Deadwood can impact forest food webs through the creation of habitat and the provision of resources, but impacts may differ based on initial differences in wood characteristics and over time. Bark beetles like the southern pine beetle (Dendroctonus frontalis Zimmerman) attack and wound pine trees, inoculating them with Ophiostomataceae (Ascomycota; hereafter “blue stain fungi”). Blue stain fungi may increase termite presence in deadwood and in the surrounding leaf litter, potentially leading to increased abundances of leaf litter invertebrates over time. The effects of deadwood in general, and simulated bark beetle–generated deadwood (i.e., deadwood inoculated with blue stain fungi) or non–bark beetle–generated deadwood (i.e., deadwood inoculated with just H₂O controls) were tested on leaf litter communities after one and seven years to measure both short– and long–term effects. The presence of deadwood led to distinct leaf litter communities compared to no wood across both collection years. However, there was no difference in community composition under logs between deadwood treatments. However, predatory beetles and non-ant Hymenoptera were indicators of bluestained wood. Taxa abundance differed by wood treatment, but richness and detritivore/predator ratio were greater under deadwood versus no deadwood, particularly after seven years. These results contribute to the mounting evidence that deadwood has important impacts on forest biodiversity and that long–term studies are necessary to fully understand deadwood impacts on forest ecosystems.

To examine impacts of deadwood presence, simulated bark beetle–generated deadwood (bluestained wood), and collection year on leaf litter invertebrate community structure, all leaf litter down to top soil was collected from 0.5 m x 0.5 m (0.25 m²) quadrats below one) uncaged bluestained logs (n=10; one per plot), two) uncaged control logs (n=10; one per plot), and three) equisized quadrats of leaf litter where no logs were present (n=10; one per plot) separated by ≥ 3 m from experimental logs on the same plot in October of years one (2015) and seven (2021) (Fig. 1). We randomly selected logs to sample leaf litter invertebrates from under each year, and quadrats sampled for litter invertebrates sampled where no logs were present were selected haphazardly to ensure random collection with no deadwood in quadrats and sufficient distance from existing experimental deadwood. Thus, in total as part of our study, leaf litter invertebrates were collected from under ten random (one per plot) uncaged bluestained logs, ten random uncaged water control logs (one per plot), and ten random plots without deadwood (n=30 total litter invertebrate samples; three per plot) in year one and again in year seven. Logs were removed during sampling as part of the larger experiment and were thus not repeatedly sampled across years. Before collection of leaf litter communities, litter depth was measured in each quadrat corner by inserting a thin metal flag through the leaf litter and measuring the distance from the top of the leaf litter to the top of the mineral soil. Then, all leaf litter in quadrats was collected down to the top of the mineral soil and placed in a hand sifter (1.5 cm x 1.5 cm mesh) to remove larger debris but retain leaf litter invertebrates.

Leaf litter, debris, and invertebrates that passed through mesh (siftate) were then placed in muslin bags and transported to the laboratory for invertebrate extraction using Berlese funnels with a 40 W halogen light bulb for 24 hours. Due to logistical time constraints, in year seven collections, total volume of the siftate from each quadrat was measured, but a random 250ml subsample of each quadrat’s siftate was used in Berlese funnels, which ran for 24 hours. Invertebrates were collected and preserved in a container with 70% ethanol until identification. To estimate the total number of invertebrates collected in year seven to compare with year one, the abundance of each focal taxon group was calculated using A𝑇 = 𝑇𝑆𝑉 ∗ (𝐴𝑀/250 𝑚𝑙). Here, AT is the estimated total abundance, and AM is the measured taxon abundance from the 250 mL subsample, and TSV is the total siftate volume (mL) collected from the field. Logs were destructively sampled each year as part of the larger experiment; consequently, leaf litter invertebrate communities were sampled from under different logs in year one than year seven (e.g., samples are not repeated measures). Abundances of Acari and Collembola were estimated, because they numbered in the hundreds to thousands in these samples. Specifically, we extracted invertebrate samples from Berlese funnels and placed them placed in 148.2 mm diameter gridded (2 cm x 2 cm grid squares) Petri dishes in 70% ethanol. First, all non–Acari and Collembola were removed from the Petri dish and identified, then, the remaining taxa across the Petri dish area were homogenized by gently swirling the dish and using forceps to evenly spread taxa. After which, the number of non–predatory mites, predatory mites, and Collembola in each of the ten randomly chosen grid squares were counted. Abundance of each taxon was estimated by calculating the average of each taxon among the ten squares and multiplying the average by the total area of the Petri dish (151.9 cm²).