close

Вход

Забыли?

вход по аккаунту

?

oik.04549

код для вставкиСкачать
Accepted Ar tic le
Non-breeding range size predicts the magnitude of population trends in transSaharan migratory passerine birds
Jaroslav Koleček1, Petr Procházka1, Christina Ieronymidou2, Ian J. Burfield2, Jiří Reif3,4
1
Inst. of Vertebrate Biology, Academy of Sciences of the Czech Republic, v.v.i., Květná
8, CZ-60365 Brno, Czech Republic
2
BirdLife Int., The David Attenborough Building, Cambridge, UK
3
Inst. for Environmental Studies, Faculty of Science, Charles Univ., Prague, Czech
Republic
4
Dept of Zoology and Laboratory of Ornithology, Faculty of Science, Palacký Univ. in
Olomouc, Olomouc, Czech Republic
Corresponding author: Jaroslav Koleček, Inst. of Vertebrate Biology, Academy of
Sciences of the Czech Republic, v.v.i., Květná 8, CZ-60365 Brno, Czech Republic.
Email: j.kolecek@gmail.com
Decision date: 20-Oct-2017
This article has been accepted for publication and undergone full peer review but has
not been through the copyediting, typesetting, pagination and proofreading process,
which may lead to differences between this version and the Version of Record. Please
cite this article as doi: [10.1111/oik.04549].
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
ABSTRACT
Understanding why populations of some migratory species show a directional change
over time, i.e. increase or decrease, while others do not, remains a challenge for
ecological research. One possible explanation is that species with smaller non-breeding
ranges may have more pronounced directional population trends, and their populations
are thus more sensitive to the variation in environmental conditions in their nonbreeding quarters. According to the serial residency hypothesis, this sensitivity should
lead to higher magnitudes (i.e. absolute values) of population trends for species with
smaller non-breeding ranges, with the direction of trend being either positive or
negative depending on the nature of the environmental change. We tested this
hypothesis using population trends over 2001–2012 for 36 sub-Saharan migratory
passerine birds breeding in Europe. Namely, we related the magnitude of the species'
population trends to the size of their sub-Saharan non-breeding grounds, whilst
controlling for factors including number of migration routes, non-breeding habitat niche
and wetness, breeding habitat type and life-history strategy. The magnitude of species’
population trends grew with decreasing absolute size of sub-Saharan non-breeding
ranges, and this result remained significant when non-breeding range size was expressed
relative to the size of the breeding range. After repeating the analysis with the trend
direction, the relationship with the non-breeding range size disappeared, indicating that
both population decreases and increases are frequent amongst species with small nonbreeding range sizes. Therefore, species with small non-breeding ranges are at a higher
risk of population decline due to adverse factors such as habitat loss or climatic
extremes, but their populations are also more likely to increase when suitable conditions
appear. As non-breeding ranges may originate from stochasticity of non-breeding site
selection in naive birds (‘serial-residency’ hypothesis), it is crucial to maintain a
network of stable and resilient habitats over large areas of birds’ non-breeding quarters.
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
INTRODUCTION
Understanding population changes in long-distance migratory species remains a challenge for ecological
research (Newton 2004). Long-distance migrants visit different ecological communities during their
annual cycle (Bauer and Hoye 2014), and their populations are thus shaped by factors acting in different
parts of the world (Webster and Marra 2005, Hewson and Noble 2009, Harrison et al. 2011). Therefore, it
is difficult to assess the relative importance of the numerous factors at play. At the same time, longdistance migrants face more threats than sedentary species due to their movements over large distances
(Kanyamibwa et al. 1990, 1993, Berthold et al. 1993, Böhning-Gaese and Bauer 1996; but see Koleček et
al. 2010). Better understanding of the processes crucial for the populations of long-distance migrants is
therefore essential for effective conservation of these species.
Animal populations often fluctuate due to demographic and environmental stochasticity; however, some
species show directional temporal shifts in population size, i.e. population trends, due to the pervasive
influence of some external factors such as habitat loss (Gregory et al. 2005). Species’ population trends
are of great interest to conservationists, because population declines can result in extinction if the
populations remain under permanent negative pressures (Greenwood 2003). Although many studies have
focused on determining the external drivers of long-term declines and increases (see Reif 2013 for
review), factors that would explain their species-specific sizes (magnitudes) regardless of the direction of
these changes remain unclear. This is rather surprising, as this measure may be crucial to explaining
population resilience, long-term fluctuations and responses to environmental changes (Cuervo and Møller
2017).
