We evaluated the use of + OM amendments at a wetland restoration site. Amendments have been recommended for restoration sites where the starting SOM content is low (NRCS 2010; Stelk et al. 2017; Ott et al. 2020). Our sandy loam soil, previously an agricultural field, fit the low SOM criteria (0.6% total carbon). The site was restored without the aid of + OM by re-establishing wetland hydrology, which allowed us to experimentally examine the effects of different types and loading rates of + OM.
Increased CH4 emissions are an important negative consequence of + OM treatments and was evident at our site. The control plots emitted an estimated 13.8 g CH4•m− 2•yr− 1 (Table 1), comparable to natural freshwater wetland emissions of 7–8 g CH4•m− 2•yr− 1 (Bridgham et al. 2006). Treatments with + OM had emissions as high as 100.9 g CH4•m− 2•yr− 1 (Table 1). These rates were on the high side compared to the other reported emission for natural and restored wetlands (4–126 gCH4•m− 2•yr− 1) (Nahlik and Mitsch 2010)(Bortolotti et al. 2016). Methane emissions were significantly higher in plots receiving high + OM loading rates, but low loading rates were not significantly different; therefore, limiting low loading rates would not be detrimental.
Whether + OM had been composted (i.e., aged outdoors) prior to application also impacted CH4 emissions. The highest CH4 emissions occurred with fresh + OM, manure and hay, whereas cured amendments (biosolids and wood chips) produced less. Higher CH4 production with fresh + OM has been noted previously (McNicol et al. 2020; Scott et al. 2022). For example, straw (Ballantine et al. 2015) and peat bales (Green et al. 2014) increase CH4 emissions, whereas composted wood and yard waste did not increase CH4 emissions (Winton and Richardson 2015). Increased CH4 emissions from fresh versus aged + OM has also been observed when the OM source was aboveground plant litter (Kandel et al. 2019), likely due to the higher cellulose content (Zak et al. 2015; Li et al. 2018). Organic material is commonly cured or composted to remove plant pathogens and to reduce the amount of cellulosic material (Hubbe et al. 2010), which competes for oxygen and contributes to phytotoxicity (Saidpullicino et al. 2007; Hu et al. 2017). Curing produces humic acids and increases the nominal oxidation state of the carbon (Guo et al. 2019). When cured material is subjected to anaerobic conditions, less CH4 is produced (Yao and Conrad 1999). Emissions of CH4 during early wetland ecosystem development is generally high even without + OM (Lee et al. 2017), and the CH4 emissions can be a primary factor that determines how long it will take for wetlands to switch from being sources of global warming to syncs. Without + OM, the transition from a net warming to a net cooling system can take decades, but elevated CH4 (from + OM) can increase the return period to centuries or longer (Hemes et al. 2018).
Our site experienced wet and dry periods, but local hydrology varied and Block C was almost continuously saturated (Fig. 1). With prolonged saturation CH4 emissions were lower overall (Fig. 1c) and did not differ due to temperature in either dry (p = 0.06) or wet (p < 0.001) months. The production of CH4 responded instead to rapidly changing water levels. Following soil re-saturation, CH4 production rebounded quickly (within a day or two, Fig. 1D) without the lag period predicted by others (Zak et al. 2015; Limpert et al. 2020; Scott et al. 2022), but consistent with a recent study showing soil oxidation events increase CH4 after re-wetting (Wilmoth et al. 2021), which could explains why Block C (constantly saturated), had lower annual CH4 emissions than Blocks A, B & D (Table 2). Fresh forms of + OM (M and H) produced the highest levels of CH4, suggesting + OM quality was a primary determining factor for CH4 emissions (Bridgham et al. 2013; Ballantine et al. 2015), although other factors may have contributed, including competition from alternate electron acceptors, diffusive flux from high water levels, or aerenchyma transport, particularly in cattail dominated plots (Bridgham et al. 2013; Brown et al. 2014).
We considered the effect of + OM on plant productivity. Productivity is associated with many ecosystem functions of wetlands, including carbon sequestration potential (Valach et al. 2021), but is not usually a primary metric for mitigation wetland success (Matthews and Endress 2008). Average above-ground biomass was higher with + OM (Table 1, Table S5 & Figure S1). This finding is expected as + OM add nutrients to wetland soils, leading to higher productivity (Steinbachová-Vojtíšková et al. 2006). We did not observe significant change in root growth with + OM, whereas other studies have reported reduced root growth (Dickinson 2007).
Plant diversity is often lower in amended soils due to excess nutrients. None of the differences in Shannon-Wiener Index (SWI) were statistically significant (p < 0.05); however, they were consistent with previous observations in that + OM with higher N & P (M & B) had lower SWI and + OM with the lowest N & P (H) had the highest SWI (Table 1 & S1) (Bedford et al. 1999; Zedler and Kercher 2004; Bailey et al. 2007; Scott et al. 2020). While an index such as SWI describes overall plant diversity, it has been suggested that the Floristic Quality Assessment Index (FQAI) is a better measure because it identifies the adverse impact of high nutrient environments, which favors non-native plant species (Kutcher and Forrester 2018). This was the case in our study as the manure amended plots, with the highest levels of both N and P, had a lower FQAI (p < 0.01). Hydrology also has a strong influence on the FQAI (Gallaway et al. 2019), which we also observed in our study (p < 0.0001) (Fig. 3). Block C (continuously saturated) had the lowest FQAI (5.05 ± 0.43), primarily due to the presence of cattail. Block D (dry) was also low (6.53 ± 0.56) due to the presence of barnyard grass, which, like cattail, has a low coefficient of conservation (leading to low FQAI). The two blocks (A & B) with higher FQAI (8.53 ± 0.59 and 7.60 ± 0.39, respectively) had intermediate hydrology – less inundation but did not dry. Recent work has shown that even hydrologically similar sites can have differing plant communities (Sueltenfuss and Cooper 2019): the hydrologic differences at our site were comparatively small, and would have fit that publication’s definition of “similar”. Our study suggests that + OM can accentuate differences caused by minor differences in hydrology, and combined may lead to different plant community compositions.
