Ph.D Thesis

Ph.D StudentKent Refael
SubjectRelationships between Environmental Parameters and Species
Composition Patterns and their Implications for
Biodiversity Surrogates - a Multiple
Scale Study
DepartmentDepartment of Civil and Environmental Engineering
Supervisor PROF. Yohay Carmel
Full Thesis textFull thesis text - English Version


Relationships between species and environment are a central issue in ecology. Understanding these relationships can contribute to biodiversity conservation. However, there is a lack of data to conduct large scale species composition analyses, needed for large scale conservation planning. This work attempts to promote scientific tools of biodiversity conservation along two lines. The first principal objective of this work was to quantify the variation in the relationships between species composition and environmental determinants as a function of spatial scale. The main problem was data availability. At large scales, the only available data is presence-only data, which is considered inappropriate for such analyses. A simulation study was conducted, with virtual species based on real environmental data and true species occurrence data from the Global Biodiversity Information Facility, to test whether presence-only data are sufficient for large scale species composition analyses. Results of the simulation tests allowed me to commence with the next step which was analyses of the environmental determinants of terrestrial mammal composition in the contiguous USA, at varying spatial scales. Climatic parameters, such as temperature and precipitation, were the most important environmental variables determining species composition. Land-use - land-cover parameters were also relatively influential at small and intermediate scales; however their effect decreased at larger scales. Topography and primary productivity were of lesser importance at the scales studied, and their effect further decreased with increasing scale. Second, methods to produce biological and environmental surrogates for biodiversity were compared, and new measures of surrogate efficiency were proposed. Results indicated that the most efficient classification method for biodiversity surrogates production is Ward's method. Additionally, a new type of surrogates was developed, termed here the bio-environmental surrogates, combining biological and environmental data. There surrogates improved the representativeness of surrogates, representing almost all target species, regardless of the specific approach taken to produce them. The two lines of progress achieved by this work may be combined to enhance the abilities of conservation scientists and practitioners to protect biodiversity at large scales. In light of the extremely high species extinction rate, estimated at ~100 - 1000 times higher than the background rate, and in light of latest climatic changes and habitat loss through land-use alterations both processes directly or indirectly related to human activity in past century, urgent actions are needed in order to protect the remaining biodiversity. This work contributes to these efforts.