Mammal ecology topics

IL240 Influence of habitat structure on small mammal assemblage composition in Sulawesi, Indonesia

(Weeks 2 - 8; need to have completed IL001)

Most of Sulawesi’s remarkable diversity of endemic mammals are small, terrestrial rodents and insectivores, but as with many Sulawesi vertebrates, very little is known about them.  The general aim of this project will be to examine relationships between forest habitat structure and small mammal assemblage composition (rodents and shrews). As part of the above overall research and monitoring project, you will assist the ecologists in the collection of mammal data at approximately 45 sampling sites stratified into areas with differing levels of forest disturbance and habitat type.  The surveys at each site will be completed using pitfall traps and baited small mammal traps, which will be checked daily.  These data could be supplemented by taking habitat structural measurements at each site, using standard measurements previously developed and the spatial patterns of assemblage composition and individual species with respect to habitat characteristics determined.  This project could focus on either the pitfall data, or mammal trap data.  Alternatively, this project could focus upon an evaluation of comparative trapping methods.

further information

IL241 Factors affecting bat assemblage composition in lowland forests of Indonesia (Weeks 2 - 8; need to have completed IL001)

The bat research programme has been capturing bats around Labundo village and at three forest sites since 2000, with most efforts concentrated in the Kakenauwe Reserve. Forest interior bat species are adapted to life in dense vegetation and are therefore expected to be especially vulnerable to the effects of logging, rattan extraction and other types of forest disturbance. The best student bat projects should be able to use data from every animal captured. Potential projects could focus on the determinants (e.g. species, sex, age, reproductive status) of individual parasite load, the application of acoustic methods to characterise activity levels of some key species, or – for students with a few weeks on site – projects to characterise the community. Students should contact the bat staff early to discuss equipment and logistics.

IL242 Ecology of the top mammalian predator in the forests of Sulawesi, Indonesia (Weeks 2 - 8; need to have completed IL001)

The Malay civet is the principal (possibly the only) mammalian carnivore in the Lambusango forest. Very little is known about the population dynamics of civets or indeed any other rainforest carnivore, and this project provides an excellent study system to investigate population processes in this ecologically important group.  The project builds on a 7-year data set working on a marked population of civets occupying a 4 km² area of forest.  So far, more than 110 civets have been tagged, and the project is beginning to build up a unique picture of civet ecology. Dissertation projects can add to this long-term study by repeating annual surveys which involves a 7-8 week capture-mark-recapture (CMR) study to estimate population density and to identify which animals are still remaining in the population from previous surveys.  Projects could focus on comparing observed population metrics (e.g. density, age structure, average body mass) of Malay civets in the Lambusango forests with those published in studies from elsewhere in their range under different ecological conditions, or even with other similar-sized mammalian carnivores living in contrasting habitats.  Project students can also examine elements of civet survival (e.g. what attributes of a civet make it more likely to be found in the population in the following year?) using data from the previous survey season (data includes physical attributes such as age, size, and sex and behavioral attributes such as the number of times it was captured within a season).  This project provides experience in how to carry out a CMR study for a small carnivore and how to estimate population density from CMR data using CMR models (e.g. CAPTURE, Jolly-Seber).  These are standard tools in wildlife ecology and a must for any budding wildlife biologist.

further information

IL243 The ecology of Sulawesi cuscus, arboreal marsupials of Sulawesi

(Weeks 2 – 8; need to have completed IL001)

The Australasian radiation of marsupials reaches its western range limit in Sulawesi.  Two species of cuscus, arboreal possums, are known from the Sulawesi archipelago.  One of these species, the bear cuscus (Ailurops ursinus), represents a monotypic genus due to its relatively primitive characteristics and distinct morphology that set it apart from other cuscus species.  At present, the ecology of Sulawesi cuscus species are poorly understood with only one published study from mainland Sulawesi.  The bear cuscus is a folivore and is predominantly diurnal whereas the dwarf cuscus (Strigocuscus celebensis) is frugivorous and predominantly nocturnal.  Both species occur on Buton Island and the bear cuscus is frequently seen foraging in the forest and adjacent farms and plantations.  We plan to undertake a study of the ecology of cuscus within and adjacent to the Lambusango forest reserve on Buton Island, Sulawesi. Population densities, feeding ecology and habitat preferences of the species will be established using a combination of observational and Distance sampling techniques.  Transects will be established and walked repeatedly over a number of weeks in a variety of habitats within and outside of the reserve.  Basic ecological data will be recorded, characteristics of trees and surrounding habitat structure, location in tree and activity, to improve our understanding of what factors influence distribution and abundance of these species.

