Content area
ABSTRACT
Intensifying the management of grazed pasturelands can lead to higher greenhouse gas emissions, loss of biodiversity, and a general decline in ecosystem function. Yet, maintaining the abundance and activity of soil organisms, such as dung beetles, in pasturelands may improve soil ecosystem function. The tunneling dung beetle,
Introduction
Aboveground management in ecosystems can strongly impact belowground ecosystem function (A'Bear et al. 2014; Hooper et al. 2000). This is particularly evident in agricultural systems, such as grazed pastures, when land managers shift aboveground management practices that, in turn, change ecosystem properties to improve soil health. When managed with a stocking rate that is appropriate for the land area and that includes adequate rotations, and thus minimizes overgrazing, pasture systems can store more carbon in their soil and become a net negative greenhouse gas system (Bossio et al. 2020; Lorenz and Lal 2018). However, when overgrazed, pastures can be net emitters of greenhouse gases (Conant et al. 2017). Emissions from grazed pasture systems include both nitrous oxide and methane released from manure and carbon dioxide released from overgrazed soils (Gerber et al. 2013). However, animal dung can also serve as an effective fertilizer that increases soil organic carbon (Maillard and Angers 2014). How dung is incorporated can determine whether a grazed pasture system emits greenhouse gases, or stores carbon in the soil (Hammer et al. 2016). Dung beetles are integral in this process, as their burrowing and nesting activities can dry out dung, decreasing its emissions, and also incorporate dung into the soil, providing a nutrient-dense substance that increases soil health (Nichols et al. 2008; Slade, Riutta, et al. 2016).
Dung beetles are functionally important in pastures due to their role in the promotion of soil health, as well as their importance as secondary seed dispersers and their ability to reduce fly abundance (Nichols et al. 2008). Pastureland management can impact dung beetle abundance and diversity (Barrágan et al. 2021), which can, in turn, impact ecosystem function (Menéndez et al. 2016; Noriega et al. 2023). There are three different functional types of dung beetles—dwellers, tunnelers, and rollers—and all have complementary effects on ecosystem functionality in pastures. All adult beetles consume the liquid portion of the dung and all beetle larvae grow in dung (Nichols et al. 2008). Because dung is the main feeding resource for both adults and larvae from all three functional groups of dung beetles, competition has facilitated the development of different strategies for exploiting the dung. Dwellers reside within the dung, where they deposit their eggs, which subsequently develop in the dung pats. Tunnelers dig directly underneath the dung pat, bringing a ball of dung with them downwards, and using this as a “broodball,” where they deposit their eggs. Finally, rollers form dung balls, which are rolled far away from the source, where they bury them (generally at a shallower depth than the tunnelers) using them for egg deposition and feeding of their developing larvae (Hanski and Cambefort 1991). All these strategies represent functional differences that potentially contribute to dung beetle complementarity in pasture systems.
During nesting, dung beetles affect soil nutrients, in particular nitrogen, phosphorus, and potassium either through direct pathways (providing nutrients through dung burial) or indirectly by modifying soil microbial communities (Hasan et al. 2024, Lipton et al. 2023, Stanbrook et al. 2021, Yokoyama et al. 1991). Their impacts on carbon, however, are less clear (Whitehead et al. 2018). There is evidence that tunneling dung beetles increase carbon in the surface layer of the soil after 3 days of dung tunneling activity (Yokoyama et al. 1991), and that the functional diversity of dung beetles impacts the amount of carbon leachate permeating the soil profile (Menéndez et al. 2016). However, the effects of dung beetle activities on total soil carbon, as well as soil fractions including mineral associated organic matter (MAOM) and particulate organic matter (POM), are understudied. When considering carbon sequestration in the soil, it is necessary to consider the type of carbon being measured, as MAOM is associated with long-term nutrient storage because of the mineral associations, while POM provides a more readily available source of nutrients, including carbon, for plants and soil microbes (Lavallee et al. 2020).
Dung beetles affect the soil microbial community at the soil surface (0–1 cm), increasing the relative abundance of bacterial and fungal taxa involved in carbon and nitrogen cycling (Lipton et al. 2023). Dung beetles also change the composition of soil bacterial communities at greater depths (8–9 cm), though the compositional changes are dependent on the species present, with lower evenness in soil bacterial communities when both tunnelers and dwellers are present as opposed to when just dwellers are present (Slade, Roslin, et al. 2016). Carbon and other nutrients are mediated by soil microbes, which drive decomposition by feeding on soil organic matter, while also contributing to soil organic matter through their extracellular metabolites and, importantly, through microbial necromass (Cotrufo et al. 2013; Schmidt et al. 2011). Few studies have looked at the relationship between dung beetles and soil microbes using next generation sequencing metabarcoding to identify the particular microbial taxa that dung beetles promote (Lipton et al. 2023; Slade, Roslin, et al. 2016). However, no studies have examined the relationships among dung beetles, soil microbes, and carbon.
We employed enclosure field experiments across three sites in California, USA, to investigate the relationships between tunneling dung beetle abundance, soil microbial composition, and soil organic carbon content. Specifically, we sought to identify which taxa within the soil microbiome are affected by the presence of dung beetles and to determine if changes in soil organic carbon content are associated with shifts in microbial community structure. We also investigated whether the abundance of any of the fungal taxa and other microbes associated with soil organic carbon was likewise associated with dung beetle abundance.
We asked the following questions: (1) Does the abundance of tunneling dung beetles influence soil organic carbon content, measured in total or in fractions? (2) Does tunneling dung beetle abundance influence soil microbial composition? (3) Which microbial families are driving observed differences in the soil microbial community composition treated with dung beetles? and, (4) Are specific microbial groups associated with both dung beetle treatment and soil organic carbon content?
Methods
Study Sites
We conducted the experiment across three ranches in and around the Central Coast of California, located in Pescadero, Watsonville, and Paicines, California (Figure 1a). The field sites are in a Mediterranean climate and range from coastal, which is wetter and colder than the other sites (Pescadero), to slightly inland with some coastal fog in the morning but warmer temperatures than the coast (Watsonville), to far inland where it is hotter and drier (Paicines). Pescadero soils are Dublin clay, Watsonville soils are Los Osos loam, and Paicines soils were at the bottom of a dried riverbed and are classified as Vallecitos Loam using the UC Davis Soil Web Survey app (O'Geen et al. 2017). Over the course of the experiment the average temperature in Pescadero was 20.98°C and the average relative humidity was 68.44%, the average temperature in Watsonville was 21.65°C and the average relative humidity was 62.29%, and the average temperature in Paicines was 22.23°C and the average relative humidity was 51.92%. Samples were taken in the summer months when there was no rain. The mean annual temperature in Pescadero in 2021 was 12.7°C and mean annual precipitation was 623.16 mm, the mean annual temperature in Watsonville was 14.6°C and mean annual precipitation was 709.05 mm, and the mean annual temperature in Paicines was 14.6°C and mean annual precipitation was 374.2 mm (PRISM Climate Group, Oregon State University).
[IMAGE OMITTED. SEE PDF]
Study Species
We collected
Dung Beetle Enclosures
We established enclosures on recently grazed pastures across the three sites with three replicates of each treatment at each site: high dung beetle, low dung beetle, and no dung beetles (control) for a total of 27 enclosures. The three sites were at least 44 km apart from each other. Each site was relatively flat with similar soil consistency throughout the site. We constructed enclosures from 30.48 cm diameter, 22.73 L plastic buckets with the bottoms cut out. We dug the buckets into the ground to a depth of approximately 7 cm with minimal disturbance to the soil and vegetation inside of the enclosures. Following the establishment of the enclosures, we placed dung and dung beetles into the enclosures (except for the control, which only had dung) and covered them with a mesh window screen netting secured with a rubber band. This prevented the beetles from flying away. In the week before the experiment, we collected fresh cow dung from each ranch that had been recently deposited on the field (within the previous 6 h). We homogenized dung by ranch, allocated the dung into 400 g portions, froze it for at least 1 week at −20°C and then defrosted the dung before placement in the traps, following Manning et al. (2016). At each site we also included one temperature and relative humidity sensor (HOBO) inside one enclosure (without dung beetles) set to measure temperature and relative humidity every hour for the duration of the experiment.
We based the high and low dung beetle treatment numbers off of results from a pilot experiment (Lipton et al. 2023). The results of the pilot experiment showed no difference in the effect of six (low) and 22 (high) dung beetle treatments; we therefore increased the magnitude of difference in dung beetle abundance from 3.6 times to 5 times, with three beetles in the low abundance treatment and 15 beetles in the high abundance treatment to see if this created a difference in treatment effect.
We set up the enclosures in June 2021 within a 4-day period (June 16: Watsonville; June 18: Pescadero; June 19: Paicines), and let them sit for 60 days, until August, 2021.
