- Email: [email protected]
Limited nest site availability helps seabirds to survive cat predation on islands
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/ecolmodel
Limited nest site availability helps seabirds to survive cat predation on islands Dominique Pontier a,∗ , David Fouchet a , Jo¨el Bried b , Narg`es Bahi-Jaber a a
UMR CNRS 5558 “Biom´etrie et Biologie Evolutive”; Universit´e de Lyon; Universit´e Lyon 1, 43 Boulevard du 11 Novembre 1918, 69622 Villeurbanne Cedex, France b Departamento de Oceanografia e Pescas, Centro do IMAR da Universidade dos Ac¸ores, 9901-862 Horta, Ac¸ores, Portugal
a r t i c l e
i n f o
a b s t r a c t
Article history:
Introduced cats Felis catus have a high detrimental impact on native seabirds on islands,
Received 17 March 2007
especially when alien preys, like rabbits Oryctolagus cuniculus, co-occur. Seabirds are highly
Received in revised form
vulnerable because of their long reproductive cycles, slow turn-over of generations and
27 January 2008
the absence of efficient behaviour against terrestrial predators, especially in some burrow-
Accepted 15 February 2008
nesting species. Through a deterministic modelling approach, we explored a neglected
Published on line 3 April 2008
mechanism that may explain the resistance of some seabird species to cat predation. It was indeed observed that seabirds may compete for nest sites. As a consequence, part of
Keywords:
the breeders foregoes breeding when nest sites are a limiting resource. Our model linked
Oceanic islands
the dynamics of cats with that of seabird species. We showed that the annual impact of cats
Domestic cat
on seabirds was lower when seabirds faced competition for burrows than when the latter
Seabirds
were not a limiting resource. This was due to the fact that limited nest site availability pre-
Nest site availability
vents an optimal growth of the cat population. Cats in turn cannot manage to exterminate
Alternative preys
all the prospecting birds during the same breeding season. The limitation of the number of
Prey-predator relationships
nest sites generates a mechanism leading the bird population to conserve a large pool of
Deterministic model
sexually mature individuals while only slightly reducing the production of juveniles in the colony. This pool of floaters may play an important role in natural populations by buffering the decrease in colony size during years with harsh environmental conditions on land. In combination with buffer mechanisms, the limitation of the number of nest sites may greatly improve the chances of survival of bird populations facing predation. © 2008 Published by Elsevier B.V.
1.
Introduction
Insular ecosystems are extremely sensitive to anthropogenic perturbations (Jarvis, 1979; Byrne, 1980; Dowding and Murphy, 2001). Most of the bird extinctions attributable to introduced predators, if not all, have occurred on islands (Stattersfield and Capper, 2000). Amongst introduced predators, cats Felis catus (L.) have played a major role in the extirpation of indigenous seabirds on islands (Moors and Atkinson, 1984; Dowding
∗
Corresponding author. Tel.: +33 4 72 43 13 37; fax: +33 4 72 43 13 88. E-mail address: [email protected] (D. Pontier). 0304-3800/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.ecolmodel.2008.02.010
and Murphy, 2001). Their impact on the native avifauna was the most dramatic on islands where alien preys co-occur (Atkinson, 1985), like for instance rabbits Oryctolagus cuniculus (L.) (Weimerskirch et al., 1989). Because they evolved in areas inaccessible, or poorly accessible, to mammals (that is, oceanic islands or inaccessible cliffs on mainlands), seabirds generally lack the adaptations, both in their behaviour and their life-history strategies, that would enable them to cope successfully with alien predators (Lack,
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
1968; Moors et al., 1992). Moreover, most seabird species show quasi-absolute fidelity to their breeding island (review in Bried and Jouventin, 2002), which will increase the probability of extermination by alien predators. The risk is higher for burrowing petrels which leave their single chick unattended in their burrows when a few days old (Warham, 1990). Seabirds also have a slow turn-over of generations: they generally have a low fecundity, the extreme being a single-egg clutch without replacement (especially in all albatrosses and petrels), a delayed sexual maturity (the minimum age at first breeding ranges from 2 to 8 years) and they can live up to several tens of years (reviews in Warham, 1996; Bried and Jouventin, 2002; Appendix 2 in Schreiber and Burger, 2002). They also have long breeding cycles, that is, long periods of colony attendance: from several weeks to more than one year (review in Bried and Jouventin, 2002). Because seabirds are long-lived, the dynamics of their populations is extremely sensitive to a small increase in adult mortality (Weimerskirch et al., 1987; Cuthbert et al., 2001). In addition, introduced predators have an indirect impact on seabirds by increasing the incidence of nest switching and divorce, which can result in the loss of several breeding years (Bried and Jouventin, 1999). Studying the impact of introduced predators is of paramount importance for conservation biology, especially on subantarctic islands where the native avifauna shows high levels of endemism and is often made up essentially by seabirds (Marchant and Higgins, 1990; Higgins and Davies, 1996; Shirihai, 2002). The existence of simplified trophic webs on many subantarcic islands but also on tropical and temperate islands free of native mammals where avian communities exhibit similar characteristics makes such islands excellent models to assess the impact of introduced predators on the marine avifauna. Previous theoretical studies, despite increasing accuracy (Courchamp et al., 2000; Gaucel et al., 2005; Zhang et al., 2006), tended to neglect important aspects of seabird population dynamics. In particular, while competition for nest sites is common in seabirds (Burger and Shisler, 1978; Ramos et al., 1997; Quintana and Yorio, 1998), the role of nest site availability in the response of seabirds to predation has never been investigated. In this paper, using a modelling approach, we show that the impact of alien predators on seabirds may differ if we account for the limitation in the number of nest sites. We consider a situation similar to that on the main island of the Kerguelen archipelago, southern Indian Ocean, for which there are reliable estimates of cat densities (Say et al., 2002), and because 33 out of the 35 avian species breeding at this locality are seabird species (Weimerskirch et al., 1989). Our model incorporates the cat as a top-level predator and a “theoretical” cavity-nesting seabird species (whose demographic parameters fall within the range of those observed in burrow-nesting seabirds and where breeders show absolute year-to-year fidelity to their colony) that faces variations in nest site availability. We also consider that cats prey on birds only during the breeding period of the latter, and that nests are easily accessible to cats. If too few nest sites are available, all the birds that are in breeding condition cannot breed during a given breeding season and cats suffer from the decrease in prey availability. The growth of the cat population is limited, resulting in a decreased predation pressure on birds. This rather simple mechanism creates
317
a scenario that may enable seabird species to persist longer than they would if all potential breeders could easily obtain a nest. Finally, we discuss the implications of our results for seabird conservation.
2.
Materials and methods
2.1.
Considerations and definitions
In our model, and like on the Kerguelen Islands, seabird colony size varies between 103 and 105 pairs, depending on species (see, e.g., Weimerskirch et al., 1989; Marchant and Higgins, 1990), and cat density varies between 1 and 2 cats/km2 (Say et al., 2002). Our model takes the limiting resources for seabirds and the seasonality of reproduction of prey species into account. Birds leave the colony at the end of the breeding season and remain at sea until the onset of the next breeding cycle. Cats prey on birds but also on alien preys, for instance rabbits and mice Mus musculus which are present in the same area on the main island of the Kerguelen archipelago (Pontier et al., 2002) and provide feral cats with a primary prey species, helping them to maintain their numbers during the winter and to spread (Moors et al., 1992; Pontier et al., 2002; Gaucel et al., 2005). Because of their opportunistic behaviour, cats take birds at the beginning of the breeding season (i.e., from the pair formation period onwards) and then, the proportion of birds in their diet decreases towards the end of the breeding period (Bloomer and Bester, 1990). The birds’ life cycle is modelled as follows (Fig. 1). Each individual belongs to only one of the following categories: (1) juveniles (Bj ) that is, young from hatching until their departure to sea, (2) immatures (Bi ) that is, individuals between fledging and the year when they will become socially mature, (3) “socially mature” individuals (Bm ) that is, all the birds that have returned to the colony at least once since fledging (some of them will remain at sea without returning ashore during the currently starting breeding season), plus the birds that will return ashore for the first time of their life during the currently starting breeding season, (4) prospectors (Bp ), i.e., all the individuals that come ashore during the current breeding season (this category includes the former breeders that have regained breeding condition, the individuals reaching breeding condition for the first time and the individuals that are not physiologically able to breed but that come ashore to secure a nest and/or a mate for the next year; see, e.g., Nelson, 1983; Warham, 1990) and (5) breeders (Bn ), that is, the prospectors that effectively attempt to breed (i.e., that lay at least an egg or whose female lays at least an egg). The prospectors that do not attempt to breed during the current breeding season are thereafter referred to as “floaters”. At the end of the breeding season, breeders return to the Bm compartment. Let bB be the birth rate of birds. We note 1 the rate at which immature becomes socially mature. By definition, only prospectors return to the colony and their number is limited by Kb through food availability at sea. Indeed, the distribution and abundance of marine organisms (Waluda et al., 1999), linked to variations in oceanographic conditions (e.g., Deacon, 1977), directly influence the amount of body reserves stored by the individuals, as well as their breeding perfor-
318
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Fig. 1 – Schematic life cycle of seabirds.
