Assistants

Rare and endangered carnivores of Amazonia: Research and Conservation of the giant otter and the short-eared dog

Supported by

This project aims to promote conservation of the Giant River Otter (Pteronura brasiliensis) and the Short-Eared Dog (Atelocynus microtis), in one of the largest and most pristine rainforest areas of the world, in the Manú National Park (Perú) and the adjacent Alto Purús reserved zone. The choice to focus on these two species is based on two overriding concerns. The first is the extreme paucity of basic ecological data that exists on these two top carnivores, and the second is their known susceptibility to human presence, in particular because they are vulnerable to diseases spread by domestic dogs (i.e., canine distemper and parvovirus). The proposal requests funding to support a research team of American, Brazilian and Perúvian researchers, who will investigate the animals' basic ecology, habitat needs, regional distribution and potential threats. Based on results of this research a strategy for conservation of these two species will be proposed to the park authorities. Part of our conservation plan is to improve long-term viability of the populations of these two predators by highlighting their special needs, including protection from diseases of domestic dogs, and large, intact areas of habitat. Ultimately we hope to engage park authorities in a plan to elevate the status of the Alto Purús reserve zone to a stronger category of protection, and through education and outreach, to elevate the appreciation of these charismatic animals to the status of "flagship species" for the newly-created protected area.

INTRODUCTION, GOALS AND OBJECTIVES

The Center for Tropical Conservation (CTC) has long focused its scientific research programs on understanding the interaction of top-level animals such as carnivores with their environment. Carnivores are considered to be critical for maintaining the biological integrity of ecosystems, yet they are often the most threatened members of animal assemblages (Greene, 1988; Terborgh, 1990; Soulé and Terborgh, 1999). Large predators play an important role in regulating the distribution and abundance of smaller carnivores and other prey species, and are believed to have a dramatic indirect effect on plant communities (Terborgh et al., 1988; Crooks and Soulé, 1999; and many others). Aspects of the biology and ecological role of some Neotropical carnivores have been studied over the last three decades but other species are barely studied and their ecological roles unknown. Giant otters (Pteronura brasiliensis) and the short-eared dog (Atelocynus microtis) can be considered among the most poorly-understood carnivores in the Neotropics. Preliminary studies begun in 1999 and 2000 by Leite and Davenport have established the existence of populations of these animals around the Cocha Cashu Biological Station, which can be readily studied. Nevertheless, much more work needs to be completed to gain even a basic understanding of their needs and ecological roles in their terrestrial and aquatic environments. In addition to the lack of data on the animals' ecology, it is strongly suspected that these rare carnivores are threatened by diseases spread by domestic dogs kept by local people for hunting and as pets. Domestic dogs are known worldwide to host viruses such as rabies, Canine Distemper (CDV) and Parvovirus (CPV) that impact a variety of wildlife species. Some of these diseases have been responsible for infecting lions, seals, ferrets, African wild dogs, Ethiopian wolves, crab-eating fox, North American otters, and gray wolves (Bengston et al, 1991, Alexander et al, 1996, Sillero-Zubiri et al. 1996). This project aims to study both the basic ecology and the threats that may imperil these carnivores' long-term survival. The recent (2000) expansion of the regional conservation network that includes Manú National Park and the adjacent Alto Purús reserved zone (Figure 1) poses a unique opportunity for these Neotropical carnivores and other mammals that are at great danger of extinction due to their sensitivity to human disturbance (Pierret, 2000). The CTC, with its long history of work in the Manú and with its connections to both scientific and conservation communities, is in a prime position to take maximal advantage of this opportunity, and to realize projects with high scientific and conservation relevance. We have pinpointed the "Rare and Endangered