Soil Collection
We collected soil for microbial, carbon, and nitrogen analysis at several points across the study. Using a 2.54 cm diameter soil probe, we took soil samples to a 10 cm depth (referred to as “core soil”) at days 0, 21, and 60.
Soil Carbon and Nitrogen
We measured soil carbon and nitrogen from the 0 to 10 cm soil core at 0 and 60 days. Soil samples were air dried until they maintained a constant weight. We separated approximately 5 g subsamples of soil into size-separated mineral-associated organic matter (MAOM) and particulate organic matter (POM) using a wet sieving method and a 53 μm sieve, following the methods of Cotrufo et al. (2019). We then ground these samples using a Retsch 20.745.0001 Mixer Mill MM400, grinding soils to a fine powder following the manufacturer's instructions. MAOM samples did not need to be ground and these were instead shaken in their plastic vials until they broke apart into a fine powder. We took 5–6 mg subsamples of the ground and homogenized soils and packaged them into Costech (Valencia, CA) 5 × 9 mm tin capsules. The packaged soil samples were sent for carbon and nitrogen elemental analysis at the University of California, Santa Cruz's Stable Isotope Laboratory on a Carlo Erba EA1108.
Extraction, Library Preparation, and Sequencing
We extracted DNA from 250 mg of soil samples using the Qiagen DNeasy Powerlyzer PowerSoil Kit (Qiagen, Germantown, MD) following the manufacturer's instructions and extracted two replicates from each sample (Figure 1). Soil samples used for microbial analyses of core samples at days 0 and 60 were subsamples of the samples used for nutrient analysis. We included a blank extraction with every batch of extractions, which included 12–24 samples. Extracted DNA was stored at −20°C before amplification. We amplified bacteria and archaea using the 16S marker (515F and 806R, Caporaso et al. 2011) and fungal ITS1 marker (ITS5 and 5.8S, White et al. 1990, Epp et al. 2012). We amplified samples in triplicate and included eight template-free negative controls (see Figure 1b for a visual of this process). Triplicate PCR replicates were ultimately combined by sample, though we kept extraction replicates for each sample separate. We followed the methods for amplification in Lin et al. (2021) using three 15 μL reaction mixtures with 7.5 μL of Qiagen Multiplex PCR Plus 2× Master Mix, 6.2 μL of deionized water, 0.15 μL of each primer, and 1 μL of template DNA. We conducted PCR cycling for 16S and ITS1 in a thermocycler using a touchdown program of initial denaturation for 15 min at 95°C, followed by 13 denaturation cycles at 94°C for 30 s, beginning annealing for 30 s at 69.5°C, and then decreasing the temperature by 1.5°C every cycle until it reached 50°C followed by extension at 72°C for 1 min. We then carried out an additional 30 cycles at an annealing temperature of 50°C, instead of 35 cycles as in Lin et al. (2021) to avoid overamplification, and followed this with a final extension at 72°C for 10 min. We confirmed amplification of each sample replicate using gel electrophoresis and subsequently pooled sample replicates by sample and then cleaned them with MagBio HighPrep PCR Clean-up beads (MagBio, Gaithersburg, MD) at a ratio of 1.2 beads: sample. We quantified pooled samples with Qubit dsDNA BR assay kit on a Varioskan LUX microplate reader (Thermo Scientific, Waltham, MA) and then pooled 16S and ITS1 amplicons to equimolar levels by sample. We indexed the pools with Nextera Unique Dual Indexes, bead cleaned again with a ratio of 1.1 beads: sample, quantified them again, and then pooled for an equal number of molecules per sample before sequencing on an Illumina NextSeq 550 (Illumina, San Diego, CA) for 2 × 150 reads at the UCSC Paleogenomics Sequencing Lab where they also sequenced other dual unique indexed metabarcoding libraries on the same run and added 20% PhiX to the sequencing runs. We aimed for a targeted depth of 40,000 reads/sample/marker.
Bioinformatics
We processed Fastq results in Anacapa (Curd et al. 2019) to remove low-quality reads and assign taxonomy. We set minimum read quality to 32, used updated CRUX “v2” 16S and Fungal ITS1 “FITS” reference databases, and otherwise used default settings. Following taxon assignment, we created phyloseq objects using the phyloseq package (McMurdie and Holmes 2013) for downstream analysis in R version 4.5.0 (R Core Team 2024). We decontaminated samples with the Decontam R package (Davis et al. 2018) with the Prevalence method and a threshold of 0.1 and then removed negative controls (i.e., extraction blanks and PCR blanks) and taxa with fewer than 10 total reads.
Data Analysis
We determined whether dung beetles affect soil MAOM carbon (C), soil POM C, and total C by subtracting the percent weight of carbon in the soil (wt %C) at day 0 from the percent weight of carbon in the soil at day 60 in each enclosure and running a linear mixed effect model (LME) to determine if the results varied with dung beetle abundance treatment. We did the same for MAOM nitrogen (N), POM N, and total N. For each type of soil nutrient measurement, we had 54 samples. We ran the LME using the lme function in the nlme package (Pinheiro et al. 2023) and subsequent analyses using multcompView (Graves et al. 2024) and emmeans (Lenth 2024). We set the ranch as the random intercept to account for variability within ranches. We also ran additional LMEs with the data separated by ranch and the enclosure as the random intercept.
We measured the effect of dung beetle abundance on microbial composition using a Bray-Curtis PERMANOVA, with the default number of permutations of 999, with ranch as strata to account for non-independence between samples from the same ranch, and including day in the experiment, dung beetle treatment, total soil C and N, POM C and N, and MAOM C and N as explanatory variables, using soil samples with a minimum of 4000 reads. We chose 4000 reads after examining rarefaction curves (Figure S1) created using phyloseq and ggplot2 (Wickham 2011), and decided to use a minimum number of reads instead of rarefying the samples so that we could preserve more data (McMurdie and Holmes 2014). We had a total of 54 samples for each PERMANOVA (two replicates × 27 total enclosures) To visualize these differences, we used canonical correspondence analyses (CCAs) in the vegan package with a phyloseq wrapper. We also disaggregated the samples by ranch to see if PERMANOVA effects were more apparent. In this case the PERMANOVAs also included day as a variable so that we could increase our sample size to 36 (two replicates × 9 enclosures/ranch × 2 days).
We also ran a Bray-Curtis PERMANOVA of the surface soils including just dung beetle treatment, day in the experiment, and an interaction effect of day and dung beetle treatment on samples with a minimum of 4000 reads. Surface samples were taken at one ranch at days 21 and 60; we were not able to include soil C and N factors in these PERMANOVAs because we did not take samples for C and N at day 21. We used the adonis2 function in vegan (Oksanen et al. 2025) for this analysis. We used principal coordinate analyses to visualize the effects of dung beetle treatment and day on the surface soils.
We also investigated which fungal families changed in relative abundance based on dung beetle treatment using the DESeq2 package available for R (Love et al. 2014). DESeq2 provides the log2fold change in relative abundance of reads between two treatments. Binning read abundance at the family level reduces incorrect taxonomic assignment issues while being useful for interpreting patterns in a functional context. While there is a lot of ecological plasticity within families, many fungal families share ecological traits (Powell et al. 2009). We prepared the data for this test by filtering samples to only include taxa that have more than 3 reads in at least 10% of the samples. The DESeq object was made using a Wald test, local fitType, poscounts of sfType and the significance below an alpha of 0.01.
To predict potential associations between the presence or absences of soil-associated fungal taxa, organic carbon content, and dung beetle treatment we ran random forest models, using the R package randomForest (Liaw and Wiener 2002). The results show fungal families associated with multiple predicting factors based on their presence/absence, and not their change in relative abundance differential between communities under different treatments, as with DESeq2. For these analyses we used soil samples with a minimum of 4000 reads. To build and evaluate all random forest models we randomly selected 80% of the data for model training, with the remaining 20% used for testing. In this study we ran 100 model iterations. We used the default number of trees, which is 500, and used the default number of variables tried at each split (Liaw and Wiener 2002). To evaluate model accuracy we calculated the mean and standard deviation for the True Skills Statistic (TSS) scores of these 100 iterations. The TSS is interpreted as the true positive rate plus the true negative rate, minus one. According to Thuiller et al. (2019), if the mean TSS score exceeds 0.4 the model is considered accurate, and the evaluation of the presence/absence of the fungal family is considered an accurate indicator of the environmental data in the model. For all models we calculated the relative importance of predictor variables using the importance function in the randomForest package. We calculated the relative importance of predictor variables in all models with their mean decrease in node impurity, quantified by their Gini indices. A higher rank importance indicates that the specific variable has a greater effect on the presence or absence of a specific fungal family.