mances, age at sexual maturity, colony attendance and/or the proportion of breeders (Drent and Daan, 1980; Erikstad et al., 1998; Weimerskirch, 2002). Futhermore, we assume that birds become physiologically able to breed – for the first time or to resume breeding – only at the end of the non-breeding season, resulting in a seasonal rate of acquisition of breeding condition 2 (t). Without competition, the rate at which prospectors obtain a nest is (t). Bird reproduction is regulated by the number of available nest sites Kn , i.e., only a given proportion of prospecting adults perform breeding attempts. The rate of departures from the colony after the breeding period (during incubation, chickrearing or when the chicks fledge, depending on the outcome of the breeding attempts) is given by (t), chosen in order to have birds attending the colony approximately for 6 months (see below). Finally, we note mJ the mortality rate of chicks and mA that of immature and adult birds. Cats (C) may prey on both birds (in the nest) and alien preys, e.g., rabbits, which are not explicitly modelled here. The per capita bird intake by cats depends on bird density via the functional response o = (1 − e−d(Bj +Bn ) ) (Barlow and Choquenot, 2002), where is the maximum per capita bird intake, and d represents the easiness to capture birds by describing how quickly the intake decreases with declining prey density. Among all the birds they consume, cats eat a proportion Bj /(Bj + Bn ) of Bj and a proportion Bn /(Bj + Bn ) of Bn . The growth rate of the cat population is an affine function of the per capita intake rC (1 − e−d(Bj +Bn ) ) − mC , where mC is the rate of decline of the population when birds are absent from the colony during the non-breeding period and rC −mC is the maximum growth rate of the cat population. We assume that in the absence of birds, cats are able to maintain their numbers at a low density Cmin , thanks to the presence of alternative preys, mainly rabbits (Pontier et al., 2002). We assume that this is the only way in which alternative preys influence the system. To simplify we do not model them explicitly. We modify the growth rate of
the cat population by introducing the term Cmin that ensures that C is always greater than Cmin . The system of equations reads dBj dt
= bB Bn − (mJ + (t))Bj −
Bj Bj + Bn
(1 − e−d(Bj +Bn ) )C
dBi = (t)Bj − 1 Bi − mA Bi dt
Bp + Bn dBm = (t)Bn + 1 Bi − 2 (t) 1 − dt Kb
dBp Bp + Bn = 2 (t) 1 − dt Kb
Bm − mA Bm
Bm − mA Bp − (t) 1 −
dBn Bn = −(t)Bn + (t) 1 − dt Kn −
Bn Kn
Bp
Bp − ma Bn
Bn (1 − e−d(Bj +Bn ) )C Bj + Bn
dC = rC (1 − e−d(Bj +Bn ) )C − mC (C − Cmin ) dt
2.2.
Simulations
Basic values for the parameters are given in Table 1. We define the parameter values on a yearly scale and take the seasonal variations into account. The rate at which birds attend the colony is (t)=0 (|cos(2t)| + cos(2t)) with 0 = 10 year−1 . We vary the parameter Kn from small values representing high competition for nesting to high values for modelling the absence of competition. The rate at which birds become socially mature is 1 = 0.25 year−1 and the rate at which they reach breeding condition is bounded by 2 (t) = 0 (|cos(2(t + 0.5))|+cos(2(t + 0.5))) with 0 = 10 year−1
319
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Table 1 – Basic values of the parameters (time unit is the year) Parameter
Symbol
Bird hatching rate Bird juvenile mortality rate Bird adult mortality rate Maximum number of prospector Number of nests available Maximum rate at which birds become socially mature Maximum rate at which birds reach breeding condition Maximum rate at which birds attend the colony Maximum rate at which birds leave the colony after breeding Rate of cat decay during the absence of birds Bird predation-induced maximum growth rate Minimum number of cats Maximum prey intake per cat Coefficient of cat predation according to the number of birds
and Kb = 105 birds. The rate at which breeders leave the colony once their breeding attempt ends (i.e., during incubation or chick-rearing if they fail, when the young fledge if they are successful) is given by (t) = 0 (|cos(2(t + 0.5))| + cos(2(t + 0.5))), and 0 = 10 year−1 . Bird hatching and mortality rates are bB = 1 year−1 , mJ = 2.4 year−1 (the mean hatching rate is one chick per breeding pair and the probability of fledging is approximately equal to e−2.4×0.5 = 0.3), and mA = 0.09 year−1 (corresponding to an annual adult survival rate of 0.91). The rate of decrease in cat population size in the absence of birds is mC = 2 year−1 and its maximum growth rate is rC −mC = 8 year−1 . The cat population can decline to Cmin = 3 cats when birds are absent from the colony. The maximum prey intake by cats is = 365 birds year−1 and d = 10−5 bird−1 . Simulations of the model were performed with Matlab (The MathWorks Inc.) using an explicit Runge–Kutta method.
3.
Results
The system reaches a periodic state for all tested sets of parameters. The model is analysed at the cyclic state. In the following section, we will distinguish two ways of measuring bird abundance: population size, Bj + Bi + Bm + Bp + Bn , and colony size, Bj + Bp + Bn . The latter is made up by the number of all observable birds at the colony, i.e., juveniles, floaters (non-breeding prospectors) and breeders.
3.1.
Demography of seabirds without predation
In the absence of predation, the birds’ distribution among the different categories of the population depends on the number of nest sites. When there are plenty of nests available (Fig. 2a), most of the birds in breeding condition obtain a nest and thus achieve reproduction, which leads to a high number of juveniles and then immatures. Only a small proportion of the latter gets a sufficient condition to return ashore (to prospect or to breed), due to high competition for food at sea. The population is mainly made up of immature birds which remain at sea (here approximately 2/3). On the contrary, when the number of nest sites is low (Fig. 2b), fewer young are produced, which, in the long term, results in decreased competition for food at sea. This leads to a lower proportion of immatures (as defined
bB mJ mA Kb Kn 1 0 0 0 mC rC Cmin d
Basic value 1 2.4 0.09 105 Variable 0.25 10 10 10 2 10 3 365 10−5
in Section 2) in the population compared to the previous situation (here only 1/3). Interestingly, colony size (Bj + Bp + Bn ) is less affected than population size by a decrease in the number of nest sites, especially when the number of nest sites is high enough (Fig. 2c).
3.2.
Effect of predation on birds
A low number of nest sites leads to a lower food supply for cats, limiting the growth of cat population (Fig. 3a). Under these conditions, the birds that attend the colony (i.e., floaters and breeders) are thus little affected by cats (Fig. 3b). As cats exert a low predation pressure, floaters have difficulties to obtain a nest and they still represent a high proportion of the bird population (see Fig. 3b). The other parts of the bird population are all becoming more important as the number of nest sites increases (see Fig. 3b). Unsurprisingly, an increase in the number of nest sites results in higher breeding numbers (see Fig. 3b) and hence in more juveniles, immature and mature birds (see Fig. 3b, the three categories are gathered since they show very similar responses to the variations in the number of nest sites). Interestingly, the reduction in predation pressure resulting from a decreased number of nest sites leads to a hump in both population and colony sizes (Fig. 3c). In other words, the negative consequences of a low number of nest sites for births are more than compensated by the positive impact of the reduction in cat numbers. The reduction of the cat population resulting from decreased nest availability is the key factor allowing observing the hump: if we assume a constant number of cats, bird population/colony size becomes a regularly increasing function of the number of nest sites (Fig. 3d). This means that a nest shortage always reduces seabird population/colony size, cat population size being equal. The simple fact that floaters suffer lower mortality than breeders does not explain why a decrease in nest site availability may lead to an increase in bird numbers. The fact that floaters can replace breeders when the latter die is not necessarily beneficial in the situation invoked here. As observed in Fig. 3b, between the optimal number of nests in terms of population size and unlimited number of nests, there is only an increase by around 10% in seabird breeding numbers. In other words, providing more nests does not help to
320
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Fig. 2 – Seabird population in the absence of cat predation. (a and b) Number of birds in each category: juveniles (Bj ), individuals that remain at sea (Bi + Bm ) and individuals that come ashore (i.e., breeders and non-breeding prospectors, Bn + Bp ) and total number of birds in the population (BT ) when nest site availability is not limiting (a) and for 2 × 104 nest sites (b). (c) Mean population (solid line) and colony (dashed line) sizes as a function of the number of nest sites (Kn ).
increase the number of birds that can reproduce. At this stage, the occupancy of vacant nests by floaters only leads to higher predation pressure. Filling vacancies consumes the pool of non-breeding prospectors while only slightly increasing the number of breeders. Several parameters were found to change significantly the consequences of a limited number of nest sites (Fig. 4). We defined as an index of the impact of nest shortage the population and colony size obtained when the number of nest sites is optimal. For example, Fig. 3c reveals a maximum number of birds of 9.6 × 104 (for 3.6 × 104 nest sites), versus 7.9 × 104 when Kn is infinity, i.e., when the number of nest site available is several order of magnitude higher than the number of birds that can reproduce. Thus, a nest shortage can lead up to an increase by 22% (9.6/7.9 = 1.22) in seabird numbers. This index allows estimating the sensitivity of the results to changes in parameter values. The number of birds reaching social maturity strongly depends on bB , mJ , mA and 1 , whereas the minimal impact of cats on birds depends on Cmin , mA and bB . When large numbers of individuals reach social maturity, socially mature birds act as a reservoir that may replace predated birds the next year. In such a context, a poor supply of nests allows conserving the reservoir. In contrast, when the number of nest sites is not limiting, the reservoir is emptied rapidly, and fewer individuals are able to reproduce.