In a seminal paper, Cresswell (2014) developed a theoretical framework predicting that the populations of
bird species with restricted non-breeding ranges are less efficiently buffered against local environmental
perturbations than those occupying larger areas. This is because any local change on more restricted nonbreeding grounds will affect a larger proportion of the breeding population of such species. By contrast,
in species with large non-breeding grounds, these perturbations will be diluted by the many individuals
occupying various non-breeding sites not subject to the local change.
Recently, Gilroy et al. (2016) found that migratory species occupying larger non-breeding ranges relative
to breeding ranges were less likely to decline than those with relatively smaller non-breeding ranges.
However, they focused on direction of population trends, but not on their absolute values, preventing
them from testing Cresswell’s (2014) hypothesis. Their analyses also included residents and partial
migrants, which may obscure variability within long-distance migrants and thus hinder understanding of
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
the factors limiting their populations. Finally, Gilroy et al. (2016) did not design their study considering
the reasoning of Cresswell’s hypothesis, which concerns the absolute size of the non-breeding range.
Testing the relationships between population trend magnitudes and attributes of their non-breeding
grounds proposed by Cresswell (2014) may help us to better understand whether and which species may
respond more sensitively to environmental changes and may consequently deserve more conservation
attention. Our study aims to fill these knowledge gaps by performing a comprehensive test of various
potential environmental and life-history agents underlying the absolute values of population trends
(hereafter called ‘magnitude of population trends’) as a measure of population stability for 36 bird species
breeding in Europe and spending the non-breeding period in tropical Africa. For this purpose, we use
recently collated data on bird species’ population trends in Europe between 2001 and 2012 (BirdLife
International 2015). We used data on passerines that breed in Europe and form a monophyletic, but highly
diversified group, which comprises a suitable dataset to test the following predictions:
(i) Large sub-Saharan non-breeding grounds of some species will more probably cover regions with
varying conditions, driving both negative as well as positive population changes (Cresswell 2014).
Therefore, across species ranges, declines of some populations will be compensated by increases of
others, and the overall magnitude of population trend will tend to be low. In contrast, species with
restricted non-breeding areas are more likely to be exposed to either predominantly adverse or favourable
conditions, and thus the magnitude of their population trends will be higher – i.e. more steeply decreasing
or increasing. We also tested whether this relationship depends on the size of the non-breeding range
relative to the size of the breeding area, following Gilroy’s et al. (2016) approach. Similarly, we predict
that (ii) species following more migration routes (Møller et al. 2011) have a greater potential to reduce the
effect of environmental conditions and thus have lower magnitude of population trends than those with a
single migration route. (iii) We also expect that generalists with a broader non-breeding habitat niche
(utilising more habitats) will respond only weakly to specific habitat changes and consequently have
lower magnitude of population trends than habitat specialists (utilising fewer habitats). (iv) Further, we
predict lower magnitude of population trends for species that spend the non-breeding period in forests
than in open habitats, due to more stable climatic conditions in the tropical forest zone than in grasslands
(Schneider et al. 2014). Finally, we predict that (v) species occupying wet non-breeding habitats will have
higher magnitude of population trends than species using dry habitats, because most wetlands depend
upon precipitation levels and their water level and habitat quality vary considerably over time (Zwarts et
al. 2009). Populations of long-distance migrants naturally also depend on habitat conditions on their
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
breeding grounds (see Reif 2013 for review). We therefore control our analyses for species’ breeding
habitat use and life-history strategies.
METHODS
Magnitude of species’ population trends
We considered 36 long-distance migratory passerines (see Supplementary material Appendix 1, Table A1
in Supporting Information) that regularly breed in at least one European country (BirdLife International
2015) and have non-breeding grounds predominantly south of the Sahara (BirdLife International 2016).
We omitted those long-distance migrants wintering to a large extent in southern and western Europe, such
as the blackcap (Sylvia atricapilla), and species with unknown pan-European population trends according
to BirdLife International (2015).
For magnitudes of population trends, we used BirdLife International (2015), which provides the only
recent data covering all passerines breeding in Europe. However, because these pan-European trends are
expressed in relatively simple categories discriminating only population decline, increase and stability,
we used the underlying fine-scale national magnitudes of population trends (for a similar approach, see
Sanderson et al. 2006) – i.e. the overall change in population size of a given species in a given country (in
%) over the time period from 2001 to 2012. In total, we obtained 987 magnitudes of population trends
from combinations of species and countries. Species-country combinations with unknown magnitudes
were not included, and fluctuations around a stable population size were set as a magnitude of zero (see
Donald et al. 2007). For each species, we then derived an overall magnitude of population trend at the
European level as a weighted mean of its national population trend magnitudes across the countries where
the species was present, weighted by each country’s population size so that larger national populations of
a species contributed more to its overall magnitude at the European level. In the same way, we calculated
European level population trends while taking their direction into account. Since we calculated the
European level population trend magnitudes and directions separately, the resulting values differ in more
than just sign (Table A1).
