Resource selection in an apex predator and variation in response to local landscape characteristics
Introduction
Rapid changes in landscape composition and structure represent major threats to biodiversity worldwide (Candolin and Wong, 2012; Ripple et al., 2014, Ripple et al., 2015). Typically leading to habitat loss (Barnosky et al., 2011), fragmentation (Fahrig, 2003), and increased barrier effects (Seidler et al., 2015), these changes may limit animal movement and dispersal and result in widespread reduction of core ecological processes (Haddad et al., 2015; Tucker et al., 2018). Investigating changes in species' resource selection behavior in response to landscape alteration is critical for developing relevant conservation management plans to facilitate species survival (Van Buskirk, 2012).
Natural or anthropogenic change in landscape structure, however, often vary across broad spatial scales (e.g. regions, biomes, continents), affecting individuals and populations in different ways (Mysterud and Ims, 1998). As a result, responses to temporally-dynamic and/or spatially heterogeneous resources may vary considerably across a species' geographic range, presenting significant challenges for land managers and decision makers (Boyce et al., 2002; Roever et al., 2012). Conversely, some resources may be used consistently across a species' geographic range, regardless of their underlying availability. African elephants (Loxodonta africana), for example, show consistent affinity for areas of low slope, high tree cover, and greater distance from human settlement (Roever et al., 2012). Despite the growing number of studies investigating how species “adjust” their resource selection across rapidly changing environments, few studies have performed analyses at a scale that encompasses a species' full geographic range (although see Parmesan, 2006), largely because range-wide, regional or continental-scale data are usually not available.
Resource selection functions (RSFs) have been widely applied to investigate species-habitat relationships (e.g., Boyce and McDonald, 1999; Gillies et al., 2006; Roever et al., 2012; Lehman et al., 2016; Stabach et al., 2016). For instance, RSFs are frequently used to compare used resources at observed telemetry locations (i.e., ‘use’ locations) with attributes of randomly-selected locations that are potentially available (i.e., ‘pseudo-absence’ locations). Resource selection functions may differ based on how the representative sample of available locations is selected, potentially influencing the scale of biologic inference (for a review of RSFs, see Lele et al., 2013).
Over the past decade, discrete choice models have emerged as a promising alternative to traditional logistic regression approaches for assessing resource preferences (Cooper and Millspaugh, 1999; McDonald et al., 2006; Thomas et al., 2006; Carter et al., 2010; Rota et al., 2014). Unlike logistic regression models, discrete choice models allow the researcher to develop different ‘choice sets’ for each independent observation. In doing so, discrete choice models provide a robust framework for accounting for changes in resource quality or availability through space and time. Furthermore, adjusting the spatial scale at which locally-available alternatives are selected allows the researcher to identify and compare drivers of resource selection behavior across a range of spatial scales (McDonald et al., 2006; Bonnot et al., 2011). Discrete choice models also provide an opportunity to test for potential functional responses, or context-dependent resource selection, by relating individual behavior to local habitat characteristics (e.g., Roever et al., 2012).
Jaguar (Panthera onca) is a widely-distributed apex predator, ranging across the Americas and inhabiting a wide variety of habitats, stretching from tropical moist forest to tropical dry forest or xeric areas (Sanderson et al., 2002). The species is found across a gradient of human disturbance (Jędrzejewski et al., 2018), which has been shown to affect its movement patterns and increase local extinction risk (Morato et al., 2016; Morato et al., 2013). Across many regions, jaguar populations have plummeted over the past few decades and are considered threatened in several countries (De La Torre et al., 2017a). The species, however, is also highly adaptable, making them a good candidate to study how apex predators respond to a gradient of land use change. While many resource selection studies exist on jaguar (e.g, Conde et al., 2010; Foster et al., 2010; Colchero et al., 2011; Cullen Jr et al., 2013; De La Torre et al., 2017b), none have addressed resource selection across a large range of biomes due to data limitations and the difficulty of combining datasets across a range of disturbance.
Here, we applied a Bayesian framework to the analysis of discrete choice models in order to characterize patterns of resource selection of the jaguar (Panthera onca), incorporating the largest telemetry dataset available to date for the species that covers approximately 30% of the species distribution. We developed resource selection models to compare resource selection between two different scales of inference: home range scale (coarse) and foraging scale (fine), corresponding to Johnson's (1980) 3rd and 4th order of resource selection, respectively. We were most interested in: 1) identifying primary drivers of jaguar resource use at multiple spatial scales and 2) examining how resource selection behaviors differ among gender and individual as a result of differences in resource availability across the landscape.