The minimal impact of cats, i.e., the impact of Cmin cats on the bird population, is also important. It is mainly driven by three parameters. Cmin is the most intuitive because it acts directly on the number of predated birds. Bird growth rate parameters (mainly bB and mA ) are also important since a bird population with a more positive balance between births (or recruitment into the breeding population) and deaths can bear a more important predation pressure. When the minimal impact of cats is high, a higher predation pressure is not sustainable for the birds, and in turn decreasing bird numbers reach a threshold value beyond which cat numbers cannot increase any more. The beneficial effect of a decrease in cat population size is thus lower. The maximum number of prospectors (Kb , which is related to food availability at sea) is found to have only a little impact on the system. Multiplying Kb by two does not change anything (see Fig. 4). This is simply because the production of juveniles is never high enough to allow the number of prospectors to get closer to Kb . In other words, the food resources at sea are not limiting. Dividing Kb by two makes the resources limiting, but again, this only slightly affects seabird population size (Fig. 4). Fewer birds reach breeding condition, which explains why colony size is decreased (Fig. 4). Because, however, we assumed that food supply at sea affects the proportion of birds reaching breeding condition but not bird survival, dividing Kb by two only slightly affects population size (Fig. 4).
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
321
Fig. 3 – Interaction between predation and nest site availability. (a) Average number of cats at the equilibrium according to the number of available nests (Kn ); (b) average number of birds in each category at the equilibrium: non-breeding birds (Bj + Bi + Bm ), prospectors (Bp ) and breeders (Bn ), according to the number of nest sites (Kn ); (c) population (bold solid lines) and colony (bold dashed lines) sizes against the number of nest sites (Kn ) for usual values of the parameters as defined in Table 1 and (d) same as (c) but with a constant number of cats C = 50. Similar simulations with different numbers of cats were conducted and provided similar results (results not shown). In (c) and (d) asymptotic values, i.e., when the number of nest sites in the colony is several orders of magnitude higher than the number of birds that can reproduce, are given by solid and dashed horizontal lines for the population and the colony, respectively.
Fig. 4 – Effect on the maximum increase in population and colony sizes (when nest sites are in limited supply) when parameters are divided or multiplied by 2. Notations: mJ /2 means that mJ is here twice as low as its usual value. The system is particularly sensitive to mA and bB , because these parameters drive both the proportion of birds becoming socially mature and the impact of the minimal number of cats Cmin .
The model also predicts that bird population can become extinct through predation (results not shown). Extinction generally occurs when the predation exerted by the minimal number of cats (Cmin ) added to bird mortality from other causes is not balanced by births or recruitment. Poor nest availability cannot reduce the number of cats below Cmin since cats can survive by eating other preys. As a consequence,
a shortage of nests cannot prevent bird extinction in this case.
4.
Discussion
The aim of this paper was to assess the consequences of nest site availability for bird populations suffering predation. It is
322
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
important to note that the results presented here are mainly qualitative. Parameter values may vary greatly among bird species and/or colony sites. Interestingly, the limitation of the number of nest sites has been found to have positive consequences for the bird population despite decreased offspring productivity. Limited nest site availability decreases the impact of predation by limiting prey availability for cats. Cats will not exterminate all the adult birds during the same breeding season as long as the number of nest sites remains insufficient, like for example in mixedspecies colonies where a large number of birds compete for the same breeding sites (Burger and Shisler, 1978; Quintana and Yorio, 1998). The presence of a reservoir of individuals (floaters here) that can quickly replace effective breeders is often considered positive for the bird population because it allows breeding numbers to remain stable during a certain time lag (Walters et al., 2002; Grimm et al., 2005). This is not the case when considering the mechanism described here, where the turn-over of birds in vacant nests enables cat numbers to increase, which in turn limits the number of breeding birds. Finally, close to the optimal number of nest sites, the number of breeding birds is only slightly increased when vacancies are quickly filled. The cost of this slight increase is a large decrease in the number of floaters. This may have very important consequences for the stochastic extinction of the bird population (not modelled here), through a mechanism known as the buffer effect (Grimm et al., 2003, 2005; Penteriani et al., 2006). The fact that our model does not predict extinction does not mean that it will not occur. Bird numbers exhibit important inter-annual variations. During a year with harsh environmental conditions on land, an intense predation pressure can drive the bird population toward extinction. The large proportion of sexually mature birds that do not breed acts like a buffer, allowing to compensate for losses in the subpopulation of breeders and thus improves the persistence of the bird population during years with harsh environmental conditions. Analogous buffer mechanisms have been described in some other species (Grimm et al., 2003, 2005; Penteriani et al., 2006). The buffer effect and the control of predators are two different but complementary mechanisms that result from variations in nest site availability and that can highly enhance the chances of persistence for the bird population (Fig. 5). Our model shows that this is especially true for species with a relatively low mortality and/or a high fecundity or at localities where cats cannot maintain significant numbers in the absence of birds. The results presented here have implications for seabird conservation, not only on subantarctic islands but also on tropical and temperate islands where seabirds represent the bulk of the avian community and where there are no indigenous terrestrial mammals. They may explain the contrasted responses to predator removal campaigns, with some studies showing an increase in seabird breeding numbers following predator elimination whereas others show no effect or ˆ e´ and Sutherland, 1997). The relationship even decreases (Cot between predation and bird population dynamics can vary among seabird species, here through variations in colony carrying capacity. Population biology studies are thus important to estimate the level of competition for nests and/or nest site
Fig. 5 – Interplay between the control of predation mechanism and the buffer effect. Limited nest site availability controls predation, finally leading to a larger pool of prospectors. This pool may enable seabird breeding numbers to remain stable, decreasing the risks of extinction for the bird population through a buffer effect.
availability at a given locality in order to identify the most vulnerable native seabird species when alien predators occur. Management effort could then be directed on these particular seabird species. Furthermore, one aspect not considered here is the fact that the saturation of the breeding habitat (Chastel et al., 1993; Inchausti and Weimerskirch, 2002; Burg et al., 2003) and/or the disturbance due to alien predators (Thibault, 1995; Oro et al., 1999) may make a seabird colony less attractive for prospectors and/or elicit bird dispersal (Austin, 1940). The flux of immigrants (essentially immatures in the species where breeders are faithful to their colony) from distant sources is rarely invoked to explain the demographic patterns observed in seabirds. Inchausti and Weimerskirch (2002) and Jenouvrier et al. (2003) showed that even low immigration rates (as low as 3%) each year could affect seabird population dynamics at distant localities. By encouraging bird dispersal, the saturation of the breeding habitat could thus allow a higher persistence of bird species facing predation. Even if small numbers of individuals colonize an area where alien predators occur, this might be sufficient to maintain locally both the prey and the predator. The existence of distant sources might explain why the breeding numbers of yelkouan shearwaters Puffinus yelkouan (Acerbi) seem to remain stable on Port-Cros Island despite huge predation by cats which were introduced in the 17th century (Bourgeois, 2004). Coupling nest site availability with the occurrence of bird dispersal when studying metapopulation dynamics, could improve our understanding of bird-predator relationships.