Carnivores" project as a priority program that will both fill in huge gaps in scientific work on Neotropical carnivores, and that will generate conservation programs for some of the most threatened carnivores on the continent. The proposed project will assemble a multi-disciplinary team of a conservation biologist (Lisa Davenport, USA), a veterinarian (Dr. Renata Leite, Brazil) and multiple Perúvian collaborators (esp. Blgo. Ciprian Aguirre), to conduct the first systematic ecological study of the giant-otter and short-eared dog, including their habitat needs, feeding ecology and threats to their conservation. Figure 1 - Protected Areas of West of Perú. Modified from Protected Areas of Peru, INRENA, 2000. The project will form part of a larger long-term survey and conservation program involving several international and local conservation institutions with which the CTC already interacts. These include the Frankfurt Zoo Giant Otter Project (FZGOP), INRENA (Perúvian Parks and Forestry Dept.), Associação Pró-Carnivoros (Brazilian conservation group), IUCN Mammal Specialist Group, the IUCN Otter Specialist Group and IDEAWILD. Initial research by Leite and Davenport on the short-eared dog and the giant otter, respectively, are already underway, but lack sufficient support to build a cohesive conservation program. The results of this initial research represent the first ecological data collected to date on the short-eared dog (Leite, 2000; Leite at al., 2000, Leite and Williams, in press), and the only intensive study of giant otter microhabitat use and feeding ecology for the region (Davenport, 1999 and Davenport 2000). Based on the research results obtained, we will diagnose conservation needs and propose a conservation strategy for maintaining viable populations within the Manú and adjacent reserves. The full project is expected to last a minimum of 5 years, including scientific and outreach goals. The current proposal requests funding to begin the first continuous work on the project, which to this point has only been realized during brief periods of research under the Principal Investigators' own initiative and with little outside support. Ecological work on the giant otter will contribute to the PhD thesis of Ms. Lisa Davenport, a long-time Research Associate of the CTC, who will maintain a part-time affiliation with the CTC while pursuing a PhD in the Biology Department of the University of North Carolina at Chapel Hill. She expects to complete her degree 2006. Dr. Renata Leite will maintain her position with the CTC for the duration of the project, and will be the administrative head of the project. Research Components and Questions To be able to elaborate a conservation strategy for the giant otter and the short-eared dog, we need to answer some basic but important questions: 1) What habitats do the animals use, and at what density do they exist? What prey species do they consume in different microhabitats and can it be determined what constitutes prime habitat types? Does the otter actually reduce populations of human-preferred fish species, or does it in fact support higher diversity and increased populations as do other keystone predators (including sea otters)? What is the availability of prime habitat within the Manú and adjacent reserved zones? 2) What threats to these animals are most prevalent, and what options exist to prevent conflict between local people and wildlife? Specifically, are domestic dogs brought in by local inhabitants carrying canine viruses that could endanger the otters, short-eared dogs, and other wildlife? Do local people view the giant otter as a fish competitor, and how severe is their persecution of otters? What other threats constitute a threat to survival of these species (mercury, over-hunting, over-fishing, exotic fish introductions, etc)? These two overriding project components form the core of our research and conservation proposal. Detailed methods and discussion of the different subcomponents are enumerated below, in addition to a description of our plan to conduct one of the first expeditions into the Alto Purús that will focus on evaluating the potential of the larger protected zone for maintaining viable populations of the endangered giant otter and short-eared dog. Finally, we discuss our plan for conservation, education and outreach that will comprise the final component of the project, which we will develop based upon findings from our ecological work.

