Harnessing Diptera Diversity as a Bioindicator for Carbon and Nitrogen Loss During Litter Decomposition

Results

Mass Loss

Residue mass declined over time, except for in fine-mesh bags in the business-as-usual corn management system (Figure 2). Significant two-way interactions were observed between sampling date and crop (F_{4, 898} = 33.26, P < 0.001) and between sampling date and mesh size (F_{1, 896} = 260.07, P < 0.001. The model explained 74% of the variance in mass loss (marginal R² = 0.74; conditional R² = 0.76). Cover crop residue decomposition was fastest in the restored prairie for both coarse- and fine-mesh bags, although these rates did not differ significantly from those in the aspirational systems (Table 3). In coarse-mesh bags, decomposition of corn and soybean was 41% faster under aspirational management than under business-as-usual management. In fine-mesh bags, management effects were less pronounced: residue decomposed faster in aspirational corn than in business-as-usual corn, whereas decomposition rates did not differ among soybean management systems. Estimated residue half-lives ranged from 27 to 43 days in coarse-mesh and 46 to 173 days in fine-mesh. Half-life could not be estimated for fine-mesh business-as-usual corn because no measurable mass loss was observed.

Figure 2. Predicted cover crop residue decomposition (± 95% confidence intervals) over a 60-day period for aspirational, business-as-usual, and restored management, separated by mesh size. Significant two-way interactions were observed between sampling date and crop (F_{4, 898} = 33.26, P < 0.001) and between sampling date and mesh (F_{1, 896} = 260.07, P < 0.001).

Table 3. Model-estimated exponential decay rates (k) and corresponding half-lives of cover crop residue across crop and mesh size during the 60-day decomposition period.

^a Negligible mass loss in fine-mesh business-as-usual corn prevented the calculation of a half-life.

Realized Microbial Respiration

Realized microbial respiration rates were low (Figure 3), ranging from 0.000 to 8.321 mg CO₂ g^-1 biomass day^-1 (1^st Quartile = 0.011, Median = 0.044, 3^rd Quartile = 0.528). Mesh significantly affected realized microbial respiration (F_{1, 671} = 4.03, P = 0.045), and a significant two-way interaction between sampling date and crop was observed (F_{4, 671} = 6.31, P < 0.001). The model explained 21% of the variance in realized microbial respiration (marginal R² = 0.21; conditional R² = 0.57). Across all systems, respiration was 40% greater in the fine-mesh than in the coarse-mesh bags. Within fine-mesh bags placed in soybean under business-as-usual management, respiration was greater than under aspirational management on days 40 and 60. No other significant differences among systems were detected across sampling date and mesh size.

              Moisture was a primary driver of realized microbial respiration (Figure 4). A significant two-way interaction was observed between moisture content and crop (F_{4, 569} = 3.77, P = 0.005). The model explained 39% of the variance in realized microbial respiration (marginal R² = 0.38; conditional R² = 0.53). Aspirational corn exhibited the strongest positive relationship between moisture content and respiration (β = 0.201, 95% CI: 0.155–0.247); however, the slope was only significantly different from that of business-as-usual corn (β = 0.116, 95% CI: 0.079–0.153). Slopes for aspirational soybean (β = 0.140, 95% CI: 0.103–0.177), business-as-usual soybean (β = 0.142, 95% CI: 0.102–0.183), and restored prairie (β = 0.146, 95% CI: 0.112–0.179) did not differ significantly between either corn management system.

Figure 3. Mean realized microbial respiration (mg CO₂ g^-1 biomass day^-1 ± standard error) from decomposing cover crop residue in fine- and coarse-mesh measured at 20, 40, and 60 days across aspirational, business-as-usual, and restored management systems. Letters indicate significant differences among groups within each panel. Mesh significantly affected realized microbial respiration (F_{1, 671} = 4.03, P = 0.045), and a significant two-way interaction between sampling date and crop was observed (F_{4, 671} = 6.31, P < 0.001).

Figure 4. Realized microbial respiration (mg CO2 g-¹ biomass day^-1 ± 95% confidence intervals) during cover crop residue decomposition as a function of moisture content across a 60-day period for aspirational corn, business-as-usual corn, and restored prairie systems. A significant two-way interaction was observed between moisture content and crop (F_{4, 569} = 3.77, P = 0.005). Aspirational and business-as-usual soybean were omitted because their responses were statistically similar to those of the restored prairie and were visually redundant.

