Predicting ecological outcomes using fuzzy interaction webs.

The past 100 years of empirical research in ecology have generated tremendous knowledge about the component interactions that structure ecological communities. Yet, we still lack the ability to reassemble these puzzle pieces to predict community responses to perturbations, a challenge that grows increasingly urgent given rapid global change. We summarize key advances in community ecology that have set the stage for modeling ecological systems and briefly review the evolution of ecological modeling efforts to identify critical hurdles to progress. We find that while Robert May demonstrated that quantitative models could theoretically predict community interactions nearly 50 years ago, in practice, we still lack the ability to predict ecological outcomes with reasonable accuracy for three reasons: (1) quantitative models require precise data for parameterization (often unavailable) and have restrictive assumptions that are rarely met; (2) estimating interaction strengths for all network components is extremely challenging; and (3) determining which species are essential to include in models is difficult (model structure uncertainty). We propose that fuzzy interaction webs (FIW), borrowed from the social sciences, hold the potential to overcome these modeling shortfalls by integrating quantitative and qualitative data (e.g., categorical data, natural history information, expert opinion) for generating reasonably accurate qualitative predictions sufficient for addressing many ecological questions. We outline recent advances developed for addressing model structure uncertainty, and we present a case study to illustrate how FIWs can be applied for estimating community interaction strengths and predicting complex ecological outcomes in a multitrophic (plants, herbivores, predators), multi-interaction-type (competition, predation, facilitation, omnivory) grassland ecosystem. We argue that incorporating FIWs into ecological modeling could significantly advance empirical and theoretical ecology.

Parasites as ecosystem modulators: foliar pathogens suppress top-down effects of large herbivores.

Parasites can catalyze or inhibit interactions between their hosts and other species, but the ecosystem-level effects of such interaction modifications are poorly understood. We conducted a large-scale field experiment in temperate grasslands of China to understand how foliar fungal pathogens influenced top-down effects of cattle on plant diversity and productivity. When foliar pathogens were suppressed, cattle grazing strongly reduced biomass of the dominant grass, Leymus chinensis, generating competitive release that significantly increased community-level species richness and evenness. In the absence of grazing, pathogen attack on L. chinensis had no measurable effect on host biomass. However, pathogens disrupted top-down effects of herbivory by inhibiting grazing effects on plant biomass and species richness. Mechanistically, fungal pathogens were linked to increased alkaloid and reduced nitrogen levels in leaf tissue, which appeared to deter cattle grazing on L. chinensis. In conclusion, foliar pathogens can suppress top-down effects of large herbivores on grassland community composition and ecosystem function by modifying the strength of their host's interactions with dominant consumers. Parasites may act as modulators of ecosystem function when their direct effects on host abundance are overshadowed by powerful influences on host traits that modify their interactions with competitors, herbivores, or predators.

Seed Size, Seed Dispersal Traits, and Plant Dispersion Patterns for Native and Introduced Grassland Plants.

Most terrestrial plants disperse by seeds, yet the relationship between seed mass, seed dispersal traits, and plant dispersion is poorly understood. We quantified seed traits for 48 species of native and introduced plants from the grasslands of western Montana, USA, to investigate the relationships between seed traits and plant dispersion patterns. Additionally, because the linkage between dispersal traits and dispersion patterns might be stronger for actively dispersing species, we compared these patterns between native and introduced plants. Finally, we evaluated the efficacy of trait databases versus locally collected data for examining these questions. We found that seed mass correlated positively with the presence of dispersal adaptations such as pappi and awns, but only for introduced plants, for which larger-seeded species were four times as likely to exhibit dispersal adaptations as smaller-seeded species. This finding suggests that introduced plants with larger seeds may require dispersal adaptations to overcome seed mass limitations and invasion barriers. Notably, larger-seeded exotics also tended to be more widely distributed than their smaller-seeded counterparts, again a pattern that was not apparent for native taxa. These results suggest that the effects of seed traits on plant distribution patterns for expanding populations may be obscured for long-established species by other ecological filters (e.g., competition). Finally, seed masses from databases differed from locally collected data for 77% of the study species. Yet, database seed masses correlated with local estimates and generated similar results. Nonetheless, average seed masses differed up to 500-fold between data sources, suggesting that local data provides more valid results for community-level questions.

