Uq201d603 OA.Pdf

1 The mesoscavenger release hypothesis and implications for ecosystem

2 and human well-being

4 Christopher J. O’Bryan1,4,*, Matthew H. Holden2,3,4, and James E.M. Watson1,4,5 5 6 1School of Earth and Environmental Sciences, The University of Queensland, Brisbane QLD 4072, Australia 7 2ARC Centre of Excellence for Environmental Decisions, The University of Queensland, Brisbane, QLD 4072, 8 Australia 9 3Centre for Applications in Natural Resource Mathematics, School of Mathematics and Physics, The University 10 of Queensland, Brisbane, QLD 4072, Australia 11 4Centre for Biodiversity and Conservation Science, The University of Queensland, Brisbane, QLD 4072, 12 Australia 13 5Global Conservation Program, Wildlife Conservation Society, 2300 Southern Boulevard, Bronx, New York, 14 USA 15 *Corresponding author 16 Christopher J. O’Bryan e-mail: [email protected] 17 Matthew H. Holden e-mail: [email protected] 18 James E.M. Watson e-mail: [email protected] 19 20 Keywords: scavenger, vulture, predator, cascade, dynamic model, conservation, human- 21 wildlife conflict, food webs, trophic level, top-down release 22 Article type: Ideas & Perspectives 23 Number of words in the abstract: 135 24 Number of words in the main text: ~3,800 25 Number of references: 58 26 Number of figures and tables: 4 Figures, 1 Table 27 28 Corresponding author full details: Address: Christopher J. O’Bryan, 8 Cottenham Street, 29 Fairfield QLD, Australia 4103. Phone: +61 449 599 035. E-mail: [email protected] 30 31 Data accessibility: no new data Author Manuscript 32

This is the author manuscript accepted for publication and has undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/ELE.13288

This article is protected by copyright. All rights reserved 33 Statement of authorship: CO and JW conceived the idea of the manuscript. CO, MH, and 34 JW designed the research. CO conducted the literature review and wrote the manuscript. MH 35 developed the model. CO and MH modified the model, wrote model results, and developed 36 figures. 37 38 39 40 Abstract 41 Many apex scavenger species, including nearly all obligate scavengers, are in a state of rapid 42 decline and there is growing evidence these declines can drastically alter ecological food 43 webs. Our understanding of how apex scavengers regulate populations of mesoscavengers, 44 those less-efficient scavengers occupying mid-trophic levels, is improving; yet, there has 45 been no comprehensive evaluation of the evidence around the competitive release of these 46 species by the loss of apex scavengers. Here we present current evidence that supports the 47 mesoscavenger release hypothesis, the increase in mesoscavengers and increase in carrion in 48 the face of declining apex scavengers. We provide two models of scavenger dynamics to 49 demonstrate that the mesoscavenger release hypothesis is consistent with ecological theory. 50 We further examine the ecological and human well-being implications of apex scavenger 51 decline, including carrion removal and disease regulation services. 52 53 Introduction 54 Apex scavengers are functionally dominant at scavenging, meaning they can find and 55 consume carcasses more efficiently than other scavengers (Sebastián-González et al. 2016), 56 and can be either obligate or facultative scavengers. Obligate scavengers (i.e. Old and New 57 World vultures) are dependent entirely on carrion, but facultative scavengers rely partly on 58 carrion (Ogada et al. 2012a). Apex scavengers (i.e. vultures and functionally dominant 59 facultative scavengers) are facing unprecedented declines due to direct persecution, human 60 disturbance, collision with infrastructures and electrocution, poisons and other dietary toxins, 61 human disturbance, habitat loss and degradation, food shortage caused by sanitary 62 regulations, or abandonment of traditional farming practices (Ripple et al. 2014; Buechley & Author Manuscript 63 Şekercioğlu 2016). As such, it is imperative to understand the ecological and human well- 64 being impacts of apex scavenger declines (Buechley & Şekercioğlu 2016). There is growing 65 evidence of scavenger competitive release across Earth, where mid-sized, less efficient 66 scavengers (i.e., mesoscavengers) can increase in abundance in the absence of competition

This article is protected by copyright. All rights reserved 67 from more efficient apex scavengers (the mesoscavenger release hypothesis; Figure 1) 68 (Butler & du Toit 2002; Sekercioğlu et al. 2004; Markandya et al. 2008; Ogada et al. 2012b; 69 Buechley & Şekercioğlu 2016; Morales-Reyes et al. 2017). The population effects of 70 mesoscavenger release from apex scavengers can be qualitatively similar to well documented 71 patterns observed in predatory systems, where the absence of apex predators releases 72 mesopredators (Crooks & Soulé 1999; Ritchie & Johnson 2009; Ripple et al. 2014; Newsome 73 et al. 2017). However, the mechanism behind mesoscavenger release is different; it is caused 74 by reduced competition over a shared resource, not the loss of top-down control. Here, we 75 present evidence pointing to the release of mesoscavengers by the loss of apex scavengers, 76 and we discuss potential ecosystem and human well-being implications of mesoscavenger 77 release. 78 79 Empirical support for the mesoscavenger release hypothesis 80 Mesoscavengers have been shown to be more abundant and diverse in areas that are absent of 81 apex scavengers (Table 1), which can affect ecosystem structure. For example, red foxes 82 (Vulpes vulpes), which are facultative mesoscavengers, were significantly more abundant in 83 areas of south-eastern Spain that lack vultures (Gyps spp.) compared to areas with vultures 84 (Morales-Reyes et al. 2017). Morales-Reyes and colleagues contend that mesoscavengers had 85 increased scavenging opportunities and thus consumed more carrion in the absence of 86 vultures, therefore resulting in increased abundance of foxes. An increase in mesoscavengers 87 has also been observed in India, where growing feral dog (Canis lupus) and rodent 88 populations have been linked to the widespread decline of vultures caused by ingestion of 89 veterinary pharmaceuticals (i.e. diclofenac) (Markandya et al. 2008). The authors suggest that 90 the concomitant rise of carrion without vultures resulted in a spike in mesoscavenger 91 populations (Markandya et al. 2008). In Tasmania, Australia, areas where Tasmanian devils 92 (Sarcophilus harrisii) have declined due to facial tumour disease have resulted in an 93 increased abundance of feral cats (Felis catus) and forest ravens (Corvus tasmanicus) 94 (Cunningham et al. 2018). The authors show that forest ravens have increased across all of 95 Tasmania during the period of Tasmanian devil decline. In areas of Spain and South Africa 96 where apex scavengers were lacking, species richness and composition drove the Author Manuscript 97 consumption of carrion; however, context-dependent effects (i.e. species abundance) had a 98 greater effect where apex scavengers were common (Mateo-Tomás et al. 2017). These apex 99 scavengers not only included globally widespread species such as wild boar (Sus scrofa), but 100 also imperilled species like gyps vultures (Gyps spp.) and African lions (Panthera leo)

