Papers by Aniruddha V Belsare
Wildlife Society Bulletin, 2025
Models have become an integral component of wildlife conservation and management, and lie at the ... more Models have become an integral component of wildlife conservation and management, and lie at the foundation of decision-making in our field today. Yet, like too much of anything, there are risks with their overuse. Herein, we argue that despite their value, our increasing reliance on, and incorporation of, models into our data analysis, education of future wildlife profesionals, and development of management decisions has inherent risks. We believe that increased reliance on models has resulted in a detachment from the species and systems that we model, which threatens the quality of our science and, ultimately, decision-making. The impact of models, both pros and cons, on the wildlife field should be an often-discussed topic at wildlife conventions and within research agencies and academic institutions.
Landscape Ecology, 2025
Managing wild populations in rapidly changing, human-dominated landscapes requires models that ac... more Managing wild populations in rapidly changing, human-dominated landscapes requires models that accommodate complex interactions among climate, land use, disease, and evolution. Agent-based models (ABMs) are well suited to this task but are often difficult to parameterize, calibrate, and interpret at management-relevant scales. Objectives: We discuss how artificial-intelligence (AI) techniques, including machine-learning regression, data-mining diagnostics, geospatial informatics, and large-language-model code aides, can streamline ABM parameter estimation and scenario testing, enhance extraction of decision-support metrics, and broaden the accessibility of ABMs for conservation planning. Methods: We considered examples of AI use in ecology and evolution, including where AI was paired with ABMs, highlighting use cases such as calibration, rule discovery, data fusion, and code generation. Results: We show how supervised machine learning can supplement parameterization by learning relationships between empirical observations and model outputs. Data-mining methods may also be useful to identify parameters that drive most output variance. In addition, deep-learning remote-sensing products in ABMs allows landscape dynamics to be represented at ecologically relevant resolutions. Despite this, key obstacles, such as limited long-term ecological data, high computational demand, and the need for explainable-AI safeguards against biased predictions remain. Conclusions: Expanding the use of AI in ABMs will require interdisciplinary collaborations that pair ecologists with computer and geo-information scientists and explicit workflows for auditing AI decisions. However, leveraging AI enhanced ABMs will improve predictive modeling of species responses to environmental change, optimize conservation strategies, and develop more effective data-driven management.
Pathogens, 2025
Lyssaviruses are RNA viruses in the Family Rhabdoviridae, Genus Lyssavirus. They represent the ca... more Lyssaviruses are RNA viruses in the Family Rhabdoviridae, Genus Lyssavirus. They represent the causative agents of acute, progressive encephalitis, known historically as rabies. Regardless of specific etiology, their collective viral morphology, biochemistry, pathobiology, associated clinical signs, diagnosis, epizootiology, and management are essentially the same. Despite centuries of clinical recognition, these quintessential neurotropic agents remain significant pathogens today, with substantive consequences to agriculture, public health, and conservation biology. Notably, the singular morbidity caused by lyssaviruses is incurable and constitutes the highest case fatality of any viral disease. All warm-blooded vertebrates are believed to be susceptible. The dog is the only domestic animal that serves as a reservoir, vector, and victim. In contrast, felids are effective vectors, but not reservoirs. All other rabid domestic species, such as livestock, constitute spillover infections, as a bellwether to local lyssavirus activity. Frequently, professional confusion abounds among the veterinary community, because although the viral species Lyssavirus rabies is inarguably the best-known representative in the Genus, at least 20 other recognized or putative members of this monophyletic group are known. Frequently, this is simply overlooked. Moreover, often the ‘taxonomic etiology’ (i.e., ‘Lyssavirus x’) is mistakenly referenced in a biopolitcal context, instead of the obvious clinical illness (i.e., ‘rabies’). Global consternation persists, if localities believe they are ‘disease-free’, when documented lyssaviruses circulate or laboratory-based surveillance is inadequate to support such claims. Understandably, professional chagrin develops when individuals mistake the epidemiological terminology of control, prevention, elimination, etc. Management is not simple, given that the only licensed veterinary and human vaccines are against rabies virus, sensu lato. There are no adequate antiviral drugs for any lyssaviruses or cross-reactive biologics developed against more distantly related viral members. While representative taxa among the mammalian Orders Chiroptera, Carnivora, and Primates exemplify the major global reservoirs, which mammalian species are responsible for the perpetuation of other lyssaviruses remains a seemingly academic curiosity. This zoonosis is neglected. Clearly, with such underlying characteristics as a fundamental ‘disease of nature’, rabies, unlike smallpox and rinderpest, is not a candidate for eradication. With the worldwide zeal to drive human fatalities from canine rabies viruses to zero by the rapidly approaching year 2030, enhanced surveillance Pathogens 2025, 14, 586 https://doi.org/10.3390/pathogens14060586Pathogens 2025, 14, 586 2 of 60 and greater introspection of the poorly appreciated burden posed by rabies virus and diverse other lyssaviruses may manifest as an epidemiological luxury to the overall global program of the future.
