South temperate birds have higher apparent adult survival than

advertisement
Journal of Avian Biology 45: 493–500, 2014
doi: 10.1111/jav.00454
© 2014 The Authors. Journal of Avian Biology © 2014 Nordic Society Oikos
Subject Editor: Wesley M. Hochachka. Accepted 10 March 2014
South temperate birds have higher apparent adult survival than
tropical birds in Africa
Penn Lloyd, Fitsum Abadi, Res Altwegg and Thomas E. Martin­
P. Lloyd (penn@baamecology.com), Percy FitzPatrick Inst., DST/NRF Centre of Excellence, Univ. of Cape Town, P/Bag X3, Rondebosch, 7701,
South Africa, and Biodiversity Assessment and Management Pty Ltd, PO Box 1376, Cleveland 4163, Australia. – F. Abadi, South African
National Biodiversity Inst., P/Bag X7, Claremont 7735, South Africa, and School of Statistics and Actuarial Science, Univ. of Witwatersrand,
Johannesburg, P/Bag X3, Wits 2050, South Africa. – R. Altwegg, South African National Biodiversity Inst., P/Bag X7, Claremont 7735, South
Africa, and African Climate and Development Initiative, Univ. of Cape Town, South Africa, and Centre for Statistics in Ecology, Environment
and Conservation, Dept of Statistical Sciences, Univ. of Cape Town, Rondebosch 7701, South Africa. – T. E. Martin, U. S. Geological Survey
Montana Cooperative Wildlife Research Unit, Univ. of Montana, Missoula, MT 59812, USA.­
Life history theory predicts an inverse relationship between annual adult survival and fecundity. Globally, clutch size
shows a latitudinal gradient among birds, with south temperate species laying smaller clutches than north temperate
species, but larger clutches than tropical species. Tropical birds often have higher adult survival than north temperate
birds associated with their smaller clutches. However, the prediction that tropical birds should also have higher adult
survival than south temperate birds because of smaller clutch sizes remains largely untested. We measured clutch size
and apparent annual breeding adult survival for 17 south temperate African species to test two main predictions. First,
we found strong support for a predicted inverse relationship between adult survival and clutch size among the south
temperate species, consistent with life-history theory. Second, we compared our clutch size and survival estimates with
published estimates for congeneric tropical African species to test the prediction of larger clutch size and lower adult
survival among south temperate than related tropical species. We found that south-temperate species laid larger clutches,
as predicted, but had higher, rather than lower, apparent adult survival than related tropical species. The latter result may
be an artefact of different approaches to measuring survival, but the results suggest that adult survival is generally high in
the south temperate region and raises questions about the importance of the cost of reproduction to adult survival.
A trade-off between adult survival and clutch size is a fundamental theorem of life history theory (Charlesworth 1994,
Roff 2002) and a negative correlation between clutch
size and adult survival is observed among north-temperate
songbirds (Saether 1988, Martin 1995, Martin and Clobert
1996). This negative correlation also is well-expressed
between north temperate and tropical birds, where tropical
birds typically have smaller clutches correlated with higher
adult survival (Martin 1996, Ghalambor and Martin
2001, Peach et al. 2001). Here, we ask whether the negative
correlation is similarly expressed among south temperate
songbirds and in comparisons with the tropics.
This question is of interest because the north temperate
region is quite different from the rest of the world in
terms of avian life history strategies (Martin 1996, 2004).
Globally, clutch size increases with latitude (Cardillo 2002,
Jetz et al. 2008), but the slope of the increase is shallower in
the southern than northern hemisphere (Young 1994,
Martin 1996). Thus, species in the temperate southern
hemisphere lay smaller clutches than species in the temperate northern hemisphere (Young 1994, Evans et al.
2005), but often lay larger clutches than related species in
the tropics (Moreau 1944, Young 1994, Martin et al. 2006).
The observation of larger clutch sizes in the south temperate
than tropics leads to the prediction of lower adult survival
compared with the tropics, under the trade-off expected
from the cost of reproduction hypothesis of life history theory.
Annual adult survival of birds has been compared between
tropical and north temperate regions, with studies finding
survival to be higher in the tropics than in north temperate
regions (Faaborg and Arendt 1995, Johnston et al. 1997,
Peach et al. 2001, Francis and Wells 2003, McGregor et al.
2007, but see Karr et al. 1990). Whether survival is lower in
the south temperate region compared with the tropics, however, is less clear. Studies of south temperate species have
found relatively high adult survival (Robinson 1990, Magrath
and Yezerinac 1997, Gardner et al. 2003, Radford 2004,
Armstrong et al. 2005, Doerr and Doerr 2006), but comparisons with the tropics are lacking. Thus, the prediction
that adult survival of south temperate birds should be intermediate between that of north temperate and tropical species remains untested. Moreover, existence of a relationship
between clutch size and adult survival in south temperate
songbirds is largely unknown. We examine both issues here.
493
We measured clutch size and annual adult apparent survival for 17 south temperate African species, first, to test the
prediction of larger clutch size and lower adult survival
among south temperate species when compared with tropical species in Africa, and second, to test the strength of the
predicted inverse relationship between clutch size and adult
survival.
Methods
Field study
We studied annual adult survival and clutch sizes of an
assemblage of 17 species inhabiting the 2900 ha Koeberg
Nature Reserve (33°41′S, 18°26′E, elevation 10 m), on
the west coast of South Africa. The study area has a
Mediterranean climate with hot, dry summers and cool,
wet winters (Low and Rebelo 1996). Annual rainfall, as
measured over the period 1980–2007 at a weather station
located within the study area averaged 375  77 mm
(range 242 to 640 mm). The vegetation is low coastal shrubland, with an average shrub height of 1–2 m (Nalwanga
et al. 2004). The 17 species included 16 passerines and one
non-passerine (Colius colius, Order Coliiformes). The nonpasserine was included because it is an altricial, opennesting species with similar life history to the passerines
included. Indeed, it fits in the relationships and is not an
outlier affecting relationships.