Evaluating unintended consequences of intentional species introductions and eradications for improved conservation management.

Increasingly intensive strategies to maintain biodiversity and ecosystem function are being deployed in response to global anthropogenic threats, including intentionally introducing and eradicating species via assisted migration, rewilding, biological control, invasive species eradications, and gene drives. These actions are highly contentious because of their potential for unintended consequences. We conducted a global literature review of these conservation actions to quantify how often unintended outcomes occur and to elucidate their underlying causes. To evaluate conservation outcomes, we developed a community assessment framework for systematically mapping the range of possible interaction types for 111 case studies. Applying this tool, we quantified the number of interaction types considered in each study and documented the nature and strength of intended and unintended outcomes. Intended outcomes were reported in 51% of cases, a combination of intended outcomes and unintended outcomes in 26%, and strictly unintended outcomes in 10%. Hence, unintended outcomes were reported in 36% of all cases evaluated. In evaluating overall conservations outcomes (weighing intended vs. unintended effects), some unintended effects were fairly innocuous relative to the conservation objective, whereas others resulted in serious unintended consequences in recipient communities. Studies that assessed a greater number of community interactions with the target species reported unintended outcomes more often, suggesting that unintended consequences may be underreported due to insufficient vetting. Most reported unintended outcomes arose from direct effects (68%) or simple density-mediated or indirect effects (25%) linked to the target species. Only a few documented cases arose from more complex interaction pathways (7%). Therefore, most unintended outcomes involved simple interactions that could be predicted and mitigated through more formal vetting. Our community assessment framework provides a tool for screening future conservation actions by mapping the recipient community interaction web to identify and mitigate unintended outcomes from intentional species introductions and eradications for conservation.

Evaluación de las Consecuencias Involuntarias de las Introducciones y Erradicaciones Intencionales de Especies para el Manejo Mejorado de la Conservación Resumen Actualmente se despliegan estrategias cada vez más intensas para mantener la biodiversidad y la función del ecosistema como respuesta a las amenazas antropogénicas mundiales, incluyendo la introducción y erradicación intencionales de especies por medio de la migración asistida, el retorno a la vida silvestre, el control biológico, la erradicación de especies invasoras y la genética dirigida. Estas acciones son muy polémicas por el potencial que tienen para generar consecuencias involuntarias. Realizamos una revisión de la literatura mundial sobre estas acciones de conservación para cuantificar cuán seguido ocurren las consecuencias involuntarias y cuáles son sus causas subyacentes. Para evaluar los resultados de conservación, desarrollamos un marco de trabajo de evaluación comunitaria para mapear sistemáticamente el rango de posibles interacciones para 111 estudios de caso. Con la aplicación de esta herramienta cuantificamos el número de tipos de interacción consideradas en cada estudio y documentamos la naturaleza y la fuerza de los resultados involuntarios. Se reportaron los resultados voluntarios en 51% de los casos, una combinación de resultados voluntarios e involuntarios en 26% de los casos y estrictamente los resultados involuntarios en el 10% de los casos. Por lo tanto, los resultados involuntarios fueron reportados en el 36% de todos los casos evaluados. En la evaluación general de los resultados de conservación (sopesando los efectos voluntarios y. los involuntarios), algunos efectos involuntarios fueron bastante inocuos en relación con el objetivo de conservación, mientras que otros resultaron en consecuencias involuntarias severas para las comunidades receptoras. Los estudios que evaluaron un mayor número de interacciones comunitarias con la especie objetivo reportaron resultados involuntarios con mayor frecuencia, lo que sugiere que las consecuencias involuntarias pueden estar subvaloradas debido al escrutinio insuficiente. La mayoría de los resultados involuntarios reportados surgieron de los efectos directos (68%) o de los efectos indirectos o mediados por la densidad (25%) vinculados con la especie diana. Solamente unos cuantos casos documentados surgieron de interacciones más complejas (7%). Por lo tanto, la mayoría de los resultados involuntarios involucran interacciones simples que podrían ser pronosticadas y mitigadas por medio de un escrutinio más formal. Nuestro marco de trabajo de evaluación comunitaria proporciona una herramienta para la revisión de las acciones de conservación en el futuro mediante el mapeo de la red de interacciones entre comunidades receptoras y para la mitigación de los resultados involuntarios surgidos de las introducciones y erradicaciones intencionales de especies a favor de la conservación.

