A computational study of episodic events and historical context in long-term ecological process: climate and grazing in the northern Chihuahuan Desert

TitleA computational study of episodic events and historical context in long-term ecological process: climate and grazing in the northern Chihuahuan Desert
Publication TypeJournal Article
Year of Publication1992
AuthorsConley W.H., Conley MR, Karl TR
JournalCoenoses
Volume7
Pagination55-60
Date Published1992
Accession NumberJRN00147
Call Number00430
Keywordsarticle, articles, cattle grazing, climate, temperature, climate, trends, desertification,climate, grassland,climate, journal, journals, livestock,also SEE <GRAZING,CATTLE GRAZING>, model,monte-carlo simulations, precipitation,effects on vegetation, shrub, increase of
AbstractWe examined recent climate data from the Mesilla Valley in southern New Mexico, USA. Monthly data for total precipitation, maximum and minimum temperatures, and mean annual temperatures were analyzed from the Jornada Experimental Range Headquarters station (July 1914-Dec 1984) and NMSU station (Jan 1892-Dec 1984). We examined data at monthly, seasonal, and annual intervals. Seasons were developed that corresponded to vegetation and animal phenologies from the study area. Our intent was to examine the hypothesis that changes in climate patterns over the past 100 years were the causal agent in the observed shift in vegetation of the Jornada Basin from perennial grasslands to shrublands. An alternate hypothesis was that this vegetation shift was caused by the introduction and effect of grazing by domestic livestock. We conducted statistical analyses that included descriptive and exploratory techniques, linear and quadratic regression, de-trended autocorrelation analyses, and Monte-Carlo simulation studies. There were no statistically significant trends in precipitation in 11 of 12 months, with only May showing a significant quadratic fit. There are no statistically significant linear or quadratic trends in seasonal or annual precipitation totals from either the Jornada headquarters or the NMSU stations. Significant autocorrelation terms were obtained for lag times of 2, 4, 16, and 17 years. Only lag 17 showed agreement between both the Jornada headquarters and the NMSU stations. Spatial variability was demonstrated between the Jornada and the NMSU stations for both precipitation and temperature. This is particularly so for the summer season, where the Jornada precipitation is quite different from that at the NMSU station. Simulation studies allowed the examination of probabilities of occurrence of segments of data that constituted climatic events in the sense of being of interest biologically. These analyses allowed us to examine such events in the context of the recent climate history of the valley. We were able to demonstrate any consistent statistical rationale for invoking climate as a causal mechanism in the observed vegetative change of the valley. An examination of the history of grazing by introduced livestock leads to acceptance of an alternate hypothesis that this vegetation shift was caused by grazing effects from introduced livestock. These analyses also demonstrate that the meteorological and environmental "context" within which ecological studies are conducted can have significant influence on the results obtained.