HTR	was	91.2	±	3.7	mm.	In	subsequent	analyses,	all	seahorses	smaller	than	the	mean	HTM	were	considered	juveniles,	and	those	larger	were	considered	adults.	Adult	seahorses	without	a	pouch	were	considered	to	be	females.	In	the	same	way,	mean	HTR	was	the	cut-off	for	reproductive	maturity.	HTM	and	HTR	models	were	run	in	R	3.3.1	(www.r-project.org).																				 143	Appendix	B:	Tables	to	support	methods	and	results	in	Chapter	2		Table	B1	Seagrass	and	substrate	summary	statistics	for	the	seven	plots.	Mean	and	standard	error	of	the	mean	(SE)	calculated	from	all	quadrats	at	a	specific	plot,	regardless	of	strata.	Using	a	gridded-quadrat,	the	percent	cover	of	P.	australis	and	substrate	was	estimated	by	counting	the	number	of	grid	intersections	each	substrate	type	accounted	for.	Substrate	is	the	percentage	cover	that	is	not	sand	or	rock.	SHG	=	Seahorse	Gardens.		 Depth	(m)	 Seagrass	density	(shoots	0.25m-2)		 Seagrass	height	(cm)	 %	P.	australis	 %	Substrate	Plot	 Mean	 SE	 Mean	 SE	 Mean	 SE	 Mean	 SE	 Mean	 SE	Shoal	Bay	 0.3	 0.1	 68	 25	 32	 2	 74	 8	 88	 4	Little	Beach	 1.0	 0.2	 41	 7	 38	 2	 70	 9	 80	 5	Fly	Point	 1.1	 0.3	 50	 6	 48	 3	 80	 5	 90	 3	SHG	1	 1.7	 0.3	 24	 4	 55	 6	 61	 9	 70	 10	SHG	2	 1.7	 0.2	 26	 4	 48	 7	 67	 8	 79	 4	Pipeline	 2.0	 0.1	 26	 6	 41	 2	 40	 10	 86	 4	Dutchies	 1.5	 0.2	 38	 6	 41	 2	 84	 4	 90	 3		 144	Table	B2	Summary	of	the	model-averaged	statistics	for	the	top	models	predicting	height	within	Little	Beach	for	(a)	all	seahorses,	(b)	all	reproductively	active	(RA)	seahorses,	(c)	females,	(d)	RA	females,	(e)	males,	and	(f)	RA	males.	LL	=	log-likelihood,	AICc	=	corrected	Akaike	information	criterion,	ΔAICc	=	difference	in	model	AICc	with	that	of	the	top	model,	wi	=	Akaike	weight,	df	=	number	of	model	parameters	including	intercepts	and	residuals.	The	following	abbreviations	have	been	made:	DPTH	=	depth,	DENS	=	seagrass	density,	HGHT	=	seagrass	height,	TPT	=	prey	types,	TPI	=	prey	density,	FLNG	=	fouling,	PRED	=	total	predators,	and	TPRED	=	types	of	predators.		Model	and	parameters	included	 LL	 AICc	 ΔAICc	 wi	 df		(a)	Height	-	All	Seahorses	 	 	 	 	 		 	 DPTH,	HGHT,	TPT,	PRED,	TPRED	(global	model)	 -1347.57	 2713.71	 0.00	 0.31	 9		 	 DPTH,	TPT,	PRED,	TPRED	 -1349.37	 2715.19	 1.48	 0.15	 8		 	 DPTH,	HGHT,	TPT,	PRED	 -1349.43	 2715.31	 1.60	 0.14	 8		 	 DPTH,	HGHT,	PRED,	TPRED	 -1349.44	 2715.34	 1.63	 0.14	 8		 	 DPTH,	HGHT,	PRED	 -1351.20	 2716.75	 3.04	 0.07	 7		 	 DPTH,	TPT,	PRED	 -1351.21	 2716.77	 3.06	 0.07	 7		 	 DPTH,	PRED,	TPRED	 -1351.24	 2716.83	 3.12	 0.07	 7		 	 DPTH,	HGHT,	TPT,	TPRED	 -1350.24	 2716.93	 3.22	 0.06	 8		 (b)	Height	-	RA	Seahorses	 	 	 	 	 		 	 DPTH,	HGHT,	TPT,	PRED,	TPRED	(global	model)	 -1128.95	 2276.51	 0.00	 0.27	 9		 	 DPTH,	HGHT,	PRED,	TPRED	 -1130.50	 2277.49	 0.97	 0.17	 