Microbial Respiration Potential

Microbial respiration potential rates were consistently higher than realized microbial respiration and declined over time (Figure 5). A significant three-way interaction was observed among sampling date, crop, and mesh size (F_{4, 663} = 2.43, P < 0.05). The model explained 59% of the variance in microbial respiration potential (marginal R² = 0.59; conditional R² = 0.66).

At 20 days, microbial respiration was highest in fine-mesh bags placed in aspirational corn (13.11 ± 2.38 mg CO₂ g^-1 biomass day^-1) but was comparable to the restored prairie (9.79 ± 1.84 mg CO₂ g^-1 biomass day^-1). Aspirational corn exceeded aspirational soybean and both business-as-usual systems (5.13–8.82 mg CO₂ g^-1 biomass day^-1). Aspirational management increased microbial respiration in corn by 156% relative to business-as-usual but had no significant effect on soybean. In coarse-mesh bags, differences were less pronounced; the only significant comparison was between aspirational corn (7.75 ± 1.40 mg CO₂ g^-1 biomass day^-1) and business-as-usual soybean (5.03 ± 0.91 mg CO₂ g^-1 biomass day^-1).

              After 40 days, microbial respiration declined, though trends in the fine-mesh remained consistent. Aspirational corn again exhibited the highest microbial respiration (6.32 ± 1.04 mg CO₂ g^-1 biomass day^-1), not differing from restored prairie (5.17 ± 0.89 mg CO₂ g^-1 biomass day^-1), but exceeding aspirational soybean and both business-as-usual systems (2.35–4.53 mg CO2 g^-1 biomass day^-1). Aspirational management increased microbial respiration by 169% in corn and 69% in soybean relative to business-as-usual. Trends for residue in coarse-mesh bags mirrored those observed in fine-mesh bags.

By 60 days, microbial respiration reached its lowest levels across corn, soybean, and prairie, regardless of management or mesh size. In fine-mesh bags, aspirational corn did not differ from restored prairie and aspirational soybean (2.32–3.04 mg CO2 g^-1 biomass day^-1); however, both aspirational systems remained higher than the business-as-usual systems (1.07–1.16 mg CO2 g^-1 biomass day^-1). Aspirational management increased microbial respiration in corn by 184% and in soybean by 100% relative to business-as-usual. In coarse-mesh bags, aspirational corn maintained the highest microbial respiration (3.06 ± 0.56 mg CO₂ g^-1 biomass day^-1), exceeding restored prairie, aspirational soybean, and both business-as-usual systems (1.07–1.75 mg CO2 g^-1 biomass day^-1). Aspirational corn exhibited 75% greater respiration than restored prairie and 186% greater than business-as-usual corn. Aspirational soybean did not differ from business-as-usual soybean.

Figure 5. Mean microbial respiration potential (mg CO₂ g^-1 biomass day^-1 ± standard error) from decomposing cover crop residue in fine- and coarse-mesh measured at 20, 40, and 60 days across aspirational, business-as-usual, and restored management systems. Letters indicate significant differences among groups within each panel. A significant three-way interaction among sampling date, crop, and mesh was observed (F_{4, 663} = 2.43, P < 0.05).

Total Carbon and Nitrogen

Echoing the mass loss, total nitrogen declined across all management systems, and mesh size had a strong impact over time (Figure 6). For nitrogen, significant two-way interactions were observed between sampling date and mesh size (F_{1, 675} = 201.40, P < 0.001) and between crop and mesh size (F_{4, 675} = 260.07, P = 0.003). The model explained 72% of the variance in nitrogen release (marginal R² = 0.72; conditional R² = 0.77). .

 nitrogen release were consistently greater in coarse-mesh bags placed in the restored prairie and aspirational systems than in the business-as-usual systems. After 20 days, restored prairie (31.2 ± 3.5%) and aspirational soybean (29.1 ± 3.1%) retained the least nitrogen in the remaining residue, followed by aspirational corn (43.8 ± 4.6%), while business-as-usual corn (61.1 ± 6.5%) and soybean (63.8 ± 6.8%) retained the greatest amount of nitrogen. Carbon dynamics followed a similar trend, but system differences were weaker, with no significant differences between restored prairie and the aspirational systems. Fine-mesh bags retained substantially more carbon and nitrogen. At 20 days, remaining nitrogen ranged from 58.1–88.6%.  Aspirational corn (59.1 ± 6.1%) was significantly lower than business-as-usual corn (88.6 ± 9.2%) but did not differ from the restored prairie (59.9 ± 6.8%). No significant differences were observed between soybean management systems. For carbon, the only significant difference observed was between the restored prairie (68.1 ± 6.3%) and business-as-usual corn (88.8 ± 7.9%).

