Using local autocorrelation analysis to identify conservation areas: an example considering threatened invertebrate species in Spain

Olga Lucía Hernández-Manrique,David Sánchez-Fernández,José R. Verdú,Catherine Numa,Eduardo Galante,Jorge M. Lobo
DOI: https://doi.org/10.1007/s10531-012-0303-5
IF: 3.4
2012-05-15
Biodiversity and Conservation
Abstract:Local autocorrelation statistics offer new opportunities for the discrimination of important conservation areas since the spatial dependence of local values upon neighbouring ones may assist conservation decisions. We exemplify the use of local autocorrelation statistics for conservation purposes using data on Spanish threatened invertebrates to identify areas composed of similarly species-rich localities (hot spots), species-rich “islands”, cold spots and species-poor “islands”. In order to assess the probable causes of the detected patterns differences in environmental, land use and protected area variables were examined between the different regions. Distributional data for threatened invertebrate species in Spain at 100 km2 UTM cell resolution were used. After defining a neighbouring area, statistically significant local autocorrelation values (both positive and negative) were estimated. Kruskal–Wallis ANOVA by rank test was used to compare the environmental, land use and protected area percentages between the cells of the different regions. Around 11 and 2 % of total cells can be considered hot spots and rich “islands”, respectively. Hot spots are characterized by a lower percentage of anthropic land uses and a higher percentage of current protected area. However, approximately a third part of these cells possess at least 98 % of their area unprotected. Rich “island” cells are not environmentally different from those considered as cold spots, though experiencing a lower rate of anthropization and higher proportion of protected area. Unfortunately, almost 70 % of these rich “island” cells have <2 % of their areas currently protected. The use of local autocorrelation statistics on species richness values may complement conservation decisions by discriminating interconnected sites facilitating local persistence (hot spots) as well as isolated and vulnerable sites (rich “islands”). The study of different variables associated with these regions allows us to suggest determinant causal factors. Our results suggest that land use changes due to human activities are the main cause of threats.
environmental sciences,biodiversity conservation,ecology
What problem does this paper attempt to address?