Thesis (PhD)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: Widespread, human-induced ecosystem degradation and the associated biodiversity loss pose a direct threat to human wellbeing. While there is no substitute for healthy, self-sustaining ecosystems, ecological restoration offers an attractive, and indeed inevitable, supplement where conservation alone is not sufficient to support ecosystem integrity. Restoration is undergoing a revolution, where evidence-based (EB) practice is emerging as a new approach to increase the chances of successfully achieving restoration goals. EB practice is based on the notion that implementation decisions need to be based on the appraisal and use of evidence of effectiveness of alternative options. The point of departure of this thesis is the contention that EB practice need not be dependent only on research evidence. The work presented herein thus addresses the production and use of evidence of effectiveness in restoration practice. Using ten restoration programs in South Africa, the quality of evidence produced in practice was assessed. Three components of evidence production that were evaluated were (i) baseline condition measurement; (ii) goal setting and (iii) monitoring. Results showed poor definition of goals; a bias towards the use of socio-economic goals and indicators; more monitoring of inputs than impact; and inconsistent and short-term monitoring of biophysical indicators. Practitioners regarded the evidence base as adequate, but cited a few challenges associated with planning and resource availability as attributing factors to the gaps observed. I propose that practitioners’ perception of the current evidence base poses an additional threat to the generation of a strong evidence base. In addition to the production of evidence, access to said evidence is a vital component of EB practice. In an exploration of how evidence is made available by practitioners, it became evident in that a considerable amount of the information that was not easily accessible in documented form was known by the practitioners. This highlights the need for a shift in practice culture towards the valuing and rewarding of the dissemination of information. An assessment of EB restoration would have been incomplete without a deliberate consideration of social factors. I thus conducted a case study of an invasive alien plant clearing program, to determine what drives the use of scientific evidence in decision making. I observed that organizational structure, policies, priorities and capacity influence, and even limit, the use of scientific evidence to inform decisions. The challenges to making restoration evidence-based are diverse in nature, ranging from poor planning of restoration work, which points to limited appreciation of the need to produce a strong evidence base, to a lack of instruments and incentives to drive the generation, dissemination and use of evidence that spans both the biophysical and social aspects of restoration. These challenges are largely rooted in the conventional way of approaching restoration from individual disciplinary perspectives, thus artificially simplifying and compartmentalizing a naturally complex problem like degradation. I end by proposing transdisciplinarity, which focuses on a holistic world view and the production of knowledge that embraces complexity, as a possible vehicle to help move the practice of restoration towards being evidence-based. / AFRIKAANSE OPSOMMING: Wydverspreide, mens-veroorsaakte agteruitgang van ekostelsels en die gepaardgaande verlies aan biodiversiteit hou 'n direkte bedreiging vir menslike welsyn in. Alhoewel daar geen plaasvervanger is vir 'n gesonde, selfonderhoudende ekostelsel is nie, bied ekologiese restourasie 'n aantreklike en inderdaad onvermydelik, vul waar bewaring alleen nie voldoende ekosisteem integriteit kan ondersteun nie. Restourasie ondergaan tans ‘n revolusie, waar bewys gebaseerde (BG) praktyk 'n opkomende nuwe benadering om die kanse van die suksesvolle bereiking van herstel doelwitte te verhoog. BG praktyk is gebaseer op die idee dat die uitvoering van besluite gebaseer moet word op die evaluering en die gebruik van bewyse van die effektiwiteit van alternatiewe opsies. Die punt van vertrek van hierdie proefskrif is die bewering dat BG praktyk nie noodwendig afhanklik van navorsings bewys hoef te wees nie. Die werk wat hier aangebied word spreek tot die produksie en gebruik van bewyse van effektiwiteit in die restourasie praktyk. Deur die gebruik van tien restourasie programme in Suid-Afrika is die kwaliteit van die bewyse in die praktyk geassesseer. Die drie komponente van bewyse produksie wat geëvalueer is sluit in (i) basislyn toestand meting, (ii) doelwitstelling en (iii) monitering. Resultate toon 'n swak definisie van doelwitte; 'n vooroordeel