Table 5. Total bird abundance in the study area for six forest-bird species, due to (the various components of) forest fragmentation.

	Total bird abundance (#individuals)		Changes in bird abundance (% of initial abundance) caused by:
	1985	2008	Forest loss (ΔF)	Changed patch features (ΔD)	Correlation term (ΔF*ΔD)	Forest Fragmentation (ΔF+ΔD+ΔF*ΔD)
Chucao	222,443	224,429	−31,864 (−14.3%)	+39,509 (+17.8%)	−5,659 (−2.6%)	+1,986 (+0.9%)
Elaenia	686,273	571,738	−98,305 (−14.3%)	−18,943 (−2.8%)	+2,713 (+0.4%)	−114,535 (−16.7%)
Hummingbird	622,950	638,974	−89,234 (−14.3%)	+122,858 (+19.7%)	−17,599 (−2.8%)	+16,025 (+2.6%)
Rayadito	935,236	801,268	−133,968 (−14.3%)	-	-	−133,968 (−14.3%)
Blackbird	342,595	293,520	−49,075 (−14.3%)	-	-	−49,075 (−14.3%)
Thrush	1,381,746	1,183,817	−197,928 (−14.3%)	-	-	−197,928 (−14.3%)

All estimates are based on relationships between bird density and patch features measured in the 2008 field survey, and patch features measured in Landsat images from 1985 and 2008. For the three species in which bird density did not vary with patch characteristics, changes in bird abundance caused by changed patch features (ΔD) and the correlation term (ΔF*ΔD) are predicted to be zero, thus they are left blank. Sign before the figures indicates decreased (−) or increased (+) bird abundances.