Oecologia (2018) 186:129–139 https://doi.org/10.1007/s00442-017-4016-z POPULATION ECOLOGY – ORIGINAL RESEARCH Successional changes in trophic interactions support a mechanistic model of post‑fre population dynamics Annabel L. Smith1 Received: 7 December 2016 / Accepted: 16 November 2017 / Published online: 22 November 2017 © Springer-Verlag GmbH Germany, part of Springer Nature 2017 Abstract Models based on functional traits have limited power in predicting how animal populations respond to disturbance because they do not capture the range of demographic and biological factors that drive population dynamics, including variation in trophic interactions. I tested the hypothesis that successional changes in vegetation structure, which afected invertebrate abundance, would infuence growth rates and body condition in the early-successional, insectivorous gecko Nephrurus stel- latus. I captured geckos at 17 woodland sites spanning a succession gradient from 2 to 48 years post-fre. Body condition and growth rates were analysed as a function of the best-ftting fre-related predictor (invertebrate abundance or time since fre) with diferent combinations of the co-variates age, sex and location. Body condition in the whole population was positively afected by increasing invertebrate abundance and, in the adult population, this efect was most pronounced for females. There was strong support for a decline in growth rates in weight with time since fre. The results suggest that increased early- successional invertebrate abundance has fltered through to a higher trophic level with physiological benefts for insectivorous geckos. I integrated the new fndings about trophic interactions into a general conceptual model of mechanisms underlying post-fre population dynamics based on a long-term research programme. The model highlights how greater food availability during early succession could drive rapid population growth by contributing to previously reported enhanced reproduction and dispersal. This study provides a framework to understand links between ecological and physiological traits underlying post-fre population dynamics. Keywords Disturbance · Fire management · Functional traits · Habitat accommodation model · Pyrodiversity Introduction layers to the already complex suite of factors driving popu- lation dynamics (Burgess and Maron 2016; Davies et al. Temporal population dynamics depend on a range of inter- 2012). However, changes in post-fre population density acting demographic, biological and environmental factors are often interpreted as responses to changing habitat com- (Hodges et al. 2006) including predator–prey cycles (Rad- plexity with time since fre (e.g. Nimmo et al. 2012; Smith chuk et al. 2016), climatic fuctuations (Letnic and Dick- et al. 2013b), following the habitat accommodation model of man 2010) and density-dependent competition for resources animal succession (Caughley 1985; Fox 1982). Few studies (Forero et al. 2002). In fre-prone ecosystems, periodic dis- consider the interacting biological and demographic mecha- turbance and subsequent habitat succession add additional nisms which drive population dynamics and how the mecha- nisms themselves change with time since fre. This means we still lack the ability to predict how ecological communities Communicated by Jean-François Le Galliard. respond to variation in fre regimes, such as increases in fre Electronic supplementary material The online version of this frequency or changes in the spatial confguration of fres article (https://doi.org/10.1007/s00442-017-4016-z) contains (Kelly et al. 2011; Westgate et al. 2012). This predictive supplementary material, which is available to authorized users. capacity is essential to plan and implement fre manage- * Annabel L. Smith ment strategies that will conserve biodiversity (Driscoll et al. [email protected] 2010). It also underpins our knowledge of how ecosystems will respond to the rapid changes in fre regimes that are 1 Department of Zoology, School of Natural Sciences, Trinity College Dublin, Dublin 2, Ireland Vol.:(0123456789)1 3 130 Oecologia (2018) 186:129–139 occurring with climate change (Moritz et al. 2012; Silvério rates and body condition in a disturbance specialist, insec- et al. 2013). tivorous gecko Nephrurus stellatus. The gecko species Models based on functional ecological traits have proven has a strong and regionally consistent response to fre: valuable in understanding plant responses to varying fre population density increases for 10–15 years after fre regimes (e.g. Keith et al. 2007; Pausas 2015), but have lim- then declines sharply to less than 10% of peak density ited power in predicting disturbance responses in animal 30 years or more after fre (Driscoll and Henderson 2008; communities (Bargmann et al. 2016; Hu et al. 2013; Kelly Driscoll et al. 2012; Smith et al. 2013b). This refects vari- et al. 2011; Santos et al. 2014). One reason is that these ation in population density, not detectability (Smith et al. models do not capture the complex interactions that afect 2012). At the same study sites, invertebrate abundance is animals during post-fre succession such as rainfall (Green- higher in the frst 5 years after fre, compared with habi- ville et al. 2016), grazing (Driscoll et al. 2012), soil type tat that has not burned for over 30 years (Teasdale et al. (Davies et al. 2012) and inter-specifc interactions (St. Clair 2013). As burrowers, N. stellatus can probably survive et al. 2016). For example, Nimmo et al. (2014) showed that fre and recently burnt habitat might allow them to for- regional diferences in fre–vegetation relationships obscured age more efciently in open, sandy spaces (Smith et al. generalisable patterns of post-fre abundance in animal spe- 2013b) although direct evidence for this is currently lack- cies. Another reason is the severe lack of ecological and ing. Successional changes in factors such as predation life-history data for the majority of animal species targeted pressure (Hawlena et al. 2010), parasitism (McCoy et al. by these studies. Coarse ecological trait groups such as 2012) and thermal specialisation (Hossack et al. 2009) ‘low’ or ‘high mobility’ are often used (necessarily, given might also drive changes in body condition and growth scarce data, e.g. Santos et al. 2014) which might not provide rates. However, links between invertebrate abundance and enough resolution to detect complex ecological responses. growth trajectories in geckos would provide support for In other cases, so little is known about some species that the hypothesis that changes in trophic interactions infu- they cannot be assigned to even coarse groups (Smith et al. ence post-fre population dynamics. 2013b). Improved ecological data for a range of species in Long-term mechanistic studies of single species have the community are needed. At the individual species level, greatly advanced ecological knowledge of temporal popu- data showing how ecological traits and demographic rates lation dynamics [classic examples include studies of voles change during succession will provide deeper insight into (Radchuk et al. 2016), lemmings (Therrien et al. 2014) the mechanisms underlying post-fre population dynamics. and snowshoe hares (Hodges et al. 2006)]. The fre ecol- A range of processes might contribute to post-distur- ogy of N. stellatus has now been studied for over a dec- bance population dynamics within animal species includ- ade (Driscoll and Henderson 2008; Driscoll et al. 2012; ing changes in dispersal rates (Templeton et al. 2011), sur- Smith et al. 2012, 2013a, b) and the species could make vival and reproductive rates (Sanz-Aguilar et al. 2011) and a valuable contribution to our understanding of post-fre intraspecifc social interactions (Banks et al. 2012). Trophic population dynamics. As population density increases interactions are another potentially important driver (Bow- after fre, survival rates are low while reproductive rates man et al. 2016). For example, increasing reproduction and almost double, ofsetting the lower survival and contrib- local abundance of plant species after fre can be partially uting to the rapid increase in population density (Smith driven by disrupted interactions with antagonistic herbivores et al. 2012). Spatial genetic structure is greatest during (García et al. 2016; St. Clair et al. 2016). Much of the work very early succession, probably refecting rapid population on post-fre trophic interactions has focussed on plant–her- expansion from a very small number of individuals at the bivore relationships (e.g. Caut et al. 2014; Cherry et al. time of the fre (Smith et al. 2016b). As population density 2016) or predation of native mammals by exotic carnivores declines, genetic diversity and rates of dispersal and move- (e.g. Leahy et al. 2016; McGregor et al. 2014). Knowledge ment also decline (Smith et al. 2016b). Given this detailed of interactions at a wider range of trophic levels is needed body of knowledge, the second aim of the current study to develop a general understanding of how this process was to integrate the new fndings about body condition contributes to post-fre population dynamics (Di Stefano and growth rate into a general conceptual model of mecha- et al. 2014). Furthermore, understanding how successional nisms behind post-fre population dynamics. This in-depth changes in trophic interactions relate to other mechanisms investigation of a single species provides a framework to driving population dynamics is necessary to advance basic understand