Skip to main content

Wildfires Alter Rodent Community Structure Across Four Vegetation Types in Southern California, USA


We surveyed burned and unburned plots across four habitat reserves in San Diego County, California, USA, in 2005 and 2006, to assess the effects of the 2003 wildfires on the community structure and relative abundance of rodent species. The reserves each contained multiple vegetation types (coastal sage scrub, chaparral, woodland, and grassland) and spanned from 250 m to 1078 m in elevation. Multivariate analyses revealed a more simplified rodent community structure in all burned habitats in comparison to unburned habitats. Reduction in shrub and tree cover was highly predictive of changes in post-fire rodent community structure in the burned coastal sage scrub and chaparral habitats. Reduction in cover was not predictive for the less substantially burned woodlands and grasslands, for which we hypothesized that interspecific competition played a greater role in post-fire community structure. Across vegetation types, generalists and open habitat specialists typically increased in relative abundance, whereas closed habitat specialists decreased. We documented significant increases in relative abundance of the deer mouse (Peromyscus maniculatus Wagner) and Dulzura kangaroo rat (Dipodomys simulans Merriam). In contrast, we found significant decreases in relative abundance for the California mouse (Peromyscus californicus Gambel), San Diego pocket mouse (Chaetodipus fallax Merriam), desert woodrat (Neotoma lepida Thomas), and brush mouse (Peromyscus boylii Baird). Currently, our research program involves assessment of whether habitat conservation plans (HCPs) in southern California provide long-term protection to HCP covered species, as well as preserve ecosystem function. The scenario of increased wildfires needs to be incorporated into this assessment. We discuss our results in relation to management and conservation planning under a future scenario of larger and more frequent wildfires in southern California.


Wildfires have long been a part of the natural and human-altered environments of southern California, USA. However, the frequency and scope of these fires have substantially increased in comparison to historic levels (Wells et al. 2004). Keeley et al. (2004) related this phenomenon to a combination of fire suppression efforts and an increase in ignition sources associated with human population growth. Although native vegetation communities of southern California evolved various survival strategies to adapt to fire (Hanes 1971, Vogl and Schorr 1972, Keeley and Keeley 1981, Zedler et al. 1983, Keeley and Fotheringham 2001), large and repeated fires can drastically and permanently alter the habitat, converting diverse chaparral and coastal sage scrub into simple non-native grasslands (Zedler et al. 1983, Keeley 2005). Concomitantly, these fires can have short- and long-term impacts on animal communities in a region already recognized as being one of the most at risk areas for loss of biodiversity (Mittermeier et al. 1997).

In October 2003, wildfires burned approximately 300 000 hectares (ha) in southern California. This included approximately 130 000 ha in San Diego County from the Cedar and Otay fires that resulted in the loss of nearly 5000 structures and in 15 human fatalities (CDF 2003). A network of planned reserve habitat and open space totaling almost 80 000 ha, known as the Multiple Species Conservation Plan (MSCP; City of San Diego 1997), was heavily impacted by these two fires. Half of the protected lands within the MSCP were affected by the fires and some protected habitats were entirely burned.

Concern over the recovery of these habitats and the species within them motivated our efforts to conduct this research. As part of a larger long-term study of the effects of wildfires on multiple taxonomic groups (Mendelsohn et al. 2008, Rochester et al. 2010a, Matsuda et al. 2011), we examined the short-term effects of the Cedar and Otay fires on native rodent communities. Rodents occupy a central position in the food web and perform many ecological functions (Polis et al. 1997). As a food source for many predators, they serve as important links between producers and higher level consumers. Rodent burrows provide shelter for many species, serve to aerate the soil, and aid in surface water absorption. Finally, by consuming and distributing seeds, they also affect successional patterns and vegetation characteristics in many habitats (Tevis 1956, Price and Jenkins 1986, Brock and Kelt 2004).

Currently, three rodent species (San Diego pocket mouse [Chaetodipus fallax fallax Merriam], Dulzura pocket mouse [C. californicus femoralis Allen], and desert woodrat [Neotoma lepida Thomas]) in our study are listed as Species of Special Concern by the California Department of Fish and Game (Williams 1986).

Prior to our investigation, studies of fire effects on rodent and small mammal communities have largely been limited to single sites or habitat types without replication over the larger landscape (Cook 1959, Lawrence 1966, Price and Waser 1984, Wirtz et al 1988, Schwilk and Keeley 1998, Converse et al. 2006). The size and scope of the 2003 wildfires presented a unique opportunity to sample across four natural reserves within San Diego County. The reserves span low to moderate elevations and each contained a mixture of vegetation types representing the natural landscapes of southern California. Our objectives were: 1) to determine the short-term effects of fire on rodent community structure across four vegetation types (coastal sage scrub, chaparral, grassland, and woodlands), 2) to determine if changes in post-fire community structure were associated with changes in vegetation structure, 3) to determine if individual species responses were predictable based upon habitat preferences and life-history characteristics, and 4) to identify species at risk in consideration of a changing fire regime in southern California.


Study Sites

Our study sites were Rancho Jamul Ecological Reserve-Hollenbeck Canyon Wildlife Area (Rancho Jamul), Otay Mountain Wilderness-Little Cedar Ridge (Little Cedar), Santa Ysabel Open Space Preserve (Santa Ysabel), and Elliott Chaparral Reserve (Elliott) in San Diego County, California (Figure 1).

Figure 1

The four study sites in San Diego County, California, USA: Little Cedar Ridge (Little Cedar), Elliott Chaparral Reserve (Elliott), Rancho Jamul Ecological Reserve (Rancho Jamul), and Santa Ysabel Open Space Preserve (Santa Ysabel).

Rancho Jamul is in the lower foothills of southwestern San Diego County, approximately 30 km from the coast. Coastal sage scrub, oak and sycamore woodlands, native and nonnative grasslands, and riparian vegetation are the dominant vegetation types. The most common plant species at the site include brome grass (Bromus spp. L.), wild oats (Avena spp. L.), California buckwheat (Eriogonum fasciculatum Benth.), California sage (Artemisia californica Less.), coast live oak (Quercus agrifolia Née), western sycamore (Platanus racemosa Nutt.), laurel sumac (Malosma laurina [Nutt.] Nutt. ex Abrams), and San Diego sunflower (Viguiera laciniata A. Gray). The property is 2797 ha in size with an average elevation of 250 m. The Otay Fire burned 37 % of the southern portion of this site in October 2003.

Little Cedar is 6 km south of Rancho Jamul on the northern slopes of Otay Mountain. It is dominated by coastal sage scrub, mixed chaparral, and cypress forest. The most common plant species at the site are chamise (Adenostoma fasciculatum Hook. and Arn.), tecate cypress (Cupressus forbesii Jeps.), annual non-native grasses, brome grass and wild oats, and California buckwheat (Eriogonum fasciculatum Benth.). The property is 6800 ha in size, ranges from 250 m to 1000 m in elevation, and was completely burned in the Otay Fire of 2003. The average elevation of our study plots in the Little Cedar site is 400 m.