Species traits
To test our hypotheses about the magnitudes of population trends, we used the following predictors: (i)
Area of sub-Saharan non-breeding range south of 20°N (km2) was calculated using R (R Core Team
2016), packages ‘geosphere’, ‘maps’, ‘maptools’, ‘raster’ and ‘rgeos’ (Bivand and Lewin-Koh 2013,
Hijmans 2013, 2014, Bivand and Rundel 2014), based on maps of the species’ non-breeding ranges from
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
BirdLife International and NatureServe (2014) and log-transformed. Additionally, we calculated (ii)
dispersion as the ratio between non-breeding and breeding range sizes (Gilroy et al. 2016): dispersion =
(log10non-breeding area – log10breeding area) / log10breeding area. The breeding ranges were also
extracted from BirdLife International (2016), and we constrained the areas to Eurasia west of 52°E (see
Gilroy et al. 2016) and north of 36°N to correspond with the area where the information on population
changes was collected. (iii) We established the number of main migration routes - from 1 to 3 routes,
leading through the Western (roughly Morocco, Algeria), Central (Tunisia and western Libya) and
Eastern Mediterranean (eastern Libya, Egypt and the Middle East) using Zink and Bairlein (1987–1995)
and Cramp (1988–1994). For non-breeding habitat we distinguished between (iv) niche position and (v)
niche breadth. We first assessed species’ habitat requirements based on Cramp (1988–1994) along a
seven-class gradient of vegetation structural complexity and density (Böhning-Gaese and Oberrath 2003):
(1) closed forest, (2) open forest, (3) forest edge, (4) savannah, orchard, garden, (5) scrubland or human
settlement, (6) open country with solitary trees or shrubs and (7) open country without trees or shrubs.
Then, we calculated niche breadth as the difference between the two habitat classes farthest apart (ranging
between 0 for species occupying a single habitat class to 6 for those utilising the whole range of habitat
classes), and niche position as the mean habitat class. This classification of species’ niches corresponds to
the alternative approaches based on field monitoring data and has been used in numerous studies on
comparative analysis of bird species’ traits (e.g. Reif et al. 2011, Laube et al. 2013, Koleček et al. 2014).
(vi) Non-breeding habitat wetness was scored according to Cramp (1988–1994) from (1) dry habitats (e.g.
desert, semi-desert, dry savannah, dry forest) through (2) wet habitats (e.g. wet savannah, rainforest,
marshland) to (3) aquatic habitats (e.g. rivers, lakes, reservoirs).
Furthermore, we controlled for life-history strategy and habitat use on breeding grounds, discriminating
four habitat types: forest, wetland, urban and open habitat (based on Cramp 1988–1994 and Koleček et al.
2010), as they may drive interspecific differences in population trends (Sæther and Bakke 2000, Reif
2013). We characterised life-history strategy from body mass, egg mass, number of broods per year,
clutch size and length of incubation, for which mean values were extracted and log-transformed from
Cramp (1988–1994). As these traits are tightly correlated, we reduced their number into two independent
axes using principal component analysis (PCA). Each species was positioned along the two most
important ordination axes, and the resulting PC scores were used for further analyses (Fig. A1). The first
ordination axis (PC 1, explaining 46.5% of the variability among species, eigenvalue = 2.33) expressed a
gradient from ‘slower strategy’ (K-selected) species (i.e. those having larger body mass and eggs, and
longer incubation period) to ‘faster strategy’ (r-selected) species. The second axis (PC 2, explaining
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
20.6% of the variability, eigenvalue = 1.03) expressed an independent gradient from species allocating
most of their energy to just one breeding attempt per season (i.e. having a single brood and larger clutch
size) to species spreading their investments across multiple breeding attempts per season (i.e. having
multiple broods and shorter incubation length).
To account for regional differences not captured by the trait variables described above, we defined four
regions based on the approximate location of species’ non-breeding ranges: (1) West-central region – the
area from Senegal along the Gulf of Guinea to South Sudan; (2) West-east region – from the west-central
region to Eastern Africa; (3) Central-south region – from Cameroon to the Indian Ocean and South
Africa; and (4) ‘Entire’ region – ranges that cover all three regions above.
Data analysis
We tested the effect of individual species trait variables (explanatory variables) on the magnitude of
species' population trends (response variable) using linear models in R. To improve normality, trend
magnitudes were log transformed before analyses. To obtain comparable model parameter estimates
(Schielzeth 2010) we standardized continuous explanatory variables using the ‘scale’ function, so that the
mean was zero and the standard deviation 1.
