A pervasive limiting resource in aridland ecosystems is water. In central New Mexico, precipitation inputs vary seasonally, annually and on decadal time scales. In the southwestern US, the amount and timing of seasonal and annual precipitation are influenced by two major climate cycles, the El Niño Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO). ENSO regulates variability in winter precipitation with high precipitation occurring during El Niño periods, and low precipitation during La Niña periods. ENSO events typically occur every 3-4 years and usually last only through one winter season. More recently it has been suggested that a longer-term climatic event, the Pacific Decadal Oscillation, may have profound effects on regional climate in the southwestern United States (Gutzler et al. 2002). The PDO, which oscillates on approximately 50-year cycles, modulates ENSO events and it may be the cause of periodic, extended, severe droughts in the region (Milne et al. in press).
Available soil moisture is not only a function of precipitation inputs but also temperature. Mean annual temperature from 1989-2002 at the Deep Well meteorological station on McKenzie Flats, a grassland site on the Sevilleta, is 13.2° C, with a low of 1.6°C in January and a high of 25.1°C in July. In addition, this site receives approximately 250 mm of precipitation annually, about 60% of which falls during the summer monsoon season from June through September and the remainder primarily from winter frontal systems. However, the relative contribution of summer monsoon and winter rains varies from year to year, creating a highly variable seasonal pattern of water inputs from year to year. A climate diagram based on data from Socorro, NM, south of the Sevilleta shows that on average, the lower elevations in this region are in moisture deficit most of the year, with potential surpluses only during August, December, and January.
In the piñon-juniper woodlands in the upper elevations of the Los Piños Mountains east of the Deep Well site, annual precipitation is about 365 mm and average annual temperature is 12.7°C with a low of 2.5°C in January and a high of 23.0°C in July. At this upper elevation site, there is a longer annual period of water surplus, on average, than in the lower elevation grasslands at the base of the mountains.
I. Monitoring studies. Our monitoring studies quantify ecosystem responses to natural pulse events at large spatial and multiple temporal scales (single events to decades).
A. Meteorological monitoring. Climate is the key driver of ecosystem processes in aridland environments at daily, seasonal, interannual and decadal time scales (North American Monsoon, ENSO, the Pacific Decadal Oscillation) (Gutzler et al. 2002, 2010, Milne et al. 2003). We maintain 10 comprehensive meteorological stations plus three partial stations and nine Hobo rain gauges across the SNWR. Comprehensive met stations measure precipitation, air and soil temperature, wind speed and direction, solar radiation, relative humidity and soil moisture. Meteorological and soil measurements are saved hourly and reported daily on the SEV website and ClimbDB. These stations, combined with the NOAA CRN site, the NRCS SCAN station, and met stations at some experiments allow us to gather spatially extensive meteorological data in all major habitats at SEV.
B. Riparian/river dynamics. Terrestrial and aquatic components of the landscape are periodically coupled through trigger-transfer events (Ludwig et al. 2005). The goal of our river and riparian measurements is to understand how pulse events (fires, floods) move materials from terrestrial to aquatic components of the landscape affecting water quantity and quality in the Middle Rio Grande. Continuing activities include a sonde network that measures temperature, pH, conductivity, turbidity, and dissolved oxygen at 48 locations from the Valles Caldera through the Jemez River to the Rio Grande south of SEV. Currently we are documenting the effects of the Las Conchas Fire that burned from June-August 2011 in the Jemez Mountains north of SNWR. Once the monsoon started, pulses of nutrients and organic matter flowed into the Rio Grande and its tributaries. The sonde network is a partnership between SEV, Office of Experimental Program to Stimulate Competitive Research (EPSCoR) and the United States Army Core of Engineers (USACE). A second major focus linked to EPSCoR is the deployment and testing of continuous measurements of nitrate and phosphate in river water along with sonde data.
C. Sevilleta Ecological Observatory Network (SEON). The varied topography and large elevation gradients that characterize central New Mexico create a wide range of climatic conditions - and associated biomes - within relatively short distances, providing an ideal study system to determine the effects of climate variability on ecosystem dynamics. Such studies are critical given that our region has already experienced altered precipitation patterns (Mote et al. 2005, Petrie et al. in prep), and will likely experience a warmer, drier and variable climate in the future (Seager et al. 2007; Ting et al. 2007, Gutzler et al. 2010). To predict regional changes in C storage, hydrologic partitioning and water resources in response to climate change, it is critical to understand how both temperature and soil moisture affect total C uptake (GPP), ecosystem respiration (Reco), evaportranspiration (ET), and net ecosystem exchange (NEE) of C, water and energy across a range of biomes. SEV maintains a network of 11 eddy covariance towers (SEON) which includes 9 towers in the New Mexico Elevation Gradient: 2 grassland sites, creosote shrubland, grass-shrub transition, juniper savanna, 2 piñon-juniper woodlands sites, ponderosa pine, mixed conifer ecosystems (Kurc & Small 2004, 2005, Anderson-Teixeira et al. 2011) and 2 riparian sites (Cleverly et al. 2006). The SEON network measures the exchange of C, water and energy between each ecosystem and the atmosphere and associated meteorological instruments in each biome. The primary goal of SEON is to use the elevation gradient in central New Mexico – ranging from hot and arid at low elevations to mesic riparian or cool, high elevation forests – to determine how processes in these climate-controlled ecosystems will respond to climate change. In particular, we are focusing on the interactive effects of temperature and soil moisture pulses on ecosystem processes, as temperature decreases and soil moisture increases along the gradient, and through time within sites.