Banding of adults with a unique combination of three
colour bands and a numbered metal band commenced in
2001 and continued annually to 2007, but was most
intensive in the years 2001–2004. Most birds were caught
opportunistically using mist-nets at any time of the year
within a 90 ha core area of the 260 ha study area. However
some breeding adults were also targeted at their nests during the breeding season (August–October), using mistnets or spring traps baited with a mealworm, or in the case
of Anthoscopus minutus catching, by hand, birds roosting
in their enclosed nests. Body mass was measured for all
banded birds. Resighting effort was confined to the
3-month breeding season, August–October 2001–2007,
when intensive monitoring of nests took place throughout
the study area. Nests were located using parental behaviour, usually during the building stage, and checked at 1–4
d intervals to determine clutch size and fate (Martin et al.
2006). The number of nests monitored each year ranged
from 650 to 1640. Most species (except for Crithagra
albogularis, C. flaviventris and Colius colius) nested within
well-defined territories, which were mapped across the
study area each year on the basis of nest locations and
resighting observations. Any individuals resighted with
missing colour bands were targeted for re-trapping and
band replacement during each breeding season. We aimed
to estimate survival of breeding adults (including helpers
in co-operatively breeding species); therefore capture
histories were entirely restricted to resighting and recapture information obtained during the breeding season
(August–October). Thus, although banding occurred
opportunistically during the non-breeding season, individuals only entered the survival analysis dataset in the
494
year they were first resighted during the breeding season or
trapped as an adult attending a nest. This meant that for
some individuals banded in the non-breeding season and
not resighted or recaptured during the subsequent breeding season(s), their capture histories only started one or
more years after their initial banding once they were first
resighted or recaptured during the breeding season. Furthermore, the subsequent capture histories for all individuals included only information derived from resightings or
recaptures during the breeding season (August–October);
any opportunistic recaptures or resightings during the
non-breeding season were ignored in the creation of the
capture histories to avoid inflating resighting probability
(p) and the apparent survival estimate (f).
After 2004, field effort was scaled back, meaning that
monitoring of certain territories for certain species was discontinued and individuals occupying such territories could
no longer be re-encountered, resulting in recapture probabilities of zero. We accounted for the heterogeneity in recapture
probabilities caused by this design (see further detail below).
Survival analysis
The analysis of annual adult apparent survival proceeded in
two stages, a preliminary analysis conducted independently
for each species to identify factors to be included in a multispecies hierarchical model, followed by the fitting of a
multi-species hierarchical model to estimate apparent annual
adult survival rates. The rationale for developing a multispecies hierarchical model was to use it as a platform for
investigating the relationship between survival and clutch
size across the 17 species we studied while fully accounting
for the uncertainty in the survival estimates.
For the preliminary analysis, we fitted a standard
Cormack–Jolly–Seber model (CJS, Lebreton et al. 1992) for
each species separately to examine the effects of time
on apparent survival (f) and resighting probability (p). The
fit of each species-specific CJS model was tested by running
a parametric bootstrap goodness-of-fit test (100 simulations)
in program MARK (White and Burnham 1999). As our
primary interest was to investigate mean annual adult survival across species, we kept survival constant and treated
year as a random effect on resighting probabilities for all
species in the subsequent hierarchical analysis.
The first step in developing the multi-species hierarchical
model involved constructing the likelihood for the CJS
model for each species using a state-space formulation
based on state and observation process equations (Royle
2008). The state process describes whether an individual i of
species s is alive (z(s(i), t)  1) or dead (z(s(i), t)  0) given
it was alive at time t 2 1 using the equation:
z( s(i ), t ) | z( s(i ), t 1 ∼ Bernoulli ( z ( s(i ), t 1)  φs ,t )
where fs,t is the survival probability of species s between
year t 2 1 and t. Note that at the time of first encounter (f ),
z(s(i), t  f ) is equal to 1 with probability 1. The observation
process describes whether an individual i of species s
is resighted (y(s(i), t)  1) or not (y(s(i), t)  0) at time t
conditional on being alive at time t using the equation:
y( s(i ), t ) | z( s(i ), t ) ∼ Bernoulli ( z( s(i ), t )  ps ,t  e s ,t )
Where ps,t is the resighting probability of species s in year t,
and es,t is an effort matrix which takes a value of 1 if the individual resided in a territory that was monitored in year t
or a value of 0 otherwise. This structure accounted for individuals that resided in parts of the study area in which
monitoring was discontinued part-way through the study.
Next, we adopted the approach of Lahoz-Monfort
et al. (2011) to fit a multi-species hierarchical model with
species-specific, constant survival probabilities:
logit (φs ,t )  µ s
and species-specific, time-dependent resighting probabilities:
logit ( ps ,t )  γ s  ηt
where ms and gs are the logit of mean survival and resighting
probabilities of species s, ht is a normally distributed, random time effect with mean 0 and variance s2. We report
survival estimates based on this model. We implemented
the model within the Bayesian framework, specifying noninformative prior distributions (Uniform (25,5)) for species-specific means (ms) and a Uniform (0,5) prior for
standard deviation of the time effect (s). To assess the convergence of the Markov Chain Monte Carlo (MCMC)
chains, we ran three parallel chains of length 20 000 with a
burn-in 10 000, and thinning of every 20th value. The
Brooks–Gelman–Rubin diagnostic test (R-hat, Brooks and
Gelman 1998) confirmed convergence (i.e. all R-hat values
were less than 1.01). We then ran a single chain of 50 000
iterations with a burn-in of 20 000 and thinning of 20 to
compute the posterior summary statistics. All analyses were
performed using the software WinBUGS (Spiegelhalter
et al. 2003) called from R (R Core Team) via the package
R2WinBUGS (Sturtz et al. 2005). The R and WinBUGS
codes are provided in the supplementary material
(Supplementary material Appendix 2).