RESEARCH COMPONENT 1:

Habitat Use, Abundance and Predator-Prey Ecology

So little is known about the giant otter and the short-eared dog, that it is still not clear what habitats constitute prime habitat for these animals. The short-eared dog is almost completely unstudied, although most visual records of it come from areas adjacent to rivers. The giant otter, of course, is tightly restricted to rivers and oxbow lakes, but even these show great variation in size, depth, pH, turbidity, oxygen content, and presence of underwater vegetation. The fish fauna of these different lake types is variable, but poorly studied, and will be a focus for the giant otter study of habitat-related factors in feeding ecology and microhabitat use. Here, we explain the methods to be used for the studying the animals and for answering the questions posed above. Giant Otter Research Status and abundance. The giant river otter (Pteronura brasiliensis) is one of the most beloved residents of Perú's Manú National Park, being a major draw for the park's many international tourists. Yet it is also one of the most threatened, existing at very low densities where it is found. Outside the park, it faces threats from local fishing, gold-mining, aquaculture, non-native fish introductions, disease risks and tourism activities (Schenck and Staib 1996, Schenck and Staib 1998; Schenck and Staib 2001). Few protected areas are large enough to maintain viable populations of giant otters, and even the population of the Manú is still not secure. The entire Manú Biosphere Reserve of approximately 2 million hectares currently supports only about 65 otters (based on surveys by the Frankfurt Zoo Giant Otter Project) and the status of the Alto Purús populations is completely unknown (Hajek pers. comm.). For the giant otter, status assessment will concentrate within the Alto Purús zone, using the same methodology that Frankfurt Zoo developed for the Manú. Giant otters are diurnal, noisy, and curious, and thus directly observable where present. They have conspicuous den sites and latrines, and also display white throat markings that allow identification of individuals. The methodology devised for surveying giant otters therefore is straightforward, and comprehensive, so long as all potential waterways and oxbow lakes in an area are reached and assessed. Video-taping observed individuals is the best method for avoiding the potential problem of double-counting individuals or family groups. Habitat Use and Predator-prey ecology. Local people consider giant otter to be a competitor for fish, yet studies of temperate aquatic systems have shown that top predators often increase both abundance and diversity of herbivorous fish, by predating on mid-level (or "meso") predatory fish (an example of top-down regulation) (Carpenter and Kitchell 1993). In addition, studies of nutrient enrichment have revealed that where top predators such as caiman or otters are removed, supposedly to increase fish abundance, fish populations actually decline (Fittkau, 1973). Declines occur as an indirect consequence of reduced primary productivity in the aquatic system because the principle source of nutrients is the feces input from these carnivores. Alternatively, true bottom-up processes such as oxygen availability or geological formations that permit nutrient sinks may control fish communities. Clearly, the system is too little understood to comprehend what habitats are required for the long-term survival of otters. In this vacuum of knowledge, it is quite likely that the interactions between otters and fish may be completely misconstrued by local people. Yet, to date there exist no comprehensive studies on tropical lake ecology and dynamics to determine what top-down, bottom-up, or other processes control fish communities. To complete such a study will require multiple periods of sampling over several years, monitoring changes in vegetation, fish, and zooplankton communities, and conducting experimental tests on the effect of excluding predators from some portion of the lakes. The topic will comprise the PhD dissertation research of Ms. Lisa Davenport, and some preliminary work has already been completed. Project Results to Date: In the fall of 1999 and 2000, the boundaries of Cocha Cashu and Cocha Totora were mapped with GIS, bathymetric profiles were measured, and edge and underwater vegetation classified and mapped. Still required are detailed determinations of the presence of features such as logs, caiman lairs, and topography which would affect fish habitat and hunting conditions for otters. Such observations are possible with "fish-finder" sonar instruments currently available. Observations of otter hunting behavior and hunting success indicate that the microhabitat does affect the fishing method used, the size and type of prey caught, and the handling time, although considerably more data is required. In addition pilot experiments with the fishes "boca chico" (Prochilodus magdalanae) and "bujurki" (Satanoperca sp.) determined successful marking techniques for a catch-release program on the fish species of primary interest. Technical difficulties with a generator impeded a preliminary assessment of electrosampling in 2000. Included in the budget is a request for an improved, portable model which should allow further study of the fish of Cocha Cashu and 8 other regional lakes. Mark and release work will begin in 2001, and will be monitored throughout the duration of the study. Much of the essential equipment required for the study has already been donated or acquired, including two portable kayaks, material for blinds, image stabilizing binoculars, stopwatches, a digital video camera, electroshocker, gill nets and fish marking equipment.

