No abstract

An extensive study of vegetation changes as a consequence of ¢re and grazing pressure and their e¡ect on small mammal populations inside the Maasai Mara National Reserve, Kenya, was carried out during MayĴ une 1997. Comparison of vegetation maps from 1979 and1998 suggested that vegetation in 46% of the Reserve area converted from shrubland to grassland, possibly as a result of ¢re and grazing pressure.We tested the hypothesis that in areas with high ¢re and grazing impact the population of small mammals was negatively a¡ected. A low density of rodents was recorded in all habitats except in areas of human activity, where arti¢cial resources are constantly present. Capture e¡orts were unsuccessful in grasslands. Our results con¢rm those of Norton-Gri¤ths (1979) and Dublin (1995), i.e. that ¢re and grazing pressure impact the vegetation of the Serengeti^Mara ecosystem and limit the natural regeneration of woodlands. This indirectly a¡ects the small mammal community, which is limited in its long-term establishment.

Majella National Park in central Italy is known to be an endemic-rich area, but distributions of its endemics have not been comprehensively studied. Endemics with 10 or more records and spatial uncertainties at 55 km were extracted from the Central-Apennine floristic geodatabase and the MNP Seed Index. Nine environmental predictor layers were prepared at 90 and 30 m resolution. A stepwise Maximum Entropy (Maxent) model was generated per endemic to achieve the most parsimonious result at an area under the curve 4 0.8. Arctic-alpine elevation, edaphic barrens and low open-vegetation, individually or in pairs, were found to be predictive for endemics. Forty-eight endemics, 10 of which exclusive, were recorded and Maxent-predicted for the Majella massif. Subsets of 38 endemics were recorded on other mountains in proportion to their arctic-alpine area, thus conforming to the Island Theory. Maxent confirmed its strengths also at fine resolutions and, in addition, showed to be robust across predictor layers at both resolutions. A linear species-area relationship appeared superior to the Maxent model in predicting the number of endemics per arctic-alpine ''island''. Our findings suggest the need for a proactive management of the botanical biodiversity contained in the alpine and montane barrens and low-open vegetation.

Crop-raiding elephants affect local livelihoods, undermining conservation efforts. Yet, crop-raiding patterns are poorly understood, making prediction and protection difficult. We hypothesized that raiding elephants use corridors between daytime refuges and farmland. Elephant counts, crop-raiding records, household surveys, Bayesian expert system, and least-cost path simulation were used to predict four alternative categories of daily corridors: (1) footpaths, (2) dry river beds, (3) stepping stones along scattered small farms, and (4) trajectories of shortest distance to refuges. The corridor alignments were compared in terms of their minimum cumulative resistance to elephant movement and related to crop-raiding zones quantified by a kernel density function. The ''stepping stone'' corridors predicted the crop-raiding patterns. Elephant presence was confirmed along these corridors, demonstrating that small farms located between refuges and contiguous farmland increase habitat connectivity for elephant. Our analysis successfully predicted elephant occurrence in farmland where daytime counts failed to detect nocturnal presence. These results have conservation management implications.

The conversion of closed forest (CCF) in Carrasco Province, Bolivia, was monitored using a series of four midresolution satellite images from 1986 to 2002. The conversion of forests into nonforests from 1986 to 2002 was 1.5% annually. Inclusion of conversions into open forest doubles the annual CCF rate to 3.1%. Five predictors of CCF were tested in a spatial model: land tenure regime, distance from roads, distance from settlements, topography, and soil suitability for farming. Only three out of the five predictors tested were found to be reliable predictors of CCF: land tenure regime, distance from roads, and distance from settlements. University reserve and indigenous land show substantially less CCF than national park and untitled land. In addition the spatial model shows that the greater the distance of forest from roads or settlements, the less CCF. Topography and soil suitability for farming lack predictive power for CCF and are therefore excluded from the spatial model.