Lipid‐enhanced Oilcane does not impact soil carbon dynamics compared with wild‐type Sugarcane

The carbon neutral potential of bioenergy relies in part on the ability of feedstocks to sequester carbon (C) in the soil. Sugarcane is one of the most widely used bioenergy crops, yet there remain unknowns about how it impacts soil C dynamics. In addition, Oilcane, a genetically modified version of Sugarcane has been produced to accumulate more energy‐dense oils and less soluble lignin, which enhances conversion efficiency but may also impact soil C cycling. Thus, our objectives were to examine the impact of Sugarcane litter decomposition on soil C formation and losses and determine if the genetic modifications to produce Oilcane alter these dynamics. To do this, we incubated bagasse (processed stem litter) and leaf litter from Sugarcane and Oilcane in microcosms with forest soil for 11 weeks. We used differences in natural abundance δ13C between C3 forest soil and C4 litter to trace the fate of the litter into respiratory losses as well as stable and unstable soil C pools. Our results show that genetic modifications to Oilcane did not substantially alter soil C dynamics. Sugarcane and Oilcane litter both led to net soil C gains dominated by an accumulation of the added litter as unstable, particulate organic C (POC). Oilcane litter led to small but significantly greater net soil C gains than Sugarcane litter due to greater POC formation, but the formation of stable, mineral associated organic matter (MAOC) did not differ between crop types. Sugarcane and Oilcane had opposing effects on tissue type where Sugarcane bagasse formed more MAOC, while Oilcane leaves preferentially remained as POC which may have important management implications. These results suggest that genetic modifications to Sugarcane will not significantly impact soil C dynamics; however, this may not be universal to other crops particularly if modifications lead to greater differences in litter chemistry.


