Data set ID:
*We designed a study to examine the long-term spatial and temporal patterns of mass loss, changes in nitrogen content and chemical composition during decomposition of roots in a desert watershed. Because nitrogen availability has been hypothesized to affect decomposition rates of buried litter in deserts, we studied decomposition of roots along a transect on a desert watershed that had been fertilized with ammonium nitrate and compared these results with the decomposition of roots along a transect that had not been fertilized.
*Roots were collected from areas on the watershed adjacent to the study transects. The roots were gently washed to remove adhering soil particles, then dried to a constant mass at 50 C. Roots were collected from species that comprise more than 10% of the vegetative cover. Between one and three root segments of the woody sub-shrub, Gutierrezia sarothrae, and the most common summer herbaceous annual, Baileya multiradiata, were tethered with fine wire attached to an aluminum tag labeled with the identity code and intial mass. Approximately 1 g of roots fo each species: black grama grass (Bouteloua eriopoda), fluff grass (Erioneuron pulchellum) and the grass from the dry lake basin (Panicum obtusum), were placed in each 12 cm X 12 cm fiberglass mesh bag (mesh size=1mm). We made 960 bundles of G. sarothrae roots and 960 bags of E. pulchellum roots. There were 160 bags or root bundles of the other species. Each bundle or bag was attached to a unique number tag connected to a 30 cm wire. The study was designed for the retrieval fo 8 units of each species from each watershed location and each transect on ten sample dates. Roots were buried in short trenches (3 m long) separated by approximately 3 m between trenches. Each bag or root bundle was separated from others in a trench by a minimum of 50 cm. The trenches were located at the appoximate mid-section of the fertilized transect of the basin slope and 80 bundles were buried in the mid-section of the un-fertilized transect of the basin slope. The roots of P. obtusum were placed in the dry lake basin, the roots of B. eriopoda were placed in the black grama grasss zone at teh top of the watershed, and the roots of B. multiradiata were placed in teh basin slope. Roots of G. sarothrae and E. pulchellum were placed in all vegetation zones because these species occur in all the zones on the watershed. The other species are restricted to one zone, and although B. multiradiata occurs in four zones, its highest abundance is in the basin slope zone. At all sites, roots and root bags were buried at 5-10 cm soil depth. Eight samples of each species were retrieved after 1, 3, 6, 12, 18, 24, 30, and 36 months in the field. The sequence of removal was determined by drawing numbers from the remaining bags at random. Although this study was designed for 48 months duration, parts of teh study had to be terminated early. At 42 months the trenches were completely and carefully excavated and the remaining material recovered. We were unable to retrieve complete samples of roots except for G. sarothrae. The small number of root samples that were collected in month 42 were pooled by site and treatment in order to obtain sufficient material for the chemical analyses. After retrieval, roots were removed from the bags, gently washed in distilled water to remove adhering soil particles and dried to a constant mass at 50C. After the final dry mass measurement, the roots were ground in a laboratory mill equipped with a 1 mm screen. An exponential decay model was used to estimate the mass loss rate: X/Xo = ln -kt where X = mass remaining at time t (months), Xo = original weight, ln is the natural logarithm and k is the decomposition constant (Olson 1963). For the analysis of total nitrogen, sub-samples of ground roots were treated by Kjeldahl digestion in a block digestor. Ammonium nitrogen was analyzed by an automated salicylate procedure in a Scientific Instruments autoanalyzer (Fisher et al. 1987). Three organic matter fractions (water and non-polar solubles, acid solubles, and acid insolubles) were analyzed in the remaining root material. Non polar and water soluble substances were extracted using dichloromethane followed by hot (100 C) water (McGlaugherty et al. 1984). Acid soluble fiber, which was largely holocellulose, was separted from acid-soluble fiber (largely lignin and suberin) using a two-stage digestion in sulfuric acid (Effland 1977).
In the spring of 1982, as part of the establishment of the Jornada Long-Term Ecological Research site in southern New Mexico, a 135 ha portion of a 1500 ha, internally drained, watershed was exclosed from grazing by domestic livestock. Prior to exclosure the watershed, as well as the rest of the Jornada basin, had been moderately to heavily grazed for the past 100 years. Concurrent with grazing, the vegetation had undergone a dramatic change from desert grassland, with an almost continuous cover of C4 perennial grasses, to isolated patches of the original grassland in a mosaic with desert shrub dominated plant communities (Buffington and Herbel, 1965). The exclosure lies along a northeast facing piedmont slope at the base of a steep isolated mountain peak, and covers a variety of component landforms from the foot of the mountain to the basin floor. The northeast side of the exclosure is immediately upslope of the College Playa located near the NMSU College Ranch. Three parallel transects (2.7 km in length) run from the middle of the College Playa up into the foot of Mt. Summerford. The Control transect is to the west, the Treatment transect on the east side of the Control transect, and the Alternate Control to the east of the Treatment transect. Each transect is 30 meters wide with a 45 meter buffer zone between each transect. The Treatment transect was treated annually until 1987 with NHNO3 in a concentration equal to 10g N/m2. The station markers at 30 meter intervals along each transect.
Once or twice a year