Coincident shifts in riparian ground-active arthropod diversity and soil nutrients under an introduced symbiotic N2-fixing treeBenjamin D. Duval1*, Evangelina Carabotta1, Sergio de Tomas-Marin1, David Lightfoot21Biology Department, New Mexico Institute of Mining and Technology, Socorro, NM 878012Museum of Southwestern Biology, University of New Mexico, Albuquerque, NM 87131*corresponding author, ORCID 0000-0001-7692-4400Biology Department801 Leroy PlaceNew Mexico Institute of Mining and Technology01-575-835-5820, [email protected] nitrogen-fixing plants such as Russian olive, can significantly impact soil chemistry and invertebrate biodiversity in riparian ecosystems. Here, the effects of Russian olive on soil chemical properties and invertebrate communities in riparian zones of the southwestern U.S. were investigated. Russian olive stands were compared to native cottonwood stands and restoration sites by analyzing soil nitrogen (N), phosphorus (P) and moisture levels, and arthropod diversity and abundance. Sites where Russian olive is present led to a net increase in soil nitrogen, a decrease in soil phosphorus, and greater soil moisture compared to both native cottonwood stands and restoration sites. Native cottonwood stands showed lower soil N and higher P levels, as well as higher arthropod diversity. This increased diversity could be linked to the soil’s nutrient stoichiometry, as there is a negative correlation between taxonomic diversity and the soil N:P ratio. Moreover, there was a greater abundance of detritivorous arthropods in Russian olive stands compared to native vegetation. Soil nitrate (NO3-) levels showed a strong positive correlation with detritivorous arthropod abundance, but only a moderate correlation with herbivores, and NO3- was unrelated to predator abundance. These results suggest that Russian olive stands can alter soil chemistry in ways that disproportionately benefit detritivores, potentially disrupting the balance of arthropod communities and reducing overall biodiversity in riparian ecosystems. The study underscores the need for careful management of invasive, symbiotic N2-fixing plant species to preserve the ecological integrity of riparian habitats.Keywords: nitrate, phosphate, Rio Grande, Russian olive, soilINTRODUCTIONA persistent question in invasive species ecology is what are the functional roles of the invaders and their effect in the biodiversity and functional structure and composition of the recipient community (Renault et al., 2022; Galán Díaz et al., 2023). Invasive species can have cascading effects through different trophic levels and alter ecosystem functions, processes and services (McCary et al., 2016; Castro Díez and Alonso Fernández, 2017). One of the main effects of invasive plant species is the disruption of nutrient cycling and availability (Weidenhamer and Callaway, 2010; Afzal et al., 2023) which is especially relevant in invasions produced by woody and N2-fixing plants (Liao et al. 2008).Trees in symbiotic relationships with N2-fixing actinobacteria are globally reliable as invasive species and known to induce substantial effects on the N cycle relative to native vegetation (Vitousek et al. 1987, Nsikani et al. 2018, Collette and Pither 2015). These trees typically increase soil mineral N pools relative to native vegetation, and should be considered part of the ongoing anthropogenic global disruption of the nitrogen (N) cycle (Galloway, 1998; Galloway et al., 2008). Given that increased N is occurring with relatively less change to phosphorus (P) cycling, ecosystem responses are best understood in the context of increased N:P of major biotic and abiotic components (Penuelas et al., 2020).The ratio of N to P in leaves has long been understood to impact decomposition processes and nutrient cycling (Koerselman and Meuleman, 1996; Güsewell and Freeman, 2005). As part of the initial decomposition processes, soil and surface-active arthropods physically and chemically alter litter, and multiple experiments have demonstrated food preferences for litter based on a variety of chemical and structural properties (David, 2014). The direct mechanisms by which microbial use of arthropod frass contributes to soil organic material cycling are not fully clear, but there is evidence that microbial litter colonization directly increases litter palatability to arthropods when microbes increase litter nutritive value (Gerlach et al., 2012), or indirectly when microbes detoxify plant secondary compounds that hinder arthropod consumption and digestion of said litter (David 2014). Another established link between arthropods and soil microbes is that N mineralization is consistently increased following arthropod litter processing (David, 2014). Therefore, it is reasonable to assume that altered litter and soil N:P are filters on arthropod communities, which is plausible given that arthropods are likely phylogenetically constrained in their stoichiometry and feeding guilds (Martinson et al., 2008; Ross et al., 2022), and different community assemblages and functional groups are expected to inhabit this chemically shifted environment (Tie et al., 2021; Deng et al., 2022; Nessel et al., 2023). If detritivores, and to some extent herbivores, increase in abundance from higher N litter, feedbacks to the soil system may promote further increases in N:P via decomposed frass and dung inputs (Wolters, 2000), and increased N mineralization rates (David 2014). Changes