Partial migration: growth varies between resident and migratory fish

Downloaded from http://rsbl.royalsocietypublishing.org/ on July 7, 2015
Marine biology
rsbl.royalsocietypublishing.org
Partial migration: growth varies between
resident and migratory fish
Bronwyn M. Gillanders1, Christopher Izzo1, Zoe¨ A. Doubleday1 and Qifeng Ye2
1
Research
Cite this article: Gillanders BM, Izzo C,
Doubleday ZA, Ye Q. 2015 Partial migration:
growth varies between resident and migratory
fish. Biol. Lett. 11: 20140850.
http://dx.doi.org/10.1098/rsbl.2014.0850
Received: 16 October 2014
Accepted: 2 February 2015
Subject Areas:
ecology
Keywords:
migration, residency, partial migration, fish,
otolith chemistry, otolith growth
Southern Seas Ecology Laboratories, School of Biological Sciences, University of Adelaide, Adelaide,
South Australia 5005, Australia
2
South Australian Research Development Institute Aquatic Sciences, Adelaide, West Beach,
South Australia 5024, Australia
Partial migration occurs in many taxa and ecosystems and may confer survival
benefits. Here, we use otolith chemistry data to determine whether fish from a
large estuarine system were resident or migratory, and then examine whether
contingents display differences in modelled growth based on changes in
width of otolith growth increments. Sixty-three per cent of fish were resident
based on Ba : Ca of otoliths, with the remainder categorized as migratory,
with both contingents distributed across most age/size classes and both
sexes, suggesting population-level bet hedging. Migrant fish were in slightly
better condition than resident fish based on Fulton’s K condition index.
Migration type (resident versus migratory) was 56 times more likely to explain
variation in growth than a model just incorporating year- and age-related
growth trends. While average growth only varied slightly between resident
and migratory fish, year-to-year variation was significant. Such dynamism in
growth rates likely drives persistence of both life-history types. The complex
relationships in growth between contingents suggest that management of
species exhibiting partial migration is challenging, especially in a world subject
to a changing climate.
1. Introduction
Author for correspondence:
Bronwyn M. Gillanders
e-mail: [email protected]
One contribution to the special feature
‘Frontiers in marine movement ecology’
organized by Lee Fuiman, Benjamin Walther
and Pablo Munguia.
Electronic supplementary material is available
at http://dx.doi.org/10.1098/rsbl.2014.0850 or
via http://rsbl.royalsocietypublishing.org.
Invited to commemorate 350 years
of scientific publishing at the Royal
Society.
Migration is a powerful force in shaping the distribution and abundance of
animals in space and time. It can range from daily vertical migrations of zooplankton and fish [1] to extensive annual migrations of Arctic terns from the high Arctic
to the Southern Ocean [2]. Even within marine waters, vast oceanic journeys are
undertaken by organisms, such as humpback whales [3] and green turtles [4],
which can migrate thousands of kilometres. Although variation in migratory tendency occurs between species and between populations within a species, within
population variation has been less well studied [5].
Populations may comprise both resident and migratory individuals, which
is known as partial migration. Partial migration occurs in all major vertebrate
groups [5] and some invertebrate taxa [1], as well as across all major ecosystems
[6]. Bird migration has been most widely researched, and among fish taxa, salmonids and Atlantic cod are the predominant species investigated [7]. Partial
migration may influence population dynamics and persistence [8], as it may
confer survival benefits through reduced predation risk (e.g. [9]) and/or
increased food availability [6]. Although these factors are frequently cited as
the ultimate cause of movement, the proximate cause is debated. Effects of
partial migration on demographic parameters such as growth have been investigated but usually only for a portion of the life history or as an integrated
signature over an individual’s lifespan.
Here, we first use otolith Ba : Ca profiles to categorize fish as either resident
or migratory, a technique that has been previously validated (e.g. [10,11]). Next,
we determine whether the two contingent types lead to differences in growth
based on relative changes in widths of otolith growth increments.
