Bird and arthropod response to herbicide and grazing: Trade‐offs and time lags in invasive plant management

Managing ecologically harmful plant species requires an understanding of the impacts of complex spatial and temporal disturbances on ecological communities. Actions aimed at controlling non‐native plants can have dynamic effects on patterns of succession, with lag effects and trade‐offs possible. In grasslands, invasive plants can be reduced using intensive grazing during the peak growth of target species or applying herbicide when non‐target plants are dormant. It is unclear how combining herbicide with grazing impacts fauna and whether there are lag effects as ecosystem recovery occurs. Using landscape‐scale experiments, we studied how these techniques affect multiple trophic levels several years after management disturbance, focusing on grassland birds (2015–2018) and their arthropod prey (2015–2017). Our large study patches ( n = 54, mean size = 10 ha) were invaded by tall fescue ( Schedonorus arundinaceus ), an ecologically harmful non‐native grass. We first conducted a large‐scale controlled experiment on a subset of 21 patches, which were grouped spatially into threes and treated with (1) herbicide (glyphosate), (2) herbicide and native seeding, or (3) no treatment (control). When examining the entire sample of 54 patches, we were able to explore the interactive effects of herbicide and (1) intensive‐early stocking (double‐stocking cattle April–June), (2) season‐long stocking (moderate stocking April–September), or (3) no grazing. Birds and arthropods were monitored for 3–4 years post‐treatment, allowing us to examine temporal shifts after these management disturbances. In our experiment, we found that herbicide‐treated patches, in comparison with controls, had an increased abundance of several native bird species and arthropod orders in Year 2 or 3 after treatment, with some negative responses in Year 1. These lag times indicate that applying herbicide over broad areas without providing refugia may have short‐term negative consequences, especially for arthropods. When examining the interactive effects of treatments, we found that herbicide in conjunction with grazing resulted in declines for several disturbance‐sensitive bird species. Improved understanding of lag effects, trade‐offs, and taxon‐specific responses to treatment combinations can inform restoration that limits deleterious effects of treatments on native wildlife.