Seed predator effects on plants: Moving beyond time-corrected proxies.

We previously demonstrated that small mammals impact plant recruitment globally via size-dependent seed predation, generating a unimodal pattern across ecosystems. Chen et al. (2021) critiqued our seed removal analysis, advocating corrections for exposure time. We show such manipulations are unwarranted and argue for increased emphasis on plant recruitment metrics.

Occupancy modeling and resampling overcomes low test sensitivity to produce accurate SARS-CoV-2 prevalence estimates.

BACKGROUND: We evaluated whether occupancy modeling, an approach developed for detecting rare wildlife species, could overcome inherent accuracy limitations associated with rapid disease tests to generate fast, accurate, and affordable SARS-CoV-2 prevalence estimates. Occupancy modeling uses repeated sampling to estimate probability of false negative results, like those linked to rapid tests, for generating unbiased prevalence estimates. METHODS: We developed a simulation study to estimate SARS-CoV-2 prevalence using rapid, low-sensitivity, low-cost tests and slower, high-sensitivity, higher cost tests across a range of disease prevalence and sampling strategies. RESULTS: Occupancy modeling overcame the low sensitivity of rapid tests to generate prevalence estimates comparable to more accurate, slower tests. Moreover, minimal repeated sampling was required to offset low test sensitivity at low disease prevalence (0.1%), when rapid testing is most critical for informing disease management. CONCLUSIONS: Occupancy modeling enables the use of rapid tests to provide accurate, affordable, real-time estimates of the prevalence of emerging infectious diseases like SARS-CoV-2.

Soil engineering by ants facilitates plant compensation for large herbivore removal of aboveground biomass.

The interplay between top-down and bottom-up processes determines ecosystem productivity. Yet, the factors that mediate the balance between these opposing forces remain poorly understood. Furthering this challenge, complex and often cryptic factors like ecosystem engineering and trait-mediated interactions may play major roles in mediating the outcomes of top-down and bottom-up interactions. In semiarid grasslands of northeastern China, we conducted a large-scale, three-year experiment to evaluate how soil engineering by ants and plasticity in plants independently and jointly influenced the top-down effects of grazing by a ubiquitous herbivore (cattle) on aboveground standing biomass of the dominant perennial grass, Leymus chinensis. Herbivory had strong top-down effects, reducing L. chinensis AB by 25% relative to baseline levels without cattle or ants. In contrast, soil engineering by ants facilitated weak bottom-up effects in the absence of herbivory. However, in the presence of herbivory, soil engineering effects were strong enough to fully offset herbivore removal of aboveground biomass. This outcome was mediated by L. chinensis's plasticity in reallocating growth from below- to aboveground biomass, a result linked to additive effects of engineers and herbivores increasing soil N availability and engineering effects improving soil structure. Soil engineering increased soil N by 12%, promoting aboveground biomass. Herbivores increased soil N by 13% via defecation, but this increase failed to offset their reductions in aboveground biomass in isolation. However, when combined, engineers and herbivores increased soil N by 26% and engineers improved soil bulk density, facilitating L. chinensis to shift resource allocations from below- to aboveground biomass sufficiently to fully offset herbivore suppression of aboveground biomass. Our results demonstrate that soil engineering and trait-mediated effects of plant plasticity can strongly mediate the outcome of top-down and bottom-up interactions. These cryptic but perhaps ubiquitous processes may help to explain the long-debated phenomenon of plant compensatory responses to large grazers.

Seed size predicts global effects of small mammal seed predation on plant recruitment.