We assessed the performance of candidate models with individual combinations of explanatory variables
using the Akaike information criterion corrected for small sample sizes (AICc; Burnham and Anderson
2002). Given our sample size (n = 36 species), we limited the maximum number of parameters (k) in each
candidate model to four, following the recommendation of n/k = 10 to avoid overfitting and to obtain
conclusive inference (Burnham and Anderson 2002). The most supported models were those with lowest
AICc values (ΔAICc < 2). Model averaging was applied across all these models, taking model weight
into account (see Johnson and Omland 2004) in order to reveal parameter estimates for the explanatory
variables (Bartoń 2016).
To check whether the relative non-breeding range size had the same effect on trend magnitude as the
absolute non-breeding range size, we repeated the analyses with ‘dispersion’ instead of absolute nonbreeding range size (sensu Gilroy et al. 2016). Finally, we re-ran all analyses with direction of population
trend as the response variable (instead of trend magnitude). For each model, we checked the normality of
residuals, constant error variance and presence of outliers.
Although it is highly unlikely that past evolutionary history has played a role in shaping species’
population trends over recent decades, we tested for phylogenetic autocorrelation in the residuals of the
best performing models by calculating Moran’s I values using the R package ‘ape’ (Paradis et al. 2004,
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
Paradis 2009) and using phylogenetic information from Jetz et al. (2012, see Fig. A2). Non-significant
autocorrelation in the residuals implies that phylogenetic autocorrelation did not affect the results.
RESULTS
Based on the values of AICc, 5 out of 256 candidate models met the ΔAICc < 2 threshold. The size of the
sub-Saharan non-breeding range and the number of migration routes were both retained in the five best
models (Table 1, see Table A2 for the full set of models), and therefore represent the strongest predictors
of the magnitude of population trends (Table 2). Namely, species with larger sub-Saharan non-breeding
ranges had lower trend magnitudes than those using more restricted non-breeding grounds (Table 2). A
separate analysis revealed that the relationship between the trend magnitudes and dispersion based on the
most supported model (containing ‘dispersion’, ‘number of migration routes’ and ‘non-breeding region’)
was also significant (estim. = –0.910, P < 0.001). This indicates that both absolute and relative sizes of
non-breeding ranges are related to the magnitudes of species᾽ population trends. Unexpectedly, species
with more migration routes exhibited higher magnitude of population trends than those with a single route
(Table 2). Neither non-breeding niche breadth, breeding habitat requirements, position on the gradient
from K-selected to r-selected species nor non-breeding region appeared in the five most supported
models.
Residuals of the five best performing models did not show a significant phylogenetic autocorrelation
(Table 1). This also holds for the most supported model containing ‘dispersion’ (Moran’s I = –0.02, P =
0.850).
Finally, none of the predictors that influenced species’ magnitudes of population trends was significantly
related with species’ population trends in the analysis of directional trends. The population trends were
only associated with the species’ position along the first life-history axis (Table A3). Namely, species
with slower life history strategies had more positive population changes than species with faster
strategies.
DISCUSSION
We investigated factors linked to the magnitude of population trends for European long-distance
migratory passerines while controlling for the effects of species’ breeding-ground habitat use, life-history
and phylogenetic relatedness. Specifically, we tested the hypothesis that larger non-breeding ranges and
multiple migration routes stabilize bird populations by buffering the effects of environmental factors,
which should result in lower magnitudes of population trends in these species. The hypothesis was
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
supported in the case of both absolute and relative non-breeding range size, but the opposite pattern was
found in the case of the number of migration routes. However, both effects disappeared after considering
trend direction.
As hypothesized, species occupying more restricted non-breeding ranges had higher trend magnitudes
than those with more extensive non-breeding areas. This suggests that species with restricted nonbreeding ranges are more prone to population changes due to variation in environmental conditions. In
other words, they are more likely to be affected by environmental changes affecting a particular region,
and thus more likely to undergo a significant directional population change. Conversely, species
occupying larger non-breeding grounds will be affected by beneficial environmental changes in some
areas, but detrimental changes in others, leading to less pronounced overall population changes.
Our results emphasize the importance of the absolute size of non-breeding ranges, as their effect remains
significant even when we included the ratio between breeding and non-breeding range sizes. Gilroy et al.
(2016) suggested that this ‘dispersion’ can provide the same buffering effects as we hypothesized for
absolute non-breeding range sizes. Yet species with larger non-breeding ranges may generally have lower
magnitudes of population change than species with more restricted non-breeding ranges, even if their
dispersion (sensu Gilroy et al. 2016) is equal.
