Environmental correlates of fine-scale juvenile steelhead trout (Oncorhynchus mykiss) habitat use and movement patterns in an intermittent estuary during drought

Abstract

We used acoustic telemetry and environmental monitoring to elucidate preferred microhabitats of juvenile steelhead trout (Oncorhynchus mykiss) in a Central California intermittent estuary (IE) during historic drought. We collected over half a million fish locations in the Pescadero IE (San Mateo County, CA) across 15 weeks during an extended sandbar-closed period which permitted quantification of fine scale habitat use and movement patterns. Tagged juvenile steelhead expressed strong site fidelity, especially at night when core habitat area - defined as the 50% probability of being present in an area - contracted by over one order of magnitude. The rate of movement was slow overall (~0.4 to 0.6 lengths·s−1) and remained at baseline levels at night (~40 mm∙s−1). The daytime rate of movement generally tracked solar radiation levels and appeared to be moderated by water temperatures. Spikes in the rate of movement occurred during crepuscular periods and the maximum hourly rate of movement (138 mm∙s−1) was observed during the early study period from 10:00 to 11:00 when water temperatures were physiologically optimal (17–18 °C). Water quality worsened upstream when water temperatures exceeded 18 °C and dissolved oxygen concentrations declined below 7.0 mg·L−1. Fish tag detections at stationary receivers in the upper estuary declined linearly with deteriorating water quality conditions. Qualitative analysis of juvenile steelhead habitat utilization indicated a strong preference for two microhabitat features in the estuary during the study; both were shallow (~1.5 m), wind-protected, and possessed cover and sandy substrates that occurred within the fresh or near fresh epilimnion where lagoon water quality was best and benthic prey was likely most abundant. Upstream movement occurred in late fall for over half of the tagged cohort, which likely enhances population resiliency by allowing these fish to escape lethal water quality conditions coincident with the transition from closed to open estuary in late fall. Climate projections for California’s Central Coast predict an increase in extreme dry events and the information presented here can help natural resource managers prepare for the future, such as the critical need to promote development of a sufficiently oxygenated epilimnion during extended sandbar-closed ecosystem states.

Full text

1 23
Environmental Biology of Fishes
ISSN 0378-1909
Volume 103
Number 5
Environ Biol Fish (2020) 103:509-529
DOI 10.1007/s10641-020-00971-y
Environmental correlates of fine-scale
juvenile steelhead trout (Oncorhynchus
mykiss) habitat use and movement
patterns in an intermittent estuary during
drought
Eric R.Huber & Stephanie M.Carlson

1 23
Your article is protected by copyright and
all rights are held exclusively by Springer
Nature B.V.. This e-offprint is for personal
use only and shall not be self-archived
in electronic repositories. If you wish to
self-archive your article, please use the
accepted manuscript version for posting on
your own website. You may further deposit
the accepted manuscript version in any
repository, provided it is only made publicly
available 12 months after official publication
or later and provided acknowledgement is
given to the original source of publication
and a link is inserted to the published article
on Springer's website. The link must be
accompanied by the following text: "The final
publication is available at link.springer.com”.

Environmental correlates of fine-scale juvenile steelhead
trout (Oncorhynchus mykiss) habitat use and movement
patterns in an intermittent estuary during drought
Eric R. Huber &Stephanie M. Carlson
Received: 5 April 2019 /Accepted: 7 April 2020
#Springer Nature B.V. 2020
Abstract We used acoustic telemetry and environmental
monitoring to elucidate preferred microhabitats of juvenile
steelhead trout (Oncorhynchus mykiss)inaCentralCali-
fornia intermittent estuary (IE) during historic drought. We
collected over half a million fish locations in the Pescadero
IE (San Mateo County, CA) across 15 weeks during an
extended sandbar-closed period which permitted quantifi-
cation of fine scale habitat use and movement patterns.
Tagged juvenile steelhead expressed strong site fidelity,
especially at night when core habitat area - defined as the
50% probability of being present in an area - contracted by
over one order of magnitude. The rate of movement was
slow overall (~0.4 to 0.6 lengths·s
−1
) and remained at
baseline levels at night (~40 mm∙s
−1
).Thedaytimerate
of movement generally tracked solar radiation levels and
appeared to be moderated by water temperatures. Spikes in
the rate of movement occurred during crepuscular periods
and the maximum hourly rate of movement (138 mm∙s
−1
)
was observed during the early study period from 10:00 to
11:00 when water temperatures were physiologically opti-
mal (17–18 °C). Water quality worsened upstream when
water temperatures exceeded 18 °C and dissolved oxygen
concentrations declined below 7.0 mg·L
−1
. Fish tag detec-
tions at stationary receivers in the upper estuary declined
linearly with deteriorating water quality conditions. Qual-
itative analysis of juvenile steelhead habitat utilization
indicated a strong preference for two microhabitat features
in the estuary during the study; both were shallow
(~1.5 m), wind-protected, and possessed cover and sandy
substrates that occurred within the fresh or near fresh
epilimnion where lagoon water quality was best and ben-
thic prey was likely most abundant. Upstream movement
occurred in late fall for over half of the tagged cohort,
which likely enhances population resiliency by allowing
these fish to escape lethal water quality conditions coinci-
dent with the transition from closed to open estuary in late
fall. Climate projections for California’s Central Coast
predict an increase in extreme dry events and the informa-
tion presented here can help natural resource managers
prepare for the future, such as the critical need to promote
development of a sufficiently oxygenated epilimnion dur-
ing extended sandbar-closed ecosystem states.
Keywords Acoustic telemetry .VEMCO positioning
system (VPS) .Mobile tracking .Microhabitat mapping .
Diel effects .Water quality
Introduction
Pacific salmon have immense economic, cultural, and
social value and over half of the management units
within the United States (Evolutionarily Significant
https://doi.org/10.1007/s10641-020-00971-y
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s10641-020-00971-y)contains
supplementary material, which is available to authorized users.
E. R. Huber (*):S. M. Carlson
University of California, Berkeley, Department of Environmental
Science, Policy, and Management, University of California,
Berkeley, CA 94720, USA
e-mail: eric.r.huber@gmail.com
Environ Biol Fish (2020) 103:509–529
/Published online: 29 May 2020
Author's personal copy

Units, ESUs) are listed as threatened or endangered
under the US Endangered Species Act (USESA)
(Ruckelshaus et al. 2002). Abundances of the threatened
Central California Coastal (CCC) steelhead (Oncorhyn-
chus mykiss) ESU have declined by approximately one
order of magnitude over the past 50 years and climate
change and estuary alteration are considered primary
threats to the viability and persistence of the population
complex (Moyle et al. 2017). Extinction of the entire
CCC steelhead complex is possible in the next century
without significant investments in monitoring and re-
search to inform management decisions and guide res-
toration efforts (Moyle et al. 2017).
Of California’s 577 estuaries, 48% are considered
“bar-built estuaries”or “coastal river mouth lagoons”
(Clark and O'Connor 2019). These intermittent estuaries
(IEs) are commonly encountered in small- to medium-
sized watersheds with energetic wave climates and wet/
dry seasonality like the Mediterranean-climate regions of
South Africa, Australia, and the Pacific coast of North
America. These systems are known regionally as bar-
built estuaries, temporarily open/closed estuaries
(TOCEs), or intermittently closed and open lakes and
lagoons (ICOLLs) (Perissinotto et al. 2010). Intermittent
estuaries shift between open (full tidal exchange), partial-
ly open (muted tidal exchange), and closed or lagoonal
(no tidal exchange) ecosystem states depending on the
presence or absence of a coastal sandbar at the mouth and
the sandbar’s size in relation to ocean water surface
elevation fluctuations. Ecosystem functioning in IEs is
mainly governed by the sandbar regime, freshwater in-
flows, and mixing levels (Perissinotto et al. 2010). During
the closed state, water volume generally increases but
water quality usually declines (Perissinotto et al. 2010;
Largier et al. 2015) and an understanding of the quantity
versus quality tradeoff for rearing organisms is needed to
guide conservation and management practices.
California IEs are considered “high-risk–high-reward”
habitat for juvenile O. mykiss (Satterthwaite et al. 2012).
During favorable environmental conditions, California’s
IEs provide opportunities for rapid juvenile steelhead
growth during both the closed (i.e., “lagoon”,Smith
1990; Hayes et al. 2008) and open (Fuller 2011; Huber
2018) states and lagoon-rearing fish exhibit strong size-
dependent marine survival advantages (Bond et al. 2008).
Mid- to late-summertime growth rates for CCC steelhead
can be 20 to 80 times greater in lagoons compared to
upstream habitats (Smith 1990; Hayes et al. 2008)be-
cause of abundant prey (Robinson 1993; Martin 1995)
and the cool coastal climate. Estuarine-rearing juvenile
CCC steelhead represent an important life history type of
the population complex (Smith 1990; Hayes et al. 2011)
and the increased biocomplexity likely enhances popula-
tion resiliency via the portfolio effect (Hilborn et al. 2003;
Schindler et al. 2010). However, predation pressure is
considered much greater in the estuary compared to up-
stream habitats (Satterthwaite et al. 2012). Indeed,
Osterback et al. (2013) estimated that the median proba-
bility of predation by just a single bird species, Western
gull (Larus occidentalis), on juvenile steelhead in three
Central California Coast watersheds (Waddell, Gazos,
and Scott Creeks) to be 30%, and to range from ~7 to
80% depending on the watershed and year. Beyond pre-
dation, chronically degraded water quality during extend-
ed closure may diminish nursery function (Huber 2018)
or cause it to cease altogether (Jankovitz 2015). Finally,
animal mass mortality events occasionally occur in IEs
(Whitfield 1995), particularly during the transition from
closed to open estuary. For example, regular breach-
induced fish kills occur in the Pescadero IE along
California’s Central Coast (Sloan 2006; Largier et al.
2015) and are caused by acute and widespread chemical
oxygen demand during the transition from closed to open
estuary (Smith 2009;Richardsetal.2018).
There is growing concern that anticipated increases
in extreme drought (Langridge 2018)willlikelycause
Central California Coast lagoon waters to become too
warm and oxygen poor for juvenile steelhead (Moyle
et al. 2017). Therefore, the identification of habitat
features favored by juvenile steelhead - particularly
during periods of prolonged drought-induced sandbar
closure - will help inform natural resource management
decisions. Determining the microhabitat use for mobile
aquatic animals is difficult because of the opacity and
instability of the underwater environment and limita-
tions of traditional sampling methods. Recent advances
in acoustic telemetry technology, however, have trans-
formed the ability to quantify and relate aquatic animal
distributions and movements to habitat features over
fine spatiotemporal scales (reviewed by Hussey et al.
2015). We used a combination of passive and mobile
acoustic telemetry to identify microhabitat features used
by multiple juvenile O. mykiss rearing in a Central
California IE during an extended sandbar-closed state.
Specifically, we (1) deployed an array of stationary
receivers to triangulate and automatically log fish posi-
tions for 15 weeks in the lower and middle estuary and
detect presence in the upper estuary, (2) conducted
Environ Biol Fish (2020) 103:509–529
510
Author's personal copy