Recent studies demonstrate that by focusing on traits linked to fundamental plant life-history trade-offs, ecologists can begin to predict plant community structure at global scales. Yet, consumers can strongly affect plant communities, and means for linking consumer effects to key plant traits and community assembly processes are lacking. We conducted a global literature review and meta-analysis to evaluate whether seed size, a trait representing fundamental life-history trade-offs in plant offspring investment, could predict post-dispersal seed predator effects on seed removal and plant recruitment. Seed size predicted small mammal seed removal rates and their impacts on plant recruitment consistent with optimal foraging theory, with intermediate seed sizes most strongly impacted globally - for both native and exotic plants. However, differences in seed size distributions among ecosystems conditioned seed predation patterns, with relatively large-seeded species most strongly affected in grasslands (smallest seeds), and relatively small-seeded species most strongly affected in tropical forests (largest seeds). Such size-dependent seed predation has profound implications for coexistence among plants because it may enhance or weaken opposing life-history trade-offs in an ecosystem-specific manner. Our results suggest that seed size may serve as a key life-history trait that can integrate consumer effects to improve understandings of plant coexistence.

Seedling recruitment correlates with seed input across seed sizes: implications for coexistence.

Understanding controls on recruitment is critical to predicting community assembly, diversity, and coexistence. Theory posits that at mean fecundity, recruitment of highly fecund small-seeded plants should be primarily microsite limited, which is indicated by a saturating recruitment function. In contrast, species that produce fewer large seeds are more likely to be seed-limited, which is characterized by a linear recruitment function. If these patterns hold in nature, seed predation that disproportionately affects larger-seeded species can limit their establishment. We tested these predictions by comparing recruitment functions among 16 co-occurring perennial forb species that vary by over two orders of magnitude in seed size. We also assessed how postdispersal seed predation by mice influenced recruitment. We added seeds at densities from zero to three times natural fecundity of each species to undisturbed plots and examined spatial variation in recruitment by conducting experiments across 10 grassland sites that varied in productivity and resource availability. Consistent across two replicated years, most species had linear recruitment functions across the range of added seed densities, indicative of seed-limited recruitment. Depending on year, the recruitment functions of only 19-37% of target species saturated near their average fecundity, and this was not associated with seed size. Recruitment was strongly inhibited by rodent seed predation for large-seeded species but not for smaller-seeded species. Proportional recruitment was more sensitive to spatial variation in recruitment conditions across sites for some small-seeded species than for large-seeded species. These results contradict the common belief that highly fecund small-seeded species suffer from microsite-limited recruitment. Rather, they imply that, at least episodically, recruitment can be strongly correlated to plant fecundity. However, proportional recruitment of small-seeded species was inhibited at productive sites to a greater extent than large-seeded species. Results also show that in a system where the dominant granivore prefers larger seeds, low-fecundity large-seeded species can suffer from even greater seed-limited recruitment than would occur in the absence of predators.

Secondary invasion re-redefined: The distinction between invader-facilitated and invader-contingent invasions as subclasses of secondary invasion.

Linked Article: https://doi.org/10.1002/ece3.3964.

The fluctuating resource hypothesis explains invasibility, but not exotic advantage following disturbance.

Invasibility is a key indicator of community susceptibility to changes in structure and function. The fluctuating resource hypothesis (FRH) postulates that invasibility is an emergent community property, a manifestation of multiple processes that cannot be reliably predicted by individual community attributes like diversity or productivity. Yet, research has emphasized the role of these individual attributes, with the expectation that diversity should deter invasibility and productivity enhance it. In an effort to explore how these and other factors may influence invasibility, we evaluated the relationship between invasibility and species richness, productivity, resource availability, and resilience in experiments crossing disturbance with exotic seed addition in 1-m2 plots replicated over large expanses of grasslands in Montana, USA and La Pampa, Argentina. Disturbance increased invasibility as predicted by FRH, but grasslands were more invasible in Montana than La Pampa whether disturbed or not, despite Montana's higher species richness and lower productivity. Moreover, invasibility correlated positively with nitrogen availability and negatively with native plant cover. These patterns suggested that resource availability and the ability of the community to recover from disturbance (resilience) better predicted invasibility than either species richness or productivity, consistent with predictions from FRH. However, in ambient, unseeded plots in Montana, disturbance reduced native cover by >50% while increasing exotic cover >200%. This provenance bias could not be explained by FRH, which predicts that colonization processes act on species' traits independent of origins. The high invasibility of Montana grasslands following disturbance was associated with a strong shift from perennial to annual species, as predicted by succession theory. However, this shift was driven primarily by exotic annuals, which were more strongly represented than perennials in local exotic vs. native species pools. We attribute this provenance bias to extrinsic biogeographic factors such as disparate evolutionary histories and/or introduction filters selecting for traits that favor exotics following disturbance. Our results suggest that (1) invasibility is an emergent property best explained by a community's efficiency in utilizing resources, as predicted by FRH but (2) understanding provenance biases in biological invasions requires moving beyond FRH to incorporate extrinsic biogeographic factors that may favor exotics in community assembly.