& 2015 The Author(s) Published by the Royal Society. All rights reserved.
Downloaded from http://rsbl.royalsocietypublishing.org/ on July 7, 2015
4
2
rsbl.royalsocietypublishing.org
Ba : Ca
3
2
1
0
50
100
150
distance along otolith
200
0
50
100
150
distance along otolith
200
Figure 1. Representative migrant and resident Ba : Ca (mmol mol21) profiles for co-occurring black bream individuals aged 5þ (LHS) and 10þ (RHS); solid line,
resident; dashed line, migrant. The distance along the otolith has been standardized to a common value with zero representing the early life and more than 200
representing the collection time. The peaks in Ba : Ca suggest movement of fish.
Table 1. Otolith samples examined and number of increments analysed to assess variation in growth. Seven fish were unable to be classified as migrant
or resident.
movement type
no. fish
age range
total length range (mm)
year of capture range
no. increment
measurements
all fish
migrant
173
63
5 – 32
5 – 17
283– 470
300– 470
2008– 2012
2008– 2012
1332
630
resident
104
5 – 32
283– 462
2008– 2012
659
2. Material and methods
Black bream (Acanthopagrus butcheri) were captured from the
Murray River estuary, which is the terminus of Australia’s largest
river, the Murray – Darling. Fish were collected from 14 sites
throughout the estuary between October 2008 and September
2013 (table 1). The salinity of each collection site was estimated
from the WaterConnect website or measured directly while in
the field.
All fish were measured (total length, mm), weighed (body
weight, g) and sexed (electronic supplementary material,
table S1). Gonads were removed and weighed (gonad weight,
g) and sagittal otoliths extracted, cleaned and stored. Fulton’s
K condition factor was estimated by dividing body weight
minus gonad weight by length3 and multiplying by a constant.
Otoliths were prepared for elemental analysis and analysed
using a laser ablation inductively coupled plasma-mass spectrometer (Resonetics). Sectioned otoliths were examined under a
compound microscope (Leica DMLB), photographed and growth
increment widths measured across the otolith. Thus, a temporally
resolved proxy of somatic growth was obtained [12] (see the electronic supplementary material for more details on otolith
preparation and elemental analysis, as well as growth analyses).
A series of mixed effects models were used to investigate
differences in annual growth among resident and migratory fish
based on increment widths. A two-stage process was adopted
where first the random effect structure was optimized, followed
by the fixed effects structure (see the electronic supplementary
material for more details on growth modelling). The relative
support for each model was assessed using Akaike’s information criterion corrected for small sample sizes [13]. Analyses
were performed in R [14].
3. Results
All locations comprised both resident and migrant fish
where more than two fish were collected (with the exception of
a single location) (figures 1 and 2a). In addition, residents and
migrants were caught from freshwater to hypersaline waters
(0.3–74 ppt), although the majority of fish were collected from
marine salinities. Age at first migration ranged from 0 to 5þ,
with the majority of fish (94%) moving before they were 2þ. Of
the fish analysed, 62% were resident and 38% were migratory,
with both sexes showing similar patterns. In addition, migrants
and residents were found for most size and age classes, but
migrant fish tended to be larger and older than resident fish (electronic supplementary material, figure S1). Condition of fish as
estimated by Fulton’s K was marginally greater in migrant fish
than in resident fish, but was not significantly different
(electronic supplementary material, table S2).
Inclusion of migration type in the fixed model structure
was the highest ranked model and was 56 times more
likely to explain variation in growth among all fish than a
model just incorporating age (electronic supplementary
material, table S3). Resident and migrant fish did not share
a common age–growth relationship and adding a contingent-specific directional growth term did not improve the
model (electronic supplementary material, table S4). When
modelled independently, the best explanatory mixed effects
model structure differed slightly between resident and
migrant fish (table 2). The random year effects show different
growth trajectories for resident and migrant fish (figure 2b).
