Fire History and Stand Structure of a central Nevada Pinyon-Juniper Woodland EXECUTIVE SUMMARY AND FINAL REPORT A Report to the Bureau of Land Management, Nevada State Office September, 2006 Peter J. Weisberg John M. Bauer Natural Resources and Environmental Science University of Nevada Reno 1000 Valley Road, M.S. 186 Reno NV 89557 UNR-BLM Cooperative Agreement: FAA010017 Through the Great Basin Cooperative Ecosystems Study Unit (CESU)
EXECUTIVE SUMMARY Pinyon-juniper (PJ) expansion in the Great Basin has been ongoing over much of the Holocene period, but has accelerated over the past century or so. Landscape restoration requires a long-term view toward the range of historic variability for PJ woodland dynamics, as well as a sound understanding of causal factors for recent landscape change. Several factors have been implicated as causes for pinyon-juniper woodland expansion, including fire exclusion due to grazing, direct effects of grazing on plant competitive interactions, active fire suppression, climate change, response to post-settlement tree harvesting, and carbon dioxide enrichment. These factors have likely interacted in complex ways that vary with geography and location. Fire history has seldom been studied for Great Basin landscapes, for pinyon-juniper forest types, or for sagebrush vegetation which pinyon-juniper has recently invaded. Yet such studies could help us understand whether woodland expansion and infilling result from 19 th and 20 th Century fire exclusion. A quantification of the historic range of variability of fire can also inform land managers using fire for restoration of PJ systems. Our initial intention was to compare fire history and woodland structure for areas with varying levels of historical grazing. We had planned to locate relictual landscapes that did not endure the early period of unmitigated grazing. When this proved infeasible, we broadened the study to encompass a generalized description of historical fire regimes for a representative area of PJ woodland. The study focused on the following questions: What was fire s role in shaping the pre-settlement woodland and landscape structure? How has fire regime varied across space, considering effects of topography, landforms and adjacent vegetation types? How has fire regime varied over time, according to variability in climate and land use practices? The study employed multiple fire history methods, including novel methods that should serve as a model for other tree-ring-based, PJ fire history studies conducted in areas without good fire scar recorder tree species such as ponderosa or Jeffrey pine. While the study area was one 4600-acre watershed on Forest Service land in the Shoshone Range, results should be broadly generalizable to large areas of PJ woodland in central Nevada. The study area, Barrett Canyon, was chosen because (1) it includes a wide range of PJ stand structures and age classes; (2) it includes fire-scarred pinyon trees; and (3) it allowed us to collaborate and pool field data with Robin Tausch (USFS-RMRS) and so greatly expand the scope of the study. Principal results are listed below, divided into the following categories: Methodological Issues; Historical and Current Fire Regime; Changes in Woodland Structure; Management Implications and Recommendations. For more detailed information as to methodology and results, including relevant tables, maps and graphical figures, refer to Bauer (2006). The Py et al. (2006) study cited in the Research Products section was also 2
funded in the context of this Cooperative Agreement. This latter study was focused on dendrochronological methodology for assessing pinyon growth responses to disturbance, and is not summarized here because of its limited relevance to land managers. Methodological Issues for PJ Fire History Fires in PJ woodland leave dateable, basal fire scars on single-needle pinyon pine trees, although such evidence is generally sparse. An investigation of recent fires suggests that such scars occur mainly on the edges of areas that have burned in high-severity crown fire. A total of 94 fire scarred-trees in the Barrett Canyon study area revealed 44 cross-dated fire years spanning from 1445 to 1947 AD. PJ fire history is best reconstructed from multiple lines of evidence, including relating maximum stand ages of woodland patches to nearby fire scar dates, and using a reconstructed time-since-fire map to develop statistical models of fire cycle. Historical and Current Fire Regime Prior to settlement, the PJ woodland experienced a high severity fire regime with a fire cycle of approximately 400 years. The fire cycle estimates the fire rotation, which indicates the average number of years required to burn an area the size of the study area, where some portions of the landscape may burn multiple times and others not at all. Historical fire regime was dominated by high-severity fire; there is no support for spreading, low-severity surface fires. Of 90 plots sampled, a total of 71 (79%) showed direct evidence of high-severity fire sometime over the past 700 years. Some of these plots represent old-growth woodland, while others represent younger woodland. A total of 9 plots (10%) are old-growth with no evidence of having burned