Comparison between South Africa and Malawi
To test the hypothesis that south-temperate birds have larger
clutch size and lower annual adult survival than tropical
birds, we compared our data (latitude 34°S) with published
data for a suite of species in tropical Malawi (latitude 16°S,
elevation 60 m). We sourced estimates of adult survival
from Peach et al. (2001) and measurements of clutch size
specific to Malawi, or the tropical zone generally, from Fry
and Keith (1988–2004), Hockey et al. (2005) and Dowsett
et al. (2008). To control for the effect of phylogenetic
similarity on the traits of interest, we restricted our analysis
to eight genera with species occurring at each of the two
sites, using formal meta-analysis techniques. These data
have three sources of variability: among genera, among species within a genus, and the uncertainty associated with
each survival estimate. To account for this hierarchical structure, we adopted a Bayesian approach similar to that
of McCarthy and Masters (2005) to estimate the mean difference between the estimates from South Africa and
those from Malawi. The basic structure of the model was
similar to a regular linear mixed effects model with region
(south-temperate South Africa vs tropical Malawi) as a fixed
effect, genus as a random effect, and normally distributed
residuals. However, instead of treating the survival rates as
observed without error, we modelled them as originating
from a normal distribution using the means and standard
errors estimated in the hierarchical model for South Africa
or reported by Peach et al. (2001) for Malawi. No measure
of uncertainty was available for the clutch size estimates for
the tropics, therefore the comparison of clutch size between
regions accounted for variation among genera and species
within each genus only. We used non-informative priors
(Normal (0,107)) for the mean and regional difference in
demography, and Uniform (0,100) for the standard deviation of the genus and species effects, and ran three MCMC
chains of 40 000 iterations, discarding the first 20 000 as
burn-in. The Gelman–Rubin diagnostic indicated that these
models converged quickly and all R-hat values were below
1.01. The WinBUGS code for this analysis is provided in
the supplementary material (Supplementary material
Appendix 3).
Analysis of the relationship between survival and
clutch size
To test the predicted inverse relationship between annual
adult survival and clutch size among the 17 species studied,
we extended the multi-species hierarchical model to include
clutch size as a covariate. We also included body mass
as a covariate in the analysis because it might influence survival among species (Ghalambor and Martin 2001). This
leads to the following model:
logit (φs ,t )  β0  β1  clutch size s  β2  body mass s
where φs,t is the survival probability of species s between year
t 2 1 and t, and b0 to b2 are the regression coefficients. The
model was implemented within the Bayesian framework as
described earlier, also specifying non-informative prior distributions (Uniform (25,5)) for regression coefficients (b).
Results
We found no evidence for lack of fit of the CJS models,
with ĉ  1.5 for all species except two (Cossypha caffra:
1.77, and Cisticola subruficapilla: 2.09). There was no evidence for a time effect on apparent survival with the exception of one species, Crithagra flaviventris (Supplementary
material Appendix 1). While the best-fitting model for
Crithagra flaviventris included a time effect on apparent
survival, the model with constant survival was still a reasonable model, having ΔAICc  2. Time dependence on
resighting probabilities was evident in three species
(Supplementary material Appendix 1). The hierarchical
model holding survival constant and treating time as a random effect on resighting probabilities for all species estimated apparent annual adult survival as ranging from 34 to
90% among our 17 species, with mean inter-annual resighting probabilities ranging between 60 and 97% (Table 1).
Mean clutch sizes varied from 2.1 to 4.9 eggs across
the 17 south temperate species (Table 1). The model including clutch size and body mass as covariates revealed a
strong negative relationship between annual adult survival and clutch size, while controlling for body mass
495
Table 1. Summary of apparent annual adult survival (f) and annual resighting probability (p) estimates, number of adults trapped and
resighted as breeders (N), and number of encounter occasions in years (yr) for survival models for 17 south-temperate South African species.
The estimates of f and p were obtained from a multi-species hierarchical model with species-specific and constant survival across years, and
species-specific and random time effect in resighting probabilities.