multiple mobile fish tracking and water quality surveys
throughout the lagoon during the late summer and fall,
and (3) automatically monitored water quality condi-
tions at multiple depths at a fixed location and multiple
locations at fixed depths. We used these data streams to
characterize relationships between physical habitat, wa-
ter quality, and fish size with habitat utilization areas
and fish movement rates.
Methods
Study site
The Pescadero Intermittent Estuary is located in Central
California, approximately 60 km south of San Francisco
(Fig. 1). The watershed originates in the heavily forested
Santa Cruz mountains and is comprised of two main
sub-basins - Pescadero Creek (150 km
2
) and Butano
Creek (60 km
2
) - which share a confluence in the upper
estuary (Fig. 1). Total watershed relief is 732 m and
mean annual precipitation ranges from 50 cm near the
coast to 130 cm or greater at higher elevations (Bozkurt
Frucht et al. 2018). Nearly all precipitation occurs from
November to April because of the region’s Mediterra-
nean climate. Usually sandbar formation occurs during
the late summer followed by full breaching with the
onset of the fall rains. However, closure periods may
persist much longer during drought because of the com-
bined effects of low freshwater inflows and the absence
of energetic storm swells. At its fullest when adjacent
salt marsh wetlands are flooded (Fig. 1), the lagoon’s
surface area is approximately 1.00∙10
6
m
2
, which is one
to two orders of magnitude larger than other IEs in the
region (Beck et al. 2006).
Fish tagging
Juvenile O. mykiss were sampled by beach seine in the
middle estuary near station 07 (Fig. 1) on September 9,
2013. Sampling occurred in the middle estuary where,
unlike the lower estuary, deep water was available to
hold fish in live wells during pre- and post-surgical
periods. The first 35 fish sampled below a 15% tag size
to fork length (FL) ratio (Chittenden et al. 2009) were
surgically implanted with uniquely coded acoustic
transmitters (VEMCO V5–180 kHz; nominal code
transmission delay: 30–90 s range). The tagged fish
ranged in size from 139 to 206 mm (x
̅
=175 mm; Table 1)
FL which corresponds to a tag length to FL ratio of 6.2%
to 9.1% (x
̅
=7.3%). The tag weight in air to estimated
body weight ratio ranged from 0.6% to 2.1% (x
̅
=1.0%)
Fig. 1 Location of Pescadero Intermittent Estuary (IE) in San
Mateo County, California. VEMCO Positioning System (VPS)
stations (st) with a receiver only are indicated in red font. VPS
stations with both a receiver and synctag are in orange font and
stations with synctags only are presented in green font. Receivers
at st13 (Pescadero Creek) and st14 (Butano Creek) detected fish
presence in the upper estuary but were not used to calculate
positions since they were located too far upstream of the VPS
array. The lower estuary is downstream of st06 whereas the middle
estuary is between st06 and the confluence of Pescadero and
Butano Creeks. The Google Earth (V 7.3.2.5776; earth.google.
com/web/) satellite image was accessed 9-Jan, 2020 and is from
28-Mar, 2015 at 37.267099°N, 122.411496°W and an eye altitude
of 431 m
Environ Biol Fish (2020) 103:509–529 511
Author's personal copy

Table 1 Summary information for individually tagged juvenile steelhead (“individual sample”) including each unique tag ID and fork length (FL) at tagging as well as information on total
number of VPS positions, total detections upstream of the VPS array in Pescadero (st13) and Butano (st14) Creek arms of the upper estuary, and total number of mobile positions. First and
last detection dates by passive and active telemetry are also provided. Kernel Density Estimates (KDE) representing estimates of central core (25% KDE area), core (50% KDE area), and
core plus activity space (75% KDE area) habitat use area for the 17 fish providing nearly all of all HPE-filtered position data for the pooled sample are shown. Lastly, overall rate of
movement (ROM) mean and standard error (SE) calculations are presented for all fish providing position data
Tag
ID
FL
(mm)
Tot VPS
pos
Tot st13
det
Tot st14
det
Tot
mob
pos
First VPS
pos
Last VPS
pos
First st13
det
Last st13
det
First st14
det
T01 139 22 2 3 9/1013 20:54 9/10/13 15:05 9/10/13 17:24
T02 140 30569 1618 548 23 9/11/13 6:39 11/19/13 19:37 9/10/13 12:47 11/12/13 15:49 9/13/13 1:48
T03 206 36923 645 628 18 9/23/13 13:15 12/19/13 8:03 9/26/13 10:30 12/6/13 4:47 9/12/13 4:41
T04 492 2661 309 9/11/13 6:43 9/12/13 19:00 9/10/13 17:20 9/16/13 6:56 9/13/13 3:26
T05 152
T06 185 25844 29361 1414 22 9/12/13 20:26 12/20/13 22:25 9/10/13 10:24 12/22/13 3:44 9/10/13 10:50
T07 168 42293 23 9/12/13 0:17 12/22/13 22:07
T08 182 59073 2019 167 20 9/11/13 17:05 12/15/13 2:37 12/14/13 8:22 12/20/13 23:46 9/18/13 14:57
T09 177 57383 1035 370 30 9/11/13 15:23 12/9/13 3:51 9/12/13 5:45 12/19/13 17:54 9/11/13 3:07
T10 172 37811 3563 28072 22 9/11/13 7:55 12/21/13 15:59 9/14/13 6:39 12/16/13 7:02 9/14/13 8:29
T11 189 1092 1408 11 9/11/13 14:26 9/15/13 15:47 9/10/13 17:02 9/12/13 15:37 9/11/13 20:04
T12 179 54500 2942 200 30 9/11/13 16:55 12/8/13 1:53 11/4/13 12:09 12/22/13 14:33 9/26/13 15:29
T13 188 57681 6229 702 35 9/17/13 16:01 12/22/13 14:10 9/11/13 4:59 9/17/13 9:23 9/11/13 7:30
T14 172 62 461 9/11/13 18:43 9/16/13 3:07 9/11/13 9:56
T15 188 295 141 9/10/13 9:52 9/10/13 19:25 9/11/13 10:41
T16 170 32170 761 2269 19 9/11/13 7:43 12/6/13 1:40 9/10/13 11:08 12/6/13 4:30 9/10/13 5:16
T17 170 112 9/11/13 2:02 9/10/13 20:36
T18 173 20247 7477 99 4 9/19/13 6:17 12/1/13 2:16 10/21/13 1:28 12/9/13 2:50 9/23/13 14:28
T19 195 3194 1086 45124 25 9/11/13 7:42 11/27/13 16:37 9/10/13 10:40 11/20/13 18:05 9/10/13 8:47
T20 189 296 37 9/10/13 3:36 9/10/13 17:51 9/9/13 23:35
T21 170 29 9/10/13 4:00 9/11/13 20:37
T22 172 62306 556 22271 30 9/11/13 8:15 12/22/13 13:13 9/28/13 3:40 11/6/13 1:43 9/25/13 16:11
T23 195 97 178 9/10/13 11:07 9/10/13 23:04 9/11/13 1:11
T24 162 60 73013 289 14 10/2/13 4:00 10/2/13 5:59 9/10/13 14:24 12/22/13 0:17 9/10/13 7:44
T25 172 1134 9/11/13 7:43 9/16/13 6:35
T26 189 158 414 70 9/12/13 21:20 9/13/13 6:41 9/10/13 12:48 9/15/13 1:28 9/10/13 11:50
T27 159 250 118 36 9/11/13 12:49 9/13/13 7:23 9/10/13 17:25 9/13/13 7:47 9/11/13 14:40
T28 150 1358 238 1 9/11/13 17:18 9/15/13 10:44 9/10/13 20:03 9/10/13 18:29 9/15/13 21:12
T29 139 32511 512 314 25 9/11/13 17:19 12/21/13 2:02 9/10/13 13:50 12/15/13 18:58 9/11/13 2:27
T30 172 13643 23956 18423 20 9/14/13 20:46 12/19/13 22:29 9/10/13 10:36 12/22/13 4:26 9/12/13 12:42
T31 166 84521 2 9/11/13 11:14
Environ Biol Fish (2020) 103:509–529
512
Author's personal copy

Table 1 (continued)
Tag
ID
FL
(mm)
Tot VPS
pos
Tot st13
det
Tot st14
det
Tot
mob
pos
First VPS
pos
Last VPS
pos
First st13
det
Last st13
det
First st14
det
T32 201 3749 30592 50 21 9/18/13 9:25 12/5/13 18:00 9/10/13 8:09 12/16/13 19:57 9/11/13 4:19
T33 168 65074 9825 442 45 9/12/13 19:45 12/22/13 12:00 9/10/13 13:38 12/18/13 22:50 9/11/13 3:25
T34 200 67057 462 33 9/11/13 15:30 12/18/13 17:27 9/19/13 14:43 12/10/13 7:09
T35 200 71228 32 29 9/11/13 6:57 12/14/13 15:50 9/12/13 2:38
Tag ID Last st14 det First mob pos Last mob det KDE
25
(m
2
)KDE
50
(m
2
)KDE
75
(m
2
) Mean ROM (mm⋅s
–1
)SEROM(mm⋅s
–1
)
T01 9/10/13 17:25 10/1/13 12/14/13
T02 11/16/13 17:23 10/1/13 11/7/13 579.2 1655.3 4997.0 76.9 1.4
T03 11/13/13 16:07 11/7/13 12/14/13 114.4 279.8 1847.6 54.3 0.4
T04 9/14/13 8:20 23.1 1.9
T05
T06 12/5/13 12:17 10/1/13 12/14/13 61.7 220.7 1500.1 47.9 0.5
T07 11/7/13 12/14/13 20.8 49.4 95.4 43.3 0.5
T08 12/15/13 20:51 10/1/13 12/14/13 43.0 118.9 262.0 49.7 0.5
T09 12/6/13 1:45 10/1/13 12/14/13 185.7 741.4 3416.6 84.5 0.5
T10 12/22/13 1:43 10/1/13 11/21/13 217.4 1135.1 4427.8 72.9 0.5
T11 9/12/13 16:57 22.1 2.1
T12 12/20/13 5:32 10/1/13 12/14/13 65.6 163.4 960.0 65.3 0.8
T13 10/4/13 23:40 10/1/13 12/14/13 264.3 991.1 4610.2 49.2 0.5
T14 9/15/13 16:20
T15 9/11/13 13:55
T16 12/6/13 11:49 11/7/13 12/14/13 105.3 243.0 1364.1 65.9 0.7
T17
T18 12/1/13 13:05 12/14/13 12/14/13 48.4 213.6 730.7 47.6 0.6
T19 11/26/13 13:05 10/1/13 11/21/13 69.1 2.0
T20 9/10/13 18:54
T21
T22 11/7/13 0:22 10/1/13 12/14/13 35.5 81.0 201.9 43.4 0.4
T23 9/11/13 5:25
T24 10/3/13 0:58 11/7/13 12/14/13 104.5 31.1
Environ Biol Fish (2020) 103:509–529 513
Author's personal copy