| INTRODUCTION
Maximizing the carbon-neutral potential of bioenergy as a fuel source relies on the ability of the feedstocks to sequester carbon (C) in the soil (Mathews, 2008). Currently, Saccharum officinarum (herein Sugarcane) is one of the most widely used bioenergy feedstock due to its high biomass yields, conversion efficiency, and adaptability for a wide range of growing conditions (Hoang et al., 2015;Lam et al., 2009). The conventional Sugarcane harvest method of burning the field to remove the leaves for easier harvest of the stems diminishes the crop's carbon-neutral potential by increasing C emissions. Recently, there has been a shift toward more sustainable harvesting practices that leave the leaf litter on the field rather than burning it as well as adding the processed stem litter (bagasse) back as a soil amendment. These sustainable harvesting practices significantly increase the amount of biomass left on the field which has the potential to increase soil C stocks. However, it remains unknown whether these sustainable practices significantly enhance soil C (Cerri et al., 2011;de Resende et al., 2006). In addition, there have been recent efforts to genetically modify Sugarcane to enhance its conversion efficiency and optimize the bioenergy crop (Lam et al., 2009). One successful effort is the development of Oilcane, which has altered litter chemistry owing to genetically enhanced oil production in the stems and leaves (Cheng & Timilsina, 2011;Limayem & Ricke, 2012;Parajuli et al., 2020;Zale et al., 2016). As litter chemistry regulates the microbial activity and shifts soil C dynamics (Cotrufo et al., 2015), it is critical to quantify and compare the abilities of both Sugarcane and Oilcane litter to form soil C.
The ability of sustainable harvesting practices to increase soil C stocks relies on the degree to which the litter and bagasse inputs lead to a net priming or a net stabilization effect. Overall, the resulting balance between priming and stabilization is likely driven by differences in litter quality between Sugarcane and Oilcane ( Figure 1). Lower-quality Sugarcane litter with a high C:N ratio decomposes slowly and can increase soil C stocks by remaining in the soil as undecomposed or partially decomposed particulate organic C (POC; Phukongchai et al., 2022). However, POC is largely unprotected and subject to further decomposition making it less ideal for building stable soil C stocks (Cotrufo et al., 2013). By contrast, Oilcane litter is likely of higher quality than Sugarcane owing to a lower C:N ratio due to a 44%-59% reduction in soluble lignin resulting in more rapid decomposition (Parajuli et al., 2020). This rapid decomposition has the potential to both prime soil C losses and stabilize soil C. On the one hand, Oilcane litter may prime the decomposition of soil organic C (SOC) and lead to greater respired C losses (Strickland et al., 2015;Talbot et al., 2012). On the other hand, enhanced decomposition in Oilcane may increase the stabilization of simple, microbially derived C in mineral-associated organic C (MAOC) that is highly stable and physically protected from microbial attack (Cotrufo et al., 2013;Lehmann & Kleber, 2015). Recent advances in soil stabilization theory suggest the more microbially accessible Oilcane litter may promote a higher microbial C use efficiency (CUE) and enhance the production of microbial products that preferentially form MAOC in grassland and crop ecosystems (Angst et al., 2021;Bradford & Crowther, 2013;Liang et al., 2017; F I G U R E 1 Hypothesized impacts of Sugarcane and Oilcane litter on soil carbon stocks. (1) Lower-quality Sugarcane litter will lead to soil carbon gains that are mostly particulate organic C (POC). (2) Higher-quality Oilcane litter will lead to increased soil carbon losses of primed soil organic matter. (3) Higherquality Oilcane litter will be decomposed with a higher carbon use efficiency by the microbial communities resulting in soil carbon gains that are mostly mineral associated organic matter (MAOC). Created with BioRe nder.com. Manzoni et al., 2012). Therefore, the sustainable harvesting practices with Sugarcane versus Oilcane litter may have disparate priming and stabilizing effects due to differences in litter quality and the resulting impacts on microbial decomposition.
Given the potential for sustainable harvesting practices to enhance soil C, it is important to investigate how Sugarcane litter builds stable soil C and impacts existing SOC. As such, our first objective was to quantify the balance between Sugarcane litter priming the loss of existing soil C stocks versus forming new stable SOC. In addition, genetic alterations to increase the energy density may increase the lability of the litter and impact microbial activity. Therefore, our second objective was to quantify differences between Sugarcane and Oilcane litter in the balance of soil C priming losses versus stable soil C gains. To meet our two objectives, we conducted a lab incubation experiment to trace the fate of Sugarcane and Oilcane bagasse and litter and tested the following hypotheses: (1) Due to its high C:N ratio, Sugarcane litter will reduce microbial decomposition resulting in more of the litter forming POC than MAOC compared with Oilcane litter. Due to the two competing mechanisms in which Oilcane litter could either enhance or reduce soil C stocks, we have two competing hypotheses: (2) The lower soluble lignin content of Oilcane will result in faster decomposition than the Sugarcane litter and result in greater priming-induced losses of soil C as a result of increased microbial activity, or (3) The greater decomposition of the more labile Oilcane litter will result in a higher microbial CUE and greater accumulation of MAOC.

| Experimental design
To test our hypotheses, we compared the decomposition of Sugarcane and Oilcane litter in a laboratory microcosm experiment. We did this by incubating C 4 Sugarcane and Oilcane bagasse and leaf litter in C 3 forest soils that differed in their natural abundance 13 C isotopic signatures. Our study included control of soil with no litter (S) and 4 litter treatments with 10 replicates each (total of 50 incubations): soil and Sugarcane bagasse (S-B), soil and Sugarcane leaves (S-L), soil and Oilcane bagasse (O-B), and soil and Oilcane leaves (O-L). We measured the concentration and δ 13 C signature of CO 2 to quantify how much of the added litter versus existing SOC was respired. After 11 weeks, we density fractionated the soil to quantify the added litter C in POC and MAOC by assessing the δ 13 C signature of these soil C fractions.

| Soil collection
We collected 10 kg of fine-loamy, mixed, active, mesic Ultic Hapludalfs soil in September 2021 from the top 15 cm of multiple locations in a 20 × 20 m forest plot dominated by the arbuscular mycorrhizal (AM) tree species, sugar maple (Acer saccharum), and tulip poplar (Liriodendron tulipifera), at Tom's Run Nature Preserve in Morgantown, West Virginia. We brought the soil back to the lab and sieved it through a 2-mm sieve to remove large roots and rocks and to homogenize the soil. We sieved the soil within 1 week of collection and stored it at 5°C prior to incubation.