II. Manipulative experiments. We experimentally alter precipitation pulses, the drivers of our Threshold-Delay-Nutrient Dynamics (TDND) model, to gain a mechanistic understanding of the consequences of increased climatic variability and extreme events on ecosystem processes and biotic responses.
A. Monsoon rainfall manipulation experiment (MRME). Global change models predict more variable precipitation regimes in the future (IPCC 2007, Diffenbaugh et al. 2009). We established MRME to experimentally manipulate the amount and timing of monsoon precipitation events. The primary goal of MRME is to determine how changes in the temporal pattern of summer precipitation affect plant and microbial community structure and interactions, and ecosystem processes in desert grassland. MRME contains five replicate plots (9mx14m) of the following treatments imposed from July through September: 1) ambient plus a weekly addition of 5 mm rainfall, and 2) ambient plus a monthly addition of 20 mm rainfall. At the end of the monsoon season, each experimental treatment has received the same total amount of precipitation, but in events that differ in size and frequency. In addition, three plots receive only ambient precipitation. Plots are instrumented with soil temperature, moisture and CO2 probes, minirhizotrons, and root donuts (belowground NPP). Subplots in each plot receive 5gN m-2y-1 as NH4NO3 prior to the monsoon season. Plant species composition and NPP are measured twice per year. These treatments create distinctly different patterns of rainfall that mimic future scenarios for climate change in our region (Thomey et al. 2011, Vargas et al. in review). Results thus far support predictions that ecological processes in arid ecosystems will exhibit positive responses to increased climate variability (Knapp et al. 2008). In addition, in 2011 soils were collected from all replicates for metagenomic analysis to determine how precipitation variability alters microbial community composition, structure and functioning.
B. Warming-El Nino-Nitrogen Deposition Experiment (WENNDEx). Climate change coupled with increased resource availability alters plant growth and potentially drives long-term shifts in community composition and ecosystem processes (LeBauer & Treseder 2008, Smith et al. 2009, Allen et al. 2010). The primary goal of WENNDEx is to determine how nighttime warming, winter precipitation pulses, and N deposition affect plant and soil microbial communities, Rs, and NPP in desert grassland. WENNDEx is a multi-factorial, fully crossed (n=5) field experiment. Treatments include increased nighttime temperatures (1.5-2.0C during winter), N deposition (2gN m-2y-1 as NH4NO3) and winter precipitation (+50% of long term average). Plots (2mx3m) are instrumented with soil temperature, moisture, and CO2 sensors, and air temperature probes. Treatments mimic conditions expected to occur in the next 50 years. Although warming should favor the desert grass, B. eriopoda, N deposition and increased winter precipitation may favor the northern grass B. gracilis and C3 species. How these contrary forces will interact to determine community composition is unknown. Initial results support these predictions (Collins et al. 2010), however, this experiment was burned in the 2009 wildfire and we are now tracking post-fire ecosystem reconstruction under this environmental change scenario. Also, in 2011 soils were collected from all replicates for metagenomic analysis to determine how these drivers alters microbial community composition, structure and functioning.
C. Piñon-Juniper Rainfall Experiment (PJREx). Warmer temperatures, increased precipitation variability and decreasing total precipitation may create conditions that exceed the physiological tolerances of some species while increasing productivity and abundance of others. Large-scale vegetation change could occur as catastrophic mortality during climate extremes removes some dominant species and alters growth and reproduction of other species under shifting climate regimes. For example, across the southwestern US, piñon pine (Pinus edulis) has recently experienced 40-95% mortality across large areas while co-occurring juniper (Juniperus monosperma) experienced lower (2-25%) mortality (Breshears et al. 2005, 2010). This differential mortality has decreased woody plant cover, altered species distributions and altered key ecosystem functions like C uptake and water balance. The primary goals of PJREx are to determine the causes of pine mortality under chronic drought and the potential for pine to outperform juniper under conditions of increased annual precipitation. PJREx is a large-scale manipulative experiment to study the mechanistic basis of responses of piñon-juniper woodland to altered rainfall regimes in 40 x 40m plots (n=12, 4 treatments). Using a 45% rainfall exclusion treatment, we are using a hydraulic model (Sperry et al. 1998, 2002) to evaluate water transport limitations in piñon and juniper during severe drought (McDowell et al. 2008). Irrigation (150% of mean annual precipitation) in adjacent plots allows us to measure the effects of increased rainfall. Plots are instrumented to measure soil moisture, temperature, Rs, along with piñon and juniper sap flow. The study will allow us to distinguish among a number of hypothesized mechanisms for the observed differential mortality during drought and responses during periods of above-average rainfall that may actually pre-dispose the system to more catastrophic responses during drought. Results to date suggest that C starvation is a key factor in pine mortality during periods of prolonged drought.