Santa Ysabel spans the upper foothills and lower mountains of north-central San Diego County approximately 60 km from the coast and is dominated by oak and pine woodlands, native and non-native grasslands, coastal sage scrub, chaparral, and riparian vegetation. The most common plant species in the various habitats are coast live oak, brome grass, chamise, Engelmann oak (Quercus engelmannii Greene), wild oats, and white sage (Salvia apiana Jeps.). This property is 2188 ha in size, is split into east and west properties, and has an average elevation of 1078 m. In 2003, 47 % of the property was burned in the Cedar Fire.

Elliott is located approximately 25 km from the coast. It is predominately coastal sage scrub and chaparral, with chamise, California buckwheat, and annual non-native brome grass and wild oats. The property, while only covering 43 ha within its boundaries, is bordered by expansive open space in the Marine Corps Air Station Miramar and has an average elevation of 195 m. The entire property was burned in October 2003 by the Cedar Fire.

Rodent Surveys

In the summer of 2001 and fall of 2003, we surveyed rodents at 16 study plots in the coastal sage scrub and grasslands at Rancho Jamul, giving us a limited amount of pre-fire data. In 2005 and 2006, two and three years following the 2003 wildfires, we surveyed for rodents in burned and unburned habitats at 63 study plots across all four sites using the same methods. We sampled each site four times per year, twice in the spring and twice in the fall. All plots were within one of four vegetation types: chaparral, coastal sage scrub, grassland, and woodland (Table 1).

Table 1 Number of impact and reference plots sampled by study site and vegetation type. Data are presented as number of plots sampled before the fire followed by number of plots sampled after the fire, in brackets. Both the Elliot and Cedar sites burned completely and thus do not contain reference plots. We collected pre-burn data at Rancho Jamul only.

We surveyed all plots using an array of seven pitfall traps, three funnel traps, and 18 Sherman live-traps. For the pitfall trap array, we buried seven 19 L buckets flush with the ground and connected them using shade cloth drift fence. These arrays formed a “Y” shape, with each arm being approximately 15 m long and the buckets buried 7.5 m apart as described by Fisher et al. (2008). We placed one funnel trap along each arm between the two outer buckets (Fisher et al. 2008). We then placed nine pairs of Sherman live-traps approximately 1 m and 16 m from the center bucket between each arm of the drift fence and approximately 9 m from the center bucket along each arm of the drift fence. Each pair was composed of one Sherman XLK trap (8 cm × 9 cm × 30.5 cm) and one larger Sherman XLF15 trap (10 cm × 11.5 cm × 38 cm). We left pitfall and funnel traps unbaited as passive traps for mammals, and baited Sherman traps with steam-rolled oats. We placed raised bucket lids over the pitfall traps and wood roof tiles over the funnel and Sherman traps to provide daytime shading and cover for captured animals. We placed a baseball sized piece of polyester batting in all of the trap types to insulate captured animals from colder nighttime temperatures. For each session, we sampled traps each morning for four consecutive days. Including all trap types, our sampling effort totaled 168 trap-nights per array before the fire and 336 trap-nights per array after the fire.

We handled and processed all animals in a humane manner as approved by USGS Animal Care and Use Committee and in accordance with ASM guidelines (Sikes and Gannon 2011). We identified all rodents to species and temporarily marked them by hair clipping to identify recaptures within each session.

Vegetation Surveys

We measured the vegetation at each plot before the 2003 wildfires and again in the spring of 2005 and 2006. We recorded species composition, vegetation height, substrate, and leaf litter depth at 0.5 m intervals along a 50 m linear transect for a total of 100 data points per plot (Sawyer and Keeler-Wolf 1995, Laakkonen et al. 2001, Fisher et al. 2002). We laid out all vegetation transects in a north-south direction centered at the center bucket of each trapping array. Results from linear vegetation transects have been shown to be comparable to more time-consuming quadrats, especially in chaparral (Bauer 1943).

Data Classification and Reduction

We used a before-after reference-impact approach (BARI) to classify the data for univariate tests (Stewart-Oaten et al. 1986, Underwood 1994, Smith 2002). In this process, we classified each sample as either a reference plot or an impact plot. We classified all plots burned during the 2003 wildfires as impact plots and all plots not burned during the wildfires as reference plots. Before-after classification was relevant only to Rancho Jamul because it was the only study area where we sampled before and after the fires (Table 1).

For the multivariate analyses, we pooled the fire classification into two categories: razed and non-razed. This classification allowed for sufficient statistical power to test the significance of fire effects on rodent community structure within each vegetation type. In terms of the BARI classifications described above, non-razed refers to all of the data that we collected in the pre-fire surveys (before reference and before impact) plus the post-fire surveys at the unburned plots (after reference). Razed refers to all of the data that we collected in the impact plots (after impact) after the 2003 wildfires (Table 1). We reviewed the data using nonmetric multidimensional scaling (MDS) to ensure the validity of our pooling of the unburned reference and impact plots (i.e., did not group as separate clusters).

Statistical Analysis

To examine changes to the vegetation, we computed averages and standard errors for the percentages of cover types (grasses, forbs, shrubs + trees) in reference plots and fire-impacted plots, both before and after the wildfire. Pearson’s correlations were also conducted as appropriate to explore relationships between cover types.

For our index of species abundance, we used the minimum number known alive (i.e,. number of unique individuals; MNA). We calculated MNA by removing all recaptures within each four-day trapping session at each array. Although MNA can be biased as an abundance estimator, it has been shown to be proportional to population sizes of small mammals and is thus a reasonable index of abundance (Slade and Blair 2000). To reduce variability due to seasonal or other environmental factors, we averaged the MNA for each species across all pre-fire sample sessions and across all post-fire sample sessions. We refer to this average as a capture rate (average number of individuals captured per session). We used the averaged pre-fire and post-fire capture rates with sample plots treated as replicates for the basis for all further analyses.

Before conducting formal significance tests, we initially explored large scale patterns in the data using a combination of approaches. We averaged capture rates by vegetation type and fire condition (razed and non-razed), and a created a Bray-Curtis similarity matrix of the data. We used MDS to illustrate similarities among sample plots and fire condition. In an MDS plot, samples that are relatively close together can be interpreted as more similar, while those that are more distant are interpreted as more dissimilar. We used hierarchical agglomerative cluster analysis with group-average linking to examine the degree of similarity between razed rodent communities in comparison to the non-razed communities. We then overlaid vectors on the MDS plot to visualize the relative magnitude and directional response (trajectory) of individual species abundances in relation to razed and non-razed vegetation types (Clarke and Warwick 2001, Clarke and Gorley 2006).