| Litter processing
The Oilcane was genetically modified as detailed by Parajuli et al. (2020). The Sugarcane and Oilcane were grown through direct organogenesis in media, transferred to a temperaturecontrolled greenhouse (27 ± 2°C) for 2 months, and then transplanted into the field at the Plant Science Research and Education Unit in Citra, FL where they grew from April 2020 until October 2020. The crops were harvested in the Fall of 2020. During harvest, the leaves were separated from the stems for collection. In the lab, the bagasse was produced by extracting the juice from the stems using a benchtop juice extractor (JuiceMattic SC-3, Juicernet, FL, USA). Both the bagasse and leaves were oven dried at 50°C and then ground using a hammer mill with a 2-mm sieve (W-8-H, Schutte-Buffalo Hammermill). Lipid content, C:N ratio, and δ 13 C are provided in Table 1. Complete details on the growth and processing of the litter can be found in Maitra et al. (2022).

| Microcosm incubation
For the incubation, we used wide-mouthed glass mason jars (930 mL) with rubber septa installed into each lid. To set up the incubation, we mixed the sieved soil and added 70 ± 5.0 g of field moist soil (45% gravimetric water content) to all 50 jars. For the 10 control jars, we did not add any litter. For the remaining 40 jars, we added 0.5 g of dry litter according to the treatment (10 jars received Sugarcane bagasse, 10 jars received Sugarcane leaves, 10 jars received Oilcane bagasse, and 10 jars received Oilcane leaves). We sealed the mesocosms with the lids and incubated for 11 weeks in the dark (under blackout curtains) at room temperature (∼22°C).

| Respiration and δ 13 C of CO 2 measurements
To determine the amount of added litter C and SOC that was respired, we sampled the microcosm headspace weekly for 11 weeks including measurements on days 1, 3, and 10, for a total of 13 time points. We measured the total microbial production of CO 2 using an infrared gas analyzer (LI-6400, LI-Cor Biosciences Inc.) and the δ 13 C signature of the respiration with a Picarro G2201 (Picarro Inc.). Due to the C isotopic differences between the C 3 forest soil and C 4 litter, we were able to use the δ 13 C signature to calculate the proportion of the total respiration that came from the added litter versus SOC using the two end-member mixing model using the following equation: where pCO 2 litter is the proportion of the total CO 2 attributed to microbial respiration of the litter, the δ 13 C signatures of each sample and the soil controls were measured weekly on the Picarro, and the δ 13 C signature of the litter was obtained using a Thermo Fisher Delta V+ isotope ratio mass spectrometer interfaced with a Carlo Erba NC2500 Elemental Analyzer at the University of Maryland Center for Environmental Science Appalachian Laboratory. To calculate the amount of CO 2 attributed to the respiration of the added litter, we multiplied the pCO 2 litter by the total CO 2 respired. We then calculated the CO 2 attributed to the respiration of soil organic matter by subtracting the CO 2 derived from the added litter from the total CO 2 respired (Morrissey et al., 2017;Ridgeway et al., 2022).
After the gas samples were taken, we aerated the jars for 20 min before resealing them and then placing them back under a blackout curtain at room temperature (∼22°C). At the end of the 11 weeks, we terminated the incubation and air-dried the soils for 2 months at room temperature (∼22°C).

| Recovery of Sugarcane and Oilcane litter in soil organic carbon fractions
We density fractionated the dry soils to determine the amount of litter C that ended up in the POC versus the MAOC fractions as outlined in Lavallee et al. (2020). We performed a water extraction followed by an extraction with 1.85 g/mL sodium polytungstate (SPT) solution to separate out the light fraction of the POC. We then wetsieved the remaining heavy fraction using a 53-μm sieve to separate the heavy fraction of the POC and MAOC (POC > 53 μm, MAOC < 53 μm). We dried each fraction at 60°C and weighed them to calculate the recovery. We calculated mass recovery by comparing the total mass recovered in all three fractions to the initial sample mass prior to fractionating. All the samples had recoveries of 100% ± 5% except for 3 that were 100% ± 10%. Lastly, we ground and analyzed each fraction for %C, %N, and δ 13 C using a Thermo Fisher Delta V+ isotope ratio mass spectrometer interfaced with a Carlo Erba NC2500 Elemental Analyzer at the University of Maryland Center for Environmental Science Appalachian Laboratory. We scaled the results to the mass of dry soil in each jar to determine the proportion of total added litter that ended up in each fraction.