Acknowledgements This work was supported by IPEV (programme “Popchat” n◦ 279), the “Environnement & Sustainability” Department of CNRS (“Zone Atelier de Recherche sur l’Environnement
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Antarctique et Subantarctique”) and LIA 197 “ReaDiLab”. It was also part of JB’s postdoctoral contract at the Instituto do Mar (FCT grant SFRH/BPD/20291/2004). We thank J.-M. Gaillard, M. Guiserix and two anonymous referees for their helpful comments on an earlier version of the paper.
references
Atkinson, I.A.E., 1985. The spread of commensal species of Rattus rattus to oceanic islands and their effects on islands avifaunas. In: Moors, P.J. (Ed.), Conservation of Island Birds. Technical Publication No. 3. ICBP, Cambridge, UK, pp. 35–81. Austin, O.L., 1940. Some aspects of individual distribution in the Cape Cod Tern colonies. Bird Banding 11, 155–159. Barlow, N., Choquenot, D., 2002. Predicting the impact and control of stoats: a review of modelling approaches. Sci. Conserv. 191, 46. Bloomer, J.P., Bester, M.N., 1990. Diet of a declining feral cat Felis catus population on Marion Island. S. Afr. J. Wildl. Res. 20, 1–4. Bourgeois, K., 2004. Ecologie et conservation d’un oiseau marin ´ ´ ´ Puffinus yelkouan-predation ´ endemique de Mediterran ee, par ´ le chat haret et selection de l’habitat dans le Parc National de ˆ ´ Port-Cros. ‘Diplome d’Etudes Superieures’ Thesis. University of Aix-Marseille 3. Bried, J., Jouventin, P., 1999. Influence of breeding success on fidelity in long-lived birds: an experimental study. J. Avian Biol. 30, 392–398. Bried, J., Jouventin, P., 2002. Site and mate choice in seabirds: an evolutionary approach. In: Schreiber, E.A., Burger, J. (Eds.), Biology of Marine Birds. CRC Press, Boca Raton, FL, pp. 263–305. Burg, T.M., Lomax, J., Almond, R., Brooke, M., de, L., Amos, W., 2003. Unravelling dispersal patterns in an expanding population of a highly mobile seabird, the northern fulmar (Fulmarus glacialis). Proc. R. Soc. Lond. B 270, 979–984. Burger, J., Shisler, J., 1978. Nest-site selection and competition interactions of herring and laughing gulls in New Jersey. Auk 95, 252–266. Byrne, R., 1980. Man and the variable vulnerability of island life. A study of recent vegetation change in the Bahamas. Atoll Res. Bull. 240, 200 pp. Chastel, O., Weimerskirch, H., Jouventin, P., 1993. High annual variability in reproduction success and survival of an Antarctic seabird, the snow petrel (Pagodroma nivea). Oecologia 94, 278–285. ˆ e, ´ I.M., Sutherland, W.J., 1997. The effectiveness of removing Cot predators to protect bird populations. Conserv. Biol. 11, 395–405. Courchamp, F., Langlais, M., Sugihara, G., 2000. Rabbits killing birds: modelling the hyperpredation process. J. Anim. Ecol. 69, 154–164. Cuthbert, R., Fletcher, D., Davis, L.S., 2001. A sensitivity analysis of Hutton’s shearwater: prioritizing conservation research and management. Biol. Conserv. 100, 163–172. Deacon, G.E.R., 1977. Seasonal variations in the water temperature and salinity near South Georgia 1925–1937. Report 49. Institute of Oceanography Science, Wormley, UK. Dowding, J.E., Murphy, E.C., 2001. The impact of predation by introduced mammals in endemic shorebirds in New Zealand: a conservation perspective. Biol. Conserv. 99, 47– 64. Drent, R.H., Daan, S., 1980. The prudent parent: energetic adjustment in avian breeding. Ardea 68, 225–252.
323
Erikstad, K.E., Fauchald, P., Tveraa, T., Steen, H., 1998. On the cost of reproduction in long-lived birds: the influence of environmental variability. Ecology 79, 1781–1788. Gaucel, S., Langlais, M., Pontier, D., 2005. Invading introduced species in insular heterogeneous environments. Ecol. Modell. 18, 62–75. Grimm, V., Dorndorf, N., Frey-Roos, F., Wissel, C., Wyszomirski, T., Arnold, W., 2003. Modelling the role of social behavior in the persistence of the alpine marmot Marmota marmota. Oikos 102, 124–136. Grimm, V., Revilla, E., Groeneveld, J., Kramer-Schadt, S., Schwager, M., Tews, J., Wichmann, M., Jeltsch, F., 2005. Importance of buffer mechanisms for population viability analysis. Conserv. Biol. 19, 578–580. Higgins, P.J., Davies, S.J.J.F., 1996. Handbook of Australian, New Zealand and Antarctic Birds, 3. Oxford University Press, Melbourne. Inchausti, P., Weimerskirch, H., 2002. Dispersal and metapopulation dynamics of an oceanic seabird, the wandering albatross, and its consequences for its response to long-line fisheries. J. Anim. Ecol. 71, 765–770. Jarvis, P.T., 1979. The ecology of plant and animal introductions. Prog. Phys. Geogr. 3, 187–214. Jenouvrier, S., Barbraud, C., Weimerskirch, H., 2003. Effects of climatic variability on the temporal population dynamics of southern fulmars. J. Anim. Ecol. 72, 576–587. Lack, D., 1968. Ecological Adaptations for Breeding in Birds. Methuen, London. Marchant, S., Higgins, P.J., 1990. Handbook of Australian, New Zealand and Antarctic Birds, 1A–1B. Oxford University Press, Melbourne. Moors, P.J., Atkinson, I.A.E., 1984. Predation on Seabirds by Introduced Animals, and Factors Affecting its Severity. International Council for Bird Preservation, Cambridge, UK, pp. 667–690, ICBP Techn. Publ. 2. Moors, P.J., Atkinson, I.A.E., Sherley, G.H., 1992. Reducing the rat threat to island birds. Bird Conserv. Int. 2, 93–114. Nelson, J.B., 1983. Contrasts in breeding strategies between some tropical and temperate marine pelecaniformes. Stud. Avian Biol. 8, 95–114. Oro, D., Pradel, R., Lebreton, J.-D., 1999. Food availability and nest predation influence life history traits in Audouin’s gull Larus audouinii. Oecologia 118, 438–445. Penteriani, V., Otalora, F., Ferrer, M., 2006. Floater dynamics can explain positive patterns of density-dependent fecundity in animal populations. Am. Nat. 168, 697–703. Pontier, D., Say, L., Debias, F., Bried, J., Thioulouse, J., Micol, T., Natoli, E., 2002. The diet of feral cats (Felis catus L.) at five sites on the Grande Terre, Kerguelen archipelago. Polar Biol. 25, 833–837. Quintana, F., Yorio, P., 1998. Competition for nest sites between kelp gulls (Larus dominicanus) and terns (Sterna maxima and S. eurygnatha) in Pentagonia. Auk 115, 1068–1071. Ramos, J.A., Monteiro, L.R., Sola, E., Moniz, Z., 1997. Characteristics and competition for nest cavities in burrowing Procellariiformes. Condor 99, 634–641. Say, L., Gaillard, J.-M., Pontier, D., 2002. Spatio-temporal variation in cat population density in a Sub-Antarctic environment. Polar Biol. 25, 90–95. Schreiber, E.A., Burger, J., 2002. Biology of Marine Birds. CRC Press, Boca Raton, FL. Shirihai, H., 2002. A Complete Guide to Antarctic Wildlife. Alula Press Oy, Degerby. Stattersfield, A.J., Capper, D.R., 2000. Threatened Birds of the World. BirdLife International and Lynx Edicions, Barcelona. Thibault, J.-C., 1995. Effect of predation by the black rat Rattus rattus on the breeding success of Cory’s shearwater Calonectris diomedea in Corsica. Mar. Ornithol. 23, 1–10.
324
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Walters, J.R., Crowder, L.B., Priddy, J.A., 2002. Population viability analysis for red-cockaded woodpeckers using an individual-based model. Ecol. Appl. 12, 249–260. Waluda, C., Trathan, P., Rodhouse, P., 1999. Influence of oceanographic variability on recruitment in the Ilex argentinus Cephalopoda: Ommastrephidae fishery in the South Atlantic. Mar. Ecol. Progr. Ser. 183, 159–167. Warham, J., 1990. The Petrels. In: Their Ecology and Breeding Systems. Academic Press, London. Warham, J., 1996. The Behaviour, Population Biology and Physiology of the Petrels. Academic Press, London.
Weimerskirch, H., 2002. Seabird demography and its relationships with the marine environment. In: Schreiber, E.A., Burger, J. (Eds.), Biology of Marine Birds. CRC Press, Boca Raton, FL, pp. 115–135. Weimerskirch, H., Clobert, J., Jouventin, P., 1987. Survival in five southern albatrosses and its relationship with their life history. J. Anim. Ecol. 56, 1043–1055. Weimerskirch, H., Zotier, R., Jouventin, P., 1989. The avifauna of the Kerguelen Islands. Emu 89, 15–29. Zhang, J., Fan, M., Kuang, Y., 2006. Rabbits killing birds revisited. Math. Biosci. 203, 100–123.
