Population Ecology

, Volume 56, Issue 2, pp 417–425

With Allee effects, life for the social carnivore is complicated

Authors

    • Department of Fish and Wildlife Conservation, Fralin Life Science InstituteVirginia Polytechnic Institute and State University
    • Center for African Resources: Animals, Communities and Land Use (CARACAL)
  • Sarah Elizabeth Jobbins
    • Department of Fish and Wildlife Conservation, Fralin Life Science InstituteVirginia Polytechnic Institute and State University
    • Center for African Resources: Animals, Communities and Land Use (CARACAL)
  • Kathleen Ann Alexander
    • Department of Fish and Wildlife Conservation, Fralin Life Science InstituteVirginia Polytechnic Institute and State University
    • Center for African Resources: Animals, Communities and Land Use (CARACAL)
Forum

DOI: 10.1007/s10144-013-0410-5

Cite this article as:
Sanderson, C.E., Jobbins, S.E. & Alexander, K.A. Popul Ecol (2014) 56: 417. doi:10.1007/s10144-013-0410-5

Abstract

Anthropogenic modification of the landscape, resultant habitat loss, and decades of persecution have resulted in severe decline and fragmentation of large carnivore populations worldwide. Infectious disease is also identified as a primary threat to many carnivores. In wildlife species, population demography and group persistence are strongly influenced by group or population size. This is referred to as the Allee effect, in which a population or group is at an increased risk of extinction when the number or density of individuals falls below some threshold due to ecological and/or genetic factors. However, in social mammalian species, the relationship between the number of individuals and the risk of extinction is complicated because aggregation may enhance pathogen exposure and transmission. Although theoretical studies of the interaction between infectious disease transmission and Allee effects reveal important implications for carnivore management and population extinction risk, information about the interaction has yet to be synthesized. In this paper, we assess life history strategies of medium to large carnivore species (≥2.4 kg) and their influence on population dynamics, with a special focus on infectious disease. While declining population trends are observed in 73 % of all carnivores (both social and solitary species), infectious disease is identified as a significant cause of population decline in 45 % of social carnivores and 3 % of solitary carnivores. Furthermore, where carnivores suffer a combination of rapid population decline and infectious disease, Allee effects may be more likely to impact social as compared to solitary carnivore populations. These potentially additive interactions may strongly influence disease transmission dynamics and population persistence potential. Understanding the mechanisms that can result in Allee effects in endangered carnivore populations and the manner in which infectious disease interfaces at this nexus may define the outcome of developed conservation strategies.

Keywords

African wild dog Anthropogenic change Cooperative breeders Group size Infectious disease

Introduction

Humans and wildlife are inextricably linked across ecosystems and landscapes, with human and animal health reliant on the ecosystems in which they live and the global environment (WHO 2011). Humans play a central role in this system and their activities can have cascading consequences. For example, anthropogenic change has been identified as a primary driver of infectious disease emergence (Daszak et al. 2001), climate change (Rosenzweig et al. 2008), species invasions (Vitousek et al. 1997), as well as the loss of biodiversity (Brooks et al. 2002).

Free ranging carnivores represent one group of animals particularly sensitive to human influences, as their key life-history traits may increase vulnerability to anthropogenic change and extinction (Purvis et al. 2000). For example, large-bodied carnivores often live at low densities and have extensive home ranges making them vulnerable to habitat loss, landscape alteration and “edge effects” (Woodroffe and Ginsberg 1998). Additionally, many carnivores have slow reproductive rates causing populations to remain at low densities for extended periods of time following high mortality events (Weaver et al. 1996).

Carnivores are critical to the structure and function of many ecosystems and encourage biodiversity through the generation of resources and the facilitation of trophic flow (reviewed in Sergio et al. 2005). In many systems, carnivores are considered both umbrella species, where their presence and persistence in an environment is a predictor of overall ecosystem health, and keystone species, where they play a significant role in maintaining the structure of ecological communities (Noss et al. 1996; Caro and O’Doherty 1999). The loss of predators from such networks can be extremely detrimental, resulting in potentially important changes to food web structure and ecosystem function (reviewed in Noss et al. 1996). The persistence of carnivores in natural systems is an important ecological and aesthetic imperative, but one that is increasingly difficult to achieve due to continued habitat loss, persecution, and negative perceptions (Winterbach et al. 2013).