We used a multivariate analysis of variance routine (PERMANOVA; Anderson 2001, McArsle and Anderson 2001) to test whether the factors of burn status (razed or non-razed), site, vegetation type, burn status × vegetation type, and the percent cover of shrubs and trees predicted any differences in rodent community structure. This was based on a dissimilarity matrix of species abundance data at each sample array. The PERMANOVA routine performs a partitioning of the total sum of squares and calculates a distance-based pseudo-F statistic for each term in the model. We used 999 permutations to calculate P-values. We followed up any significant burn status × vegetation interaction by analyzing each vegetation type independently. For these analyses, to determine how much of the fire-related change in rodent community structure was associated with changes in vegetation structure, we calculated the proportion of variance from the burn status term that was also explained by the cover term (shrub + tree cover). For each vegetation type, we accomplished this by testing two models with ordered terms: 1) burn status then cover, and 2) cover then burn status. The proportion of variance due to fire that was explained by changes in shrub and tree cover was the difference in the component of variation (COV) due to burn status in these two models divided by the total COV due to burn status in the first model.

Finally, we tested for effects of the fire on the relative abundance of individual species within vegetation types in two ways. First, using the post-fire data from each vegetation type, we generated models of relative abundance with site, species, and burn status as fixed factors. We conducted contrast tests to determine whether relative abundance of a species at the impact plots differed from that at the reference plots. For this, we pooled variances across species and sample plots within each vegetation type. Second, for the impact and reference plots where we had both pre-fire and post-fire data (Rancho Jamul: coastal sage scrub and grassland), we subtracted the pre-fire capture rate from the post-fire capture rate for each species at each plot to produce a “net change after fire” statistic. We then ran a two-way ANOVA with species and burn status (reference vs. impact) as fixed factors. We used the differences of least squares means to test whether the change in rodent abundance at the impact plots differed from the change in abundance at the reference plots.

For all analyses, we set the alpha level at 0.10. We performed all multivariate analyses using Primer-E software (Primer E Ltd. 2006, Plymouth, United Kingdom) and univariate analyses using SAS software (SAS Institute Inc. 2002–2003, Cary, North Carolina, USA).


Vegetation Structure

Coastal sage scrub and chaparral fire-impacted plots had 77 % and 46 % reduced shrub and tree cover compared to reference plots, respectively. In woodlands, there was less reduction in shrub and tree cover for fire-impacted plots (−26 %). Forb cover was an average of 30 % greater in coastal sage scrub fire-impacted plots. Over both reference and impact coastal sage scrub plots, increased forb cover was correlated with decreased shrub cover (Pearson’s r = −0.87, P < 0.001). Forb cover did not differ between reference and impact plots for the other vegetation types. Grass cover did not differ between reference and impact plots in any vegetation type (Table 2). Grasses were composed of predominantly non-native Bromus spp. and Avena spp. before and after the fires in all vegetation types (Rochester et al. 2010b).

Table 2 Means and standard errors for the percent cover of forbs, grass, shrubs, and trees over reference and impact plots.

Overall Trapping Results

We captured 16 rodent species across all vegetation types for a total of 4444 captures pre-fire and post-fire. The most common captures included the deer mouse, Dulzura kangaroo rat, San Diego pocket mouse, and the harvest mouse (Reithrodontomys megalotis Baird). Other species captured were the California pocket mouse, California vole (Microtus californicus Peale), desert woodrat, big-eared woodrat (Neotoma macrotis Thomas), brush mouse, California mouse, and the cactus mouse (Peromyscus eremicus Baird). We also documented the presence of Botta’s pocket gopher (Thomomys bottae Eydoux and Gervais), house mouse (Mus musculus L.), and the California ground squirrel (Otospermophilus beecheyi Richardson). We captured these last three species only rarely and therefore did not use them in any analyses.

Post-Fire Multivariate Community Analyses

Multidimensional scaling graphs and hierarchical clustering showed that across vegetation types, the rodent communities from razed plots more closely resembled each other than the corresponding non-razed rodent communities (Figure 2). Using group-average linking, razed coastal sage scrub, chaparral, grassland, and woodland merged into a single group at a similarity level of 59.2 %. This is in contrast to the non-razed communities that merge at a lower similarity level of 38.2 % (Figure 2). Vectors of individual species abundance in relation to the razed and non-razed rodent communities showed positive directional responses for both deer mouse and Dulzura kangaroo rat to the razed communities, whereas California mouse, brush mouse, San Diego pocket mouse, California pocket mouse, California vole, desert woodrat, and cactus mouse vectors showed positive directional responses to non-razed communities. The harvest mouse response was mixed, with increased abundance along a directional vector toward both non-razed and razed woodlands and razed grasslands (Figure 2).

Figure 2

Multivariate Community Analysis. A: Nonmetric multi-dimensional scaling (MDS) with B: hierarchical clustering showing how the rodent communities in razed plots (R) of coastal sage scrub (CSS), chaparral (CHAP), grassland (GRASS), and woodland (WOOD) shifted away from the communities in the non-razed plots (NR). C: Vectors of individual species abundances in relation to the razed (✱) and non-razed () communities showed positive responses for both Peromyscus maniculatus (PEMA), and Dipodomys simulans (DISI) toward the fire-impacted communities, whereas Peromyscus californicus (PECA), Peromyscus boylii (PEBO), Chaetodipus fallax (CHFA), Microtus californicus (MICA), Neotoma lepida (NELE), Chaetodipus californicus (CHCA), and Peromyscus eremicus (PEER) showed positive responses toward non-razed communities.

The PERMANOVA resulted in a significant burn condition × vegetation interaction (Pseudo-F 5, 80 = 1.76, P = 0.001), indicating that the response of the rodent communities to fire varied among vegetation types. Therefore, we did the follow-up analyses of the burn effects for each vegetation type separately. These analyses showed that rodent community structure in razed plots was significantly different than non-razed plots in all four vegetation types; coastal sage scrub (Pseudo-F 1, 23 = 6.51, P = 0.001, 30.5 % of total variation), chaparral (Pseudo-F 1, 11 = 4.37, P = 0.026, 18.5 % of total variation), grassland (Pseudo-F 1, 15 = 4.53, P = 0.004, 21.6 % of total variation), and woodland (Pseudo-F 1, 13 = 2.73, P = 0.016, 14.0 % of total variation). The proportion of variation (POV) of the rodent community structure explained by both burn condition and decreased shrub and tree cover (i.e., shared variation) was high for coastal sage scrub (80 %) and chaparral (58 %), but relatively low for grassland (12.1 %) and woodland (4.6 %) habitats.

Differences in Relative Abundance of Species

In coastal sage scrub unburned reference plots, the relative abundance of 11 rodent species was evenly distributed (i.e., all within a factor of four) (Figure 3). However, the distribution was highly skewed in the burned impact plots. Examining the difference in relative abundance of all of our captured species revealed that two rodent species responded positively to fire: the deer mouse (Diff = 2.53, se = 0.92, t209 = 7.49, P = 0.006) and the Dulzura kangaroo rat (Diff = 1.69, t209 = 3.33, se = 0.92, P = 0.069). Conversely, two other species responded negatively to fire, with the California mouse (Diff = −1.61, se = 0.92, t209 = 3.00, P = 0.084) and desert woodrat (Diff = −2.01, se = 0.92, t209 = 4.74, P = 0.031; Figure 3) being less abundant in the impact plots. All other species were less abundant on average, but not significantly, on plots impacted by the fires. Tests of paired contrasts between the four impact and four reference plots in coastal sage scrub at Rancho Jamul also showed a significant increase in abundance of the Dulzura kangaroo rat (Diff = 3.34, se = 1.914, t41 = 1.75, P = 0.088), but not the deer mouse (Diff = −0.69, se = 1.56, t41 = 0.44, P = 0.662), which also increased at the reference plots in the same sampling year.