We ran two different types of random forest models to first test whether fungal families are predicted by certain factors (including soil nutrients), or if their presence can predict soil carbon content. In random forest model 1, we looked at samples taken across all sites (Paicines, Watsonville, Pescadero) at day 60, after the full enclosure trial period. For this first model all soil factors (total C, total N, MAOM C, MAOM N, POM C, and POM N), as well as dung beetle treatment, ranch, extraction batch, and enclosure (i.e., which enclosure was sampled) were set as predictor variables and the presence/absence of fungal families was set as the response variable. We grouped samples across all sites in order to have enough samples to meet a threshold of 30 samples (van Proosdij et al. 2016); thus we had a total of 54 samples (two extraction replicates × 27 enclosures at day 60). For the first model we evaluated the accuracy using a TSS score (Allouche et al. 2006). The rank importance of model variables was calculated from their Gini indices, then visualized using heat maps generated using ggplot2 (Wickham 2011). For random forest model 2, we used the presence/absence of each family as predictor variables to predict the response of MAOM, POM, and total carbon content. This model also contained samples from day 60 and all sites (Paicines, Watsonville, Pescadero). For this model, we evaluated accuracy using a Pearson correlation between the predicted and actual carbon content within the testing data. In order to summarize the correlation coefficients, we first Fisher transformed the values, then calculated the mean and standard deviation, and then reverse Fisher transformed the resulting values, using the DescTools package (Signorelli et al. 2024). See Figure S2 for a flow chart of the samples used in random forest analyses and respective analyses.
Results
Does the Abundance of Tunneling Dung Beetles Influence Soil Organic Carbon Content?
Using a linear mixed effects model, we found that the abundance of tunneling dung beetles had no effect on the change in MAOM, POM, or total carbon or nitrogen over time when aggregated or disaggregated by site (Table 1 and Table S3; Figures 2 and Figures S3, S4).
TABLE 1 Linear mixed effects model results for the effect of dung beetles on carbon and nitrogen, MAOM, POM, and total.
| Treatment comparison | Estimate | SE | df | t ratio | p value |
| (a) POM C | |||||
| High–Low | −1.228 | 0.88 | 22 | −1.395 | 0.388 |
| High–None | 0.009 | 0.445 | 22 | 0.02 | 0.9999 |
| Low–None | 1.237 | 0.927 | 22 | 1.335 | 0.4209 |
| (b) MAOM C | |||||
| High–Low | −0.169 | 0.475 | 22 | −0.356 | 0.9483 |
| High–None | −0.183 | 0.501 | 22 | −0.366 | 0.9453 |
| Low–None | −0.014 | 0.347 | 22 | −0.042 | 0.9994 |
| (c) Total C | |||||
| High–Low | −0.497 | 0.266 | 22 | −1.868 | 0.1842 |
| High–None | 0.241 | 0.292 | 22 | 0.826 | 0.7283 |
| Low–None | 0.738 | 0.381 | 22 | 1.935 | 0.1633 |
| (d) POM N | |||||
| High–Low | −0.062 | 0.053 | 22 | −1.166 | 0.5196 |
| High–None | 0.006 | 0.025 | 22 | 0.223 | 0.9803 |
| Low–None | 0.068 | 0.052 | 22 | 1.297 | 0.442 |
| (e) MAOM N | |||||
| High–Low | −0.018 | 0.044 | 22 | −0.402 | 0.9336 |
| High–None | −0.023 | 0.045 | 22 | −0.523 | 0.8872 |
| Low–None | −0.006 | 0.035 | 22 | −0.16 | 0.99 |
| (f) Total N | |||||
| High–Low | −0.0333 | 0.0224 | 22 | −1.488 | 0.3396 |
| High–None | 0.0211 | 0.0252 | 22 | 0.836 | 0.7222 |
| Low–None | 0.0544 | 0.0326 | 22 | 1.669 | 0.2573 |
[IMAGE OMITTED. SEE PDF]
Does the Tunneling Dung Beetle Abundance Influence the Soil Microbial Composition?
After decontamination, our 16S dataset had 4,173,158 reads with 55 phyla, 118 classes, 245 orders, 456 families, 1271 genera, and 3245 species identified (Table S1). There were an average of 30,082 reads per sample and 14.9% of reads were unassigned. The decontaminated fungal ITS1 dataset had 7,283,190 reads, with 8 phyla, 43 classes, 126 orders, 308 families, 860 genera, and 1584 species identified (Table S2). There were an average of 55,374 reads per sample and 0.67% of reads were unassigned. The reads reported are after decontamination through the Decontam package (Davis et al. 2018).
Core Soil (0–10 cm)
Using Bray-Curtis PERMANOVAs with ranch set as strata, we found that dung beetle abundance had significant influences on the composition of fungal (fungal ITS1, here shortened as “FITS”) communities, but not bacterial/archaeal 16S communities in soil samples collected from 0–10 cm depth; similarly, only fungal composition varied with changes in soil C and N, as POM, MAOM, and total (Table 2). At day 60, neither soil factors nor dung beetle treatment had significant effects on the compositions of bacterial/archaeal communities.
TABLE 2 PERMANOVA output for soils across all ranches showing relationships between dung beetle treatment, day of experiment, and several soil carbon and nitrogen factors and (a) FITS microbial composition in all sites, (b) 16S microbial composition in all sites, and (c), relationships between dung beetle treatment, day in experiment (days 21 and 60 only), and the interaction between dung beetle treatment and day in experiment and 16S and FITS microbial composition in surface soil communities at Paicines Ranch.
| (a) FITS PERMANOVA output for soils in all ranches | ||||||
| FITS for day 60 only | ||||||
| df | Sum of squares | R 2 | F | Pr(>F) | ||
| Dung Beetle | 2 | 0.495 | 0.043 | 1.656 | 0.001 | |
| MAOM C | 1 | 1.938 | 0.168 | 12.958 | 0.057 | |
| POM C | 1 | 1.344 | 0.117 | 8.987 | 0.001 | |
| Total C | 1 | 0.346 | 0.030 | 2.313 | 0.036 | |
| MAOM N | 1 | 0.343 | 0.030 | 2.291 | 0.002 | |
| POM N | 1 | 0.228 | 0.020 | 1.524 | 0.037 | |
| Total N | 1 | 0.252 | 0.022 | 1.686 | 0.030 | |
| Residual | 44 | 6.579 | 0.571 | |||
| Total | 52 | 11.523 | 1 | |||
| (b) 16S PERMANOVA output for soils in all ranches | ||||||
| 16S for day 60 only | ||||||
| df | Sum of squares | R 2 | F | Pr(>F) | ||
| Dung Beetle | 2 | 0.163 | 0.028 | 1.042 | 0.256 | |
| MAOM C | 1 | 1.285 | 0.218 | 16.410 | 0.149 | |
| POM C | 1 | 0.505 | 0.086 | 6.456 | 0.867 | |
| Total C | 1 | 0.227 | 0.038 | 2.899 | 0.086 | |
| MAOM N | 1 | 0.097 | 0.016 | 1.239 | 0.343 | |
| POM N | 1 | 0.131 | 0.023 | 1.676 | 0.143 | |
| Total N | 1 | 0.121 | 0.021 | 1.549 | 0.195 | |
| Residual | 43 | 3.366 | 0.571 | |||
| Total | 51 | 5.895 | 1 | |||
| (c) 16S and FITS surface soil Paicines ranch days 21 and 60 PERMANOVA results | ||||||
| 16S | FITS | |||||
| R 2 | F | P | R 2 | F | P | |
| Dung Beetle Treatment | 0.200 | 4.328 | 0.001 | 0.109 | 2.007 | 0.001 |
| Day | 0.062 | 2.684 | 0.016 | 0.070 | 2.606 | 0.002 |
| Interaction | 0.043 | 0.926 | 0.505 | 0.064 | 1.186 | 0.183 |
Site had a large effect on the microbial community for both 16S and FITS regions (Figure 3) so we also examined the impacts of dung beetles, MAOM C and N, POM C and N, and Total C and N, for each ranch separately (Tables S4 and S5). However, in looking at the effects of day in treatment, dung beetle treatment, and all soil factors by ranch, we find that dung beetles significantly affected the fungal communities across all ranches and still did not affect the bacterial/archaeal community (Tables S4 and S5). Some soil factors had a significant effect on the soil bacterial/archaeal communities at different ranches, but the effects were not consistent across all ranches (Tables S4 and S5).
[IMAGE OMITTED. SEE PDF]
Surface Soil
In contrast to the core soil, we found that dung beetles have a strong effect on both the 16S and FITS communities in the soil collected from the top 0–1 cm of soil (Table 2c). The surface soil samples were taken at only one ranch (Paicines), at days 21 and 60. In surface soils at Paicines, the effect of dung beetle treatment is stronger than the effect of the day in the experiment for both 16S and FITS communities (Table 2 and Figure S5).
Which Families Are Driving the Changes in Fungal Communities by Dung Beetles?