ten gunste van die gebruik van sosio-ekonomiese doelwitte en aanwysers; meer monitering van insette as die impak; en teenstrydige en kort-termyn monitering van biofisiese aanwysers. Beofenaars het die gebruik van bewys gebaseerde inligting as voldoende beskou, maar 'n paar uitdagings wat verband hou met die beplanning en die beskikbaarheid van bronne is aangehaal as kenmerkende faktore in die gapings wat tans waargeneem word. Ek stel voor dat beoefenaars se persepsie van die huidige bewysbasis praktyk 'n bykomende bedreiging vir die generasie van 'n sterk bewybasis praktyk inhou. Benewens die produksie van bewyse, is die toegang tot bewyse 'n belangrike komponent van die BG praktyk. In die verkenning van hoe bewyse beskikbaar gestel word deur beoefenaars, is dit duidelik dat 'n aansienlike aantal inligting wel bekend is aan beofenaars maar nie maklik toeganklik in gedokumenteerde vorm is nie. Dit beklemtoon die behoefte vir 'n verskuiwing in die praktyk kultuur tot die waardering en beloning van die verspreiding van inligting. 'n Beoordeling van die BG herstel sou onvolledig wees sonder 'n doelbewuste oorweging van sosiale faktore. Ek het dus 'n gevallestudie van 'n indringerplant verwyderings program uitgevoer om vas te stel wat die gebruik van wetenskaplike bewyse in besluitneming aandryf. Ek het opgemerk dat die organisatoriese struktuur, beleid, prioriteite en kapasiteit die gebruik van wetenskaplike bewyse kan beinvloed, en selfs beperk. Die uitdagings om herstelwerk bewys-gebaseerd te maak is uiteenlopend van aard, dit wissel van swak beplanning van herstel werk, wat dui op beperkte waardering van die behoefte om 'n sterk bewyse basis te produseer, 'n gebrek aan instrumente en aansporings vir die generasie van besyse, verspreiding en gebruik van bewyse wat strek oor beide die biofisiese en maatskaplike aspekte van die restaurasie. Hierdie uitdagings is grootliks gegrond op die konvensionele manier van restaurasie wat gebaseer is op individuele dissiplinêre perspektiewe,wat lei tot die kunsmatige simplifiseering van ‘n uiteraand komplekse problem soos agteruitgang. Ek eindig af deur die gebruik van transdissiplinariteit, wat fokus op 'n holistiese wêreldbeskouing en die produksie en kennis van kompleksiteit insluit voor te stel, as 'n moontlike voertuig om die skuif in praktyk van restourasie na n bewys-basis te vergemaklik.

<p> Throughout the United States, and globally, there has been recent interest in large-scale monitoring of bats, driven largely by the many threats that bats currently face such as climate change, white nose syndrome, habitat loss, and wind energy development. Additionally, many human activities generate sensory disturbances including anthropogenic light and noise pollution that have been shown to affect habitat use and foraging efficiency in bats and other wildlife. My research took place in the Piceance Basin of northwestern Colorado, where there has been considerable development of natural gas resources in recent years. During the drilling phase for natural gas, drill rigs run continuously for weeks to months at each well development site. In addition to the physical disturbance and increased human presence at the well pad, drill rigs are brightly lit, and also emit high amplitude anthropogenic noise. The light and noise from active drill rigs can travel many miles from the source, far beyond areas where wildlife habitat has been physically disturbed. The goal of my research was to isolate noise from the other associated forms of disturbance, and investigate what, if any, effects drilling noise is having on the bat community. </p><p> My dissertation is comprised of three stand-alone chapters, as follows. In Chapter 1, I test two different strategies for deploying bat detectors to determine which strategy yields higher detection and species identification rates. In Chapters 2 and 3, I present the results of noise playback experiments. Chapter 2 compares bat activity levels at control sites, and at treatment sites where noise was added experimentally. In Chapter 3, I monitor bat activity patterns both during and after a noise playback experiment to determine how rapidly activity levels recover post-exposure. A brief summary of each chapter follows. </p><p> In Chapter 1, I focus on the methods for recording bat echolocation calls, and identifying free-flying bats in the field. Unlike capture techniques, current acoustical methods for bat monitoring do not provide information about unique individuals, age, sex, or reproductive status. What acoustical monitoring can provide is information about bat activity levels, habitat use, and species identification in some cases, without interfering with bat movement, foraging, or other activities. The commercially available technology for recording bat echolocation calls has rapidly advanced, and