Woody-plant encroachment has been occurring in US rangelands since the 1900s. Overgrazing has been regarded as the primary cause, however, there is evidence that changes in precipitation may play a major role. Although woody-plant encroachment has been studied extensively, little is known about how their seedlings establish in grass-dominated landscapes. To better understand this process, we focused on the seedling establishment phase and ask: How do precipitation, grass-shrub competition, and herbivory interact to influence the probability of woody-plant (Prosopis glandulosa) establishment in grasslands? We hypothesized that unique combinations of grazing, seed and seedling predation and water availability are required for successful shrub recruitment. To test this hypothesis, we conducted a field-based germination experiment at the Chihuahuan Desert Jornada LTER. Using Automatic Rainfall Manipulation Systems, we established 30-2.5x5m plots where 10 received 20% of ambient precipitation, 10 received 100% of ambient precipitation (control), and 10 received 180% of ambient precipitation. The vegetation within half of each plot was clipped to 15cm height to simulate grazing; half was left intact. Scarified shrub seeds (n=5/exclosure) added to clipped and unclipped plots were placed either inside or outside of exclosures (n=16/plot) to prevent access by ants, rodents, or ant+rodents, to quantify germination and establishment.
Results/Conclusions
Prosopis seedling emergence increased with precipitation (X2=6.54, df=2, p<0.05), with precipitation treatments simulating wet years having the highest levels (seedling/m2; 30.0±7.8), and those simulating drought the lowest (6.7±6.7). Herbivory access significantly influenced Prosops recruitment (X2=13.7, P<0.001), with greatest recruitment occurring when seeds were protected from rodents (80.0±19.4) and lowest recruitment occurring when seeds were accessible to rodents (13.3±5.4). The presence or absence of ants had no effects on recruitment. Shrub recruitment on clipped (33.3±19.9) and unclipped plots (13.3±5.4) was comparable (Z=1.02, P>0.05). Recruitment was influenced by the interaction between grazing and precipitation (X2=14.5, df=6, p<0.05), wherein clipping effects were significant on drought and control precipitation treatments, but not in the wet treatment. Our results suggest that grazing influences on the intensity of grass competition with shrub seedlings is modulated by precipitation, that seed/seedling predation by ants is of little consequence and that Prosopis recruitment will be enhanced when years of average to above-average rainfall coincide with lows in rodent populations. Results from this experiment help inform us of the unique combination of drivers necessary for woody-plants to establish on grass-dominated rangelands.