Bioenergy crops, such as switchgrass (*Panicum virgatum* L.) and gamagrass (*Tripsacum dactyloides* L.), offer a potential solution to reduce fossil fuel consumption. Nitrogen (N) fertilization enhances bioenergy crop yields, but its impact on belowground processes, particularly soil extracellular enzymes, remains understudied. Soil extracellular glycosidases, produced by microorganisms, are crucial for decomposing labile soil organic carbon. These enzymes are significantly affected by N fertilization, yet the impact on their spatial distribution and the variation of effects with different bioenergy crop species are not well understood. Understanding the spatiotemporal dynamics of soil extracellular enzymes is essential for improving soil health management practices, especially considering the widespread overuse of N fertilizers globally. Extracellular glycosidases, including α-glucosidase (AG), β-glucosidase (BG), β-xylosidase (BX), and cellobiohydrolase (CBH), are key indicators of microbial activity involved in organic carbon cycling. Previous studies have shown varying effects of N fertilization on glycosidase activities, with some reporting increases while others show decreases or no effect, depending on factors such as enzyme type, soil depth, and crop species. The spatial distribution of soil extracellular enzymes also varies significantly, influencing the efficiency of nutrient cycling and decomposition processes. However, studies specifically examining the spatial heterogeneity of soil glycosidases under N fertilization in bioenergy crops are lacking. This study aimed to investigate the impact of N fertilization and bioenergy crop species on both the central tendency (mean activity) and spatial heterogeneity of four key glycosidases in a three-year field experiment.

The literature reveals inconsistent effects of nitrogen fertilization on soil glycosidase activities. Some studies show little to no change in AG, BX, and CBH, but significant reductions in BG in topsoil (2–12 cm) in acidic forest soil. Other studies found no effect on glycosidase activities in topsoil (0–10 cm) of alpine grasslands, while in other studies, N addition significantly reduced AG, BG, and BX in deeper soil layers. Further inconsistencies include minor effects on a wide range of soil extracellular enzyme activities in Chinese forests, and significant cropping system effects on AG and BG, with higher activities observed under meadow or oat and lowest under corn and soybean. The effect of N fertilization on BG is complex, with some studies showing positive effects attributed to plant growth and litter input, while others show negative effects associated with a shift in fungal-bacterial ratios. A meta-analysis indicated that N fertilization stimulated AG, BG, BX, CBH, and C acquisition. The spatial distribution of soil extracellular enzymes has been documented across various scales, with studies indicating similar spatial variations at microsite and plot scales for BG. Studies on other hydrolytic enzymes involved in nutrient acquisition showed similar spatial variations. However, the spatial heterogeneity of extracellular enzyme activities is often more evident in grassland and forest soils than in agricultural soils, possibly due to contrasting root morphology and chemistry. There is a paucity of research on the spatial patterns of soil extracellular enzymes in bioenergy crops under N fertilization, with only one study showing that high N deposition tended to homogenize spatial patterns in a semi-arid Mediterranean shrubland.

A three-year N fertilization experiment was established in 2011 at the Tennessee State University Main Campus Agriculture Research and Education Center (AREC). The experiment included two bioenergy crop species, switchgrass (SG) and gamagrass (GG), and three N levels: no N input (NN), low N input (LN: 84 kg N ha⁻¹ year⁻¹), and high N input (HN: 168 kg N ha⁻¹ year⁻¹). Each treatment had four replicated plots. On June 6th, 2015, soil samples (0–15 cm) were collected from 12 plots using a spatially explicit sampling design (24 cores per plot, 288 cores total). Soil gravimetric moisture content, water-extractable soil pH were measured, and four glycosidase activities (AG, BG, BX, CBH) were quantified using fluorimetric enzymatic assays. Statistical analyses included two-way ANOVA to test the effects of N fertilization, crop species, and their interaction on glycosidase activities. Cochran’s C test was used to evaluate variance homogeneity. Geostatistical analysis included trend surface analysis (TSA), Moran’s I index to quantify spatial autocorrelation, and inverse distance weighting (IDW) interpolation to generate spatial maps of enzyme activities.

Significant interactive effects of N fertilization and crop type were observed for BX activity. In switchgrass, LN and HN significantly increased BX activity by 14% and 44%, respectively, compared to NN. Significant crop type effects were also observed for BG, BX, and marginally for CBH, with GG showing 15–39% higher activities than SG (except for AG). Within-plot variances were generally higher in SG than GG, with little difference among N fertilization levels. Spatial patterns were more evident in LN or HN plots than NN plots for BG in SG and CBH in GG. The sample size requirement (SSR) was generally higher for NN than LN or HN, and higher for SG than GG. Trend surface analysis showed varying significant linear or non-linear trends among treatments and crop types. Moran’s I analysis revealed varying numbers and distances of significant spatial autocorrelations, with fertilized treatments generally exhibiting higher numbers of significant autocorrelations. IDW maps visually illustrated higher enzyme activities in GG than SG across all treatments, and more pronounced spatial variations in LN and HN plots in SG compared to NN.

The significant positive response of BX to N fertilization in switchgrass, but not in gamagrass, could be due to several factors: higher increases in SOC under N fertilization in switchgrass, the correlation between BX and soil C content, and the lack of correlation between BX and microbial biomass. The higher glycosidase activities in gamagrass, despite similar root volumes, may be attributed to contrasting root chemistry, leading to different nutrient competition strategies with soil microbes. The increased spatial heterogeneity of glycosidases under N fertilization likely resulted from the uneven distribution of fertilizer, creating hotspots of microbial activity and SOC. The lack of correlation between microbial biomass C and glycosidases, coupled with the significant correlation between SOC and glycosidases, indicates that substrate availability, rather than microbial abundance, may be the primary driver of glycosidase heterogeneity. The enzyme-specific responses to N fertilization highlight the complexity of interactions between enzyme characteristics, plant roots, soil, and microbial communities.

This study demonstrated that N fertilization significantly increased the mean activity and spatial heterogeneity of soil glycosidase activities in bioenergy croplands. However, these effects varied depending on the crop species and enzyme type. Future studies should focus on specific enzymes within particular bioenergy croplands when evaluating the effects of N fertilization.

The study focused on a single year of sampling after three years of fertilization. Longitudinal studies would provide better understanding of temporal dynamics. The spatial sampling design, while spatially explicit, might not capture the full extent of small-scale heterogeneity. Further investigation into the underlying microbial mechanisms driving the observed patterns would enhance the understanding.