Similar differences in remaining carbon and nitrogen persisted through days 40 and 60. By day 60, coarse-mesh bags in the restored prairie and aspirational systems retained only 19.1–21.2% of initial carbon and 9.8–14.6% of initial nitrogen, compared with 26.2–30.8% carbon and 20.0–23.4% nitrogen in the business-as-usual systems. The remaining nitrogen was lowest in the restored prairie and the aspirational soybean, followed by the aspirational corn, with both aspirational systems exhibiting significantly less total nitrogen than business-as-usual systems. The remaining carbon did not differ significantly between aspirational corn and soybean but did between business-as-usual corn and soybean. In fine-mesh bags, restored prairie and aspirational systems retained 40.3–50.5% of initial carbon and 42.6–59.6% of initial nitrogen, whereas business-as-usual corn and soybean systems retained 50.9–56.3% carbon and 61.4–69.5% nitrogen. The restored prairie retained significantly less carbon and nitrogen than the aspirational and business-as-usual systems, which did not differ significantly from one another.

Figure 6. Mean proportion of nitrogen remaining (± standard error) in cover crop residue in fine- and coarse-mesh measured at 20, 40, and 60 days across aspirational, business-as-usual, and restored management systems. Letters indicate significant differences among groups within each panel. Significant two-way interactions were observed between sampling date and mesh (F_{1, 675} = 201.40, P < 0.001) and between crop and mesh (F_{4, 675} = 260.07, P = 0.003).

Diptera Biodiversity

During 2023 and 2024, a total of 88,716 flies from 70 families were collected across soil emergence and malaise traps, revealing distinct trophic composition between sampling approaches (Figure 7). Detritivores and herbivores dominated emergence trap communities, while malaise trap communities were primarily comprised of detritivores and nectarivores. Nearly half of all flies within the malaise trap community lacked an assignable trophic guild, driven primarily by Cecidomyiidae, Chironomidae, and Sciaridae.

Figure 7. Relative Diptera trophic guild composition sampled using soil emergence and malaise traps in the Long-Term Agroecosystem Research (LTAR) Aspirational Cropping Systems Experiment (ACSE) during 2023–2024. Emergence—detritivore (47.8%), herbivore (44.6%), microbivore (4.3%), predator (2.5%), and parasitoid (0.8%). Malaise—nectarivore (23.6%), detritivore (19.8%), predator (2.8%), hematophage (2.0%), herbivore (1.6%), microbivore (1.2%), parasite (< 0.1%), and none (48.9%),

Multivariate GLMs revealed that the trophic guild composition of emergence trap communities was significantly affected by management system (Deviance = 213.25, df = 7, p = 0.001) and year (Deviance = 80.42, df = 1, p = 0.001), while block had no significant effect (Deviance = 11.44, df = 3, p = 0.777). Management system influenced all five trophic guilds: detritivores (Deviance = 60.00, p = 0.001), herbivores (Deviance = 51.25, p = 0.001), microbivores (Deviance = 31.60, p = 0.003), parasitoids (Deviance = 26.95, p = 0.004), and predators (Deviance = 52.46, p = 0.001). Malaise trap trophic guild composition was similarly affected by management system (Deviance = 173.34, df = 7, p = 0.001) and year (Deviance = 198.70, df = 1, p = 0.001), as well as block (Deviance = 43.06, df = 3, p = 0.036). At the guild level, management system influenced detritivores (Deviance = 29.03, p = 0.005), microbivores (Deviance = 48.16, p = 0.001), nectarivores (Deviance = 38.57, p = 0.002), but not hematophages (Deviance = 12.54, p = 0.337), herbivores (Deviance = 15.05, p = 0.315), parasites (Deviance = 14.06, p = 0.337), or predators (Deviance = 15.94, p = 0.304).

Univariate GLMMs confirmed that management system significantly affected the abundance of all trophic guilds in the emergence trap communities: detritivores (χ² = 77.90, df = 7, p < 0.001), herbivores (χ² = 81.49, df = 7, p < 0.001), microbivores (χ² = 34.54, df = 7, p < 0.001), parasitoids (χ² = 35.55, df = 7, p < 0.001), and predators (χ² = 59.98, df = 7, p < 0.001; Figure 8). Detritivores were significantly more abundant in all the aspirational systems except for aspirational forage compared to the restored prairie (1.7–2.2× increase), and only lower in business-as-usual corn (54.8% decrease). Herbivore abundance was only lower than the restored prairie in business-as-usual corn (73.3% decrease), with no other management systems differing significantly. Microbivores were more abundant in aspirational corn, aspirational wheat, aspirational canola, and business-as-usual soybean (5.6–8.6×). Parasitoids were consistently less abundant across all aspirational and business-as-usual (72.7–90.9% decrease) management systems. Predator abundance was higher in aspirational corn, aspirational canola, and business-as-usual soybean (2.6–3.6× increase).