8		 	 DPTH,	HGHT,	TPT,	PRED	 -1130.53	 2277.54	 1.03	 0.16	 8		 	 DPTH,	TPT,	PRED,	TPRED	 -1130.84	 2278.16	 1.65	 0.12	 8		 	 DPTH,	HGHT,	PRED	 -1132.02	 2278.42	 1.91	 0.11	 7		 	 DPTH,	PRED,	TPRED	 -1132.39	 2279.15	 2.64	 0.07	 7		 	 DPTH,	TPT,	PRED	 -1132.76	 2279.9	 3.39	 0.05	 7		 	 DPTH,	HGHT,	TPT,	TPRED	 -1131.76	 2280.01	 3.49	 0.05	 8		 (c)	Height	-	Females	 	 	 	 	 		 	 DPTH,	HGHT,	PRED,	TPRED	(global	model)	 -630.13	 1277.18	 0.00	 0.46	 8		 	 DPTH,	HGHT,	PRED	 -631.94	 1278.57	 1.40	 0.23	 7		 	 DPTH,	PRED,	TPRED	 -632.08	 1278.86	 1.68	 0.20	 7		 	 DPTH,	PRED	 -633.78	 1280.08	 2.90	 0.11	 6		 (d)	Height	-	RA	Females	 	 	 	 	 		 	 DPTH,	HGHT,	TPT,	PRED,	TPRED	(global	model)	 -576.81	 1172.81	 0.00	 0.48	 9		 	 DPTH,	TPT,	PRED,	TPRED	 -578.83	 1174.61	 1.80	 0.20	 8		 145		 	 DPTH,	HGHT,	TPT,	PRED	 -578.94	 1174.84	 2.03	 0.18	 8		 	 DPTH,	HGHT,	PRED,	TPRED	 -579.82	 1176.59	 3.78	 0.07	 8		 	 DPTH,	HGHT,	TPT,	TPRED	 -579.86	 1176.66	 3.85	 0.07	 8		 (e)	Height	-	Males	 	 	 	 	 		 	 DPTH,	HGHT,	TPT,	PRED,	TPRED	(global	model)	 -557.14	 1133.59	 0.00	 0.31	 9		 	 DPTH,	HGHT,	TPT,	PRED	 -559.09	 1135.22	 1.63	 0.14	 8		 	 DPTH,	TPT,	PRED,	TPRED	 -559.21	 1135.46	 1.87	 0.12	 8		 	 DPTH,	HGHT,	TPT,	TPRED	 -559.29	 1135.62	 2.03	 0.11	 8		 	 DPTH,	HGHT,	PRED,	TPRED	 -559.45	 1135.94	 2.36	 0.10	 8		 	 DPTH,	TPT,	PRED	 -560.95	 1136.71	 3.12	 0.07	 7		 	 DPTH,	HGHT,	TPT	 -560.99	 1136.78	 3.19	 0.06	 7		 	 DPTH,	TPT,	TPRED	 -561.23	 1137.26	 3.67	 0.05	 7		 	 HGHT,	TPT,	PRED,	TPRED	 -560.28	 1137.59	 4.00	 0.04	 8		 (f)	Height	-	RA	Males	 	 	 	 	 		 	 DPTH,	HGHT,	TPT,	PRED,	TPRED	(global	model)	 -523.69	 1066.71	 0.00	 0.35	 9		 	 DPTH,	HGHT,	TPT,	PRED	 -525.55	 1068.16	 1.45	 0.17	 8		 	 DPTH,	TPT,	PRED,	TPRED	 -525.72	 1068.49	 1.78	 0.14	 8		 	 DPTH,	HGHT,	TPT,	TPRED	 -525.87	 1068.81	 2.10	 0.12	 8		 	 DPTH,	TPT,	PRED	 -527.33	 1069.48	 2.77	 0.09	 7		 	 DPTH,	HGHT,	PRED,	TPRED	 -526.43	 1069.92	 3.21	 0.07	 8			 		 DPTH,	HGHT,	TPT	 -527.86	 1070.53	 3.82	 0.05	 7							 146	Appendix	C:	Tables	to	support	methods	and	results	in	Chapter	3		Table	C1	Summary	of	morphological	characteristics	for	syngnathids	used	in	this	study,	and	the	references	used	to	generate	the	data.	a	Neutens	et	al.	(2014).	b	Data	from	FishBase	(Froese	&	Pauly,	2017)	with	the	following	exceptions:	^	Storero	&	Gonzalez,	(2008),	*	Dawson	(1982).	If	mentioned,	specimens	were	measured	at	the	Australian	Museum	in	Sydney,	Australia.					 		 Body	form	and	tail	traits	a	 		 		 Head	length	(HL)	 Snout	length	(SnL)	 Snout	depth	(SnD)	 References	Genus	 Species	 Body	form	 Caudal	fin	 Prehensile	tail	Caudal	fin:	Body	area	ratio	 Max	StL	(mm)	b	 Ave.	(mm)	 StL:HL	 Ave.	(mm)	 HL:SnL	 Ave.	