Discussion
Our study uniquely assessed the simultaneous responses of plants, insects, and soil microbes to gradients of invasion at multiple study sites with two different invasive brome species. Overall, invasive annual brome abundance was more strongly related to plant community composition than insect and soil microbial communities. B. tectorum also had stronger associations with insects and soil microbes, suggesting communities show differential sensitivity to invasion based on taxa and that even apparently similar invasive species (congeneric C3 winter annual grasses of the same genus) may impact communities in unique ways.
Plant species richness significantly decreased with brome invasion, and community and functional composition shifted with invasion, aligning with previous work (Gasch et al. 2013), as annual bromes often have a competitive advantage over native species. B. arvensis andB. tectorum germinate in the fall or spring depending on climatic conditions, allowing them to utilize available nutrients and moisture earlier than perennial species (Germino et al. 2016). We found declines with invasion in C4 perennial grasses across all gradients. C4 perennials, growing later in the season, are sensitive to light availability (Still et al. 2003). As annual bromes use available moisture and decrease light earlier in the season, they outcompete native C4 species, reducing C4 abundance (Chambers et al. 2007). In Montana, most plant functional groups were negatively associated with increasing invasion, suggesting B. arvensis may create wide-ranging plant compositional shifts in this system.
While insect richness did not vary consistently with invasion, insect community and functional composition significantly differed with invasion across all gradients except one (functional composition in Wyoming B. arvensis gradients). As annual bromes shift plant composition, insect communities are also likely to change, given the feedback between available plant species and arthropod presence (Biere and Bennett 2013). For both insects and plants, functional composition results largely paralleled community composition results, suggesting a lack of functional redundancy with potential consequences for ecosystem services such as palatable forage availability, pollination, and pest control (Memmott and Waser 2002, Stout and Morales 2009, Kaiser-Bunbury et al. 2011). When insect functional groups were considered independently, we saw declines in leaf-chewing and sap-sucking herbivores in Montana and increases in parasitoids and sap-sucking herbivores in Wyoming B. arvensis gradients. The declines in certain herbivores may be a reflection of less heterogenous habitat (Germino et al. 2016) with increased invasion cover and/or a preference for native forage species (Cumberland et al. 2016). However, the increase in parasitoids supports a preference for annual brome species as an oviposition site (Perez-Mendoza et al. 2006). Positive relationships between annual brome species and insect abundance in Wyoming also match prior work suggesting that bromes are associated with higher insect abundance in this short-structured prairie (Duchardt et al. 2021).
Insect herbivory was less common on annual bromes than native species. Interestingly, across all years and study sites, invasion level was never related to total insect biomass (Figure 4, Appendix S1: Figure S4, Tables S5, S6). Based on these observations, we measured herbivory in two years in Montana. We found that insect herbivores, under any invasion level, preferred native forage species to invasive bromes. While there is some evidence that certain insect herbivores prefer native species over bromes later in the growing season, insect herbivores will consume bromes earlier in the season (Cumberland et al. 2016). However, we found that both later in the season (2021) and early in the growing season (2022), insect herbivores avoided consuming invasive bromes. This supports our previous finding of a decline in certain herbivore functional group abundances in Montana as well. Combined with a lack of response of total insect biomass to invasion, this could indicate substantial added pressure on native forage species, as insects compete with livestock for forage (Branson and Haferkamp 2014).
We predicted that compositional and functional changes with invasion would occur across plants, insects, and soil microbes. However, soil microbial communities were markedly less sensitive to brome invasion than other taxa considered as we saw marginal changes in B. tectorum gradients only. At the functional group scale, we found strong declines in functional groups associated with phototrophy, asBromus spp. can increase leaf litter (and thatch) and decrease light availability (Still et al. 2003, Bennett et al. 2014). Given that B. tectorum also utilizes large amounts of soil nitrogen (Blank and Morgan 2012), observed negative associations betweenB. tectorum and various nitrogen processing microbes are also logical. While B. tectorum has previously been linked to decreased soil microbial abundance (Gasch et al. 2013) and altered community composition (Nasto et al. 2022), potentially due to soil moisture differences (Gasch et al. 2013), our results newly suggest more limited associations between microbial functional groups and B. arvensis , implying that the microbial communities may be resilient to even large shifts in B. arvensis abundance, protecting services associated with microbial communities like nutrient cycling (Batten et al. 2006). Alternatively, our sequencing methodology and functional groupings may only partly account for functional differences relative to other methods (e.g., soil enzyme activity or transcriptomics), potentially overinflating functional redundancy in our study (Louca et al. 2016).
We found plants, insects, and soil microbes did not relate to B. arvensis and B. tectorum equally. Given the functional similarity of these two species (both are C3 winter annual grasses), we expected fairly similar associations between the invasives and plants, insects, and microbes. Our results suggest thatB. tectorum may have more substantial and far-reaching consequences for community structure than B. arvensis . While studies comparing B. arvensis and B. tectorum are lacking, evidence suggests niche differentiation between the species, particularly in terms of microsite preference (Porensky and Blumenthal 2016). Underlying differences in microsite, including light availability (bare ground cover) soil texture, moisture, and nutrient availability, may help explain the soil microbial differences seen between the species (Porensky et al. 2018). Encouragingly, consequences of invasion for soil functioning may not be too damaging under invasion by B. arvensis . In general, this supports the idea that invasion studies and management must consider invasive species identity, as well as functional group (Roscher et al. 2009).
Overall, this work sheds light on the understudied consequences of annual brome invasion in northern mixed-grass prairies for plants, insects, and soil microbes. Species interactions shape ecosystem functions, and these dynamics are influenced by biodiversity (Thébault and Loreau 2005) and invasion. Annual brome invasion related to plant and insect community structure and function, highlighting the potential for cascading consequences of invasion by B. arvensis andB. tectorum for rangeland ecosystems. To the contrary, soil microbial communities appeared to be more resistant to change in relation to invasive abundance, suggesting resilience to invasive bromes. Responses to invasion vary across taxa, so further studies are needed to examine cross-community change. Understanding how invasion alters community composition and functional group diversity across rangeland communities is vital not only for long-term rangeland sustainability (DiTomaso 2000), but also for advancing invasion science and management (Ricciardi et al. 2017).