there are many ultrasound detector&ndash;recorder systems (hereafter, &lsquo;bat detectors&rsquo;) available with a wide range of recoding options. Due to rapid attenuation of ultrasound signals, one of the challenges to acoustical monitoring is the relatively limited recording range of bat detectors. To increase this range, I took advantage of a bat detector that had the ability to record on two channels (in stereo). By attaching microphone extension cables, I was able to increase the distance between the left and right channel microphones, thereby increasing the acoustical sampling space. When this data collection effort took place, the SM2BAT+ detector from Wildlife Acoustics, Inc. was the only commercially available bat detector that had the two-channel recording capability. I deployed two identical bat detectors at each study site, and compared the recordings made using the stereo option to recordings made from a single channel. </p><p> In general, the stereo setup outperformed the single-channel systems. With the stereo microphones separated by approximately 10 m, the bat detectors that recorded in stereo produced 2.7 times more recordings overall. The increased number of recordings resulted in a higher number of calls that could be identified to species. The benefit of the stereo setup was not equal for all species. With the stereo microphones only about 10 m apart, there was some overlap between the calls that were identified on the left and right channels. The highest rate of overlap (19.5%) was in big brown bats (<i>Eptesicus fuscus </i>). Rates of overlap for species in the Myotis genus were all less than 5%, and none of the recordings of pallid bats (<i>Antrozous pallidus </i>) were identified on both channels for the same bat-pass-event. The stereo option is a promising way to increase the number of bat recordings, which may be a particularly useful when surveying for rare species. </p><p> In Chapter 2, I used a noise playback experiment to isolate noise from other forms of anthropogenic disturbance, and monitored the bat activity level response. I recorded the sounds of an active drill rig, and played these recordings at treatment sites. I measured sound pressure levels at the drill site, and estimated the sound pressure levels at the noise playback sites. Using outdoor speakers, I was unable to project the drill rig noise at the same amplitude of an actual rig, but I was able to significantly elevate the sound levels at treatment sites. The noise levels at treatment sites roughly corresponded to noise levels that can be experienced approximately 100 m from a drill rig. This distance from a drill rig is typically beyond the well pad, in habitat that is not physically disturbed. There is widespread recognition that noise, light, and other sensory disturbances can affect the behavior and physiology of wildlife. The goal of the experiment was to determine if noise alone impacted the activity levels of bats, after being separated from the other forms of disturbance at a drilling site. I projected noise at treatment sites that were not already developed, and paired these treatment sites with control sites with no added noise. I conducted this experiment in 2013 and 2014, and present the results from 20 sites each year (10 control-treatment pairs annually). </p><p> Both years, there was an overall decrease in bat activity at treatment sites, when compared to control sites. In 2013, 8 of the 10 treatment sites had lower estimated bat activity levels. In 2014, all 10 treatment sites had lower estimated activity levels, although for some control-treatment pairs there was overlap in the credible intervals. Multiple species showed signs of reduced activity at treatment sites. For both years, <i>M. ciliolabrum </i> and <i>L. cinarius</i> had reduced activity levels at treatment sites. The response of other species was more idiosyncratic, with reduced activity in one field season, and inconclusive or no response during the other year. The species that did respond to the noise treatment have very different life histories, making it difficult to generalize about how any given species may respond to noise. </p><p> Chapter 3 focuses on bat activity level trends over time during a two period cross-over experiment. The classic two-period crossover experiment consists of two treatments (i.e., treatments &lsquo;A&rsquo; and &lsquo;B&rsquo;), where each site is exposed to both treatments, and the order of the treatments is randomly assigned. For this study, &lsquo;A&rsquo; refers to no added noise, and &lsquo;B&rsquo; refers to a noise treatment consisting of the projected recording of drilling noise. A total of 12 sites were randomly assigned to the A:B sequence, and 13 sites were assigned to the B:A sequence. I acoustically monitored bat activity throughout