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/ecolmodel
Limited nest site availability helps seabirds to survive cat predation on islands Dominique Pontier a,∗ , David Fouchet a , Jo¨el Bried b , Narg`es Bahi-Jaber a a
UMR CNRS 5558 “Biom´etrie et Biologie Evolutive”; Universit´e de Lyon; Universit´e Lyon 1, 43 Boulevard du 11 Novembre 1918, 69622 Villeurbanne Cedex, France b Departamento de Oceanografia e Pescas, Centro do IMAR da Universidade dos Ac¸ores, 9901-862 Horta, Ac¸ores, Portugal
a r t i c l e
i n f o
a b s t r a c t
Article history:
Introduced cats Felis catus have a high detrimental impact on native seabirds on islands,
Received 17 March 2007
especially when alien preys, like rabbits Oryctolagus cuniculus, co-occur. Seabirds are highly
Received in revised form
vulnerable because of their long reproductive cycles, slow turn-over of generations and
27 January 2008
the absence of efficient behaviour against terrestrial predators, especially in some burrow-
Accepted 15 February 2008
nesting species. Through a deterministic modelling approach, we explored a neglected
Published on line 3 April 2008
mechanism that may explain the resistance of some seabird species to cat predation. It was indeed observed that seabirds may compete for nest sites. As a consequence, part of
Keywords:
the breeders foregoes breeding when nest sites are a limiting resource. Our model linked
Oceanic islands
the dynamics of cats with that of seabird species. We showed that the annual impact of cats
Domestic cat
on seabirds was lower when seabirds faced competition for burrows than when the latter
Seabirds
were not a limiting resource. This was due to the fact that limited nest site availability pre-
Nest site availability
vents an optimal growth of the cat population. Cats in turn cannot manage to exterminate
Alternative preys
all the prospecting birds during the same breeding season. The limitation of the number of
Prey-predator relationships
nest sites generates a mechanism leading the bird population to conserve a large pool of
Deterministic model
sexually mature individuals while only slightly reducing the production of juveniles in the colony. This pool of floaters may play an important role in natural populations by buffering the decrease in colony size during years with harsh environmental conditions on land. In combination with buffer mechanisms, the limitation of the number of nest sites may greatly improve the chances of survival of bird populations facing predation. © 2008 Published by Elsevier B.V.
1.
Introduction
Insular ecosystems are extremely sensitive to anthropogenic perturbations (Jarvis, 1979; Byrne, 1980; Dowding and Murphy, 2001). Most of the bird extinctions attributable to introduced predators, if not all, have occurred on islands (Stattersfield and Capper, 2000). Amongst introduced predators, cats Felis catus (L.) have played a major role in the extirpation of indigenous seabirds on islands (Moors and Atkinson, 1984; Dowding
∗
Corresponding author. Tel.: +33 4 72 43 13 37; fax: +33 4 72 43 13 88. E-mail address: [email protected] (D. Pontier). 0304-3800/$ – see front matter © 2008 Published by Elsevier B.V. doi:10.1016/j.ecolmodel.2008.02.010
and Murphy, 2001). Their impact on the native avifauna was the most dramatic on islands where alien preys co-occur (Atkinson, 1985), like for instance rabbits Oryctolagus cuniculus (L.) (Weimerskirch et al., 1989). Because they evolved in areas inaccessible, or poorly accessible, to mammals (that is, oceanic islands or inaccessible cliffs on mainlands), seabirds generally lack the adaptations, both in their behaviour and their life-history strategies, that would enable them to cope successfully with alien predators (Lack,
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
1968; Moors et al., 1992). Moreover, most seabird species show quasi-absolute fidelity to their breeding island (review in Bried and Jouventin, 2002), which will increase the probability of extermination by alien predators. The risk is higher for burrowing petrels which leave their single chick unattended in their burrows when a few days old (Warham, 1990). Seabirds also have a slow turn-over of generations: they generally have a low fecundity, the extreme being a single-egg clutch without replacement (especially in all albatrosses and petrels), a delayed sexual maturity (the minimum age at first breeding ranges from 2 to 8 years) and they can live up to several tens of years (reviews in Warham, 1996; Bried and Jouventin, 2002; Appendix 2 in Schreiber and Burger, 2002). They also have long breeding cycles, that is, long periods of colony attendance: from several weeks to more than one year (review in Bried and Jouventin, 2002). Because seabirds are long-lived, the dynamics of their populations is extremely sensitive to a small increase in adult mortality (Weimerskirch et al., 1987; Cuthbert et al., 2001). In addition, introduced predators have an indirect impact on seabirds by increasing the incidence of nest switching and divorce, which can result in the loss of several breeding years (Bried and Jouventin, 1999). Studying the impact of introduced predators is of paramount importance for conservation biology, especially on subantarctic islands where the native avifauna shows high levels of endemism and is often made up essentially by seabirds (Marchant and Higgins, 1990; Higgins and Davies, 1996; Shirihai, 2002). The existence of simplified trophic webs on many subantarcic islands but also on tropical and temperate islands free of native mammals where avian communities exhibit similar characteristics makes such islands excellent models to assess the impact of introduced predators on the marine avifauna. Previous theoretical studies, despite increasing accuracy (Courchamp et al., 2000; Gaucel et al., 2005; Zhang et al., 2006), tended to neglect important aspects of seabird population dynamics. In particular, while competition for nest sites is common in seabirds (Burger and Shisler, 1978; Ramos et al., 1997; Quintana and Yorio, 1998), the role of nest site availability in the response of seabirds to predation has never been investigated. In this paper, using a modelling approach, we show that the impact of alien predators on seabirds may differ if we account for the limitation in the number of nest sites. We consider a situation similar to that on the main island of the Kerguelen archipelago, southern Indian Ocean, for which there are reliable estimates of cat densities (Say et al., 2002), and because 33 out of the 35 avian species breeding at this locality are seabird species (Weimerskirch et al., 1989). Our model incorporates the cat as a top-level predator and a “theoretical” cavity-nesting seabird species (whose demographic parameters fall within the range of those observed in burrow-nesting seabirds and where breeders show absolute year-to-year fidelity to their colony) that faces variations in nest site availability. We also consider that cats prey on birds only during the breeding period of the latter, and that nests are easily accessible to cats. If too few nest sites are available, all the birds that are in breeding condition cannot breed during a given breeding season and cats suffer from the decrease in prey availability. The growth of the cat population is limited, resulting in a decreased predation pressure on birds. This rather simple mechanism creates
317
a scenario that may enable seabird species to persist longer than they would if all potential breeders could easily obtain a nest. Finally, we discuss the implications of our results for seabird conservation.
2.
Materials and methods
2.1.
Considerations and definitions
In our model, and like on the Kerguelen Islands, seabird colony size varies between 103 and 105 pairs, depending on species (see, e.g., Weimerskirch et al., 1989; Marchant and Higgins, 1990), and cat density varies between 1 and 2 cats/km2 (Say et al., 2002). Our model takes the limiting resources for seabirds and the seasonality of reproduction of prey species into account. Birds leave the colony at the end of the breeding season and remain at sea until the onset of the next breeding cycle. Cats prey on birds but also on alien preys, for instance rabbits and mice Mus musculus which are present in the same area on the main island of the Kerguelen archipelago (Pontier et al., 2002) and provide feral cats with a primary prey species, helping them to maintain their numbers during the winter and to spread (Moors et al., 1992; Pontier et al., 2002; Gaucel et al., 2005). Because of their opportunistic behaviour, cats take birds at the beginning of the breeding season (i.e., from the pair formation period onwards) and then, the proportion of birds in their diet decreases towards the end of the breeding period (Bloomer and Bester, 1990). The birds’ life cycle is modelled as follows (Fig. 1). Each individual belongs to only one of the following categories: (1) juveniles (Bj ) that is, young from hatching until their departure to sea, (2) immatures (Bi ) that is, individuals between fledging and the year when they will become socially mature, (3) “socially mature” individuals (Bm ) that is, all the birds that have returned to the colony at least once since fledging (some of them will remain at sea without returning ashore during the currently starting breeding season), plus the birds that will return ashore for the first time of their life during the currently starting breeding season, (4) prospectors (Bp ), i.e., all the individuals that come ashore during the current breeding season (this category includes the former breeders that have regained breeding condition, the individuals reaching breeding condition for the first time and the individuals that are not physiologically able to breed but that come ashore to secure a nest and/or a mate for the next year; see, e.g., Nelson, 1983; Warham, 1990) and (5) breeders (Bn ), that is, the prospectors that effectively attempt to breed (i.e., that lay at least an egg or whose female lays at least an egg). The prospectors that do not attempt to breed during the current breeding season are thereafter referred to as “floaters”. At the end of the breeding season, breeders return to the Bm compartment. Let bB be the birth rate of birds. We note 1 the rate at which immature becomes socially mature. By definition, only prospectors return to the colony and their number is limited by Kb through food availability at sea. Indeed, the distribution and abundance of marine organisms (Waluda et al., 1999), linked to variations in oceanographic conditions (e.g., Deacon, 1977), directly influence the amount of body reserves stored by the individuals, as well as their breeding perfor-
318
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Fig. 1 – Schematic life cycle of seabirds.
