Quantifying biotic resistance of different native plant functional groups at different seeding densities is a critical step in enhancing wetland revegetation (Sloey et al. 2023). In this experiment, we manipulated native plant functional group identity and native seeding density to determine how those factors individually and interactively might drive biotic resistance. We found three key findings from this research. First, surprisingly and unfortunately, it appears that extremes in water depth during the two study years (drought in 2022 and flooding in 2023) overwhelmed the seeding treatments to drive plant community composition and reduce any effect of biotic resistance. Yet, these hydrologic extremes, which are becoming increasingly common due to climate change and human water withdrawals (IPCC 2023; Wurtsbaugh et al. 2017), provided an opportunity to evaluate the recovery potential of the various functional groups. In turn, we identified a subset of species (B. maritimus and S. acutus from the bulrush functional group and D. spicata from the grass functional group) that were able to survive the extreme weather conditions of both growing seasons. Second, we found indications that different emergent wetland restoration sites can have widely diverging plant community recovery despite geographic proximity. Although in the same watershed, the two field sites differed in final cover and overall plant community. We hypothesize that these site-specific differences are related to differences in salinity and hydrology at the two locations. Finally, we suggest that these findings underscore the importance of seeding diverse seed mixes as a bet-hedging strategy. As unpredictable extreme weather events and hydrologic fluctuations become more common in wetlands, there will be a greater need for bet-hedging strategies to promote native growth regardless of the abiotic conditions that arise.
Water depth appeared to overwhelm seeding treatments
Drought can be a major constraint to successful restoration as water availability can lead to rapid mortality of the more vulnerable seed and seedling stages of plants relative to more established plants (Doherty and Zedler 2015; Buisson et al. 2021; Hebert 2022). On the other hand, flooding can drive wetland community assembly because different species germinate and survive at different water depths with associated variations in light, temperature, and oxygen (Casanova and Brock 2000; Fraser and Karnezis 2005; Webb et al. 2012; Rosbakh et al. 2020). Flooding that is too long or too deep, however, can constrain seeds and seedlings of many wetland species due to anoxia-driven inhibition of germination (Fraser et al. 2014; Kettenring and Tarsa 2020; Rosbakh et al. 2020) and mortality of seedlings that do not yet have sufficient aerenchyma developed to support oxygen exchange and aerobic respiration (Moor et al. 2017). Reduced light levels and muted temperature fluctuations, also associated with deeper flooding, inhibit seed germination and seedling survival of many wetland species (Fraser et al. 2014; Kettenring et al. 2006; Kettenring and Tarsa 2020; Rosbakh et al. 2020).
In the present study, hydrologic extremes, both severe drought and prolonged, and at times deep, flooding, appeared to overwhelm the functional group and seeding density treatments at both sites during both growing seasons. The ability to compare hydrologic data to plant growth outcomes is complicated to evaluate and has rarely been tested (Brudvig et al. 2017; Sueltenfuss and Cooper 2019). These methodological limitations (e.g., linking hydrology to delayed plant responses; collecting plant and hydrology data on compatible spatial and temporal scales) are particularly problematic given that the inability to manage hydrology adequately for plant communities is a major reason that restoration projects fail (Bohnen and Galatowitsch 2005; Galatowitsch and Bohnen 2021) Although we cannot statistically test for the effect of inundation on the plant community outcomes here—because the plant and water level data were collected on different spatial scales (plot vs. block level)—common sense suggests that the extreme hydrologic conditions (severe drought and prolonged flooding) restricted the growth of native plants (i.e., total plant cover was low for both years). At Farmington Bay, most of the seeded species never emerged and the majority of those that germinated in the first growing season were not present during the second growing season. Although emergent species can survive a range of hydrologic conditions, the set of conditions under which seedlings can germinate and grow is narrower (Rosbakh et al. 2020). One study of germination from a wetland seed bank found that growth was reduced across the entire seed bank when water was 6 cm below the soil surface, and greatly reduced when water was 2 cm or greater above the soil surface (Fraser and Karnezis 2005). In the present experiment water levels were 8–76 cm below the soil surface in 2022 and 8–25 cm above the soil surface in 2023, suggesting that few seedlings were able to germinate and survive in 2022 and any that were able to emerge in 2022 were likely unable to establish in 2023. At Utah Lake, the plots were flooded by > 1.5 m of water in year 2 over the entire growing season so the plots themselves were not identifiable and certainly the seedlings would not have survived if they had withstood the drought in year 1.