the experiment, with particular interest in understanding the activity level dynamics post-exposure to the noise treatment. Most studies that investigate the impacts of noise on wildlife have focused on the response to noise during a noise treatment period, or noise event. Only a handful of previous studies have addressed the post-exposure period after noise ends. In these studies, noise treatments or events were relatively short in duration (a few minutes), and the corresponding recovery period was monitored over a similarly brief timeframe. My study differs in both the duration of the noise treatment (continuous noise over six days/nights), and in the duration of the post-exposure monitoring period (also six days/nights). I focused on the response of four bat species, <i>Myotis ciliolabrum, Myotis evotis, Lasiurus cinareus,</i> and <i>Taderida brasiliensis.</i> </p><p> Of the four focal species, <i>M. evotis</i> showed no substantial response to the noise treatment. The responses of <i>M. ciliolabrum</i> and <i>T. brasiliensis</i> were somewhat challenging to interpret. The overall activity levels of these species were lower during the treatment period in the A:B sequence, but they also had declining activity levels throughout the pre-exposure period. The activity patterns of <i>L. cinareus</i> provided the most convincing evidence of noise avoidance. Furthermore, <i> L. cinareus</i> activity levels did not show signs of recovery after the noise treatment ended in the B:A sequence. This suggests that after the six-day post-exposure period, there were still lingering effects of noise on <i>L. cinareus</i> activity levels. This study provides evidence that the effects of noise can linger for multiple days post-exposure. Future wildlife studies that assess periods of post-exposure could contribute meaningfully to this area of research, and aid in the development of conservation and mitigation efforts.</p>

<p> Survival of the endangered Light-footed Ridgway&rsquo;s Rail (Rallus obsoletus levipes) depends on marsh conservation and restoration efforts that supply suitable nesting habitat. Previous research indicates that rails preferentially nest in low marsh areas with tall, dense Spartina foliosa (Pacific cordgrass). While canopy architecture (e.g. height and density) is important, additional microhabitat and landscape-level metrics may be involved in nest-site selection. This project characterized microhabitat parameters (e.g. vegetation and food availability) and landscape-level parameters (e.g. spatial configuration) for 40 nests and 40 non-nest sites in two California wetlands, Upper Newport Bay and the Tijuana Slough, to identify habitat features that predict reproductive success. Assessment of the microhabitat and landscape-level parameters suggests that vegetation structure (e.g. height, stem density) is important, but that invertebrate density and composition, tidal connectivity, and elevation are also important to Light-footed Ridgway&rsquo;s Rail nesting.</p>

<p> Individual variation and fitness are the cornerstones of evolution by natural selection. The trophic niche represents an important source of phenotypic variation on which natural selection can act. Although individual variation is fundamental to species-level ecological and evolutionary change, individual variation is often ignored in population-level approaches to wildlife ecology, conservation and management. Failing to link individual resource use to fitness or to biological outcomes related to fitness limits us to managing for the average resource needs of a population, which may be insufficient for protecting the diversity of resource use within populations and the underlying eco-evolutionary processes that generate that diversity. My goals were to provide insights into the mechanisms that generate and constrain intrapopulation trophic niche variation, evaluate whether linkages exist between individual biological outcomes and variation in food habits across the range of resources consumed within generalist consumer populations and examine how that variation manifests in population-level responses. </p><p> I investigated the causes and physiological consequences of intrapopulation trophic niche variation in two generalist consumers, the American black bear (<i>Ursus americanus</i>) and brown bear (<i>U. arctos</i>) across three sites in British Columbia, CAN and at one site in Alaska, USA. My primary tools included stable isotope analysis to estimate diet, enzyme-linked immunoassay of hair to quantify the hormone cortisol for indexing physiological stress, and genetic analyses to identify individuals, species, and sex and to estimate ancestry. I found that individual differences in resource use can result in similar biological outcomes and that similar resource use can result in different biological outcomes. Intra- and interspecific competition, sex-based differences in nutritional and social constraints and annual variation in food availability all influenced trophic niche variation and the resultant biological outcomes. I also found evidence of a link between intrapopulation trophic niche variation and population genetic structure. My results highlight the diverse ecological drivers and diverse consequences of trophic niche variation, which further illuminates why the trophic niche is a nexus for eco-evolutionary dynamics.</p>