This article is protected by copyright. All rights reserved 101 (Mateo-Tomás et al. 2017). Likewise, in the Mendocino National Forest of California, foxes, 102 corvids, and rodents had significantly higher species richness at deer carcasses in the absence 103 of black bears and puma – the apex facultative scavenger and apex predator, respectively 104 (Allen et al. 2014). The authors argue that the nestedness, or structure of the scavenger 105 community, increased at carcasses where large carnivores were present (Allen et al. 2014). 106 107 In the absence of apex scavengers, mesoscavengers are thought to be less effective at locating 108 carrion, resulting in a longer carcass decomposition time. For instance, when turkey vultures 109 (Cathartes aura) and black vultures (Coragyps atratus) were experimentally excluded from 110 carrion in South Carolina, USA, 80 percent of carcasses were not scavenged by 111 mesoscavengers, resulting in a ten-fold increase in carrion that were not fully scavenged 112 compared to controls (Hill et al. 2018). A similar pattern was observed in Australia where 113 nearly 70 percent of carrion were not scavenged by mesoscavengers (rats, dogs, foxes, and 114 corvids) in the absence of apex facultative scavenging species such as kites (Haliastur spp.) 115 and white-bellied sea eagles (Haliaeetus leucogaster) (Huijbers et al. 2015). Similarly, 116 carrion persisted 2.6 times longer in areas where Tasmanian devils declined due to a facial 117 tumor disease (Cunningham et al. 2018). Carcasses were scavenged three times slower in the 118 absence of vultures in the Laikipia District of central Kenya (Ogada et al. 2012b) and thirteen 119 times slower in areas without vultures in south-eastern Spain (Morales-Reyes et al. 2017). 120 Not only are vultures more efficient at locating and consuming carrion in the Masai Mara 121 National Reserve in Kenya, they have also been shown to aid mesoscavengers in locating 122 carrion (Kane & Kendall 2017). As such, the loss of apex scavengers can result in increased 123 available carrion biomass and slower decomposition time likely due to a lower scavenging 124 efficiency by mesoscavengers. 125 126 Theoretical support for the mesoscavenger release hypothesis 127 Unlike the mesoscavenger release hypothesis, the mesopredator release hypothesis has been 128 shown to be consistent with the outputs of classic predator-prey models (Crooks & Soulé 129 1999). To determine if the mesoscavenger release hypothesis is also consistent with 130 ecological theory, we built two simple dynamic models, one of apex obligate scavengers and Author Manuscript 131 mesoscavengers competing over carcasses for food, and the second model where the apex 132 scavenger is facultative. 133 134 Apex obligate scavenger model

138 = + , 1 + �� ―�� 139 �� ℎ��

140 = 1 + , ( ) 1 +� �� ― 141 �� ℎ��

142 = , 1 + 1 + �� ― �� ― ― 143 �� ℎ�� ℎ�� 144 where A, M, and C are apex scavenger, mesoscavenger, and carrion biomasses, respectively. 145 For the sake of the model above, we consider all apex scavengers to be obligate scavengers 146 that die in the absence of carrion at rate μ. Mesoscavengers are assumed to have alternate 147 food sources, and therefore in the absence of carrion grow logistically at rate r, with carrying 148 capacity k. Apex scavengers and mesoscavengers convert food into increased reproduction at

149 rates ga and gm, respectively. The efficiency at which they find carrion is ea and em, and

150 scavenger handling time is ha and hm. Note that ea will always be greater than em under 151 mesoscavenger release due to apex scavengers being functionally dominant at scavenging 152 relative to mesoscavengers. Carrion increase due to animal death at a constant rate p and 153 decay at rate δ. The model assumes that carrion are from non-modelled species and are 154 therefore mathematically similar in structure to chemostat models of resource dynamics 155 where resources enter and exit a system at constant rates (Smith & Waltman 1995). While 156 mesoscavenger and apex scavenger carcasses can contribute to scavenged carrion, this 157 contribution is generally inconsequential compared to the carcasses of other species, and are 158 often avoided due to the coevolutionary relation between carnivores and their parasites 159 (Moleón et al. 2017). 160 161 We parameterize our model using known mortality, efficiency, decay, and carrion availability Author Manuscript 162 rates and scavenger handling times (Table S1). Known parameters on apex obligate 163 scavengers are obtained from studies on cape vultures (Gyps coprotheres) (Komen 1992), 164 griffon vultures (Gyps fulvus) (Houston 1974), and multi-scavenger systems with vultures 165 present (Morales-Reyes et al. 2017). Known parameters on mesoscavengers are obtained

