Biochar improves soil quality prior to prairie restoration

Intensive agriculture has led to widespread losses of native grasslands. Restoration projects are trying to reestablish native communities. However, the poor quality and altered food webs of former agricultural soils act as barriers to establishing diverse plant communities. Soil amendments such as biochar have been used to improve soil quality and increase crop yield, but little research has been on biochar in restoration settings. Eastern Washington University is attempting to restore a former wheat field to native Palouse prairie vegetation. In this study, we compared (1) soil abiotic properties and food webs between Palouse prairie remnants and Eastern Washington University's (EWU) restoration site and (2) tested whether biochar can move soil conditions at EWU's restoration site closer to those of remnant sites. We compared soils from three Palouse prairie remnants and EWU's restoration site, and compared soils at the restoration site that were treated with different levels of biochar. Soil nematodes were extracted and sorted to functional groups as an assay of the structure and functional diversity of the soil food web. The soil at EWU's restoration site had significantly lower nematode abundance, pH, and organic matter compared to prairie remnants. In addition, nematode abundance and pH increased significantly at the EWU site with the addition of biochar, both increasing to conditions similar to what is found at the prairie remnant sites. These results suggest biochar can improve some soil conditions in former agricultural lands even prior to native plant restoration.


Introduction
The physical properties of soil, such as moisture, organic matter, pH, and texture, act as strong filters on plant diversity and community composition (Faucon 2020).Higher soil moisture often favors greater plant diversity (Deng et al. 2016) and soil organic matter affects the water holding capacity of soil, soil structure, and nutrient retention (Chapin et al. 2012).Furthermore, the availability of nutrients is strongly influenced by pH, with the majority being most available in neutral soils (Brady 1990).
Soil texture, pH, moisture, and organic matter also strongly influence soil microbial communities, including bacteria, fungi, and nematodes (Kandji et al. 2001;Fierer 2017).These soil food webs maintain primary production and plant diversity in ecosystems through their roles in nutrient cycling and plant-soil feedbacks (van der Heijden et al. 2008).Plants are generally nitrogen and phosphorous limited, and rely on the mineralization of these nutrients by microbes within the soil (Elser et al. 2007).In addition, the soil microbe and plant communities are closely connected; removing one species or functional group out of one community can alter the composition of the other (van der Putten et al. 2013).
Conventional agricultural practices, including the use of intensive tillage, chemical fertilizers, pesticides, and monoculture cropping have significant negative impacts on the functioning and quality of soils (Arriaga et al. 2017).As soils are aerated by tilling, microbial activities spike leading to an increase in the breakdown of soil organic matter (Alvarez et al. 1995).This loss of organic matter can lead to a decrease in soil pH (Bai et al. 2018) and reduce water holding capacity (Chapin et al. 2012).The nitrification of ammonium-based fertilizers by bacteria decreases soil pH (Vitousek et al. 1997).Compared to crop-rotation, monocropping also leads to reduced soil organic matter, nutrients, and structure, and an increased need for fertilizers (Ewel et al. 1991;Salaheen & Biswas 2019).Although these practices have contributed to the tremendous productivity of agriculture in recent decades, the negative impacts of conventional agricultural methods are felt throughout the ecosystem, from soil microbes, to the plants which depend on the soil food web, and the aboveground communities they support.
Over time, restoring native plant communities on land used for crop production can reconstruct the organic carbon pool in the soil and support beneficial microbial activity, improving soil quality (Guzman & Al-Kaisi 2010;Maher et al. 2010).However, depending on how degraded the soil is, it may be hard for native plants to establish.Soil that has been in crop production may not have the microbial community or quality needed to support diverse native plant, such that it may take many years to reestablish a diverse native plant community.
Restoration of soil food webs and native plants may be accelerated by focusing on restoring soil quality at a site through the use of soil amendments.Biochar is a soil amendment that is made by burning biomass in the absence of oxygen, a process called pyrolysis (Porter & Laird 2021).Biochar has been used by many farmers to improve the health of their soils and increase crop yield (Rosenani et al. 2014).Agricultural fields tend to have acidic soils and by incorporating biochar, which is slightly alkaline, the pH of the soil can be raised, increasing the availability of essential nutrients within the soil.Biochar has been shown to increase soil carbon sequestration, moisture retention, and nutrient content (Zhang et al. 2020), increasing the overall quality of crop fields with poor soil conditions (Liu et al. 2020).Although there is much less research on using biochar in restoration settings, it has been shown to benefit grassland restoration in extreme conditions when used in combination with compost and mycorrhizal inoculation (Ohsowski et al. 2018) and to a native prairie grass without benefitting invasive Sericea lespedeza (Adams et al. 2013).Other recent restoration projects have found that both incorporation and mulch application of biochar may benefit seedling establishment, while also having a strong and rapid effect on plant communities and soil nutrients (Van de Voorde et al. 2014;Phillips et al. 2020).
The Palouse prairie is one of the most endangered ecosystems in the world.Beginning in the mid-1800s, this prairie was converted from diverse grasses, shrubs, and wildflowers to agriculture (Black et al. 1998).Remaining prairie is highly fragmented and overrun with non-native plants.In response to this major loss of biodiversity, Eastern Washington University (EWU) is planning to restore 52-ha of farmland back to native Palouse prairie.A 6-ha pilot site has been developed to test the use of biochar in restoring the Palouse prairie.
The goals of this project were to compare soil abiotic properties and the soil food webs between Palouse prairie remnants and EWU's prairie restoration site and test whether biochar can move soil conditions at EWU's restoration site closer to the conditions of remnant soils, in the hopes of enabling better plant establishment.Soil nematodes were used as a measure of soil food webs because they interact closely with soil microbes, play important roles in nutrient cycling (Zhu et al. 2018), have easily identifiable functional groups (Yeates 1999), and have long been used to assess soil conditions (Bongers 1999).
We hypothesized that: (1) Soil quality in terms of pH, soil moisture, and soil organic matter would be lower at EWU's restoration site compared to intact prairie remnants.(2) Soil nematodes would be more abundant and nematode functional group composition would be more diverse at intact prairie remnant compared to EWU's restoration site.
(3) Biochar amendments would improve soil quality and nematode community composition at EWU's restoration site.