mances, age at sexual maturity, colony attendance and/or the proportion of breeders (Drent and Daan, 1980; Erikstad et al., 1998; Weimerskirch, 2002). Futhermore, we assume that birds become physiologically able to breed – for the first time or to resume breeding – only at the end of the non-breeding season, resulting in a seasonal rate of acquisition of breeding condition 2 (t). Without competition, the rate at which prospectors obtain a nest is (t). Bird reproduction is regulated by the number of available nest sites Kn , i.e., only a given proportion of prospecting adults perform breeding attempts. The rate of departures from the colony after the breeding period (during incubation, chickrearing or when the chicks fledge, depending on the outcome of the breeding attempts) is given by (t), chosen in order to have birds attending the colony approximately for 6 months (see below). Finally, we note mJ the mortality rate of chicks and mA that of immature and adult birds. Cats (C) may prey on both birds (in the nest) and alien preys, e.g., rabbits, which are not explicitly modelled here. The per capita bird intake by cats depends on bird density via the functional response o = (1 − e−d(Bj +Bn ) ) (Barlow and Choquenot, 2002), where is the maximum per capita bird intake, and d represents the easiness to capture birds by describing how quickly the intake decreases with declining prey density. Among all the birds they consume, cats eat a proportion Bj /(Bj + Bn ) of Bj and a proportion Bn /(Bj + Bn ) of Bn . The growth rate of the cat population is an affine function of the per capita intake rC (1 − e−d(Bj +Bn ) ) − mC , where mC is the rate of decline of the population when birds are absent from the colony during the non-breeding period and rC −mC is the maximum growth rate of the cat population. We assume that in the absence of birds, cats are able to maintain their numbers at a low density Cmin , thanks to the presence of alternative preys, mainly rabbits (Pontier et al., 2002). We assume that this is the only way in which alternative preys influence the system. To simplify we do not model them explicitly. We modify the growth rate of
the cat population by introducing the term Cmin that ensures that C is always greater than Cmin . The system of equations reads dBj dt
= bB Bn − (mJ + (t))Bj −
Bj Bj + Bn
(1 − e−d(Bj +Bn ) )C
dBi = (t)Bj − 1 Bi − mA Bi dt
Bp + Bn dBm = (t)Bn + 1 Bi − 2 (t) 1 − dt Kb
dBp Bp + Bn = 2 (t) 1 − dt Kb
Bm − mA Bm
Bm − mA Bp − (t) 1 −
dBn Bn = −(t)Bn + (t) 1 − dt Kn −
Bn Kn
Bp
Bp − ma Bn
Bn (1 − e−d(Bj +Bn ) )C Bj + Bn
dC = rC (1 − e−d(Bj +Bn ) )C − mC (C − Cmin ) dt
2.2.
Simulations
Basic values for the parameters are given in Table 1. We define the parameter values on a yearly scale and take the seasonal variations into account. The rate at which birds attend the colony is (t)=0 (|cos(2t)| + cos(2t)) with 0 = 10 year−1 . We vary the parameter Kn from small values representing high competition for nesting to high values for modelling the absence of competition. The rate at which birds become socially mature is 1 = 0.25 year−1 and the rate at which they reach breeding condition is bounded by 2 (t) = 0 (|cos(2(t + 0.5))|+cos(2(t + 0.5))) with 0 = 10 year−1
319
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Table 1 – Basic values of the parameters (time unit is the year) Parameter
Symbol
Bird hatching rate Bird juvenile mortality rate Bird adult mortality rate Maximum number of prospector Number of nests available Maximum rate at which birds become socially mature Maximum rate at which birds reach breeding condition Maximum rate at which birds attend the colony Maximum rate at which birds leave the colony after breeding Rate of cat decay during the absence of birds Bird predation-induced maximum growth rate Minimum number of cats Maximum prey intake per cat Coefficient of cat predation according to the number of birds
and Kb = 105 birds. The rate at which breeders leave the colony once their breeding attempt ends (i.e., during incubation or chick-rearing if they fail, when the young fledge if they are successful) is given by (t) = 0 (|cos(2(t + 0.5))| + cos(2(t + 0.5))), and 0 = 10 year−1 . Bird hatching and mortality rates are bB = 1 year−1 , mJ = 2.4 year−1 (the mean hatching rate is one chick per breeding pair and the probability of fledging is approximately equal to e−2.4×0.5 = 0.3), and mA = 0.09 year−1 (corresponding to an annual adult survival rate of 0.91). The rate of decrease in cat population size in the absence of birds is mC = 2 year−1 and its maximum growth rate is rC −mC = 8 year−1 . The cat population can decline to Cmin = 3 cats when birds are absent from the colony. The maximum prey intake by cats is = 365 birds year−1 and d = 10−5 bird−1 . Simulations of the model were performed with Matlab (The MathWorks Inc.) using an explicit Runge–Kutta method.
3.
Results
The system reaches a periodic state for all tested sets of parameters. The model is analysed at the cyclic state. In the following section, we will distinguish two ways of measuring bird abundance: population size, Bj + Bi + Bm + Bp + Bn , and colony size, Bj + Bp + Bn . The latter is made up by the number of all observable birds at the colony, i.e., juveniles, floaters (non-breeding prospectors) and breeders.
3.1.
Demography of seabirds without predation
In the absence of predation, the birds’ distribution among the different categories of the population depends on the number of nest sites. When there are plenty of nests available (Fig. 2a), most of the birds in breeding condition obtain a nest and thus achieve reproduction, which leads to a high number of juveniles and then immatures. Only a small proportion of the latter gets a sufficient condition to return ashore (to prospect or to breed), due to high competition for food at sea. The population is mainly made up of immature birds which remain at sea (here approximately 2/3). On the contrary, when the number of nest sites is low (Fig. 2b), fewer young are produced, which, in the long term, results in decreased competition for food at sea. This leads to a lower proportion of immatures (as defined
bB mJ mA Kb Kn 1 0 0 0 mC rC Cmin d
Basic value 1 2.4 0.09 105 Variable 0.25 10 10 10 2 10 3 365 10−5
in Section 2) in the population compared to the previous situation (here only 1/3). Interestingly, colony size (Bj + Bp + Bn ) is less affected than population size by a decrease in the number of nest sites, especially when the number of nest sites is high enough (Fig. 2c).
3.2.
Effect of predation on birds
A low number of nest sites leads to a lower food supply for cats, limiting the growth of cat population (Fig. 3a). Under these conditions, the birds that attend the colony (i.e., floaters and breeders) are thus little affected by cats (Fig. 3b). As cats exert a low predation pressure, floaters have difficulties to obtain a nest and they still represent a high proportion of the bird population (see Fig. 3b). The other parts of the bird population are all becoming more important as the number of nest sites increases (see Fig. 3b). Unsurprisingly, an increase in the number of nest sites results in higher breeding numbers (see Fig. 3b) and hence in more juveniles, immature and mature birds (see Fig. 3b, the three categories are gathered since they show very similar responses to the variations in the number of nest sites). Interestingly, the reduction in predation pressure resulting from a decreased number of nest sites leads to a hump in both population and colony sizes (Fig. 3c). In other words, the negative consequences of a low number of nest sites for births are more than compensated by the positive impact of the reduction in cat numbers. The reduction of the cat population resulting from decreased nest availability is the key factor allowing observing the hump: if we assume a constant number of cats, bird population/colony size becomes a regularly increasing function of the number of nest sites (Fig. 3d). This means that a nest shortage always reduces seabird population/colony size, cat population size being equal. The simple fact that floaters suffer lower mortality than breeders does not explain why a decrease in nest site availability may lead to an increase in bird numbers. The fact that floaters can replace breeders when the latter die is not necessarily beneficial in the situation invoked here. As observed in Fig. 3b, between the optimal number of nests in terms of population size and unlimited number of nests, there is only an increase by around 10% in seabird breeding numbers. In other words, providing more nests does not help to
320
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Fig. 2 – Seabird population in the absence of cat predation. (a and b) Number of birds in each category: juveniles (Bj ), individuals that remain at sea (Bi + Bm ) and individuals that come ashore (i.e., breeders and non-breeding prospectors, Bn + Bp ) and total number of birds in the population (BT ) when nest site availability is not limiting (a) and for 2 × 104 nest sites (b). (c) Mean population (solid line) and colony (dashed line) sizes as a function of the number of nest sites (Kn ).
increase the number of birds that can reproduce. At this stage, the occupancy of vacant nests by floaters only leads to higher predation pressure. Filling vacancies consumes the pool of non-breeding prospectors while only slightly increasing the number of breeders. Several parameters were found to change significantly the consequences of a limited number of nest sites (Fig. 4). We defined as an index of the impact of nest shortage the population and colony size obtained when the number of nest sites is optimal. For example, Fig. 3c reveals a maximum number of birds of 9.6 × 104 (for 3.6 × 104 nest sites), versus 7.9 × 104 when Kn is infinity, i.e., when the number of nest site available is several order of magnitude higher than the number of birds that can reproduce. Thus, a nest shortage can lead up to an increase by 22% (9.6/7.9 = 1.22) in seabird numbers. This index allows estimating the sensitivity of the results to changes in parameter values. The number of birds reaching social maturity strongly depends on bB , mJ , mA and 1 , whereas the minimal impact of cats on birds depends on Cmin , mA and bB . When large numbers of individuals reach social maturity, socially mature birds act as a reservoir that may replace predated birds the next year. In such a context, a poor supply of nests allows conserving the reservoir. In contrast, when the number of nest sites is not limiting, the reservoir is emptied rapidly, and fewer individuals are able to reproduce.