in litter input can also change soil moisture quantity and dynamics, which would feedback to the above scenarios (Wang et al., 2020; Liu et al., 2021). Thus, an invader with impacts on soil N dynamics may be expected to also alter arthropod communities in ways that further change soil N pools and transformations.Invasive N2-symbiont trees therefore provide a natural experiment to evaluate the effect of N:P shifts on surface-active arthropod diversity and potential feedbacks to soil. Arid riparian zones in the southwestern USA have been significantly altered by the establishment of non-native vegetation (Dukes and Mooney, 2004; Harms and Hiebert, 2006), including by a tree with N2-fixing actinobacterial symbionts, Russian olive (Elaeagnus angustifolia ; Bertrand and Lalonde 1985). This species was introduced to riparian woodlands along the Rio Grande of New Mexico, USA, between 1900 and 1915 (Hink and Ohmart 1984). A major ecological disruption in the form ofreduced disturbance was a consequence of dams, levees and diversions that altered dynamic flood regimes of the river, and the relatively static channels of the Rio Grande are less suitable habitat for vegetation communities evolved to frequent overbank flooding and sedimentation (Rood and Mahoney, 1990; Shah and Dahm, 2008). The tree rapidly spread throughout the Rio Grande and became a dominant component of riparian vegetation by 1960 (Campbell and Dick-Peddie 1964). Native N2-fixing symbiotic trees are known in the region (Shepherdia argentea ; silver buffaloberry) but are locally rare and exhibit much lower fixation rates than Russian olive (Petrides 1998, Follstad Shah et al. 2010).Several reports document N cycling alterations following Russian olive invasion, including increased N loss from litter mass decomposed compared to cottonwood (Simons and Seastedt 1999). De Cant (2008) isotopically demonstrated N2-fixation from Russian olive in a Rio Grande bosque, and ~5-fold increase in foliar N compared to adjacent cottonwoods (Populus deltoides var.wislizenii ), but the native trees did not utilize increased soil N from fixation inputs. Decomposing Russian olive litter (C:N = 13) was observed to produce spikes in N2O not observed from cottonwood litter (C:N = 22) under laboratory conditions (Duval et al. 2020). That study also reports increased N-processing enzyme (leucine amino peptidase, LAP) activity from cottonwood soils, suggesting potential microbial N limitation not found with Russian olive soils. However, Russian olive soils exhibited slightly higher acid phosphatase activity compared to cottonwood (Duval et al. 2020), suggesting P limitation.Russian olive introductions clearly have impacts on soil N inputs and cycling, and recent evidence suggests the presence or removal of the tree alters riparian tree-dwelling arthropod communities (West et al., 2023). Ground-dwelling arthropods are excellent taxa to explore invasive tree effects on ecosystem processes, because they are sensitive to fine-scale environmental conditions, and these conditions themselves are vital contributors to decomposition and nutrient cycling in riparian areas (Perry and Herms 2017). Indeed, detritivorous arthropods generally respond well with increased nitrogen and phosphorus levels, with isopods increasing in abundance (Coccia and Fariña 2022). Ellis et al. (1999) found the species composition and richness of middle Rio Grande bosque ground-dwelling arthropods to be similar between native cottonwood and saltcedar habitats, but cottonwood habitats supported greater densities of non-native isopods. However, it is generally unknown what the arthropod community looks like in areas dominated by Russian olive and areas where it has been removed. To improve our knowledge of the ecosystem-level impacts of Russian olive, we designed a study to compare ground dwelling arthropod relative abundance and diversity (taxonomic and functional) in Russian olive dominated stands versus native cottonwood woodlands, and the effect of restoration efforts by comparing Russian olive stands with plots where the tree had been mechanically removed. This allows us to evaluate longer-term (decadal) influence of an invasive tree compared to historic vegetation, as well as short-term impacts on the arthropod community driven by physical removal of plants (and associated ecosystem effects) where the soil chemistry has perhaps not yet appreciably changed.Because we are centering our work on shifting stoichiometry due to a plant invasion, we quantify functional (feeding) diversity of surface-active arthropods as well as taxonomic diversity. We hypothesize that cottonwood stands will have significantly different arthropod communities (beta diversity; taxonomic and functional diversity) than Russian olive stands, and those will be related to soil chemical changes induced by the invasive tree, such as greater N pools and increased N:P (DeCant 2008; Peñuelas et al. 2020). Stands where Russian olive has been removed are predicted to be chemically similar to extant Russian olive stands, but arthropod communities will differ between these areas due to physical character alterations from plant removal (moisture retention, soil temperature). Addressing these questions will provide additional insight into the ecosystem effects of altered N:P on biogeochemically relevant biota, and feedbacks related to invasive, symbiotic N2-fixing vegetation in an arid riparian zone.MATERIALS & METHODS
