COS 119-4 - The hidden costs of fragmentation: disrupted species-area relationships and biotic homogenization

Abstract: Background/Question/Methods

Despite multiple studies indicating that biodiversity is affected by habitat fragmentation and loss, there remains insufficient empirical evidence to support the generalization of the known impacts of habitat size on alpha diversity at the patch scale to the landscape scale. The SLOSS debate continues to rumble over whether single large or several small patches increases species diversity, but it fails to explain that, in addition to habitat size, habitat spatial configuration and matrix quality also play a role in biodiversity variation. We investigated whether landscape variables influence (i) the relationship between habitat size and species richness at the patch scale, as well as (ii) beta and gamma diversity at the landscape level. We compared the abundances and identities of 7,672 vertebrates, invertebrates, and plant species in 1,314 forest patches using a global database compiled from 89 studies. We used buffers (500 to 2,000 m) to extract landscape variables (e.g., number of patches, forest percentage, and connectivity). We used a subset of the datasets to comparer alpha, beta and gamma diversity between continuous and fragmented forest.

Results/Conclusions

We show that as patch size increased, standardized species richness and evenness increased across all taxa. However, as the amount of forest remaining in the landscapes decreased and the number of patches increased from low to high, the relationship between patch size and species richness disappeared. In addition, landscape fragmentation had a negative effect on beta diversity. We found that forest fragments had lower alpha diversity and higher beta diversity than continuous forests. However, despite beta diversity being positively related to the number of small fragments, higher beta diversity did not translate into a higher gamma diversity at the landscape scale. We found that forest fragments had lower alpha diversity and higher beta diversity than continuous forests. Higher beta diversity in fragmented forests did not translate into a higher gamma diversity at the landscape scale.
These findings suggest that synthesizing species-area relationships between patch size and species diversity is difficult without taking into account the landscape in which those patches are embedded. Habitat patches cannot be treated as equivalent to isolated islands in an unsuitable aquatic matrix. Conservation strategies might prioritize the amount of available habitat in the landscape as it benefits local diversity in larger patches, and patches in continuous areas or with clumped fragments as fragmentation per se weakens the positive effect of larger patches on local diversity.