In contrast to Gilroy et al. (2016), who found more negative population trends in species with relatively
smaller non-breeding ranges, we did not find any relationship between direction of population trend and
either absolute or relative non-breeding range size. This indicates that the effect of a small non-breeding
range size is not exclusively related to population declines, but that both population declines and
increases are frequent in sub-Saharan migrants with high magnitudes of population trends. The observed
difference between our and Gilroy᾽s et al. (2016) results may result from differences in how population
changes were expressed and between the sets of species used for the analyses, e.g. while we focused
exclusively on passerines migrating long-distances, Gilroy et al. (2016) considered also partial migrants
and residents.
Large non-breeding ranges can result from several mechanisms, which may operate simultaneously.
Firstly, a large breeding range may translate into a large non-breeding range (e.g. Møller and Szép 2011,
Cormier et al. 2013, Hahn et al. 2013; our dataset: correlation between ranges: rS = 0.66). Secondly, the
extent of non-breeding distribution may reflect large-scale intratropical movements in response to, for
example, rapidly changing conditions following seasonal shifts of the intertropical convergence zone (e.g.
Koleček et al. 2016, Thorup et al. 2017). Finally, Cresswell (2014) suggested that the extent of nonbreeding grounds (and subsequent population stability) is related to the amount of stochasticity in initial
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
discovery and use of non-breeding sites by juveniles. After becoming adults, the birds maintain their
formerly used migration stopovers and non-breeding grounds. Thus, populations with greater first-year
stochasticity in occupying sub-Saharan sites will be expected to cover broader non-breeding areas and
also be more resilient to larger-scale change, as they will encounter sites with both favourable and
unfavourable conditions (Cresswell 2014). We suggest that such populations can more successfully
respond to climate-induced habitat changes and shifts in the long term as they may still locate the
appropriate habitat, even though geographically shifted. This may be especially true for species with large
non-breeding grounds which potentially more likely include also the shifted habitats. Of course, when
habitats (or conditions) change at continental scales, the size of the non-breeding area might become
irrelevant. For instance, the extensive droughts in the 1970s and 1980s within the whole Sahel zone
caused considerable declines in many sub-Saharan migrants (e.g. Kanyamibwa et al. 1990, 1993) and any
potential stabilizing effect of range size was unlikely to be observed during that period.
The magnitude of population trends was higher in species with more migration routes. This result is
somewhat surprising because we would have expected the same stabilizing effect as for non-breeding
range size. Although non-breeding range size and the number of migration routes are correlated across
species in our dataset (rS = 0.59), this value is below the 0.7 threshold used for detection of possible
multicollinearity (Dormann et al. 2013), and their simultaneous presence in the highest ranking models
suggests their independent statistical effects, which can hardly result from the same underlying
mechanism for both variables. From this perspective, the counterintuitive lack of support for stabilizing
effect of the number of migration routes is not so surprising. Due to the correlation of this variable with
the non-breeding range size, the part of the variability possibly linked to the stabilizing effect of the
number of migration routes on trend magnitude has been likely outperformed by the effect of nonbreeding range size. The mechanism causing higher magnitude of trends in species with more migration
routes for a given non-breeding range size remains unclear. We suggest that specific environmental
conditions along a route (or routes) a given species uses can play an important role – e.g. risks may be
greater on western routes (see Hewson et al. 2016). For example, if western flyways are less risky than
the eastern flyways due to conservation legislation and its enforcement in particular countries, then
species with a prevailing south-eastward migration direction would have more declining populations and
thus a higher trend magnitude than species migrating south-westwards. Unfortunately, testing this would
involve detailed information about the migration routes of particular sub-populations, which remain
poorly known. Further attention is also needed to better understand how different conditions on individual
migration routes affect the whole population.
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
We are aware that conditions on the non-breeding grounds and en route are not the only drivers
influencing migrants’ populations, and that factors operating on breeding grounds are also important
(Newton 2004). Since our central focus was on the conditions at non-breeding grounds, we accounted for
effects at breeding sites only by including variables describing habitat requirements (Koleček et al. 2010)
and life-history strategies (Sæther and Engen 2002) into our models. None of these factors, as well as the
habitat requirements and location of non-breeding regions, was an important predictor of population trend
magnitudes.
When we used trend direction instead of trend magnitude in the models, species᾽ life history strategy
emerged as an important trait, with steeper declines being associated with species with faster life
histories. This result accords with several other recent studies focused on the population trends of
European birds (e.g. Jiguet et al. 2007, Reif et al. 2010, Koleček et al. 2014). It seems that under current
conditions of environmental perturbations, species with slower life histories can wait for longer to find
suitable conditions for reproduction, while species with faster life histories cannot and thus decline
steeply (Sol et al. 2012).