(mm)	 SnL:SnD	 For	HL,	SnL,	SnD	measurements	 For	caudal	fin:	body	area	measurements	Acentronura		tentaculata	pipefish	 no	 yes	 0.000	 63	 9.9	 6.4	 3.1	 3.2	 1.2	 2.5	 Dawson	(1985)	 -	Anarchopterus	 criniger	pipefish	 yes	 no	 0.025	 100	 9.1	 11.0	 2.7	 3.4	 1.6	 1.7	 Dawson	(1982)	 Whitehead	et	al.	(1984)	Apterygompus	epinnulatus	pipefish	 yes	 no	 -	 30	 2.6	 11.4	 0.7	 3.9	 0.6	 1.2	 Dawson	(1985)	 -	Bryx		 dunckeri	pipefish	 yes	 no	 -	 75	 6.9	 10.9	 2.1	 3.3	 1.2	 1.7	 Dawson	(1982)	 -	Choeroichthys	brachysoma	pipefish	 yes	 no	 -	 70	 14.0	 5.0	 6.7	 2.1	 1.4	 4.8	 Dawson	(1985)	 -	Choeroichthys	 sculptus	pipefish	 yes	 no	 -	 85	 14.2	 6.0	 6.3	 2.3	 2.2	 2.9	 Dawson	(1985)	 -	Corythoichthys	amplexus	pipefish	 yes	 no	 -	 100	 9.8	 10.2	 4.0	 2.5	 0.9	 4.6	 Dawson	(1985)	 -	Corythoichthys	intestinalis	pipefish	 yes	 no	 -	 160	 18.0	 8.9	 9.0	 2.0	 1.3	 6.9	 Dawson	(1985)	 -	Cosmocampus	 elucens	pipefish	 yes	 no	 -	 150	 19.7	 7.6	 9.9	 2.0	 1.5	 6.4	 Dawson	(1982)	 -	Doryrhamphus	excisus	excisus	pipefish	 yes	 no	 0.095	 70	 15.9	 4.4	 7.2	 2.2	 1.3	 5.5	 Dawson	(1985)	 Dawson	(1985)	Doryrhamphus	 janssi	pipefish	 yes	 no	 -	 140	 29.8	 4.7	 18.1	 1.7	 1.8	 10.1	 Dawson	(1985)	 -	Dunckerocampus		 baldwini	pipefish	 yes	 no	 -	 140	 28.3	 5.0	 17.7	 1.6	 1.6	 11.0	 Dawson	(1985)	 -	Dunckerocampus		chapmani	pipefish	 yes	 no	 -	 85	 22.7	 3.8	 13.3	 1.7	 1.7	 8.1	 Dawson	(1985)	 -	Dunckerocampus		dactyliophorus	pipefish	 yes	 no	 0.054	 190	 45.8	 4.2	 28.6	 1.6	 1.8	 15.7	 Dawson	(1985)	 Dawson	(1985)		 147	Dunckerocampus		pessuliferus	pipefish	 yes	 no	 -	 160	 42.1	 3.8	 28.1	 1.5	 2.7	 10.5	 Dawson	(1985)	 -	Festucalex	 cinctus	pipefish	 yes	 no	 -	 130	 14.9	 8.7	 7.0	 2.2	 1.6	 4.5	 Dawson	(1985)	 -	Festucalex	 scalaris	pipefish	 yes	 no	 -	 180	 19.5	 9.3	 8.5	 2.3	 2.4	 3.6	 Dawson	(1985)	 -	Filicampus	 tigris	pipefish	 yes	 no	 0.024	 296	 29.6	 10.0	 12.9	 2.3	 2.3	 5.6	 Dawson	(1985)	 Dawson	(1985)	Halicampus	 brocki	pipefish	 yes	 no	 -	 120	 11.5	 10.5	 4.3	 2.7	 1.2	 3.5	 Dawson	(1985)	 -	Halicampus	 dunckeri	pipefish	 yes	 no	 -	 150	 12.3	 12.2	 3.3	 3.7	 1.4	 2.5	 Dawson	(1985)	 -	Halicampus	macrorhynchus	pipefish	 yes	 no	 -	 180	 36.7	 4.9	 22.3	 1.7	 1.8	 12.6	 Dawson	(1985)	 -	Halicampus	 nitidus	pipefish	 yes	 no	 -	 73	 7.7	 9.5	 2.3	 3.3	 1.0	 2.3	 Dawson	(1985)	 -	Heraldia	nocturna	pipefish	 yes	 no	 -	 92	 18.4	 5.0	 7.8	 2.4	 2.4	 3.3	 Dawson	(1985)	 -	Hippichthys	heptagonus	pipefish	 yes	 no	 -	 150	 14.1	 10.7	 6.0	 2.4	 1.7	 3.5	 Dawson	(1985)	 -	Hippichthys	penicillus	pipefish	 yes	 no	 -	 180	 26.5	 6.8	 14.3	 1.9	 2.6	 5.6	 Dawson	(1985)	 -	Hippichthys		cyanospilos	pipefish	 yes	 no	 0.038	 160	 18.7	 8.6	 8.7	 2.2	 2.5	 3.5	 Measured	from	specimens	(N	=	22)	 Kuiter	(2000)	Hippocampus	abdominalis	seahorse	 no	 yes	 0.000	 350	 60.2	 5.8	 23.0	 2.6	 5.9	 3.9	 Lourie,	unpublished	data	 -	Hippocampus	breviceps	seahorse	 no	 yes	 0.000	 150	 31.1	 4.8	 10.0	 3.1	 3.7	 2.7	 Lourie,	unpublished	data	 -	Hippocampus	coronatus	seahorse	 no	 yes	 0.000	 108	 21.3	 5.1	 8.5	 2.5	 2.5	 3.4	 Lourie,	unpublished	data	 -	Hippocampus	 erectus	seahorse	 no	 yes	 0.000	 190	 40.4	 4.7	 14.8	 2.7	 5.0	 2.9	 Lourie,	unpublished	data	 -	Hippocampus	guttulatus	seahorse	 no	 yes	 0.000	 215	 43.7	 4.9	 16.9	 2.6	 4.8	 3.5	 Lourie,	unpublished	data	 -	Hippocampus	hippocampus	seahorse	 no	 yes	 0.000	 150	 27.5	 5.5	 9.2	 3.0	 3.7	 2.5	 Lourie,	unpublished	data	 -	Hippocampus	mohnikei	seahorse	 no	 yes	 0.000	 80	 14.1	 5.7	 3.8	 3.7	 2.3	 1.6	 Lourie	et	al.	(1999)	 -	Hippocampus	patagonicus	seahorse	 no	 yes	 0.000	 162^	 34.6	 4.7	 11.3	 3.1	 4.2	 2.7	 Lourie,	unpublished	data	 -	Hippocampus	 reidi	seahorse	 no	 yes	 0.000	 175	 40.2	 4.4	 18.4	 2.2	 4.5	 4.1	 Lourie,	unpublished	data	 -	Hippocampus	spinosissimus	seahorse	 no	 yes	 0.000	 172	 33.9	 5.1	 15.1	 2.2	 3.7	 4.1	 Lourie	et	al.	(1999)	 -	Hippoca subelon seaho no	 yes	 0.000	 200	 47.5	 4.2	 22.2	 2.1	 4.0	 5.5	 Lourie,	 -		 148	mpus	 gatus	 rse	 unpublished	data	Hippocampus	trimaculatus	seahorse	 no	 yes	 0.000	 220	 40.5	 5.4	 18.0	 2.2	 3.7	 4.9	 Lourie	et	al.	(1999)	 -	Hippocampus	 whitei	seahorse	 no	 yes	 0.000	 130	 32.8	 4.0	 14.5	 2.3	 3.0	 4.9	 Lourie,	unpublished	data	 -	Hippocampus	 zosterae	seahorse	 no	 yes	 0.000	 50	 13.9	 3.6	 3.2	 4.3	 1.6	 2.0	 Lourie,	unpublished	data	 -	Hippocampus		 kuda	seahorse	 no	 yes	 0.000	 140	 26.9	 5.2	 11.6	 2.3	 3.1	 3.8	 Lourie	et	al.	(1999)	 -	Histiogamphelus	 briggsii	pipefish	 yes	 no	 -	 225	 24.9	 9.1	 11.3	 2.2	 5.1	 2.2	 Dawson	(1985)	 -	Histiogamphelus	 cristatus	pipefish	 yes	 no	 0.055	 265	 29.3	 9.1	 13.3	 2.2	 6.0	 2.2	 Dawson	(1985)	 Kuiter	(2000)	Hypselognathus	rostratus	pipefish	 yes	 no	 -	 305	 50.4	 6.1	 34.8	 1.5	 3.6	 9.7	 Dawson	(1985)	 -	Kaupus	 costatus	pipefish	 yes	 no	 -	 129	 14.7	 8.8	 5.2	 2.8	 2.0	 2.7	 Dawson	(1985)	 -	Leptoichthys	fistularius	pipefish	 yes	 no	 -	 630	 132.6	 4.8	 98.2	 1.4	 5.2	 19.0	 Dawson	(1985)	 -	Lissocampus	 caudalis	pipefish	 yes	 yes	 0.009	 100	 7.5	 13.4	 2.3	 3.2	 1.7	 1.4	 Dawson	(1985)	 Dawson	(1985)	Lissocampus	 runa	pipefish	 yes	 no	 -	 94	 7.4	 12.8	 2.4	 3.1	 1.5	 1.6	 Dawson	(1985)	 -	Maroubra	perserrata	pipefish	 yes	 no	 -	 72	 11.7	 6.2	 5.4	 2.2	 1.1	 5.1	 Dawson	(1985)	 -	Micrognathus	andersonii	pipefish	 yes	 no	 -	 85	 9.2	 9.3	 3.0	 3.1	 1.4	 2.2	 Dawson	(1985)	 -	Micrognathus	 crinitus	pipefish	 yes	 no	 -	 150	 14.6	 10.3	 4.7	 3.1	 2.1	 2.2	 Dawson	(1982)	 -	Micrognathus	 natans	pipefish	 yes	 no	 -	 60	 8.2	 7.4	 3.5	 2.4	 1.0	 3.5	 Dawson	(1985)	 -	Microphis	 deocata	pipefish	 yes	 no	 -	 150	 20.1	 7.5	 11.8	 1.7	 1.6	 7.2	 Dawson	(1985)	 -	Microphis	ocellatus	pipefish	 yes	 no	 -	 125	 13.6	 9.2	 5.9	 2.3	 1.4	 4.2	 Dawson	(1985)	 -	Mitotichthys	meraculus	pipefish	 yes	 no	 0.009	 222	 28.1	 7.9	 14.8	 1.9	 3.1	 4.8	 Dawson	(1985)	 Web	photo	Mitotichthys	semistriatus	pipefish	 yes	 no	 0.007	 268	 38.6	 7.0	 22.0	 1.8	 2.6	 8.5	 Dawson	(1985)	 Web	sketch	Nannocampus	 pictus	pipefish	 yes	 no	 -	 100	 8.2	 12.2	 2.7	 3.0	 1.6	 1.7	 Dawson	(1985)	 -	Nerophis	lumbriciformis	pipefish	 no	 yes	 0.000	 170	 14.0	 12.1	 5.1	 2.7	 2.3	 2.2	 Measured	from	specimens	(N	=	9)	 -	Nerophis	 ophidion	pipefish	 no	 yes	 0.000	 300	 17.9	 16.8	 7.3	 2.5	 2.5	 3.0	 Measured	from	specimens	(N	=	6)	 -		 149	Oostethus	 jagorii	pipefish	 yes	 no	 -	 160	 23.2	 6.9	 11.0	 2.1	 1.9	 5.7	 Dawson	(1985)	 -	Phoxocampus	diacanthus	pipefish	 yes	 no	 -	 87	 12.2	 7.2	 5.1	 2.4	 1.9	 2.7	 Dawson	(1985)	 -	Phycodurus	 eques	seadragon	 no	 no	 -	 350	 76.1	 4.6	 46.1	 1.7	 7.0	 6.6	 Dawson	(1985)	 -	Phyllopteryx	taeniolatus	seadragon	 no	 no	 0.000	 460	 96.8	 4.8	 62.5	 1.6	 6.1	 10.2	 Dawson	(1985)	 -	Pugnaso		curtirostris	pipefish	 yes	 no	 0.007	 182	 17.1	 10.7	 6.2	 2.8	 2.4	 2.6	 Dawson	(1985)	 Dawson	(1985)	Siokunichthys	nigrolineatus	pipefish	 yes	 no	 -	 80	 5.1	 15.6	 1.4	 3.7	 1.3	 1.1	 Dawson	(1985)	 -	Solegnathus	hardwickii	pipefish	 no	 yes	 0.000	 400	 56.3	 7.1	 33.1	 1.7	 4.1	 8.1	 Dawson	(1985)	 -	Solegnathus	spinosissimus	pipefish	 no	 yes	 0.000	 490	 79.7	 6.2	 46.9	 1.7	 6.4	 7.4	 Dawson	(1985)	 -	Stigmatopora	 argus	pipefish	 no	 yes	 0.000	 254	 42.4	 6.0	 28.1	 1.5	 2.0	 14.1	 Measured	from	specimens	(N	=	53)	 -	Stigmatopora	 nigra	pipefish	 no	 yes	 0.000	 162	 26.4	 6.1	 16.5	 1.6	 1.4	 12.0	 Measured	from	specimens	(N	=	54)	 -	Stipecampus		 cristatus	pipefish	 yes	 no	 -	 220	 16.2	 13.6	 4.8	 3.4	 3.3	 1.5	 Dawson	(1985)	 -	Syngnathoides	biaculeatus	pipefish	 no	 yes	 0.000	 290	 55.0	 5.3	 33.4	 1.6	 5.1	 6.5	 Measured	from	specimens	(N	=	12)	 -	Syngnathus	 abaster	pipefish	 yes	 no	 0.026	 210	 29.5	 7.1	 14.9	 2.0	 3.3	 4.5	 Measured	from	specimens	(N	=	4)	 Web	photo	Syngnathus	 acus	pipefish	 yes	 no	 0.040	 500	 61.3	 8.2	 31.5	 2.0	 5.2	 6.0	 Dawson	(1985)	 Dawson	(1985)	Syngnathus	californiensis	pipefish	 yes	 no	 -	 500	 66.2	 7.6	 37.8	 1.8	 4.9	 7.7	 Dawson	(1982)	 -	Syngnathus	 floridae	pipefish	 yes	 no	 0.023	 250	 38.5	 6.5	 22.6	 1.7	 3.1	 7.2	 Dawson	(1982)	 Web	photo	Syngnathus	 folletti	pipefish	 yes	 no	 0.029	 200.5*	 21.2	 9.5	 10.9	 2.0	 1.8	 6.2	 Dawson	(1982)	 Web	sketch	Syngnathus	 fuscus	pipefish	 yes	 no	 0.043	 330	 39.8	 8.3	 18.9	 2.1	 3.7	 5.1	 Dawson	(1982)	 Web	photo	Syngnathus	louisianae	pipefish	 yes	 no	 0.034	 380	 54.3	 7.0	 31.9	 1.7	 3.1	 10.2	 Dawson	(1982)	 Web	photo	Syngnathus	 schlegeli	pipefish	 yes	 no	 0.039	 300	 36.6	 8.2	 20.3	 1.8	 2.4	 8.5	 