We've Moved!
Visit SDSU’s new digital collections website at https://digitalcollections.sdsu.edu
Description
Many rodents in southern California shrublands are granivorous and can affect the distribution of seeds after seed dispersal through seed predation and caching. Rodent seed preferences are hypothesized to affect plant community composition, as rodents tend to prefer larger, exotic seeds over smaller, native seeds. Traditionally, small mammal studies use live trapping to determine community composition and density. Some of these studies couple trapping data with seed predation studies, using indirect inference to relate seed removal patterns to the local granivore community. Using this approach, researchers must assume the rodents that are trapped on large grids are proportional to the individuals removing seed from the dishes and driving seed preference (if any). The only way to evaluate this critical assumption is through direct observation. In this project, I used direct observation and performed a seed predation study. I presented native and non-native seed mixtures in partitioned Petri dishes at stations across fourteen sites at Rancho Jamul Ecological Reserve (RJER; San Diego County, California). At each station, one seed dish was open and thus accessible to all granivores. A second dish was placed inside a wire cage with a PVC tube entrance and was only accessible to rodents. I recorded all seed removal using a custom built, portable infrared digital camera and digital video recorder system. The objectives of this study are: (1) to quantify selective granivory (native versus exotic) (2) If so, to assess whether selectivity varies with animal guild (genus, family) and (3) to evaluate what new insights can be gleaned from direct observation of granivory in this system. Total weight of seed removed varied with season, time of day, dish placement (open or tube), and the presence/absence of certain taxa. Selective granivory (preference for nonnative seed) was common and varied with taxon. Peromyscus and heteromyids were the strongest indicators of seed removal, and also the strongest indicators of selective seed predation: when either Peromyscus or heteromyids were present during a two-night study, significantly more total seed was removed and significantly more non-native seed than native seed was removed. The tendency for a granivore to visit a seed dish varies with taxon. Peromyscus and Chaetodipus were more likely than the other common genera to visit a dish once they had been detected by the video camera. Additionally, the PVC tube entrance to the rodent-only exclosure was assumed to be a benign component of the station and equally used by all rodent species. I documented tube avoidance behavior by all but one (Peromyscus) rodent genus that visited the stations. This study shows that direct observation provides much more detailed and nuanced information on seed dish visitation and preferential seed predation. As technology improves, it is important to consider new and/or alternative methods that can improve our understanding of ecological interactions through direct, detailed observation.