Methods Study Area
The Palouse prairie is characterized by rolling hills of deep loess soils, average annual precipitation of 25-76 cm, mostly occurring during the winter, and average annual temperature of 7-12 C (Daubenmire 1992).In this study, we collected soil from EWU's prairie restoration site and three Palouse prairie remnants: Kamiak Butte County Park, Steptoe Butte State Park, and Department of Natural Resource (DNR) land near Medical Lake, WA, U.S.A. (Fig. 1).EWU's restoration site is located in Cheney, WA, U.S.A (47.488N 117.594W).Strips (7 Â 72 m) of biochar (Ag Energy Solutions, Spokane Valley, WA, U.S.A.) were applied during the fall of 2020 as part of a 6-ha pilot site.We set up 12 biochar strips, arranged in four experimental blocks.Each strip was top-dressed with biochar 6 months prior to sampling and minimal mixing occurred.Qualterra, a clean ag-tech company based in Pullman, WA, provided the biochar used in this project.The biochar was produced through a process of gasification (800-1,000 C).The high temperature produces a char with a high percentage of crystalline carbon and less free carbon.Within each experimental block, strips were assigned to each of three treatments: high biochar concentration (6,725 kg/ha), low biochar concentration (673 kg/ha), and control with no biochar.Low biochar treatment was chosen to mimic typical applications in agricultural settings in the area, and the high biochar treatment was intended to see if there were any negative impacts if trying to maximize carbon sequestration.EWU's restoration pilot site was seeded in the fall of 2021, after all of the samples in this study were collected.Kamiak Butte County Park is one of the largest remaining Palouse prairie remnants (Looney & Eigenbrode 2012).It is located at 46.867 N 117.152W in Whitman County, WA and sits at an approximate elevation of 967 m.Steptoe Butte is located in Colfax, WA, at 1045 m of elevation.The portion of Steptoe Butte that we sampled was located at 47.034 N 117.289W. The DNR land that we sampled was located at 47.568 N 117.636W, 819 m in elevation, just east of Medical Lake, WA.The three remnant sites have silty loam soil, while the restoration site has silty clay loam (Table S1).The prairie remnants had high plant richness (80 species collectively) and were dominated, in cover and richness, by native perennial species.EWU's restoration site, by contrast, had only 5 plant species, primarily invasive annuals, and low total plant cover (Snyder 2022).

Field Methods
We sampled EWU's restoration site monthly from March and July, 2021, by collecting three soil cores (subsamples) from each biochar strip.Soil cores were collected using a 5.08 cm diameter soil corer at a depth of 10 cm.The corer was cleaned between strips to eliminate contamination between biochar treatments.We stored each of the soil cores at 8 C until analysis.
Each of the three remnant sites were sampled using a stratified random sampling design.Remnants were sampled in June and July, 2021.Each site was blocked into north facing slopes, south facing slopes, ridges, and troughs.Three soil cores per aspect were taken at randomized locations, using ArcGIS Pro's random point generator, for a total of 12 soil cores per remnant site.Remnant sites were sampled the same weeks as EWU's restoration site.Soil cores were collected with the same equipment and were stored using the same methods used to store cores from EWU's pilot site.Soil was stored in a cold room for no more than 2 weeks prior to analysis.