Figure 3

Relative abundance of rodent species from post-fire surveys in reference (A) and impact (B) plots in coastal sage scrub (CSS) communities. Species are ordered from most abundant to least abundant. C) Species differences in relative abundance (impact minus reference) in post-fire CSS communities are presented in the order of greatest change to smallest change. Relative abundance is the minimum number known alive (MNA) per array per sample session. Error bars are ±1 standard error. Levels of significance are indicated as follows: *** P < 0.01, ** P < 0.05, * P < 0.10. Species codes are as in Figure 2, except for Reithrodontomys megalotis (REME).

In grassland reference plots, the San Diego pocket mouse and deer mouse were the two most abundant species (Figure 4). In grassland burned impact plots, the relative abundance of the San Diego pocket mouse was substantially lower (Diff = −1.83, se = 0.57, t99 = 10.45, P = 0.002). The abundance of the San Diego pocket mouse also decreased at the four impact and four grassland reference plots at Rancho Jamul (San Diego pocket mouse Diff = −5.44, se = 0.98, t19 = 5.54, P < 0.001).

Figure 4

Relative abundance of rodent species from post-fire surveys in reference (A) and impact (B) plots in grassland communities. Species are ordered from most abundant to least abundant. C) Species differences in relative abundance (impact minus reference) in post-fire grassland communities are presented in the order of greatest change to smallest change. Abundance index, error bars, and significance levels are as defined in Figure 2.

In chaparral reference plots, the California mouse and cactus mouse were the two most abundant species (Figure 5), whereas the deer mouse and Dulzura kangaroo rat were both highly abundant in the fire-impacted plots. Results in chaparral were very similar to coastal sage scrub, as the deer mouse (Diff = 5.35, se = 1.23, t154 = 19.04, P = <0.001) and Dulzura kangaroo rat (Diff = 2.76, se = 1.23, t154 = 5.10, P = 0.025) responded positively to fire, whereas the California mouse (Diff = −2.24, se = 1.228, t154 = 3.32, P = 0.07) responded negatively to fire. Other species showed very little differences in abundance across reference and impact plots (Figure 5).

Figure 5

Relative abundance of rodent species from post-fire surveys in reference (A) and impact (B) plots in chaparral communities. Relative abundance is the minimum number known alive (MNA) per array per sample session. Species are ordered from most abundant to least abundant. C) Species differences in relative abundance (impact minus reference) in post-fire chaparral communities are presented in the order of greatest change to smallest change. Error bars are ±1 standard error. Levels of significance are indicated as follows: *** P < 0.01, ** P < 0.05, * P < 0.10. Species codes are as in Figures 2 and 3.

In the woodland reference plots, the deer mouse and brush mouse were similarly high in their relative abundances. In the fire-impacted plots, however, the relative abundance of the deer mouse was double that of the reference plots (Diff = 2.42, se = 0.63, t143 = 14.50, P = 0.002), and brush mouse abundance was 90 % lower than in the reference plots, making it a minor component of the post-fire community (Diff = −1.77, se = 0.63, t143 = 7.68, P = 0.006). Other species showed very little differences in abundance across reference and impact plots (Figure 6).

Figure 6

Relative abundance of rodent species from post-fire surveys in reference (A) and impact (B) plots in woodland communities. Species are ordered from most abundant to least abundant. C) Species differences in relative abundance (impact minus reference) in post-fire woodland communities are presented in the order of greatest change to smallest change. Abundance index, error bars, and significance levels are as defined in Figure 2.


Rodent Community Response to Fire

The results of our study indicate that the coastal southern California rodent community exhibits a strong response to wildfire within the first two to three years after fire. Multivariate tests for the effects of fire on rodent communities showed significant differences in rodent community structure in coastal sage scrub, grassland, chaparral, and woodland vegetation types. Overall, the rodent communities in the burned habitats were much more similar to each other than to rodent communities in the unburned habitats.

This response to fire was consistent across vegetation types in that the generalist deer mouse became the most abundant species in the rodent community at all four vegetation types. Other rodents largely responded in a manner consistent with their microhabitat and food preferences. For instance, the open habitat and seed specialist, Dulzura kangaroo rat, increased in multiple habitats, whereas the closed habitat shrub specialists, California mouse, desert woodrat, and brush mouse, decreased.

Habitat for closed shrub specialists was substantially reduced in the burned coastal sage scrub and chaparral communities, with losses of 77 % and 46 % shrub and tree cover, respectively. In contrast, the wildfires had less notable impacts on riparian woodlands with an average reduction of 26 % shrub and tree cover. The riparian environments contain high levels of moisture, which likely allowed them to maintain more of their pre-fire habitat structure and possibly act as refugia for the animals sheltering within it (Chang 1996). Although the rodent community in burned grasslands was significantly different than in unburned grasslands, the vegetation structure did not change as measured by shrub, forb, or grass cover. Nor were there significant changes in the proportion of native to non-native grasses (Rochester et al. 2010) that may affect the rodent community.

Previous research reported negative relationships between small mammal species richness and loss of vegetation structure in human-degraded coastal sage scrub and chaparral (Bolger et al. 1997, Savajout et al. 1998, Diffendorfer et al. 2007a). In our study, reduction in shrub and tree cover (strongly correlated with increased forb cover) explained most of the variation in fire-altered rodent community structure in coastal sage scrub (80 %), but less in chaparral (33 %) and woodland communities (5 %). Thus, loss of vegetation structure appeared to be a more important driver of fire-related change in rodent community structure in the more heavily burned habitats. We hypothesize that the remaining variation should be largely explained by differences in initial mortality (especially for above-ground nesters), species interactions, and predation pressures, which would appear to be more important drivers in grasslands and in the lesser burned habitats. The major post-fire changes in woodland communities were an increase in deer mouse abundance and a decrease in brush mouse abundance. Similarly, the primary post-fire changes in grassland communities were an increase in deer mouse abundance and a decrease in San Diego pocket mouse abundance. In southern California habitats, this points to potential competitive advantage by the generalist deer mouse in all post-fire environments, even those maintaining a considerable amount of habitat structure.

