Planning Site Restoration of Southwestern Willow Flycatcher Habitat with High Resolution (1m) Flycatcher Niche Models Incorporating Classification of Tamarisk, Willow, and Cottonwood from Aerial Imagery
James L. Tracy1 and Robert N. Coulson
1 Texas A&M University, College Station, TX, USA
Site restoration planning for habitat of the federally endangered Southwestern Willow Flycatcher (Empidonax traillii extimus, flycatchers) requires planning optimal configuration of willow patches in light of the existing vegetation, hydrology, the impact of tamarisk beetles, and potential costs. Habitat suitability index (HSI) and MaxEnt niche models were compared for projecting flycatcher habitat suitability at a site on Tonto Creek, Arizona. Niche models incorporated detailed mapping of riparian vegetation. Vegetation was mapped using classification of multi-temporal high resolution (1 m) aerial imagery from USDA National Agricultural Imagery Program and Google Earth. Riparian vegetation was classified to greater than 80% accuracy using random forest in conjunction with the aid of feature selection from a large number of spectral indices. The resulting flycatcher habitat models for the Tonto Creek site were then used to assess current habitat quality and develop future simulation scenarios for (1) tamarisk beetle-induced defoliation and dieback, and (2) restoration of willow habitat patches. Flycatcher habitat suitability simulations can guide planning of willow restoration activities to improve flycatcher habitat suitability in tamarisk beetle affected habitats.