Species
Mass
(g)1
Clutch size
mean  SD (n)
Survival (f)
mean  SE
Resighting (p)
mean  SE
N
(yr)
Emberiza capensis (Cape bunting)
Crithagra albogularis (white-throated canary)
Crithagra flaviventris (yellow canary)
Cinnyris chalybeus (southern double-collared sunbird)
Cercotrichas coryphaeus (Karoo scrub-robin)
Cossypha caffra (Cape robin-chat)
Zosterops capensis (Cape white-eye)
Apalis thoracica (bar-throated apalis)
Cisticola subruficapilla (grey-backed cisticola)
Prinia maculosa (Karoo prinia)
Pycnonotus capensis (Cape bulbul)
Parisoma subcaeruleum (chestnut-vented tit-babbler)
Sphenoeacus afer (Cape grassbird)
Sylvietta rufescens (long-billed crombec)
Anthoscopus minutus (Cape penduline tit)
Telophorus zeylonus (bokmakierie)
Colius colius (white-backed mousebird)2
19.7
28.1
16.3
8.0
22.2
29.6
11.4
11.9
10.2
9.2
37.0
13.7
30.5
12.1
6.6
63.7
45.3
3.07  0.69 (88)
3.82  0.86 (73)
3.69  0.75 (397)
2.26  0.59 (364)
2.87  0.41 (768)
2.07  0.42 (321)
2.33  0.61 (140)
2.91  0.50 (253)
3.18  0.64 (292)
3.64  0.71 (835)
2.75  0.53 (177)
2.56  0.54 (150)
2.54  0.60 (90)
2.16  0.37 (63)
4.91  0.92 (75)
2.94  0.76 (35)
3.70  0.89 (168)
0.623  0.07
0.670  0.07
0.596  0.04
0.794  0.03
0.777  0.02
0.903  0.02
0.742  0.04
0.764  0.02
0.629  0.05
0.706  0.02
0.786  0.04
0.810  0.03
0.838  0.05
0.788  0.04
0.337  0.06
0.785  0.07
0.721  0.08
0.969  0.03
0.871  0.09
0.798  0.08
0.944  0.03
0.973  0.01
0.905  0.04
0.882  0.06
0.897  0.04
0.863  0.08
0.959  0.02
0.880  0.06
0.899  0.05
0.918  0.05
0.885  0.05
0.878  0.11
0.896  0.08
0.595  0.13
25 (7)
24 (7)
95 (7)
55 (7)
364 (8)
83 (7)
41 (7)
134 (8)
54 (8)
212 (8)
45 (7)
62 (8)
21 (7)
41 (8)
53 (6)
13 (7)
62 (5)
Family
Emberizidae
Fringillidae
Fringillidae
Nectariniidae
Muscicapidae
Muscicapidae
Zosteropidae
Cisticolidae
Cisticolidae
Cisticolidae
Pycnonotidae
Sylviidae
Sylviidae
Sylviidae
Remizidae
Malaconotidae
Coliidae
­2Average adult body mass of species captured for banding at our study site.
1
A non-passerine (Order Coliiformes).
(b̂1 (95% Credible Interval  20.640 (20.800 to 20.483))
(Fig. 1). At the same time, annual adult survival tended to
increase with body mass while controlling for clutch size
(b̂2 (95% CI)  0.009 (20.001 to 0.018)) (Fig. 1).
Controlling for the variation among genera, south temperate species from our study site laid larger clutch sizes
(mean  2.9 versus 2.4, 95% CI for difference: 0.3 to
0.7), and had higher apparent annual adult survival
(mean  0.75 versus 0.67, 95% CI for difference: 0.02 to
0.14) than the same species or other species in the same
genus in the tropics (Table 2).
Discussion
Mean clutch sizes varied fairly extensively among our south
temperate species, ranging from clutch sizes typical of many
tropical songbirds (2.1–2.7 eggs) (Moreau 1944, Skutch
1985) to clutch sizes typical of many north temperate species
(3.5–4.9 eggs) (Martin 1995). However, we found that
south temperate birds laid average clutches less than one
egg larger than average clutches of congeneric relatives in the
tropics in Africa, in agreement with published studies
from the Americas (Young 1994, Martin et al. 2006) and
Africa (Moreau 1944). Our comparison was not influenced
by altitudinal effects on life-history (Badyaev and
Ghalambor 2001), because both sites were at low elevation.
The larger clutch sizes of the south temperate species
yields an expectation of lower adult survival based on the
cost of reproduction hypothesis (Williams 1966, Saether
1988, Charlesworth 1994, Martin 1995). Yet, we found
that south-temperate species had higher apparent annual
adult survival than their tropical counterparts. The higher
estimates may be influenced by methodological differences
caused by our focus on breeding adults and inclusion of both
496
resighting and recapture (Martin et al. 1995, Sandercock
et al. 2000, Ghalambor and Martin 2001) compared to the
mark–recapture-only study of Peach et al. (2001), as discussed in more detail below. Yet, the high apparent annual
adult survival rates of our study species (x‒  72%, n  17
species) were generally similar to other estimates from
mark–resight studies of tropical species (Sandercock et al.
2000). Moreover, mark–resight studies of individual species
in southern South America (Ghalambor and Martin 2001)
and south temperate Australia (Robinson 1990, Magrath
and Yezerinac 1997, Radford 2004, Doerr and Doerr 2006)
have also yielded similarly high adult survival estimates.
Thus, south temperate species generally show adult survival
rates that appear to be similar to the tropics, despite larger
clutch sizes in the south.
The absence of a decrease in adult survival probability
despite an increase in clutch sizes in the southern hemisphere raises interesting questions. It opposes general theory
regarding clutch size as a determinant of adult survival
through the cost of reproduction (Williams 1966, Saether
1988, Charlesworth 1994, Martin 1995). Of course, clutch
size does not represent total seasonal fecundity because of
re-nesting efforts, and annual fecundity may be the most
appropriate expression of reproductive effort for adult
survival (Martin 1995). Tropical birds may have more nesting attempts because of longer breeding seasons to cause
similar or even higher annual fecundity compared to
southern hemisphere birds, thereby explaining the similar
adult survival estimates. Yet, this explanation is not fully
satisfactory for a couple of reasons. First, many tropical
species are largely single brooded with limited nesting
attempts per year (Ahumada 2001, Cox and Martin
2009). Second, many southern hemisphere birds are multibrooded, with up to three broods and many re-nesting
attempts and still high adult survival (Magrath et al. 2000).