Individual Species Responses

The deer mouse was the most abundant species in fire-impacted plots across all four vegetation types. This is in comparison to the reference plots, where the deer mouse was not abundant. This species is a habitat and dietary generalist (Jameson 1952, Cook 1959, M’Closkey 1972, Holbrook 1978) and has been shown to invade post-fire habitats in other studies (Cook 1959, Lawrence 1966, Meserve 1976b, Holbrook 1978, Kaufman et al. 1988, Wirtz et al. 1988, Schwilk and Keeley 1998, Converse et al. 2006). A negative relationship was found between the deer mouse and fire severity in Yosemite National Park; however, sampling ranged from two to 15 years post-fire and may have represented a longer term response (Roberts et al. 2008). A combination of reproductive traits such as postpartum estrous and large litter sizes enable the deer mouse to reproduce faster than all other species in this study, except for the similarly fecund harvest mouse (Rood 1966, Richins et al. 1974, Jameson and Peeters 2004). In contrast to the harvest mouse, this species occupies all vegetation types. Therefore, the deer mouse is well equipped to respond rapidly to the abundant seed bank resources, new plant growth, and decreased competition from other rodents offered in these southern California post-fire habitats.

There was a significant increase in the relative abundance of the Dulzura kangaroo rat in burned coastal sage scrub and chaparral. Post-fire increases of this species have been previously documented in coastal sage scrub (Price and Waser 1984). In both vegetation types, this bipedal heteromyid rodent occupies open habitat on gentle slopes, and its diet is primarily composed of seeds (Meserve 1976a, 1976b; Jameson and Peeters 2004, Kelt et al. 2005). By removing dense vegetation and leaf litter and exposing seed resources (Wright and Bailey 1982, Kaufman et al. 1988), the fire created more suitable habitat conditions for this species.

The desert woodrat significantly decreased in coastal sage scrub. The woodrat builds large nests above ground in rocks, trees, and shrubs using woody materials and other flammable debris. The decrease in the relative abundance in burned coastal sage scrub was likely due to direct mortality from the combustion of its above-ground nests as the fires occurred (Wirtz et al. 1988). However, some desert woodrats living in rocky outcrops with sufficient retreats may have survived. The populations may be depressed into the near future due to lack of suitable materials with which to build new nests (Wirtz et al. 1988). We did not have enough data to evaluate the response of the big-eared woodrat, as it was present in only a few of our sites. Because of its similar life history, we would expect this species to experience similar declines (Bolger et al. 1997).

We observed significant decreases in the relative abundance of the California mouse within burned chaparral and coastal sage scrub communities. This species has strong food preferences for shrub fruits, seeds, and flowers (Meserve 1976a, 1976b). For habitat, this species prefers chaparral and dense coastal sage scrub (M’Closkey 1972) and shelters in burrows and woodrat nests (Grinnell and Orr 1934, Merritt 1974). Mortality from fire and reduced habitat suitability from the thinning of burned chaparral and coastal sage scrub habitats are likely reasons for the significant decreases observed for this species. Post-fire declines in this species have been documented in other studies (Lawrence 1966, Wirtz et al. 1988).

There was a significant decrease in the relative abundance of the brush mouse in woodland plots. The brush mouse is arboreal, commonly using trees, logs, and shrub canopies for foraging and nesting (Holbrook 1979, Kalcounis-Rüppell and Millar 2002, Jameson and Peeters 2004). Its post-fire decrease may be due to reduced suitable habitat as it prefers areas with medium to high densities of shrub and tree cover (Holbrook 1979, Quinn 1990, Roberts et al. 2008). As discussed, densities may have also been depressed due to increased competition from the deer mouse.

The San Diego pocket mouse decreased in burned grassland. This species is a granivore with a diet consisting of large amounts of seeds from both annual forbs and grasses (Meserve 1976a). The preferred habitats of the San Diego pocket mouse tend to be sparsely vegetated areas with rocky or sandy soils suitable for digging burrows (Price and Waser 1984, Lackey 1996). Since this species is typically associated with open scrub habitats, we expected an increase in numbers due to an increase in suitable habitat. However, negative relationships of San Diego pocket mouse abundance with fragmentation and fire-related disturbance have been reported (Price and Waser 1984, Bolger et al. 1997). We hypothesize that this species requires an intermediate amount of shrub cover and that densities may have been depressed due to a combination of habitat alteration and increased competition from the deer mouse.

In burned coastal sage scrub and grasslands, there were insignificant but marked decreases in the relative abundance of the California vole. In compact clay soils, typical in southern California grasslands, the California vole largely prefers to forage and nest above-ground (Stark 1963). This species also has small home ranges and displays high site fidelity (Stark 1963). Work by Cook (1959) and Pearson (1959) indicate that the California vole requires at least one year’s worth of grass growth before cover is suitable. Since our study took place in the second and third years after the fire, when grasses were abundant, we hypothesize that this species was negatively impacted by direct mortality from the fires.

We did not have the power to assess the impact of the fires on some of the less abundant members of the rodent community such as the cactus mouse, harvest mouse, and big-eared woodrat. Results from a concurrent study of both a chaparral and a coastal sage scrub site affected by the 2003 wildfires were consistent with our findings (Diffendorfer et al. 2007b)


Despite suppression efforts and management plans, wildland fires continue to occur from both natural ignition sources and those associated with increasing human population (Keeley et al. 2004). Our results indicate that a more frequent and intense fire regime will simplify rodent communities and thus result in a loss of biodiversity for the region and its ecosystems. The result may be rodent communities dominated by generalists and disturbance specialists, such as the deer mouse and the Dulzura kangaroo rat.

By understanding the responses of rodent species to fire according to their habitat, nesting, and food preferences, our findings can help managers to identify those species at risk of population decline or extirpation from wildfires, as well as to predict responses of unstudied species. The rodent species at greatest risk from large and intense fires are those associated with mature woodland or shrubland habitats that require some degree of shrub and tree structure for their habitat or dietary needs. In San Diego County, these species include the desert and big-eared woodrats, the California mouse, and the brush mouse. Although data were lacking in this study, we would also expect the California pocket mouse to be at moderate risk for decline. We have found similar patterns with responses of herpetofauna species to these large scale fires, with responses consistent with habitat preferences and life history characteristics (Rochester et al. 2010).

The spatial extent and severity of wildfires are important for determining the level of risk to rodent species and populations (Shaffer and Laudenslayer 2006). Smaller, patchy, less severe wildfires may allow more species to survive while also leaving some portions of the vegetation intact. These types of fires, occurring at natural intervals, should allow the persistence of a diverse rodent community. Currently, however, most fires are ignited by human sources during the fall when dry Santa Ana winds act to spread the flames quickly. These large, intense fires increase mortality and, more drastically, alter wide swaths of habitat. As a consequence, they are more likely to have a greater and more sustained impact on small animal communities. In our study, the fires at Santa Ysabel and Rancho Jamul were patchy, whereas the Elliott and Little Cedar sites experienced intense and complete burns at a much greater distance from un-burned refugia. We expect that the communities at these latter two sites will take more time to recover.

In addition to fire severity concerns, fire frequency will be paramount to determining risk and long-term fire impacts to the southern California habitats. Under repeated short-return interval fire events, the potential exists that both the vegetation and animal communities will transition to grassland communities and not rebound to those of intact, unburned chaparral and coastal sage scrub (Zedler et al. 1983, Keeley 2005). In our study areas, non-native grasses were present and naturalized in all vegetation types prior to the 2003 wildfires, likely already affecting vegetation fuel properties and thus fire frequency (Brooks et al. 2004). In 2007, San Diego County experienced a second series of large scale wildfires that affected most of Rancho Jamul and one array at Santa Ysabel. We are currently resampling the vegetation and rodent communities at all four study sites to assess current conditions and states of post-fire recovery.