This article is protected by copyright. All rights reserved 166 from a study on scavenger systems with vultures absent (Morales-Reyes et al. 2017), with 167 information on carrion availability and decay rates from field experiments in Africa and the 168 USA, respectively (Houston 1985; Carter et al. 2006). Population growth rate of the 169 mesoscavenger in the absence of carrion, and the food conversion rates for both scavengers 170 are arbitrarily set equal to the mortality rate of apex scavengers in the absence of food. We 171 vary the carrying capacity of the mesoscavengers in the absence of carcasses from 0.5 – 2.0, 172 to capture differing dependence and/or preference for non-carrion food sources. We also 173 consider the case where scavenging efficiency of the mesoscavenger is doubled in the 174 presence of an apex obligate scavenger due to evidence of following behaviour by 175 mesoscavengers on vultures (Kane & Kendall 2017). We note that this model could be non- 176 dimentionalised to reduce the number of parameters, but we present the model here in its 177 most biologically interpretable form, for clarity. 178 179 We define mesoscavenger release as the increase in functionally less dominant 180 mesoscavengers in the absence of more functionally dominant apex scavengers, where 181 functional dominance is determined by a scavenger’s relative ability to efficiently locate and 182 consume carrion (Mateo-Tomás et al. 2017). Under the assumed conditions (i.e. that apex 183 scavengers are stronger competitors than mesoscavengers) a reduction of apex scavengers 184 leads to an increase in mesoscavengers and carrion in our model, which is consistent with the 185 mesoscavenger release hypothesis. These results therefore suggest that a decrease in apex 186 obligate scavengers may lead to increases in mesoscavengers and carrion, which is also 187 consistent with observational data (e.g. Morales-Reyes et al. 2017). When apex obligate 188 scavengers are removed from the system, mesoscavenger equilibrium densities increase along 189 with carrion density. For example, in the baseline parameterisation, with an apex obligate 190 scavenger 18 times more efficient (e.g. Gyps fulvus) than the typical mesoscavenger 191 assemblage, removing the apex scavenger causes a 13-fold increase in equilibrium carrion 192 density (Figure 3). The effect is strongest when apex scavengers are substantially more 193 efficient than the mesoscavengers, and when mesoscavenger carrying capacity is low 194 (although mesoscavenger release occurs across a range of carrying capacities; Figure 3 & 195 Figure S1). For example, removing an apex scavenger that is only three times as efficient as Author Manuscript 196 the typical mesoscavenger increases carrion density two-fold. Furthermore, when 197 mesoscavengers have a high carrying capacity in the absence of carrion, apex obligate 198 scavengers still reduce carrion density by over half, due to their relatively high search 199 efficiency (Figure 3).

This article is protected by copyright. All rights reserved 200 201 When doubling mesoscavenger search efficiency in the presence of apex scavengers, the 202 mesoscavenger release response is qualitatively similar to the baseline case (Figure 3). The 203 only differences are that mesoscavenger density slightly increases (4.7% increase) and 204 carrion density drops by 5.3% when apex scavengers improve mesoscavenger search 205 efficiency (measured at the baseline apex scavenger search efficiency). A full local sensitivity 206 analysis of the model parameterisation is presented in the Supplementary Material (Figure 207 S1). 208 209 Our models do not consider interactive effects of other carnivore species at modulating the 210 ability of vultures to exploit carrion. For example, apex predators may have differing effects 211 on scavenging behavior and consumption patterns of both apex scavengers and/or 212 mesoscavengers at and around carcasses (Selva & Fortuna 2007; Olson et al. 2012; Barton et 213 al. 2013; Allen et al. 2014; Moleón et al. 2014; Cunningham et al. 2018; Sivy et al. 2018). 214 More research is needed to investigate the role of other carnivore species at regulating 215 scavenger behavior and population dynamics. 216 217 Apex facultative scavenger model 218 As apex scavengers can also be facultative, consider a similar dynamic model, but with apex 219 scavengers capable of alternate feeding strategies. Apex scavengers now grow logistically at

220 rate ra, with carrying capacity ka in the absence of carrion, 221 222 (2)

223 = 1 + . ( ) 1 +� �� ― 224 �� ℎ�� 225 We parameterize our model using known efficiency and handling times from a study on 226 Tasmanian devils (Cunningham et al. 2018). Other parameters, describing carrion decay and 227 carrion availability, are set to the baselines in the previous model (Table S1). We explore the 228 parameter space for this system by varying handling times, search efficiencies, growth rates, Author Manuscript 229 and carrying capacities against equilibrium population densities for the Tasmanian system 230 (Figure S2 & S3). We also consider a hypothetical system using the vulture parameterisation. 231 The vulture parameterisation allows us to see how the mesoscavenger release generated from

This article is protected by copyright. All rights reserved 232 the obligate scavenger model would be affected by allowing an apex facultative scavenger 233 with considerably high search efficiency to consume non-carrion food sources. Here, all 234 parameters are set to match the vulture parameterisation in the obligate apex scavenger 235 model. We vary carrying capacity of the apex facultative scavenger in the absence of carrion 236 to range between 1/100th of mesoscavenger carrying capacity to mesoscavenger carrying 237 capacity, as we do not expect apex facultative scavengers to have higher densities than 238 mesoscavengers in the absence of carrion. 239 240 We find that mesoscavenger release occurs in facultative systems. For example, in the 241 vulture-parameterized model in Figure 4, mesoscavenger and carrion densities drop 3-fold 242 when the apex facultative scavenger’s carrying capacity in the absence of carrion reaches 243 1/10th of the mesoscavengers’ carrying capacity (Figure 4). However, in the Tasmanian devil- 244 parameterized model, mesoscavenger densities are only slightly impacted, likely due to the 245 relatively close search efficiencies between the apex scavengers and mesoscavengers (Figure 246 4). This suggests that apex facultative scavengers with higher search efficiencies (i.e. more 247 functionally dominant) relative to mesoscavengers tend to have greatest impact on 248 mesoscavenger and carrion densities, especially at lower apex scavenger carrying capacities 249 in the absence of carrion (Figure 4). A full local sensitivity analysis of the model 250 parameterisation is presented in the Supplementary Material (Figures S2 and S3). 251 252 We assume that facultative scavengers do not decrease their search effort on carrion when the 253 carrion are at low densities. It may be true that facultative scavengers reduce their search 254 efficiency for carrion when carrion densities are low, and alternatively increasingly target live 255 prey if more available, which is consistent with optimal foraging theory (Kane et al. 2017; 256 Margalida et al. 2017). However, it is empirically unclear how facultative scavenger search 257 efficiency fluctuates depending on food availability. Food switching could be modelled as a 258 discontinuous functional response, or even approximated by a Holling’s Type III sigmoidal 259 functional response, which is an approach taken in food-source-switching models of 260 predation (van Baalen et al. 2001). More research on facultative scavenger behaviour is 261 needed to realistically explore the effect of food switching on the mesoscavenger release Author Manuscript 262 hypothesis. 263 264 Known consequences of losing apex scavengers