Section snippets
Study areas
Our study areas spanned the southern portion of the jaguar's distribution, covering areas in Brazil and NE Argentina (~900,000 ha), and included eight study sites encompassing five biomes: 1) Mamirauá Sustainable Reserve (MSR, Amazon, 1,124,000 ha), 2) Serra da Capivara National Park (CA, Caatinga, 100,000 ha), 3) private lands in the Cerrado (CE), 4) Iguazú and Iguaçu National Parks (PNI, Atlantic Forest, 200,000 ha), 5) Invinhema State Park (ISP, Atlantic Forest, 73,000 ha), 6) Morro do Diabo
Resource selection
At the home range scale, the relative probability of a jaguar using a location was positively associated with the local percent forest cover for both females (μ5= 0.51, 95% Highest Posterior Density Interval (HPDI) – [0.20, 0.86]) and males (μ5= 0.71, 95% HPDI = [0.30, 1.15]). Jaguars also selected for areas closer to rivers, with relative probability of use decreasing with increasing distance to river in both females (μ3= −5.95, 95% HPDI – [−9.86, −2.06]) and males (μ3= −9.16, 95% HPDI –
Discussion
Jaguars are widely recognized as a focal and umbrella species for biodiversity conservation planning at regional and countrywide scales (Rabinowitz and Zeller, 2010; Silveira et al., 2014). However, the effectiveness of jaguar conservation strategies depends on understanding how resource selection relates to landscape characteristics and how the response to these characteristics differs across the species' geographic range (Silveira et al., 2014; Watkins et al., 2015). Our approach allowed us
Acknowledgements
We thank Antonio Carlos Csermak Junior, Apolonio N. S. Rodrigues, Cristina Gianni, Dale Anderson, Jorge Luiz Pegoraro, Ivan Carlos Baptiston, Leanes Silva, Mariella Burti, Paulo Roberto Amaral, Paulo Baptista, Pollyana Motinha, Rafael Garay, Raphael Xavier, Rogério Silva de Jesus, Silvano Gianni, Tarcizio Paula, Thiago Luczinski, Valdomiro Lemos, Wendy Debbas and the members and volunteers of Proyecto Yaguarete for supporting on animal capture and monitoring.
Studies were funded by FAPESP (
References (78)
- et al.
Relating populations to habitats using resource selection functions
Trends Ecol. Evol.
(1999) Evaluating resource selection functions
Ecol. Model.
(2002)Modelling the risk of livestock depredation by jaguar in the Transamazon Highway, Brazil
Basic Appl. Ecol.
(2015)Sex matters: modeling male and female habitat differences for jaguar conservation
Biol. Conserv.
(2010)Understanding species persistence for defining conservation actions: a management landscape for jaguars in the Atlantic Forest
Biol. Conserv.
(2013)Habitat availability and connectivity for jaguars (Panthera onca) in the Southern Mayan Forest: conservation priorities for a fragmented landscape
Biol. Conserv.
(2017)Human-predator-prey conflicts: ecological correlates, prey losses and patterns of management
Biol. Conserv.
(2005)Predicting carnivore distribution and extirpation rate based on human impacts and productivity factors; assessment of the state of jaguar (Panthera onca) in Venezuela
Biol. Conserv.
(2017)- et al.
Predicting ranchers' intention to kill jaguars: case studies in Amazonia and Pantanal
Biol. Conserv.
(2012) - et al.
A range-wide model of landscape connectivity and conservation for the jaguar, Panthera onca
Biol. Conserv.
(2010)
Space-use and habitat associations of black-backed woodpeckers (Picoides arcticus) occupying recently disturbed forests in the Black Hills, South Dakota
For. Ecol. Manag.
A spatially explicit agent-based model of the interactions between jaguar populations and their habitats
Ecol. Model.
Spatial Model of Livestock Predation by Jaguar and Puma in Mexico: Conservation Planning
Base Hidrográfica Ottocodificada Multiescalas
Ecologia e conservação da onça pintada e da onça parda no Parque Nacional da Serra da Capivara, Piauí
Comparative ecology of jaguars in Brazil
CAT News
Living in extreme environments: modeling habitat suitability for jaguars, pumas, and their prey in a semiarid habitat
J. Mammal.
Has the earth's sixth mass extinction already arrived?
