• Energy Research
• biological sciences close

Fragmentation poses one of the greatest threats to tropical forests with short‐term changes to the structure of forest canopies affecting microclimate, tree mortality, and growth. Yet the long‐term effects of fragmentation are poorly understood because (1) most effects require many decades to materialize, but long‐term studies are very rare, (2) the effects of edges on forest canopy structure as a function of fragment size are unknown, and (3) edge effects are often confounded by fragment shape. We quantified the long‐term (centennial) effects of fragmentation on forest canopy structure using airborne light detection and ranging (LiDAR) of 1060 Hawaiian rain forest fragments ranging in size from 0.02 to 1000 ha, created more than 130 years ago by flowing lava. Along with distance from edge, we developed a metric, minimum span, to gain additional insight into edge effects on three measures of canopy structure: canopy height, height variation, and gap fraction.Fragment size was a strong determinant of the three structural variables. Larger fragments had greater average height, larger variation in height, and smaller gap fraction. Minimum span had a large effect on the depth and magnitude of edge effects for the three structural variables. Locations associated with high span values (those surrounded by more forest habitat) showed little effect of distance to fragment edge. In contrast, locations with low span values (those more exposed to edges) were severely limited in canopy height, showed lower height variation, and were associated with greater gap fraction values. The minimum span attribute allows for a more accurate characterization of edge as well as fragment‐level effects, and when combined with high resolution imagery, can improve planning of protected areas for long‐term ecological sustainability and biodiversity protection.