Dawson	(1985)	 Web	photo	Syngnathus	 scovelli	pipefish	 yes	 no	 0.043	 183	 24.1	 7.6	 10.5	 2.3	 2.8	 3.7	 Dawson	(1982)	 Web	photo	Syngnathus	taenionotus	pipefish	 yes	 no	 0.042	 190	 26.7	 7.1	 13.0	 2.1	 2.2	 5.9	 Measured	from	photo	(N	=	1)	 Web	photo	Syngnath typhle	 pipefi yes	 no	 0.032	 350	 59.3	 5.9	 34.4	 1.7	 6.3	 5.4	 Measured	from	 Web	photo		 150	us	 sh	 specimens	(N	=	9)	Trachyrhamphus	bicoarctatus	pipefish	 yes	 no	 -	 400	 34.9	 11.5	 20.0	 1.8	 2.4	 8.2	 Dawson	(1985)	 -	Urocampus	carinirostris	pipefish	 yes	 no	 0.001	 100	 8.8	 11.4	 2.7	 3.2	 1.8	 1.6	 Dawson	(1985)	 Dawson	(1985)	Vanacampus	margaritifer	pipefish	 yes	 no	 -	 159	 19.2	 8.3	 9.3	 2.1	 1.5	 6.2	 Dawson	(1985)	 -	Vanacampus	 phillipi	pipefish	 yes	 no	 0.006	 184	 22.4	 8.2	 10.1	 2.2	 2.4	 4.2	 Measured	from	specimens	(N	=	50)	 Web	photo	Vanacampus	 vercoi	pipefish	 yes	 no	 -	 105	 10.4	 10.1	 3.9	 2.7	 1.3	 3.0	 Dawson	(1985)	 -	Vanacampus		poecilolaemus	pipefish	 yes	 no	 0.007	 261	 35.5	 7.4	 18.7	 1.9	 2.4	 7.7	 Dawson	(1985)	 Web	photo																												 151		Table	C2	Summary	of	syngnathid	feeding	kinematics	in	the	literature.	Time	=	0	was	calculated	as	being	one	frame	before	the	onset	of	hyoid	movement	for	all	studies	unless	species	name	is	denoted	by	^,	which	indicates	time	=	0	was	calculated	as	being	one	frame	before	onset	of	head	rotation.	*Measurements	made	on	one	individual.					 Species	 Doryrhamphus	dactyliophorus	 Doryrhamphus	melanopleura	 Entelurus	aequoreus	 Hippocampus	erectus^	 Hippocampus	reidi	 Hippocampus	reidi		 Age	 	 	 	 	 	 0-3	days		 No.	of	repetitions	 20	 20	 	 25	 14	 			 Individuals	 2	 2	 		 4	 5	 		Approach	 	 	 	 	 	 	 		 Mean	approach	speed	(successful	strike;	mm/s)	 	 	 	 	 	 		 Mean	approach	speed	(unsuccessful	strike;	mm/s)	 	 	 	 	 	 	Onset	of	activity	 	 	 	 	 	 		 Time	to	start	of	epaxial	muscle	activity	(ms)	 	 	 	 	 	 		 Time	to	start	of	hyoid	rotation	(ms)	 	 	 	 	 0.5	 		 Time	to	start	of	head	rotation	(ms)	 	 	 	 	 0.64	 		 Time	to	start	of	mouth	opening	(ms)	 	 	 	 	 0.75	 	Intial	prey	orientation	 	 	 	 	 	 		 Initial	prey	distance	(mm)	 2.46	 1.835	 	 7.8*	 6.71	 		 Mean	initial	prey	distance	(successful	strike;	mm)	 	 	 	 	 	 		 Mean	initial	prey	distance	(unsuccessful	strike;	mm)	 	 	 	 	 	 		 Initial	prey	angle	(deg)	 117.425	 113.2	 	 	 	 	Hyoid	rotation	 	 	 	 	 	 		 Total	hyoid	rotation	(deg)	 	 	 	 	 68.45	 96		 Total	hyoid	rotation	time	(ms)	 	 	 	 4.7	 17.29	 4.5		 Max.	hyoid	rotation	velocity	(deg/s)	 	 	 	 	 2985	 		 152	Head	rotation	 	 	 	 	 	 		 Total	head	rotation	(deg)	 9.65	 9.8	 	 29.1	 31.1	 42		 Total	head	rotation	time	(ms)	 6.65	 6.5	 	 6.5	 18.46	 2.5		 Mean	head	velocity	(deg/s)	 	 	 	 	 	 		 Max.	head	velocity	(deg/s)	 4250	 4700	 	 	 13880	 		 Prey	distance	after	head	