As the differences in the core soil sample microbial communities driven by dung beetle treatment were only found in fungal communities, we only proceeded with further analyses of the FITS community. Overall, we found several differences in the representation of different families in enclosures with different dung beetle abundances at day 60 (Figure 4, Table S6, see Table S7 for day 0 differences). Across all sites, Morchellaceae was overrepresented in samples where dung beetles were present, compared to the controls without dung beetles (Figure 4a,b). However, fungal communities were quite varied in the different ranches, and we therefore examined the data for each ranch separately as well. We found differences between high-, low-, and no-dung beetles at each ranch. At the Paicines site, at day 60, Didymellaceae, Ascodesmidaceae, Morchellaceae, and Trichiidae were all overrepresented in the low- versus no dung beetle treatments (Figure 4d), and Pyronemataceae, Psathyrellaceae, Niessliaceae, Morchellaceae, Clavariaceae, Hyaloscyphaceae, Ascodesmidaceae, and Trichiidae were all overrepresented in the high- vs. no-dung beetle treatment (Figure 4e). At Pescadero, at day 60, Herpotrichiellaceae and Helotiaceae, as well as the order Coniochaetales, were overrepresented in the low- compared with no-dung beetle treatment (Figure 4g), and Helotiaceae was overrepresented in the high- versus no-dung beetle treatment (Figure 4h). Finally, at Watsonville, we found overrepresentation of Ambisporaceae, Exidiaceae, and Ascobolaceae in the low- compared with the no-dung beetle treatment (Figure 4j). Moreover, Ascobolaceae and Exidiaceae were overrepresented in the high- compared with the no-dung beetle treatment (Figure 4k). There were also differences between the low- versus high- dung beetle treatments at all sites, but by day 60 there were generally fewer differential families than in the no- versus high- or no- versus low- dung beetle comparisons (4f, 4i, 4 L).
[IMAGE OMITTED. SEE PDF]
Which Taxa Are Associated With Dung Beetle Treatment or Changes in Carbon?
Random forest model 1 predicted the presence/absence of fungal families based on dung beetle treatment, carbon and nitrogen factors, extraction batch of soil samples, and enclosure number of soil samples at all ranches at day 60. Several fungal families that are overrepresented with the presence of dung beetles in DESeq2 were also reported as response variables in random forest model 1; the families with the highest rank importance for dung beetles in the random forest model were Exidiaceae (7), Mucoraceae (5), and Diversisporaceae (4) (Figure 5; Table S8). The enclosure number had a rank importance of 10 for almost all fungal families, except for Microdochiaceae (4), and Trichomeriaceae (9). Soil C factors were generally more important than soil N factors and total C and total N were more important than MAOM or POM for C or N. Ranch ranged in importance from 1 to 9 for all families.
[IMAGE OMITTED. SEE PDF]
Which Taxa Are Associated With Changes in Carbon?
In random forest model 2, we set carbon level as the response variable and determined which taxa are indicators for total, MAOM and POM carbon level with the best fit model to determine if there are specific fungal families associated with carbon in the soil (Table 3). The Mean Decrease Gini coefficient reported here is an indicator of how important the presence or absence of each fungal family is in observing a change in carbon levels. The correlation coefficient between the actual predicted values for total C, MAOM C, and POM C is 0.971 ± 0.229, 0.942 ± 0.168, and 0.954 ± 0.240, respectively. We only show the top important families until there is a significant drop in the Mean Decrease Gini, indicating that the family is less important to the model. The top families in predicting MAOM C are also the top families in predicting total C. Only one family, Microdochiaceae, appears as an indicator family for both POM C and MAOM C.
TABLE 3 Random forest 2 results. Mean and standard deviations (SD) of decreases in Gini Index values that show the fungal families that increase the likelihood of observing (a) changes in Total C, (b) changes in POM C, and (c) changes in MAOM C as determined with a random forest model. Higher values indicate greater importance of the carbon pool to the model. We included the families with the highest Gini indices and did not report those once the Gini index dropped precipitously.
| Variable | Mean decrease Gini | SD decrease Gini |
| (a) Total C | ||
| Microdochiaceae | 34.773 | 4.188 |
| Trichomeriaceae | 21.393 | 6.389 |
| Rhynchogastremataceae | 13.586 | 5.260 |
| Pluteaceae | 12.025 | 5.124 |
| Hyaloriaceae | 10.096 | 3.509 |
| (b) POM C | ||
| Dothioraceae | 36.013 | 15.801 |
| Hyaloriaceae | 35.893 | 10.749 |
| Didymellaceae | 16.217 | 9.597 |
| Wallemiaceae | 15.637 | 8.309 |
| Ceratobasidiaceae | 14.435 | 8.193 |
| Cuniculitremaceae | 13.475 | 7.197 |
| Microdochiaceae | 13.378 | 6.844 |
| (c) MAOM C | ||
| Microdochiaceae | 17.409 | 2.247 |
| Trichomeriaceae | 12.061 | 2.840 |
| Rhynchogastremataceae | 7.164 | 2.295 |
| Pluteaceae | 6.837 | 1.949 |
| Erysiphaceae | 3.047 | 1.405 |
| Polyporaceae | 2.015 | 1.111 |
Discussion
Our results show that dung beetles strongly influenced fungal community composition both at the soil surface and down to 10 cm depth, while effects on bacterial and archaeal communities were restricted to the surface; however, dung beetles did not directly affect soil organic carbon in the time frame of this experiment. The lack of effect on soil organic carbon may reflect the relatively short duration of the experiment, which may not have been long enough to capture consistent changes, that can take several years to emerge (Schuman et al. 2002), or the sampling schedule may have missed shorter-term pulses as seen in other experiments (Evans et al. 2019; Smith 2004; Yokoyama et al. 1991). It is also possible that the warm temperature led to higher microbial death and recycling of DNA so some effects are missed, as soil eDNA is both in-cell and free DNA from both the living and dead microbes (Schnecker et al. 2024). Another possibility is that the effects were not detectable at the depth we sampled. Evans et al. (2019) found an effect of dung beetles on carbon in 0–10 cm but not 10–20 cm depth soil cores; this difference was also most notable after 14 days, compared to the other days measured (1, 3, 7, 28, and 56 days). Their study shows that both sample depth and timing are important. Similarly, another study that found dung beetles impact soil carbon and nitrogen at the 0–5 cm depth under dung pats after 40 days (Owen et al. 2006) as compared to the 60 days that we measured. Several studies have also shown that dung removal and nutrient cycling changes with dung beetle diversity, which may reflect the specific depths that different beetles tunnel, or the amount of dung that different beetles are able to incorporate into the soil (Noriega et al. 2023, Menéndez et al. 2016, Stanbrook and King 2022).
The effect of dung beetles on the fungal, bacterial, and archaeal surface communities confirms the effect seen in a previous pilot project of a similar design (Lipton et al. 2023). Dung beetles may change the soil microbial community by changing soil porosity, increasing air, water flow, and nutrient flow through soil (Brown et al. 2010; Menéndez et al. 2016). Increased porosity changes the soil structure and allows for different fungi and bacteria to enter the soil (Erktan et al. 2020), which may explain the changes seen in the fungal community when dung beetles are present. Changes may also occur with the incorporation of dung into the soil, shifting the soil microbial community to mimic the dung microbiome (Slade, Roslin, et al. 2016), or they may modify the microbial community through excretions because dung beetles have their own unique microbiome that is transferred via maternal lineage (Parker et al. 2020).
There are several reasons that we see an effect of dung beetles on the bacterial community at the surface, but not at a 10 cm depth. Changes in the microbial community due to management, heat, and water are more detectable at the surface than in lower soil profiles (Barbour et al. 2022; Dove et al. 2021). Additionally, the sampling method that we used of compiling cores in and around the dung beetle nest means that we are incorporating more soil that may not have been affected by the dung beetle treatment, decreasing any signal from the treatment. Finally, we could only compare samples at days 0 and 60, so we may have missed any short-lived shift in the bacterial community due to dung beetle activity. This can be seen in the surface soil results, where the difference in soil microbial communities by dung beetle treatment at day 21 is more pronounced, but is less clear at day 60 (Figure S5).
It is notable that we do see an effect of dung beetle abundance on the fungal community at 10 cm. In general, bacterial and archaeal communities turn over more quickly than fungal communities (Hannula et al. 2019), so the dung beetle effect may not have lasted as long in the bacterial/archaeal communities and would have decreased with time as noted above. However, in an experiment on the effects of leaf litter addition to soil, the abundance of fungal communities increased more than those of bacterial communities, likely because fungi specialize in degrading larger organic molecules, whereas bacteria primarily utilize the simpler monomers that are generated from these processes (Habtewold et al. 2020). It is possible that fungal communities are similarly more susceptible to changes in nutrient inputs from dung.