Figure 8. Pairwise contrasts (± 95% confidence intervals) of aspirational and business-as-usual management systems relative to the restored prairie for each significant trophic guild in emergence trap communities: detritivores (χ² = 77.90, df = 7, p < 0.001), herbivores (χ² = 81.49, df = 7, p < 0.001), microbivores (χ² = 34.54, df = 7, p < 0.001), parasitoids (χ² = 35.55, df = 7, p < 0.001), and predators (χ² = 59.98, df = 7, p < 0.001). Points represent estimated marginal contrasts for each management system; values greater than zero indicate higher abundance than the restored prairie, and values less than zero indicate lower abundance. Solid points denote significant differences.

Similarly, univariate GLMMs also corroborated that the management system significantly influenced detritivore (χ² = 35.43, df = 7, p < 0.001), microbivore (χ² = 65.91, df = 7, p < 0.001), and nectarivore (χ² = 47.68, df = 7, p < 0.001) abundance in malaise trap communities (Figure 9). Detritivore abundance was significantly higher in aspirational corn, aspirational soybean, aspirational forage, business-as-usual corn , and business-as-usual soybean than in the restored prairie (1.5–1.6× increase). Microbivores were more abundant in nearly every aspirational corn, aspirational wheat, and aspirational forage as well as business-as-usual corn and business-as-usual soybean (2.4–6.2× increase). Nectarivore abundance was higher in aspirational corn, aspirational canola, aspirational forage, and business-as-usual corn (1.4–1.9× increase).

Figure 9. Pairwise contrasts (± 95% confidence intervals) of aspirational and business-as-usual management systems relative to the restored prairie for each significant trophic guild in malaise trap communities: detritivores (χ² = 35.43, df = 7, p < 0.001), microbivores (χ² = 65.91, df = 7, p < 0.001), and nectarivores (χ² = 47.68, df = 7, p < 0.001) . Points represent estimated marginal contrasts for each management system; values greater than zero indicate higher abundance than the restored prairie, and values less than zero indicate lower abundance. Solid points denote significant differences.

 Detritivore phenology varied significantly across management systems throughout the growing season in emergence trap communities (χ² = 87.74, df = 28, p < 0.001), with differences most pronounced from May through August and dissipating by September (Figure 10). Business-as-usual corn consistently harbored the fewest detritivores each month. Early in the season, aspirational corn and aspirational wheat had the greatest detritivore activity (33.2 ± 12.6 and 32.2 ± 12.3, respectively), outpacing all other management systems (5.9–10.0) except aspirational canola (14.0 ± 5.4). Aspirational corn retained the highest abundance into June (30.9 ± 11.8), with aspirational soybean also increasing in activity (20.7 ± 7.9), while aspirational wheat (19.3 ± 7.4) was only greater than the restored prairie (5.49 ± 2.2) and business-as-usual corn (3.8 ± 1.6). By July, business-as-usual soybean increased to support the greatest detritivore abundance (25.7 ± 9.8), exceeding the restored prairie (8.2 ± 3.3) and business-as-usual corn (4.6 ± 1.9). Peak August abundance shifted to aspirational canola and aspirational soybean (25.5 ± 9.7 and 23.8 ± 9.1, respectively), while business-as-usual corn remained the lowest (3.0 ± 1.3). By September, differences among management systems were subtler, with only business-as-usual (2.5 ± 1.2) remaining lower than the restored prairie (10.7 ± 4.2) and aspirational soybean (13.1 ± 5.1).