according to a power transformation for Pescadero sub-
adult steelhead sizes (mass = 8.53·10
−6
·FL
3.06
;53–
442 mm range; R
2
= 0.99; Huber 2018). Acoustic tags
were inserted into the peritoneal cavity through an inci-
sion above the ventral midline. The incision was closed
with two absorbable, sterile, and plain gut interrupted
surgical sutures (Chromic Gut 2–0, Ethicon Inc., Som-
erville, NJ). All fish were carefully monitored and re-
leased back into the environment after recovering from
surgery.
Telemetry
Passive acoustic telemetry (lower and middle estuary
positions) The VEMCO Positioning System (VPS) was
comprised of 12 strategically-placed omnidirectional
acoustic receivers (VEMCO VR2W-180 kHz) with
overlapping ranges capable of triangulating fine-scale
fish positions according to preliminary range testing
(Fig. 1). All receivers were secured to a line with floats
of unequal buoyancies at each end that was threaded
through a weighted mooring and deployed with
hydrophone-down in the middle of water column
(~0.4 m to 0.7 m water depths) in the lower and middle
estuary (Fig. 1). Eight receiver time synchronization
transmitters (hereafter “synctags”;VEMCOV6–
180 kHz; nominal code transmission delay: 500–700 s
range) were strategically deployed at stations (st) with or
without VPS receivers (Fig. 1). Synctags are needed by
the VPS to improve positioning accuracy by correcting
for clock drift between stationary receivers. Due to
unexpected station movements, the VPS software was
used to determine the calculate receivers and synctag
locations based on the observed arrival time differences
between signals at the receivers. Determinations of the
geographic latitude and longitude coordinates of fish
positions from VPS receiver detections were post-
processed and provided by VEMCO (Bedford, Nova
Scotia).
Passive acoustic telemetry (upper estuary
detections) Two stationary receivers were positioned
upstream of the VPS in the Pescadero (st13) and Butano
(st14) Creek arms of the estuary (Fig. 1). Detections
from these receivers allowed us to determine presence
of acoustically tagged fish outside of our main array in
the upper estuary, but did not allow us to triangulate
positions. Rather, we used data from these two station-
ary receivers to infer upstream movement.
Table 1 (continued)
Tag ID Last st14 det First mob pos Last mob det KDE
25
(m
2
)KDE
50
(m
2
)KDE
75
(m
2
) Mean ROM (mm⋅s
–1
)SEROM(mm⋅s
–1
)
T25 41.2 6.8
T26 9/15/13 9:25 26.8 3.7
T27 9/12/13 17:48 76.9 32.0
T28 9/15/13 21:12 25.1 3.0
T29 12/16/13 3:11 10/1/13 12/14/13 138.7 1477.8 5154.3 72.3 1.0
T30 11/30/13 10:48 10/1/13 12/14/13 126.5 477.6 2890.7 53.5 0.9
T31 12/21/13 15:31 11/21/13 12/14/13
T32 12/4/13 11:46 10/1/13 12/14/13 114.3 4.0
T33 12/21/13 13:31 10/1/13 12/14/13 248.0 841.2 3123.6 72.4 0.6
T34 10/1/13 12/14/13 267.6 1084.0 3924.2 69.3 0.6
T35 11/26/13 6:18 10/1/13 12/14/13 23.3 70.5 573.7 65.5 0.4
Environ Biol Fish (2020) 103:509–529
514
Author's personal copy

Active acoustic telemetry (all navigable water
positions) To supplement the information from the
VPS array, fish positions throughout the lower, middle,
and upper estuary were determined by mobile tracking
surveys conducted during the afternoon hours on October
1, November 7 and 21, and December 14 in 2013
throughout all navigable waters using a kayak-mounted
acoustic receiver (VEMCO VR100) and omnidirectional
hydrophone (VEMCO VH165).
Position error VEMCO provides a unitless horizontal
position error (HPE) with each fish tag and synctag
position. The HPE is a relative measure of the confi-
dence estimate for each position and is calculated ac-
cording to the relationship between theoretical position
error sensitivities and observed synctag measurement
errors since synctag positions are known. Error potential
is greater beyond the boundsof the VPS array (Espinoza
et al. 2011a;Smith2013) and highHPE values - defined
as those greater than the 75th percentile in the data set -
were excluded from all analyses in order to improve the
accuracy of fish position measurements (Roy et al.
2014).
A study-specific linear transformation is needed
to convert HPE to an estimate of measured error, in
meters (HPE
m
). Following Smith (2013), we derived
HPE
m
by applying a linear transformation between
HPE and HPE
m
for calculated synctag positions. We
assume that the relationship between HPE and HPE
m
is representative of the same relationship for fish
tags. We found that 75% of all fish positions (here-
after “HPE-filtered”) possessed HPE values less than
3.8. An HPE value of 3.8 correlates to an HPE
m
value of 1.4 m for this study. Therefore, the horizon-
tal distance between a synthesized position and the
known location of the transmitter is estimated to be
1.4 m or less for this study.
Environmental data
Water depth, temperature, salinity, pH, DO, and light
were measured throughout the study area, including at
fixed water depths at multiple locations and at multiple
water depths at fixed locations in order to investigate
potential water quality effects on fish microhabitat uti-
lization. Specifically, conductivity-temperature-depth
(CTD) sensor clusters (XR-420 CTD, RBR Ltd., Otta-
wa, Ontario) were secured to weighted moorings and
positioned approximately 0.25 m above the bottom of
the estuary throughout the VPS array (stations 01, 02,
06, 09, 11, and 12; Fig. 1). A CTD was also deployed
0.25 m below the water surface in the middle estuary at
station 06 (Fig. 1). Periodic measurements of a land-
surveyed stage gauge attached to the CA Hwy-1 bridge
in the lower estuary (Fig. 1) were used as reference points
to convert CTD depth data to water surface elevations
(WSEs). Lagoon WSE data were converted to meters
using the NAVD88 vertical datum (GEOID 2009 model).
Solar radiation and temperature sensors (HOBO® Pendant
Temperature/Light Data Loggers, Onset Computer Corp.,
Bourne, MA) were deployed aerially in the upper estuary
(37.260807°N, 122.407951°W) and within the VPS and
non-VPS zones of the stationary array at water depths of
0.75 m (stations 02, 03, 06, 08, 12, 13), 1.25 m (stations
02, 06, 08, 12, 13, 14), 1.50 m (st01), and 1.75 m (stations
08, 12) (Fig. 1). Additional temperature sensors (HOBO®
Pendant Temperature Data Loggers, Onset Computer
Corp., Bourne, MA) were deployed within the VPS and
non-VPS zones of the stationary array at water depths of
0.25 m (stations 01, 02, 03, 05, 07, 08, 09, 10, 13, 14),
0.75 m (stations 04, 05, 10, 11), 1.25 m (stations 04, 05,
09, 10, 11), and 1.75 m (stations 08, 12) (Fig. 1).
Temperature, salinity, pH, and DO concentrations
were measured by multiparameter water quality sondes
(YSI 6600, YSI Inc., Yellow Springs, OH) moored to
stations in the lower (st02) and upper (stations 13, 14)
estuary (Fig. 1) at 1.25 m depth. Water quality sondes
(Hydrolab MS5, OTT Hydromet-Hach® Co., Loveland,
CO) were also deployed throughout the water column at
0.60, 1.10, 1.60, 2.10, and 2.35 m depths in the middle
of the VPS zone at station 08 (Fig. 1). During the
September 25 (both pre-noon and post-noon hours)
and November 21 (afternoon hours only) mobile
trackimg fish surveys, lagoon water temperatures, salin-
ity, pH, and DO were measured every 4 s using a
compact water quality sonde (YSI 600XLM, YSI Inc.)
that was slowly bobbed up and down throughout the
water column in front of a foot pedal-powered kayak.
Geographic coordinates were recorded during water
quality cruises using a handheld GPS device (Garmin
GPSMAP 78, Garmin Ltd., Olathe, KS). Mobile survey
data are presented using Google Fusion Tables (Google
LLC, Mountain View, CA). All geographic data are
presented using the WGS84 datum for coordinates and
all times are UTC-8 h. Fixed environmental sensor data
were automatically logged at 15 min intervals except
those obtained from the station 08 sonde cluster (30 min
pulse rate). Some data gaps exist because of instrument
Environ Biol Fish (2020) 103:509–529 515
Author's personal copy