This article is protected by copyright. All rights reserved 265 The literature we review in this paper, along with our dynamic models, suggest that without 266 apex scavengers, organic waste may be left unscavenged for longer periods and at higher 267 biomass, and this can increase mesoscavenger abundance (biomass in our models) due to the 268 increase in carrion biomass. Such increases in mesoscavengers and waste can change 269 ecosystem structure (Sebastián-González et al. 2016) and impact human well-being 270 (Markandya et al. 2008; Braczkowski et al. 2018; O’Bryan et al. 2018). Mesoscavengers are 271 often pest species. Their population increases can lead to a loss of species at lower trophic 272 levels (Ackerman et al. 2006), increased invasive species (Brown et al. 2015), increased 273 associated pest control costs (Buechley & Şekercioğlu 2016), and disease risk (Markandya et 274 al. 2008). There is evidence that some facultative mesoscavengers have faster reproductive 275 rates and can therefore increase in population size with more available resources compared to 276 apex obligate scavengers, assuming that increased carrion will result in increased fecundity 277 (Buechley & Şekercioğlu 2016). Research in the Canary Archipelago showed that increased 278 scavenging opportunities at “vulture restaurants” resulted in increased predation of native 279 ground-nesting birds by facultative scavengers (Cortez-Avanzada et al. 2009). Similarly, the 280 California gull (Larus californicus) depredated 61 percent of American avocet chicks 281 (Recurvirostra americana) and 23 percent of black-necked stilt chicks (Himantopus 282 mexicanus) in San Francisco Bay as a result of an increase in available refuse (Ackerman et 283 al. 2006). Furthermore, with the loss of apex scavengers and the increased availability of 284 carcasses under a mesoscavenger release scenario, it is likely that invertebrate scavengers 285 will have higher importance in reducing carcass biomass. Future investigations on the 286 relationship between mesoscavengers and invertebrate scavenger resource exploitative 287 competition would be informative. For example, it is plausible that invertebrate scavenger 288 release may occur with the loss of apex scavengers, but these dynamics are poorly understood 289 (but see DeVault et al. 2004). As a result, examination of the mesoscavenger release 290 hypothesis and the impacts across multiple taxonomic and functional groups would be an 291 important future research agenda. 292 293 Burgeoning mesoscavenger populations may affect human well-being (Braczkowski et al. 294 2018; O’Bryan et al. 2018), such as the spread of bubonic plagued rats of the 1300’s (Keeling Author Manuscript 295 & Gilligan 2000). More recently, rabies risk has increased dramatically following the rapid 296 decline of India’s apex obligate scavengers, vultures (Markandya et al. 2008) and more than 297 20,000 Indian citizens have died from rabies each year since 1985 (Menezes 2008). There 298 have been numerous recent calls for vulture conservation and feral dog sterilization

This article is protected by copyright. All rights reserved 299 (Markandya et al. 2008). Not only does increased carrion result in increased mesoscavengers, 300 but it can also result in amplified carcass-borne diseases, such as spongiform 301 encephalopathies found in unconsumed livestock remains in the European Union (Gwyther et 302 al. 2011). Additionally, in Kenya, carcasses without vultures had a three-fold increase in 303 interactions between facultative mesoscavengers (Ogada et al. 2012b), and the authors 304 contend a potential change in patterns of disease transmission between mesoscavengers as a 305 result. Thus, the loss of apex scavengers can alter scavenging assemblages that can cause risk 306 to ecosystem structure and human health either by increasing disease hosts or by increasing 307 carcass decay time. However, empirical evidence that apex scavengers alter mesoscavenger 308 dynamics at carcasses and thus disease dynamics on its own does not indicate that similar 309 patterns hold at the population-level. As such, the mesoscavenger release hypothesis should 310 be vigorously tested, especially around the effect of removing apex scavengers on 311 mesoscavengers at the population-level. 312 313 Apex scavenger conservation in the 21st century 314 Vultures and many apex facultative scavengers are among the most threatened functional 315 groups worldwide (Estes et al. 2011; Buechley & Şekercioğlu 2016; Ogada et al. 2016). A 316 better understanding of the effects of apex scavengers on mesoscavenger communities is 317 important to illuminate where and when to undertake conservation action. This has been 318 clearly demonstrated for apex predators and mesopredator release, with significant efforts 319 now underway to rewild and re-introduce populations of apex predators in North America 320 and Europe (Svenning et al. 2016). These efforts could be expanded to include the protection 321 of apex scavengers, particularly in places where mesoscavenger release could have negative 322 consequences for humans or ecosystems. Areas such as southern and eastern Africa, South 323 Asia, and the Iberian Peninsula appear to be of high priority for Old World vulture 324 conservation given current threats and lack of protection (Santangeli et al. 2019). 325 Additionally, developing nations may be particularly susceptible to the consequences of apex 326 scavenger loss due to a lack of waste disposal infrastructure and increased disease risk from 327 harmful mesoscavengers. 328 Author Manuscript 329 Since scavengers feed on organic waste, they are often found in human-dominated areas 330 where anthropogenic waste is prevalent. Indeed, avian scavengers in the Middle East and 331 Africa strongly select for habitats associated with humans, such as highways, power 332 distribution lines, and towns (Buechley et al. 2018). Apex scavenger persistence in shared