<p> Habitat loss and the resulting fragmentation can alter animal movements and disrupt animal seed dispersal mutualisms; however, these effects on spatial patterns of seed dispersal are not well understood. To explore the effects of habitat loss and fragmentation on seed dispersal distances and seed dispersion (aggregation), I created a spatially-explicit, individual-based model of an animal dispersing seeds (SEADS &ndash; Spatially-Explicit Animal Dispersal of Seeds). For Chapter 1, I ran SEADS in a theoretical landscape of 0-90% habitat loss to determine the effects of habitat loss on three animal traits relevant to seed dispersal: movement distance, gut retention time, and time between movements. SEADS results revealed a complex interaction involving all animal traits and habitat loss on dispersal distances and dispersion, driven by a novel underlying mechanism of fragment entrapment. Unexpectedly, intermediate habitat loss could increase dispersal distances and dispersion relative to low and high habitat loss for some combinations of animal traits. In Chapter 2, I developed a new method (TriAD) to estimate animal locations from triangulation data to improve triangulations for toucans. Location estimates calculated with TriAD averaged 23-63 m closer to true locations for three of four distance classes compared to the traditional method (Lenth&rsquo;s maximum likelihood estimator, LMLE). Additionally, over 99% of 570 triangulations converged for TriAD compared to 85.8% for the LMLE. For toucan data, 98.0% of 4445 triangulations converged in TriAD compared to 88.6% for the LMLE. In Chapter 3, I parameterized SEADS with data from toucan movements and gut retention trials collected in Costa Rica to evaluate the effects of landscape, body size, and social organization on seed dispersal. The larger, less social species dispersed seeds 17-28% farther and in less aggregated patterns than the smaller, social species. This pattern was driven by longer movements in the larger species, suggesting that body size is more important than social organization for seed dispersal in our system. Habitat use and home range sizes limited dispersal distance and dispersion for both species. Findings from our simulations illustrate the importance of separating and quantifying the effects of disperser behaviors and landscape effects that shape seed deposition in fragmented areas.</p><p>

<p> Wildlife management and conservation frequently rely on understanding mechanisms that influence distribution and abundance of animals. I quantified space use for a population of inland riverine adult male alligators in Mississippi. Results indicated habitat selection is a scale-dependent process and aquatic vegetation, water depth, and water temperature may be important factors influencing alligator foraging and thermoregulation. Apparent habitat suitability and low alligator density did not manifest in an observed body size-based dominance hierarchy. I also analyzed long-term Mississippi alligator spotlight survey data for trends and effects of environmental covariates on counts. Model results indicated alligator counts have increased over time. This response likely reflects benefits accrued from decades of protection and wetland conservation. Distance sampling does not appear to be a feasible monitoring technique for riverine alligator populations. Nevertheless, it is important that survey protocols and monitoring programs account for imperfect detection and model important covariates.</p>

<p> The bog turtle (<i>Glyptemys muhlenbergii</i>) is listed as threatened under the U.S. Endangered Species Act primarily because of the loss and degradation of its specialized wetland habitat. Adequate nesting habitat may be a limiting factor for bog turtle recovery, and nesting habitat creation or expansion may be an effective conservation tool but only if managers understand nesting behaviors, habitat requirements, and threats to nest success. The objective of this study was to understand the nesting ecology of the bog turtle and to use this information to suggest improvements to bog turtle management. I used radio telemetry to locate nesting females. I then used nest location and maternal genetic microsatellite data to investigate nest-site fidelity and