Herbivore phenology followed a broadly similar seasonal trend, though with some notable differences in which management systems stood out each month (χ² = 81.76, df = 28, p < 0.001). As with detritivores, herbivores were consistently less abundant in business-as-usual throughout the season. Aspirational wheat drove the strongest early-season signal (35.2 ± 14.0), supporting higher herbivore abundance than aspirational soybean (5.0 ± 2.1), aspirational forage (7.6 ± 3.2), business-as-usual corn (4.1 ± 1.8), and business-as-usual soybean (6.8 ± 2.8). Aspirational wheat retained the highest abundance into June (24.4 ± 9.7), though the gap among systems narrowed considerably, with only aspirational soybean (6.1 ± 2.6) and business-as-usual corn (2.2 ± 1.0) remaining lower. A shift occurred in July, when business-as-usual soybean rose to the highest herbivore abundance (39.9 ± 15.8), exceeding aspirational corn (11.2 ± 4.6) and business-as-usual corn (5.9 ± 2.5). In August, aspirational canola (31.6 ± 12.5) and aspirational forage (23.3 ± 9.3) became the most abundant systems, while aspirational corn (3.3 ± 1.5) and business-as-usual corn (2.5 ± 1.2) remained the least abundant again. By September, herbivore abundances declined substantially across all systems, with only business-as-usual corn (0.4 ± 0.3) remaining lower than aspirational wheat (5.0 ± 2.1) and aspirational corn (4.1 ± 1.8).

Figure 10. Estimated marginal means (± standard error) of detritivore and herbivore abundance in emergence trap communities across aspirational, business-as-usual, and restored prairie management systems from May through September. Letters indicate significant differences among groups within each panel. Detritivore (χ² = 87.74, df = 28, p < 0.001) and herbivore (χ² = 81.76, df = 28, p < 0.001) abundance were significantly affected by management system.

Detritivore phenology also varied significantly across management systems throughout the growing season in malaise trap communities (χ² = 50.15, df = 28, p = 0.006), though differentiation among systems was considerably weaker than in emergence traps (Figure 11). Aspirational corn had the greatest abundance early in the season (50.9 ± 16.3) but only exceeded aspirational wheat (18.5 ± 6.0) and the restored prairie (15.7 ± 5.1) in May. Business-as-usual soybean surged in abundance in June (96.4 ± 30.6), well above aspirational wheat (36.8 ± 11.8) and aspiration canola (29.3 ± 9.4), while remaining indistinguishable from all other systems. Differences in detritivore abundance among management systems effectively disappeared from July through August.

Nectarivore phenology exhibited a markedly stronger response among management systems throughout the season (χ² = 1446.69, df = 28, p < 0.001). In May, aspirational corn supported the greatest nectarivore abundance (147.7 ± 11.7), more than all other management systems. Restored prairie (68.2 ± 5.8), aspirational canola (58.4 ± 5.1), and aspirational wheat (52.6 ± 4.6) contained the fewest nectarivores. Aspirational forage led nectarivore abundance in June (115.7 ± 9.3), well above aspirational wheat (51.8 ± 4.6), which remained the lowest. By July, aspirational forage, business-as-usual corn, and aspirational canola converged with the highest abundance (93.9 ± 7.7, 92.8 ± 7.6, and 91.2 ± 7.5, respectively). August brought a prominent reversal, and aspirational canola abundance was greater than every other management system (89.9 ± 7.4) and aspirational corn fell to the lowest (16.7 ± 1.9). Through September, nectarivore activity declined broadly but aspirational forage and aspirational wheat retaining the highest abundances (18.6 ± 2.1 each) and business-as-usual soybean finishing the season lowest (5.1 ± 0.9).

Figure 11. Estimated marginal means (± standard error) of detritivore and nectarivore abundance in malaise trap communities across aspirational, business-as-usual, and restored prairie management systems from May through September. Letters indicate significant differences among groups within each panel. Detritivore (χ² = 50.15, df = 28, p = 0.006) and (χ² = 1446.69, df = 28, p < 0.001) abundance were significantly affected by management system.

Discussion

Integrating soil health practices into a sustainably intensified cropping system increased decomposition, microbial functional capacity, and nutrient release relative to conventional management. After only two to three years of establishment, aspirational systems approached the level of ecosystem functioning observed in a restored prairie, suggesting that management diversification can restore ecological processes often attenuated by conventional agricultural intensification. More importantly, these findings indicate that agricultural management influences not only the rate of cover crop decomposition but also reorganizes the mechanistic structure of the detrital food web. Synergistic above- and belowground interactions can enhance ecosystem function in intensively managed agroecosystems (Giller et al., 1997), but these interactions emerge from an intricate detrital food web constrained by environmental conditions.