Community Assembly Theory as a Framework for Biological Invasions.

Biological invasions present a global problem underlain by an ecological paradox that thwarts explanation: how do some exotic species, evolutionarily naïve to their new environments, outperform locally adapted natives? We propose that community assembly theory provides a framework for addressing this question. Local community assembly rules can be defined by evaluating how native species' traits interact with community filters to affect species abundance. Evaluation of exotic species against this benchmark indicates that exotics that follow assembly rules behave like natives, while those exhibiting novel interactions with community filters can greatly underperform or outperform natives. Additionally, advantages gained by exotics over natives following disturbance can be explained by accounting for extrinsic assembly processes that bias exotic traits toward ruderal strategies.

Negative plant-soil feedbacks increase with plant abundance, and are unchanged by competition.

Plant-soil feedbacks and interspecific competition are ubiquitous interactions that strongly influence the performance of plants. Yet few studies have examined whether the strength of these interactions corresponds with the abundance of plant species in the field, or whether feedbacks and competition interact in ways that either ameliorate or exacerbate their effects in isolation. We sampled soil from two intermountain grassland communities where we also measured the relative abundance of plant species. In greenhouse experiments, we quantified the direction and magnitude of plant-soil feedbacks for 10 target species that spanned a range of abundances in the field. In soil from both sites, plant-soil feedbacks were mostly negative, with more abundant species suffering greater negative feedbacks than rare species. In contrast, the average response to competition for each species was unrelated with its abundance in the field. We also determined how competitive response varied among our target species when plants competed in live vs. sterile soil. Interspecific competition reduced plant size, but the strength of this negative effect was unchanged by plant-soil feedbacks. Finally, when plants competed interspecifically, we asked how conspecific-trained, heterospecific-trained, and sterile soil influenced the competitive responses of our target species and how this varied depending on whether target species were abundant or rare in the field. Here, we found that both abundant and rare species were not as harmed by competition when they grew in heterospecific-trained soil compared to when they grew in conspecific-cultured soil. Abundant species were also not as harmed by competition when growing in sterile vs. conspecific-trained soil, but this was not the case for rare species. Our results suggest that abundant plants accrue species-specific soil pathogens to a greater extent than rare species. Thus, negative feedbacks may be critical for preventing abundant species from becoming even more abundant than rare species.

Linking Native and Invader Traits Explains Native Spider Population Responses to Plant Invasion.

Theoretically, the functional traits of native species should determine how natives respond to invader-driven changes. To explore this idea, we simulated a large-scale plant invasion using dead spotted knapweed (Centaurea stoebe) stems to determine if native spiders' web-building behaviors could explain differences in spider population responses to structural changes arising from C. stoebe invasion. After two years, irregular web-spiders were >30 times more abundant and orb weavers were >23 times more abundant on simulated invasion plots compared to controls. Additionally, irregular web-spiders on simulated invasion plots built webs that were 4.4 times larger and 5.0 times more likely to capture prey, leading to >2-fold increases in recruitment. Orb-weavers showed no differences in web size or prey captures between treatments. Web-spider responses to simulated invasion mimicked patterns following natural invasions, confirming that C. stoebe's architecture is likely the primary attribute driving native spider responses to these invasions. Differences in spider responses were attributable to differences in web construction behaviors relative to historic web substrate constraints. Orb-weavers in this system constructed webs between multiple plants, so they were limited by the overall quantity of native substrates but not by the architecture of individual native plant species. Irregular web-spiders built their webs within individual plants and were greatly constrained by the diminutive architecture of native plant substrates, so they were limited both by quantity and quality of native substrates. Evaluating native species traits in the context of invader-driven change can explain invasion outcomes and help to identify factors limiting native populations.