| Estimation of the balance between soil C losses versus new soil C gains
To determine if the treatments had net soil C losses or gains, we calculated the difference between soil C lost through priming and the new soil C gained from the added litter. We calculated the soil C losses through priming by subtracting the average cumulative respiration in the control jars without litter from the cumulative soil respiration in each jar with litter. Finally, we calculated the new soil C gains from the added litter by summing the litter C recovered in the POC and MAOC fractions.

| Statistical analysis
To determine the extent to which the genetic modifications to oilcane altered microbial respiration of the litter and SOC as well as the fate of the litter into the different soil C fractions, we performed a two-way analysis of (1) pCO 2 litter = δ 13 C sample − δ 13 C soil ∕ δ 13 C litter − δ 13 C soil

| Respiratory losses and soil priming effect of added litter
After 11 weeks of respiration measurements, we found no differences in the cumulative total (soil + litter), soil, or litter respiration between Oilcane and Sugarcane. However, there was more total respiration in jars with bagasse than leaves (Figure 2a, Tissue Type p < 0.01), which was due to significantly more soil-derived respiration (Figure 2c, Tissue Type p < 0.0001). We also found there was a trend between crop and tissue type on the amount of leaf litter and bagasse litter respired in which we identified a trend for Oilcane but not for Sugarcane ( Figure 2b, Crop * Tissue Type p < 0.1). We found that litter additions led to the priming of existing SOC. All the litter addition treatments had greater respiration of SOC than the control treatment ( Figure S1; p < 0.0001). In addition, this priming varied by tissue type with greater SOC priming in jars with bagasse than leaves (Figure 2d, Tissue Type p < 0.0001).

| Recovery of the added litter in soil C fractions
On average, we recovered 98% of the added litter C after fractionating the soil organic matter. Across all treatments, the greatest amount of litter C was recovered in the POC (~55%), followed by CO 2 (~37%), and then MAOC (~7%). Of the proportion of added litter C that was respired, there was an interactive effect between crop and tissue type. There was significantly more litter C respired with Oilcane leaves than Oilcane bagasse (Figure 3a, Crop * Tissue Type p < 0.001). There were main effects of crop and tissue type on the proportion of litter that was recovered in the POC fraction (Figure 3b, Crop p < 0.05, Tissue Type p < 0.05). However, there was an interactive effect of crop and tissue type where the proportion of litter C recovered in POC was significantly higher for Oilcane leaves than the other treatments (Figure 3b, Crop * Tissue Type p < 0.05). Lastly, tissue type had a significant main effect on the proportion of litter C recovered in the MAOC fraction (Figure 3c, Tissue Type p < 0.0001). In addition, there was a significant interactive effect of crop and tissue type (Figure 3c, Crop * Tissue Type p < 0.05) where there was a significant difference in the proportion of the litter C recovered between Sugarcane but not Oilcane tissue types. Sugarcane bagasse led to significantly more MAOC formation than Sugarcane leaves.

| Net soil C gains
All litter treatments had net soil C gains, where the litter C incorporated into SOC exceeded the soil C respired. Overall, Oilcane had greater net soil C gains than Sugarcane (Figure 4, Crop p < 0.05). In addition, there was an interactive effect between crop and tissue type where leaves had greater net soil C gains than bagasse for Oilcane ( Figure 4, Crop * Tissue Type p < 0.05).