natal homing. I collected microhabitat data at each nest and random points in both spring and fall at three spatial scales to understand nest-site selection. I also studied nest success with and without predator excluder cages to determine if predator excluders reduced predation or altered nest conditions. I found evidence that females display nest-site fidelity to nest-site areas but no evidence for natal homing. Bog turtles nested in relatively small patches of habitat in close proximity to water or saturated soil with low densities of woody stems and low percent cover of woody vegetation, forbs, and ferns. Predator excluders reduced predation by &sim;40% while having no effect on nest temperatures, incubation periods, or underlying (disregarding predation) nest success. Given the results of this research, I conclude that the creation of new nesting areas could be a viable conservation tool, as females do not always return to the same nest-site area. Managers should create or maintain small patches of open-canopy habitat with reduced woody vegetation in areas that have consistent moisture from spring to fall. Increasing the availability of nesting areas may eventually lead to increases in recruitment, but if predation pressures are high, managers should also deploy predator excluders. This is the first study to examine the behavioral drivers of bog turtle nest-site selection and the efficacy of predator excluders on bog turtle nests.</p>

<p>A fundamental condition of the struggle for existence is resource limitation. Optimal foraging theory describes how individuals refine behavior to most efficiently exploit available resources. For colonial breeding animals, such as seabirds, competition for limited resources is amplified by a high density of competitors at feeding grounds near the colony, which can lead to a diversification of foraging strategies that best exploit available resources. Laysan and black-footed albatrosses (Phoebatria immutabilis and P. nigripes, respectively) breed sympatrically and synchronously in dense colonies on small atolls throughout the northwestern Hawaiian Islands. While the distributions and habitat preferences of these congeners are well described, the variability in foraging strategies driven by niche partitioning is not. In chapter 1, I used fine-scale behavioral data from GPS data-loggers to tease apart subtle behavioral nice partitioning between and within Laysan and black-footed albatrosses constrained to poor-nutrient tropical waters during the most energetically-demanding reproductive stage in birds: the brood-guard. Albatrosses showed discrete behavioral partitioning both between and within species, which was driven by differences in nocturnal and diurnal foraging and by sex-specific strategies. Black-footed albatrosses exhibited greater variability in foraging patterns suggesting they may experience strong intraspecific competition at Tern Island. In chapter 2, I employed a novel adaptation of a biochemical diet analysis, quantitative fatty acid analyses, to investigate dietary niche partitioning across the breeding season. This analysis characterized diet of individual adult albatrosses across the breeding season, from incubation to the chick-brood, for the first time in these species. Additionally, this research extended beyond diet characterization to quantify dietary dependence on fisheries-associated resources for these two species. One of the primary threats to albatross is the incidental mortality of birds from fisheries bycatch, therefore, understanding the incidence of fisheries-associated food in the diet of a breeding colony of albatrosses could have important conservation implications. While older birds tended to consume more squid than younger birds, we did not identify an age bias in birds that exploited fisheries resources. Parallel with results from the behavioral analysis in chapter 1, I found that black-footed albatrosses had greater dietary flexibility than Laysan albatrosses at a population level, and a greater degree of dietary specialization at the individual level. In the final chapter, I looked at how the obligatory act of flight feather molt impacted activity levels and space use across the post-breed migrations, when albatrosses are freed from central place foraging constraints, but limited by demands of self-maintenance. I found that activity levels, but not habitat use, were impacted by molt extent. Habitat use appeared to be primarily driven by consistent individual preferences, with birds showing a high degree of site fidelity to their post-breeding molting grounds. A clear decrease in home range size with age indicated that birds spend less time searching for profitable feeding grounds when they are older, likely a product of accumulated memory and experience in these long-lived species.

