Graduate Thesis Or Dissertation


Landscape features affecting genetic diversity and structure in East African ungulate species Public Deposited

Downloadable Content

Download PDF


Attribute NameValues
  • Habitat loss and fragmentation is a crisis affecting wildlife worldwide. In Tanzania, East Africa, a dramatic and recent (<80 years) expansion in human settlement and agriculture threatens to reduce gene flow among protected areas for many species of large mammals. Wildlife linkages can mitigate population isolation, but linkage designs lacking empirical justification may be controversial and ineffective. Connectivity conservation requires an understanding of how biogeographic factors shaped gene flow prior to habitat loss or fragmentation, however the history of interaction among populations is rarely known. The goal of my study was to provide context for connectivity conservation in central and southern Tanzania by identifying landscape features that have shaped gene flow for three ungulate species with different dispersal capabilities. I investigated historical patterns of connectivity for Maasai giraffe (Giraffa camelopardalis tippelskirchi), impala (Aepyceros melampus), and eland (Tragelaphus oryx) by estimating genetic structure among four to eight protected areas per species. Genetic structure changes very slowly among large populations and thus is likely to reflect historical processes instead of recent anthropogenic influences. I collected noninvasive DNA samples and generated microsatellite genotypes at 8 to 15 loci per species, then estimated genetic diversity metrics (allelic richness, AR, and expected heterozygosity, H[subscript E]) for each population (defined by reserve). I also calculated genetic distance (F[subscript ST] and Nei's unbiased genetic distance, D[subscript hat]) and an estimate of gene flow (Nm) between all population pairs for each species. To elucidate the possible causes of genetic structure between these populations, I tested for isolation by distance and isolation by resistance based on a suite of biogeographic factors hypothesized to affect gene flow. To do this, I created GIS-based resistance surfaces that assigned different costs of movement to landscape features. I created one or more resistance surfaces for each hypothesis of landscape effect. I used circuit theory to estimate the cumulative resistance between each pair of reserves for each weighting scheme, and then performed Mantel tests to calculate the correlation between these resistances and the observed population pairwise genetic distances (D[subscript hat]). I chose the optimal resistance model for each species as the model that was most highly correlated with observed genetic patterns. To verify that the correlation of resistance models with genetic distance was not an artefact of geographic distance, I performed partial Mantel tests to calculate correlation while controlling for the effect of geographic distance. Finally, I compared historical gene flow patterns to the distribution of contemporary human activity to predict areas that are at risk of a loss of connectivity. Indices of genetic diversity were moderate for all three species and comparable to previously reported values for other savannah ungulates. Diversity (both H[subscript E] and A[subscript R]) was highest in eland and lowest in giraffe for these populations, and was not consistently correlated with reserve size as has been reported for other species in East Africa. Although patterns in genetic distance were broadly similar across all three species there were also striking differences in connectivity, highlighting the importance of cross-species comparisons in connectivity conservation. At this scale, resistance models based on slope strongly predicted population structure for all three species; distance to water was also highly correlated with genetic distance in eland. For all three species, the greatest genetic distances occurred between populations separated by the Eastern Arc Mountains, suggesting that the topography of this area has long acted as a barrier to gene flow, but this effect is present in varying degrees for each species. I observed high levels of historical gene flow between centrally located populations (Ruaha National Park and Rungwa Game Reserve) and those in the southwest (Katavi National Park and Rukwa Game Reserve). Although human settlement in this area has been low relative to other areas, the connection between the Katavi/Rukwa and Ruaha ecosystems may be threatened by increased human activity and warrants conservation. High levels of historical gene flow were also seen between reserves in the northeast (Tarangire National Park, Swagaswaga Game Reserve) and the central and southwest populations. These connections appear highly threatened due to current land use practices, and may have already suffered a loss of gene flow. Field surveys in the lands surrounding the northeastern reserves are needed to quantify current levels of connectivity and determine whether corridors could be established to maintain or restore gene flow with other reserves.
Resource Type
Date Available
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Committee Member
Academic Affiliation
Non-Academic Affiliation
Rights Statement
Peer Reviewed
Additional Information
  • description.provenance : Made available in DSpace on 2012-04-03T23:41:55Z (GMT). No. of bitstreams: 1 CrowhurstRachelS2012.pdf: 1920194 bytes, checksum: 5c93b0a1162806b27b4487746e5be8a3 (MD5) Previous issue date: 2012-02-27
  • description.provenance : Submitted by Rachel Crowhurst ( on 2012-04-02T04:52:23Z No. of bitstreams: 1 CrowhurstRachelS2012.pdf: 1920194 bytes, checksum: 5c93b0a1162806b27b4487746e5be8a3 (MD5)
  • description.provenance : Approved for entry into archive by Julie Kurtz( on 2012-04-03T19:56:46Z (GMT) No. of bitstreams: 1 CrowhurstRachelS2012.pdf: 1920194 bytes, checksum: 5c93b0a1162806b27b4487746e5be8a3 (MD5)
  • description.provenance : Approved for entry into archive by Laura Wilson( on 2012-04-03T23:41:55Z (GMT) No. of bitstreams: 1 CrowhurstRachelS2012.pdf: 1920194 bytes, checksum: 5c93b0a1162806b27b4487746e5be8a3 (MD5)



This work has no parents.

In Collection: