A comparison of ground-based methods for estimating canopy closure for use in phenology research

Abstract Climate change is influencing tree phenology, causing earlier and more prolonged canopy closure in temperate forests. Canopy closure is closely associated with understorey light, so shifts in its timing have wide-reaching consequences for ecological processes in the understorey. Widespread monitoring of forest canopies through time is needed to understand changes in light availability during spring in particular. Canopy openness, derived from hemispherical photography, has frequently been used as a proxy for understorey light. However, hemispherical photography is relatively resource intensive, so we tested a range of inexpensive alternatives for monitoring variability in canopy closure (visual estimation, canopy scope, smartphone photography, smartphone photography with fisheye attachment; and image analysis with specialist hemispherical photography software or with simpler, open access image analysis software). Smartphone photography with an inexpensive fisheye lens attachment proved the most reliable estimator of canopy closure. We found no significant difference in canopy estimations from three widely-owned smartphone models with differing resolutions and fields of view, and no significant effect of camera operator on the results. ImageJ, a free image analysis software, detected canopy variability in a similar way to HemiView specialist hemispherical photography software. We recommend a combination of smartphone photography with fisheye attachment and analysis with ImageJ for identifying changes in the timing of canopy closure (but not for estimating absolute canopy closure). We discuss how large-scale citizen science using this approach could generate meaningful and comparative data on the timings of canopy closure in different forests, year-to-year.Abstract Climate change is influencing tree phenology, causing earlier and more prolonged canopy closure in temperate forests. Canopy closure is closely associated with understorey light, so shifts in its timing have wide-reaching consequences for ecological processes in the understorey. Widespread monitoring of forest canopies through time is needed to understand changes in light availability during spring in particular. Canopy openness, derived from hemispherical photography, has frequently been used as a proxy for understorey light. However, hemispherical photography is relatively resource intensive, so we tested a range of inexpensive alternatives for monitoring variability in canopy closure (visual estimation, canopy scope, smartphone photography, smartphone photography with fisheye attachment; and image analysis with specialist hemispherical photography software or with simpler, open access image analysis software). Smartphone photography with an inexpensive fisheye lens attachment proved the most reliable estimator of canopy closure. We found no significant difference in canopy estimations from three widely-owned smartphone models with differing resolutions and fields of view, and no significant effect of camera operator on the results. ImageJ, a free image analysis software, detected canopy variability in a similar way to HemiView specialist hemispherical photography software. We recommend a combination of smartphone photography with fisheye attachment and analysis with ImageJ for identifying changes in the timing of canopy closure (but not for estimating absolute canopy closure). We discuss how large-scale citizen science using this approach could generate meaningful and comparative data on the timings of canopy closure in different forests, year-to-year