This article is protected by copyright. All rights reserved 333 landscapes will therefore require tolerance from local people who live alongside these 334 species. However, apex scavengers are frequently persecuted and viewed as nuisance animals 335 as they can be found around these human dwellings (Buechley & Şekercioğlu 2016). An 336 improved understanding of the importance of apex scavengers and the benefits they provide 337 could help raise their profile and make their conservation a global priority. For example, there 338 is evidence of a variety of human communities that tolerate apex scavengers. In the Tigray 339 region of Ethiopia, for instance, spotted hyenas (Crocuta crocuta), facultative scavengers, are 340 tolerated because of the traditional belief that they eat evil spirits as they remove livestock 341 remains and other garbage in and around waste dumps (Baynes-Rock 2015; Yirga et al. 342 2015). Similarly, Egyptian vultures (Neophron percnopterus) are thriving in urban areas of 343 Socotra, Yemen where they dispose of 22.4 percent organic waste produced in towns 344 annually (Gangoso et al. 2013). Nevertheless, when apex scavenger species are no longer 345 tolerated, facultative mesoscavengers may increase in abundance and richness, which is 346 likely to result in increased human-wildlife conflict. For example, the loss vultures in India 347 resulted in an increase in feral dog bites on humans (Markandya et al. 2008).As such, it is 348 imperative that future research focus on the relationship between scavenger trophic 349 interactions and human tolerance (Morales-Reyes et al. 2018). 350 351 Conclusion 352 While there is both empirical and theoretical evidence for apex scavengers releasing 353 mesoscavengers (the mesoscavenger release hypothesis), there is still much to learn about the 354 impacts of different apex scavengers on mesoscavenger assemblages. We find in our simple 355 dynamic models that the relationship between apex facultative scavengers and 356 mesoscavengers vary depending on their handling times, search efficiencies, and carrying 357 capacities (Figures 4, S1, & S2). We recommend that future work explore the dynamics of 358 these parameters as they pertain to different feeding strategies. Further to this, there is much 359 debate on how scavenger assemblages are arranged in time and space, such as the influence 360 of carcass species and type (Olson et al. 2016; Moleón et al. 2017), carcass size (Moleón et 361 al. 2015), season (Pereira et al. 2014), or location (Smith et al. 2017) among others on 362 community dynamics. There may be variable effects of apex scavenger removal on obligate Author Manuscript 363 versus facultative mesoscavengers. Additionally, most of the apex scavenger and 364 mesoscavenger examples illustrated in this manuscript, and the ones used in our models, 365 pertain to terrestrial scavengers, with vulture systems dominating the literature. As such, our 366 models do not consider the contribution of invertebrate scavengers relative to vertebrate

This article is protected by copyright. All rights reserved 367 scavengers (DeVault et al. 2003; Griffiths et al. 2018), and we recommend future research 368 incorporate invertebrates in models investigating the mesoscavenger release hypothesis. 369 Thus, our simple models are useful as a starting point for future studies exploring the effects 370 of apex scavengers on mesoscavengers across a multitude of systems. 371 372 Apex scavenger conservation is especially important given the potential negative impacts of 373 released mesoscavengers on ecosystem and human health. Based on a global meta-analysis, 374 there is already skewed dominance towards mesopredators and mesoscavengers that is likely 375 due to altered top-down control mechanisms from declining apex scavenger and predator 376 populations (Mateo-Tomás et al. 2015). Considering nearly 80 percent of obligate scavenger 377 species are currently in a rapid state of decline (Buechley & Şekercioğlu 2016), and 378 widespread evidence that human pressures alter apex scavenger distributions and population 379 viability (Ogada et al. 2012a; Di Marco & Santini 2015; Huijbers et al. 2015; Buechley & 380 Şekercioğlu 2016; S̜ekercioğlu et al. 2016), we urge proactive management tailored to apex 381 scavengers similar to approaches applied as a result of mesopredator release. By building the 382 research base associated with the mesoscavenger release hypothesis, we will likely better 383 understand the vulnerability of critical ecosystem and human well-being services that apex 384 scavengers provide and can therefore administer more effective conservation action. 385 386 Acknowledgements 387 C.J.O. would like to thank his mentor, J. Wallace Coffey, for his wisdom and guidance 388 leading to the generation of ideas that resulted in this manuscript. His legacy will not be 389 forgotten. The authors would like to thank James R. Allan, Sean Maxwell, and Calum 390 Cunningham for their feedback and assistance on this manuscript. This work was funded 391 partly by an Invasive Animal Cooperative Research Centre top-up scholarship, an Australian 392 International Postgraduate Research Scholarship, and funds provided by the Wildlife 393 Conservation Society. 394 395 396 397 Author Manuscript 398 399 400

This article is protected by copyright. All rights reserved Apex scavenger(s) Mesoscavenger(s) Impacts on mesoscavenger(s) Location Source (obligate and/or facultative) (obligate and/or facultative) Griffon vulture (Gyps fulvus) Red fox (Vulpes vulpes) Abundance of foxes increased with South-eastern (Morales-Reyes et al. vulture absence Spain 2017) Black bear (Ursus americanus) Bobcat (Lynx rufus), gray fox (Urocyon Total feeding time and presence of Mendocino (Allen et al. 2014); cinereoargenteus), western spotted skunk mesoscavengers on carrion decreased National Forest, (Allen et al. 2015) (Spirogale gracilis), ringtail (Bassariscus with bear presence California astutus), common raven (Corvus corax) Long-billed vulture (Gyps indicu), Feral dogs (Canis familiaris) Numbers of dogs significantly India (Markandya et al. slender-billed vulture (Gyps increased with vulture declines 2008) tenuirostris), oriental white-backed vulture (Gyps bengalensis) Wild boar (Sus scrofa), griffon Marten (Martes spp), red fox (Vulpes vulpes), Species richness and composition Mediterranean (Mateo-Tomás et al. vulture (Gyps fulvus), white-back feral dog (Canis familiaris), Eurasian jay drove carcass consumption in Spain and 2017) vulture (Gyps africanus), African (Garrulus glandarius), azure-winged magpie ecosystems where apex scavengers subtropical lion (Panthera leo) (Cyanopica cyanus), Eurasian magpie (Pica were rare, but context dependent South Africa pica), crows (Corvus spp) factors (e.g. species abundance) drove carcass consumption where apex scavengers were common.

Palm-nut vulture (Gypohierax Hyenas (Crocuta & Haena spp), black- Increase in contacts between Laikipia (Ogada et al. 2012b) angolensis), hooded vulture backed jackal (Canis mesomelas), Egyptian mesoscavengers and number of species District, Kenya (Necrosyrtes monachus), white- mongoose (Herpestes ichneumon) at carcasses without vultures Author Manuscript

This article is protected by copyright. All rights reserved backed vulture (Gyps africanus), Rüppell's vulture (G. rueppellii), lappet-faced vulture (Torgos tracheliotus) Griffon vulture (Gyps fulvus) Red kite (Milvus milvus), black kite (Milvus Vultures dominated carcasses in Northern Spain (Cortés-Avizanda et migrans), common raven (Corvus corax), predictable locations, reducing species al. 2012) marsh harrier (Circus aeruginosus), golden abundance of mesoscavengers eagle (Aquila chrysaetos) Tasmanian devil (Sarcophilus Feral cat (Felis catus), forest raven (Corvus Mesoscavengers increased carrion Tasmania, (Cunningham et al. harrisii) tasmanicus), spotted-tailed quoll (Dasyurus consumption in areas where Australia 2018) maculatus) Tasmanian devils have declined. Raven populations increased 2.2 fold following devil declines. 409 Table 1. Studies showing the impacts of apex scavengers on mesoscavenger abundance, presence at carrion, and species richness. Author Manuscript