Multi-Trophic Effects on Decomposition

The conventional cover crop decomposition model frames residue turnover primarily as a function of substrate quality and environmental conditions, while multi-trophic interactions between microorganisms and soil fauna often overlooked or left unevaluated (Thapa et al., 2022; Adhikari et al. 2024). The initial quality of cover crop residue, expressed as the C:N ratio, was 18.8 in 2024 and 22.2 in 2025. Based on these values, predicted decomposition constants (k) in a no-till agronomic system with soil faunal exclusion range from 0.014 to 0.015 ( ; Thapa et al., 2022). The similarity between these predicted and observed decomposition rates from fine-mesh litterbags in the restored prairie and aspirational systems suggests that these systems exhibit the expected functional capacity given substrate quality (Table 3). In contrast, decomposition rates in the business-as-usual systems were substantially lower than predicted, falling below even the minimum reported by Thapa et al. (2022). Because the fine mesh excluded mesofauna and macrofauna, this discrepancy is most parsimoniously attributed to differences in microbial communities, consistent with the lower microbial respiration potential observed in business-as-usual systems.

Notably, decomposition constants from the coarse-mesh litterbags told a different story: values in business-as-usual systems converged toward predicted levels, while those in the restored prairie and aspirational systems were considerably higher than expected. This pattern suggests where conventional management suppresses microbial-mediated decomposition, soil fauna may partially compensate by assuming an expanded role in residue turnover. A key implication is that faunal access further homogenized decomposition rates across systems, narrowing the differences observed under fine-mesh exclusion. While microbial communities are often depicted as the foundation of the detrital food web (Hendrix et al., 1986), these results suggest that soil fauna act as a independent driver of residue decomposition rather than simply amplifying microbial functional capacity.

 Managing Residue for Nutrient Efficiency

The agronomic value of decomposition depends in part on synchronizing nutrient release with crop demand (Drinkwater & Snapp 2007). Therefore, differences in decomposition have direct implications for nutrient-use efficiency. Contrary to our findings, previous studies have reported that reduced or conventional tillage practices—such as double disking—either have no effect on decomposition rates or accelerate them (Lupwayi et al., 2004; Poffenbarger et al., 2015; Singh et al., 2020). However, comparisons between no-till and tilled systems are complicated by residue incorporation, which accelerates decomposition by placing residue within the soil profile (Jahanzad et al., 2016; Sievers & Cook, 2018). While residue incorporation reflects common agricultural practice, it also introduces a confounding factor that obscures the effect of management on ecosystem function. By standardizing residue placement on the soil surface, we were able to isolate the influence of management on decomposition dynamics.

              Approximately 50% of cumulative nitrogen uptake in corn occurs between the V8 and VT growth stages, while sidedress nitrogen applications typically target V6 (Bender et al., 2013). Relative to this timeline, when faunal access was permitted, residue in aspirational corn release 45% more nitrogen by V6 (20 days), and 7% more by VT (60 days) than business-as-usual corn. The narrowing of this difference by VT might suggest nitrogen release becomes comparable between systems over time; however, nitrogen release does not equate to plant uptake, which itself depends on the detrital food web. Microbial communities, rather than soil fauna, mineralize nitrogen (Hopkins & Dungait, 2010). Soil fauna, in fact, consume residue while in competition with microbial decomposers. By excluding mesofauna and macrofauna using fine-mesh litterbags, we were able to isolate decomposition mediated by microbial activity. Under faunal-exclusion, residue in aspirational corn released 268% more nitrogen by V6 (~20 days), and 60% more by VT (~60 days) than in business-as-usual corn. This larger divergence may more accurately reflect differences in nitrogen mineralization between systems and is consistent with higher microbial respiration potential in aspirational corn relative to business-as-usual corn. Even so, real system dynamics do not occur in isolation from soil fauna, and the relative contributions of fauna versus microbes to nitrogen release in the field remain unclear.

Trophic Complexity and Ecosystem Function

Biodiversity-ecosystem function theory predicts that increasing trophic complexity produces correspondingly complex ecosystem properties (Loreau et al., 2001; Hooper et al., 2005). Our results offer partial support for this prediction, though trophic complexity also made it more difficult to disentangle the relative contributions of soil fauna and microbial communities to decomposition. Coarse-mesh exclusion did not allow us to isolate whether faunal effects on decomposition operated directly, through physical fragmentation and ingestion of litter, or indirectly, by altering the microbial communities responsible for residue turnover. Once soil fauna were granted access, decomposition reflected the combined output of faunal processing and microbial activity, and these pathways could not be readily distinguished. This is further complicated by the observation that the influence of fauna on microbial activity appeared context dependent. Employing complementary approaches that monitor energy flux through soil fauna and microbial communities under exclusion using stable isotopes is necessary to address ongoing priorities for research on soil biodiversity and ecosystem functioning (Eisenhauer et al., 2026).