Ecological Modelling, Feb 1, 2020
Epidemiological surveillance for important wildlife diseases often relies on samples obtained fro... more Epidemiological surveillance for important wildlife diseases often relies on samples obtained from hunter-harvested animals. A problem, however, is that although convenient and cost-effective, hunter-harvest samples are not representative of the population due to heterogeneities in disease distribution and biased sampling. We developed an agent-based modeling framework that i) simulates a deer population in a user-generated landscape, and ii) uses a snapshot of the in silico deer population to simulate disease prevalence and distribution, harvest effort and sampling as per user-specified parameters. This framework can incorporate real-world heterogeneities in disease distribution, hunter harvest and harvest-based sampling, and therefore can be useful in informing wildlife disease surveillance strategies, specifically to determine population-specific sample sizes necessary for prompt detection of disease. Application of this framework is illustrated using the example of chronic wasting disease (CWD) surveillance in Missouri's white-tailed deer (Odocoileus virginianus) population. We show how confidence in detecting CWD is grossly overestimated under the unrealistic, but standard, assumptions that sampling effort and disease are randomly and independently distributed. We then provide adjusted sample size recommendations based on more realistic assumptions. Wildlife agencies can use these open-access models to design their CWD surveillance. Furthermore, these models can be readily adapted to other regions and other wildlife disease systems.
Review for "An outbreak of tuberculosis in endangered northern pig‐tailed macaques ( <i>Macaca leonina</i> ) and milu deer ( <i>Elaphurus davidianus</i> ) from a zoo in China
Review for "Challenges and opportunities using hunters to monitor chronic wasting disease among wild reindeer in the digital era
Supplementary material from "Harvest strategies for the elimination of low prevalence wildlife diseases
The intensive harvesting of hosts is often the only practicable strategy for controlling emerging... more The intensive harvesting of hosts is often the only practicable strategy for controlling emerging wildlife diseases. Several harvesting approaches have been explored theoretically with the objective of lowering transmission rates, decreasing the transmission period or specifically targeting spatial disease clusters or high-risk demographic groups. Here, we present a novel model-based approach to evaluate alternative harvest regimes, in terms of demographic composition and rates, intended to increase the probability to remove all infected individuals in the population during the early phase of an outbreak. We tested the utility of the method for the elimination of chronic wasting disease based on empirical data for reindeer (<i>Rangifer tarandus</i>) in Norway, in populations with (Nordfjella) and without (Hardangervidda) knowledge about exact disease prevalence and population abundance. Low and medium harvest intensities were unsuccessful in eliminating the disease, even at low prevalence. High-intensity harvesting had a high likelihood of eliminating the disease, but probability was strongly influenced by the disease prevalence. We suggest that the uncertainty about disease prevalence can be mitigated by using an adaptive management approach: forecast from models after each harvest season with updated data, derive prevalence estimates and forecast further harvesting. We identified the problems arising from disease surveillance with large fluctuations in harvesting pressure and hence sample sizes. The elimination method may be suitable for pathogens that cause long-lasting infections and with slow epidemic growth, but the method should only be attempted if there is a low risk of reinfection, either by a new disease introduction event (e.g. dispersing hosts) or due to environmental reservoirs. Our simulations highlighted the short time window when such a strategy is likely to be successful before approaching near complete eradication of the population.
Leopards in Human dominated Areas: a Spillover from Sustained Translocations Into Nearby Forests?