This article is protected by copyright. All rights reserved 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 Author Manuscript 440 Figure 1. (A) Apex scavengers are more effective at detecting and consuming carrion than 441 mesoscavengers (large solid arrow), resulting in less carrion available to mesoscavengers 442 (small dotted arrow), and thus resulting in fewer mesoscavengers (small solid and dotted 443 arrows). This may result in indirect effects such as less disease risk, pest prevalence, and

This article is protected by copyright. All rights reserved 444 invasion potential that can negatively impact humans and ecosystem structure (small dotted 445 arrows). (B) The loss of apex scavengers can result in mesoscavenger release, which is 446 primarily caused by increased carrion availability due to a reduction in competition (large 447 solid arrow). Mesoscavenger release can result in indirect effects such as increased disease 448 risk, pest prevalence, and invasion potential that can negatively impact humans and 449 ecosystem structure (large dotted arrows). 450

451

452 Figure 2. Graphical illustration of our dynamic model to test theoretical support of the 453 mesoscavenger release hypothesis. In the model, carcasses enter the system via animal death 454 and leave the system through decay or by scavenging. Both apex scavengers and 455 mesoscavengers consume carcasses with respective efficiencies and handling times. In the 456 dynamic model, mesoscavenger and apex facultative scavenger populations have a logistic 457 growth rate (combination of births and deaths) in the absence of carcasses (apex facultative 458 scavenger logistic growth is denoted by the blue coloring and dashed arrows), whereas apex

459 obligate scavengersAuthor Manuscript have a mortality rate in the absence of carcasses.

460

461

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483 Figure 3. Equilibrium population densities of carcasses (green dotted line), mesoscavengers Author Manuscript 484 (red dashed line) and apex obligate scavengers (black solid line) from the dynamic model 485 (eqn 1) as a function of apex scavenger search efficiency. The curves for each plot start at 486 mesoscavengers search efficiency. The black open circle for each plot denotes the search

This article is protected by copyright. All rights reserved 487 efficiency of an apex scavenger, the griffon vulture (Gyps fulvus), which is 18 times more 488 efficient than a mesoscavenger assemblage. The red and green open circles on the 489 equilibrium axis denote the equilibrium densities of mesoscavengers and carcasses, 490 respectively, when the apex scavengers are absent. Generally, the more efficient (i.e. 491 functionally dominant) the apex scavenger, the more they suppress mesoscavenger 492 populations and carcass densities. The first column of plots is for the baseline mesoscavenger

493 search efficiency, em = 1. The second column of plots is for the case where mesoscavenger 494 search efficiency is doubled when apex obligate scavengers are present, potentially aiding 495 mesoscavengers in finding carcasses - as documented in some vulture systems (Kane & 496 Kendall 2017).

497

498

499

500

501

502

503

504

505

506

507

508

509

510

511 Author Manuscript 512

513

514

This article is protected by copyright. All rights reserved 515 Figure 4. Equilibrium population density of carcasses (green dotted line), mesoscavengers 516 (red dashed line) and apex facultative scavengers (black solid line) from the dynamic model 517 (eqn 2) as a function of apex scavenger carrying capacity in the absence of carcasses. Values 518 on the x-axis range from 1/100th of the mesoscavenger carrying capacity value to the 519 mesoscavenger carrying capacity value. The top model is parameterized for a known 520 facultative scavenger, the Tasmanian devil (Cunningham et al. 2018). The bottom is 521 parameterized for vulture systems (Morales-Reyes et al. 2017). The key difference between 522 the two parameterisations is different mesoscavenger and apex scavenger search efficiencies, 523 which are displayed in the top right of each plot. An imaginary effect of vultures surviving 524 off of alternative food sources is displayed for comparison with Figure 3. As apex scavengers

525 are able to sustain higher populations in the absence of scavenging (increasing ka), the more 526 they suppress mesoscavenger populations and carcass densities; however, apex facultative 527 scavengers with higher search efficiencies relative to mesoscavengers tend to have greater 528 impact at lower carrying capacities compared to apex scavengers with smaller search 529 efficiencies relative to mesoscavengers.

530

531

532

533 Literature cited:

534 Ackerman, J.T., Takekawa, J.Y., Strong, C., Athearn, N. & Rex, A. (2006). California Gull 535 Distribution, Abundance, and Predation on Waterbird Eggs and Chicks in South San Francisco 536 Bay Final Report.

537 Allen, M.L., Elbroch, L.M., Wilmers, C.C. & Wittmer, H.U. (2014). Trophic Facilitation or 538 Limitation? Comparative Effects of Pumas and Black Bears on the Scavenger Community. PLoS 539 One, 9, e102257.

540 Allen, M.L., Elbroch, L.M., Wilmers, C.C. & Wittmer, H.U. (2015). The comparative effects of large 541 carnivores on the acquisition of carrion by scavengers. Am. Nat., 185, 822–33.

542 van Baalen, M., Krivan, V., van Rijn, P.C. & Sabelis, M.W. (2001). Alternative food, switching Author Manuscript 543 predators, and the persistence of predator-prey systems. Am. Nat., 157, 512–24.

544 Barton, P.S., Cunningham, S.A., Lindenmayer, D.B. & Manning, A.D. (2013). The role of carrion in 545 maintaining biodiversity and ecological processes in terrestrial ecosystems. Oecologia, 171,

547 Baynes-Rock, M. (2015). Among the bone eaters : encounters with hyenas in Harar. The 548 Pennsylvania State University Press.

549 Braczkowski, A.R., O’Bryan, C.J., Stringer, M.J., Watson, J.E., Possingham, H.P. & Beyer, H.L. 550 (2018). Leopards provide public health benefits in Mumbai, India. Front. Ecol. Environ., 16, 551 176–182.