Biol. Lett. 11: 20140850
0
Downloaded from http://rsbl.royalsocietypublishing.org/ on July 7, 2015
For both resident and migrant fish, predicted growth
decreased with increasing age; however, the rate of decrease
was greater in migrant fish than resident fish (figure 2c). For
migrant fish, age at capture had a positive influence on
growth, which suggests that older samples represent faster
growing fish. For resident fish, the growth trend was
representative of the entire sample.
120
80
40
0
1900
predicted growth (mm)
(b)
predicted growth (mm)
2000
2005
growth year
2010
1995
2000
2005
growth year
2010
0.8
0.4
0
–0.4
–0.8
1900
(c)
1995
0.4
0.3
0.2
0.1
0
0
5
10
15
20
age (years)
25
30
Figure 2. (a) Numbers of increments representing each growth year. (b) Predicted growth based on year random effect estimates (þs.e.). Only years
with more than or equal to five increment measurements are shown (see
dashed line in (a)). Dashed line represents the mean growth of the time
series (model intercept). (c) Predicted growth based on age fixed effect estimates (+95% CI). Year and age effects are derived from a base model
including age, random intercept for year and random age slopes for each
random fishID intercept for each contingent type. Black columns and lines,
migrant fish; grey columns and lines, resident fish.
Throughout the late 1990s and early 2000s resident fish had
increased growth compared with migrant fish, but this changed around 2005 when growth of migrant fish increased.
There was clear evidence for resident and migratory contingents within the black bream population. Such flexibility in
life history has now been identified in many fish taxa and
reflects a balance between the costs and benefits of migration
in terms of fitness. Fish that migrate may have access to
increased food availability, leading to greater growth and
increased fitness. Fish may also move to avoid adverse environmental conditions. Indeed, conditions within the River Murray
estuary deteriorated as the drought intensified and salinities of
up to 120 ppt were recorded [15]. Although there was variation
in salinity throughout the estuary, the whole estuary experienced some impact from drying conditions; therefore growth
of both migrant and resident fish was likely affected. Deteriorating conditions may explain the reduction in growth of
resident black bream through time. At the same time, otolith
growth of migratory fish increased slightly, presumably reflecting the greater ability of migratory fish to find more favourable
conditions and source food.
Our study suggested small differences in average growth
of fish over the chronological time series, but large differences
in year-to-year variation, as well as age-specific variation,
especially after the first few years of life. Mixed results have
been found in relation to growth variation among resident
and migratory fish of other species. Migratory white perch
had higher juvenile and adult growth rates than resident
fish [16,17], but a study on southern flounder found no
advantage of specific residency periods based on several
growth proxies [18]. The latter study only focused on first
year growth.
Minimal overall variation in growth between contingents
may also be expected from a biological and evolutionary
point of view. Dynamism in year-to-year growth variation
probably drives persistence of both contingent types and
the ratio of migrants to residents may be the outcome of
population-level bet hedging. Growth variation may also be
important for understanding resilience of populations to
natural and anthropogenic change. One strategy may be
more favourable under some conditions (e.g. migrants in
drought conditions), whereas an alternative strategy may be
more favourable under a different regime (e.g. [16,19]).
Analysis of growth increments allowed continuous
annually resolved growth histories of the two migratory contingents to be reconstructed. As otolith growth reflects fish
growth, the otolith chronologies are accurate proxies of population-level fluctuations in somatic growth. Further work is
required to determine how growth may influence population
viability, including reproductive output and survival [20].
Our study highlights complex relationships in growth between
resident and migratory contingents, making management of
species exhibiting partial migration challenging, especially in
a world subject to a changing climate.