further information

IL244 Sleeping site selectivity by the Buton Tarsier, Sulawesi, Indonesia

(Weeks 2 - 8; need to have completed IL001)

Tarsiers live in small groups which emerge from their sleeping sites in trees, tree falls, thickets or holes in the ground at dusk and return at dawn. Previous work on Buton has shown that tarsiers are using sleeping sites, including strangling figs, rock crevices and vine tangles for sleeping sites.  By observing tarsiers returning to and emerging from these sites it is possible to map them, measure group size and assess the key features of sleeping sites. Occurrence of sleep sites will be assessed, using dawn triangulation/quadrangulation sampling over 0.75 Ha plots.  Colony sizes will be estimated using dusk sleep site monitoring.  Quality and availability of potential roost sites will be evaluated in the vicinity of each plot. This topic could be developed into a series of research questions. For example, how does the quality and availability of sleep sites influence tarsier group size and group density?  Alternatively, how is availability of sleep site resources distributed with respect to landscape, habitat variability and human disturbance?  Techniques could be explored to assess relative abundance/biomass of invertebrates, perhaps using torch light traps with sticky cards installed adjacent to sleeping sites to yield indices of tarsier food availability at the various sites.

IN245 Density and distribution of Sulawesi megafauna, Indonesia

(Weeks 2 - 8; need to have completed IL001)

Four of the largest and most charismatic vertebrates that are present in the forests of Buton are endemic species of conservation concern: the anoa, a highly endangered dwarf buffalo, Sulawesi wild pig, an endemic pig species that can be a crop pest, the Buton macaque, a poorly studied endemic primate, and the knobbed hornbill which is often captured for the pet trade. A survey team will assess the abundance of anoa, macaques and hornbills on 3km transect lines at the forest node camps.  Various projects could be completed as part of this topic. Anoa and pig distribution could be assessed using patch occupancy analysis and hornbills and macaques through distance sampling. Forest structure could be assessed through standardized sampling at points along the transects and the distribution of the target species related to forest structure and levels of human activity.

further information 

HM246 Distribution patterns of frugivorous bat communities in cloud forest, Honduras: Abundance, ecology and contribution to forest regeneration

(Weeks 2 - 8; need to have completed HM001)

The bat population at Cusuco National Park has been monitored between June and August each year since 2006 using mist net surveys. For each bat captured in the net, the species, sex, reproductive status, weight and wing morphology is recorded before the bat is released. As most insectivorous bats can detect mist nets using echolocation and avoid capture, the Cusuco bat surveys are focussed on frugivorous bats (which tend to rely on their keen sense of smell to search for food rather than echolocation). Cusuco National Park is an incredibly complex landscape with huge variations in elevation, temperature and rainfall resulting in a wide range of habitats. Standardized mist net surveys are conducted at a number of surveys sites across the park, and each mist net site is associated with a habitat plot. In each habitat plot, tree type (pine or broadleaf), DBH, tree height, and canopy cover are recorded. One project could therefore investigate the abundance and diversity of frugivorous bats in relation to habitat variables. In addition, the characteristics of the bats caught (weight and wing morphology) could be investigated in relation to forest structure. Using the long term data set, trends in abundance and proportion of bats caught that were reproductively active could be investigated to provide insight into the health of the bat population.  Projects could also investigate dietary preferences of frugivorous bats by identifying seeds contained in faecal samples.

SP247 The use of behavioural studies to assist with management decisions for a large elephant population in a small private game reserve, South Africa

(Weeks 3 - 8) 