Figure 1. Relationship between apparent annual adult survival (95% credible intervals indicated by dashed lines) and each of clutch
size and body mass. The solid lines show the fitted relationship between apparent annual adult survival and each of clutch size controlling
for adult body mass (top) and adult body mass controlling for clutch size (bottom) obtained from the multi-species hierarchical model.
Indeed, nest predation rates in our South Africa site are
exceptionally high (Martin et al. 2006), all species re-nested
rapidly after failure and most were multi-brooded if their
first attempt was successful, with the result that the largest
recorded number of nesting attempts by a single female in a
season was eleven (PL unpubl.). Thus, the commonly
invoked cost of reproduction may not be a satisfactory explanation of adult survival rates in southern hemisphere birds.
The apparent hemispheric asymmetry in annual adult
survival between north temperate and south temperate
zones may be related to hemispheric differences in climate
(cf. Chown et al. 2004), particularly seasonality and
minimum winter temperature. Minimum winter temperatures have an important influence on annual bird mortality,
particularly in the north temperate region (Cawthorne
and Marchant 1980, Altwegg et al. 2006), and annual
mortality is hypothesised to be greater in more seasonal
environments (Ricklefs 1980). Oceans have an important
moderating influence on climate (Bonan 2002) and the
ratio of ocean to land differs markedly between north
temperate and south temperate regions, being approximately 1:1 between latitudes 30° and 60° north but 16:1
between 30° and 60° south (Chown et al. 2004). The greater
oceanicity of the south temperate zone moderates the climate such that mean minimum temperature in the coldest
month typically ranges between 0 and 10°C between latitudes 30° and 60° south but ranges between –40 and 0°C
between latitudes 30° and 60° north (Chown et al. 2004).
Mean minimum temperate in the coldest month at our south
temperate South African study site (7°C) is not substantially
497
Table 2. Summary of apparent annual adult survival estimates and clutch size used in paired contrast analyses comparing species within the
same genus between our south-temperate study site and a tropical site in Africa.
South Africa (34°41′S)
South-temperate
Malawi (16°16′S)
Tropical
f1  SE
Clutch1
Species
f2  SE
Clutch3
Crithagra albogularis
0.69  0.07
3.8
0.65  0.08
2.7
Crithagra flaviventris
Cinnyris chalybeus
0.58  0.04
0.79  0.03
3.7
2.3
Cossypha caffra
Apalis thoracica
Cisticola subruficapilla
Prinia maculosa
Pycnonotus capensis
Sylvietta rufescens
0.90  0.02
0.76  0.02
0.63  0.05
0.71  0.02
0.79  0.04
0.79  0.04
2.1
2.9
3.2
3.6
2.8
2.2
Crithagra mozambica
Crithagra sulphurata
Cinnyris bifasciatus
Cinnyris cupreus
Cinnyris venustus
Cossypha heuglini
Apalis flavida
Cisticola erythrops
Prinia subflava
Pycnonotus tricolor
Sylvietta rufescens
0.52  0.07
0.76  0.09
0.60  0.07
0.55  0.08
0.83  0.07
0.68  0.07
0.53  0.08
0.60  0.08
0.74  0.04
0.80  0.07
2.9
1.8
1.8
1.8
2.0
2.5
2.6
3.3
2.5
1.8
Species
­ This study; 2Peach et al. (2001); 3Fry and Keith (1988–2004), Hockey et al. (2005) and Dowsett et al. (2008).
1
lower than that experienced at the tropical Malawian study
site (10°C; World Climate Online 2013), which might
explain the similarity in mean annual adult survival of birds
between these sites. The influence of climate seasonality and
minimum winter temperature on global variation in annual
adult mortality of resident birds deserves further study.
Different approaches to dealing with the influence of
transients and immature birds may cause differences in
apparent adult survival estimates between studies. The
inclusion of transients (greater permanent emigration) and
immatures or pre-breeding adults (lower survival than
breeding adults) in any mark–recapture data set, and trap
shyness, tends to underestimate annual adult survival
(Cilimburg et al. 2002, Francis and Saurola 2002, Burton
and DeSante 2004, Nur et al. 2004). These effects are
expected to be more acute for longer-lived species for two
reasons. First, young adults may take two or more years to
secure a breeding vacancy among species with annual adult
survival greater than 70–80% (Jullien and Thiollay 1998,
Ricklefs 2000, Tarwater et al. 2011), and be more dispersive
and potentially exposed to greater mortality as adults
during the pre-breeding years (Ricklefs 2000, Nur et al.
2004). Second, long-lived species that occupy permanent,
fixed territories are expected to be risk averse (Ghalambor
and Martin 2001) and learn to avoid mist nets, particularly
if they are operated frequently at fixed sites (Burton and
DeSante 2004, Faaborg et al. 2004). Most mark–recapture
or mark–resight studies control for the effects of transients
using ‘two-age’ models that test whether survival through
the first sampling interval after banding differs from subsequent sampling intervals, under the assumption that
‘transients’ emigrate permanently during the first sampling
interval and only ‘residents’ are present in subsequent
sampling intervals (Cooch and White 2012). The use of
‘two-age’ models to control for the influence of transients
on adult survival estimates by effectively excluding data
from the first sampling interval where significant agedependence is found is similar to our approach of excluding
from analysis any data from individuals before the time that
they were confirmed to be breeders.
Peach et al. (2001) netted at fixed sites on most
( 95%) d of the entire 16-yr study period and included
498
both immatures (generally greater than six months in age)
and adults in their dataset, but controlled for age-dependence
in capture histories for ten of 21 species in which an effect
of age in ‘two-age’ models was found and trap-dependence
in one species in which an effect of trap shyness was found.