unavailability, technical difficulties, or sensor
biofouling.
Statistical analyses
Utilization areas To determine habitat utilization, the
Spatial Analyst extension in ArcGIS 10.5® (Esri™,
Redlands, CA) was used to calculate two-dimensional
kernel density estimation (KDE) isopleths using the
Silverman’s method to select bandwith (Silverman
1986). Utilization area determination using KDE analy-
sis is a popular and widely used nonparametric method
to map habitat use due to its ease of application and
suitability for large data sets (Kie et al. 2010). The KDEs
represent a probability distribution for a tagged juvenile
steelhead being located within a given area during a
specific time (Worton 1989; Seaman and Powell
1996). Quartile analyses of KDE isopleth areas is a
popular technique to estimate utilization areas of
telemetered animals (e.g., Rechetelo et al. 2016;
Rooker et al. 2018) and we consider 0–25%, 0–50%,
50–75%, and 75–100% KDE contour areas to represent
central core (25% probability of presence in area), core
(50% probability of presence in area), activity space
(outside core area and 25% probability of presence),
and excursion space (outside core and activity space
areas and 25% probability of presence) utilization areas,
respectively. The areas were generated from fish position
data for both individually-tagged fish (hereafter “indi-
vidual sample”) and all tagged fish combined (hereafter
“pooled sample”). We summarized utilization areas at
weekly time scales for the pooled sample and full study
time scales for the pooled and individual samples. Mea-
sures of central tendency (e.g., mean for pooled sample
and median for individual sample) and dispersion (e.g.,
total range for pooled sample and interquartile ranges for
individual sample) are presented. The large fish position
dataset also permitted finer scale analyses of habitat use
and we investigate potential diel microhabitat use shifts
according to decile analyses of KDE isopleth areas dur-
ing day (05:00:00 to 18:59:59) versus night (19:00:00 to
04:59:59) for the pooled sample.
We present relationships between mean weekly water
quality parameters (temperature, salinity, pH, DO) and
mean core habitat utilization area for the pooled sample
as scatterplots for ease of visual inspection and examined
quantitatively using Pearson r correlation analysis (α=
0.05). Mean weekly water temperatures were first calcu-
lated as depth-averaged daily mean values from 0.25 m,
0.75 m, 1.25 m, and 1.75 m water depths at all VPS
receiver locations. Next, weekly mean water tempera-
tures were calculated from the depth-averaged daily
mean values. The salinity, pH, and DO data used for
Pearson r correlation analyses and scatterplot display
were obtained from the middle of the VPS monitoring
zone (st08; Fig. 1) at mid-water column (1.60 m depth).
We related core and activity space areas for individuals
to fish fork length at tagging using least squares linear
regression (α= 0.05). We explored juvenile steelhead
microhabitat preferences using comparisons of KDE
decile isopleths to a previously published bathymetric
map of the Pescadero IE (Williams 2014). Finally, we
compared KDE decile isopleths to large woody debris
deposits that we mapped during mobile tracking surveys.
Rate of movement The Tracking Analyst extension in
ArcGIS 10.5® was used to determine distances from
successively ordered fish positions for individual fish to
calculate distance traveled per unit time or the “rate of
movement”(ROM, sensu Espinoza et al. 2011b). All
ROM analyses were restricted to instances when the
time interval between successive position measurements
was less than or equal to the maximum fish tag delay
(90 s) to minimize error due to deviations from straight
path movements. We related mean ROM for individual
fish to their fork length at tagging using least squares
linear regression (α= 0.05). Similar to analyses for core
habitat utilization area, relationships between mean
weekly water quality parameters (temperature, salinity,
pH, DO) and mean ROM for the pooled sample were
presented as scatterplots for ease of visual inspection and
examined quantitatively using Pearson r correlation
analysis (α= 0.05). Categorical analyses of ROM data
are made according to expected conditions in the VPS
monitoring zone based on long term monitoring of the
Pescadero IE (Sloan 2006;Smith2009;Huber2018)
including potentially deadly (e.g., <2 mg·l
−1
DO), stress-
ful (e.g., 2 to 5 mg·l
−1
DO), tolerable (e.g., 5–6 mg·l
−1
DO; <14 °C), preferred (e.g., >7 mg·l
−1
), and optimal
(e.g., 18 °C) water quality conditions for juvenile Onco-
rhynchus spp. (Brett 1979;Brett1995;Molony2001).
Upper estuary detections Fish presence upstream of the
VPS array was quantified according to total number of
detections by stationary receivers located in the
Pescadero (st13) and Butano (st14) Creek arms of the
upper estuary (Fig. 1). Categorical analysis of upper
estuary fish presence is made according to expected
Environ Biol Fish (2020) 103:509–529
516
Author's personal copy

conditions based on long term monitoring of the
Pescadero IE (Sloan 2006;Smith2009; Huber 2018).
Summary statistics (e.g., mean, median, total range, stan-
dard deviation) and least squares linear regression analy-
ses (α= 0.05) of upper estuary fish detections were ana-
lyzed using the same approach as for the ROM analyses
(see above).
Environmental data Beyond the fish location data, we
also summarize the environmental data including mean
aerial and underwater photon flux densities at multiple
water depths (lux, dependent variable) and times-of-day
(proportion of 24 h, independent variable) from 0.25 to
0.75 (06:00 to 18:00). We used a Gaussian function to fit
curves of lux values (independent variable) versus time-
of-day (dependent variable). We characterized the extent
of light attenuation using a linear regression of modeled
lux (independent variable) and sensor depth (dependent
variable).
Multicollinearity of mean water temperature, salinity,
pH, and DO was investigated using Pearson r correlation
analysis (α= 0.05). We present these data in a correlation
matrix with positive or negative correlations presented as
upward-trending or downward-trending ellipses, respec-
tively. The water temperatures used for multicollinearity
analysis were calculated as depth-averaged daily mean
values from 0.25 m, 0.75 m, 1.25 m, and 1.75 m water
depths at all VPS receiver locations (Fig. 1). The salinity,
pH, and DO data used multicollinearity analysis were
obtained from the middle of the VPS monitoring zone
(st08; Fig. 1) at mid-water column (1.60 m depth).
All statistical analyses were conducted using PAST
version 3.18 (Hammer et al. 2001) and original data are
available from the Dryad Data Repository (Huber and
Carlson 2020).
Results
Environmental conditions
Water surface elevations The low-lying, flat, and wide
sandbar permitted frequent sheet-like marine over-wash
events which produced frequent and small daily water
surface elevation (WSE) fluctuations (Fig. S1-3a and SI
Appendix A, Fig. A5 in Huber 2018). Mean daily WSE
was 2.72 m and ranged from a minimum of 2.55 m on
September 19 to 2.95 m on December 14 (Fig. S1-3a).
Solar radiation Luminous flux data were only available
from September 9 to October 2 because of extensive
biofouling of sensor housing by macroalgae. The daily
photon flux density pattern was approximately normally
distributed; minimum non-zero aerial solar radiation
occurred during the morning at 05:30 (x
̅
=1.10 lx) and
the evening at 18:45 (x
̅
=0.75 lx) (Table S1-1, Fig. S1-
3b). Maximum illumination occurred at 12:00
(x
̅
=111,149.21 lx) (Table S1-1, Fig. S1-3b). Pro-
nounced light attenuation-at-depth was observed and
daytime darkness likely occurred at depths greater than
2 m, at least during the first 24 days of the study
(Table S1-1, Fig. S1-3b).
Temperature Water temperatures were highly homoge-
nous in space but variable over time (Fig. 2a). The depth-
averaged mean water temperature, calculated as the over-
all mean of mean daily temperatures at 0.25 m, 0.75 m,
1.25 m, and 1.75 m depths at all VEMCO Positioning
System (VPS) receiver locations (Fig. 1), declined steadi-
ly from 18.5 °C to 9.7 °C across the study period (Fig.
S1–5) reflecting the seasonal shift in air temperature.
Overall, water temperatures in the lower lagoon (station
02, 1.25 m depth) were 0.3 °C and 0.7 °C cooler than
temperatures in the Pescadero Creek (st13) and Butano
Creek (st14) portions of the upper estuary, respectively
(Fig. S1-4a). Water temperature measurements through-
out all navigable waters of the Pescadero IE are presented
in Supplement 1 (Fig. S1–2a).
Salinity Mean salinities in the VPS monitoring zone
increased with depth from 2.3 ppt near the water
surface to 24.3 ppt near 3.0 m depth (Figs. 2b,S1–
1). Salinity generally decreased over the study period;
the halocline was approximately 0.35 m deeper during
themobilesurveyonNovember21comparedtoSep-
tember 25 (Fig. S1–2b). Saline conditions (≥25 ppt)
were restricted to areas deeper than the halocline (Fig.
S1–2b) such as those found along the thalweg and
various deep water pockets (Fig. 3c). Aside from a
moderate salinity spike (≥8 ppt) from the benthos up
to ~1.5 m water depth in early and mid-November
(Fig. 2b), the frequent sheet-like over-wash events over
the wide and flat sandbar (Fig. S1-3a and SI Appendix
A, Fig. A5 in Huber 2018) did not substantially alter
the lagoon’s salinity regime.
pH Aside from weakly acidic conditions in bottom
waters (Fig. S1–2c), pH levels were neutral and weakly
Environ Biol Fish (2020) 103:509–529 517
Author's personal copy

alkaline and at preferred levels for rearing salmonids
(Molony 2001;Figs.2c,S1–2c,S1-4c).
Dissolved oxygen Normoxic conditions (≥7.0 mg·L
−1
)
prevailed from the water surface to approximately 1.5 to
2.0 m water depths throughout the lower and middle
estuary and the lower portion of the upper estuary (Figs.
2d,4b,S1–2d,S1-4d). Mean DO concentrations de-
clined with depth from 8.87 mg∙l
−1
at 0.60 m to
1.86 mg∙l
−1
at 2.35 m (Fig. 2d). Mean minimum daily
DO concentrations varied spatially within the lagoon;
the mean minimum DO value at the station closest to the
mouth (st02) at 1.25 m depth was 7.83 mg∙l
−1
and
considerably greater than those observed in the upper
estuary at stations 13 (5.34 mg∙l
−1
) and 14 (5.23 mg∙l
−1
)
(Fig. S1-4d). The lowest DO values were observed in
the Butano Creek arm of the upper estuary (Fig. 4b).
Mean daily DO in both the VPS and non-VPS zones of
the stationary array tended to increase over the course of
the study (Fig. S1-4d). Daily DO range was greater in
the upper estuary than the lower estuary and DO values
were higher during the afternoon than night and morn-
ing (Figs. S1–2d,S1-4d).
Telemetry patterns
VEMCO positioning system (VPS) performance The
12 receivers in the lower and middle estuary VPS mon-
itoring zone (Fig. 1) recorded a total of 5,906,869 fish tag
detections (Fig. 5a) and 841,103 synctag detections (Fig.
5b) during the study period from September 9, 2013 to
December 22, 2013. According to VEMCO, receiver
time synchronization was “good”for the duration of the
study. A small lack of synchronization occurred at station
06 (Fig. 1) before September 20, 2013 due to a code map
change. Thirty-three unique fish tags were registered by
stationary receivers and total detections per fish tag
ranged from 162 (T23) to 476,175 (T35). Seven tags
Fig. 2 Time series of (a) tem-
perature, (b) salinity, (c) pH, and
(d) dissolved oxygen (DO) at
multiple depths in a fixed location
(st08) in the middle of the VPS
monitoring zone. Water
quality measurements were made
at 0.60, 1.10, 1.60, 2.10, and 2.35
m below the water surface and the
interpolated two-dimensional data
matrix plot uses and a color scale
with dark blue for lowest values
and dark red for highest
values. Missing physico-
chemical data are plotted as
blanks
Environ Biol Fish (2020) 103:509–529
518
Author's personal copy