In many animal systems, particularly the carnivore guild, population demography and group persistence is strongly influenced by group size or density (Courchamp et al. 1999a). Low density or small groups are more susceptible to the negative effects associated with environmental and demographic stochasticity, as well as loss of genetic diversity due to genetic drift and inbreeding (Frankham et al. 2002). One phenomenon brought about by low population density is referred to as the Allee effect (AE), or inverse density dependence at low density, which results in an increased risk of extinction (Courchamp et al. 1999b). However, in social mammalian species, the relationship between the number of individuals and the risk of extinction is complicated. Host aggregation may prevent detrimental AEs on the one hand, but increase the risk of pathogen invasion and transmission (Deredec and Courchamp 2006), due to the proximity and high contact rate between susceptible hosts (Alexander 1974; Schmid-Hempel and Crozier 1999), on the other hand. Empirical studies suggest that a small pre-epidemic group size is a strong determinant of disease-induced extinction risk in the host (De Castro and Bolker 2005). This could be due to the direct effects of disease or through the additional influence of AEs on group or population persistence. However, studies directed at understanding the implications of AEs on infectious disease dynamics are surprisingly lacking (Courchamp et al. 1999b; Deredec and Courchamp 2006; Yakubu 2007). We review the current understanding of AEs, assessing carnivore life history strategies and vulnerability to AEs, and the interacting influence of infectious disease on group and population dynamics.

Allee effects

Allee effects (AEs) have been studied across many animal systems (reviewed in Kramer et al. 2009), and can increase the impact of an array of biological phenomena and ecological processes affecting host populations when they are small and below a certain threshold size or density (Courchamp et al. 1999b). AEs can be classified into two groups: component AEs (CAEs) and demographic AEs (DAEs) (Stephens et al. 1999). CAEs arise when one or more components of individual fitness decline with a decrease in population size or density, but do not necessarily cause a population-wide loss of fitness (Stephens et al. 1999). In contrast, DAEs lead to inverse density dependence resulting in a low, and occasionally negative, per capita growth rate (Stephens et al. 1999). They can further be classified into strong DAEs, which are associated with a critical threshold population size or density under which population growth is suppressed and extinction can occur (“extinction threshold”), or weak DAEs, which are not (Berec et al. 2007).

Not all carnivores are equal when it comes to AEs

Allee effects operate on two different scales: the population level and the group level (Angulo et al. 2013). Population level AEs result from a loss of individual density and most commonly manifest in limited opportunities to find a mate, especially in species with large territories (Kramer et al. 2009). For example, mate scarcity at low densities (a CAE) is thought to have slowed re-colonization rates for the grey wolf (Canis lupis) in the Greater Yellowstone ecosystem (US) (Hurford et al. 2006) and hampered re-colonization efforts for the highly endangered Ethiopian wolf (Canis simensis) in northern Africa, following a disease-induced population crash in the 1990s (Sillero-Zubiri et al. 1996). Similarly, mate scarcity has been linked to the slow recovery of the Fennoscandian arctic fox (Vulpes lagopus) in northwest Russia following the species’ historical persecution (Herfindal et al. 2010). This CAE may also cause major reproductive crashes in already established populations, as modeled in polar bears (Ursus maritimus) (Molnár et al. 2008). Furthermore, mate scarcity may force individuals from one species to mate with others from closely related species (Muñoz-Fuentes et al. 2010). This issue is extremely important in the conservation of many carnivore species, including the Ethiopian wolf, grey wolf, red wolf (Canis rufus), Mexican wolf (Canis lupis baileyi), African wildcat (Felis silvestris), and Scottish wildcat (Felis silvestris grampia) (Beaumont et al. 2001), where the potential to hybridize with closely related species that share similar geographical areas threatens species persistence (Muñoz-Fuentes et al. 2010).

Population AEs can affect all species with low population size or density. In contrast, group AEs manifest exclusively at the group level, and are particularly detrimental to social species (Angulo et al. 2013). In these species, a threshold number of conspecifics are critical. If the population falls below this critical threshold, social interactions necessary for group survival and population persistence are lost (Courchamp and Macdonald 2001).