Our results are subject to several caveats. First, our focus on large spatial scale naturally needed a
correlative approach, which, unlike manipulative experiments, cannot effectively test for causal
mechanisms (Quinn and Keough 2002). Other, especially stochastic, conditions could also contribute to
the observed variability in population changes (Sæther and Engen 2002). Second, though simplified
classification of some variables as categories is a standard approach in correlative ecological studies, it
might blur some results potentially relevant for specific groups of species. Finally, the group of species
entering the analyses is limited and although we carefully checked the model performance, our results
might be more sensitive to the violation of standard assumptions.
To sum up, small non-breeding ranges, and especially those of species utilising more migration routes,
make populations of long-distance migrants more prone to larger population changes and thus potentially
more responsive to environmental changes. Therefore, we should accept that fluctuations might be natural
in migrant species and will continue or even intensify with climate-induced habitat shifts. The presumably
wide dispersal of juveniles across the non-breeding range (see Cresswell 2014) may provide a resilience
mechanism to climate change; however, this reaches its limits when the availability of suitable habitat
shrinks and a greater proportion of offspring cannot reach their target habitats. Better understanding of
juvenile settlement mechanisms, impact of habitat degradation and climatic circulation in the tropics on
population changes (e.g. Schneider et al. 2014) is thus needed to safeguard bird populations.
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
Our results suggest that it would be beneficial to maintain stable and resilient habitats over extensive
areas of Africa to preserve the potential for species to occupy large non-breeding ranges. This is a
difficult and perhaps unattainable task in rapidly developing regions such as sub-Saharan Africa. One
solution may, paradoxically, be the adoption of modern and more intensive technologies, which reduce
further demands for space (Janssen and Rutz 2012). For example, current intensive agricultural
cultivation can increase food supply without further expanding agricultural land (Phalan et al. 2016).
However, it is critical to have more detailed information about the habitat preferences of migrants prior to
adopting any land sparing or sharing policy for their support. Modern tracking technologies, such as
advanced types of geolocators, could be employed more effectively to identify areas with rapidly
changing conditions and to improve our knowledge of carry-over effects of events operating throughout
the annual cycle.
ACKNOWLEDGEMENTS
This study was supported by the Czech Science Foundation (13-06451S) and through the Institutional
Research Plan (RVO: 68081766). JR was supported by Charles University, Prague (PRIMUS/17/SCI/16).
The population data used (BirdLife International 2015) derive from the results of the European Red List
of Birds project, which was funded by a contract from the European Commission to BirdLife
International. We are indebted to the thousands of volunteers who collected data on bird population sizes
and trends across Europe, and to the national coordinators who collated and provided them. We also
thank Silke Bauer, Will Cresswell, Chris M. Hewson and the anonymous reviewers for their helpful
comments on previous versions of the manuscript.
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
REFERENCES
Bartoń, K. 2016. MuMIn: multi-model inference. – R package ver. 1.15.6. <https://CRAN.Rproject.org/package=MuMIn>.
Bauer, S. and Hoye, B. J. 2014. Migratory animals couple biodiversity and ecosystem functioning
worldwide. – Science 344: 1242552.
Berthold, P. at al. 1993. Analyse von Fangzahlen im Hinblick auf die Bestandsentwicklung von
Kleinvögeln nach 20jährigem Betrieb der Station Mettnau, Süddeutschland. – J. Ornithol. 134: 283–299.
BirdLife International 2015. European Red List of birds. Office for Official Publications of the European
Communities, Luxembourg. –
<http://datazone.birdlife.org/userfiles/file/Species/erlob/EuropeanRedListOfBirds_June2015.pdf>.
BirdLife International 2016. IUCN Red List for birds. – <http://www.birdlife.org/datazone/species>.
BirdLife International and NatureServe 2014. Bird species distribution maps of the world. – BirdLife
International and NatureServe.
Bivand, R. and Lewin-Koh, N. 2013. maptools: Tools for reading and handling spatial objects. – R
package ver. 0.8-27. <http://CRAN.R-project.org/package=maptools>.
Bivand, R. and Rundel, C. 2014. rgeos: Interface to Geometry Engine – Open Source (GEOS). – R
package ver. 0.3-8. <http://CRAN.R-project.org/package=rgeos>
Böhning-Gaese, K. and Bauer, H. G. 1996. Changes in species abundance, distribution, and diversity in a
central European bird community. – Conserv. Biol. 10: 175–187.
Böhning-Gaese, K. and Oberrath, R. 2003. Macroecology of habitat choice in long-distance migratory
birds. – Oecologia 137: 296–303.
Burnham, K. P. and Anderson, D. M. 2002. Model selection and multimodel inference: a practical
information-theoretic approach. – Springer.
