Abstract A siliceous permineralized peat block recovered from Hopen in the Svalbard archipelago hosts a low-diversity Late Triassic flora dominated by autochthonous roots and stems of bennettitaleans and lycophytes, and parautochthonous leaves, sporangia, spores and pollen from a small range of pteridophytes and gymnosperms. Some parenchymatous bennettitalean root cells show interactions with chytrid fungi and bacteria; the remains of other fungi and fungi-like organisms are dispersed within the peat’s detrital matrix. Cavities excavated through some roots and compacted detritus contain abundant coprolites probably derived from sapro-xylophagous oribatid mites, although no body fossils have yet been identified. Sparse larger coprolites containing leaf fragments attest to the presence of invertebrate folivores in the ancient ecosystem. The low-diversity flora, relatively few trophic levels and simple nutritional web, together with sedimentological aspects of the host formation and the peat structure, collectively favour accumulation of the organic mass as a fibric (root-dominated) peat within a temperate (high middle-latitude), well-aerated mire.

Abstract Biological soil crusts, composed of soil surfaces stabilized by a consortium of cyanobacteria, algae, fungi, lichens, and/or bryophytes, are common in most deserts and perform functions of primary productivity, nitrogen fixation, nutrient cycling, water redistribution, and soil stabilization. The crusts are highly susceptible to disturbance. The degree of perturbation is governed, at least in part, by the nature, intensity, and spatial and temporal distribution of the disturbance, as well as the soil type and soil moisture content at the time of disturbance. When disturbed, biological soil crusts lose their capacity to perform their ecological functions. Natural recovery of disturbed crusts can range from several years to millennia. Several strategies have been attempted to accelerate recovery of crusts. At present, artificial recovery is not economically feasible on large tracts of disturbed desert landscape. Management options available to the military on arid landscapes include: (1) eliminating or minimizing training in desert ecosystems, (2) avoiding critical seasons, (3) avoiding critical areas, (4) artificially restoring damaged crusts, and (5) considering desert training lands as “sacrifice areas.” Given the need to train in environments representative of the locations of many current and projected world conflicts, the first option is untenable. At this time, the most plausible alternative is to consider desert training lands as “sacrifice areas.” However, it is recommended that attempts be made to avoid critical seasons and areas inasmuch as logistically feasible, and that the military continue to support research into the development of cost-effective technologies for biological soil crust restoration.