On Friday, January 29th, we disussed two papers: Carvalho et al. 2010 and and Keith et al. 2009 (see full references below). The first tried to identify strategies for systematic conservation planning with incomplete datasets, while the latter highlighted the likelihood of non-analogous community formation in response to rapid climate change.
We appreciated the great amount of effort by Carvalho et al. trying to address several questions simultaneously. They were artificially reducing the amount of data on species distribution to 50, 25 and 10% of the original dataset (which had presence/absence records of amphibians and reptiles and covered the Iberian peninsula). These datasets (and the original one) they used in four different ways: as such, fitting species distribution models to them and using the probabilities of occurrence output by the model predictions, using model predictions transformed into presences and absences and compiling “combined” datasets of predicted and observed distributions. Finally, they used the minimum set and target-based planning approaches to perform reserve selection. Ways to use data were compared according to species representation in and cost-efficiency of resulted reserve networks. They found that no one strategy was the best in all cases, but rather the preferable approach varied with data comprehensiveness.
We were a bit concerned that the numerous analyses and steps along the process may have resulted in difficulties to interpret the results of Carvalho et al. It is not always clear what exactly is causing the observed patterns. The target-based planning algorithm in Zonation raised special concerns, as it is known to come up with suboptimal results compared to continuous benefit functions. A question that remained open is which, if any, of the planning approaches resulted in better species representation than what would be expected by random. However, the main conclusion of the paper highlighted the importance of knowing (or being able to estimate) the quality of species distribution data to make good decisions, which seems to be very important.
Keith et al. point out that many of today’s conservation strategies aim at maintaining communities as they are and assign conservation status according to community composition. They raise the question of whether this strategy is meaningful in the face of climate change that affects species distributions and might drive the formation of communities whose composition is different from any communities known today.
We shared Keith et al.’s concern over assigning conservation status to areas based on the occurrence of certain species or species combinations. This may lead to unintentional degradation of biodiversity conservation, if areas lose their conservation status when species occupying them change. Perhaps a more holistic approach to evaluating conservation success than just looking at certain species at certain locations should be adopted. The benchmarking should, however, be based on solid measures to avoid woolliness. If the policies in place are suboptimal for conserving biodiversity for the future, they should be reformed to better serve their purpose.