Still required are 1) a 1000 KW generator for electroshocking; 2) an electronic fishfinder, Lowrance X-65 for mapping the lake bottom and features and for obtaining an independent measure of fish abundance; and 3) additional Visible Implant Elastomer, used in marking fish that are captured and released. Short-term goals for the 2001-2002 period include: · To document giant otter hunting behavior and hunting success in the various microhabitats of Cocha Cashu and Cocha Totora with the observational techniques already established; · To quantify limnological characteristics and relative fish abundances in at least 9 lakes of the Manú (3 lakes with turbid water and no underwater aquatics; 3 with clear water and underwater aquatics; 3 with year-round connectivity to the river); · To quantify otter prey preferences and consumption by directly observing prey captures; · To initiate a mark-recapture program on boca chico and bujurki, and to estimate the sampling effort required for a full population monitoring program in Cocha Cashu and Cocha Totora; · To survey local people about otter interactions and contact with domestic dogs in the Manú and Alto Purús. Methodology: Document Hunting Behavior, Prey Preferences and Hunting Success Rates in Various Microhabitats: Cocha Cashu otters are already partially habituated due to their long association with researchers at the station. Otters resident on Cocha Cashu and Cocha Totora will be followed and observed from small portable kayaks for whole days, three days a week. Hunting activity will be documented in 3-minute intervals. This procedure was field-tested in 1999/2000, and was found to be a workable way to quantify hunting success. In each time period recorded, the exact location, hunting method and hunting success will be noted. When prey is captured we will try to identify the species and approximate the size. Video taping will be used to aid in prey and size identification and handling time. Compare Limnology and Fish Community Structure in Differing Lake Systems: In 1999, fish community sampling with gill nets proved unsuccessful. Specifically, piranha were attracted in large numbers to fish struggling in the nets, strongly biasing the sample. In 2001, we propose to sample the fish community of Cocha Cashu and other lakes with electrofishing methods. Electroshock methods work well in shallow waters, and all known lakes in the region with the exception of one (Cocha Salvador, which is 4-5m in depth) are characteristically = 2m in depth. An electrofishing unit donated by Idea Wild will be used. The electrofisher model acquired allows the investigator to vary the waveform of the electrical pulses propagated in the water. This capability allows us to conduct accurate sampling in a number of modes, and to minimize mortality. Shock pulses can be tuned optimally for particular species, or can be produced in a form that is relatively non-selective. While we cannot hope to estimate viable population size estimates for any particular species without a long-term mark-recapture study, we should be able to detect relative variations in the fish fauna across lakes of different types in the current field season. Fish of species of particular importance to otters will be marked for future studies on population size and variation. Marking will be conducted by injecting small quantities of visible implant elastomer (VIE) in suitable locations on the fish. VIE is a cost-effective method for marking thousands of individuals with high retention rates (Bailey et al., 1998). In 2000, the research team found that VIE markings could be reliably injected and retained at a rate of approximately 93%. Basic limnological data will also be collected, including oxygen availability, pH, turbidity, and conductivity. We envision that we will visit 9 lakes during the season, with each lake being visited once at the beginning of the field season, once in the middle, and once at the end of the field season. Statistical analysis used will be Non-Metric Multi-Dimensional Scaling (NMDS), and Canonical Correspondence Analysis (CCA) incorporating environmental variables. Interactions with Local People. It is unknown to what extent native residents of the park persecute giant otters either directly, or indirectly through their keeping of domestic dogs. It is known that Machiguengas consider otters a pest, and a competitor for fish, and may harass and even occasionally kill them. It is not known how other indigenous groups in the Alto Purús view otters and interact with them. While conducting the proposed trips within the Manú and the Alto Purús sampling domestic dogs for disease, we will also have the opportunity to query local people about attitudes and practices with respect to giant otters and the short-eared dog. In-depth interviews with local leaders, women, and NGO's that work on behalf of indigenous communities will help guide our understanding and recommendations for activities that could be implemented to minimize human and wildlife conflicts. Short-Eared Dog Reseach Little is known about South America's elusive medium-sized rain forest canid, the