were detected but insufficient data precluded calculation
of triangulated positions. A total of 777,800 fish positions
for 26 tagged individuals (Table 1) and 103,229 synctag
tag positions were provided by VEMCO. Each fish tag
transmission was detected by 3.8 receivers, on average,
and at least three receivers detected 69.7% of all fish tag
transmissions (Fig. 5a). Overall, 92.6% of synctag trans-
missions were logged on three or more receivers and each
synctag transmission was detected over seven times on
average (Fig. 5b).
Passive acoustic telemetry (lower and middle estuary
positions) As expected, many of the HPE-filtered posi-
tions were located outside the VPS receiver polygon (Fig.
S2–1) in the shallow sandy region of the lower estuary
(Figs. 1,3c) where receivers were intentionally absent
due to concerns of vandalism to moored equipment. Total
HPE-filtered positions per fish tag ranged from 53 (T24)
to 55,928 (T35). One fish tag (T05) was never detected
by passive or active acoustic telemetry and another tag
(T31) was detected outside the stationary array by mobile
Fig. 3 (a) Bathymetry map of study site. Map of (b) diurnal
(05:00:00 to 18:59:59) and (c) nocturnal (19:00:00 to 04:59:59)
KDE decile areas. The bathymetry map is redrawn from Williams
(2014) and for November 2010 when the lagoon water surface
elevations were approximately 0.25 m lower than November
2013. The KDEs represent a probability distribution for a tagged
juvenile steelhead being located within a given area during a
specific time
Environ Biol Fish (2020) 103:509–529 519
Author's personal copy

tracking efforts only (Table 1). The median number of
days between first and last detection by either passive or
active telemetry was 94.6 days and ranged from 0.2 (T17)
to 102.9 (T33) days (Table 1). Nine of the 26 tags
producing position data failed to provide more than
2829 HPE-filtered positions. The 17 highest ranking fish
tags, in terms of data yield, represent 98.6% of all HPE-
filtered fish positions in the pooled sample. These 17 fish
are chosen to represent the individual sample because,
unlike the lowest ranking fish, they were detected beyond
the first week of the study (Table 1).
From November 20 to the end of the study period on
December 22, the total number of HPE-filtered daily
fish positions declined linearly at a rate of approximate-
ly 193 positions∙day
−1
. Consistent with the decline of
stationary receiver detections (Fig. 5a) and HPE-filtered
daily fish position data during fall, we detected an
increased presence of tagged fish in the uppermost
reaches of the estuary (Fig. 4a). Mean daily detections
(and standard deviation) during the first, middle, and last
thirds of the study period were 1507 (851), 1940 (347),
and 2307 (1129) for station 13, respectively, and 1215
(806), 2743 (376), and 2020 (364) for station 14, re-
spectively. Together, these data suggest increasing up-
stream movement out of the study reach over the course
of the study.
Active acoustic telemetry (all navigable water
positions) Mean duration of the four mobile tracking
surveys was 150.2 min and ranged from 116.2 min on
December 14 to 178.1 min on November 21. A total of
493 positions representing 22 unique fish tags were
detected during the four mobile tracking surveys
(Table 1,Fig.4a). Sixteen unique fish tags were detect-
ed on October 1 and 19 unique tags were detected each
on November 7, November 21, and December 14 (Fig.
4a). On December 14, seven tagged fish were detected
in the lower lagoon, two were detected just upstream of
the Pescadero-Butano confluence in the lower portions
of the upper estuary and 10 were detected in the upper-
most portion of the estuary in Pescadero Creek (Fig. 4a).
Overall, tagged fish were detected more in the
Pescadero Creek arm of the upper estuary than the
Butano Creek arm during mobile surveys (Fig. 4a).
Utilization areas and rates of movement
Utilization areas (lower and middle estuary) Juvenile
O. mykiss heavily favored two distinct microhabitat
complexes in the Pescadero IE. One was located at
the point bar along the inside bend of the middle
estuary and another at the flood-tide delta region of
the lower embayment (Figs. 5a, b). The upper micro-
habitat complex possessed extensive LWD structure
and woody material along the shore that was recruited
by rising water levels (SI appendix A, Figs. A7, A8 in
Huber 2018) and was heavily favored at night by
juvenile steelhead (Fig. 3b). The lower microhabitat
complex was largely devoid of LWD but occurred
adjacent to a 12 m tall vertical cliff face.
Fig. 4 Fish positions during all four mobile tracking surveys (a)
and dissolved oxygen (DO) concentrations at 1.0 to 2.0 m water
depths on September 25, 2013 and November 21, 2013 (b)
Environ Biol Fish (2020) 103:509–529
520
Author's personal copy

We observe an exponential relationship between ker-
nel density estimate (KDE) decile endpoint and corre-
sponding areas for each 10% KDE increment during both
day and night for the pooled sample (Fig. S2–2). The
KDE areas for the pooled sample was lower at night than
during the day, including for the central core (184.9 m
2
vs. 33.5 m
2
), core (1185.3 m
2
vs. 113.3 m
2
), and core
plus activity space (7290.3 m
2
vs. 959.4 m
2
)habitat
utilizatiom areas (Figs. 3a, b,S2–2). Habitat utilization
areas for the pooled sample peaked during weeks 5, 10,
11, and 14 when median (and interquartile range) KDE
areas were 2827.1 m
2
(525.0 m
2
to 12,964.6 m
2
),
2278.5 m
2
(300.4 m
2
to 9885.6 m
2
), 2883.5 m
2
(312.4 m
2
to 13,363.2 m
2
), and 1447.4 m
2
(301.1 m
2
to
7804.8 m
2
), respectively (Fig. 6a).
The median (and interquartile range) of 25%, 50%,
and 75% kernel density estimate (KDE) areas for the
individual sample was 114.4 m
2
(48.4 m
2
to 217.4 m
2
),
279.8 m
2
(163.4 m
2
to 991.1 m
2
), and 1847.6 m
2
(730.7 m
2
to 3924.2 m
2
)(Table1and S1 Appendix B
Fig. 5 (a) Total fish tag and (b) synchronization
transmitter detections across all VPS receivers over time. Detec-
tions are binned into 4-h intervals. The uptick in synchronization
transmitter detections (b) is due to the effect of water temperature
and salinity on the speed-of-sound and indicates that the
VEMCO Positioning System (VPS) was properly functioning
throughout the 15-week study period. The linear decline of fish
tag detections starting in late November and increased presence of
tagged fish in the uppermost limits of the estuary in mid-December
(Fig. 4a) suggests increasing upstream movement of juvenile
steelhead over the course of the study (see text for more details)
Environ Biol Fish (2020) 103:509–529 521
Author's personal copy

in Huber 2018). Central core area ranged from 20.8–
579.2 m
2
, core area ranged from 49.4–1655.3 m
2
,and
activity space ranged from 46.0–3676.4 m
2
,respective-
ly (Table 1). All remaining wetted areas within the VPS
monitoring zone for the 17 fish comprising the individ-
ual sample consisted of excursion habitat utilization
areas (SI Appendix B in Huber 2018). Core area (C,
m
2
) declined significantly with fork length (FL, mm) at
tagging (C = −14.9·FL + 3205.5; R
2
= 0.26, p< 0.05,
df = 16). Similarly, activity space was negatively, but
not significantly, related to fork length at tagging.
Rate of movement (lower and middle estuary) The me-
dian (and interquartile range) of mean daily ROM
values for the pooled sample was 56.9 mm∙s
−1
(49.3 mm∙s
−1
to 76.7 mm∙s
−1
)(Fig.6a). Minimum and
maximum mean daily ROM for the pooled sample oc-
curred on September 11 (35.3 mm∙s
−1
)andOctober4
(109.1 mm∙s
−1
), respectively (Fig. 6a). Above average
mean daily ROM (x
̅
=62.1 mm∙s
−1
)occurredfromSep-
tember 15 to October 18, October 21, 24 to October 25,
and November 3 to 4 (Fig. 6a).
The median (and interquartile range) of mean daily
ROM for the individual sample was 65.3 mm·s
−1
(49.2 mm·s
−1
to 72.3 mm·s
−1
) and ranged from
43.3 mm·s
−1
(T07) to 84.5 mm·s
−1
(T09) (Table 1).
Mean ROM was negatively, but not significantly,
relatedtoforklengthattagging.MeanROM
(mm·s
−1
) was positively and significantly related to
core area (C = 26.1·ROM - 1010.4; R
2
= 0.41,
p<0.01, df = 16). Similarly, ROM (mm·s
−1
)waspos-
itively and significantly related to activity space area
(A, m
2
; A = 61.1·ROM - 1933.9; R
2
=0.36, p<0.05,
df =16).
Above average ROM values occurred during the day
whereas below average movement rates occurred at
Fig. 6 (a) Mean weekly central
core (0–25% KDE, black bar),
core (0–50% KDE, black and
light grey bar), and activity space
(50–75% KDE, grey) areas for all
acoustically tagged fish
combined (‘pooled sample’).
Mean daily rate of movement
(ROM, red line) for the
pooled sample is also presented;
(b) Mean hourly rate of move-
ment (ROM) during the first third
(study weeks 1–5, red), middle
third (weeks 6–10, blue), last third
(weeks 11–15, green), and entire
(weeks 1–15, black) study periods
Environ Biol Fish (2020) 103:509–529
522
Author's personal copy