Cormier, R. L., Humple, D. L., Gardali, T. and Seavy, N. E. 2013. Light-level geolocators reveal strong
migratory connectivity and within-winter movements for a coastal California Swainson’s thrush
(Catharus ustulatus) population. – Auk 130: 283–290.
Cramp, S. E. 1988–1994. The birds of the Western Palearctic: Vols. V–IX. – Oxford Univ. Press.
Cresswell, W. 2014. Migratory connectivity of Palaearctic–African migratory birds and their responses to
environmental change: the serial residency hypothesis. – Ibis 156: 493–510.
Cuervo, J. J. and Møller, A. P. 2017. Colonial, more widely distributed and less abundant bird species
undergo wider population fluctuations independent of their population trend. – PLoS ONE 12: e0173220.
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
Donald, P. F. et al. 2007. International conservation policy delivers benefits for birds in Europe. – Science
317: 810–813.
Dormann, C. F. et al. 2013. Collinearity: a review of methods to deal with it and a simulation study
evaluating their performance. – Ecography 36: 27–46.
Gilroy, J. J. et al. 2016. Migratory diversity predicts population declines in birds. – Ecol. Lett. 19: 308–
317.
Greenwood, J. J. 2003. The monitoring of British breeding birds: a success story for conservation
science? – Sci. Tot. Environ. 310: 221–230.
Gregory, R. D. et al. 2005. Developing indicators for European birds. – Phil. Trans. Roy. Soc. B. 360:
269–288.
Hahn, S. et al. 2013. Strong migratory connectivity and seasonally shifting isotopic niches in
geographically separated populations of a long-distance migrating songbird. – Oecologia 173: 1217–
1225.
Harrison, X. A. et al. 2011. Carry-over effects as drivers of fitness differences in animals. – J. Anim.
Ecol. 80: 4–18.
Hewson, C. M. and Noble, D. G. 2009. Population trends of breeding birds in British woodlands over a
32‐ year period: relationships with food, habitat use and migratory behaviour. – Ibis 151: 464–486.
Hewson, C. M. et al. 2016. Population decline is linked to migration route in the Common Cuckoo. –Nat.
Commun. 7: 12296.
Hijmans, R. J. 2013. raster: Geographic data analysis and modeling. – R package ver. 2.1-66.
<http://CRAN.R-project.org/package=raster>.
Hijmans, R. J. 2014. geosphere: Spherical trigonometry. – R package ver. 1.3-11. <http://CRAN.Rproject.org/package=geosphere>.
Janssen, R. and Rutz, D. 2012. Bioenergy for sustainable development in Africa. – Springer.
Jetz, W. et al. 2012. The global diversity of birds in space and time. – Nature 491: 444–448.
Jiguet, F. et al. 2007. Climate envelope, life history traits and the resilience of birds facing global change.
– Glob. Change Biol. 13: 1672–1684.
Johnson, J. B. and Omland, K. S. 2004. Model selection in ecology and evolution. – Trends Ecol. Evol.
19: 101–108.
Kanyamibwa, S. et al. 1990. Changes in adult annual survival rates in a western European population of
the White Stork Ciconia ciconia. – Ibis, 132, 27–35.
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
Kanyamibwa, S. et al. 1993. Comparison of survival rates between populations of the White Stork
Ciconia ciconia in central Europe. – Orn. Scand. 24: 297–302.
Koleček, J. et al. 2010. Changes in bird distribution in a Central European country between 1985–1989
and 2001–2003. – J. Ornithol. 151: 923–932.
Koleček, J. et al. 2014. Birds protected by national legislation show improved population trends in
Eastern Europe. – Biol. Conserv. 172: 109–116.
Koleček, J. et al. 2016. Cross-continental migratory connectivity and spatiotemporal migratory patterns in
the great reed warbler. – J. Avian Biol. 47: 756–767.
Laube, I. et al. 2013. Towards a more mechanistic understanding of traits and range sizes. – Glob. Ecol.
Biogeogr. 22: 233–241.
Møller, A. P. and Szép, T. 2011. The role of parasites in ecology and evolution of migration and
migratory connectivity. – J. Ornithol. 152: 141–150.
Møller, A. P. et al. 2011. Migratory divides and their consequences for dispersal, population size and
parasite–host interactions. – J. Evol. Biol. 24: 1744–1755.
Newton, I. 2004. Population limitation in migrants. – Ibis 146: 197–226.
Paradis, E. 2009. Moran’s autocorrelation coefficient in comparative methods. – R Foundation for
Statistical Computing.
Paradis, E. et al. 2004. ape: analyses of phylogenetics and evolution in R language. – Bioinformatics 20:
289–290.
Phalan, B. et al. 2016. How can higher-yield farming help to spare nature? – Science 351: 450–451.