While the majority of carnivores are solitary, approximately 15 % of all recognized carnivore species occur in social groups (Gittleman 1989), where the aggregation of conspecifics is highly advantageous and can outweigh the costs of intraspecific competition. These species can persist as highly cohesive groups, such as seen in the African wild dog (AWD, Lycaon pictus), or adhere to fission–fusion group dynamics as seen in other carnivores such as lions (Panthera leo), brown hyenas (Parahyaena brunnea), and spotted hyenas (Crocuta crocuta) (Smith et al. 2008). Species with fission–fusion group structures display high levels of cooperation during hunting and defense of resources, yet continual conspecific aggregation is not essential for group persistence and is abandoned when resources are scarce (Smith et al. 2008). These species, therefore, have a lower critical threshold or a lower minimum number of conspecifics required for group persistence (Courchamp et al. 1999a, b). In contrast, highly cohesive social species, like AWDs, require strict participation from all group members and rely on the aggregation of conspecifics for contribution in all areas of life, including predator avoidance, reproductive success, hunting, and survivorship (Jennions and Macdonald 1994). This life history strategy can result in enhanced fitness benefits for the group, but also a higher critical threshold for extinction or AEs (Courchamp et al. 1999b). In these species, multiple CAEs on group reproduction and survival can theoretically combine to result in a DAE, and population-wide loss of fitness (Berec 2008). The outcome of these effects may create a positive feedback loop, ultimately resulting in an “extinction vortex” (Courchamp and Macdonald 2001). Accordingly, highly social carnivores have higher extinction rates than solitary carnivores or those living in small family groups, and AEs are identified as an important contributing factor in this phenomenon (Muñoz-Durán 2002).

Assessment of carnivores: current conservation status

We evaluated 45 carnivore species and reviewed key life history attributes, current population trends, and identified the presence of factors increasing the potential for AEs. Species were chosen based on the following criteria: a carnivorous diet (consisting of >60 % vertebrates) and medium to large body size (≥2.4 kg) (Muñoz-Durán 2002). The data indicate that carnivores as a whole are of great conservation concern. The majority (67 %) of carnivores assessed currently have a negative International Union for Conservation of Nature (IUCN) status (defined as critically endangered, endangered, threatened, near threatened, or vulnerable), and only 33 % are classified as of least concern (IUCN 2011).

Of the carnivore species evaluated, 73 % were considered to have declining population trends. Human-wildlife conflict was an important characteristic noted among these species. For example, 60 % of these species were associated with some degree of human conflict, with 22 % associated with high to severe conflict (Table 1).
Table 1

Conservation information for 45 species of carnivores (≥2.4 kg), spanning five families (Felidae n = 31, Canidae n = 8, Hyaenidae n = 2, Ursidae n = 1 and Dasyuridae n = 3), aiding the assessment of AE vulnerability

Group

Genera

Species

Common name

Conservation status

Population trend

Conflict with humans

Sociality

Infectious disease

References

Felids

Nefelis

N. nebulosa

Clouded leopard

Vulnerable

Decreasing

Low

Solitary

 
  

N. diardi

Sunda clouded leopard

Vulnerable

Decreasing

Low

Solitary

 
 

Panthera

P. uncia

Snow leopard

Endangered

Decreasing

High

Solitary

 
  

P. pardus

Leopard

Threatened

Decreasing

Severe

Solitary

 
  

P. tigris

Tiger

Endangered

Decreasing

Severe

Solitary

 
  

P. onca

Jaguar

Near threatened

Decreasing

High

Solitary

 
  

P. leo

Lion

Vulnerable

Decreasing

Severe

Social

33 % pop. decline due to CDV

Serengeti lions (Roelke-Parker et al. 1996), Ngorogoro lions (Munson et al. 2008)

 

Leopardus

L. weidii

Margay

Near threatened

Decreasing

Low

Solitary

 
  

L. pardalis

Ocelot

Least concern

Decreasing

Low

Solitary

 
 

Acinonyx

A. jubatus

Cheetah

Vulnerable

Decreasing

Moderate

Solitary

 
 

Puma

P. concolor

Puma

Least concern

Decreasing

High

Solitary

 
 

Lynx

L. rufus

Bobcat

Least concern

Stable

Low

Solitary

 
  

L. lynx

Eurasian lynx

Least concern

Stable

High

Solitary

 
  

L. canadensis

Canadian lynx

Least concern

Stable

No conflict

Solitary

 
  

L. pardinus

Iberian lynx

Critically endangered

Decreasing

No conflict

Solitary

 
 