rotation	(mm)	 1.33	 0.595	 	 	 	 		 Prey	angle	after	head	rotation	(deg)	 177.35	 160.55	 	 	 	 	Snout	and	mouth	movement	 	 	 	 	 	 		 Total	snout	movement	(mm)	 2.45	 1.685	 	 	 	 		 Total	snout	rotation	(deg)	 	 	 	 	 	 		 Total	snout	movement	time	(ms)	 9.75	 9.35	 	 	 	 		 Max.	linear	snout	velocity	(m/s)	 0.885	 0.705	 	 	 	 		 Max.	gape	opening	(mm)	 	 	 	 2.7*	 2.7	 		 Time	to	max.	mouth	opening	(ms)	 	 	 	 4.9	 3.5	 	Body	rotation	 	 	 	 	 	 		 Total	body	rotation	(deg)	 3.75	 3.65	 	 	 	 		 Total	body	rotation	time	(ms)	 4.175	 4.45	 	 	 	 		 Max.	body	velocity	(deg/s)	 2000	 1925	 	 	 	 	Prey	capture	 	 	 	 	 	 	 		 Total	prey	capture	time	(ms)	 5.5	 3.25	 ~5	ms	 5.8	 5.5	 		 Max.	prey	velocity	(m/s)	 0.195	 0.13	 	 	 0.27	 		 Time	to	max.	prey	velocity	(ms)	 4	 2.125	 	 	 5.1	 		 	 	 	 	 	 	 			 Reference	 Van	Wassenbergh	et	al.	(2011a)	 Van	Wassenbergh	et	al.	(2011a)	 Muller	&	Osse	(1984)	 Bergert	&	Wainwright	(1997)	 Roos	et	al.	(2009a)	 Van	Wassenbergh	et	al.	(2009)										 153	Table	C2	(Continued)		 		 Species	 Hippocampus	reidi	 Hippocampus	reidi	 Hippocampus	zostera		 Age	 <1	week	 1	week	 2	weeks	 3	weeks		 Adult	 	 		 No.	of	repetitions	 10	 13	 13	 14	 12	 10	 15			 Individuals	 		 		 		 		 3	 2	 6	Approach	 	 	 	 	 	 	 	 		 Mean	approach	speed	(successful	strike;	mm/s)	 	 	 	 	 	 	 8.4		 Mean	approach	speed	(unsuccessful	strike;	mm/s)	 	 	 	 	 	 	 14.1	Onset	of	activity	 	 	 	 	 	 	 		 Time	to	start	of	epaxial	muscle	activity	(ms)	 	 	 	 	 	 	 		 Time	to	start	of	hyoid	rotation	(ms)	 	 	 	 	 	 	 		 Time	to	start	of	head	rotation	(ms)	 	 	 	 	 	 0.4	 		 Time	to	start	of	mouth	opening	(ms)	 	 	 	 	 	 1.6	 	Intial	prey	orientation	 	 	 	 	 	 	 		 Initial	prey	distance	(mm)	 	 	 	 	 	 	 		 Mean	initial	prey	distance	(successful	strike;	mm)	 	 	 	 	 	 	 0.89		 Mean	initial	prey	distance	(unsuccessful	strike;	mm)	 	 	 	 	 	 	 1.12		 Initial	prey	angle	(deg)	 	 	 	 	 	 	 	Hyoid	rotation	 	 	 	 	 	 	 		 Total	hyoid	rotation	(deg)	 	 	 	 	 	 116	 		 Total	hyoid	rotation	time	(ms)	 	 	 	 	 	 20.25	 		 Max.	hyoid	rotation	velocity	(deg/s)	 	 	 	 	 	 	 	Head	rotation	 	 	 	 	 	 	 		 Total	head	rotation	(deg)	 44	 43	 37.7	 33.3	 25	 26.3	 		 Total	head	rotation	time	(ms)	 3.3	 3.11	 3.36	 3.78	 6.4	 5.45	 		 Mean	head	velocity	(deg/s)	 	 	 	 	 	 	 		 Max.	head	velocity	(deg/s)	 32500	 33000	 26500	 21900	 15300	 	 		 Prey	distance	after	head	rotation	(mm)	 	 	 	 	 	 	 		 Prey	angle	after	head	rotation	(deg)	 	 	 	 	 	 	 124.3	Snout	and	mouth	movement	 	 	 	 	 	 	 		 Total	snout	movement	(mm)	 	 	 	 	 	 	 		 Total	snout	rotation	(deg)	 	 	 	 	 	 	 		 154		 Total	snout	movement	time	(ms)	 	 	 	 	 	 	 		 Max.	linear	snout	velocity	(m/s)	 	 	 	 	 	 	 		 Max.	gape	opening	(mm)	 	 	 	 	 	 4.45	 		 Time	to	max.	mouth	opening	(ms)	 	 	 	 	 	 3.6	 	Body	rotation	 	 	 	 	 	 	 		 Total	body	rotation	(deg)	 	 	 	 	 	 	 		 Total	body	rotation	time	(ms)	 	 	 	 	 	 	 		 Max.	body	velocity	(deg/s)	 	 	 	 	 	 	 	Prey	capture	 	 	 	 	 	 	 	 		 Total	prey	capture	time	(ms)	 	 	 	 	 	 5.75	 <1	ms		 Max.	prey	velocity	(m/s)	 	 	 	 	 	 	 		 Time	to	max.	prey	velocity	(ms)	 	 	 	 	 	 	 		 	 	 	 	 	 	 	 			 Reference	 Roos	et	al.	