Yet, their estimates of annual adult survival from mark–
recapture were still lower than our estimates derived
from a mark–resight study restricted to breeding adults. In
territorial resident species, permanent emigration is more
likely among pre-breeding individuals. Thus, our approach
of including only breeding adults may provide a more rigorous control against the influence of pre-breeding dispersal
and presence of transients on survival estimates. This methodological difference may partly explain our finding of
higher survival in south temperate African species in comparison with the tropical species in the study of Peach et al.
(2001), particularly if there is age-specific variation in the
survival of adults aged one year and older, for two reasons.
First, our approach effectively excludes pre-breeding adults
(except in two facultative cooperatively breeding species,
Cercotrichas coryphaeus and Anthoscopus minutus, where prebreeding adults acted as helpers to the breeding pair) in the
estimation of annual adult survival. Second, our approach
of starting the capture history for any individual only
once it was resighted as a breeder may reduce the proportion
of birds in younger age classes when birds are not resighted
for the first time in the breeding season following their
first capture and banding. However, this effect is likely to
be slight as only 4.6% of individuals in the dataset were
resighted as breeders one or more years after first capture.
Furthermore, the ‘two-age’ models used in mark–recapture
studies to control for transients suffer from a similar source
of potential bias, since the survival estimates for ‘residents’
are based on survival after the first sampling interval once
‘transients’ have been excluded. Consequently, while the
higher apparent annual adult survival of our south temperate species compared with tropical species in Africa may
result in part from different approaches to measuring
adult survival, the influence of these methodological differences on the survival estimates is likely to be relatively small,
and does not detract from our finding that breeding adult
survival is generally high in the south temperate region.
Our reliance on mark–resight coupled with intensive
resighting effort at nests during the breeding season, rather
than mark–recapture circumvents the problem of trap avoidance. The mark–recapture approach consistently provides
lower estimates of survival than either the mark–resight
approach (Sandercock et al. 2000, Nur et al. 2004) or use of
age ratios in museum collections (Ricklefs et al. 2011). These
results highlight the need for better standardisation of methodologies for controlling for the influence of trap avoidance,
inclusion of pre-breeding individuals and permanent emigration on survival estimates. Obtaining accurate estimates
of the survival of breeding adults that both control for the
influence of transients and the differential survival of prebreeding adults is important for comparative studies and
demographic modelling.
Despite an apparent inability of clutch size to predict
adult survival between tropical and southern latitudes, we
found strong support for an inverse relationship between
annual adult survival and clutch size within our site (Fig. 1),
as Peach et al. (2001) also found within their site. While
these results are consistent with the cost of reproduction
hypothesis (Williams 1966, Charlesworth 1994) within each
region, the lack of correspondence between regions may
indicate that extrinsic adult mortality instead influences
clutch size in concert with other local factors (Ghalambor
and Martin 2001, Martin 2004). Clearly, the question
of reproduction as the driver of variation in adult survival
probability among latitudes requires more study.
­­
Acknowledgements
– We thank volunteer banders from the
Tygerberg Bird Club for extensive assistance with colour-banding
birds, particularly Margaret McCall, Bob Ellis, Lee Silks, and
Bridget de Kok. Many field assistants and co-workers helped locate
and monitor nests and resight the colour-band combinations of
breeding adults each year, particularly Sonya Auer and Ron Bassar,
Simon Davies, Andrew Taylor, Corine Eising, David Nkosi, Joseph
Fontaine, Davide Gaglio, Pierre-Yves Perroi, Justin Shew, Anna
Chalfoun, Riccardo Ton, Adams Chaskda, Alexander Neu, Julia
Taubman, and Bettina Christ. We thank Gert Greef and Hilton
Westman for permission to work at ESKOM’s Koeberg Nature
Reserve, and Wes Hochachka for comments that improved the
manuscript. This work was supported in part through National
Science Foundation grants (INT-9906030, DEB-0841764, DEB1241041 to TEM), National Research Foundation grants (to PL
and RA) and a Claude Leon Foundation fellowship (to FA).
Capture and banding activities were licensed by Cape Nature and
SAFRING, the South African bird-banding scheme that issued the
numbered metal bands, and approved by the Animal Ethics
Committee, Univ. of Cape Town and IACUC #05910TMMCWRU at the Univ. of Montana. Any use of trade, firm,
or product names is for descriptive purposes only and does not
imply endorsement by the U.S. Government. Although the
WinBUGS program has been used by the U.S. Geological Survey
(USGS), no warranty, expressed or implied, is made by the USGS
or the U.S. Government as to the accuracy and functioning of the
program and related program material nor shall the fact of
distribution constitute any such warranty, and no responsibility is
assumed by the USGS in connection therewith.
References
Ahumada, J. A. 2001. Comparison of the reproductive biology of
two Neotropical wrens in an unpredictable environment
in northeastern Colombia. – Auk 118: 191–210.
Altwegg, R., Roulin, A., Kestenholz, M. and Jenni, L. 2006.
Demographic effects of extreme winter weather in the
barn owl. – Oecologia 149: 44–51.
Armstrong, D. P., Davidson, R. S., Perrott, J. K., Roygard, J. and
Buchanan, L. 2005. Density-dependent population growth
in a reintroduced population of North Island saddlebacks.
– J. Anim. Ecol. 74: 160–170.
Badyaev, A. V. and Ghalambor, C. K. 2001. Evolution of
life histories along elevational gradients: trade-off between
parental care and fecundity. – Ecology 82: 2948–2960.
Bonan, G. 2002. Ecological climatology: concepts and applications.
– Cambridge Univ. Press.
Brooks, S. P. and Gelman, A. 1998. General methods for monitoring
convergence of iterative simulation. – J. Comput. Graph.