Decoupling Realized Activity from Functional Capacity in Microbial Communities

Realized microbial respiration was primarily driven by moisture content rather than management. Across systems, respiration increased with moisture content, with aspirational corn exhibiting the strongest positive relationship, likely reflecting the addition of organic amendments, which have been shown to increase crop yields by 27% because of microbial function (Luo et al., 2018). Soil moisture strongly regulates microbial metabolism because water availability governs substrate diffusion, nutrient transport, and microbial access to decomposable carbon sources. The dominant role of moisture observed here is consistent with previous studies identifying water availability as a primary control on soil respiration and decomposition (Orchard & Cook, 1983; Thapa et al., 2022). Given that ambient moisture inherently constrains realized respiration, we also measured respiration potential by saturating litter. In contrast to realized respiration, microbial respiration potential significantly differed among management systems. Aspirational and restored systems exhibited higher respiration potential than business-as-usual systems. These result suggest management either increases abundance or metabolic capacity of microbial communities, but this capacity may not be expressed under dry field conditions. Respiration potential also declined over time across all systems, likely reflecting progressive depletion of labile carbon pools during decomposition (de Graaff et al., 2010). Early decomposition is typically characterized by rapid consumption of readily available substrates, whereas later stages increasingly involve more recalcitrant compounds requiring greater energetic investment to decompose. This temporal decline suggests that the management-driven differences in respiration potential observed early in decomposition may narrow as the labile carbon pool is depleted across all systems, regardless of management.

 Restoring Function to Production Systems

Restoration of agricultural land to native habitat is increasingly promoted as an alternative to agricultural production (Rey Benayas & Bullock, 2012), offering benefits such as increased biodiversity and enhanced provisioning of ecosystem services (Schulte et al., 2017).  This approach is framed as a viable option for marginal lands to render them ecologically, if not economically, productive (Csikós and Tóth, 2023). Less often considered is the possibility of restoring ecosystem function and economically viable. There is no one-size-fits-all approach to management, and the degree of landscape simplification resulting from agricultural intensification should determine whether the goal is to restore integrity, increase multifunctionality, or pursue sustainable intensification (Landis, 2017). Sustainable intensification offers a viable alternative for increasing the functionality of agricultural land while remaining economically predictable and supporting ecosystem services through systems-based management (Giller et al., 2015).

Economic outcomes from 2023-2025 suggest that ecological benefits were not achieved at a fixed economic cost, and in some cases coincided with equal or greater profitability. In 2024, the business-as-usual system outperformed the aspirational system ($154 ha^-1 versus $123 ha^-1), driven by higher soybean returns ($192 ha^-1 versus $93 ha^-1), despite lower corn returns ($117 ha^-1 versus $154 ha^-1; KBS Long-Term Agroecosystem Research, 2025). However, aspirational soybean also experienced significant slug herbivory that went untreated, resulting in additional replants and poor stands—issues that may have been avoided with appropriate management.  In 2025, this relationship reversed: the aspirational system outperformed business-as-usual system ($131 ha^-1 versus $86 ha^-1), with substantially higher returns for both corn ($181 ha^-1 versus $102 ha^-1) and comparable soybean returns ($81 ha^-1 versus $71 ha^-1; KBS Long-Term Agroecosystem Research, 2026). While economic performance varied across years, these results suggest that sustainable intensification can narrow the gap between ecological restoration and continued agricultural production, offering an alternative management framework to recover ecosystem function.

Management System Shapes Above- and Belowground Diptera Biodiversity

Belowground dipteran communities demonstrated high sensitivity to management practices that dictate soil disturbance, residue retention, and increased agrochemical inputs. The severe suppression of detritivores and herbivores in conventional corn relative to all other systems strongly supports our first hypothesis and aligns with established literature documenting the destructive impacts of inversion tillage and broad-spectrum inputs on soil invertebrates (De Graaff et al., 2015). Mechanical tillage physically disrupts the larval microhabitats and fragile puparia of soil-dwelling flies, while burying organic matter reserves away from surface-active colonizers (Wardle, 1995). Conversely, the elevated abundance of detritivores across the aspirational systems in sustainably intensified corn and sustainably intensified wheat relative to the restored prairie suggests that sustainable intensification generates resource-rich conditions highly favorable to larval flies. The combination of no-till management and high crop residue retention provides a continuous supply of decaying plant material, promoting microbial biomass that nourishes microbivorous and detritivorous larvae (Mathew et al., 2012; Zuber & Villamil, 2016). Interestingly, the sustainably intensified perennial forage phase was an exception, remaining indistinguishable from the restored prairie. This likely stems from the establishment of a stable plant community and undisturbed soil matrix, which fosters a more structurally balanced Diptera community mirroring natural reference states rather than the volatile resource pulses of annual crops.