night (Fig. 6b). Minimum and maximum ROM values
occurred during the 4-5 h (x
̅
=39.0 mm∙s
−1
)and6-7h
(x
̅
=93.6 mm∙s
−1
) hourly periods, respectively (Fig. 6b).
There were also seasonal changes in ROM, with ROM
generally decreasing over the course of the study. Dur-
ing the first third of the study period, meanROM peaked
at mid-day (Fig. 6b). After mid-October (week 6 of
study), mean ROM peaked at dawn and a smaller uptick
also occurred during dusk (Fig. 6b).
Environmental correlates of fish habitat use and rates
of movement
The physico-chemical variables were highly correlated
(Fig. 7). In general, water temperatures were negatively
correlated to pH values and DO concentrations, and
positively correlated to salinity levels (Fig. 7). The
positive relationship between pH and DO was strong
(Fig. 7), and both were negatively correlated to salinity
(Fig.7). Mean weekly core habitat areas for the
pooled sample was positively correlated to mean weekly
temperature and negatively correlated to mean weekly
salinity, pH, and DO but none of the relationships were
significant (Figs. 7a,8a-d). Similarly, mean daily ROM
was positively correlated to mean daily temperature and
negatively correlated to mean daily salinity, pH, and DO
(Figs. 1,7b,8e-h).
At the upper extent of the study site (stations 13
and 14, Fig. 1), the total daily detections of tagged
fish were negatively correlated to water temperature
(Figs. 7a, b,8i, m) and salinity (Figs. 7a, b,8j, n)
and positively correlated with pH (Figs. 7c, d,
8k, o)andDO(Figs.7c, d,8l,p) measured at
1.25 m water depths. For example, fish tag detec-
tions at station 13 were 31% higher than the overall
daily mean value (x
̅
=1918.1 det·day
−1
) when mean
daily DO concentration was greater than 7.0 mg·L
−1
and mean daily water temperature was less than
18 °C. At station 14, detections were 20% more
than the overall mean daily value (x
̅
=1992.1
det·day
−1
) when mean daily water temperature was
less than 18 °C regardless of DO concentrations.
When mean daily water temperatures were greater
than 18 °C, fish tag detections were 63% and 60%
less than average when mean daily DO concentra-
tions were less than and greater than 7.0 mg·L
−1
,
respectively. For temperatures below 14.0 °C, the
linear relationship between mean daily ROM
(mm·s
−1
) for the pooled sample and mean daily
temperature (T, °C) was significant (ROM =
1.1∙T+40.0; R
2
= 0.10, p<0.05, df =57). For tem-
peratures above 14.0 °C, a second-order polynomial
function better described the relationship between
mean daily ROM and water temperature (ROM=
−3.4∙T
2
+120.2∙T - 959.2; R
2
=0.62, df = 44) (Fig. 8e).
Overall, these patterns are reflective of a seasonal uptick
in movement that coincided with seasonal changes in
water quality conditions.
Fig. 7 Pearson r correlation matrices for (a) mean weekly physico-
chemical measurements and mean weekly core areas (50% KDE)
for the pooled sample, (b) mean daily physico-chemical measure-
ments and mean daily ROM values for the pooled sample, and mean
daily physico-chemical measurements and total daily fish tag detec-
tions at upper estuary stations 13 (c) and 14 (d). Blue upward and red
downward sloping ellipses indicate positive and negative associa-
tions, respectively. Narrower and darker ellipses represent stronger
associations and significant relationships are indicated by filled grey
cells. Scatterplots of the relationships between core areas in (a),
ROM values in (b), and detections in (c) and (d) and water quality
parameters are presented in Fig. 8
Environ Biol Fish (2020) 103:509–529 523
Author's personal copy

Discussion
Here we use passive and active fish telemetry and ex-
tensive environmental monitoring to investigate appar-
ent habitat preferences by juvenile Central California
Coast steelhead in a large intermittent estuary during
severe drought. We discover that fish exploited a limited
portion of the overall wetted area in the estuary and that
habitat contraction was correlated with degraded water
quality conditions for juvenile O. mykiss. Given the
environmental conditions encountered during the study
period, the strong multicollinearity of physico-chemical
variables (Fig. 7), and known physiological tolerances
and preferences for O. mykiss (Molony 2001) and other
juvenile salmonids (Brett 1979;Brett1995), we focus
our analysis on the combined effects of water depth,
light, temperature, DO, and cover on habitat utilization
areas and movement rates.
The exponential diurnal and nocturnal relationships
between kernel density estimate (KDE) decile endpoint
and corresponding areas for each 10% KDE increment
(Fig. S2–2) indicate that juvenile O. mykiss utilized a
small fraction of the overall wetted habitat during the
study period, especially at night. For example, there was
a 90% chance of fish position occurring in only 10%
(night) to 33% (day) of the overall wetted area in the
Fig. 8 Relationships between physico-chemical parameters and
(a) mean weekly core areas for the pooled sample, (b) mean daily
rate of movement (ROM), (c) total daily detections at station 13,
and (d) total daily detections at station 14. Water temperatures for
a-h represent depth-averaged daily mean values from 0.25 m,
0.75 m, 1.25 m, and 1.75 m water depths at all VPS receiver
locations. Salinity, pH, and DO data for a-h are from station 08
(1.60 m depth). Water quality data for i-p are from their respective
stations at 1.25 m water depths
Environ Biol Fish (2020) 103:509–529
524
Author's personal copy

lower and middle estuary which was the most utilized
region overall according to mobile tracking surveys
(Figs. 4a). While a conservative estimate of fish swim-
ming speed, mean ROM was approximately 0.4 to 0.6
fish lengths·s
−1
when corrected for the relatively slow
(~0.25 mm·day
−1
) and uniform growth rates of the 2013
lagoon-rearing cohort (Huber 2018). At these speeds,
swimming is sustained and juvenile salmonids engage
in routine behaviors like foraging and station holding
(Beamish 1978). Thus, alternating periods of station
holding in central core habitats at night while resting
and feeding forays outside of central core habitats dur-
ing the day most likely produced the observed habitat
use and movement patterns (Figs. 3a, b,6b and SI
Appendices B and C in Huber 2018).
Territoriality, or the tendency for fish to remain in an
isolated and restricted area, is expressed by juvenile
salmonids rearing in freshwater streams (Quinn 2005)
and the segregated spatial patterning (SI Appendix B in
Huber 2018) and relatively small core areas (Table 1)of
individual fish suggests that territorial behavior may
occur in closed IEs as well and warrants further study.
Nocturnal core habitat utilization area for the
pooled sample was more than ten times smaller than
diurnal core areas (Figs. 3a, b,S2–2) and the distinct
habitat use shift suggests that fish may have adopted
more schooling behaviors at night, particularly amongst
complex large woody debris (LWD) structure along the
lagoon’s wetted margin (SI Appendix A, Fig. A7 in
Huber 2018) in a possible effort to reduce vulnerability
to predators while resting.
As mentioned previously, juvenile steelhead
expressed apparent preference for two microhabitat
complexes within the VPS monitoring zone. One was
along the inside bend of the middle estuary and another
at the flood-tide delta region of the lower embayment
(Figs. 3a, b). Within each complex, fish occupied more
peripheral zones closer to the shore at night (Figs. 3a, b).
Both microhabitats were shallow (~1.5 m water depth or
less, Fig. 3), illuminated throughout the water column
during the day, and sheltered from Pacific winds which
are strong in the region (Williams 2014). The expansion
of core and activity space utilization areas during weeks
5, 10, 11, and 14 (Fig. 6a) coincided with increased rates
of movement and utilization of the downstream micro-
habitat complex (SI Appendix C in Huber 2018).
Quiñones and Mulligan (2005) observed that juvenile
O. mykiss preferred marginal estuarine habitats, espe-
cially those associated with overhanging riparian
vegetation in a northern California IE (Smith River
estuary, Del Norte County, CA) also devoid of abundant
in-stream cover. The northeastern-facing rocky cliffs at
Pescadero provided daytime shade and possible protec-
tion for steelhead occupying the lower complex. The
apparent preference for calmer water in both microhab-
itat complexes may result from an aversion to displace-
ment by hydraulic forces which would presumably incur
energetic costs and may make fish more vulnerable to
predators, especially at night while resting (Fig. 6b).
Interestingly, tagged fish expressed distinct avoidance
of the thalweg zone (Figs. 3c,S2–1) despite the presence
of favorable water quality conditions in the epilimnion
(Figs. 2,S1–2). Consequently, O. mykiss in the Pescadero
Lagoon appear to avoid utilizing deep water areas where
the benthic substrate occurs within the dark (Table S1-1,
Fig. S1-3b) and deoxygenated (Figs. 2d,S1–2d) hypo-
limnion. Indeed, abundances of preferred benthic macro-
invertebrate prey for Pescadero juvenile O. mykiss are
more than four times greater at lagoon water depths of
0.65 m or less than at depths of 0.75 m or greater and
abundances are particularly low below the halocline
(Robinson 1993; Martin 1995). These findings are similar
to those of Eby and Crowder (2002) who observed that
fishes occupied all depths when the Neuse River estuary
was fully oxygenated but were restricted to shallow oxy-
genated zones during times of deep water hypoxia.
In addition to vertical habitat compression caused by
poor bottom water quality in the lower and middle estuary
(Fig. 3), lateral habitat compression caused by worsening
water quality-at-depth moving upstream (Figs. 4b,S1-4d)
was also observed. Of all the physico-chemical parame-
ters measured, DO concentrations were most variable and
regularly exceeded stressful (<5.0 mg∙L
−1
) and lethal
(<2.0 mg∙L
−1
) limits for juvenile O. mykiss both at-
depth and upstream (Molony 2001;Figs.2d,4b,S1–2d,
S1-4d). Minimum daily DO at 1.25 m depth was greater
than 5.0 mg∙l
−1
for all 105 study days at the station closest
to the mouth (st02) and hypoxic (≤5.0 mg∙l
−1
) for 35 days
(st13) and 41 days (st14) in the upper lagoon (Figs. 1,S1-
4d). Fish tag transmission detections at the upper estuary
receivers declined linearly when mean daily water tem-
peratures in the Pescadero and Butano arms of the upper
estuary warmed above 18 °C and mean daily DO con-
centrations declined below 7.0 mg·L
−1
(Figs. 1,8i–p). A
combination of diminished growth rewards and
heightened predation risks likely account for reduced
upper estuary utilization by juvenile Pescadero steelhead
during warm and deoxygenated lagoonal conditions.
Environ Biol Fish (2020) 103:509–529 525
Author's personal copy