(2010)	 Roos	et	al.	(2009b)	 Gemmell	et	al.	(2013)																								 155	Table	C2	(Continued)		 		 Species	 Sentriscus	scutatus	 Syngnathus	acus	 Syngnathus	floridae^	 Syngnathus	leptorhynchus	 Syngnathus	leptorhynchus^		 Age	 	 Adult	 Juvenile	 	 	 		 No.	of	repetitions	 3	 2	 1	 7	 19	 27			 Individuals	 1	 1	 1	 3	 2	 9	Approach	 	 	 	 	 	 	 		 Mean	approach	speed	(successful	strike;	mm/s)	 	 	 	 	 	 		 Mean	approach	speed	(unsuccessful	strike;	mm/s)	 	 	 	 	 	 	Onset	of	activity	 	 	 	 	 	 		 Time	to	start	of	epaxial	muscle	activity	(ms)	 	 	 	 	 271	 		 Time	to	start	of	hyoid	rotation	(ms)	 	 	 	 	 	 		 Time	to	start	of	head	rotation	(ms)	 	 	 	 	 	 		 Time	to	start	of	mouth	opening	(ms)	 	 	 	 	 	 	Intial	prey	orientation	 	 	 	 	 	 		 Initial	prey	distance	(mm)	 	 	 	 9.3*	 	 7.9		 Mean	initial	prey	distance	(successful	strike;	mm)	 	 	 	 	 	 		 Mean	initial	prey	distance	(unsuccessful	strike;	mm)	 	 	 	 	 	 		 Initial	prey	angle	(deg)	 	 	 	 	 	 	Hyoid	rotation	 	 	 	 	 	 		 Total	hyoid	rotation	(deg)	 	 ~128	 	 	 	 		 Total	hyoid	rotation	time	(ms)	 	 ~8.8	ms	 	 6.1	 	 16		 Max.	hyoid	rotation	velocity	(deg/s)	 	 	 	 	 	 	Head	rotation	 	 	 	 	 	 		 Total	head	rotation	(deg)	 ~9.3	 ~30.5	 	 29.2	 19.2	 19.2		 Total	head	rotation	time	(ms)	 ~8	 ~25.5	 	 7.5	 5.1	 17.4		 Mean	head	velocity	(deg/s)	 2005	 	 	 	 	 		 Max.	head	velocity	(deg/s)	 	 4010	 	 	 8300	 		 Prey	distance	after	head	rotation	(mm)	 	 	 	 	 	 		 156		 Prey	angle	after	head	rotation	(deg)	 	 	 	 	 	 	Snout	and	mouth	movement	 	 	 	 	 	 		 Total	snout	movement	(mm)	 	 	 	 	 	 		 Total	snout	rotation	(deg)	 	 	 	 	 	 10.7		 Total	snout	movement	time	(ms)	 	 	 	 	 	 14.8		 Max.	linear	snout	velocity	(m/s)	 	 	 	 	 	 		 Max.	gape	opening	(mm)	 	 	 	 3.2*	 	 3.6		 Time	to	max.	mouth	opening	(ms)	 	 	 	 6.8	 	 	Body	rotation	 	 	 	 	 	 		 Total	body	rotation	(deg)	 	 	 	 	 	 		 Total	body	rotation	time	(ms)	 	 	 	 	 	 		 Max.	body	velocity	(deg/s)	 	 	 	 	 	 	Preycapture	 	 	 	 	 	 	 		 Total	prey	capture	time	(ms)	 ~5.4	 ~7.4	 ~6.1	 7.9	 	 4.2		 Max.	prey	velocity	(m/s)	 	 	 	 	 	 		 Time	to	max.	prey	velocity	(ms)	 	 	 	 	 	 		 	 	 	 	 	 	 			 Reference	 de	Lussanet	&	Muller	(2007)	 de	Lussanet	&	Muller	(2007)	 		 Bergert	&	Wainwright	(1997)	 Van	Wassenbergh	et	al.	(2008)	 Flammang	et	al.	(2009)		

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0355229/manifest

Comment

Related Items