The minimal impact of cats, i.e., the impact of Cmin cats on the bird population, is also important. It is mainly driven by three parameters. Cmin is the most intuitive because it acts directly on the number of predated birds. Bird growth rate parameters (mainly bB and mA ) are also important since a bird population with a more positive balance between births (or recruitment into the breeding population) and deaths can bear a more important predation pressure. When the minimal impact of cats is high, a higher predation pressure is not sustainable for the birds, and in turn decreasing bird numbers reach a threshold value beyond which cat numbers cannot increase any more. The beneficial effect of a decrease in cat population size is thus lower. The maximum number of prospectors (Kb , which is related to food availability at sea) is found to have only a little impact on the system. Multiplying Kb by two does not change anything (see Fig. 4). This is simply because the production of juveniles is never high enough to allow the number of prospectors to get closer to Kb . In other words, the food resources at sea are not limiting. Dividing Kb by two makes the resources limiting, but again, this only slightly affects seabird population size (Fig. 4). Fewer birds reach breeding condition, which explains why colony size is decreased (Fig. 4). Because, however, we assumed that food supply at sea affects the proportion of birds reaching breeding condition but not bird survival, dividing Kb by two only slightly affects population size (Fig. 4).
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
321
Fig. 3 – Interaction between predation and nest site availability. (a) Average number of cats at the equilibrium according to the number of available nests (Kn ); (b) average number of birds in each category at the equilibrium: non-breeding birds (Bj + Bi + Bm ), prospectors (Bp ) and breeders (Bn ), according to the number of nest sites (Kn ); (c) population (bold solid lines) and colony (bold dashed lines) sizes against the number of nest sites (Kn ) for usual values of the parameters as defined in Table 1 and (d) same as (c) but with a constant number of cats C = 50. Similar simulations with different numbers of cats were conducted and provided similar results (results not shown). In (c) and (d) asymptotic values, i.e., when the number of nest sites in the colony is several orders of magnitude higher than the number of birds that can reproduce, are given by solid and dashed horizontal lines for the population and the colony, respectively.
Fig. 4 – Effect on the maximum increase in population and colony sizes (when nest sites are in limited supply) when parameters are divided or multiplied by 2. Notations: mJ /2 means that mJ is here twice as low as its usual value. The system is particularly sensitive to mA and bB , because these parameters drive both the proportion of birds becoming socially mature and the impact of the minimal number of cats Cmin .
The model also predicts that bird population can become extinct through predation (results not shown). Extinction generally occurs when the predation exerted by the minimal number of cats (Cmin ) added to bird mortality from other causes is not balanced by births or recruitment. Poor nest availability cannot reduce the number of cats below Cmin since cats can survive by eating other preys. As a consequence,
a shortage of nests cannot prevent bird extinction in this case.
4.
Discussion
The aim of this paper was to assess the consequences of nest site availability for bird populations suffering predation. It is
322
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
important to note that the results presented here are mainly qualitative. Parameter values may vary greatly among bird species and/or colony sites. Interestingly, the limitation of the number of nest sites has been found to have positive consequences for the bird population despite decreased offspring productivity. Limited nest site availability decreases the impact of predation by limiting prey availability for cats. Cats will not exterminate all the adult birds during the same breeding season as long as the number of nest sites remains insufficient, like for example in mixedspecies colonies where a large number of birds compete for the same breeding sites (Burger and Shisler, 1978; Quintana and Yorio, 1998). The presence of a reservoir of individuals (floaters here) that can quickly replace effective breeders is often considered positive for the bird population because it allows breeding numbers to remain stable during a certain time lag (Walters et al., 2002; Grimm et al., 2005). This is not the case when considering the mechanism described here, where the turn-over of birds in vacant nests enables cat numbers to increase, which in turn limits the number of breeding birds. Finally, close to the optimal number of nest sites, the number of breeding birds is only slightly increased when vacancies are quickly filled. The cost of this slight increase is a large decrease in the number of floaters. This may have very important consequences for the stochastic extinction of the bird population (not modelled here), through a mechanism known as the buffer effect (Grimm et al., 2003, 2005; Penteriani et al., 2006). The fact that our model does not predict extinction does not mean that it will not occur. Bird numbers exhibit important inter-annual variations. During a year with harsh environmental conditions on land, an intense predation pressure can drive the bird population toward extinction. The large proportion of sexually mature birds that do not breed acts like a buffer, allowing to compensate for losses in the subpopulation of breeders and thus improves the persistence of the bird population during years with harsh environmental conditions. Analogous buffer mechanisms have been described in some other species (Grimm et al., 2003, 2005; Penteriani et al., 2006). The buffer effect and the control of predators are two different but complementary mechanisms that result from variations in nest site availability and that can highly enhance the chances of persistence for the bird population (Fig. 5). Our model shows that this is especially true for species with a relatively low mortality and/or a high fecundity or at localities where cats cannot maintain significant numbers in the absence of birds. The results presented here have implications for seabird conservation, not only on subantarctic islands but also on tropical and temperate islands where seabirds represent the bulk of the avian community and where there are no indigenous terrestrial mammals. They may explain the contrasted responses to predator removal campaigns, with some studies showing an increase in seabird breeding numbers following predator elimination whereas others show no effect or ˆ e´ and Sutherland, 1997). The relationship even decreases (Cot between predation and bird population dynamics can vary among seabird species, here through variations in colony carrying capacity. Population biology studies are thus important to estimate the level of competition for nests and/or nest site
Fig. 5 – Interplay between the control of predation mechanism and the buffer effect. Limited nest site availability controls predation, finally leading to a larger pool of prospectors. This pool may enable seabird breeding numbers to remain stable, decreasing the risks of extinction for the bird population through a buffer effect.
availability at a given locality in order to identify the most vulnerable native seabird species when alien predators occur. Management effort could then be directed on these particular seabird species. Furthermore, one aspect not considered here is the fact that the saturation of the breeding habitat (Chastel et al., 1993; Inchausti and Weimerskirch, 2002; Burg et al., 2003) and/or the disturbance due to alien predators (Thibault, 1995; Oro et al., 1999) may make a seabird colony less attractive for prospectors and/or elicit bird dispersal (Austin, 1940). The flux of immigrants (essentially immatures in the species where breeders are faithful to their colony) from distant sources is rarely invoked to explain the demographic patterns observed in seabirds. Inchausti and Weimerskirch (2002) and Jenouvrier et al. (2003) showed that even low immigration rates (as low as 3%) each year could affect seabird population dynamics at distant localities. By encouraging bird dispersal, the saturation of the breeding habitat could thus allow a higher persistence of bird species facing predation. Even if small numbers of individuals colonize an area where alien predators occur, this might be sufficient to maintain locally both the prey and the predator. The existence of distant sources might explain why the breeding numbers of yelkouan shearwaters Puffinus yelkouan (Acerbi) seem to remain stable on Port-Cros Island despite huge predation by cats which were introduced in the 17th century (Bourgeois, 2004). Coupling nest site availability with the occurrence of bird dispersal when studying metapopulation dynamics, could improve our understanding of bird-predator relationships.