Invasive species are successful in part because they can survive broader and more extreme conditions than native species (Zedler and Kercher 2004; Hovick et al. 2023), making it important to identify native species traits that contribute to tolerance or resilience to a range of conditions in wetlands (Moor et al. 2017). In fact, during the second growing season, the research plots were dominated by invasive species, suggesting that these species were better able to survive the extreme hydrologic conditions while many of the native species were not. This was true even though all species that emerged at Farmington Bay during the second growing season (seeded, non-seeded native, and invasive) were all wet-loving species, as shown by their wetland indicator statuses of facultative, facultative wetland, and obligate. This finding is consistent with those of Byun et al. (2020) who, in a study of species and functional diversity effects on biotic resistance in a freshwater wetland, found that P. australis had increased invasion success in wetlands following flooding.
However, the extreme weather events presented an opportunity to identify species that were able to survive these extremes, suggesting that they may be good focal species for wetland restoration. We found that three species (B. maritimus and S. acutus from the bulrush functional group and D. spicata from the grass functional group) have potential for surviving the hydrologic extremes present in the field sites. All three species are valued by managers for their habitat benefits (Downard et al. 2017; Rohal et al. 2018), thus they align with habitat restoration goals and also represent a range of hydrologic tolerances (Table 1), likely contributing to their tolerance or resilience to hydrologic extremes. Bulrushes in particular showed high potential for growing under a variety of environmental conditions. Bulrushes (including B. maritimus and S. acutus) were the only species for which seeding had a measurable effect on native cover in the study. This finding is consistent with other studies that found that B. maritimus can tolerate water levels ranging from drought to flooding as well as highly alkaline and salty soils (Downard et al. 2017; Houde 2024), and the seeds can germinate even when below the water surface (Clevering 1995). Furthermore, S. acutus was the only seeded species to increase over time during both growing seasons in the present study. Although S. acutus typically prefers flooding (Stewart and Kantrud 1972; Downard et al. 2017), these results suggest it has the potential to establish and spread under broad hydrologic conditions. Finally, D. spicata, a perennial grass, was also able to survive both years of extreme weather at Farmington Bay. This finding is consistent with previous studies that found D. spicata to successfully survive both regular inundation (Stalter et al. 2022) and extreme drought (McKee et al. 2006).
Different species had higher performance at different sites
Even in sites that are geographically close and experiencing the same climate, different restoration outcomes can result from small changes in biotic and abiotic conditions unique to an individual site or even microsite within a site (Young et al. 2015; Fried et al. 2018). In the present study, the two sites differed somewhat substantially in their unseeded native community, overall plant cover (cover was higher at Utah Lake than Farmington Bay during the first growing season), and the relative abundances and distributions of seeded species between the two sites. Differences in the plant communities could be attributed to any number of biotic and abiotic factors but salinity or distinct hydrology are the most logical. Although not measured in this experiment, Farmington Bay likely has higher salinity than Utah Lake given its location along the hyper-saline Great Salt Lake. Although these wetlands at Farmington Bay are not currently connected hydrologically with the lake (because the lake is much reduced in elevation), Great Salt Lake wetlands have been subject to flooding by the hypersaline lake in past decades (Null and Wurtsbaugh 2020). Different wetland species exhibit different salinity tolerances (Tootoonchi et al. 2023), and elevated salinity can be an impediment to plant growth, even for plants adapted to those conditions (Dethier and Hacker 2005). Here we found that Utah Lake plant communities had a higher number of unseeded native species overall but a lower proportion of halophytic species like D. spicata. Alternatively, these differences might be explained by the hydrology of the two sites, specifically long-term lake level fluctuation. Lakes that fluctuate yearly have higher diversity as differences in abiotic conditions allow for more niches and increased germination from the seed bank (Keddy and Reznicek 1986; Wilcox and Nichols 2008), a fact that may explain why more species emerged from the background community at Utah Lake than Farmington Bay. Although both sites experienced drought in 2022, Utah Lake is known to have yearly fluctuations in lake level due to evaporation and upstream water-diversions (Richards 2019). Although Farmington Bay also experiences water level fluctuations (as was seen in this experiment), managers at Farmington Bay have coarse control over water level to lessen the effects of water level fluctuations on restoration outcomes (Downard et al. 2017), as seen when managers released water into the experiment in early August 2022. Future directions for restoration research around the Great Salt Lake Watershed should focus on the effects of a wide range of environmental variables on restoration outcomes.