Biol. Lett. 11: 20140850
4. Discussion
3
rsbl.royalsocietypublishing.org
no. increments
(a)
Downloaded from http://rsbl.royalsocietypublishing.org/ on July 7, 2015
(i) fixed
covariate
(ii) random
t-value
covariate
variance (s.d.)
corr
intercept
age
22.675 (0.073)
20.216 (0.009)
236.860
223.070
fishID
agejfishID
0.038 (0.194)
0.005 (0.073)
0.830
movement type
20.058 (0.030)
21.920
year
year class
0.065 (0.254)
0.022 (0.148)
residual
0.047 (0.216)
(a) all fish
(b) resident fish
intercept
22.609 (0.043)
260.97
fishID
0.030 (0.173)
age
20.211 (0.011)
219.18
agejfishID
year
0.006 (0.076)
0.018 (0.135)
year class
residual
0.002 (0.045)
0.055 (0.235)
fishID
agejfishID
0.010 (0.098)
0.001 (0.033)
year
residual
0.026 (0.162)
0.043 (0.208)
0.804
(c) migrant fish
intercept
age
age at capture
22.580 (0.043)
20.159 (0.009)
260.66
217.94
0.026 (0.010)
2.68
Ethics statement. Otoliths were available from archived collections
which had been previously collected via SARDI Aquatic Sciences
and adhered to South Australian state laws.
Data accessibility. Data can be found at Dryad: http://dx.doi.org/10.
5061/dryad.885sn.
Acknowledgements. We thank SARDI Aquatic Sciences staff and Vidya
Kare for preparing otoliths and Aoife McFadden at Adelaide
Microscopy for assistance with LA ICP-MS. We thank Ben Walther,
20.558
John Morrongiello and two anonymous reviewers for their
constructive comments.
Author contributions. B.M.G. designed the study, Q.Y., C.I. and Z.A.D. collected and analysed data, B.M.G., C.I., Z.A.D. and Q.Y. wrote the paper.
Funding statement. This work was financially supported by an Australian
Research Council (ARC) Discovery Project and an ARC Future
Fellowship, both to B.M.G.
Conflict of interests. We have no competing interests.
References
1.
2.
3.
4.
Hansson L-A, Hylander S. 2009 Size-structured
risk assessments govern Daphnia migration.
Proc. R. Soc. B 276, 331 –336. (doi:10.1098/rspb.
2008.1088)
Egevang C, Stenhouse IJ, Phillips RA, Petersen A,
Fox JW, Silk JRD. 2010 Tracking of Arctic terns
Sterna paradisaea reveals longest animal migration.
Proc. Natl Acad. Sci. USA 107, 2078– 2081. (doi:10.
1073/pnas.0909493107)
Stevick PT, Neves MC, Johansen F, Engel MH, Allen
J, Marcondes MCC, Carlson C. 2011 A quarter of a
world away: female humpback whale moves
10 000 km between breeding areas. Biol. Lett. 7,
299–302. (doi:10.1098/rsbl.2010.0717)
Read TC, Wantiez L, Werry JM, Farman R, Petro G,
Limpus CJ. 2014 Migrations of green turtles
(Chelonia mydas) between nesting and foraging
grounds across the Coral Sea. PLoS ONE 9, e100083.
(doi:10.1371/journal.pone.0100083)
5.
6.
7.
8.
9.
Chapman BB, Bronmark C, Nilsson J-A, Hansson L-A.
2011 The ecology and evolution of partial
migration. Oikos 120, 1764–1775. (doi:10.1111/j.
1600-0706.2011.20131.x)
Brodersen J, Adahl E, Bronmark C, Hansson L-A.
2008 Ecosystem effects of partial fish migration in
lakes. Oikos 117, 40 –46. (doi:10.1111/j.2007.00301299.16118.x)
Chapman BB, Skov C, Hulthen K, Brodersen J,
Nilsson PA, Hansson LA, Bronmark C. 2012 Partial
migration in fishes: definitions, methodologies and
taxonomic distribution. J. Fish Biol. 81, 479 –499.