Stat. 7: 434–455.
Burton, K. M. and DeSante, D. F. 2004. Effects of mist-netting
frequency on capture rates at Monitoring Avian Productivity
and Survivorship (MAPS) stations. – In: Dunn, E. H. and
Ralph, C. J. (eds), Studies in Avian Biology no. 29: monitoring
bird populations using mist nets. Cooper Ornithological
Society, Camarillo, pp. 7–11.
Cardillo, M. 2002. The life-history basis of latitudinal diversity
gradients: how do species traits vary from the poles to
the equator? – J. Anim. Ecol. 71: 79–87.
Cawthorne, R. A. and Marchant, J. H. 1980. The effects of the
1978/79 winter on British bird populations. – Bird Study
27: 163–172.
Charlesworth, B. 1994. Evolution in age-structured populations.
– Cambridge Univ. Press.
Chown, S. L., Sinclair, B. J., Leinaas, H. P. and Gaston, K. J.
2004. Hemispheric asymmetries in biodiversity: a serious
matter for ecology. – PLoS Biol. 2: 1701–1707.
Cilimburg, A. B., Lindberg, M. S., Tewksbury, J. J. and Hejl, S. J.
2002. Effects of dispersal on survival probability of
adult yellow warblers Dendroica petechia. – Auk 119:
778–789.
Cooch, E. and White, G. (eds) 2012. Program MARK: a gentle
introduction, 11th ed. – www.phidot.org/software/mark/
docs/book/.
Cox, W. A. and Martin, T. E. 2009. Breeding biology of the
three-striped warbler in Venezuela: a contrast between
tropical and temperate parulids. – Wilson J. Ornithol. 121:
667–678.
Doerr, E. D. and Doerr, V. A. J. 2006. Comparative demography
of treecreepers: evaluating hypotheses for the evolution
and maintenance of cooperative breeding. – Anim. Behav.
72: 147–159.
Dowsett, R. J., Aspinwall, D. R. and Dowsett-Lemaire, F. 2008.
The birds of Zambia. – Turaco Press, Liege.
Evans, K. L., Duncan, R. P., Blackburn, T. M. and Crick, H. Q.
P. 2005. Investigating geographic variation in clutch size
using a natural experiment. – Funct. Ecol. 19: 616–624.
Faaborg, J. and Arendt, W. J. 1995. Survival rates of Puerto
Rican birds: are islands really that different? – Auk 112:
503–507.
Faaborg, J., Arendt, W. J. and Dugger, K. M. 2004. Bird population
studies in Puerto Rico using mist nets: general patterns
and comparisons with point counts. – In: Dunn, E. H. and
Ralph, C. J. (eds), Studies in Avian Biology no. 29: monitoring
bird populations using mist nets. Cooper Ornithological
Society, Camarillo, pp. 144–150.
Francis, C. M. and Saurola, P. 2002. Estimating age-specific survival
rates of tawny owls: recaptures versus recoveries. – J. Appl.
Stat. 29: 637–647.
Francis, C. M. and D. R. Wells. 2003. The bird community at
Pasoh: composition and population dynamics. – In:
Okuda, T., Manokaran, N., Matsumoto, Y., Niiyama, K.,
Thomas, S. C. and Ashton, P. S. (eds), Pasoh: ecology of a
lowland rain forest in southeast Asia. Springer, pp. 375–393.
499
Fry, C. H. and Keith, S. (eds) 1988–2004. The birds of Africa,
Vol. 3–7. – Academic Press.
Gardner, J. L., Magrath, R. D. and Kokko, H. 2003. Stepping
stones in life: natal dispersal in the group-living
but noncooperative speckled warbler. – Anim. Behav. 66:
S21–S30.
Ghalambor, C. K. and Martin, T. E. 2001. Fecundity-survival
trade-offs and parental risk-taking in birds. – Science
292: 494–497.
Hockey, P. A. R., Dean, W. R. J. and Ryan, P. G. 2005. Roberts
birds of southern Africa. – John Voelker Bird Book Fund,
Cape Town.
Jetz, W., Sekercioglu, C. H. and Böhning-Gaese, K. 2008. The
worldwide variation in avian clutch size across species and
space. – PLoS Biol. 6: 2650–2657.
Johnston, J. P., White, S. A., Peach, W. J. and Gregory, R. D.
1997. Survival rates of tropical and temperate passerines:
a Trinidadian perspective. – Am. Nat. 150: 771–789.
Jullien, M. and Thiollay, J. M. 1998. Multi-species territoriality
and dynamic of Neotropical forest understorey bird flocks.
– J. Anim. Ecol. 67: 227–252.
Karr, J. R., Nichols, J. D., Klimkiewicz, M. K. and Brawn, J. D.
1990. Survival rates of birds of tropical and temperate
forests: will the dogma survive? – Am. Nat. 136: 277–291.
Lahoz-Monfort, J. J., Morgan, B. J. T., Harris, M. P., Wanless, S.
and Freeman, S. N. 2011. A capture–recapture model for
exploring multi-species synchrony in survival. – Methods
Ecol. Evol. 2: 116–124.
Lebreton, J.-D., Burnham, K. P., Clobert, J. and Anderson, D. R.
1992. Modelling survival and testing biological hypotheses
using marked animals: a unified approach with case studies.
– Ecol. Monogr. 62: 67–118.
Low, A. B. and Rebelo, A. G. 1996. Vegetation of South Africa,
Lesotho and Swaziland. – Dept of Environmental Affairs
and Tourism, Pretoria.