Between 1979 and 2001, over 800 elephants were reintroduced to over 58 reserves in South Africa, Pongola Game Reserve being one of them.  The elephant carrying capacity for Pongola has been estimated at 37 animals, but the numbers are at over 75 individuals. In 2008, the Disney Corporation provided funding for vasectomies of the bull elephants in an attempt to cap the population. Although the vasectomies should control the elephant population in the long-term, the landowners at Pongola are still concerned about the immediate damage to vegetation caused by the high density of elephants.  The main aims of the behavioural monitoring project are to provide data to inform management policy with regards to the elephant population, and within this to determine the utility and efficiency of potential management options.  In order to reduce the pressure from the population size, the options are to expand the range available to the elephants by dropping fences with adjacent reserves or to reduce the population size, the latter being an emotive and controversial issue.  If range expansion is not possible, translocation is an alternative, and determining which individuals to translocate requires careful consideration of population association patterns and the possible consequences to the remaining population. Consequences may include the potential for delinquency in young males where musth behaviour is not suppressed by older bulls, or stress or disruption to herd dynamics if socially cohesive groups are split.  The data collected will enable the analysis of sexual and­ behavioural maturity in the bulls from juvenile to adult, and social interactions between males, yielding information on the behavioural process of musth suppression and the appropriate age for vasectomies to take place.  The data will also allow the analysis of association patterns to identify herds and individuals that form cohesive social groups.  The Pongola elephants also exhibit unusual association patterns, with bulls spending the majority of their time associating with the herds.  In addition to this having an influence on management decisions, a further aim of the project is to investigate the underlying behavioural and ecological influences on habitat use, fission-fusion dynamics and association patterns.   Pre-existing datasets on ranging and foraging strategies combined with current on-going behavioural and ecological data collection will allow the investigation of the drivers of observed patterns in spatial ecology and behaviour.

further information

SP248 Calculating the carrying capacity of the Pongola Reserve for elephants, South Africa

(Weeks 3 - 8)

The most pressing issue on the Pongola reserve is the size of the elephant population, due to their perceived impact on the habitat.  The perception among landowners and managers is that monospecific habitat now dominates where diversity used to be greater, and there are an abundance of areas with a high proportion of dead trees and trees that have been bark-stripped.  The blame for perceived habitat degradation is placed unequivocally on the elephants, despite the fact that the effect of elephant on habitat remains a contentious issue.  In theory the elephant population growth has been stopped with the implementation of contraception, but this still leaves the population at approximately twice the estimated carrying capacity of 37.  This number has become a management standard for the Reserve, influencing population management decisions, despite the calculation being a broad generalised measure of maximum density, most likely based on data from the Kruger National Park.  Data is urgently needed to describe how the habitat in Pongola is influenced by elephants and other browsing herbivores in terms of vegetation structure, species composition and distribution, as well as quantification of damage.  Measures of habitat in conjunction with data on habitat use and foraging in the elephant population can help to re-establish a reference carrying capacity based on sound ecological data that is locally relevant.   This data set also has the potential to be used to establish a scale of impact levels (e.g. based on the Walker scale), categorising observed damage and perhaps linking this to estimates of an elephant population size at which the maximum acceptable level of damage would not be exceeded.   

further information

SW249 Factors affecting the population size and distribution of large mammals in the Waterberg Biosphere Reserve, South Africa

(Weeks 4 - 8; need to have completed SW003)

The Waterberg Biosphere area contains nutritionally poor grass species that hold a minimal grazing value for herbivores, which results in relatively low carrying capacities of these species. Consequently, it is difficult to support the population densities demanded by eco-tourist visitors to reserves, particularly if reserves want to maintain sufficient numbers of large carnivores, such as lions that prey upon these herbivores.  As tourism is a massive driver behind the recent boom in land conversion from farmland to conservation in the region, it is important to understand what factors are affecting the population dynamics and distribution patterns of herbivores so that populations can be effectively managed in the future.  Welgevonden contains a wide range of herbivore species including elephant, rhino, zebra, wildebeest, hartebeest, kudu, gazelle and other antelope. The herbivore population is monitored each month using vehicle-based surveys throughout the reserve road system and annual helicopter census. Between June and August each year, additional vehicle-based surveys are conducted each day along seven 10km long transect lines across the reserve. During the daily and monthly surveys, all visual encounters with the herbivores are recorded, noting the GPS location of the animal, the species, number of individuals, age-sex class of each individual, and habitat type. Daily records of temperature and rainfall are also recorded and GIS maps showing the distribution of vegetation types in the reserve have been produced. In addition to mammal surveys, students will assist with habitat surveys conducted at 40 different sites within the reserve as a means of assessing fire and elephant impact on vegetation over time. Data collected during the 2012 field season may be added to these long-term data sets to produce a range of different projects focusing on individual species or multiple herbivore species. For example, GPS points of target species can be uploaded on to GIS vegetation maps to investigate distribution patterns and habitat preferences of the species. Game count data may be used to investigate changes to species abundance and survival rates over time in relation to environmental variables (temperature and rainfall) or changes to habitat caused by fire and elephants. Alternatively, comparison based studies could investigate differences in population density estimates calculated from DISTANCE sampling during vehicle surveys and total population counts from helicopter surveys.  Students working at Welgevonden will spend half of their time in the field while the other half will be spent in camp doing data entry, a lecture series on African Wildlife Conservation, independent work on dissertations, and assisting with creating new data sets for the reserve management such as elephant and rhino individual ID kits.

