Water availability constrains ecosystem functioning in many ecosystems around the world. A robust assessment of the degree to which water availability impacts ecosystem processes is needed in order to predict ecosystem changes caused by expected changes in climate. Field manipulative experiments represent a straightforward way to explore relationships between water availability end ecosystem functioning but are usually limited in the number of replicates and treatment levels because of their usual high cost. The objectives of this work were: (1) to design a system that provided changes in precipitation for experimental plots from 80% reduction to 80% increase, that it was low cost, which would allow for the numerous replicates needed in field experiments, and that required low maintenance and no connection to the grid, which would allow to manipulate precipitation in remote locations; (2) to test the efficiency of rainfall manipulation and the impact on other environmental variables.
Results/Conclusions
Here, we report on the design and test of the Automated Rainfall Manipulation System (ARMS) that consists of interception and irrigation components. The interception component is a rainout shelter with adjustable interception level that intercepts water and conducts it to a 55-gallon temporary storage tank. The irrigation component consists of a solar-power water pump (12 volts, 200 gallons/hour) that is turned on by a floating switch located at the bottom of the storage tank. The tank acts as a buffer and as the connection between the two main components of the ARMS. Water is pumped to an irrigated plot through a PVC pipe with two sprinklers located at opposite corners. On 15 9 m2 plots, we installed 6 ARMS (three 50% and three 80% interception-irrigation pairs) and three control plots where we measured: 1) incoming water; 2) soil moisture; 3) air temperature; and 4) photosynthetic active radiation (PAR).
We found that 50% and 80% water addition treatments received 47% and 88% more precipitation than ambient yielding a water redistribution efficiency of 80% and 109% respectively. Soil moisture increased linearly across treatments (r2= 0.3 P< 0.01). Temperature did not differ among treatments and PAR was 99% of control under 50% interception rainout shelters and 95% of controls under 80% interception rainout shelters (p > 0.05). The cost of each interception-irrigation pair is $300 and the time required to maintain manipulations for an experiment with 5 levels of precipitation and 10 replicates is of 4 days per month.