Biodiversity losses are a major driver of global changes in ecosystem functioning. While most studies of the relationship between biodiversity and ecosystem functioning have examined randomized species losses, trait-based filtering associated with species-specific vulnerability to drivers of diversity loss can strongly influence how ecosystem functioning responds to declining biodiversity. Moreover, the responses of ecosystem functioning to diversity loss may be mediated by environmental variability interacting with the suite of traits remaining in depauperate communities. We do not yet understand how communities resulting from realistic diversity losses (filtered by response traits) influence ecosystem functioning (via effect traits of the remaining community), especially under variable environmental conditions. Here, we directly test how realistic and randomized plant diversity losses influence productivity and invasion resistance across multiple years in a California grassland. Compared with communities based on randomized diversity losses, communities resulting from realistic (drought-driven) species losses had higher invasion resistance under climatic conditions that matched the trait-based filtering they experienced. However, productivity declined more with realistic than with randomized species losses across all years, regardless of climatic conditions. Functional response traits aligned with effect traits for productivity but not for invasion resistance. Our findings illustrate that the effects of biodiversity losses depend not only on the identities of lost species but also on how the traits of remaining species interact with varying environmental conditions. Understanding the consequences of biodiversity change requires studies that evaluate trait-mediated effects of species losses and incorporate the increasingly variable climatic conditions that future communities are expected to experience.

We provide updated spatial distribution and inventory data for on-road NH₃ emissions for the continental United States (U.S.) On-road NH₃ emissions were determined from on-road CO₂ emissions data and empirical NH₃:CO₂ vehicle emissions ratios. Emissions of NH₃ from on-road sources in urbanized regions are typically 0.1-1.3tkm^-2yr^-1 while NH₃ emissions in agricultural regions generally range from 0.4-5.5tkm^-2yr^-1, with a few hotspots as high as 5.5-11.2tkm^-2yr^-1. Counties with higher vehicle NH₃ emissions than from agriculture include 40% of the U.S.

Population

The amount of wet inorganic N deposition as NH₄⁺ from the National Atmospheric Deposition Program (NADP) network ranged from 37 to 83% with a mean of 58.7%. Only 4% of the NADP sites across the U.S. had <45% of the N deposition as NH₄⁺ based on data from 2014 to 2016, illustrating the near-universal elevated proportions of NH₄⁺ in deposition across the U.S. Case studies of on-road NH₃ emissions in relation to N deposition include four urban sites in Oregon and Washington where the average NH₄-N:NO₃-N ratio in bulk deposition was 2.3. At urban sites in the greater Los Angeles Basin, bulk deposition of NH₄-N and NO₃-N were equivalent, while NH₄-N:NO₃-N in throughfall under shrubs ranged from 0.6 to 1.7. The NH₄-N:NO₃-N ratio at 7-10 sites in the Lake Tahoe Basin averaged 1.4 and 1.6 in bulk deposition and throughfall, and deposition of NH₄-N was strongly correlated with summertime NH₃ concentrations. On-road emissions of NH₃ should not be ignored as an important source of atmospheric NH₃, as a major contributor to particulate air pollution, and as a driver of N deposition in urban and urban-affected regions.