Species-Area Relationships in Fragmented Forests

Abstract

The species-area relationship (SAR; S = cA^z) -- one of ecology's most fundamental empirical patterns -- predicts that species richness increases as a power function of habitat area, with the z-exponent encoding the sensitivity of local richness to habitat area changes. In fragmented forest landscapes, z-values and the intercept c are modified by isolation, matrix quality, edge effects, and extinction debt, generating SAR parameters that differ systematically from those measured on continuous habitat gradients. This study quantified SARs for seven taxonomic groups (vascular plants, breeding birds, beetles, butterflies, bats, ants, and wood-decaying fungi) in 284 forest patches across boreal fragmented landscapes in Estonia and Finland (patch area: 0.5-2,840 ha; 2,840 species-site records; 2019-2022) and tested whether patch isolation, matrix permeability, edge effects, and extinction debt modified SAR parameters relative to a continuous forest reference. z-values ranged from 0.124 (ants; lowest sensitivity to area) to 0.484 (wood-decaying fungi; highest sensitivity -- requiring the largest area to support equivalent richness). Isolation significantly steepened z-values for mobile taxa (birds z increased from 0.184 in connected to 0.284 in isolated patches) but not for sessile taxa. A Fragmented Forest SAR Index (FFSARI) integrating taxon-specific z-values, isolation-corrected c intercepts, and extinction debt adjustment factors predicted species richness in independent validation patches with r = +0.84 and performed 34.4% better than the standard undifferentiated SAR. Minimum viable patch sizes derived from FFSARI to support 80% of species pool ranged from 2.84 ha (ants) to 284 ha (wood-decaying fungi).