تخریب پوسته های بیولوژیکی خاک: اثرات بر روی فرسایش پذیری بادی خاک های شنی بیابانی در جنوب غرب یوتا Disturbance Of Biological Soil Crusts: Impacts On Potential Wind Erodibility Of Sandy Desert Soils In Southeastern Utah

INTRODUCTION Biological soil crusts, also called `cryptobiotic’, `microbiotic’, `microphytic’ or `cyanobacterial±lichen’ soil crusts, are a dominant feature of most semiarid and arid landscapes throughout the world. These crusts di€er in species composition and occur on a variety of soils. As a result, crustal function in di€erent geographic regions might vary in regard to ecological processes such as rainfall in®ltration and seedling establishment (Harper and Marble, 1989; Johansen, 1993; West, 1990). However, most studies agree that biological soil crusts reduce wind erodibility of soil surfaces (Leys, 1990; MacKenzie and Pearson, 1979; Williams, et al., 1995), although one study found no signi®cant di€erences (Andrew and Lange, 1986). Scanning electron microscope studies done by Belnap and Gardner (1993) show that the extracellular sheath material of cyanobacteria bind soil particles together, providing soil surface protection. Biological soil crusts are highly susceptible to disturbance, especially in soils with low aggregate stability such as sands (Belnap and Gardner, 1993; Gillette, et al., 1980; Webb and Wilshire, 1983). Cyanobacterial ®laments, lichens and mosses are brittle when dry, and crush easily when subjected to compressional or shear forces by activities such as trampling or vehicular trac. Because crustal organisms are only metabolically active when wet, re-establishment time is slow in arid systems. While cyanobacteria are mobile, and can often move up through disturbed sediments to reach light levels needed for photosynthesis, lichens and mosses are incapable of such movement and often die as a result. On newly disturbed surfaces, mosses and lichens often have extremely slow colonization and growth rates. Assuming adjoining soils are stable and rainfall is average, recovery rates for lichen cover in southern Utah has been most recently estimated at a minimum of 45 years, while recovery of moss cover was estimated at 250 years (Belnap, 1993). Due to this slow recolonization of soil surfaces by the di€erent crustal components, crusts can be found in many stages of development. Wind is a major erosive force in deserts where there is little organic matter or vegetation cover to protect soil surfaces. Soil deposition by wind often exceeds that of ¯uvial deposition in these drier regions (Goudie, 1978; Williams, et al., 1995). Sediment production from soil surfaces depends on the force of wind needed to detach particles from soil surfaces (threshold friction velocity). Since wind erosion is of major concern both in the western USA and worldwide (Dregne, 1983), it is important to understand how soil surface disturbance a€ects threshold velocities. While previous studies have addressed the role soil crusts play in stabilizing desert soil surfaces, none has examined how threshold velocities might vary between stages of crustal development or how disturbance might di€erentially in¯uence various crustal types. The purpose of this study was to determine typical threshold velocities for di€erent stages of biological soil crust development and to determine the e€ects of di€erent soil surface disturbances on various stages of crustal development. METHODS The study site was located approximately 16 km south of Moab, Utah, USA, in Rizzo sandy loam soils. The dominant vegetation type is pinyon and juniper at an elevation of 1400 m. Annual precipitation is 250 mm, with 30 per cent of the rainfall occurring as late summer monsoons. Treatments were applied and measurements taken in July 1995 when soils were dry. All areas tested were located within a 300 m circle, with the same substrate type, soil depth and slope. Soils were collected and analyzed for sand, silt and clay content. Biological soil crust development was placed in one of four time categories, based on previous experiments regarding recovery rates after disturbance from four-wheel vehicles or foot trac (Belnap, 1993, unpublished data). These included: (a) Class 0: bare sand, with no visible biological crustal development, indicating very recent disturbance from vehicle or foot trac. (b) Class 1: ¯at crusts, with no visible frost heaving or lichen cover and low cyanobacterial biomass, indicating disturbance from vehicles or foot trac within one year of observation. (c) Class 2: moderately bumpy biological crusts with no lichen or moss development and moderate cyanobacterial biomass levels, indicating vehicular or foot trac disturbances 5±۱۰ years prior to observation. (d) Class 3: biological crusts were very bumpy, with full lichen and moss development and high cyanobacterial biomass, indicating no vehicular or foot trac disturbance for at least 20 years. Friction threshold velocities for movement of loose sand particles on the undisturbed surface (CON in Figure 2), and surface integrity of the crusts (SI in Figure 3) were determined for each crust type at two replicated sites. The FTV for particle movement was de®ned as the friction velocity at which surface particles were both detached from the soil surface and carried away by the generated wind. The FTV for surface integrity was the friction velocity at which large, intact chunks of the surface were detached and blown away. Because wind stress equals the square of friction velocity times the density of air, relative resistances of the di€erent crustal classes to wind erosion are de®ned and reported as the square of the ratio of threshold friction velocities between the classes being compared. Once FTVs were determined for the di€erent undisturbed crustal classes, disturbance treatments were applied to each crust class at each site. These treatments included: (1) Treatment F1: one pass by walking on crusts with lug-soled boots. (2) Treatment V1: one pass of a four-wheel drive vehicle with knobbed tires. (3) Treatment V2: two passes of a four-wheel drive vehicle with knobbed tires. Comparisons across the three crustal classes were done using a two-way ANOVA and multiple range test. T-tests were used to distinguish between disturbance treatments and controls.