552 Brown, M.B., Schlacher, T.A., Schoeman, D.S., Weston, M.A., Huijbers, C.M., Olds, A.D., et al. 553 (2015). Invasive carnivores alter ecological function and enhance complementarity in scavenger 554 assemblages on ocean beaches. Ecology, 96, 2715–2725.

555 Buechley, E.R., McGrady, M.J., Çoban, E. & Şekercioğlu, Ç.H. (2018). Satellite tracking a wide- 556 ranging endangered vulture species to target conservation actions in the Middle East and East 557 Africa. Biodivers. Conserv., 1–18.

558 Buechley, E.R. & Şekercioğlu, Ç.H. (2016). The avian scavenger crisis: Looming extinctions, trophic 559 cascades, and loss of critical ecosystem functions. Biol. Conserv., 198, 220–228.

560 Butler, J.R.A. & du Toit, J.T. (2002). Diet of free-ranging domestic dogs (Canis familiaris) in rural 561 Zimbabwe: implications for wild scavengers on the periphery of wildlife reserves. Anim. 562 Conserv., 5, 29–37.

563 Carter, D.O., Yellowlees, D. & Tibbett, M. (2006). Cadaver decomposition in terrestrial ecosystems. 564 Naturwissenschaften, 94, 12–24.

565 Cortés-Avizanda, A., Jovani, R., Carrete, M. & Donázar, J.A. (2012). Resource unpredictability 566 promotes species diversity and coexistence in an avian scavenger guild: a field experiment. 567 Ecology, 93, 2570–2579.

568 Crooks, K.R. & Soulé, M.E. (1999). Mesopredator release and avifaunal extinctions in a fragmented 569 system. Nature, 400, 563–566.

570 Cunningham, C.X., Johnson, C.N., Barmuta, L.A., Hollings, T., Woehler, E.J. & Jones, M.E. (2018). 571 Top carnivore decline has cascading effects on scavengers and carrion persistence. Proc. R. Soc. 572 B Biol. Sci., 285, 20181582.

573 DeVault, T.L., Brisbin, Jr., I.L. & Rhodes, Jr., O.E. (2004). Factors influencing the acquisition of Author Manuscript 574 rodent carrion by vertebrate scavengers and decomposers. Can. J. Zool., 82, 502–509.

575 DeVault, T.L., Rhodes, Jr., O.E. & Shivik, J.A. (2003). Scavenging by vertebrates: behavioral, 576 ecological, and evolutionary perspectives on an important energy transfer pathway in terrestrial 577 ecosystems. Oikos, 102, 225–234.

This article is protected by copyright. All rights reserved 578 Estes, J.A., Terborgh, J., Brashares, J.S., Power, M.E., Berger, J., Bond, W.J., et al. (2011). Trophic 579 downgrading of planet Earth. Science, 333, 301–6.

580 Gangoso, L., Agudo, R., Anadón, J.D., de la Riva, M., Suleyman, A.S., Porter, R., et al. (2013). 581 Reinventing mutualism between humans and wild fauna: insights from vultures as ecosystem 582 services providers. Conserv. Lett., 6, 172–179.

583 Griffiths, H.M., Ashton, L.A., Walker, A.E., Hasan, F., Evans, T.A., Eggleton, P., et al. (2018). Ants 584 are the major agents of resource removal from tropical rainforests. J. Anim. Ecol., 87, 293–300.

585 Gwyther, C.L., Williams, P.A., Golyshin, P.N., Edwards-Jones, G. & Jones, D.L. (2011). The 586 environmental and biosecurity characteristics of livestock carcass disposal methods: A review. 587 Waste Manag., 31, 767–778.

588 Hill, J.E., DeVault, T.L., Beasley, J.C., Rhodes, O.E. & Belant, J.L. (2018). Effects of vulture 589 exclusion on carrion consumption by facultative scavengers. Ecol. Evol., 8, 2518–2526.

590 Houston, D.C. (1974). Food searching in griffon vultures. Afr. J. Ecol., 12, 63–77.

591 Houston, D.C. (1985). Evolutionary Ecology of Afrotropical and Neotropical Vultures in Forests. 592 Neotrop. Ornithol., 856–864.

593 Huijbers, C.M., Schlacher, T.A., Schoeman, D.S., Olds, A.D., Weston, M.A. & Connolly, R.M. 594 (2015). Limited functional redundancy in vertebrate scavenger guilds fails to compensate for the 595 loss of raptors from urbanized sandy beaches. Divers. Distrib., 21, 55–63.

596 Kane, A., Healy, K., Guillerme, T., Ruxton, G.D. & Jackson, A.L. (2017). A recipe for scavenging in 597 vertebrates - the natural history of a behaviour. Ecography (Cop.)., 40, 324–334.

598 Kane, A. & Kendall, C.J. (2017). Understanding how mammalian scavengers use information from 599 avian scavengers: cue from above. J. Anim. Ecol., 86, 837–846.

600 Komen, J. (1992). Energy requirements of adult cape vultures (Gyps coprotheres). J Raptor Res, 26, 601 213–218.

602 Di Marco, M. & Santini, L. (2015). Human pressures predict species’ geographic range size better 603 than biological traits. Glob. Chang. Biol., 21, 2169–2178.

604 Margalida, A., Colomer, M., Sánchez, R., Sánchez, F.J., Oria, J. & González, L.M. (2017). Behavioral

605 evidence of huntingAuthor Manuscript and foraging techniques by a top predator suggests the importance of 606 scavenging for preadults. Ecol. Evol., 7, 4192–4199.

607 Markandya, A., Taylor, T., Longo, A., Murty, M.N., Murty, S. & Dhavala, K. (2008). Counting the 608 cost of vulture decline—An appraisal of the human health and other benefits of vultures in India.

610 Mateo-Tomás, P., Olea, P.P., Moleón, M., Selva, N. & Sánchez-Zapata, J.A. (2017). Both rare and 611 common species support ecosystem services in scavenger communities. Glob. Ecol. Biogeogr., 612 26, 1459–1470.

613 Mateo-Tomás, P., Olea, P.P., Moleón, M., Vicente, J., Botella, F., Selva, N., et al. (2015). From 614 regional to global patterns in vertebrate scavenger communities subsidized by big game hunting. 615 Divers. Distrib., 21, 913–924.