<p> Reliable estimates of species abundance, distribution, and population trajectories are critical in conservation and management. However, for many tropical species that information is missing. We conducted Sharp-shinned Hawk (SSHA) searches during the breeding seasons of 2013, 2014, and 2016 in eight montane forest reserves and their adjacent private lands of Puerto Rico. Further, we developed a maximum entropy model (MaxEnt) of the SSHA&rsquo;s potential distribution using the following environmental variables: aspect, canopy closure, elevation, rainfall, slope, and terrain roughness. Elevation accounted for 89.8% of model fit, predicting the greatest probability of occurrence (> 60%) at elevations above 900 m. The model estimated 0.6%Puerto Rico (56.1 km²) has the greatest probability of occurrence. We developed a periodic population matrix model to describe the influence of early life stages on population growth of the Red-tailed Hawk (RTHA) in eastern Puerto Rico. Our results suggest that the RTHA population has an average annual population increase of 5%, with rates differing between highlands (&lambda;_h = 1.05) and 27% lowlands (&lambda;_l = 1.27). Adult survival was the most important population growth parameter, with more effect in highlands (elasticity = 0.86) than in lowlands (elasticity = 0.53). Sensitivity of &lambda; to changes in nestling survival was greater than for other life-stages (sensitivity lowlands = 0.46, sensitivity highlands = 0.48). I developed an open population N-mixture model to estimate abundance, availability, and detection probability of RTHA in the Luquillo Mountains and western Cordillera Central. The abundance estimates were 0.05 RTHA/ha, with an availability of (&phis;) = ~1 RTHA/per survey point and a detection probability (r) = ~0.25. In Luquillo Mountains, abundance was positively influenced by slope and elevation. In the western Cordillera Central, abundance of RTHA was constant across elevation, slope, and canopy closure. Detection probability decreased with increasing slope and wind conditions and showed a gradual small negative decrease with slope with reduced winds conditions. My findings are a contribution to our knowledge of how population traits and ecological constraints imposed by insular environments are influencing distribution, abundance, and population dynamics of raptors to propose management or conservation schemes.</p><p>

<p> Mammalian carnivores are a vital component of many ecosystems and can be particularly sensitive to human disturbance, even within protected areas (PAs). Our objective was to understand how human disturbance affects carnivore communities in southern Arizona, U.S.A., by studying habitat occupancy based on data collected using non-invasive methods in three PAs with different levels of human disturbance. We examined the impacts of human disturbance variables and disturbance level on carnivore occupancy, co-occurrence, temporal activity, and habitat associations. Carnivore occupancy varied based on human disturbance variables (i.e., roads, trails, etc.). Edges of PAs appeared to negatively impact occupancy of nearly all carnivore species. We also found that the presence of roads and trails, and not necessarily how much they are used, had a significant negative impact on the occupancy of most carnivore species. Furthermore, the overall level of disturbance within a PA influenced how sensitive carnivores were to human disturbance variables. Carnivores were more sensitive in PAs with higher levels of disturbance and were relatively unaffected by disturbance variables in a PA with low base levels of disturbance. In areas with low levels of disturbance, we found that many carnivore species have lower than expected levels of co-occurrence, which suggests spatial partitioning. As disturbance within an area increased, spatial partitioning became less prominent, and carnivores exhibited higher levels of temporal partitioning within these areas. We found that habitat associations varied among carnivore species, and associations were often different across different scales. We also found evidence of different habitat preferences in protected areas with higher levels of disturbance (e.g., avoidance of water sources). Information on the impacts of human disturbance is important when developing conservation plans, which is especially true for protected areas, given their important role in carnivore conservation, particularly as they are experiencing ever increasing rates of visitation.</p><p>