Poluda, A. et al. 2012. First confirmed connectivity between breeding sites and non-breeding areas of the
globally threatened Aquatic Warbler Acrocephalus paludicola. – Ring. Migr. 27: 57–59.
Quinn, G. P. and Keough, M. J. 2002. Experimental design and data analysis for biologists. – Cambridge
Univ. Press.
R Core Team 2016. R: A language and environment for statistical computing. – R Foundation for
Statistical Computing. <https://www.R-project.org/>.
Reif, J. 2013. Long-term trends in bird populations: a review of patterns and potential drivers in North
America and Europe. – Acta Ornithol. 48: 1–16.
Reif, J. et al. 2010. Population changes in Czech passerines are predicted by their life-history and
ecological traits. – Ibis 152: 610–621.
Reif, J. et al. 2011. Population trends of birds across the iron curtain: brain matters. – Biol. Conserv. 144:
2524–2533.
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
Sæther, B.-E. and Bakke, Ø. 2000. Avian life history variation and contribution of demographic traits to
the population growth rate. – Ecology 81: 642–653.
Sæther, B.-E. and Engen, S. 2002. Pattern of variation in avian population growth rates. – Phil. Trans.
Roy. Soc. B. 357: 1185–1195.
Sanderson, F. J. et al. 2006. Sanderson, F. et al. 2006. Long-term population declines in Afro-Palearctic
migrant birds. Biol. Conserv. 131: 93–105.
Schielzeth, H. 2010. Simple means to improve the interpretability of regression coefficients. – Meth.
Ecol. Evol. 1: 103–113.
Schneider, T. et al. 2014. Migrations and dynamics of the intertropical convergence zone. – Nature 513:
45–53.
Sol, D. et al. 2012. Unraveling the life history of successful invaders. – Science 337: 580–583.
Thorup, K. et al. 2017. Resource tracking within and across continents in long-distance bird migrants. –
Sci. Adv. 3: e1601360.
Webster, M. S. and Marra, P. P. 2005. The importance of understanding migratory connectivity and
seasonal interactions. – In: Greenberg, R. and Marra, P. P. (eds.), Birds of two worlds: the ecology and
evolution of migration. Johns Hopkins Univ. Press.
Zink, G. and Bairlein, F. 1987–1995. Der Zug europäischer Singvögel. Vols. 1–3. – Aula Verlag.
Zwarts, L. et al. 2009. Living on the edge. Wetlands and birds in a changing Sahel. – KNNV Uitgeverij.
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
TABLE LEGENDS
Table 1: Linear models assessing the relationships between the magnitude of species' population trends
and individual explanatory variables in 36 long-distance migratory passerines breeding in Europe.
Moran’s I and the related P-values expressing phylogenetic autocorrelation in the residuals of the models
are included. The models are ordered according to their weight, as revealed using the Akaike information
criterion corrected for small sample sizes (AICc). Only the best performing models (ΔAICc < 2) are
shown. See Table A2 for the models with ΔAICc ≥ 2. * = non-breeding range size + no. of migration
routes; nbr. = non-breeding.
Model
2
R
df AICc ΔAICc weight
Moran's I
P
* + nbr. habitat wetness
0.39
5 135.2
0.00
0.11
–0.02
0.826
*
0.32
4 136.0
0.81
0.08
–0.03
0.921
* + nbr. niche position + nbr. habitat wetness 0.41
6 136.5
1.30
0.06
–0.02
0.816
* + nbr. habitat wetness + PC 2
0.41
6 136.5
1.35
0.06
–0.01
0.685
* + nbr. niche position
0.36
5 136.9
1.73
0.05
–0.03
0.978
‘This article is protected by copyright. All rights reserved.’
Accepted Ar tic le
Table 2: The model-averaged coefficients based on the Akaike information criteria corrected for small
sample sizes (AICc) of the five most supported (ΔAICc < 2) linear models assessing the relationships
between the magnitude of species' population trends and individual explanatory variables in 36 passerine
species breeding in Europe. The terms in which 95% confidence intervals (CI) did not overlap zero are
printed in bold. The model-averaged CIs are based upon unconditional standard errors.
Explanatory variable
Estimate
Lower CI
Upper CI
Non-breeding range size
–0.940
–1.548
–0.332
No. of migration routes
1.149
0.532
1.767
Non-breeding habitat wetness
0.280
–0.284
0.844
Non-breeding niche position
–0.091
–0.474
0.293
PC 2
0.047
–0.246
0.340
‘This article is protected by copyright. All rights reserved.’
Документ
Категория
Без категории
Просмотров
1
Размер файла
450 Кб
Теги
oik, 04549
1/--страниц
Пожаловаться на содержимое документа