Long-term ungulate exclusion reduces fungal symbiont prevalence in native grasslands.

When symbionts are inherited by offspring, they can have substantial ecological and evolutionary consequences because they occur in all host life stages. Although natural frequencies of inherited symbionts are commonly <100 %, few studies investigate the ecological drivers of variation in symbiont prevalence. In plants, inherited fungal endophytes can improve resistance to herbivory, growth under drought, and competitive ability. We evaluated whether native ungulate herbivory increased the prevalence of a fungal endophyte in the common, native bunchgrass, Festuca campestris (rough fescue, Poaceae). We used large-scale (1 ha) and long-term (7-10 year) fencing treatments to exclude native ungulates and recorded shifts in endophyte prevalence at the scale of plant populations and for individual plants. We characterized the fungal endophyte in F. campestris, Epichloë species FcaTG-1 (F. campestris taxonomic group 1) for the first time. Under ungulate exclusion, endophyte prevalence was 19 % lower in plant populations, 25 % lower within plant individuals, and 39 % lower in offspring (seeds) than in ungulate-exposed controls. Population-level endophyte frequencies were also negatively correlated with soil moisture across geographic sites. Observations of high within-plant variability in symbiont prevalence are novel for the Epichloë species, and contribute to a small, but growing, literature that documents phenotypic plasticity in plant-endophyte symbiota. Altogether, we show that native ungulates can be an important driver of symbiont prevalence in native plant populations, even in the absence of evidence for direct mechanisms of mammal deterrence. Understanding the ecological controls on symbiont prevalence could help to predict future shifts in grasslands that are dominated by Epichloë host plants.

Quantifying "apparent" impact and distinguishing impact from invasiveness in multispecies plant invasions.

The quantification of invader impacts remains a major hurdle to understanding and managing invasions. Here, we demonstrate a method for quantifying the community-level impact of multiple plant invaders by applying Parker et al.'s (1999) equation (impact = range x local abundance x per capita effect or per unit effect) using data from 620 survey plots from 31 grasslands across west-central Montana, USA. In testing for interactive effects of multiple invaders on native plant abundance (percent cover), we found no evidence for invasional meltdown or synergistic interactions for the 25 exotics tested. While much concern exists regarding impact thresholds, we also found little evidence for nonlinear relationships between invader abundance and impacts. These results suggest that management actions that reduce invader abundance should reduce invader impacts monotonically in this system. Eleven of 25 invaders had significant per unit impacts (negative local-scale relationships between invader and native cover). In decomposing the components of impact, we found that local invader abundance had a significant influence on the likelihood of impact, but range (number of plots occupied) did not. This analysis helped to differentiate measures of invasiveness (local abundance and range) from impact to distinguish high-impact invaders from invaders that exhibit negligible impacts, even when widespread. Distinguishing between high- and low-impact invaders should help refine trait-based prediction of problem species. Despite the unique information derived from evaluation of per unit effects of invaders, invasiveness 'scores based on range and local abundance produced similar rankings to impact scores that incorporated estimates of per unit effects. Hence, information on range and local abundance alone was sufficient to identify problematic plant invaders at the regional scale. In comparing empirical data on invader impacts to the state noxious weed list, we found that the noxious weed list captured 45% of the high impact invaders but missed 55% and assigned the lowest risk category to the highest-impact invader. While such subjective weed lists help to guide invasive species management, empirical data are needed to develop more comprehensive rankings of ecological impacts. Using weed lists to classify invaders for testing invasion theory is not well supported.

Are local filters blind to provenance? Ant seed predation suppresses exotic plants more than natives.