further information

SW250 Estimating the impact of elephants on habitat in Welgevonden Reserve, South Africa

(Weeks 4 - 8; need to have completed SW003)

This project has similar aims to SP247 in Pongola Reserve.  In Welgevonden, the elephant herds have been tracked four times a day from satellite GPS collars since 2008 and these data can be used to produce much more accurate contours of elephant usage in the reserve.  Measurements of grass volume, percentage cover of shrubs, size structure of trees and levels of damage of each shrub and tree using the Walker scale.  Lower intensity surveys concentrating on just Walker damage levels could be completed from 1km transects in the areas of different elephant usage.  These data could then be analysed to produce contours encompassing areas of the reserve with equal levels of damage.  Welgevonden is a high veld reserve with lower productivity levels than the Pongola reserve and the numbers of elephants that an area could sustain to stay below pre-determined acceptable levels of damage may well be different to those in a lowland reserve. Students working at either of these sites will spend half of their time in the field whilst the other half will be spent in camp doing data entry, a lecture series on African Conservation and independent work on dissertations.

further information

PP251 Population trends and habitat preferences of Pink and Grey River Dolphins in the Peruvian Amazon

(Weeks 2 – 8; need to have completed PP001)

The pink dolphin (Inia geoffrensis) and grey dolphin (Sotalia fluviatilis) are endemic to the Amazon rivers and function as indicator species for the general health of aquatic habitats. Dolphins make an excellent indicator species, because they rapidly move out of polluted or degraded habitats and, in turn, quickly indicate changes in the condition of aquatic systems. The dolphins are also easy to count and observe, since they frequently surface and are large-bodied and very distinctive. The river dolphin population in Pacaya Samiria has been monitored for several years using fixed-width transects along rivers, lakes and channels using small boats. During these surveys, all dolphin encounters are recorded noting the species, number of individuals, habitat in which the dolphins were seen and the dolphin behaviour. Dissertation topics could examine the health of the aquatic systems in the Peruvian Amazon by evaluating population trends of the two species of river dolphin overtime, or could focus on habitat, behaviour and group size differences between the two species.

further information 

MN252 Occupancy modelling for nocturnal lemurs, carnivores and bush pigs with camera traps in Madagascar

(Weeks 1 - 6)

In the Mahamavo dry forest, most diurnal and nocturnal lemur species are easily seen by teams walking sample routes by day and at night.  However, two nocturnal species, the fork-marked lemur Phaner pallescens and fat-tailed dwarf lemur Cheirogaleus medius, are seldom seen by the field teams.  Additionally, the forest is within the range of aye-aye Daubentonia madagascariensis, so it could occur in Mahamavo, although it has not yet been detected.  Every year, there are a small number of sightings of carnivores including the Fossa Cryptoprocta ferox, Falanouc Eupleres goudotti, and Ring-tailed mongoose Galidia elegans, but not enough observations to infer their distributions or population sizes.  The dry forests are also home to bush pigs Potamochoreus larvatus.  For cryptic species such as these, a network of camera traps is the best way to gain reliable data on distributions, densities and trends through time, without needing to trap animals.  Students choosing this project would help design the spatial and temporal sampling strategy for the cameras, select ‘best’ sites at the local scale to install them on the ground or in trees, visit the cameras to change SD cards and look at the photos and then undertake analysis of the detection histories of each species recorded by the cameras at each site using occupancy models.  This powerful approach allows occupancy of sampling units (camera locations) over the course of the season to be estimated taking account of the detectability of each species.

further information

MN253 Development of indicators for monitoring bat populations in Madagascar using bat detectors and automatic analysis of sonograms

(Weeks 1 - 6)

The dry forests in Mahamavo support a wide range of microchiropteran bats which use echolocation.  These species can be monitored by constant effort mist-netting, but this requires a huge amount of sampling effort to be able to reliably detect trends in bat populations.  An alternative approach is to walk sample routes and along the edges of the forest and wetlands at dusk with a time-expansion bat detector connected to a digital sound recorder and GPS.  The sound recording can then be viewed as a sonogram and automatically compared with recordings of known species to identify how many individuals of each species were present on a sampling occasion and their location.  It would also be possible to validate results from sound analysis with capture data from mist nets and contrast the two approaches.