Catopuma

C. temmincki

Asiatic golden cat

Near threatened

Decreasing

Low

Solitary

 
  

C. badia

Bornean bay cat

Endangered

Decreasing

Data deficient

Solitary

 
 

Leptailurus

L. serval

Serval

Least concern

Stable

Low

Solitary

 
 

Caracal

C. caracal

Caracal

Least concern

Stable

Moderate

Solitary

 
 

Oncifelis

O. geoffroyi

Geoffroy’s cat

Near threatened

Decreasing

Low

Solitary

 
  

O. colocolo

Pampas cat

Near threatened

Decreasing

Low

Solitary

 
 

Herpailurus

H. yaguarondi

Jaguarundi

Least concern

Decreasing

Low

Solitary

 
 

Pardofelis

P. marmorata

Marbled cat

Vulnerable

Decreasing

Low

Solitary

 
 

Prionailurus

P. viverrinus

Fishing cat

Endangered

Decreasing

Low

Solitary

 
  

P.bengalensis

Leopard cat

Least concern

Stable

Low

Solitary

 
 

Otocolobus

O. manul

Manul, Pallas’s cat

Near threatened

Decreasing

No conflict

Solitary

 
 

Felis

F. chaus

Jungle cat

Least concern

Decreasing

Low

Solitary

 
  

F. silvestris

Wildcat

Least concern

Decreasing

Low

Solitary

 
  

F. margarita

Sand cat

Near threatened

Unknown

No conflict

Solitary

 
  

F. libyca

African wild cat

Least concern

Decreasing

Data deficient

Solitary

 
 

Profelis

P. aurata

African golden cat

Near threatened

Decreasing

Low

Solitary

 

Hyaenids

Crocuta

C. crocuta

Spotted hyena

Least concern

Decreasing

Severe

Social

4.3 % annual pop. decline during outbreaks of the pathogenic bacterium Streptococcus equi ruminatorum

(Höner et al. 2011)

 

Parahyaena

H. brunnea

Brown hyena

Near threatened

Decreasing

Low

Social

 

Canids

Alopex

A. lagopus

Arctic fox

Least concern

Stable

Low

Social

Subspecies pop. crash due to mange

(Goltsman et al. 1996)

 

Speothos

S. venaticus

Bush dog

Near threatened

Decreasing

No conflict

Social

 
 

Lycaon

L. pictus

African wild dog

Endangered

Decreasing

High

Social

Local extinction due to CDV and rabies

CDV (Alexander et al. 1996; Goller et al. 2010), Rabies (Kat et al. 1996)

 

Cuon

C. alpinus

Dhole

Endangered

Decreasing

High

Social

 
 

Canis

C. simensis

Ethiopian wolf

Endangered

Decreasing

Low

Social

> 50 % pop. decline due to rabies

(Sillero-Zubiri et al. 1996)

  

C. mesomelas

Black-backed jackal

Least concern

Stable

Severe

Social

 
  

C. latrans

Coyote

Least concern

Increasing

Severe

Social

 
  

C. lupus

Gray wolf

Least concern

Stable

High

Social

 

Ursids

Ursus

U. maritimus

Polar bear

Vulnerable

Decreasing

High

Solitary

 

Dasyurids

Sarcophilus

S. harrisii

Tasmanian devil

Endangered

Decreasing

Moderate

Solitary

> 90 % pop. decline in affected areas due to DFTD

(McCallum et al. 2007)

 

Dasyurus

D. geoffroii

Western quoll

Near threatened

Stable

No conflict

Solitary

 
  

D. maculatus

Spotted-tailed quoll

Near threatened

Decreasing

No conflict

Solitary

 

Data was collated from IUCN (2011). Felid conflict data was derived from Inskip and Zimmermann (2009). The “–” in the column “infectious disease” does not indicate that infectious diseases are absent in these species but rather there are no recorded impacts of infectious disease on population sizes in these species

Assessment of carnivores: infectious disease

Fifty-two infectious diseases have been studied in wild large carnivores (Murray et al. 2006). The majority of those causing significant mortality were identified as viruses, with direct transmission the most common route of exposure (Murray et al. 2006). We identified that infectious disease causing significant mortality was more common in social animals, with five out of six (for which information could be found) being highly social (Table 1). This, however, does not account for effects that have not been documented or observed, or the fact that these events may be less noticeable in a solitary species.