The DESeq2 results, which show the difference in relative abundance of fungal families by dung beetle treatment, consistently showed differences between high and low dung beetle treatments throughout the experiment (Figure 4). The differences between the treatments may indicate that dung beetles are either harboring certain fungal species that grow in the soil, or changing the fungal composition of the soil by adding more nutrients (dung) to the soil. These results underscore the importance of dung beetle abundance particularly when abundances are low. This is in contrast to the pilot study for this experiment, in which we saw little to no difference between high and low dung beetle abundance treatments when the low treatment was six beetles and the high treatment was 22 beetles (Lipton et al. 2023). Dung beetle abundance depends on the landscape surrounding the pastureland, grazing management, and the season (Barragán et al. 2011; Barragán et al. 2021; Daniel et al. 2022). According to a paired observation experiment we did, in certain seasons and grazing scenarios, the low end dung beetle abundance may be closer to the abundances used in this experiment (Lipton et al. 2025). This shows how management and landscape factors influencing dung beetle abundance can have cascading effects on soil microbial communities and functions.
The canonical coordinate analyses show that differences between the sites clearly affected the soil microbial community, as evidenced by the clustering of points from each site (Figure 3). This could be because the microbial communities are heterogeneous and geographically structured, or because of inherent differences in the soil, climatic conditions, or grazing management of the sites (Drenovsky et al. 2010; Fierer and Jackson 2006; Waldrop et al. 2017). For example, the soil at Paicines was silty and sandy and also had lower C and N content than the other soils, and dung beetles had a significant effect on total nitrogen at this location (see Table S3). Soil C and N, as well as soil mineral content are primary drivers of the composition of soil microbial communities (Waldrop et al. 2017); in the PERMANOVAs disaggregated by ranch, the fungal community shifts with carbon and nitrogen content, perhaps indicating that the baseline fungal community is more affected by subtle shifts in nutrients (Table S5). However, while all C and N factors appeared to be a large driver of soil fungal communities when looking across all study sites, they had much smaller and sometimes nonsignificant effects when disaggregated by site. This could reflect that the soil microbial community was changing by site and that soil C and N factors were also quite different by site, but not different between samples at the same site.
While DESeq2 shows the difference in relative abundance of fungal families by dung beetle treatment, random forest model 1 shows the presence/absence of specific fungal families based on multiple predictor variables, including dung beetle treatment. The fungal family Morchellaceae increased in relative abundance with both low and high dung beetle treatments in the DESeq2 results across all ranches. In the random forest results, dung beetles had a rank importance of 3 for Morchellaceae, while factors such as total C and POM N had higher rank importances of 9 and 8 respectively. Therefore, the abundance of Morchellaceae increases with dung beetles, though their presence had a minimal effect compared to nutrient factors in the random forest. In soils inoculated with cultivated species of Morchellaceae, Morchellaceae rapidly colonize soils and exclude other fungi, decreasing diversity (Benucci et al. 2019; Zhang et al. 2023), which may account for its high relative abundance in this soil. Morchellaceae exhibited a high relative abundance with dung beetles in DESEq2 results when the data was aggregated across all sites, but when we disaggregated the data by site, we only saw a higher relative abundance at the Paicines site. Therefore, this result may have been a statistical artifact driven by an increase at Paicines. We believe the differences between the random forest and DESeq2 results likely arise from the different modeling approaches, and the inclusion of different predictor variables in the random forest models.
Across ranches, dung beetles increased the relative abundance of soil fungi that are implicated in nutrient cycling, or indicative of disturbance in the soil. While soil carbon did not show a detectable change during the experiment, shifts in the abundance of certain microbial families suggest alterations to the soil microbial community that may signal ongoing but still undetectable changes in soil organic carbon, or pulse changes that have already subsided. The relative abundance of Ambisporaceae increased with dung beetle presence in Watsonville; this fungal family is an AMF spore and has been correlated with degraded soil in Brazilian semiarid grasslands (da Silva et al. 2022), potentially indicating that dung beetles increase AMF which help with plant nutrient acquisition. Exidiaceae's relative abundance increased with dung beetle presence in Watsonville and is associated with white rot and wood decay (Worrall et al. 1997), indicating beetles may be increasing lignin or cellulose decomposition in the soil. Ascobolaceae also increased in relative abundance with dung beetles in Watsonville and has previously been shown to increase with compost application (Dang et al. 2021); in this case its relative increase in abundance is likely due to the influx of nutrients from dung. Herpotrichiellaceae increased with dung beetle presence in Pescadero and is correlated with N fertilization on forest floors (Weber et al. 2013) and has a documented effect on N microbial metabolism (Cui et al. 2018), showing dung beetle activity has similar effects to nitrogen fertilization. Helotiaceae also increased with dung beetle abundance in Pescadero and is positively associated with carbon accumulation in peat soils (Zhang et al. 2017), showing a potential link of dung beetles to carbon in the soil.
The presence of other fungal families was also positively associated with dung beetles in Paicines, especially in the “no vs. high” dung beetle treatment comparisons in DESeq2 (Figure 4). Many of the fungal taxa associated with the dung beetle presence are known to be associated with the addition of nutrients to soil, or the decay of organic matter. Ascodesmidaceae is a soil and dung saprotroph that was positively associated with N fertilization in corn fields over a 10-year period (Tosi et al. 2021), showing dung beetle presence may have similar effects on nitrogen cycling as N fertilization. Trichiidae is a slime mold common to decaying wood (Clissmann et al. 2015), again implying that dung beetles may increase lignin or cellulose degradation. Pyronemataceae increases with dung beetles and exhibits high laccase activity, the enzyme responsible for degrading lignin (Hadibarata and Yuniarto 2020). Niessliaceae increases with dung beetles and feeds on decaying matter or wood (Huang 2021); it was also more highly represented in grazed meadows than non-grazed or restored meadows in a study in the Tibetan grasslands (Wang et al. 2021). Hyaloscyphaceae increased with dung beetle abundance and is known to have high abilities to decay organic matter and dominate in mineral soils in a mountain beech forest (Mayer et al. 2021), potentially showing an increase in organic matter decay in these soils as well. Other fungal taxa may be evidence of the changes that dung beetles have on soil structure. Clavariaceae, which increased in relative abundance with dung beetle abundance in Paicines, is a generic saprotroph found in soil that was found to decrease in droughted greenhouse soils (Cordero et al. 2023). Psathyrellaceae, which increased in relative abundance with dung beetles at Paicines, contains genera that are saprotrophic and others that are coprophilous (Larsson and Örstadius 2008) and thrives in well aggregated soil with more oxygen (Yang et al. 2019). The increase in its presence may therefore be indicative of the effects of dung beetles incorporating dung into the soil and also aerating it with their tunneling.
In the results of the random forest 1, the presence of dung beetles was found to be an important predictor for just one family: Exidiaceae (7). According to the DESeq2 results, Exidiaceae increased in relative abundance in Watsonville as well, showing that dung beetles increase both the likelihood of presence as well as the relative abundance of this lignin and cellulose degrading fungal family. MAOM C and Total N also had moderately high rank importances for predicting Exidiaceae presence. However, Exidiaceae was not included as a predictor family for random forest model 2, which was designed to see if there are specific fungal families that can predict carbon content. Both random forest models can only imply association, because fungal species can increase with increased soil carbon or they can increase carbon via their metabolic products or necromass (Cotrufo et al. 2013; Schmidt et al. 2011). A future inoculation experiment with candidate fungal families associated with carbon could determine if the fungal families are aiding in carbon formation or increasing due to carbon.
Conclusions
We found that dung beetles do, in general, affect the soil microbial community, but the change is evident in the bacterial/archaeal community only in the surface soils, while in the fungal community the effect is smaller in 0–10 cm cores than in the surface soil. While dung beetles did not directly affect carbon or nitrogen during this experiment, we did observe shifts in the microbial community that may be indicative of changes in soil nutrients. Several of the fungal families that increased in relative abundance in the presence of dung beetles have previously been associated with shifts in carbon and nitrogen dynamics in other studies. Sixty days may not have been long enough to see detectable changes in the soil nutrient content; this could take years, or we may have missed the window when a pulse increase in nutrients happened in other similar experiments.
While the effects of dung beetles on soil carbon and nitrogen were not directly apparent, in this study, dung beetles may have indirect and longer-term effects on nutrient cycling mediated through impacts on the soil microbial community. Changes in some carbon pools can take longer to detect (e.g., several years for MAOM C), whereas other fluxes may have already occurred before our measurements (e.g., POM C). Their effects on soil microbial communities resemble those observed after the incorporation of other nutrient sources, highlighting the role of dung beetles in nutrient cycling and organic matter dynamics. This study gives incentive for land managers to find ways to increase dung beetle abundance on their land. While the specific dung beetles in this study were inoculated, other species of native and non-native dung beetles are present on these pastures and could potentially be managed. There may also be differences with other dung beetles that burrow to different depths, or dwell in the dung, with more or fewer differences in the soil microbial community and soil nutrients. It would be interesting to explore how differences among dung beetles that burrow to varying depths or dwell within the dung, in turn, result in different types of impacts on soil microbial communities and nutrient cycling. Future studies are needed to determine whether species identity or functional type alters these dynamics.