The scenario of large and more frequent wildfires is also relevant to the Single Large or Several Small (SLOSS) debate about reserve design and the conservation of biodiversity in this region (Diamond 1975, Simberloff and Abele 1976). Although larger reserves should conserve more species and be less affected by habitat fragmentation (i.e., Bolger et al. 1997), these reserves are more difficult to defend from wildfires. In contrast, smaller reserves within city boundaries may be more likely to be defended as part of firefighting efforts to defend homes in urban and rural areas. A combination of approaches is likely needed.

Currently, we are trying to assess whether current habitat conservation plans (HCPs) in southern California provide long-term protection to covered species, as well as preserve ecosystem function. Federal and state laws require that HCPs be implemented within an adaptive management framework. This framework represents an iterative approach, where an information feedback loop between monitoring and management improves understanding of species dynamics and ecological systems (Atkinson et al. 2004). The future scenario of increased wildfires should be incorporated into this framework (Regan et al. 2008). This will likely require creativity in implementing fire minimization measures, creating fire-safe refugia in natural areas, and working with local and state firefighting agencies to include plans for defense of sensitive natural areas.

Literature Cited

  1. Anderson, M.J. 2001. A new method for nonparametric multivariate analysis of variance. Austral Ecology 26: 32–46.

    Google Scholar 

  2. Atkinson, A.J., P.C. Trenham, R.N. Fisher, S.A. Hathaway, B.S. Johnson, S.G. Torres, and Y.C. Moore. 2004. Designing monitoring programs in an adaptive management context for regional multiple species conservation plans. US Geological Survey Technical Report. USGS Western Ecological Research Center, Sacramento, California, USA.

    Book  Google Scholar 

  3. Bauer, H.L. 1943. The statistical analysis of chaparral and other plant communities by means of transect samples. Ecology 24: 45–60. doi: 10.2307/1929859

    Article  Google Scholar 

  4. Bolger, D.T., A.C. Alberts, R.M. Sauvajot, P. Potenza, C. McCalvin, D. Tran, S. Mazzoni, and M. E. Soulé. 1997. Response of rodents to habitat fragmentation in coastal southern California. Ecological Applications 7: 552–563. doi: 10.1890/1051-0761(1997)007[0552:RORTHF]2.0.CO;2

    Article  Google Scholar 

  5. Brock, R.E., and D.A. Kelt. 2004. Keystone effects of the endangered Stephens’ kangaroo rat (Dipodomys stephensi). Biological Conservation 116: 131–139. doi: 10.1016/S0006-3207(03)00184-8

    Article  Google Scholar 

  6. Brooks, M.L., C. D’Antonio, D.M. Richardson, J.B. Grace, J.E. Keeley, J.M. DiTomaso, R.J. Hobbs, M. Pellant, and D. Pyke. 2004. Effects of invasive alien plants on fire regimes. BioScience 54: 677–688. doi: 10.1641/0006-3568(2004)054[0677:EOIAPO]2.0.CO;2

    Article  Google Scholar 

  7. California Department of Forestry and Fire Protection [CDF]. 2003. Large fire summary. <>. Accessed 17 August 2010.

  8. Chang, C. 1996. Ecosystem responses to fire and variations in fire regimes. Pages 1071–1099 in Sierra Nevada Ecosystem Project: final report to Congress. Volume II, Assessments and Scientific Basis for Management Options. Wildland Resources Center Report No. 37. University of California, Davis, USA.

    Google Scholar 

  9. City of San Diego. 1997. Multiple Species Conservation Program: city of San Diego MSCP sub-area plan. <>. Accessed 13 June 2011.

  10. Clarke, K.R., and R.N. Gorley. 2006. PRIMER v6: user manual/tutorial. Primer-E, Plymouth, United Kingdom.

    Google Scholar 

  11. Clarke, K.R., and R.M. Warwick. 2001. Changes in marine communities: an approach to statistical analysis and interpretation, 2nd edition. Primer-E, Plymouth, United Kingdom.

    Google Scholar 

  12. Converse, S.J., G.C. White, and W.M. Block. 2006. Small mammal responses to thinning and wildfire in ponderosa pine-dominated forest of the southwestern United States. Journal of Wildlife Management 70: 1711–1722. doi: 10.2193/0022-541X(2006)70[1711:SMRTTA]2.0.CO;2

    Article  Google Scholar 

  13. Cook Jr., S.F. 1959. Effects of fire on a population of small rodents. Ecology 40: 102–108. doi: 10.2307/1929926

    Article  Google Scholar 

  14. Diamond, J.M. 1975. The island dilemma: lessons of modern biogeographic studies for the design of natural reserves. Biological Conservation 7: 129–146. doi: 10.1016/0006-3207(75)90052-X

    Article  Google Scholar 

  15. Diffendorfer, J.E., G.M. Fleming, J.M. Duggan, R.E. Chapman, M.E. Rahn, M.J. Mitrovich, and R.N. Fisher. 2007a. Developing terrestrial, multi-taxon indices of biological integrity: an example from coastal sage scrub. Biological Conservation 140: 130–141. doi: 10.1016/j.biocon.2007.08.005

    Article  Google Scholar 

  16. Diffendorfer, J., W. Spencer, S. Tremor, J. Beyers, and G. Fleming. 2007b. Effects of fire severity and distance from unburned edge on mammalian community post-fire recovery. Joint Fire Science Program Report AFP 2004–2, Task 1, Project Number 04-2-1-94. US Department of Agriculture and US Department of the Interior Joint Fire Science Program, Boise, Idaho, USA.

    Google Scholar 

  17. Fisher, R.N, D. Stokes; C.J. Rochester, C.S. Brehme, S.A. Hathaway, and T.J. Case. 2008. Herpetological monitoring using a pitfall trapping design in southern California: US Geological Survey techniques and methods. <>. Accessed 13 June 2011.