| DISCUSSION
In support of H1 that Sugarcane would enhance the POC pool, we found that the addition of Sugarcane litter led to net soil C gains that were primarily driven by an accumulation of undecomposed litter in the POC fraction ( Figure 3b). This finding is most likely explained by the high C:N ratio of the Sugarcane litter slowing microbial decomposition of the substrate. In support, previous research shows that the beginning stage of Sugarcane bagasse decomposition is dominated by microbes degrading the more labile parts of the litter, leaving behind the more recalcitrant components (Phukongchai et al., 2022). In addition, the priming losses of soil C we observed (Figure 3) may be driven by the decomposition of the labile litter components which may have reduced the energetic constraints of soil microbes (Nottingham et al., 2009). Although the added Sugarcane litter led to priming losses of soil C (Figure 2d), these losses were outweighed by the gains of new litter C in the POC fraction (Figure 3b), resulting in net soil C gains (Figure 4). These findings are consistent with longer-term (>1 year), in-situ Sugarcane decomposition experiments (Cerri et al., 2011;Robertson & Thorburn, 2007) as well as Sugarcane decomposition models (Brandani et al., 2015;Galdos, Cerri, Cerri, Paustian, et al., 2009;Silva-Olaya et al., 2017), which also report increases in soil C following the shift from harvest via burning to leaving the Sugarcane litter on the field. However, it is uncertain how long the recalcitrant components of the litter in the POC fraction will remain undecomposed and stable in the soil. Recalcitrant POC is highly susceptible to loss under conditions that may enhance microbial decomposition, like increasing soil temperature or changes in N availability (Benbi et al., 2014;Li et al., 2018). While our findings suggest that the initial addition of Sugarcane bagasse and litter enhances soil C stocks, the stability of this soil C over longer time scales is uncertain. F I G U R E 2 (a) Total respiration (soil + litter). (b) Microbial respiration of the added litter. Microbial respiration of soil organic C (SOC; c). SOC priming (difference between soil CO 2 respired in each treatment jar versus the average of the control jars; d). Data shown are 10 replicates excluding outliers for each treatment. Created with BioRe nder.com. Our findings did not support H2 that Oilcane would increase priming losses of soil C or H3 that Oilcane would build soil C in stable MAOC. Instead, we found that Oilcane did not differ from Sugarcane in priming (Figure 2d) or the incorporation of litter into MAOC (Figure 3c). The lack of support for these hypotheses likely reflects marginal differences in the C:N ratios between the two plants. Although the genetic modifications to Oilcane decreased the plant's C:N ratio by 11.5% compared with Sugarcane (Table 1), these ratios remain relatively high, and a marginal C:N reduction may not be enough for microbes to overcome their nitrogen limitations and increase decomposition of the substrate or SOC (Moorhead & Sinsabaugh, 2006). In addition, the greater lipid content in the Oilcane litter could have increased the hydrophobicity of the litter and led to reductions in the ability of microbes to decompose Oilcane litter (Lützow et al., 2006). In support, we found that Oilcane litter additions led to greater soil C gains than Sugarcane with more of the Oilcane litter remaining undecomposed in the light POC fraction (Figure 4b). Regardless of the exact mechanisms, our results show that in the short term, the genetic modifications to Oilcane did not lead to soil C losses compared with Sugarcane and may even enhance soil C gains.

