Do we know enough? The impacts of data biases, insufficient sampling and degradation on biodiversity estimates in Tanzanian forests – implications for conservation planning

In the face of continued loss of biodiversity and limited conservation resources, conservation needs to be as efficient as possible. This relies on an unbiased knowledge of the distribution of biodiversity and the threats to it. While we cannot afford to delay conservation action until better data are available, it is essential to understand the limitations of and mitigate biases in the data that we base critical conservation decisions on. This thesis analyses the implications of data biases, insufficient data and degradation and formulates recommendations for conservation planning. While it focuses on vascular plants in the coastal forests and Eastern Arc Mountains in Tanzania, an area recognised internationally for its high biodiversity, the concepts presented are generic with application outside the study area and across other taxonomic groups. The thesis’ findings indicate that the plant data used to underpin conservation decisions in the study area are severely biased and that some of the data are insufficiently representative to capture the true patterns. Funding, botanist and sampling intensity biases, which are circularly linked, may partly explain why despite decades of research surprisingly little consensus has been reached on patterns in biodiversity. In the study area, these effects account for a substantial proportion in the variation in perceived plant diversity patterns with alarming consequences for the reliability of conservation priority assessments. This and the evidence for quickly spreading degradation highlight the urgent need for more efficient surveys and systematic conservation planning. The recommendations include focusing surveys on under-researched locations, distributing botanical expertise more equally across the study area, and using pilot studies prior to surveys to determine the minimum needed sample size. Most importantly, coordinated efforts are needed to develop conservation goals and strategically plan surveys in order to mitigate data biases, and avoid duplication and ad-hoc donor-driven conservation planning. These recommendations may be applicable to other regions and even to the global scale, where partly competing conservation prioritisation schemes determine where billions of dollars in conservation investments are spent, each based on their own data and analyses. Problems associated with these schemes are seldom communicated lest public support may be compromised; however, understanding and acknowledging the gaps in our knowledge on the distribution of biodiversity and threats to it are essential to conserve the world’s biodiversity more effectively.