We monitored several soil parameters following + OM, and found them problematic, both in their connection to other observed properties and as predictors of long-term outcomes. Soil organic matter (SOM) is a vital soil parameter and the most commonly reported value in + OM wetland studies (Scott et al. 2020). There is evidence that + OM in soil will lead to incorporation of some of that material as SOM (Bruland and Richardson 2004), so the question is whether or not those increases are sufficient to merit + OM. We saw evidence of SOM priming in control plots, likely due to site disturbance (Werkmeister et al. 2018). In October 2019, shortly after the plots were constructed, the SOM in the unamended plots was 32.2 ± 1.3 mg•cm3 which fell to 29.1 ± 1.4 mg•cm3 in November 2020 despite an estimated 1.5 ± 0.3 mg•cm3 of new root growth (Table 1). We observed that manure at 678 m3•Ha-1 and wood chips at 339 m3•Ha-1 loading rates were sufficiently beneficial to overcome short term oxidative SOM loss (Figure S3). Long-term studies are better suited to evaluate the effect of + OM on SOM, and they have consistently shown that the + OM is lost over time (Scott et al. 2020) and is not needed for soils to accumulate SOM (Ott et al. 2020). Of the other soil parameters we monitored, +OM changed bulk density (Db), but the differences were not meaningful at our site, and + OM did not influence the outcome of hydric soil tests.
Evaluating the effect of + OM is challenging because there are many complex interactive effects. At our site, cattail, a dominant nuisance species, provides an example of how these effects can compound. We saw cattail re-growth (after herbicide treatment) following the soil disturbance necessary for + OM. Disturbance is often cited as one of the main facilitators of non-native species (Lang et al. 2015; Cordell et al. 2016; Johnson 2016). Nutrients from + OM also influence cattail growth, with P increasing prevalence (Newman et al. 1998). Once present, cattail out-competes other species, decreasing diversity, richness (Lishawa et al. 2019), and floristic quality (Venterink et al. 2003). Flooding favors cattail growth (Newman et al. 1998), and our site had enough hydraulic variation that we observed this effect (p < 0.005). Once cattail is established it can become persistent. Cattail became dominant at our site and that dominance has continued through at least two subsequent seasons. Species like cattail can have a strong influence over other variables: it is a conduit for CH4 release (Lawrence et al. 2017) and the presence of rhizomes skew SOM, Db, and root growth data. In plots where cattail was present (e.g. manure 339 & 678 m3•Ha− 1), rhizomes accounted for over 90% of the total root mass – several orders of magnitude greater than medium and fine roots, and we had to exclude rhizomes to avoid skewed data. Above-ground biomass increases due to higher levels of N, but this has also been show to increase CH4 emissions (Kandel et al. 2019). One way to evaluate + OM in the context of multiple factors is to use a scoring system where all metrics are given equal weight, which we have done previously (Scott et al. 2020). Using this method, we would rank moderate hay loading (H = 170 m3•Ha− 1) and higher levels of wood chip loading (339 & 678 m3•Ha− 1), as the preferred OM amendment choices (Table S6). Even so, all amendments we tested had negative, unintended consequences such as increased CH4, lower diversity, and in the case of manure and hay, an increase in non-native species (Table S6). Managers would need to weigh the acceptability of these negative outcomes in deciding whether to use + OM (Ballantine et al. 2015).
Studies have suggested other possible benefits of using + OM amendments include are increased soil nutrient supplies and cycling, supporting plant growth, and more robust microbial communities (Richardson et al. 2016). According to the Maryland wetland mitigation guidance, OM amendments are “needed to meet hydric soil characteristics and maintain the desired plant species” (Walbeck et al. 2011). In this field study, we did not observe these benefits, and more often saw the opposite effect. The increase in nutrient supply from + OM stimulated the growth of undesired species, particularly cattail (Typha sp.), which is a noted restoration issue(Price et al. 2019; Keyport et al. 2019), reducing both diversity and floristic quality. Amendments did not meaningfully lower the soil Db, did not increase SOM, and had no observable effect on hydric soil indicators (a proxy for microbial activity). The sandy loam soils at our site were low in SOM (0.6%, Table S3), conditions where + OM is expected to have the most benefit, so we would expect fewer benefits in higher quality soils. Despite our efforts to control hydrology and focus on OM, this surface-water-fed wetland had variable water infiltration rates, and hydrology varied between study plots even within the same Block. Hydrology was the overriding factor for evaluation metrics, and + OM only made the influence of hydrology more apparent. We did not observe an influence of + OM on localized hydrology, such as increased moisture retention that delayed re-oxygenation. While our study did suggest that some + OM materials may have net benefits, the choice of amendment would necessitate professional judgment based on the specific site conditions and a willingness to accept negative consequences (Ballantine et al. 2015). Avoiding soil disturbance and greater attention to the site hydrology may be a more effective means of controlling undesired species and limiting methane emissions than adding organic amendments.