  18. Fisher, R.N., A.V. Suarez, and T.J. Case. 2002. Spatial patterns in the abundance of the coastal horned lizard. Conservation Biology 16: 205–215. doi: 10.1046/j.1523-1739.2002.00326.x

    Article  Google Scholar 

  19. Grinnell, J., and R.T. Orr. 1934. Systematic review of the californicus group of the rodent genus Peromyscus. Journal of Mammalogy 15: 210–220. doi: 10.2307/1373853

    Article  Google Scholar 

  20. Hanes, T.L. 1971. Succession after fire in the chaparral of southern California. Ecological Monographs 41: 27–52. doi: 10.2307/1942434

    Article  Google Scholar 

  21. Holbrook, S. 1978. Habitat relationships and coexistence of four sympatric species of Peromyscus of northwestern New Mexico. Journal of Mammalogy 59: 18–26. doi: 10.2307/1379871

    Article  Google Scholar 

  22. Holbrook, S. 1979. Vegetational affinities, arboreal activity, and coexistence of three species of rodents. Journal of Mammalogy 60: 528–542. doi: 10.2307/1380093

    Article  Google Scholar 

  23. Jameson Jr., E.W. 1952. Food of deer mice, Peromyscus maniculatus and P. boylii, in the northern Sierra Nevada, California. Journal of Mammalogy 33: 50–60. doi: 10.2307/1375640

    Article  Google Scholar 

  24. Jameson Jr., E.W., and H.J. Peeters. 2004. Mammals of California. University of California Press, Berkeley, USA.

    Google Scholar 

  25. Kalcounis-Rüppell, M., and J.S. Millar. 2002. Partitioning of space, food, and time by syntopic Peromyscus boylii and P. californicus. Journal of Mammalogy 83: 614–625. doi: 10.1644/1545-1542(2002)083<0614:POSFAT>2.0.CO;2

    Article  Google Scholar 

  26. Kaufman, G.A., D.A. Kaufman, and E. Finck. 1988. Influence of fire and topography on habitat selection by Peromyscus maniculatus and Reithrodontomys megalotis in ungrazed tallgrass prairie. Journal of Mammalogy 69: 342–352. doi: 10.2307/1381384

    Article  Google Scholar 

  27. Keeley, J.E. 2005. Fire as a threat to biodiversity in fire-type shrublands. USDA Forest Service General Technical Report PSW-GTR-195: 97–106. Pacific Southwest Research Station, Albany, California, USA.

    Google Scholar 

  28. Keeley, J.E., and C.J. Fotheringham. 2001. Historic fire regime in southern California shrublands. Conservation Biology 15: 1536–1548. doi: 10.1046/j.1523-1739.2001.00097.x

    Article  Google Scholar 

  29. Keeley, J.E., and S.C. Keeley. 1981. Post-fire regeneration of southern California chaparral. American Journal of Botany 68: 524–530. doi: 10.2307/2443028

    Article  Google Scholar 

  30. Keeley, J.E., C.J. Fotheringham, and M.A. Moritz. 2004. Lessons from the October 2003 wildfires in southern California. Journal of Forestry 102: 26–31.

    Google Scholar 

  31. Kelt, D.A., J.A. Wilson, and E.S. Konno. 2005. Responses of two kangaroo rats (Dipodomys) to the 1997–1998 El Niño southern oscillation event. Journal of Mammalogy 86: 265–274. doi: 10.1644/BER-127.1

    Article  Google Scholar 

  32. Laakkonen, J., R.N. Fisher, and T.J. Case. 2001. Effect of land cover, habitat fragmentation, and ant colonies on the distribution and abundance of shrews in southern California. Journal of Animal Ecology 70: 776–788. doi: 10.1046/j.0021-8790.2001.00542.x

    Article  Google Scholar 

  33. Lackey, J.A. 1996. Chaetodipus fallax. Mammalian Species 517: 1–6. doi: 10.2307/3504127

    Article  Google Scholar 

  34. Lawrence, G.E. 1966. Ecology of vertebrate animals in relation to chaparral fire in the Sierra Nevada foothills. Ecology 47: 278–291. doi: 10.2307/1933775

    Article  Google Scholar 

  35. McArdle, B.H., and M.J. Anderson. 2001. Fitting multivariate models to community data: a comment on distance based redundancy analysis. Ecology 82: 290–297. doi: 10.1890/0012-9658(2001)082[0290:FMMTCD]2.0.CO;2

    Article  Google Scholar 

  36. M’Closkey, R.T. 1972. Temporal changes in populations and species diversity in a California rodent community. Journal of Mammalogy 53: 657–676. doi: 10.2307/1379205

    Article  Google Scholar 

  37. Matsuda, T., G. Turschak, C.S. Brehme, C.J. Rochester, M.J. Mitrovich, and R.N. Fisher. 2011. Effects of large-scale wildfire on ground foraging ants (Hymenoptera: Formicidae) in southern California. Environmental Entomology 40(2): 204–216. doi: 10.1603/EN10061

    Article  Google Scholar 

  38. Mendelsohn, M.B., C.S. Brehme, C.J. Rochester, D.C. Stokes, S.A., Hathaway, and R.N. Fisher. 2008. Responses of bird communities to wildland fires in southern California. Fire Ecology 4(2): 63–82. doi: 10.4996/fireecology.0402063

    Article  Google Scholar 

  39. Merritt, J. 1974. Factors influencing the local distribution of Peromyscus californicus in northern California. Journal of Mammalogy 55: 102–114. doi: 10.2307/1379260

    Article  Google Scholar 

  40. Meserve, P.L. 1976a. Food relationships of a rodent fauna in a California coastal sage scrub community. Journal of Mammalogy 57: 300–319. doi: 10.2307/1379690

    Article  Google Scholar 

  41. Meserve, P.L. 1976b. Habitat and resource utilization by rodents in California coastal sage scrub community. Journal of Animal Ecology 45: 647–666. doi: 10.2307/3573

    Article  Google Scholar 

  42. Pearson, O.P. 1959. A traffic survey of Microtus-Reithrodontomys runways. Journal of Mammalogy 40: 169–180. doi: 10.2307/1376431

    Article  Google Scholar 

  43. Polis, G.A., W.B. Anderson, and R.D. Holt. 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annual Review of Ecology and Systematics 28: 289–316. doi: 10.1146/annurev.ecolsys.28.1.289

    Article  Google Scholar 

  44. Price, M.V., and S.H. Jenkins. 1986. Rodents as seed consumers and dispersers. Pages 191–235 in: D.R. Murray, editor. Seed dispersal. Academic, Sydney, Australia.

    Chapter  Google Scholar 

  45. Price, M.V., and N.M. Waser. 1984. On the relative abundance of species: post-fire changes in a coastal sage scrub rodent community. Ecology 65: 1161–1169. doi: 10.2307/1938324

    Article  Google Scholar 

  46. Quinn, R.D. 1990. Habitat preferences and distribution of mammals in California chaparral. USDA Forest Service Research Paper PSW-202. Pacific Southwest Research Station, Berkeley, California, USA.

    Book  Google Scholar 

  47. Regan, H.M., L.A. Hierl, J. Franklin, D.H. Deutschman, H.L. Schmalbach, C.S Winchell, and B.S. Johnson. 2008. Species prioritization for monitoring and management in regional multiple species conservation plans. Diversity and Distributions 14: 262–271.