Overall, infectious disease does not seem to be a significant cause of mortality in solitary carnivores, particularly wild populations of felids (Table 1). This may be a result of their predominantly solitary nature, which limits the number of intraspecific contacts and opportunities for parasite spread. An interesting contrast is found in the Tasmanian devil (Sarcophilus harrisii), a solitary carnivore species that has been decimated by an outbreak of devil facial tumour disease (DFTD) over the past 16 years (Hawkins et al. 2006). In this instance, DFTD transmission appears to be adapted to the solitary lifestyle of this species. Transmission occurs with the physical transfer of viable tumor cells when conspecifics encounter and bite one another (dominant route of transmission) during mating activity and territorial disputes, a regularly observed behavior among devils (Pearse and Swift 2006; Murchison 2008).

In contrast, highly social carnivores suffer from periodic outbreaks of infectious disease resulting in significant morbidity and mortality, consistent with the increased opportunity for parasite spread within gregarious species. For example, rabies outbreaks have resulted in localized extinction of AWD packs (Kat et al. 1996) and a population decline of greater than 50 % in Ethiopian wolves (Sillero-Zubiri et al. 1996; Table 1). Canine distemper virus (CDV) has also had significant impacts on several social species including the AWD (Alexander et al. 1996; Goller et al. 2010) and the African lion (Roelke-Parker et al. 1996). However, CDV outbreaks do not always cause catastrophic mortality. For example, CDV emergence in AWD populations in the Okavango Delta, Botswana did not result in any detectable adult mortality or reduction in pup survivorship but extirpated AWDs from the Chobe River front in the northern part of the country (Alexander et al. 1996, 2010). Other factors, such as co-infection, have been identified as having an important influence on the variation in population effects associated with the emergence of a pathogen such as CDV. For example, drought and associated increases in tick density and Babesia infection in lions were linked to mass CDV mortality in the population (Munson et al. 2008).

Transmission dynamics of infectious disease and the compounding nature of AEs

Epidemiological models of the transmission dynamics of infectious diseases in wildlife are largely based on the basic reproductive rate of the pathogen, denoted R 0 (Anderson and May 1991). R 0 is defined as the number of secondary infections produced from a single infected individual, over the course of its infectious stage, in a population of susceptible individuals (Diekmann et al. 1990). When R 0 > 1, a pathogen can invade a host population; if R 0 < 1 it cannot invade a host population. R 0 has important influences on disease transmission dynamics, and is influenced by contact between susceptible and infected individuals, host specificity, as well as the population dynamics of the host species (potentially multiple hosts). Thus, for directly transmitted parasites, host aggregation, as seen in social carnivores, can increase host density and contacts between susceptible and infected individuals, leading to an increase in the R 0 of the pathogen. This can increase pathogen invasion potential and the occurrence of outbreaks of infectious disease.

While aggregation of conspecifics may be beneficial for reproduction, hunting, and vigilance, social living is a disadvantage when it comes to transmission of pathogens (Alexander 1974; Schmid-Hempel and Crozier 1999; Altizer et al. 2003). Mathematical models of specialist microparasite invasion dynamics suggest that the presence of underlying AEs can have contrasting influences on infectious disease outcomes (Deredec and Courchamp 2006). In some instances, AEs may limit infectious disease outbreaks by reducing population size below thresholds necessary for pathogen establishment (Deredec and Courchamp 2006). However, once a pathogen is established within a population, the presence of a strong AE can contribute to a further reduction in host population size (Deredec and Courchamp 2006). Specialist pathogens with density dependent transmission generally only cause host population decline until an endemic equilibrium is reached or pathogen extinction ensues (Thieme et al. 2009). However, incorporation of strong AEs into these host–pathogen systems can lead to significant host declines and elevated extinction risk (Thieme et al. 2009). Even in a large population, a minimal number of infected individuals can result in sufficient disease mortality to push the population below the Allee threshold, potentially resulting in population extinction (Friedman and Yakubu 2012). While specialist pathogens with frequency-dependent transmission can cause extinction of the host even in the absence of AEs (reviewed in De Castro and Bolker 2005), incorporation of strong AEs into these models results in a more rapid population decline, where extinction becomes inevitable (Hilker 2010).

