Evaluate the effectiveness of grassland restoration through mesquiteremoval to increase abundance of grassland birds on treated areas compared to control, untreated areasin the BLM new mexico pecos…

Evaluate the effectiveness of grassland restoration through mesquiteremoval to increase abundance of grassland birds on treated areas compared to control, untreated areasin the BLM new mexico pecos district. We will use the integrated monitoring in bird conservationregions protocol to collect data on the presence of birds on 50 grids split evenly between carlsbad androswell field office areas. In each area, 15 sample locations will be placed on pastures that have beenaerially treated with herbicide to kill mesquite and ten will be placed on pastures that have a highmesquite canopy cover. In addition, in year three, we will update a model of mesquite canopy cover thatis used to plan future herbicide treatments. We will use a deep learning, neural network, machinglearning technique to train a computer to identify mesquite canopy. The result is a raster file with canopycover classified into four categories important for grassland and wildlife management. This model willalso be used as a covariate in analyses to test for differences in bird density in different categories ofmesquite canopy cover. We will deliver annual reports with statistical analyses evaluating differences indensity of grassland birds between treatments (control and treatment) and among primary habitat types(grassland, mesquite shrub, other shrubs, shinnery OAK, etc.) and levels of mesquite canopy cover (15 ). In the final year, we will deliver an updated model of mesquite canopycover. BLM and other restore new mexico partners will benefit from this work by better understandingthe effectiveness of the grassland restoration work being done in southeastern new mexico.