F I G U R E 3 Added litter C recovered in CO
Sugarcane and Oilcane had opposing effects of tissue type (i.e., bagasse vs. leaf litter) on the fate, priming, and net soil C gains that may have important management implications for building stable soil C. We found that Sugarcane bagasse formed more MAOC than Sugarcane leaves while Oilcane leaves preferentially remained as POC (Figure 3b,c). These differences have the potential to impact the amount of new litter C that can be stored and its residence time in the soil. MAOC is more stable than POC, but there is an upper limit to building MAOC because the available mineral surfaces can become saturated . By contrast, there is not a clear saturating limit to POC formation, but POC may have a shorter residence time depending on litter chemistry and microbial activity (Burns et al., 2013;Cotrufo et al., 2019;Stewart et al., 2009). Therefore, soils that are saturated or close to MAOC saturation may have the potential to build more soil C by adding Oilcane leaf litter while soils that have low C and high mineral content may have the potential to enhance soil C stocks by adding Sugarcane bagasse litter (Castellano et al., 2015). In addition to the interactive differences in litter fate, bagasse litter additions led to greater priming of native SOC than leaf litter additions ( Figure 2d). As a result, Oilcane leaves led to greater net soil C gains than Oilcane bagasse, which indicates that there may be a tradeoff between building new stable MAOC and losing SOC through priming with bagasse litter additions in certain crops. The differences observed in bagasse and leaf litter are likely attributed to structural differences between the two tissue types (e.g., C:N ratio, lipid content, and differences in the proportion of and structural characteristics of cellulose, hemicellulose, and lignin; Nottingham et al., 2009;Schmatz et al., 2020). However, regardless of the exact litter chemistry control, these results suggest that the tissue type of litter you amend the field with may impact the stability, retention time, and net effect of new soil C gains.
Sugarcane and Oilcane may build less MAOC and more POC in comparison with bioenergy crops with lower C:N ratio litter. In a similar experiment with corn and miscanthus, 22%-29.3% of the added litter C formed MAOC compared with only 7.22%-7.39% with Sugarcane and Oilcane (Ridgeway et al., 2022). In addition, Sorghum aboveground litter in the field formed more MOAC than POC (Fulton-Smith & Cotrufo, 2019). While differences in experimental conditions limit direct comparisons, our results suggest that the recalcitrant nature of Sugarcane and Oilcane litter may be limiting MAOC formation and enhancing POC to a greater degree than other, lower C:N bioenergy crops. Overall, this comparison indicates that there are emergent differences between bioenergy crops in how they build soil C.
While our results point to important differences in how Sugarcane and Oilcane build soil C, we acknowledge that there are limitations to our study. First, our study was conducted in a microcosm that excludes seasonal and diurnal fluctuations in soil temperature and moisture as well as living roots that may have an impact on litter decomposition and soil C dynamics. Although these field conditions were absent, microcosm experiments have proven important in identifying and isolating mechanisms that can help explain observations and experimental results from the field (Benton et al., 2007;Cortez et al., 1996;Craig et al., 2022;Nicolardot et al., 2007;Sokol & Bradford, 2019; F I G U R E 4 Difference between soil C lost through priming and new soil C from the added litter. Data shown are 10 replicates excluding outliers for each treatment. Created with BioRe nder.com. Strickland et al., 2009). Second, we ran our incubation experiment for 11 weeks which likely only captured the initial stages of litter decomposition. However, the initial stage of decomposition is an important indicator of long-term stabilization patterns (Craig et al., 2018). Future research should examine the long-term dynamics of new MAOC or POC formation in Sugarcane systems by following the fate of litter enriched in 13 C and 15 N in the field. Finally, although both Sugarcane and the forest stands we sampled are associated with AM symbionts, adding Sugarcane and Oilcane to forest soils may have influenced microbial responses by introducing these microbes to novel substrates (Palozzi & Lindo, 2018). While studies with isotopically enriched litters and agricultural soils can better represent real-world soil microbial community differences (e.g. Fulton-Smith & Cotrufo, 2019;Ridgeway et al., 2022), our experimental design allowed us to use leaves and bagasse substrates that directly reflect real-world amendments. Moreover, we also point to the successful use of soil transplants (i.e., C 4 soils in C 3 ecosystems) to estimate root-derived SOC in forest systems (Huang et al., 2021;McCloskey et al., 2021). Despite these limitations, our results identify important mechanisms that lay the foundation for future large-scale field experiments.

| CONCLUSION
Our results showed that the genetic modifications to Sugarcane had modest impacts on soil C dynamics in the initial stages of litter decomposition but did not negatively alter soil C stocks. These results indicate that transitioning to genetically engineered Oilcane may enhance bioenergy fuel conversion efficiency without unintended consequences on soil C cycling. However, testing how other genetically modified crops alter soil C cycling remains critical, particularly when modifications lead to greater litter chemistry differences. In addition, our observations that Sugarcane bagasse amendments formed more MAOC while Oilcane leaves preferentially remained as light POC may be important to consider when deciding how to manage bioenergy agricultural systems to meet sustainability goals. Overall, we highlight the potential for Oilcane to sustainably displace Sugarcane as a bioenergy feedstock and emphasize the remaining need to examine whether genetic modification will alter soil C dynamics over longer durations in the field, across different management strategies, and for other genetically modified bioenergy crops.