Greenfall production by animals in the Chihuahuan Desert influences decomposition by altering litter quality and distribution

Most leaf inputs into aboveground litter pools occur after plant resorption of leaf nutrients during senescence, thus reducing the concentration of decomposition-limiting nutrients like nitrogen (N) in leaf litter.  In contrast, greenfall is green leaf litter that enters the litter pool before normal leaf abscission, often as a result of herbivory. In the Chihuahuan Desert, jackrabbits (Lepus californicus), kangaroo rats (Dipodomys spp.), and woodrats (Neotoma spp.) create substantial greenfall from creosotebush (Larrea tridentata), dropseed grasses (Sporobolus spp.), and soaptree yucca (Yucca elata), respectively. Differences in nutrient concentrations of greenfall and senesced leaves may affect the rate of litter decomposition.  Animal relocation of greenfall within the patchy shrub-grass-bare ground vegetation matrix typical of arid ecosystems may also affect decomposition rate. We hypothesized that (1) greenfall would decompose more rapidly than senesced litter because greenfall has higher levels of decomposition-limiting nutrients like N and (2) all litter would decompose more rapidly in open intercanopy spaces than beneath shrub canopies due to greater UV radiation and soil-litter mixing in open areas.

To test these hypotheses, we conducted a 12-month litter decomposition experiment in the northern Chihuahuan Desert. We placed separate litterbags of green and senesced litter of L. tridentata, S. flexuosus and Y. elata in plots beneath shrub canopies and in open, intercanopy areas. Litterbags were collected 0, 1, 3, 6, and 12 months after initial placement; litter was weighed to determine mass loss and analyzed for carbon (C) and N content. Differences in the decomposition rates of green and senesced litter were inconsistent between species. After 12 months, mass loss was greater for greenfall than senesced litter of S. flexuosus (P < 0.0001), but greater for senesced litter than greenfall of Y. elata (P < 0.0001). Greenfall and senesced litter of L. tridentata did not differ in mass loss. Litter in open plots had greater mass loss than litter in shrub plots for all three species (P < 0.0001). Our results support that litter mass loss rates (1) differ between greenfall and senesced litter, depending on the species, and (2) are greater in open intercanopy areas than beneath shrub canopies. These findings suggest that by generating greenfall, animals influence decomposition rates by altering both the quality and spatial distribution of organic material in the aboveground litter pool.