Our results also highlight the heterogeneity of soil microbial communities across different soils and climates. Although our results are specific to the ecosystems studied, the consistent effect of dung beetles on fungal communities across all core samples suggests that similar patterns may occur in temperate climates as well, despite climatic variation. Future studies using metabarcoding could test whether these same shifts are evident across a broader range of landscapes.
Author Contributions
The conception or design of the study: Suzanne Lipton, Rachel S. Meyer, Stacy M. Philpott. The acquisition, analysis, or interpretation of the data: Suzanne Lipton, Rachel S. Meyer, Stacy M. Philpott, Kate M. Scow, Ariel L. Simons. Writing of the manuscript: Suzanne Lipton. Editing and additional contributions to the manuscript: Suzanne Lipton, Rachel S. Meyer, Stacy M. Philpott, Kate M. Scow, Ariel L. Simons. Supervising: Rachel S. Meyer, Stacy M. Philpott, Kate M. Scow.
Acknowledgments
Special thank you to Paicines Ranch, TomKat Ranch, and Morris Grassfed Beef for assisting with the research implementation and monitoring of the enclosures, and to Radomir Schmidt, Sandipan Samaddar, and the Meyer, Philpott, and Scow Labs for feedback on methodology.
Funding
This work was supported by the Carbon Fund at the University of California, Santa Cruz, the Ruth and Alfred Heller Endowed Chair in Agroecology at the University of California, Santa Cruz, and by undergraduate participation in field and lab work, supported by the UCSC CAMINO (Center to Advance Mentored, Inquiry-Based Opportunities) program.
Ethics Statement
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
Data Availability Statement
Sequence results, published in NCBI under project number PRJNA1354185.
A'Bear, A. D., S. N. Johnson, and T. H. Jones. 2014. “Putting the ‘Upstairs–Downstairs’ Into Ecosystem Service: What Can Aboveground–Belowground Ecology Tell Us?” Biological Control, the Impact of Global Change on Biological Control 75: 97–107. https://doi.org/10.1016/j.biocontrol.2013.10.004.
Allouche, O., A. Tsoar, and R. Kadmon. 2006. “Assessing the Accuracy of Species Distribution Models: Prevalence, Kappa and the True Skill Statistic (TSS).” Journal of Applied Ecology 43: 1223–1232. https://doi.org/10.1111/j.1365‐2664.2006.01214.x.
Anderson, J. R., and E. C. Loomis. 1978. “Exotic Dung Beetles in Pasture and Range Land Ecosystems.” California Agriculture 32: 31–32.
Barbour, K. M., C. Weihe, S. D. Allison, and J. B. H. Martiny. 2022. “Bacterial Community Response to Environmental Change Varies With Depth in the Surface Soil.” Soil Biology and Biochemistry 172: 108761. https://doi.org/10.1016/j.soilbio.2022.108761.
Barragán, F., J. Labastida, and A. Ramírez‐Hernández. 2021. “Response of Dung Beetle Diversity to Three Livestock Management Systems in Drylands of Central Mexico.” Journal of Arid Environments 193: 104598. https://doi.org/10.1016/j.jaridenv.2021.104598.
Barragán, F., C. E. Moreno, F. Escobar, G. Halffter, and D. Navarrete. 2011. “Negative Impacts of Human Land Use on Dung Beetle Functional Diversity.” PLoS One 6, no. 3: e17976. https://doi.org/10.1371/journal.pone.0017976.
Benucci, G. M. N., R. Longley, P. Zhang, Q. Zhao, G. Bonito, and F. Yu. 2019. “Microbial Communities Associated With the Black Morel Morchella Sextelata Cultivated in Greenhouses.” PeerJ 7: e7744. https://doi.org/10.7717/peerj.7744.
Bossio, D. A., S. C. Cook‐Patton, P. W. Ellis, et al. 2020. “The Role of Soil Carbon in Natural Climate Solutions.” Nature Sustainability 3, no. 5: 391–398. https://doi.org/10.1038/s41893‐020‐0491‐z.
Brown, J., C. H. Scholtz, J.‐L. Janeau, S. Grellier, and P. Podwojewski. 2010. “Dung Beetles (Coleoptera: Scarabaeidae) Can Improve Soil Hydrological Properties.” Applied Soil Ecology 46: 9–16. https://doi.org/10.1016/j.apsoil.2010.05.010.
Caporaso, J. G., C. L. Lauber, W. A. Walters, et al. 2011. “Global Patterns of 16S rRNA Diversity at a Depth of Millions of Sequences Per Sample.” Proceedings of the National Academy of Sciences 108: 4516–4522. https://doi.org/10.1073/pnas.1000080107.
Clissmann, F., A. M. Fiore‐Donno, B. Hoppe, et al. 2015. “First Insight Into Dead Wood Protistan Diversity: A Molecular Sampling of Bright‐Spored Myxomycetes (Amoebozoa, Slime‐Moulds) in Decaying Beech Logs.” FEMS Microbiology Ecology 91: fiv050. https://doi.org/10.1093/femsec/fiv050.
Conant, R. T., C. E. P. Cerri, B. B. Osborne, and K. Paustian. 2017. “Grassland Management Impacts on Soil Carbon Stocks: A New Synthesis.” Ecological Applications 27: 662–668. https://doi.org/10.1002/eap.1473.
Cordero, I., A. Leizeaga, L. C. Hicks, J. Rousk, and R. D. Bardgett. 2023. “High Intensity Perturbations Induce an Abrupt Shift in Soil Microbial State.” ISME Journal 17: 2190–2199. https://doi.org/10.1038/s41396‐023‐01512‐y.
Cotrufo, M. F., M. G. Ranalli, M. L. Haddix, J. Six, and E. Lugato. 2019. “Soil Carbon Storage Informed by Particulate and Mineral‐Associated Organic Matter.” Nature Geoscience 12, no. 12: 989–994. https://doi.org/10.1038/s41561‐019‐0484‐6.
Cotrufo, M. F., M. D. Wallenstein, C. M. Boot, K. Denef, and E. Paul. 2013. “The Microbial Efficiency‐Matrix Stabilization (MEMS) Framework Integrates Plant Litter Decomposition With Soil Organic Matter Stabilization: Do Labile Plant Inputs Form Stable Soil Organic Matter?” Global Change Biology 19, no. 4: 988–995. https://doi.org/10.1111/gcb.12113.
Cui, Y., L. Fang, X. Guo, et al. 2018. “Responses of Soil Microbial Communities to Nutrient Limitation in the Desert‐Grassland Ecological Transition Zone.” Science of the Total Environment 642: 45–55. https://doi.org/10.1016/j.scitotenv.2018.06.033.
Curd, E. E., Z. Gold, G. S. Kandlikar, et al. 2019. “Anacapa Toolkit: An Environmental DNA Toolkit for Processing Multilocus Metabarcode Datasets.” Methods in Ecology and Evolution 10: 1469–1475. https://doi.org/10.1111/2041‐210X.13214.
da Silva, D. F., J. V. Moreira, L. I. S. de Sousa, et al. 2022. “Arbuscular Mycorrhizal Fungi Community in Soils Under Desertification and Restoration in the Brazilian Semiarid.” Microbiological Research 264: 127161. https://doi.org/10.1016/j.micres.2022.127161.
Dang, Q., Y. Wang, S. Xiong, et al. 2021. “Untangling the Response of Fungal Community Structure, Composition and Function in Soil Aggregate Fractions to Food Waste Compost Addition.” Science of the Total Environment 769: 145248. https://doi.org/10.1016/j.scitotenv.2021.145248.
Daniel, G. M., J. A. Noriega, P. G. da Silva, et al. 2022. “Soil Type, Vegetation Cover and Temperature Determinants of the Diversity and Structure of Dung Beetle Assemblages in a South African Open Woodland and Closed Canopy Mosaic.” Austral Ecology 47, no. 1: 79–91. https://doi.org/10.1111/aec.13138.
Davis, N. M., D. M. Proctor, S. P. Holmes, D. A. Relman, and B. J. Callahan. 2018. “Simple Statistical Identification and Removal of Contaminant Sequences in Marker‐Gene and Metagenomics Data.” Microbiome 6: 226. https://doi.org/10.1186/s40168‐018‐0605‐2.
Dove, N. C., M. E. Barnes, K. Moreland, R. C. Graham, A. A. Berhe, and S. C. Hart. 2021. “Depth Dependence of Climatic Controls on Soil Microbial Community Activity and Composition.” ISME Communications 1: 78. https://doi.org/10.1038/s43705‐021‐00081‐5.
Drenovsky, R. E., K. L. Steenwerth, L. E. Jackson, and K. M. Scow. 2010. “Land Use and Climatic Factors Structure Regional Patterns in Soil Microbial Communities.” Global Ecology and Biogeography 19: 27–39. https://doi.org/10.1111/j.1466‐8238.2009.00486.x.