Magrath, R. D. and Yezerinac, S. M. 1997. Facultative helping
does not influence reproductive success or survival in
cooperatively breeding white-browed scrubwrens. – J. Anim.
Ecol. 66: 658–670.
Magrath, R. D., Leedman, A. W., Gardner, J. L., Giannasca, A.,
Nathan, A. C., Yezerinac, S. M. and Nicholls, J. A. 2000. Life
in the slow lane: reproductive life history of the white-browed
scrubwren, an Australian endemic. – Auk 117: 479–489.
Martin, T. E. 1995. Avian life history evolution in relation to
nest sites, nest predation and food. – Ecol. Monogr. 65:
101–127.
Martin, T. E. 1996. Life history evolution in tropical and south
temperate birds: what do we really know? – J. Avian Biol.
27: 263–272.
Martin, T. E. 2004. Avian life-history evolution has an eminent
past: does it have a bright future? – Auk 121: 289–301.
Martin, T. E. and Clobert, J. 1996. Nest predation and avian life
history evolution in Europe versus North America: a possible
role of humans? – Am. Nat. 147: 1028–1046.
Martin, T. E., Clobert, J. and Anderson, D. R. 1995. Return rates
in studies of life history evolution: are biases large? – J. Appl.
Stat. 22: 863–875.
Martin, T. E., Bassar, R. D., Bassar, S. K., Fontaine, J. J., Lloyd,
P., Mathewson, H. A., Niklison, A. M. and Chalfoun, A. 2006.
Life-history and ecological correlates of geographic variation
in egg and clutch mass among passerine species. – Evolution
60: 390–398.
McCarthy, M. A. and Masters, P. 2005. Profiting from prior
information in Bayesian analyses of ecological data. – J. Appl.
Ecol. 42: 1012–1019.
Supplementary material (Appendix JAV-00454 at www.
avianbiology.org/readers/appendix). Appendix 1–3.
500
McGregor, R., Whittingham, M. J. and Cresswell, W. 2007.
Survival rates of tropical birds in Nigeria, west Africa. – Ibis
149: 615–618.
Moreau, R. E. 1944. Clutch size: a comparative study, with
special reference to African birds. – Ibis 86: 286–347.
Nalwanga, D., Lloyd, P., du Plessis, M. A. and Martin, T. E. 2004.
Nest-site partitioning in a strandveld shrubland bird
community. – Ostrich 75: 250–258.
Nur, N., Geupel, G. R. and Ballard, G. 2004 Estimates of adult
survival, capture probability, and recapture probability:
evaluating and validating constant-effort mist netting. – In:
Dunn, E. H. and Ralph, C. J. (eds), Studies in Avian Biology
no. 29: monitoring bird populations using mist nets. Cooper
Ornithological Society, Camarillo, pp. 63–70.
Peach, W. J., Hanmer, D. B. and Oatley, T. B. 2001. Do southern
African songbirds live longer than their European counterparts?
– Oikos 93: 235–249.
Radford, J. Q. 2004. Breeding biology, adult survival and
territoriality of the white-browed treecreeper (Climateris
affinis) in north-west Victoria, Australia. – Emu 104:
305–316.
Ricklefs, R. E. 1980. Geographic variation in clutch size among
passerine birds: Ashmole’s hypothesis. – Auk 97: 38–49.
Ricklefs, R. E. 2000. Density dependence, evolutionary
optimization, and the diversification of avian life histories.
– Condor 102: 9–22.
Ricklefs, R. E., Tsunekage, T. and Shea, R. E. 2011. Annual adult
survival in several new world passerine birds based on age
ratios in museum collections. – J. Onithol. 152: 481–495.
Robinson, D. 1990. The social organisation of the scarlet robin
Petroica multicolor and flame robin P. phoenicea in southeastern
Australia. – Ibis 132: 78–94.
Roff, D. A. 2002. Life history evolution. – Sinauer.
Royle, J. A. 2008. Modelling individual effects in the
Cormack–Jolly–Seber model: a state-space formulation.
– Biometrics 64: 364–370.
Saether, B.-E. 1988. Pattern of covariation between life-history
traits of European birds. – Nature 331: 616–617.
Sandercock, B. K., Beissinger, S. R., Stoleson, S. H., Melland, R.
R. and Hughes, C. R. 2000. Survival rates of a Neotropical
parrot: implications for latitudinal comparisons of avian
demography. – Ecology 81: 1351–1370.
Skutch, A. F. 1985. Clutch size, nesting success, and predation
on nests of neotropical birds reviewed. – Ornithol. Monogr.
36: 575–594.
Spiegelhalter, D. J., Thomas, A., Best, N. G. and Lunn, D. 2003.
WinBUGS user manual version 1.4. – Medical Research
Council Biostatistics Unit, Cambridge, UK.
Sturtz, S., Ligges, U. and Gelman, A. 2005. R2WinBUGS:
a package for running WinBUGS from R. – J. Stat. Softw.
12: 1–16.
Tarwater, C. E., Ricklefs, R. E., Maddox, J. D. and Brawn, J. D.
2011. Pre-reproductive survival in a tropical bird and
its implications for avian life histories. – Ecology 92:
1271–1281.
White, G. C. and Burnham, K. P. 1999. Program MARK: survival
estimation from populations of marked animals. – Bird Study
46: 120–139.
Williams, J. B. 1966. Adaptation and natural selection. – Princeton
Univ. Press.
World Climate Online 2013. Weather data for Cape Town, South
Africa and Nchalo, Malawi. – www.worldweatheronline.
com on 26/03/2013.
Young, B. E. 1994. Geographic and seasonal patterns of clutchsize variation in house wrens. – Auk 111: 545–555.
Download