616 Moleón, M., Martínez-Carrasco, C., Muellerklein, O.C., Getz, W.M., Muñoz-Lozano, C. & Sánchez- 617 Zapata, J.A. (2017). Carnivore carcasses are avoided by carnivores. J. Anim. Ecol., 86, 1179– 618 1191.

619 Moleón, M., Sánchez-Zapata, J.A., Sebastián-González, E. & Owen-Smith, N. (2015). Carcass size 620 shapes the structure and functioning of an African scavenging assemblage. Oikos, 124, 1391– 621 1403.

622 Moleón, M., Sánchez-Zapata, J.A., Selva, N., Donázar, J.A. & Owen-Smith, N. (2014). Inter-specific 623 interactions linking predation and scavenging in terrestrial vertebrate assemblages. Biol. Rev., 624 89, 1042–1054.

625 Morales-Reyes, Z., Martín-López, B., Moleón, M., Mateo-Tomás, P., Botella, F., Margalida, A., et al. 626 (2018). Farmer Perceptions of the Ecosystem Services Provided by Scavengers: What, Who, and 627 to Whom. Conserv. Lett., 11, e12392.

628 Morales-Reyes, Z., Sánchez-Zapata, J.A., Sebastián-González, E., Botella, F., Carrete, M. & Moleón, 629 M. (2017). Scavenging efficiency and red fox abundance in Mediterranean mountains with and 630 without vultures. Acta Oecologica, 79, 81–88.

631 Newsome, T.M., Greenville, A.C., Ćirović, D., Dickman, C.R., Johnson, C.N., Krofel, M., et al. 632 (2017). Top predators constrain mesopredator distributions. Nat. Commun., 8, 15469.

633 O’Bryan, C.J., Braczkowski, A.R., Beyer, H.L., Carter, N.H., Watson, J.E.M. & Mcdonald-Madden, 634 E. (2018). The contribution of predators and scavengers to human well-being. Nat. Ecol. Evol., 635 2, 229–236.

636 Ogada, D., Shaw, P., Beyers, R.L., Buij, R., Murn, C., Thiollay, J.M., et al. (2016). Another

637 Continental VultureAuthor Manuscript Crisis: Africa’s Vultures Collapsing toward Extinction. Conserv. Lett., 9, 638 89–97.

639 Ogada, D.L., Keesing, F. & Virani, M.Z. (2012a). Dropping dead: causes and consequences of vulture 640 population declines worldwide. Ann. N. Y. Acad. Sci., 1249, 57–71.

This article is protected by copyright. All rights reserved 641 Ogada, D.L., Torchin, M.E., Kinnaird, M.F. & Ezenwa, V.O. (2012b). Effects of Vulture Declines on 642 Facultative Scavengers and Potential Implications for Mammalian Disease Transmission. 643 Conserv. Biol., 1249, 453–460.

644 Olson, Z.H., Beasley, J.C., DeVault, T.L. & Rhodes, O.E. (2012). Scavenger community response to 645 the removal of a dominant scavenger. Oikos, 121, 77–84.

646 Olson, Z.H., Beasley, J.C. & Rhodes, O.E. (2016). Carcass Type Affects Local Scavenger Guilds 647 More than Habitat Connectivity. PLoS One, 11, e0147798.

648 Pereira, L.M., Owen-Smith, N. & Moleón, M. (2014). Facultative predation and scavenging by 649 mammalian carnivores: seasonal, regional and intra-guild comparisons. Mamm. Rev., 44, 44–55.

650 Ripple, W.J., Estes, J.A., Beschta, R.L., Wilmers, C.C., Ritchie, E.G., Hebblewhite, M., et al. (2014). 651 Status and ecological effects of the world’s largest carnivores. Science, 343, 151–162.

652 Ritchie, E.G. & Johnson, C.N. (2009). Predator interactions, mesopredator release and biodiversity 653 conservation. Ecol. Lett., 12, 982–998.

654 Santangeli, A., Girardello, M., Buechley, E., Botha, A., Minin, E. Di & Moilanen, A. (2019). Priority 655 areas for conservation of Old World vultures. Conserv. Biol.

656 Sebastián-González, E., Moleón, M., Gibert, J.P., Botella, F., Mateo-Tomás, P., Olea, P.P., et al. 657 (2016). Nested species-rich networks of scavenging vertebrates support high levels of 658 interspecific competition. Ecology, 97, 95–105.

659 S̜ekercioğlu, C., Wenny, D.G. & Whelan, C.J. (2016). Why birds matter: avian ecological function 660 and ecosystem services. University of Chicago Press.

661 Sekercioğlu, C.H., Daily, G.C. & Ehrlich, P.R. (2004). Ecosystem consequences of bird declines. 662 Proc. Natl. Acad. Sci., 101, 18042–18047.

663 Selva, N. & Fortuna, M.A. (2007). The nested structure of a scavenger community. Proc. R. Soc. B, 664 274, 1101–1108.

665 Sivy, K.J., Pozzanghera, C.B., Colson, K.E., Mumma, M.A. & Prugh, L.R. (2018). Apex predators 666 and the facilitation of resource partitioning among mesopredators. Oikos, 127, 607–621.

667 Smith, H.L. & Waltman, P.E. (1995). The theory of the chemostat : dynamics of microbial

668 competition. CambridgeAuthor Manuscript University Press.

669 Smith, J.B., Laatsch, L.J. & Beasley, J.C. (2017). Spatial complexity of carcass location influences 670 vertebrate scavenger efficiency and species composition. Sci. Rep., 7, 10250.

671 Svenning, J.-C., Pedersen, P.B.M., Donlan, C.J., Ejrnæs, R., Faurby, S., Galetti, M., et al. (2016).

This article is protected by copyright. All rights reserved 672 Science for a wilder Anthropocene: Synthesis and future directions for trophic rewilding 673 research. Proc. Natl. Acad. Sci. U. S. A., 113, 898–906.

674 Yirga, G., Leirs, H., De Iongh, H.H., Asmelash, T., Gebrehiwot, K., Deckers, J., et al. (2015). Spotted 675 hyena (Crocuta crocuta) concentrate around urban waste dumps across Tigray, northern 676 Ethiopia. Wildl. Res., 42, 563.

677

678

679

680

681 Author Manuscript

ele_13288_f1.png

ele_13288_f3.png

ele_13288_f4.png