Epp, L. S., S. Boessenkool, E. P. Bellemain, et al. 2012. “New Environmental Metabarcodes for Analysing Soil DNA: Potential for Studying Past and Present Ecosystems.” Molecular Ecology 21: 1821–1833. https://doi.org/10.1111/j.1365‐294X.2012.05537.x.
Erktan, A., D. Or, and S. Scheu. 2020. “The Physical Structure of Soil: Determinant and Consequence of Trophic Interactions.” Soil Biology and Biochemistry 148: 107876. https://doi.org/10.1016/j.soilbio.2020.107876.
Evans, K. S., M. Mamo, A. Wingeyer, et al. 2019. “Soil Fauna Accelerate Dung Pat Decomposition and Nutrient Cycling into Grassland Soil.” Rangeland Ecology & Management 72, no. 4: 667–677. https://doi.org/10.1016/j.rama.2019.01.008.
Fierer, N., and R. B. Jackson. 2006. “The Diversity and Biogeography of Soil Bacterial Communities.” PNAS 103: 626–631. https://doi.org/10.1073/pnas.0507535103.
Floate, K. D., D. W. Watson, R. M. Weiss, and O. Olfert. 2017. “Bioclimatic Analyses for the Distributions of Onthophagus Nuchicornis, Onthophagus Taurus, and Digitonthophagus Gazella (Coleoptera: Scarabaeidae) in North America.” Canadian Entomologist 149: 504–524. https://doi.org/10.4039/tce.2017.20.
Gerber, P. J., H. Steinfeld, B. Henderson, et al. 2013. Tackling Climate Change Through Livestock. Food and Agriculture Organization of the United Nations.
Graves, S., H.‐P. P. Dorai‐Raj, and Luciano Selzer with help from Sundar Dorai‐Raj. 2024. “multcompView: Visualizations of Paired Comparisons.”
Habtewold, J. Z., B. L. Helgason, S. F. Yanni, H. H. Janzen, B. H. Ellert, and E. G. Gregorich. 2020. “Litter Composition Has Stronger Influence on the Structure of Soil Fungal Than Bacterial Communities.” European Journal of Soil Biology 98: 103190. https://doi.org/10.1016/j.ejsobi.2020.103190.
Hadibarata, T., and A. Yuniarto. 2020. “Biodegradation of Polycyclic Aromatic Hydrocarbons by High‐Laccase Basidiomycetes Fungi Isolated From Tropical Forest of Borneo.” Biocatalysis and Agricultural Biotechnology 28: 101717. https://doi.org/10.1016/j.bcab.2020.101717.
Hammer, T. J., N. Fierer, B. Hardwick, et al. 2016. “Treating Cattle With Antibiotics Affects Greenhouse Gas Emissions, and Microbiota in Dung and Dung Beetles.” Proceedings of the Royal Society B: Biological Sciences 283: 20160150. https://doi.org/10.1098/rspb.2016.0150.
Hannula, S. E., A. M. Kielak, K. Steinauer, et al. 2019. “Time After Time: Temporal Variation in the Effects of Grass and Forb Species on Soil Bacterial and Fungal Communities.” MBio 10: e02635‐19. https://doi.org/10.1128/mBio.02635‐19.
Hanski, I., and Y. Cambefort. 1991. Dung Beetle Ecology. Princeton University Press.
Hasan, F., K. J. Wallace, S. V. Fowler, et al. 2024. “Dung Beetles Drive Direct and Indirect Changes in Ecosystem Multifunctionality.” Functional Ecology 38, no. 9: 1971–1983. https://doi.org/10.1111/1365‐2435.14607.
Hooper, D. U., D. E. Bignell, V. K. Brown, et al. 2000. “Interactions Between Aboveground and Belowground Biodiversity in Terrestrial Ecosystems: Patterns, Mechanisms, and Feedbacks: We Assess the Evidence for Correlation Between Aboveground and Belowground Diversity and Conclude That a Variety of Mechanisms Could Lead to Positive, Negative, or no Relationship—Depending on the Strength and Type of Interactions Among Species.” Bioscience 50: 1049–1061. https://doi.org/10.1641/0006‐3568(2000)050[1049:IBAABB]2.0.CO;2.
Huang, S. 2021. “Taxonomic Studies of Coronophorales and Niessliaceae (Hypocreomycetidae).” Mycosphere 12: 875–992. https://doi.org/10.5943/mycosphere/12/1/9.
Larsson, E., and L. Örstadius. 2008. “Fourteen Coprophilous Species of Psathyrella Identified in the Nordic Countries Using Morphology and Nuclear rDNA Sequence Data.” Mycological Research 112: 1165–1185. https://doi.org/10.1016/j.mycres.2008.04.003.
Lavallee, J. M., J. L. Soong, and M. F. Cotrufo. 2020. “Conceptualizing Soil Organic Matter Into Particulate and Mineral‐Associated Forms to Address Global Change in the 21st Century.” Global Change Biology 26: 261–273. https://doi.org/10.1111/gcb.14859.
Lenth, R. 2024. “Emmeans: Estimated Marginal Means, Aka Least‐Squares Means.” R Package Version 1.11.2. https://CRAN.R‐project.org/package=emmeans.
Liaw, A., and M. Wiener. 2002. “Classification and Regression by Random Forest.” R News 2: 18–22.
Lin, M., A. L. Simons, R. J. Harrigan, et al. 2021. “Landscape Analyses Using eDNA Metabarcoding and Earth Observation Predict Community Biodiversity in California.” Ecological Applications 31: e02379. https://doi.org/10.1002/eap.2379.
Lipton, S., R. S. Meyer, G. Richardson, and S. M. Philpott. 2023. “Onthophagus taurus Increases Soil Microbes Associated With Nutrient Cycling in California Pastureland Soils.” Rangeland Ecology & Management 91: 24–29. https://doi.org/10.1016/j.rama.2023.07.007.
Lipton, S., P. Skelley, and S. Philpott. 2025. “Landscape and Grazing Management Affect Dung Beetle Communities in California.” Basic and Applied Ecology 89: 120–127.
Lorenz, K., and R. Lal. 2018. “Carbon Sequestration in Grassland Soils.” In Carbon Sequestration in Agricultural Ecosystems, edited by K. Lorenz and R. Lal, 175–209. Springer International Publishing. https://doi.org/10.1007/978‐3‐319‐92318‐5_4.
Love, M. I., W. Huber, and S. Anders. 2014. “Moderated Estimation of Fold Change and Dispersion for RNA‐Seq Data With DESeq2.” Genome Biology 15: 550. https://doi.org/10.1186/s13059‐014‐0550‐8.
Macagno, A. L. M., A. P. Moczek, and A. Pizzo. 2016. “Rapid Divergence of Nesting Depth and Digging Appendages Among Tunneling Dung Beetle Populations and Species.” American Naturalist 187, no. 5: E143–E151. https://doi.org/10.1086/685776.
Maillard, É., and D. A. Angers. 2014. “Animal Manure Application and Soil Organic Carbon Stocks: A Meta‐Analysis.” Global Change Biology 20, no. 2: 666–679. https://doi.org/10.1111/gcb.12438.
Manning, P., E. M. Slade, S. A. Beynon, and O. T. Lewis. 2016. “Functionally Rich Dung Beetle Assemblages Are Required to Provide Multiple Ecosystem Services.” Agriculture, Ecosystems & Environment 218: 87–94. https://doi.org/10.1016/j.agee.2015.11.007.
Mayer, M., B. Rewald, B. Matthews, et al. 2021. “Soil Fertility Relates to Fungal‐Mediated Decomposition and Organic Matter Turnover in a Temperate Mountain Forest.” New Phytologist 231: 777–790. https://doi.org/10.1111/nph.17421.
McMurdie, P. J., and S. Holmes. 2013. “Phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data.” PLoS One 8: e61217. https://doi.org/10.1371/journal.pone.0061217.
McMurdie, P. J., and S. Holmes. 2014. “Waste Not, Want Not: Why Rarefying Microbiome Data Is Inadmissible.” PLoS Computational Biology 10, no. 4: e1003531. https://doi.org/10.1371/journal.pcbi.1003531.
Menéndez, R., P. Webb, and K. H. Orwin. 2016. “Complementarity of Dung Beetle Species With Different Functional Behaviours Influence Dung–Soil Carbon Cycling.” Soil Biology and Biochemistry 92: 142–148. https://doi.org/10.1016/j.soilbio.2015.10.004.
Nichols, E., S. Spector, J. Louzada, T. Larsen, S. Amezquita, and M. E. Favila. 2008. “Ecological Functions and Ecosystem Services Provided by Scarabaeinae Dung Beetles.” Biological Conservation 141: 1461–1474. https://doi.org/10.1016/j.biocon.2008.04.011.