Journal of the Bombay Natural History Society, 2007
ABSTRACT In the past decade, parts of many Indian states have reported an increase in leopard (Pa... more ABSTRACT In the past decade, parts of many Indian states have reported an increase in leopard (Panthera pardus fusca) populations outside forests accompanied by large numbers of attacks on people. This high density was attributed to declining natural habitats and prey species, and the increased survival of leopards in croplands where they preyed on tended and feral domestic animals. That leopard cubs were frequently found in agricultural fields was thought to also indicate rising leopard populations. We use data from our human-leopard conflict study in Junnar, Maharashtra along with information from three other conflict sites in India to propose that the reason for this increase in leopard population and conflict is related to the sustained translocations of ‘problem’ leopards into nearby forests. That sustained releases could lead to population increases was never considered before, even though translocation is known to be a procedure aimed at increasing populations of species at or close to the site of release. Although scientists do not recommend translocations as a management strategy used in response to problem carnivores, it is currently the legally recommended method of dealing with ‘problem’ large cats in India. Such faulty policies will only further condemn the conservation of this species, which is hunted in large numbers for the illegal wildlife trade.
Preventive Veterinary Medicine, Aug 1, 2013
Mass vaccination of dogs is a mainstay for efforts to control rabies and other viral pathogens. T... more Mass vaccination of dogs is a mainstay for efforts to control rabies and other viral pathogens. The success of such programs is a function of the ability to vaccinate sufficient proportions of animals to develop herd immunity. However, fully assessing success in reaching target vaccination-levels and in understanding the outcome of mass vaccination efforts is hindered if insufficient information is available on the demographics of dog populations and the prevalence of the targeted pathogens. While such information can sometimes be gained from questionnaire surveys, greater precision requires direct assessment of the dog populations. Here we show how such information can be gained from surveys of dogs conducted in association with mass-vaccination programs. We conducted surveys of dogs in six villages in rural Maharashtra, India, between February and July 2011 as part of an effort to reduce the risk of human rabies and virus transmission from dogs to wildlife. Mass vaccination efforts were conducted in each village, and paired with blood sample collection and photographic mark-recapture approaches to gain epidemiologic and demographic data. This data in turn facilitated estimates of dog abundance, population density and structure, vaccination coverage, and seroprevalence of antibodies against canine adenovirus (CAV), canine parvovirus (CPV), and canine distemper virus (CDV). The median dog population size for the six villages was 134 (range 90-188), the median dog population density was 719 dogs per km 2 (range 526-969), and the median human:dog ratio for these six villages was 34 (range 30-47). The median household:dog ratio for the six villages was 6 (range 5-8). Following vaccination efforts, the median vaccination coverage achieved was 34% (range 24-42%). The dog populations consisted mostly of adult dogs (67-86%) and the median sex ratio for the study area was male biased (1.55 males per female; range 0.9-2.5). The seroprevalence of antibodies against CAV, CPV and CDV was 68, 88 and 73%, respectively. Mass vaccination campaigns provide an opportunity to obtain vital epidemiological and demographic data, and develop a clearer understanding of the threats and impacts of diseases and disease control measures.
Journal of the Arkansas Academy of Science, 2017
Transboundary and Emerging Diseases, Aug 1, 2014
There is an increasing concern that free-ranging domestic dog (Canis familiaris) populations may ... more There is an increasing concern that free-ranging domestic dog (Canis familiaris) populations may serve as reservoirs of pathogens which may be transmitted to wildlife. We documented the prevalence of antibodies to three viral pathogens, canine parvovirus (CPV), canine distemper virus (CDV) and canine adenovirus (CAV), in free-ranging dog and sympatric Indian fox (Vulpes bengalensis) populations in and around the Great Indian Bustard Wildlife Sanctuary, in Maharashtra, central India. A total of 219 dogs and 33 foxes were sampled during the study period. Ninety-three percentage of dogs and 87% of foxes were exposed to one or more of the three pathogens. Exposure rates in dogs were high: >88% for CPV, >72% for CDV and 71% for CAV. A large proportion of adult dogs had antibodies against these pathogens due to seroconversion following earlier natural infection. The high prevalence of exposure to these pathogens across the sampling sessions, significantly higher exposure rates of adults compared with juveniles, and seroconversion in some unvaccinated dogs documented during the study period suggests that these pathogens are enzootic. The prevalence of exposure to CPV, CDV and CAV in foxes was 48%, 18% and 52%, respectively. Further, a high rate of mortality was documented in foxes with serologic evidence of ongoing CDV infection. Dogs could be playing a role in the maintenance and transmission of these pathogens in the fox population, but our findings show that most dogs in the population are immune to these pathogens by virtue of earlier natural infection, and therefore, these individuals make little current or future contribution to viral maintenance. Vaccination of this cohort will neither greatly improve their collective immune status nor contribute to herd immunity. Our findings have potentially important implications for dog disease control programmes that propose using canine vaccination as a tool for conservation management of wild carnivore populations.
Zoos' print journal, May 21, 2003
The reasons for contraception in a zoo setting are numerous of which some have been listed below ... more The reasons for contraception in a zoo setting are numerous of which some have been listed below (Gerlofsma et al., 1994): a) short age of enclosure space; b) prevention of inbreeding or represented lines; c) prevention of passing on undesirable genetic traits; d) to increase generation length; e) to prevent mating between too young or too old animals; f) poor maternal care shown by the dam; g) as a result of veterinary indication/ intervention. Case history: A Blackbuck doe (Antilope cervicapra), acquired from the forest authorities in 1999 and kept at the Rajiv Gandhi Zoological Park, Katraj, Pune showed signs of dystokia on 23 November 2001. Efforts to induce parturition using Epidosin (Valethamate 8mg) and Pitocin (Oxytocin 10IU) intramuscularly failed, hence the animal was anaesthetized on 24 November 2001, using a combination of Xylazine @1mg kg-1 and Ketamine @1mg kg-1 in a 1:1 proportion. The dead foetus was removed by caesarean section and it was observed that the foetus had an abnormally large head and its hind limbs were kinked at the fetlock joints. With daily administration of broad-spectrum antibiotics and dressing of surgical wound, the animal had an uneventful recovery from the surgery and the stitches were removed after 15 days.
Helminthologia, May 17, 2017
Baylisascaris procyonis is a helminth parasite of raccoons Procyon lotor and represents a health ... more Baylisascaris procyonis is a helminth parasite of raccoons Procyon lotor and represents a health concern in paratenic hosts, including humans and diverse domestic and wildlife species. In North America the helminth is expanding its geographic range. To better understand patterns of infection in the Ozark region of the USA, raccoons (n = 61) were collected in 2013-2014 from fi ve counties in Missouri and Arkansas, USA and necropsied. We documented B. procyonis in all surveyed locations. The overall prevalence of B. procyonis was 44.3 % (95 % CI = 31.9-57.4) and was signifi cantly higher in females than males. There were also signifi cant differences in prevalence among raccoons sampled north and south of the Missouri River. Mean intensity was 9.9 (CI = 5.44-17.22), and parasites were highly aggregated among hosts such that approximately 20 % of hosts harbor 90 % of parasites. These levels of parasitism indicate that B. procyonis is common in the region and its impacts on paratenic hosts could be qualitatively similar to effects observed in other localities.
Parasitology International, Aug 1, 2018
Baylisascaris procyonis is a nematode of significant concern to public and domestic animal health... more Baylisascaris procyonis is a nematode of significant concern to public and domestic animal health as well as wildlife management. The population genetics of B. procyonis is poorly understood. To gain insights into patterns of genetic diversity within (infrapopulation level) and among (component population level) raccoon (Procyon lotor) hosts, and specifically to assess the relative importance of indirect and direct transmission of the parasite for explaining observed population structure, we collected 69 B. procyonis from 17 wild raccoons inhabiting five counties in Missouri and Arkansas, USA. Informative regions of mitochondrial (CO1, CO2) and nuclear (28S, ITS2) genes were amplified and the distribution and genetic variability of these genes were assessed within and across raccoons. Concatenation of the CO1 and CO2 mtDNA sequences resulted in 5 unique haplotypes, with haplotype diversity 0.456 ± 0.068. The most common haplotype occurred in 94% of raccoons and 72.5% of B. procyonis. Sequences for 28S rDNA revealed four unique nuclear genotypes, the most common found in 100% of raccoons and 82.6% of B. procyonis. ITS2 genotypes were assessed using fragment analysis, and there was a 1:1 correspondence between 28S and ITS-2 genotypes. Infrapopulation variation in haplotypes and genotypes was high and virtually all hosts infected with multiple sequenced nematodes also harbored multiple haplotypes and genotypes. There was a positive relationship between the size of the analyzed infrapopulation (i.e., the number of nematodes analyzed) and the number of haplotypes identified in an individual. Collectively this work emphasizes the importance of indirect transmission in the lifecycle to this parasite.