    Google Scholar 

  48. Richins, G.H., H.D. Smith, and C.D. Jorgensen. 1974. Western North American Naturalist 34(2): 105–120.

    Google Scholar 

  49. Roberts, S.L., J.W. van Wagtendonk, A.K. Miles, D.A. Kelt, and J.A. Lutz. 2008. Modeling the effects of fire severity and spatial complexity on small mammals in Yosemite National Park, CA. Fire Ecology 4(2): 83–104. doi: 10.4996/fireecology.0402083

    Article  Google Scholar 

  50. Rochester, C.J., C.S Brehme, D.R. Clark, D.C. Stokes, S.A. Hathaway, and R.N. Fisher. 2010a. Reptile and amphibian responses to large-scale wildfires in southern California. Journal of Herpetology 44: 333–351. doi: 10.1670/08-143.1

    Article  Google Scholar 

  51. Rochester, C.J., M.J. Mitrovich, D.R. Clark, M.B. Mendelsohn, D.C. Stokes, and R.N. Fisher. 2010b. Plant community responses to large-scale wildfires in southern California. US Geological Survey Data Summary prepared for San Diego Association of Governments, San Diego, California, USA.

    Google Scholar 

  52. Rood, J.P. 1966. Observations on the reproduction of Peromyscus in captivity. American Midland Naturalist. 76: 496–503. doi: 10.2307/2423101

    Article  Google Scholar 

  53. Sauvajot, R.M., M. Buechner, and C M. Schonewald. 1998. Patterns of human disturbance and response by small mammals and birds in chaparral near urban development. Urban Ecosystems 2: 279–297. doi: 10.1023/A:1009588723665

    Article  Google Scholar 

  54. Sawyer, J.O., and T. Keeler-Wolf. 1995. A manual of California vegetation. California Native Plant Society, Sacramento, California, USA.

    Google Scholar 

  55. Schwilk, D.W., and J.E. Keeley. 1998. Rodent populations after large wildfire in California chaparral and coastal sage scrub. Southwestern Naturalist 43: 480–483.

    Google Scholar 

  56. Shaffer, K.E., and W.F. Laudenslayer Jr. 2006. Fire and animal interactions. Pages 118–144 in: N.G. Sugihara, J.W. van Wagtendonk, K.E. Shaffer, J. Fites-Kaufman, and A.E. Thode, editors. Fire in California’s ecosystems. University of California Press, Berkeley, USA.

    Chapter  Google Scholar 

  57. Sikes, R.S., and W.L. Gannon. 2011. Guidelines of the American Society of Mammalogists for the use of wild mammals in research. Journal of Mammalogy: 92(1): 235–253. doi: 10.1644/10-MAMM-F-355.1

    Article  Google Scholar 

  58. Simberloff, D.S., and L.G. Abele. 1976. Island biogeography theory and conservation practice. Science 191: 285–286. doi: 10.1126/science.191.4224.285

    Article  CAS  Google Scholar 

  59. Slade, N.A., and S.M. Blair. 2000. An empirical test of using counts of individuals captured as indices of population size. Journal of Mammalogy: 81(4): 1035–1045. doi: 10.1644/1545-1542(2000)081<1035:AETOUC>2.0.CO;2

    Article  Google Scholar 

  60. Smith, E.P. 2002. BACI design. Pages 141–148 in: A.H. El-Shaarawi and W.W. Piegorsch, editors. Encyclopedia of environmetrics. John Wiley and Sons, New York, New York, USA.

    Google Scholar 

  61. Stark, H.E. 1963. Nesting habitats of the California vole, Microtus californicus, and microclimatic factors affecting its nest. Ecology 44: 663–669. doi: 10.2307/1933012

    Article  Google Scholar 

  62. Stewart-Oaten, A., W.M. Murdoch, and K.R. Parker. 1986. Environmental impact assessment: pseudoreplication in time? Ecology 67: 929–940. doi: 10.2307/1939815

    Article  Google Scholar 

  63. Tevis Jr., L. 1956. Effect of slash and burn on forest mice. Journal of Wildlife Management 20: 405–409. doi: 10.2307/3797152

    Article  Google Scholar 

  64. Underwood, A.J. 1994. On beyond BACI: sampling designs that might reliably detect environmental disturbances. Ecological Applications 4: 3–15. doi: 10.2307/1942110

    Article  Google Scholar 

  65. Vogl, R.J., and P. Schorr. 1972. Fire and manzanita chaparral in the San Jacinto Mountains, California. Ecology 53: 1179–1188. doi: 10.2307/1935432

    Article  Google Scholar 

  66. Wells, M.L., J.F. O’Leary, J. Franklin, J. Michaelson, and D.E. McKinsey. 2004. Variations in a regional fire regime related to vegetation type in San Diego County, CA (USA). Landscape Ecology 19:139–152. doi: 10.1023/B:LAND.0000021713.81489.a7

    Article  Google Scholar 

  67. Williams. D.F. 1986. Mammalian species of special concern in California. Report prepared for California Department of Fish and Game, Sacramento, USA. <>. Accessed 13 June 2011.

  68. Wirtz II, W.O., D. Hoekman, J.R. Muhm, and S.L. Souza. 1988. Post-fire rodent succession following prescribed fire in southern California chaparral. Pages 333–339 in: R Szaro, K. Severson, and D. Patton, editors. Management of amphibians, reptiles, and small mammals in northern America. USDA Forest Service General Technical Report RM-166. Rocky Mountain Forest and Range Experiment Station, Flagstaff, Arizona, USA.

    Google Scholar 

  69. Wright, H.A., and H.A. Bailey. 1982. Fire ecology: United States and southern Canada. John Wiley and Sons, New York, New York, USA.

    Google Scholar 

  70. Zedler, P.H., C.R. Gautier, and G.S. McMaster. 1983. Vegetation change in response to extreme events: the effect of a short interval between fires in California chaparral and coastal scrub. Ecology 64: 809–818. doi: 10.2307/1937204

    Article  Google Scholar 

Download references


The journal editor and two anonymous reviewers provided comments that improved this manuscript. This manuscript also benefitted from reviews by M.E. Mueller, D. Schwilk, J. Yee, and R. Klinger. We thank Dr. J. Yee for statistical advice. For site access, we thank all four land managers (T. Dillingham of California Department of Fish and Game; J. Stowers, B. Thompson, and J. Rundell of San Diego County Parks and Recreation Department; I. Kay of Elliott Chaparral Reserve; and J. Schlachter and J. Byergo of Bureau of Land Management). For their efforts in the field and in the office, we thank S. Hathaway, T. Matsuda, M. Mendelsohn, S. Schuster, D. Stokes, G. Turschak, and D. Wood. Funding for this project was provided in part by the San Diego MSCP Initiative, California Department of Fish and Game, The Nature Conservancy, United States Fish and Wildlife Services, United States Geological Survey, and the San Diego Association of Governments-TransNet Environmental Management Program. The use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the United States Geological Survey.

Author information



Corresponding authors

Correspondence to Cheryl S. Brehme or Denise R. Clark.

Rights and permissions

Open Access  This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.

The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

To view a copy of this licence, visit

Reprints and Permissions

About this article

Cite this article

Brehme, C.S., Clark, D.R., Rochester, C.J. et al. Wildfires Alter Rodent Community Structure Across Four Vegetation Types in Southern California, USA. fire ecol 7, 81–98 (2011).

Download citation


  • Chaetodipus
  • chaparral
  • coastal sage scrub
  • Dipodomys
  • grassland
  • Neotoma
  • Peromyscus
  • small mammals
  • wildfire
  • woodland