short-eared dog (Atelocynus microtis). Its current status with IUCN is "DD" or "Data Deficient". Apart from two published accounts of field encounters with a single individual in Brazil and another in Colombia (Peres, 1991; Defler and Santacruz, 1994), most of what is known about the species comes from scattered museum specimens. In the 1960's the short-eared dog was reported as being relatively common in Perú (Grimwood, 1969 and Gardner, pers. comm.). However, despite intensive field research in the 70's and 80's, the species was never seen in the Perúvian Amazon until 1990 (Terborgh et al., 1984; Jason and Emmons, 1990; Woodman et al., 1991; Pacheco et al., 1993; Pacheco et al., 1995). Even Emmons (1984; 1987; 1988), who carried out long-term projects censusing and trapping ocelots and other mammals at Cocha Cashu (with the same kind of traps that Gardner used), never saw or trapped this species (Terborgh, pers. comm.). For whatever reason, the species appears to have temporarily vanished from the region between 1970 and 1990. Over the last decade, sightings of the short-eared dog have again become relatively frequent in the Manú. Since 1990, at least twelve encounters with A. microtis have been reported by researchers in Madre de Dios, mostly in the vicinity of the Cocha Cashu Biological Station. The first ecological data about this species was obtained in 2000 by Dr. Leite (Leite, 2000, Leite et al. 2000 and Leite and Williams, in press). In the fall field season, we confirmed the existence of a population of the short-eared dog and did a preliminary study of its diet. Five different individuals were found in a 10 km transect, and the skeleton of one dead juvenile was collected. In the coming field season, efforts will concentrate on trapping short eared dogs to test them for exposure to CDV and CPV and for a radio-tracking study to determine their home range, activity pattern, and habitat use. The information collected will be incorporated into both a Manú-region and an Amazonia-wide conservation plan to ensure the long-term survival of the species. Methodology: Habitat use and density: Direct (visual and radiolocations) and indirect (tracks, scats, bones, corpses) observations will be used to assess the habitat used by the short-eared dog. To estimate population density we will use line or strip transects, also known as King or Keller transects (Glanz, 1982), concentrated in the places where the animal has been sighted over the last year. During the censuses, two people will each walk one half of a 10-km transect at an average speed of 1 km/h, stopping frequently to scan for animals. For each individual of A. microtis encountered, we will record the time, the location, the distance of the animal when first detected, the perpendicular distance from the animal to the trail, feeding and other behavior. Both researchers will carry tape recorders to record any calls during encounters. Scats and tracks will be collected and registered along the transects. Since the area of the transect is known, the overall population density can be estimated from the number of encountered individuals. The habitats used by the short-eared dog will be identified in a satellite image and by extrapolation of similar habitats types the population will be estimate for the Manú National Park and Alto Purús Reserved Zone. Prey species : When scats are encountered along the transect near the short-eared dog tracks, they will be collected and oven-dried at 60oC. If tracks are not found around the scats, they will be collected and examined for the presence of the target animal's hair. Hair from scats will be compared with hair from Lima and Chicago museum specimen of A. microtis. For those scats positively identified as belonging to A. microtis (i.e., if some of the hair in the scat or tracks matches that of A. microtis), diet will be studied via analysis of feces. Fragments of bones, hairs, feathers, scales and vegetable matter will be compared with my own reference collection and those of Perúvian museums. To compare the diet in different sites, I will use the Sorensen similarity index (Krebs, 1989), which varies from 0 to 1, and is based on the presence or absence of prey species in the diets. The food niche breadth will be calculated using the "B" index (Levins, 1968) and the value of the food niche breadth standard calculated using the "Bsta" index (Colwell & Futuyma, 1971). Bsta varies from 0 to 1 and makes possible the comparison between studies with different numbers of prey categories. A Bsta value of 1 means that all categories of prey were used in equal proportion, while a value near zero means that one or a few categories were used with high frequency and many with low frequency. Capturing the short-eared dog: In order to test the short-eared dogs to CDV and CPD, radio-collared them to find their contact rates with domestic dogs, and to study the habitat use and home-range, we need to trap them. To do so, we will be using 10 live or box traps and 24 snares along the 10 km transect were the species was found last year. All trapped animals that are not A. microtis will