The systematic decline of larval parasitoids across all managed systems compared to the restored prairie represents a critical functional vulnerability. Even within the structurally complex and diverse agricultural systems, parasitoid emergence dropped significantly. Dipteran parasitoids require stable host populations, overwintering refugia, and continuous floral resources to sustain their complex life cycles (Al-Dobai et al., 2012). The dramatic suppression observed here confirms our second hypothesis and strongly indicates that within-field diversification alone cannot substitute for the stable ecological conditions provided by native, unmanaged semi-natural habitats. Restored grasslands clearly act as irreplaceable source habitats for these critical natural enemies within the wider agricultural mosaic.

Adult trophic guild responses in malaise trap communities were more selective, with management system significantly affecting detritivores, microbivores, and nectarivores. Among adult guilds, nectarivores exhibited the most pronounced and dynamic shifts. Sustainably intensified forage and sustainably intensified corn supported significantly higher annual nectarivore abundances than the restored prairie. Rather than indicating a true resource deficit in the prairie, this trend highlights a seasonal phenological mismatch and competitive dynamics. Reconstructed grasslands exhibit distinct floral gaps between early spring and late summer blooms, leaving flies temporary resource-limited (Delaney et al., 2015). Furthermore, highly competitive social bees often monopolize floral resources, with potential to displace solitary adult Diptera toward adjacent crop canopies where cover crops or weeds provide accessible, underutilized nectar resources because they are not dependent on high quality resources for a hive (Inouye et al., 2015; Lindström et al., 2016; Wignall et al., 2020).

Seasonal Dynamics in Trophic Guild Abundance

Evaluating seasonal phenology revealed that annual aggregation masks critical temporal windows during which management systems exert disproportionate functional control. In emergence trap communities, our third hypothesis was partially corroborated: fall-planted sustainably intensified winter wheat along with cover-cropped corn fields drove a massive detritivore and herbivore emergence pulse in early May. This provides robust empirical evidence for the green bridge effect for a trophic guild other than herbivores, predators, and parasitoids. By maintaining active root systems and accumulating vegetative organic matter throughout the winter and early spring, these systems sustain early-season larval development long before spring-planted conventional systems have accumulated meaningful biomass. As the season progressed, resource availability shifted dynamically across the rotation. The dramatic surge of detritivores and herbivores within conventional soybean fields in July potential for soybean to serve as a high-quality resource for some herbivorous taxa that may not necessarily be crop pests. By August, peak functional activity shifted back to aspirational canola and perennial forage tracking the late-season resource accumulation unique to these extended rotations. The near-complete convergence of detritivore and herbivore emergence by September points to a seasonal homogenization of belowground functional processes. This temporal compression highlights the necessity of multi-date sampling regimes; single-date assessments carried out late in the summer would completely fail to capture the critical early-season functional contributions maintained by sustainable intensification practices.

Implications for Agroecosystem Management

Synthesizing these findings demonstrates that neither sustainable intensification nor targeted habitat restoration alone can maintain the full functional complement of agricultural Diptera. Sustainable intensification effectively maximizes internal production services by enhancing detritivorous decomposition, microbivorous nutrient cycling, and adult nectarivore activity. However, the chronic suppression of parasitoid communities across all production treatments illustrates a persistent ecological limitation of working lands. Conventionally managed corn emerged as a clear functional sink for larval trophic guilds, consistently reducing abundances across almost every trophic guild temporally. The functional deficit observed in conventional corn underscores that the ecological penalties of intensive monoculture extend far beyond simple taxonomic richness metrics to structurally compromise the foundational biotic pathways that drive long-term soil health and pest suppression. These results firmly support an integrated landscape-level conservation framework that views sustainable intensification and land sparing as highly complementary, non-exclusive strategies (Kremen, 2015). Simultaneously, interspersed patches of restored prairie are mandatory to serve as stable ecological reservoirs, provisioning the landscape with the essential natural enemy complexes required to secure robust, long-term biological control. Diptera have the potential to serve as a bioindicator, ecosystem function, and soil health in agroecosystems, and detritivores respond strongly to sustainable intensification.