Laboratory Methods
Soil moisture and organic matter were determined gravimetrically.We weighed soil samples, dried them at 50 C for 2-3 days, and then reweighed them to estimate soil moisture.Dried soil samples were then combusted at 450 C for 2 hours to estimate soil organic matter (Salehi et al. 2011).Soil pH was determined using an EcoSense pH10A probe in a 1:2 slurry of soil and DI water (Robertson et al. 1999).
From each sample, 40 g (AE2 g) of field moist soil were used to extract nematodes using Baermann funnels.Nematodes were extracted over 2 days at room temperature and then stored at 8 C until sorting.Sorting was done in two steps; first using a gridded petri dish and a dissecting microscope, we counted all nematodes to determine total abundance.Then the samples were centrifuged at 600 rpm for 6 minutes to concentrate the nematodes.We then sorted a portion of the nematodes (10% of the sample or 30 total nematodes, whichever was more) into functional groups under a compound microscope.The functional groups which we used for sorting included bacterivore, fungivore, plant-feeding, plant-associated, omnivore, and predator (Yeates 1999).All nematode abundances were expressed as nematodes per gram of dry soil.Functional group diversity of nematodes was quantified using the Shannon Index.

Analysis
To test if environmental variables were significantly different between treatments or sites, we used factorial analysis of variances with Tukey post hoc tests.First, we compared soil variables at the three prairie remnants and the control treatment from EWU's restoration site, with month and the interaction between month and site as additional fixed factors.We then compared soil variables between biochar treatments at EWU's prairie restoration site, with month and the interactions between month and treatment as fixed factors, and block as a random factor.The relationship between soil nematode abundance and soil environmental factors at the restoration site was evaluated using linear regression with month as fixed factor.All statistics were performed in R version 4.2.1.

Results
Soil pH at EWU's restoration site was significantly lower than all of the prairie remnants (Table S2; Fig. 2A, p = 0.0009).At the restoration site, soil pH was higher in the high biochar strips, but the low biochar and control strips did not differ (Table S3; Fig. 3A, p = 0.0001).Soil organic matter at the remnant sites ranged from 7.94 to 10.48%, but was lower at EWU's restoration site (3.915AE 1.42%, mean AE SD; Table S2; Fig. 2B, p = 0.0008) and did not differ between biochar treatments (Table S3; Fig. 3B, p = 0.21).Soil moisture did not differ between sites, and there was no overall difference between biochar treatment plots.
EWU's restoration site had lower soil nematode abundances than Medical Lake and Steptoe Butte, but did not differ from Kamiak Butte (Table S2, p = 0.0026).Nematode abundances did not differ between remnant sites (Fig. 2C).Soil nematode abundance increased with the addition of biochar (Table S3; Fig. 3C, p < 0.0001).As with total nematode abundance, bacterivores, fungivores, and plant feeders were more abundant in the high biochar treatment, while plant associates did not differ, and omnivores were more abundant at the low biochar treatment (Table S3).Bacterivorous nematodes were the most abundant functional group of nematodes at all sites, followed by fungivores, then plant feeders.Remnants had more omnivores than plant associates, but the restoration site had more plant associates.While bacterivores and fungivores did not differ in abundance between sites, plant feeders and associates were less abundant at EWU than at Steptoe Butte (Table S2).Omnivores were less abundant at EWU (0.40 AE 0.24 per g dry soil) than all remnant sites (2.20 AE 0.93, Table S2, p < 0.0001).Predatory nematodes were found at very low abundance at remnant sites, but were not found at the restoration site.Therefore, predators were omitted from analysis.Functional group diversity of nematodes was lower at EWU than at all of the remnant sites (Table S2, p < 0.0001).
Controlling for month, soil moisture was strongly related to soil organic matter at EWU's restoration site (Table S4; Fig. S1).Both soil moisture and organic matter also had positive relationships with nematode abundance (Table S4; Figs.S2 & S3).