be immediately released. Any captured A. microtis will be injected by blowgun with Zoletil® (3 mg/kg; active ingredients are 1.5 mg tiletamine + 1.5 mg zolazepam/kg). While the animal is anesthetized, I will take data on its body weight and measurements, sex, and reproductive status. A small blood sample (5 ml) will be collected from the radial vein with a plastic hypodermic syringe for later hematological analysis. A photograph of the animal will also be taken. An imprint of both the front and back footpads will be preserved. Two animals will be fitted with a radio collar. Once this work is complete, trapped animals will be held in a darkened covered trap until they recover from anesthesia, at which point they will be released at the same place they were trapped. Because I am a trained veterinarian and have several years of experience trapping large mammals, I can carry out this work with two field assistants. If animals are trapped and successfully collared, a trained field assistant will track the individuals on a daily basis using a mobile antenna, for two years. These will be the first data on the species' daily activities, distance moved over time, home range areas and territoriality. All telemetry data will be plotted in a satellite image. RESEARCH COMPONENT 2: Investigating Threats - Domestic dogs in Manú National Park and Alto Purús Reserved Zone Canine Distempers virus (CDB and CPV) typically infect domestic dogs but were diagnosed as responsible for decline of wild carnivores populations, as black-footed ferrets (Mustela nigripes) and lions (Panthera leo - Roelke-Parker, 1996; Williams et al. 1988). Scheik and Staid (2000) did a preliminary study in 1995 on the occurrence of these diseases in domestic dog in Manú National Park. Their results show that domestic dogs have been in contact with these diseases. Current park policies do not address the threat domestic dogs pose to wildlife, and neither vaccination programs nor educational programs have been initiated. Based on the potential risk that the Manú wildlife community faces, we feel that this is an urgent unmet conservation need, but one that requires better data to convince park authorities to act. We propose to re-sample the Manú population of domestic dogs and, importantly, to expand the sampling to the Alto Purús zone during our expedition into this reserve (see below). We plan to sample domestic dogs in Machiguenga communities, and also to take blood samples from any short-eared dog that is captured during the trapping and radio-collaring study at Cocha Cashu. Dr. Renata Leite, an experienced vet with many years experience handling zoo animals and wild animals, will personally handle the domestic and wild animals during the simple procedure required. To do so, we will collect 10 ml of blood sample from either jugular or ulnar vein of each dog and will separate the serum of each sample and will preserve them for lab analyses. Local communities, domestic dog presence and the results of viruses test will be plotted (overlay) the satellite image were the habitat for both giant otter and short-eared dog were plotted. During our search for domestic dogs in Manú National Park and surrounding reserved zones, we expected to record wild animal uses by local people, logging, mining and any other activities that could represent any potential threat. All results found will be plotted as a overlay in the satellite regional image. Expedition to the Rio Purús Given the low population densities of the giant otter and the short-eared dog within the Manú, the possibility of long-term survivorship of these animals will depend upon contact with populations from outside the park. Thus, it is important to learn much more about the status of these animals in the new Alto Purús reserve. In the coming year, we plan to mount a 2-month expedition into the protected area to learn more about the animals' occurrence there. We plan to focus on entering the upper Rio Purús, where an old airstrip still exists at the missionary outpost of Balta. From there we will hire boats with guides of the Cashinahua tribe, and travel as far into the reserve as is deemed safe by our local guides. The presence of uncontacted tribes does pose a safety threat in some parts of the reserve. In the areas entered, we will conduct river-based and land-based transects to search for otters and sign of the short-eared dog. Standard survey methods will be used, and for giant otters that are observed, video taping will be used to capture photos of each individual's signature throat markings. This method is in accordance with other survey work conducted in the Manú National Park by the Frankfurt Zoo. In addition, we will enter all indigenous settlements to attempt to take blood samples from domestic dogs living in the reserve to test for disease occurrence. Availability of prime habitat The assessment of prime habitat will be made by plotting in a local map the field locations obtained by direct and indirect observations. This map will be compared with the most recent satellite images, where the data will be plotted using the GIS program Arc View.