Acknowledgements This work was supported by IPEV (programme “Popchat” n◦ 279), the “Environnement & Sustainability” Department of CNRS (“Zone Atelier de Recherche sur l’Environnement
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Antarctique et Subantarctique”) and LIA 197 “ReaDiLab”. It was also part of JB’s postdoctoral contract at the Instituto do Mar (FCT grant SFRH/BPD/20291/2004). We thank J.-M. Gaillard, M. Guiserix and two anonymous referees for their helpful comments on an earlier version of the paper.
references
Atkinson, I.A.E., 1985. The spread of commensal species of Rattus rattus to oceanic islands and their effects on islands avifaunas. In: Moors, P.J. (Ed.), Conservation of Island Birds. Technical Publication No. 3. ICBP, Cambridge, UK, pp. 35–81. Austin, O.L., 1940. Some aspects of individual distribution in the Cape Cod Tern colonies. Bird Banding 11, 155–159. Barlow, N., Choquenot, D., 2002. Predicting the impact and control of stoats: a review of modelling approaches. Sci. Conserv. 191, 46. Bloomer, J.P., Bester, M.N., 1990. Diet of a declining feral cat Felis catus population on Marion Island. S. Afr. J. Wildl. Res. 20, 1–4. Bourgeois, K., 2004. Ecologie et conservation d’un oiseau marin ´ ´ ´ Puffinus yelkouan-predation ´ endemique de Mediterran ee, par ´ le chat haret et selection de l’habitat dans le Parc National de ˆ ´ Port-Cros. ‘Diplome d’Etudes Superieures’ Thesis. University of Aix-Marseille 3. Bried, J., Jouventin, P., 1999. Influence of breeding success on fidelity in long-lived birds: an experimental study. J. Avian Biol. 30, 392–398. Bried, J., Jouventin, P., 2002. Site and mate choice in seabirds: an evolutionary approach. In: Schreiber, E.A., Burger, J. (Eds.), Biology of Marine Birds. CRC Press, Boca Raton, FL, pp. 263–305. Burg, T.M., Lomax, J., Almond, R., Brooke, M., de, L., Amos, W., 2003. Unravelling dispersal patterns in an expanding population of a highly mobile seabird, the northern fulmar (Fulmarus glacialis). Proc. R. Soc. Lond. B 270, 979–984. Burger, J., Shisler, J., 1978. Nest-site selection and competition interactions of herring and laughing gulls in New Jersey. Auk 95, 252–266. Byrne, R., 1980. Man and the variable vulnerability of island life. A study of recent vegetation change in the Bahamas. Atoll Res. Bull. 240, 200 pp. Chastel, O., Weimerskirch, H., Jouventin, P., 1993. High annual variability in reproduction success and survival of an Antarctic seabird, the snow petrel (Pagodroma nivea). Oecologia 94, 278–285. ˆ e, ´ I.M., Sutherland, W.J., 1997. The effectiveness of removing Cot predators to protect bird populations. Conserv. Biol. 11, 395–405. Courchamp, F., Langlais, M., Sugihara, G., 2000. Rabbits killing birds: modelling the hyperpredation process. J. Anim. Ecol. 69, 154–164. Cuthbert, R., Fletcher, D., Davis, L.S., 2001. A sensitivity analysis of Hutton’s shearwater: prioritizing conservation research and management. Biol. Conserv. 100, 163–172. Deacon, G.E.R., 1977. Seasonal variations in the water temperature and salinity near South Georgia 1925–1937. Report 49. Institute of Oceanography Science, Wormley, UK. Dowding, J.E., Murphy, E.C., 2001. The impact of predation by introduced mammals in endemic shorebirds in New Zealand: a conservation perspective. Biol. Conserv. 99, 47– 64. Drent, R.H., Daan, S., 1980. The prudent parent: energetic adjustment in avian breeding. Ardea 68, 225–252.
323
Erikstad, K.E., Fauchald, P., Tveraa, T., Steen, H., 1998. On the cost of reproduction in long-lived birds: the influence of environmental variability. Ecology 79, 1781–1788. Gaucel, S., Langlais, M., Pontier, D., 2005. Invading introduced species in insular heterogeneous environments. Ecol. Modell. 18, 62–75. Grimm, V., Dorndorf, N., Frey-Roos, F., Wissel, C., Wyszomirski, T., Arnold, W., 2003. Modelling the role of social behavior in the persistence of the alpine marmot Marmota marmota. Oikos 102, 124–136. Grimm, V., Revilla, E., Groeneveld, J., Kramer-Schadt, S., Schwager, M., Tews, J., Wichmann, M., Jeltsch, F., 2005. Importance of buffer mechanisms for population viability analysis. Conserv. Biol. 19, 578–580. Higgins, P.J., Davies, S.J.J.F., 1996. Handbook of Australian, New Zealand and Antarctic Birds, 3. Oxford University Press, Melbourne. Inchausti, P., Weimerskirch, H., 2002. Dispersal and metapopulation dynamics of an oceanic seabird, the wandering albatross, and its consequences for its response to long-line fisheries. J. Anim. Ecol. 71, 765–770. Jarvis, P.T., 1979. The ecology of plant and animal introductions. Prog. Phys. Geogr. 3, 187–214. Jenouvrier, S., Barbraud, C., Weimerskirch, H., 2003. Effects of climatic variability on the temporal population dynamics of southern fulmars. J. Anim. Ecol. 72, 576–587. Lack, D., 1968. Ecological Adaptations for Breeding in Birds. Methuen, London. Marchant, S., Higgins, P.J., 1990. Handbook of Australian, New Zealand and Antarctic Birds, 1A–1B. Oxford University Press, Melbourne. Moors, P.J., Atkinson, I.A.E., 1984. Predation on Seabirds by Introduced Animals, and Factors Affecting its Severity. International Council for Bird Preservation, Cambridge, UK, pp. 667–690, ICBP Techn. Publ. 2. Moors, P.J., Atkinson, I.A.E., Sherley, G.H., 1992. Reducing the rat threat to island birds. Bird Conserv. Int. 2, 93–114. Nelson, J.B., 1983. Contrasts in breeding strategies between some tropical and temperate marine pelecaniformes. Stud. Avian Biol. 8, 95–114. Oro, D., Pradel, R., Lebreton, J.-D., 1999. Food availability and nest predation influence life history traits in Audouin’s gull Larus audouinii. Oecologia 118, 438–445. Penteriani, V., Otalora, F., Ferrer, M., 2006. Floater dynamics can explain positive patterns of density-dependent fecundity in animal populations. Am. Nat. 168, 697–703. Pontier, D., Say, L., Debias, F., Bried, J., Thioulouse, J., Micol, T., Natoli, E., 2002. The diet of feral cats (Felis catus L.) at five sites on the Grande Terre, Kerguelen archipelago. Polar Biol. 25, 833–837. Quintana, F., Yorio, P., 1998. Competition for nest sites between kelp gulls (Larus dominicanus) and terns (Sterna maxima and S. eurygnatha) in Pentagonia. Auk 115, 1068–1071. Ramos, J.A., Monteiro, L.R., Sola, E., Moniz, Z., 1997. Characteristics and competition for nest cavities in burrowing Procellariiformes. Condor 99, 634–641. Say, L., Gaillard, J.-M., Pontier, D., 2002. Spatio-temporal variation in cat population density in a Sub-Antarctic environment. Polar Biol. 25, 90–95. Schreiber, E.A., Burger, J., 2002. Biology of Marine Birds. CRC Press, Boca Raton, FL. Shirihai, H., 2002. A Complete Guide to Antarctic Wildlife. Alula Press Oy, Degerby. Stattersfield, A.J., Capper, D.R., 2000. Threatened Birds of the World. BirdLife International and Lynx Edicions, Barcelona. Thibault, J.-C., 1995. Effect of predation by the black rat Rattus rattus on the breeding success of Cory’s shearwater Calonectris diomedea in Corsica. Mar. Ornithol. 23, 1–10.
324
e c o l o g i c a l m o d e l l i n g 2 1 4 ( 2 0 0 8 ) 316–324
Walters, J.R., Crowder, L.B., Priddy, J.A., 2002. Population viability analysis for red-cockaded woodpeckers using an individual-based model. Ecol. Appl. 12, 249–260. Waluda, C., Trathan, P., Rodhouse, P., 1999. Influence of oceanographic variability on recruitment in the Ilex argentinus Cephalopoda: Ommastrephidae fishery in the South Atlantic. Mar. Ecol. Progr. Ser. 183, 159–167. Warham, J., 1990. The Petrels. In: Their Ecology and Breeding Systems. Academic Press, London. Warham, J., 1996. The Behaviour, Population Biology and Physiology of the Petrels. Academic Press, London.
Weimerskirch, H., 2002. Seabird demography and its relationships with the marine environment. In: Schreiber, E.A., Burger, J. (Eds.), Biology of Marine Birds. CRC Press, Boca Raton, FL, pp. 115–135. Weimerskirch, H., Clobert, J., Jouventin, P., 1987. Survival in five southern albatrosses and its relationship with their life history. J. Anim. Ecol. 56, 1043–1055. Weimerskirch, H., Zotier, R., Jouventin, P., 1989. The avifauna of the Kerguelen Islands. Emu 89, 15–29. Zhang, J., Fan, M., Kuang, Y., 2006. Rabbits killing birds revisited. Math. Biosci. 203, 100–123.