Noriega, J. A., J. Hortal, I. deCastro‐Arrazola, et al. 2023. “Dung Removal Increases Under Higher Dung Beetle Functional Diversity Regardless of Grazing Intensification.” Nature Communications 14: 8070. https://doi.org/10.1038/s41467‐023‐43760‐8.
O'Geen, A., M. Walkinshaw, and D. Beaudette. 2017. “SoilWeb: A Multifaceted Interface to Soil Survey Information.” Soil Science Society of America Journal 81: 853–862. https://doi.org/10.2136/sssaj2016.11.0386n.
Oksanen, J., G. L. Simpson, F. G. Blanchet, et al. 2025. “vegan: Community Ecology Package (Version 2.8‐0) [R Package].” https://cran.r‐project.org/package=vegan.
Owen, W. K., J. E. Lloyd, D. E. Legg, and R. Kumar. 2006. “Endocoprid Activity of Aphodius Fossor (Coleoptera: Scarabaeidae) Related to Bovine Dung Decomposition in a Mixed Grass Prairie.” Journal of Economic Entomology 99, no. 6: 2210–2215. https://doi.org/10.1093/jee/99.6.2210.
Parker, E. S., I. L. G. Newton, and A. P. Moczek. 2020. “(My Microbiome) Would Walk 10,000 Miles: Maintenance and Turnover of Microbial Communities in Introduced Dung Beetles.” Microbial Ecology 80, no. 2: 435–446. https://doi.org/10.1007/s00248‐020‐01514‐9.
Pinheiro, J., D. Bates, and R Core Team. 2023. “Nlme: Linear and Nonlinear Mixed Effects Models R Package Version 3.1–164.” https://CRAN.R‐project.org/package=nlme.
Powell, J. R., J. L. Parrent, M. M. Hart, J. N. Klironomos, M. C. Rillig, and H. Maherali. 2009. “Phylogenetic Trait Conservatism and the Evolution of Functional Trade‐Offs in Arbuscular Mycorrhizal Fungi.” Proceedings of the Royal Society B: Biological Sciences 276, no. 1676: 4237–4245. https://doi.org/10.1098/rspb.2009.1015.
R Core Team. 2024. “Nlme: Linear and Nonlinear Mixed Effects Models.” (version), J.P. (S, to 2007), D.B. (up, to 2002), S.D. (up, to 2005), D.S. (up, authors (src/rs.f), E., sigma), S.H. (Author fixed, sigma), B.V.W. (Programmer fixed, Ranke (varConstProp())).
Schmidt, M. W. I., M. S. Torn, S. Abiven, et al. 2011. “Persistence of Soil Organic Matter as an Ecosystem Property.” Nature 478: 49–56. https://doi.org/10.1038/nature10386.
Schnecker, J., T. Böckle, J. Horak, V. Martin, T. Sandén, and H. Spiegel. 2024. “Improving Measurements of Microbial Growth, Death, and Turnover by Accounting for Extracellular DNA in Soils.” Soil 10, no. 2: 521–531. https://doi.org/10.5194/soil‐10‐521‐2024.
Schuman, G. E., H. H. Janzen, and J. E. Herrick. 2002. “Soil Carbon Dynamics and Potential Carbon Sequestration by Rangelands.” Environmental Pollution 116, no. 3: 391–396. https://doi.org/10.1016/S0269‐7491(01)00215‐9.
Signorell, A., K. Aho, A. Alfons, et al. 2024. DescTools: Tools for Descriptive Statistics. R Packag Version 099 28:17.
Slade, E. M., T. Riutta, T. Roslin, and H. L. Tuomisto. 2016. “The Role of Dung Beetles in Reducing Greenhouse Gas Emissions From Cattle Farming.” Scientific Reports 6: 18140. https://doi.org/10.1038/srep18140.
Slade, E. M., T. Roslin, M. Santalahti, and T. Bell. 2016. “Disentangling the ‘Brown World’ Faecal‐Detritus Interaction Web: Dung Beetle Effects on Soil Microbial Properties.” Oikos 125: 5–635. https://doi.org/10.1111/oik.02640.
Smith, P. 2004. “How Long Before a Change in Soil Organic Carbon Can Be Detected?” Global Change Biology 10, no. 11: 1878–1883. https://doi.org/10.1111/j.1365‐2486.2004.00854.x.
Stanbrook, R., E. Harris, M. Jones, and C. P. Wheater. 2021. “The Effect of Dung Beetle Size on Soil Nutrient Mobilization in an Afrotropical Forest.” Insects 12: 141. https://doi.org/10.3390/insects12020141.
Stanbrook, R., and J. R. King. 2022. “Dung Beetle Community Composition Affects Dung Turnover in Subtropical US Grasslands.” Ecology and Evolution 12, no. 2: e8660. https://doi.org/10.1002/ece3.8660.
Thuiller, W., M. Guéguen, J. Renaud, D. N. Karger, and N. E. Zimmermann. 2019. “Uncertainty in Ensembles of Global Biodiversity Scenarios.” Nature Communications 10: 1446. https://doi.org/10.1038/s41467‐019‐09519‐w.
Tosi, M., W. Deen, R. Drijber, M. McPherson, A. Stengel, and K. Dunfield. 2021. “Long‐Term N Inputs Shape Microbial Communities More Strongly Than Current‐Year Inputs in Soils Under 10‐Year Continuous Corn Cropping.” Soil Biology and Biochemistry 160: 108361. https://doi.org/10.1016/j.soilbio.2021.108361.
van Proosdij, A. S. J., M. S. M. Sosef, J. J. Wieringa, and N. Raes. 2016. “Minimum Required Number of Specimen Records to Develop Accurate Species Distribution Models.” Ecography 39, no. 6: 542–552. https://doi.org/10.1111/ecog.01509.
Waldrop, M. P., J. M. Holloway, D. B. Smith, et al. 2017. “The Interacting Roles of Climate, Soils, and Plant Production on Soil Microbial Communities at a Continental Scale.” Ecology 98: 1957–1967.
Wang, J., X. Wang, G. Liu, G. Wang, and C. Zhang. 2021. “Grazing‐To‐Fencing Conversion Affects Soil Microbial Composition, Functional Profiles by Altering Plant Functional Groups in a Tibetan Alpine Meadow.” Applied Soil Ecology 166: 104008. https://doi.org/10.1016/j.apsoil.2021.104008.
Weber, C., R. Vilgalys, and C. Kuske. 2013. “Changes in Fungal Community Composition in Response to Elevated Atmospheric CO2 and Nitrogen Fertilization Varies With Soil Horizon.” Frontiers in Microbiology 4: 1–14.
White, T., T. D. Bruns, S. Lee, and J. Taylor. 1990. “Amplification and Direct Sequencing of Fungal Ribosomal RNA Genes for Phylogenetics.” In PCR Protocols: A Guide to Methods and Applications, 315–322. Academic Press.
Whitehead, D., L. A. Schipper, J. Pronger, et al. 2018. “Management Practices to Reduce Losses or Increase Soil Carbon Stocks in Temperate Grazed Grasslands: New Zealand as a Case Study.” Agriculture, Ecosystems & Environment 265: 432–443. https://doi.org/10.1016/j.agee.2018.06.022.
Wickham, H. 2011. “ggplot2.” WIREs Computational Statistics 3: 180–185. https://doi.org/10.1002/wics.147.
Worrall, J. J., S. E. Anagnost, and R. A. Zabel. 1997. “Comparison of Wood Decay Among Diverse Lignicolous Fungi.” Mycologia 89: 199–219. https://doi.org/10.2307/3761073.
Yang, C., N. Liu, and Y. Zhang. 2019. “Soil Aggregates Regulate the Impact of Soil Bacterial and Fungal Communities on Soil Respiration.” Geoderma 337: 444–452. https://doi.org/10.1016/j.geoderma.2018.10.002.
Yokoyama, K., H. Kai, T. Koga, and T. Aibe. 1991. “Nitrogen Mineralization and Microbial Populations in Cow Dung, Dung Balls and Underlying Soil Affected by Paracoprid Dung Beetles.” Soil Biology and Biochemistry 23: 649–653. https://doi.org/10.1016/0038‐0717(91)90078‐X.
Zhang, Y., S. Sun, D. Luo, et al. 2023. “Decline in Morel Production Upon Continuous Cropping Is Related to Changes in Soil Mycobiome.” Journal of Fungi 9: 492. https://doi.org/10.3390/jof9040492.
Zhang, Z., X. Zhou, L. Tian, et al. 2017. “Fungal Communities in Ancient Peatlands Developed From Different Periods in the Sanjiang Plain, China.” PLoS One 12, no. 12: e0187575. https://doi.org/10.1371/journal.pone.0187575.
© 2025. This work is published under http://creativecommons.org/licenses/by-nc/4.0/ (the "License"). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.