Discussion
Soil quality at the EWU restoration site was degraded in terms of pH, SOM, and nematode abundance compared to intact prairie remnants.Reference remnants also had a greater diversity of nematode functional groups, due to the presence of predatory nematodes and the higher relative abundance of omnivorous nematodes.The addition of biochar partially mitigated the degraded soil quality in ways that may facilitate subsequent plant restoration.The high biochar addition increased soil pH and nematode abundance at the restoration site, but did not affect SOM.
Soil series of the Palouse prairie (e.g., Palouse, Latah, Garfield) typically have neutral pH (National Cooperative Soil Survey 2013), but agricultural practices are known to acidify soils.The soil at EWU's restoration site was acidic compared to reference sites, likely due to conventional wheat production at the site.Biochar is alkaline, however, and can increase the pH of acidic soils (Chang et al. 2021).At our site, the high biochar treatment was closer to neutral, and within the range found at reference sites.Local prairie remnants averaged 133% more SOM than the restoration site.Surprisingly, the addition of biochar had no detectable impact on SOM, despite the biochar being organic material that loses 67% of its mass on ignition (McCullough unpublished data).It is likely, however, that we lacked statistical power to detect the effect of biochar addition on SOM.The difference between means of control and high biochar treatments for SOM was only 0.2%, while the standard deviation was 0.8%.We calculate that the immediate impact of high biochar addition would have been to increase SOM by 0.3-0.8%,depending on the depth to which the biochar was incorporated and the bulk density of the soil.This suggests that some of the biochar was lost from the plots in the 6-10 months after it was added, and that the background variation in SOM was large relative to the organic matter in the remaining biochar.
Prairie remnants had higher nematode abundance, suggesting that the EWU site had lower soil microbial productivity and diversity.Plant diversity has been shown to be correlated with soil mutualists, including beneficial nematodes (Dietrich et al. 2021).There was a large difference in plant species richness and abundance between the EWU restoration site and the prairie remnant sites, which could further explain differences in soil properties found between sites.Biochar amendment increased nematode abundance at the restoration site, suggesting that biochar can improve soil productivity in the short-term even in the absence of plants.Other studies have shown that biochar and other organic amendments can increase nematode abundance (Liu et al. 2020), but much of that effect is thought to be mediated by plant responses to the amendments.Our study shows the benefit is immediate and can occur prior to restoration seeding efforts.The mechanism through which biochar increased nematode abundance at the restoration site is not entirely clear.Nematode abundance was correlated with soil moisture and SOM, but biochar addition did not significantly affect those factors within our study period.Biochar improved soil pH, but nematode abundance was not correlated with pH at our site.This may indicate that nematode abundance responded to other effects of biochar, likely mediated by soil microbes.Biochar addition has been shown to affect exoenzyme activity in the soil (Fehmi et al. 2020) and soil microbes are sensitive to pH (Fierer 2017), so microbial communities likely differed substantially between treatments.
Prairie remnants had a larger proportion of omnivorous nematodes than EWU's restoration site, as well as having predatory nematodes.Although biochar addition increased the total abundance of nematodes, it did not alter the nematode functional group composition at the restoration site.Omnivorous and predatory nematodes have longer generation times and are more K-selected than other functional groups of soil nematodes (Bongers 1999).It is likely that these higher trophic level nematodes will take longer to recover.Predatory and omnivorous nematodes regulate populations of other nematodes, including plant parasites, as well as microbial community structure and function (Thakur & Geisen 2019).Future research should focus on methods of accelerating the recovery of these important functional groups, perhaps through soil inoculations from reference sites, as has been used for soil microbes (Wubs et al. 2016), or by addressing soil physical properties, such as compaction, that may impact larger soil fauna (Bouwman & Arts 2000).In addition, we were unable to test soil nutrients in this study, but it would be a good idea for future research to consider soil nutrients as well.
This study demonstrates that amending degraded soils with biochar can improve soil quality, which can be beneficial at the onset of restoration projects when the establishment of native vegetation may be hindered by initial site conditions.However, applying biochar amendments at the densities found to be successful in this study may be cost-prohibitive when scaling up to a larger restoration site.Biochar can cost anywhere between $400 and $2000 per ton, which can become costly when amending larger fields (Drost 2021).As an alternative, land managers and restoration practitioners can invest in producing their own biochar on-site.However, this alternative requires more upfront labor.More research is needed to investigate the long-term effects of lower biochar application rates on soil quality and primary production.

Figure 1 .
Figure 1.The location of sampled sites within eastern Washington, U.S.A. Soil was collected from EWU's prairie restoration site and three intact Palouse prairie remnants.

Figure 2 .
Figure 2. Soil pH (A), organic matter (B), and nematode abundance (C) at sampled sites.EWU = Eastern Washington University restoration site (in white).The other sites, in gray, are all remnant prairie sites.Bars show the mean of eight samples (four for pH), + SD.Sites with the same letter did not differ at the 0.05 level.

Figure 3 .
Figure 3. Soil pH (A), organic matter (B), and nematode abundance (C) in the three biochar treatments at the EWU restoration site.Control are plots with no added biochar.Bars show the mean of 20 samples (8 for pH), + SD.Sites with the same letter did not differ at the 0.05 level.