U.S. Joint Fire Science Program


Date of this Version


Document Type



Final report Joint Fire Science Program AFP 2004-2, Task 1. Project ID Number: 04-2-1-94


US government work.


This final report summarizes key findings relative to the five major objectives listed in our proposal, along with the crosswalk of accomplished and future deliverables. Additional details concerning objectives, methods, results, and recommendations are presented in Attachment A. Each of our major objectives is listed below with a brief summary of findings to date. Because we only recently finished the final session of data collection, all findings are preliminary and may change as more comprehensive statistical analyses are completed. 1. Evaluate the effects of fire on southern California rodent, large mammal, and bat diversity and study patterns of post fire recovery. Overview: Fire immediately alters the composition of the mammal community in chaparral and coastal sage scrub vegetation, although species responses to fire, including changes over time following fire, are highly species-specific. In chaparral, overall rodent richness is similar between burned and unburned areas, but the composition of the community is very different between these conditions. Furthermore, even after 4 years, the post-burn communities differ from their pre-fire status and from unburned communities. Fire in coastal sage scrub reduced rodent diversity to a simpler community, similar to that that of disturbed habitats dominated by annual grasses and forbs. Carnivore species appear not to be strongly affected by fire. We also detected little effect of burn status on the bat community, although at least two species appear to forage more frequently over unburned chaparral. Rodents: We examined rodent responses to fire in both chaparral (Cedar Fire; Cleveland National Forest) and coastal sage scrub (Otay Fire; Rancho Jamul Ecological Reserve) vegetation beginning about 13 months following fire and continuing until 43 months post fire. In both communities we noted substantial differences in mammal community composition between burned and unburned habitat, and in patterns of post-fire responses over time. Some species were clearly more widespread and abundant in burned than unburned chaparral, others showed the opposite pattern, and still others showed more complex associations with fire over time (Figure 1). For example, the Dulzura kangaroo rat (Dipodomys simulans; DISI) was far more common on burned than unburned plots over the nearly 4-year study duration, while the California mouse (Peromyscus californicus; PECA) was far more common on unburned plots, and the cactus mouse (Peromyscus eremicus; PEER) was equally common on burned and unburned plots. In unburned chaparral, PECA dominated the rodent community across the entire study period, with several other species varying in their relative abundance over time. For example, the California pocket mouse (Chaetodiopus californicus; CHCA) was not abundant on unburned plots during the first spring sampling session, but increased to become the second most abundant species during the last two spring samples (the third and fourth springs following fire). Conversely, the brush mouse (Peromyscus boylii; PEBO) was fairly common on unburned plots early in the study, but gradually declined to become a minor component of the community in latter sessions. Burned plots were dominated by deer mice (Peromyscus maniculatus; PEMA) during the first post-fire spring sample, with DISI second most abundant. Over time, PEMA declined in abundance on burned plots, while DISI increased to become the dominant species. The bigeared woodrat (Neotoma macrotis; NEMA) was very rare on burned plots although common on unburned plots. In contrast, the desert woodrat (N. lepida; NELE) was about equally common on burned and unburned plots. Despite gradual recovery of some chaparral species on burned plots over time, compositional differences between burned and unburned plots were still pronounced 4 years following fire, and the communities are still changing. Sampling would need to continue for more years to determine when communities return to pre-fire status.