No abstract

We compared decomposition of surface and buried, untreated, mixed desert shrub litter to that of insecticide-and fungicide-treated litter. Suppression of fungi reduced decomposition by =29%; exclusion of microarthropods reduced decomposition by =53%. Approximately 55% of the organic mass of the untreated litter disappeared during the 6-mo growing season and 23-29% disappeared in the winter months (November through March).There was a consistent pattern of microarthropod colonization of buried litter that was related to the percent organic matter lost. This sequence was tydeid mites, tarsonemid and pyemotid mites, gamasina and predatory Prostigmata, Collembola and Psocoptera, and oribatids. After I yr, large numbers of enchytraeid worms were extracted from buried litter. Decomposition of insecticide-treated litter varied directly with rainfall and soil temperature while abiotic factors accounted for <50% of the variation in decomposition of untreated buried litter. We hypothesize that microarthropods affect litter decomposition in desert ecosystems by inoculating litter with fungal spores, by grazing on fungi, and in a heretofore undescribed mode, by preying on free-living nematodes.

Carbon and nitrogen dynamics were analyzed during the decomposition of litter and roots of the desert ephemeral pepperweed (Lepidium lasiocarpum). We treated litter bags with the insecticide chlordane and the fungicides benomyl and captan to eliminate or restrict groups of soil biota.The mass losses of buried litter (51, 39, and 25% for untreated, insecticide-treated, and fungicideinsecticide-treated material, respectively) were higher than those of the respective root treatments (35, 18, and 15%) at 96 d. The mass loss of untreated material was correlated with numbers of detritivorousfungivorous microarthropods, and only a small percentage of this loss was as C0 2 : 27 and 42% for litter and roots, respectively. In the absence of microarthropods a higher percentage of mass-loss carbon could be accounted for as C0 2 : 33 and 76% for litter and roots, respectively, indicating that mass loss was due primarily to litter removal by microarthropod activity and not to mineralization. Litter removal by microarthropods was less dependent on abiotic constraints such as soil moisture (r = 0.65, P < .001) than was mass loss when microarthropods were absent (r = 0.79, P < .001). In the absence of microarthropods, mass loss was more closely coupled with biomass of grazers, such as nematodes, which require free water for activity (r = 0.99, P < .0001).Unlike mass loss, carbon mineralization was highest in untreated roots, suggesting a stimulation of microbial activity by microarthropods, while in untreated litter no stimulation was observed when compared to insecticide treatments. This difference was primarily a function of fungivorous microarthropod density, with overgrazing occurring in the untreated litter.Nitrogen budgets indicated the importance of microarthropods in the turnover of root nitrogen. In the presence ofmicroarthropods 132% of the initial root nitrogen could be accounted for after 96 d, while in the absence ofmicroarthropods 270% could be accounted for. This net immobilization of nitrogen was primarily in the soil organic fraction around the roots and was associated with fungal development.Data from this study re-emphasize the importance of microarthropods as regulators of decomposition in deserts and suggest that predation by nematodes or protozoa on bacteria and fungi contributes to rate regulation. Nitrogen flux data suggest that when spring ephemeral plant production is high, decomposition of ephemeral roots with attendant nitrogen immobilization can reduce the nitrogen available to creosotebush, Larrea tridentata, thus reducing shrub production. Higher taxa of soil biota, i.e., nematodes and microarthropods, may thus be important regulators of nitrogen fluxes and of mass loss in decomposition.

We studied changes in populations of mites, nematodes, bacteria, and fungi in buried creosote bush litter treated with selected inhibitors. Elimination of microarthropods (primarily tydeid mites) resulted in increased numbers of bacteriophagic nematodes and reduction in numbers of bacteria; elimination of both nematodes and microarthropods resulted in increased numbers of bacteria compared to untreated controls. Fungal grazing mites, Pyemotidae, and fungivorous nematodes, Aphelenchus sp., increased in numbers between days 25 and 30, reducing the fungi on untreated leaves but not on stems and petioles, while mean length of fungal hyphae increased in insecticidetreated leaves. Elimination of mites resulted in a 40% reduction in decomposition suggesting that in a desert, tydeid mites affect decomposition of buried litter by regulating the population size of the bacterial grazers, cephalobid nematodes.

We conducted studies of mass losses from surface and buried litter bags in four North American hot desert areas to test the following hypotheses: (1) leaf litter disappearance in hot deserts is independent of actual evapotranspiration, (2) buried litter disappearance is a function of actual evapotransporation, (3) the pattern of microarthropod colonization of buried leaf litter is a function of the stage of decomposition, and (4) elimination of microarthropods results in reduced rates of decomposition and increased numbers of free—living nematodes. Mass losses from surface Larrea tridentata leaf litter bags ranked highest to lowest: Chihuahuan desert, Sonoran desert, Mojave desert, Coloradan desert. Mass losses from buried litter bags were essentially equal. °40%, in each of the deserts for bags buried from March to October. There was low correlation between rainfall and mass loss of buried litter and surface litter in the North American hot deserts. Mass losses from insecticide—treated buried bags were lower than from untreated bags. There was a greater abundance of nematodes in insecticide—treated bags than in untreated bags. Tarsonemid mites were found only in litter bags from the Chihuahuan desert. The most abundant microarthropods in buried leaf litter in the other deserts were predatory raphignathids, tydeids, and arctacarids. Decomposition (litter disappearance) in North American hot deserts was highly correlated with long—term rainfall patterns, which we hypothesize have served as the selective agents for the soil biota active in the decomposition process. Thus litter disappearance does not respond to annual fluctuations in rainfall amounts.