Nature
Discrete choice modeling of shovelnose sturgeon habitat selection in the lower Missouri River
J. Appl. Ichthyol.
LandScan: Digit Dataset
Ctmm: an R package for analyzing animal relocation data as a continuous-time stochastic process
Methods Ecol. Evol.
American black bear habitat selection in northern Lower Peninsula, Michigan, USA, using discrete-choice modeling
Ursus
Predation patterns of jaguars (Panthera Onca) in a seasonally flooded forest in the southern region of Pantanal, Brazil
J. Mammal.
Jaguars on the move: modeling movement to mitigate fragmentation from road expansion in the Mayan Forest
Anim. Conserv.
The application of discrete choice models to wildlife resource selection studies
Ecology
Jaguar spacing, activity and habitat use in a seasonally flooded environment in Brazil
J. Zool.
As onças-pintadas como detetives da paisagem no coredor do Alto Paraná, Brasil
Nat. e Conserv.
Selection of habitat by the jaguar, Panthera onca (Carnivora: Felidae), in the upper Paraná River, Brazil
Zoologia
Implications of fine-grained habitat fragmentation and road mortality for jaguar conservation in the Atlantic Forest, Brazil
PLoS One
Differential impact of landscape transformation on pumas (Puma concolor) and jaguars (Panthera onca) in the upper Paraná Atlantic Forest
Divers. Distrib.
Food habits and livestock depredation of sympatric jaguars and pumas in the Iguaçu food habits and livestock depredation of sympatric jaguars and pumas in the lguagu National Park Area, South Brazil
Biotropica
The jaguar's spots are darker than they apear: assessing the global conservation status of the jaguar Panthera onca
Oryx
When roads appear jaguars decline: increased access to an Amazonian wilderness area reduces potential for jaguar conservation
PLoS One
Effects of habitat fragmentation on biodiversity
Annu. Rev. Ecol. Evol. Syst.
Rigorous home range estimation with movement data: a new autocorrelated kernel density estimator
Ecology
Habitat use by sympatric jaguars and pumas across a gradient of human disturbance in Belize
Biotropica
Application of random effects to the study of resource selection by animals
J. Anim. Ecol.
Habitat fragmentation and its lasting impact on Earth's ecosystems
Sci. Adv.
Cited by (47)
Jaguar density in the Argentine Yungas: Overcoming camera trap failure
2024, Journal for Nature ConservationAnthropogenic factors do not affect male or female jaguar habitat use in an Amazonian Sustainable Reserve
2023, Perspectives in Ecology and ConservationEffects of human-induced habitat changes on site-use patterns in large Amazonian Forest mammals
2023, Biological ConservationJaguar (Panthera onca) population density and landscape connectivity in a deforestation hotspot: The Paraguayan Dry Chaco as a case study
2022, Perspectives in Ecology and ConservationCitation Excerpt :This allows for an estimation of how the landscape covariate affects individual space use as a deviance from the symmetric assumption of the traditional SCR Euclidian distance model (Morin et al., 2017; Royle et al., 2013; Sutherland et al., 2015). We considered this important as we expected that our highly structured landscapes and fragmented forest cover, along with the preferred use of forest by jaguars (Alvarenga et al., 2021; Morato et al., 2018a, 2018b; Thompson et al., 2021), would likely result in non-circular activity areas and spatial heterogeneity in detectability. Using the resistance parameter (δ), the cost to movement across a landscape (dlcp), and the estimate of σ, the potential connectivity of a landscape (the expected frequency of use of a point in the landscape as a function of the cost to movement) can be estimated (Morin et al., 2017; Sutherland et al., 2015).
Environmental and anthropogenic factors synergistically affect space use of jaguars
2021, Current BiologyCitation Excerpt :For speed, this indicates that decisions about movement at the home range scale were dependent upon the availability of preferred habitat and that jaguars adjust their behavior to account for anthropogenic factors.11,12,62–65 Animal movements are expected to be slower in areas of higher habitat quality and reduced habitat heterogeneity and fragmentation,66,67 which has been observed in big cats,68–71 and consequently, the inverse relationship between speed and forest cover is expected given jaguars’ preference for forest cover.11,12,64,65 Mean speed was similar between sexes, but male movements were more directional, which is logical given that male home ranges are larger, and to be traversed at the same speed as females, movements would have to be less tortuous.