Journal of Zoo and Wildlife Medicine, Jun 1, 2010
In India, leopards (Panthera pardus fusca) inhabit human-dominated landscapes, resulting in encou... more In India, leopards (Panthera pardus fusca) inhabit human-dominated landscapes, resulting in encounters that require interventions to prevent harm to people, as well as the leopards. Immobilization is a prerequisite for any such intervention. Such emergency field immobilizations have to be carried out with limited tools, often amidst large uncontrollable crowds. An effective and practicable approach is discussed, based on 55 wild leopard immobilizations undertaken between January 2003 and April 2008. A xylazine hydrochloride (1.4 6 0.3 mg/kg)-ketamine hydrochloride (5 6 2 mg/kg) mixture was used for immobilization of leopards, based on estimated body weight. When weight could not be estimated, a standard initial dose of 50 mg of xylazine-150 mg of ketamine was used. Supplemental doses (50-75 mg) of only ketamine were used as required. No life-threatening adverse effects of immobilization were documented for at least 1 mo postimmobilization.
I obtained baseline demographic and epidemiologic data for village dog populations around a prote... more I obtained baseline demographic and epidemiologic data for village dog populations around a protected area in central India, with the objective of evaluating the risks dog populations present to sympatric wild carnivores. Demographic surveys indicated that dogs occurred at high densities in the study villages (> 526 dogs per km 2). Epidemiological surveys documented a high prevalence of exposure to three pathogens of conservation concern, canine parvovirus (CPV), canine distemper virus (CDV), and
Review for "An outbreak of tuberculosis in endangered northern pig‐tailed macaques ( <i>Macaca leonina</i> ) and milu deer ( <i>Elaphurus davidianus</i> ) from a zoo in China
Investigating <i>Leptospira</i> dynamics in a multi‐host community using an agent‐based modelling approach
Transboundary and Emerging Diseases, Nov 1, 2022
To vaccinate or not to vaccinate: lessons learned from an experimental mass vaccination of free-ranging dog populations
Animal Conservation, Aug 15, 2014
Domestic animal populations have the potential to act as reservoirs for multi‐host pathogens, whi... more Domestic animal populations have the potential to act as reservoirs for multi‐host pathogens, which may be transmitted to native species and cause population declines or extirpations. Domestic dogs are known reservoirs for several multi‐host pathogens that may be transmitted to native carnivores. Mass vaccination of dogs has been suggested as a tool for mitigating the disease threat to sympatric carnivores. To determine the extent to which such mass vaccination programs are practicable and appropriate for large, free‐ranging dog populations, a vaccination experiment was undertaken in six villages adjoining the Great Indian Bustard Wildlife Sanctuary in central India. Dogs from three villages were vaccinated against rabies virus, canine adenovirus (CAV), canine parvovirus (CPV) and canine distemper virus (CDV) (treatment dogs), while those from three other villages were only vaccinated against rabies virus (control dogs). Immunoglobulin G (IgG) antibody titers against CAV, CPV and CDV in control and treatment dogs were determined on four occasions during the study. When first examined, a large proportion of the initially unvaccinated adult dogs in both the control and treatment group were IgG antibody positive, indicating prior natural exposure. Dogs recovering from natural infection because of CAV, CPV and CDV develop a lifelong immunity and are thus protected against these pathogens. For each of the three viruses, background antibody seroprevalence rates were generally > 72%. Furthermore, several unvaccinated adult dogs acquired protection against these pathogens during the study. Vaccination failed to increase the proportion of adult dogs with IgG antibodies against CAV, CPV or CDV in the treatment group compared with the control group, as much of the effort was put into vaccinating dogs that were already antibody positive. In such situations, vaccination of adult dogs against these enzootic viral pathogens seems unnecessary, and would escalate the cost‐benefit ratio of dog disease control programs.
Uploads
Papers by Aniruddha V Belsare