Using stationary PIT tag antenna, Osterback et al. (2018)
demonstrates that juvenile CCC steelhead in the nearby
Scott Creek IE (Santa Cruz County, CA) repeatedly
moved between warmer lagoon and cooler lower
mainstem creek waters during an extended drought-
induced sandbar closed phase in 2015, suggesting indi-
viduals actively thermoregulated by moving between ad-
jacent habitat types. Pedersen (1987) reports that juvenile
O. mykiss require DO concentrations of approximately
7.0 mg∙L
−1
for proper growth and food conversion effi-
ciency and food consumption declines below 6.0 mg∙L
−1
.
Maximum swimming speeds for juvenile salmonids are
reduced by 30% to 43% at DO concentrations between
5.7 mg∙L
−1
to 5.9 mg∙L
−1
(Jones 1971) and there is little
capacity for anaerobic metabolism below 5.1 mg∙L
−1
(Kutty 1968) which is required for burst swimming
(Drummond and Black 1960) and, thus, predatory escape.
Not surprisingly, zero tagged O. mykiss were detected by
mobile tracking surveys in the Butano Creek arm of upper
estuary beyond ~150 m of the confluence (Figs. 4a)
where lagoon DO levels were lowest (Fig. 4b).
Aside from a major peak at dawn and a minor peak at
dusk, fish movement rates closely tracked daily changes in
luminous flux (Fig. 6b). Approximately 45% less move-
ment occurred at night than during the day and fish moved
at consistently low baseline levels at night (~40 mm·s
−1
;
Fig. 6b). Some movement while resting at night may have
been required to replenish oxygen across the gill boundary
layer (Moyle and Cech 2000). Mean ROM from 6 to 7 h
(x
̅
=94 mm∙s
−1
), when surface light intensities increased
over three orders of magnitude, was more than twice as
fast as mean ROM during the previous 12 h
(x
̅
=44 mm∙s
−1
)(Fig.6b). This elevated movement may
reflect changes in foraging activity (Brett 1995). Indeed,
the appetites of many fish species either peaks at dawn or
dusk or is spread more evenly throughout the day (Boujard
and Leatherland 1992). The crepuscular ROM spikes (Fig.
6b) may be the result of a tradeoff between the ability to
visually feed and reduced predation pressure from visual
predators. Alternatively, invertebrate prey may be more
vulnerable during dawn and dusk (Davis 1978). The larger
and more consistent peak at dawn (Fig. 6b) indicate that
preferred food items may be more available at this time or
that fish may take more foraging risks when hungry after a
nighttime fasting period.
Water temperature appears to have moderated the
effects of time-of-day on fish movement rates. We de-
tected a second degree polynomial relationship between
mean daily ROM for the pooled sample and mean daily
temperature, with maximum ROM (89 mm∙s
−1
)predict-
ed to occur at 17.4 °C and approximately 39% and 49%
less movement was predicted to occur when water tem-
peratures were 13 °C and 21 °C, respectively. The
temperature-ROM relationship observed here (Fig. 8e)
closely resembles the thermal relationships for juvenile
salmonid food consumption, food conversion efficien-
cy, and growth rate (Brett 1979,1995). Levels for all
processes increase to a maximum as water temperature
warms to an optimum value near 17 °C to 18 °C before
declining at higher temperatures (Brett 1979,1995;Fig.
8e). Cold temperatures and slow digestion rates limit
fish feeding activity regardless of food abundance (Brett
1979,1995)and juvenile salmonid appetites and activity
levels decline at hot temperatures. Ration size, scope for
activity, and growth potential are maximized for juve-
nile salmonids (Brett 1979,1995) near the observed
optimum ROM temperature. Maximum ROM occurred
at dawn when temperatures throughout the lagoon were
suboptimal and during the late morning when tempera-
tures were optimal (Figs. 6b,S1–5) and these observa-
tions suggest that juvenile O. mykiss were most active
when potential growth rewards were maximized.
The mobile tracking survey data indicate that the ap-
proximate 0.4 fish tag∙day
−1
decline from November 26,
2013 until the end of the study period December 22 and
concomitant decline in total fish tag detections (Fig. 5) can
be attributed to upstream movement by individuals of the
lagoon-rearing cohort. More than half of all remaining
tagged fish were detected in the Pescadero arm of the far
upper estuary on December 14 after not being observed
there during the first three mobile surveys (Figs. 4a). The
juvenile steelhead located far upstream may have been in
the early stage of a “twice smolting”life history pathway
that is a unique and popular CCC O. mykiss strategy
(Shapovalov and Taft 1954; Hayes et al. 2011). Hayes
et al. (2011) suggest that declining lagoon water quality at
the end of rearing period may have triggered upstream
movement of juvenile O. mykiss in the nearby Scott Creek
IE. Given that DO conditions improved over time as
temperatures declined (Figs. 2a, d,S1–2a,d,S1-4a,d),it
appears that the upstream movement may also be triggered
by declining fall day lengths, low temperatures (Figs. 2a,
S1–2a,S1-4a,S1–5), or a combination of both factors. The
10 fish located farthest upstream in mid-December were
smaller at tagging overall (x
̅
=173 mm FL, range: 139 mm
FL to 201 mm FL) than the seven located closest to the
mouth (x
̅
=182 mm FL, range: 139 mm FL to 206 mm FL)
and the two fish centrally located near the confluence
Environ Biol Fish (2020) 103:509–529
526
Author's personal copy

(x
̅
=174 mm FL) (Fig. 4a). Production of Na
+
,K
+
-ATPase,
while less than the spring peak, is expressed by CCC
O. mykiss during fall and winter and larger individuals
are generally more salt-tolerant (Hayes et al. 2011). Con-
sequently, the larger fish remaining near the mouth in
December may have been more osmotically prepared for
marine life than the upstream fish and, thus, did not
respond to upstream movement cues (Fig. 4a). Marine
survival for O. mykiss asymptotes at large smolt body sizes
(WardandSlaney1988ascitedinMangeland
Satterthwaite 2008) and it is also possible that the
O. mykiss individuals detected far upstream had not
attained threshold sizes during a potential late fall life
history decision window. Interestingly, three of the fish
detected far upstream on 14-Dec (T06, T30, T33) subse-
quently returned to the VPS array which suggests that
upstream movement is not entirely unidirectional for all
fish or that some fish may switch life history pathways
during critical decision window periods. If these interpre-
tations are correct, smaller arrival size, slower growth, later
estuarine arrival, or a combination of these factors may
promote twice smolting life history expression for CCC
steelhead.
Our study adds to a growing body of research about the
consequences of extended drought for CCC steelhead
(e.g., Smith 1990; Jankovitz 2015; Osterback et al.
2018). The historic California drought persisted until wa-
ter year 2016 and Jankovitz (2015) reports that poor water
quality during summer 2015 “terminated all O. mykiss
production and probably survival throughout the
[Pescadero] lagoon”. In 2015, sandbar formation occurred
in late July during base freshwater inflows. Initially a
shallow brackish water epilimnion formed but was lost
by late summer when saline conditions, hot water temper-
atures, and very low DO concentrations prevailed
throughout the water column (Jankovitz 2015).
Dewatered channels upstream of the Pescadero IE during
summer 2015 likely impeded upstream movement and
behavioral thermoregulation (Osterback et al. 2018).
While quasi-annual acute breach-induced fish kills
have been observed at Pescadero since the mid-1990s
(Sloan 2006; Smith 2009; Largier et a. 2015), the situation
in 2015 represents the first documented case of ecosystem
function loss caused by chronic degradation of water
quality during the sandbar-closed estuarine state. Such
extreme climatic events are expected to occur more fre-
quently in the future (Langridge 2018) and the informa-
tion provided here can help guide resource management
decisions. In particular, the findings of this study and
Jankovitz (2015) highlight the paramount importance of
providing a sufficiently oxygenated surface water layer
for juvenile O. mykiss rearing in California IEs during
prolonged drought-induced sandbar closure. The extent of
effective habitat area should increase with increasing
depth of the freshening epilimnion, especially in shallow
systems like the Pescadero IE (Fig. 3c). Water conserva-
tion practices and other actions that boost lagoon infill
are also expected to promote estuary-mainstem connec-
tivity and access to thermal refugia for juvenile O. mykiss
during severe drought (Osterback et al. 2018). Additional
actions that reduce biological oxygen demands and light
attenuation, such as limiting anthropogenic nutrient inputs
to the estuary, should further increase the area of useable
habitat for juvenile salmonids and other estuarine fishes in
impacted systems like the Pescadero IE. Habitat quality in
the cover-poor Pescadero IE could be improved immedi-
ately by strategic additions of LWD, particularly in loca-
tions that are easily accessed from preferred feeding
grounds (Grand and Dill 1997). Lastly, conservation
managers should consider mechanical sandbar breaching
as a means to maintain tolerable water quality conditions
for juvenile steelhead in Central California Coast IEs
during extreme drought conditions, at least until early
winter when “twice smolting”fish have moved upstream
and mortality risk for the population is more spread in
space.
Acknowledgements This research was made possible by gen-
erous volunteers and we are particularly grateful for the assiduous
assistance provided by David Kammerer, Frank Hubinsky, and
Pierre Tardif. We are especially thankful to Mark Stacey (UC
Berkeley) for use of CTDs and mooring materials and Megan
Williams (UC Berkeley) for providing of the bathymetry map.
Special thanks also to Walter Heady (UC Santa Cruz) for use of
the VR100 mobile receiver and VH165 hydrophone and to Esri™
and the Geospatial Innovation Facility at UC Berkeley for free use
of ArcGIS. We appreciate the helpful study design recommenda-
tions provided by Dale Webber (VEMCO) and the free loan of two
company receivers. Special thanks to the California Department of
Parks and Recreation the San Francisco Regional Water Quality
Control Board for use of water quality sondes. We are thankful to
Nicholas Demetras (National Oceanic and Atmospheric Adminis-
tration, NOAA) for help with tagging and Sean Hayes (NOAA)
who provided advice at multiple stages. We are grateful for the
helpful reviews of earlier versions of the manuscript by Patrick
Samuel (California Trout) and two anonymous referees. Funding
was provided by the UC Water Resources Center and the Depart-
ments of Environmental Science, Policy, and Management and
Civil Engineering at UC Berkeley. All handling and care proce-
dures used were reviewed and approved by the UC Berkeley
Animal Care and Use Committee (Animal Use Protocol R343-
0612).
Environ Biol Fish (2020) 103:509–529 527
Author's personal copy