Diverse seed mixes can act as a bet-hedging strategy for restoration success
Given the difficulty of predicting environmental conditions within wetlands (Bohnen and Galatowitsch 2005), it is important to incorporate diversity on multiple levels of a restoration project to serve as a bet-hedging strategy in the face of increasing unpredictability (Pastorok et al. 1997; Doherty and Zedler 2015), particularly with increasing hydrologic extremes due to climate change and human water withdrawals (IPCC 2023; Wurtsbaugh et al. 2017). The use of a bet-hedging strategy (also called a portfolio approach; Schindler et al. 2015), such as seeding an array of species, is particularly important when creating restoration plans. Biological systems are scale-dependent and thus more volatile when viewed on smaller scales (Schindler et al. 2015). Given that extreme weather events are unpredictable and restoration projects often take place on small geographic scales, portfolio approaches can be particularly important for increasing resilience to environmental extremes in restoration projects to combat this increased volatility (Zabin et al. 2022). The results of this experiment exemplify how seeding species from different functional groups and hydrologic tolerances can serve as a form of bet-hedging. Although we did not design this experiment to test the effects of bet-hedging, seeding a variety of species (i.e., across the functional group identity treatments) resulted in some native growth during both hydrologic regimes, something that may not have occurred if only flood-tolerant or drought-tolerant species had been introduced. Seed mixes should include a wide variety of species from different functional groups to ensure that restoration goals can be met regardless of weather and site conditions and to avoid biotic homogenization, which can lead to a loss of diversity and ecosystem functioning (Holl et al. 2022; Luong et al. 2023). Furthermore, seeding a variety of species allowed us to ensure native growth at both locations even without precise knowledge of the future abiotic conditions. A bet-hedging strategy can be further enhanced with reseeding sites over multiple years and breaking dormancy in only a subset of the seed mix to increase the probability of aligning seed and seedling abiotic requirements with favorable environmental conditions for establishment (Doherty and Zedler 2015; Rader et al. 2022; Zabin et al. 2022).
Interestingly, we broke dormancy for all seeds in the seed mixes under the presumption that the seeds would be sown at the optimal time for germination and seedling establishment. Yet, we did not foresee the hydrologic extremes facing the restoration sites post-seeding. Treating a fraction of the seed lot would have been an additional, but not insurmountable, logistical hurdle that might be feasible for others to implement in wetland seedings thereby building up the seed bank for the future when environmental conditions are better suited for seeds (Evans and Dennehy 2005). Re-seeding sites multiple years would add substantial costs to any restoration projects but may be increasingly required given a future with a higher probability of restoration failure in a given year with environmental extremes (Kettenring and Tarsa 2020; Groves and Brudvig 2019). Given the difficulty of predicting abiotic conditions at any given site within a short temporal scale, it is important to incorporate bet-hedging to ensure restoration success regardless of conditions. This research was a useful step in exemplifying the importance of this strategy for the many restoration projects where year to year hydrologic extremes may ultimately drive outcomes.