Effects of Plant Invasions on Birds Breeding In Desert Grasslands

Abstract

Over the past 150 years, encroachment by woody plants and invasions by nonnative grasses have altered the structure and composition of plant communities that once dominated desert and semidesert grasslands. These plant invasions have reduced the quantity and quality of habitat for many grassland-associated species, including grassland birds, which have declined more rapidly across North America than any other group of birds. Along gradients in cover of woody plants and nonnative grasses in semidesert grasslands of southeastern Arizona, we evaluated how these plant invasions affected species richness, density, habitat selection, and arthropod prey of breeding birds between 2013 and 2015. Encroachment by woody plants influenced breeding birds strongly, with distributions and densities of most grassland obligates decreasing sharply as woody cover increased. As encroachment by woody plants progressed, overall density of birds increased as shrub-associated species were recruited to the community. Bird species richness also increased along the encroachment gradient and plateaued when woody cover reached ~20%, the point at which losses of grassland-associated species equaled gains in shrub-associated species. Invasions by nonnative grasses altered settlement patterns and distributions of breeding bird species differentially, with variation among species reflecting the interplay between the plasticity of habitat preferences and the structural similarity between the invading plants and the native plants with which they evolved. These patterns reveal the importance of the role of habitat selection by individual species as a mechanism to explain broad-scale responses of birds that express at population and community scales. Additionally, invasions by nonnative grasses altered populations and communities of arthropods, the primary prey of grassland birds during their nesting period. For every 10% increase in nonnative-grass dominance, richness of arthropods decreased by 2% and total abundance decreased by 7%. Responses of arthropods varied among taxa and functional groups. Abundances of most predators and specialist herbivores decreased as dominance of nonnative grasses increased, whereas abundances of most generalist herbivores were lowest at intermediate points of nonnative dominance. These changes in the arthropod community can alter broad-scale ecological processes and reduce food resources for insectivores, which can have adverse, cascading effects on imperiled grassland ecosystems. As woody plants and nonnative grasses continue to proliferate in desert grassland ecosystems and threaten native species, monitoring the effects on species of conservation concern reliably will become increasingly important. We evaluated the influence of population attributes (animal density, availability, detection probability, and magnitude of trend) and sampling design features (number of sites, number of repeat surveys, and survey-year interval) on statistical power to detect trends in abundance within a distance-sampling framework. For a range of values common to monitoring studies for breeding grassland birds, power to detect trends in abundance was highest when the number of sites surveyed was maximized, sites were surveyed only once per year, and the interval between survey years was long, a pattern independent of density or detectability of the target species. Our results provide general guidance to help ensure that monitoring programs will have sufficient power to meet objectives effectively and efficiently.