References
Beamish FWH (1978) Swimming capacity. In: Hoar WS, Randall
DJ (eds) Fish physiology, vol 7. Academic Press, New York,
pp 101–187
Beck N, Freund E, & Mathias M (2006) Comparative lagoon
ecological assessment project (CLEAP). Santa Cruz
County, California. October 2006
Bond MH, Hayes SA, Hanson CV, MacFarlane RB (2008) Marine
survival of steelhead (Oncorhynchus mykiss) enhanced by a
seasonally closed estuary. Can J Fish Aquat Sci 65:2242–
2252
Boujard T, Leatherland JF (1992) Circadian rhythms and feeding
time in fishes. Environ Biol Fish 35:109–131. https://doi.
org/10.1139/F08-131
Bozkurt Frucht S, Napolitano M, Ponton J (2018) Total Maximum
Daily Load for Sediment and and Habitat Enhancement Plan
for Pescadero-Butano Watershed. Final Staff Report for
Proposed Basin Plan Amendment. California Regional
Water Quality Control Board, San Francisco Bay Region.
December 11, 2018
Brett JR (1979) Environmental factors and growth. In: Hoar WS,
Randall DJ (eds) Fish physiology, vol 8. Academic Press,
New York, pp 599–675
Brett JR (1995) Energetics. In: Groot C, Margolis L, Clarke WC
(eds) Physiological ecology of Pacific Salmon. University of
British Columbia Press, Vancouver, pp 3–68
Clark R, O'Connor K (2019) A systematic survey of bar-built
estuaries along the California coast. Estuar Coast Shelf Sci
226:106285. https://doi.org/10.1016/j.ecss.2019.106285
Davis ST (1978) Diel activity of benthic crustaceans in the
Columbia River estuary. MS thesis, Oregon State
University, Corvallis, OR
Drummond GI, Black EC (1960) Comparative physiology: fuel of
muscle metabolism. Annu Rev Physiol 22:169–190.
https://doi.org/10.1146/annurev.ph.22.030160.001125
Eby LA, Crowder LB (2002) Hypoxia-based habitat compression
in the Neuse River estuary: context-dependent shifts in be-
havioral avoidance thresholds. Can J Fish Aquat Sci 59:952–
965. https://doi.org/10.1139/f02-067
Espinoza M, Farrugia TJ, Webber DM, Smith F, Lowe CG
(2011a) Testing a new acoustic telemetry technique to quan-
tify long-term, fine-scale movements of aquatic animals. Fish
Res 108:364–371. https://doi.org/10.1016/j.
fishres.2011.01.011
Espinoza M, Farrugia TJ, Lowe CG (2011b) Habitat use, move-
ments and site fidelity of the gray smooth-hound shark
(Mustelus californicus gill 1863) in a newly restored southern
California estuary. J Exp Mar Biol Ecol 401:63–74.
https://doi.org/10.1016/j.jembe.2011.03.001
Fuller JA (2011) Extended residency and movement behavior of
juvenile steelhead (Oncorhynchus mykiss) in the Russian
River estuary, California. MS thesis, Humboldt State
University, Arcata, CA
Grand TC, Dill LM (1997) The energetic equivalence of cover to
juvenile coho salmon (Oncorhynchus kisutch): ideal free
distribution theory applied. Behav Ecol 8:437–447.
https://doi.org/10.1093/beheco/8.4.437
Hammer Ø, Harper DAT, Ryan PD (2001) PAST: paleontological
statistics software package for education and data analysis.
Paleontologia Electronica 4(1):9 pp Available from
http://folk.uio.no/ohammer/past/
Hayes SA, Bond MH, Hanson CV, Freund EV, Smith JJ,
Anderson EC, Amman AJ, MacFarlane RB (2008)
Steelhead growth in a small Central California watershed:
upstream and estuarine rearing patterns. Trans Am Fish Soc
137:114–128. https://doi.org/10.1577/t07-043.1
Hayes SA, Bond MH, Hanson CV, Jones AW, Ammann AJ,
Harding JA, Collins AL, Perez J, MacFarlane RB (2011)
Down, up, down and “smolting”twice? Seasonal movement
patterns by juvenile steelhead (Oncorhynchus mykiss)ina
coastal watershed with a bar closing estuary. Can J Fish
Aquat Sci 68:1341–1350. https://doi.org/10.1139/f2011-062
Hilborn R, Quinn TP, Schindler DE, Rogers DE (2003)
Biocomplexity and fisheries sustainability. Proc Natl Acad
Sci 100:6564–6568. https://doi.org/10.1073
/pnas.1037274100
Huber ER (2018) The management and ecology of Oncorhynchus
spp. and other estuarine-dependent native California fishes in
artificially and naturally disconnected aquascapes. PhD the-
sis, University of California Berkeley, Berkeley, CA
Huber ER, Carlson SM (2020) Environmental correlates of fine-
scale juvenile steelhead trout (Oncorhynchus mykiss)habitat
use and movement patterns in an intermittent estuary during
drought, Dryad, Dataset. https://doi.org/10.6078/D1X12R
Hussey NE, Kessel ST, Aarestrup K, Cooke SJ, Cowley PD, Fisk
AT, Harcourt RG, Holland KN, Iverson SJ, Kocik JF, Mills
Fleming JE, Whoriskey FG (2015) Aquatic animal telemetry:
a panoramic window into the underwater world. Science 348:
1255642. https://doi.org/10.1126/science.1255642
Jankovitz J (2015) Summary of fish sampling and water quality
monitoring: Pescadero Creek Lagoon 2015
Jones DR (1971) The effect of hypoxia and anaemia on the
swimming performance of rainbow trout (Salmo gairdneri).
J Exp Biol 55:541–551
Kutty MN (1968) Respiratory quotients in goldfish and rainbow
trout. J Fish Res Board Can 25(8):1689–1728
Langridge R (2018) Central Coast Summary Report. California’s
Fourth Climate Change Assessment. Publication number:
SUM-CCCA4-2018-006
Largier JL, Aiello IW, Jacobs D, Lacy J, Pallud C, Stacey MT,
Carlson SM, Huber E, Bowles CM (2015) Report of
Pescadero Lagoon Science Panel. 48 pp.
Mangel M, Satterthwaite WH (2008) Combining proximate and
ultimate approaches to understand life history variation in
salmonids with application to fisheries, conservation, and
aquaculture. Bull Mar Sci 83:107–130
Martin JA (1995) Food habits of some estuarine fishes in a small,
seasonal Central California lagoon. MS thesis, San Jose State
University, San Jose, CA
Molony B (2001) Environmental requirements and tolerances of
rainbow trout (Oncorhynchus mykiss) and brown trout
(Salmo trutta) with special reference to Western Australia: a
review. Department of Fisheries, Government of Western
Australia
Moyle PB, Cech JJ (2000) Fishes: an introduction to ichthyology
(4th ed), Prentice Hall, Upper Saddle River, NJ
Moyle PB, Lusardi R, Samuel P (2017) State of the Salmonids II:
Fish in hot water - status, threats and solutions for California
Environ Biol Fish (2020) 103:509–529
528
Author's personal copy

salmon, steelhead, and trout. Commissioned by California
Trout. San Francisco, CA. Available from http://caltrout.
org/sos/
Osterback AMK, Frechette DM, Shelton AO, Hayes SA, Bond
MH, Shaffer SA, Moore JW (2013) High predation on small
populations: avian predation on imperiled salmonids.
Ecosphere 4:1–21. https://doi.org/10.1890/es13-00100.1
Osterback AMK, Kern CH, Kanawi EA, Perez JM, Kiernan JD
(2018) The effects of early sandbar formation on the abun-
dance and ecology of coho salmon (Oncorhynchus kisutch)
and steelhead trout (Oncorhynchus mykiss) in a Central
California coastal lagoon. Can J Fish Aquat Sci 75:2184–
2197. https://doi.org/10.1139/cjfas-2017-0455
Pedersen CL (1987) Energy budgets for juvenile rainbow trout at
various oxygen concentrations. Aquaculture 62:289–298.
https://doi.org/10.1016/0044-8486(87)90171-2
Perissinotto R, Stretch DD, Whitfield AK, Adams JB, Forbes AT,
Demetriades NT (2010) Temporarily open/closed estuaries in
South Africa. Nova Science Publishers, New York
Quinn TP (2005) The behavior and ecology of Pacific salmon and
trout. University of Washington Press, Seattle
Quiñones RM, Mulligan TJ (2005) Habitat use by juvenile salmo-
nids in the Smith River estuary, California. Trans Am Fish
Soc 134:1147–1158. https://doi.org/10.1577/t04-092.1
Rechetelo J,Grice A, Reside AE, Hardesty BD, Moloney J (2016)
Movement patterns, home range size and habitat selection of
an endangered resource tracking species, the black-throated
finch (Poephila cincta cincta). PLoS One 11(11):e0167254.
https://doi.org/10.1371/journal.pone.0167254
Richards CM, van Puffelen JL, Pallud C (2018) Effects of sedi-
ment resuspension on the oxidation of acid-volatile sulfides
and release of metals (iron, manganese, zinc) in Pescadero
estuary (CA, USA). Environ Toxicol Chem 37:993–1006.
https://doi.org/10.1002/etc.4047
Robinson MA (1993) The distribution and abundance of benthic
and epibenthic macroinvertebrates in a small, seasonal
Central California lagoon. MS thesis, San Jose State
University, San Jose, CA
Rooker JR, Dance MA, Wells RD, Quigg A, Hill RL, Appeldoorn
RS, Padovani Ferreira B,Boswell KM, Sanchez PJ, Moulton
DL, Kitchens LL (2018) Seascape connectivity and the in-
fluence of predation risk on the movement of fishes
inhabiting a back-reef ecosystem. Ecosphere 9:e02200.
https://doi.org/10.1002/ecs2.2200
Roy R, Beguin J, Argillier C, Tissot L, Smith F, Smedbol S, De-
Oliveira E (2014) Testing the VEMCO positioning system:
spatial distribution of the probability of location and the
positioning error in a reservoir. Anim Biotelemetry 2:1–13.
https://doi.org/10.1186/2050-3385-2-1
Ruckelshaus MH, Levin P, Johnson JB, Kareiva PM (2002) The
Pacific salmon wars: what science brings to the challenge of
recovering species. Annu Rev Ecol Evol Syst 33:665–706.
https://doi.org/10.1146/annurev.ecolsys.33.010802.150504
Satterthwaite WH, Hayes SA, Merz JE, Sogard SM, Frechette
DM, Mangel M (2012) State-dependent migration timing
and use of multiple habitat types in anadromous salmonids.
Trans Am Fish Soc 141:781–794. https://doi.org/10.1080
/00028487.2012.675912
Schindler DE, Hilborn R, Chasco B, Boatright CP, Quinn TP,
Rogers LA, Webster MS (2010) Population diversity and the
portfolio effect in an exploited species. Nature 465:609–612.
https://doi.org/10.1038/nature09060
Seaman DE, Powell RA (1996) An evaluation of the accuracy of
kernel density estimators for home range analysis. Ecology
77:2075–2085. https://doi.org/10.2307/2265701
Shapovalov L, Taft AC (1954) The life histories of the steelhead
rainbow trout (Salmo gairdneri gairdneri) and silver salmon
(Oncorhynchus kisutch) with special reference to Waddell
Creek, California, and recommendations regarding their
management. California Department of Fish and Game,
Sacramento
Silverman BW (1986) Density estimation for statistics and data
analysis, vol 26. Chapman & Hall, London
Sloan RM (2006) Ecological investigations of a fish kill in
Pescadero Lagoon, California. MS thesis, San Jose State
University, San Jose, CA
Smith JJ (1990) The effects of sandbar formation and inflows on
aquatic habitat and fish utilization in Pescadero, San
Gregorio, Waddell and Pomponio Creek estuary/lagoon sys-
tems, 1985–1989. Report prepared by San Jose State
University under interagency agreement 84–04-324 for the
California Department of Parks and Recreation, Sacramento,
CA
Smith KA (2009) Inorganic chemical oxygen demand of re-
suspended sediments in a bar-built lagoon. MS thesis, San
Jose State University, San Jose, CA
Smith F (2013) Understanding HPE in the VEMCO positioning
system (VPS). Available from http://vemco.com/wp-
content/uploads/2013/09/understanding-hpe-vps.pdf
Whitfield AK (1995) Mass mortalities of fish in South African
estuaries. South Afr J Aquat Sci 21:29–34. https://doi.
org/10.1080/10183469.1995.9631360
Williams ME (2014) Hydrodynamics and salt dispersion in inter-
mittently closed bar-built estuaries. PhD thesis, University of
California Berkeley, Berkeley, CA
Publisher’snoteSpringer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional
affiliations.
Environ Biol Fish (2020) 103:509–529 529
Author's personal copy