Generalist or multi-host pathogens may have more complex interactions with AEs in carnivore populations, particularly for species where a domestic animal reservoir is present. However, empirical and theoretical data are lacking in this area despite these pathogens often representing a primary conservation concern.

Carnivore life-history and conservation data indicates that a declining population trend is commonly observed in both social systems; 26 of 33 solitary species and seven of 11 social species (Table 1; Fig. 1). The cases in which infectious disease caused significant population declines were greatly biased towards social carnivores (e.g., 45 % of social carnivores versus 3 % of solitary carnivores, Fig. 1). This may lead to an increased susceptibility to AEs in these species and, ultimately, population depression and extinction of groups or even populations. While further empirical evidence will be required to prove the relationship between social structure, the occurrence of AEs, disease prevalence, and conservation status in carnivores, the data presented here does allude to the potential for these interactions to influence carnivore population dynamics and persistence.
https://static-content.springer.com/image/art%3A10.1007%2Fs10144-013-0410-5/MediaObjects/10144_2013_410_Fig1_HTML.gif
Fig. 1

A comparison between solitary and social carnivore species, focusing on decreasing population trends and infectious disease prevalence

Case study: the African wild dog

The relationship between social structure, disease prevalence, and AEs are supported by the well-documented plight of the AWD. The highly endangered AWD is an obligate cooperative breeder that suffers from infectious disease as well as multiple group-level AEs (Courchamp and Macdonald 2001). Packs with less than four individuals are generally unable to rear pups (Woodroffe et al. 1997), and small pack size also reduces vigilance (Courchamp and Macdonald 2001) and foraging efficiency (Courchamp et al. 2002; Creel and Creel 1995). When population density drops, trade-offs are also made between specialized roles, such as pup guarding and hunting (Courchamp et al. 2002). Therefore, group extinction events in AWDs may have resulted solely from increased social contact and pathogen transmission between pack members, but may also be attributed to disease-related mortality impacts on group size and the resultant occurrence of AEs. The AWD is a primary example of the complex interactions between infectious disease and AEs, which push social carnivores ever closer to the extinction threshold.

Behavioral plasticity and AEs

One of the remarkable characteristics of carnivore ecology is their well-documented behavioral plasticity (Macdonald 1983). Evidence suggests that favorable environmental conditions, such as high or specific prey availability or minimal intraguild competition, can alter social behaviors and may counteract the formation of AEs even in small populations (Gusset and Macdonald 2010). For example, coyotes in the Rocky Mountain ecosystem hunt rodents in pairs, but in other regions they feed on ungulate carrion and congregate in packs to aid in defense against competitors (reviewed in Macdonald 1983). Alternative behavioral responses may also evolve to minimize the costs associated with pathogen invasion, such as reproducing at a younger age, as seen in the Tasmanian devil in response to DFTD (Jones et al. 2008). Although the nature of the relationship between behavioral plasticity and the dynamics of AEs and infectious disease is unclear, these potential interactions must be taken into consideration in the management of threatened carnivore populations.

Consideration of AEs in conservation planning

While carnivores are particularly susceptible to anthropogenic influences, all mammalian species are affected by habitat loss, landscape alteration, and the threat of emerging infectious disease. All of these factors can result in small population size or density, leaving wild populations prone to AEs and associated population depression. Another social species of particular conservation concern is the Western gorilla (Gorilla gorilla). Together with high levels of poaching in parts of their range, the emergence of the zoonotic Ebola virus has resulted in extinction of many social groups and currently threatens the survival of this species (Caillaud et al. 2006). Gorilla populations likely suffer from a number of AEs due to their decreasing group and population sizes, hampering future recolonization efforts. Other wildlife species may be similarly affected, broadly highlighting the far-reaching importance of AEs and infectious disease in wildlife conservation and management. Despite their significance, consideration of AEs are lacking in wildlife management and conservation planning (Courchamp et al. 1999b; Stephens and Sutherland 1999). Failure to consider AEs may result in underestimation of critical threshold population sizes or densities required for population persistence.

Acknowledgments

C.E. Sanderson and S.E. Jobbins were supported in part by the Fralin Life Science Institute at Virginia Tech and the National Science Foundation (CNH 1114953). We are grateful to J. Walters (Virginia Tech) for invaluable feedback on our manuscript.

Copyright information

© The Society of Population Ecology and Springer Japan 2013