The question of whether species' origins influence invasion outcomes has been a point of substantial debate in invasion ecology. Theoretically, colonization outcomes can be predicted based on how species' traits interact with community filters, a process presumably blind to species' origins. Yet, exotic plant introductions commonly result in monospecific plant densities not commonly seen in native assemblages, suggesting that exotic species may respond to community filters differently than natives. Here, we tested whether exotic and native species differed in their responses to a local community filter by examining how ant seed predation affected recruitment of eighteen native and exotic plant species in central Argentina. Ant seed predation proved to be an important local filter that strongly suppressed plant recruitment, but ants suppressed exotic recruitment far more than natives (89% of exotic species vs. 22% of natives). Seed size predicted ant impacts on recruitment independent of origins, with ant preference for smaller seeds resulting in smaller seeded plant species being heavily suppressed. The disproportionate effects of provenance arose because exotics had generally smaller seeds than natives. Exotics also exhibited greater emergence and earlier peak emergence than natives in the absence of ants. However, when ants had access to seeds, these potential advantages of exotics were negated due to the filtering bias against exotics. The differences in traits we observed between exotics and natives suggest that higher-order introduction filters or regional processes preselected for certain exotic traits that then interacted with the local seed predation filter. Our results suggest that the interactions between local filters and species traits can predict invasion outcomes, but understanding the role of provenance will require quantifying filtering processes at multiple hierarchical scales and evaluating interactions between filters.

Staged invasions across disparate grasslands: effects of seed provenance, consumers and disturbance on productivity and species richness.

Exotic plant invasions are thought to alter productivity and species richness, yet these patterns are typically correlative. Few studies have experimentally invaded sites and asked how addition of novel species influences ecosystem function and community structure and examined the role of competitors and/or consumers in mediating these patterns. We invaded disturbed and undisturbed subplots in and out of rodent exclosures with seeds of native or exotic species in grasslands in Montana, California and Germany. Seed addition enhanced aboveground biomass and species richness compared with no-seeds-added controls, with exotics having disproportionate effects on productivity compared with natives. Disturbance enhanced the effects of seed addition on productivity and species richness, whereas rodents reduced productivity, but only in Germany and California. Our results demonstrate that experimental introduction of novel species can alter ecosystem function and community structure, but that local filters such as competition and herbivory influence the magnitude of these impacts.

Population-level compensation impedes biological control of an invasive forb and indirect release of a native grass.

The intentional introduction of specialist insect herbivores for biological control of exotic weeds provides ideal but understudied systems for evaluating important ecological concepts related to top-down control, plant compensatory responses, indirect effects, and the influence of environmental context on these processes. Centaurea stoebe (spotted knapweed) is a notorious rangeland weed that exhibited regional declines in the early 2000s, attributed to drought by some and to successful biocontrol by others. We initiated an experiment to quantify the effects of the biocontrol agent, Cyphocleonus achates, on Ce. stoebe and its interaction with a dominant native grass competitor, Pseudoroegneria spicata, under contrasting precipitation conditions. Plots containing monocultures of each plant species or equal mixtures of the two received factorial combinations of Cy. achates herbivory (exclusion or addition) and precipitation (May-June drought or "normal," defined by the 50-year average) for three years. Cy. achates herbivory reduced survival of adult Ce. stoebe plants by 9% overall, but this effect was stronger under normal precipitation compared to drought conditions, and stronger in mixed-species plots compared to monocultures. Herbivory had no effect on Ce. stoebe per capita seed production or on recruitment of seedlings or juveniles. In normal-precipitation plots of mixed composition, greater adult mortality due to Cy. achates herbivory resulted in increased recruitment of new adult Ce. stoebe. Due to this compensatory response to adult mortality, final Ce. stoebe densities did not differ between herbivory treatments regardless of context. Experimental drought reduced adult Ce. stoebe survival in mixed-species plots but did not impede recruitment of new adults or reduce final Ce. stoebe densities, perhaps due to the limited duration of the treatment. Ce. stoebe strongly depressed P. spicata reproduction and recruitment, but these impacts were not substantively alleviated by herbivory on Ce. stoebe. Population-level compensation by dominant plants may be an important factor inhibiting top-down effects in herbivore-driven and predator-driven cascades.