(doi:10.1111/j.1095-8649.2012.03349.x)
Kerr LA, Secor DH, Piccoli PM. 2009 Partial
migration of fishes as exemplified by the estuarinedependent white perch. Fisheries 34, 114– 123.
(doi:10.1577/1548-8446-34.3.114)
Skov C, Chapman BB, Baktoft H, Brodersen J,
Bronmark C, Hansson L-A, Hulthen K, Nilsson PA.
2013 Migration confers survival benefits against
avian predators for partially migratory freshwater
fish. Biol. Lett. 9, 20121178. (doi:10.1098/rsbl.
2012.1178)
10. Elsdon TS, Gillanders BM. 2005 Alternative lifehistory patterns of estuarine fish: barium in
otoliths elucidates freshwater residency.
Can. J. Fish Aquat. Sci. 62, 1143– 1152. (doi:10.
1139/f05-029)
11. Hamer PA, Jenkins GP, Coutin P. 2006 Barium
variation in Pagrus auratus (Sparidae) otoliths: a
potential indicator of migration between an
embayment and ocean waters in south-eastern
Australia. Estuar. Coast Shelf Sci. 68, 686 –702.
(doi:10.1016/j.ecss.2006.03.017)
12. Doubleday ZA, Izzo C, Haddy JA, Lyle JM, Ye Q,
Gillanders BM. Submitted. Long-term patterns in
estuarine fish growth across two climatically
divergent regions. Oecologia
Biol. Lett. 11: 20140850
estimate (s.e.)
4
rsbl.royalsocietypublishing.org
Table 2. Optimal model parameter estimates and test statistics describing (i) fixed and (ii) random sources of growth variation in black bream. Growth
modelling was limited to years with more than or equal to five increment measurements (1995– 2011). Random age slopes for each individual are denoted by
‘agejfishID’. Corr, correlation statistic. The coefficient estimate for movement type in (a) relates to resident fish.
Downloaded from http://rsbl.royalsocietypublishing.org/ on July 7, 2015
143, 721–731. (doi:10.1080/00028487.2014.
892535)
19. 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. (doi:10.1038/
nature09060)
20. Morrongiello JR, Crook DA, King AJ, Ramsey DSL,
Brown P. 2011 Impacts of drought and predicted
effects of climate change on fish growth in
temperate Australian lakes. Glob. Change Biol.
17, 745 –755. (doi:10.1111/j.1365-2486.2010.
02259.x)
5
Biol. Lett. 11: 20140850
16. Kraus RT, Secor DH. 2004 Dynamics of white perch
Morone americana population contingents in the
Patuxent River estuary, Maryland, USA. Mar.
Ecol. Prog. Ser. 279, 247 –259. (doi:10.3354/
meps279247)
17. Kerr LA, Secor DH. 2009 Bioenergetic trajectories
underlying partial migration in Patuxent River
(Chesapeake Bay) white perch (Morone americana).
Can. J. Fish Aquat. Sci. 66, 602–612. (doi:10.1139/
f09-027)
18. Nims MK, Walther BD. 2014 Contingents of
southern flounder from subtropical estuaries
revealed by otolith chemistry. Trans. Am. Fish Soc.
rsbl.royalsocietypublishing.org
13. Burnham KP, Anderson DR. 2004 Multimodel
inference: understanding AIC and BIC in model
selection. Sociol. Methods Res. 33, 261 –304.
(doi:10.1177/0049124104268644)
14. R Development Core Team. 2011 R: a language and
environment for statistical computing. Vienna,
Austria: R Foundation for Statistical Computing. See
http://www.R-project.org/.
15. Gillanders BM, Munro AR. 2012 Hypersaline
waters pose new challenges for reconstructing
environmental histories of fish based on
otolith chemistry. Limnol. Oceanogr.
57, 1136 –1148. (doi:10.4319/lo.2012.57.4.1136)