at any time in the past. Five plots (6%) represent expansion woodland where trees have established in former shrubland, with no evidence of historical fire or previous tree dominance. Only 2 plots (2%) showed evidence of mixed-severity fire, and no plots showed evidence of spreading surface fire. (Figure 1) Over the period of record, pre-settlement woodland fires were generally small (< 35 ha). More frequent grassland and shrubland fires were recorded by fire scars on trees near valley floors. Most fires that burned PJ appear to have originated in shrub-dominated plant communities outside the PJ belt. Of 55 total fires reconstructed, 22 (40%) originated in the more productive valley bottom (sagebrush grassland), with only slight penetration into the PJ woodland. A total of 12 (22%) originated in the valley bottom and penetrated at least 100 meters upslope through the woodland. A total of 9 (15%) originated above the woodland, mainly in productive mountain big sagebrush communities, and burned down into the woodland belt. For 3 fires (5%), there was insufficient evidence to determine area of origin. Only 9 reconstructed fires (15%) appear to have originated in the woodland belt. 3
Topography strongly influenced the presettlement fire cycle. Fire cycles decreased (i.e. fires were more frequent) at higher elevations and slope positions, and on sites with more northeasterly aspects, indicating fires were more frequent on more mesic sites (Figure 2) Prior to 1890, fires were more likely during extreme drought years. The probability of fire during a given year was not significantly influenced by antecedent conditions (e.g. moist years producing heavy fuel build-up). Since approximately 1880, there have been no stand replacing fires in the watershed. There have been 3 fires during that period which burned the valley bottom and scarred trees at the lower woodland ecotone. The last of these fires occurred in 1947. (Figure 3) Broad, century-scale changes in fire regime parallel broad, century-scale changes in Great Basin climate and land use. Fires were more frequent during the period from 1300 1570 (187 year fire cycle), less frequent from 1570 1880 (427 year fire cycle), and virtually absent during the period since 1880. The first fire cycle division slightly lags the trend toward cooler, moister conditions starting in the late 15 th Century. The more recent fire cycle division may be associated with both climatic and anthropogenic changes. (Figure 4) Changes in Woodland Structure Conifer stem densities have increased dramatically in most Barrett Canyon PJ woodland sites since approximately 1880. However, there are many caveats underlying this interpretation (see Bauer 2006). This also describes a process of infilling in areas where woodland has long been established, not of expansion into sites that have long been dominated by shrub or grassland vegetation types. The latter may also be occurring in the study area, but was not a question that was directly addressed by this study. Increased tree density in established woodlands is not attributable to the absence of 20 th century fire. This inference follows from the absence of any evidence for recurring, low-severity surface fire in the fire history record. Presettlement old-growth (OG) woodlands were not confined to small rocky outcrops, but formed continuous tracts as large as 100 hectares. (Figure 5) Management Implications and Recommendations Not all younger woodlands are expansion woodlands representing new PJ establishment in areas that were formerly dominated by other vegetation types. Some younger woodlands have established following high-severity wildfire. This is an important distinction for ecological restoration activities. Some old-growth woodlands show no evidence of having burned at any time over the past centuries. Fire may not be characteristic of all PJ woodlands. The fire regime of certain PJ communities growing on unproductive substrates, or 4
protected by rock or other geologic barriers to fire spread, may be one of no fire or isolated single-tree events from lightning strikes. Woodland fire regimes vary according to topography, landform and fire occurrence in adjacent vegetation types. There should not be one-size-fits-all management prescriptions. PJ woodlands have burned infrequently, with active crown fires occurring only under extreme conditions. There seems to be little ecological justification for reintroducing fire or other disturbance to areas of persistent (i.e. historic) woodland. There is little justification for attempting to introduce low-severity fire to PJ woodlands, or for emulating low-severity fire through thinning from below management actions. This is true at least from the perspective of ecological restoration for emulating historic range of variability. Low-severity surface fires have likely not played an important role in these systems. Shrub-dominated communities adjacent to PJ woodlands, including areas now dominated by trees, are likely to have burned more frequently. This investigation was not able to develop robust estimates of fire frequency for shrubdominated areas adjacent to the PJ woodland. It is essential that restoration plans consider the variability in presettlement woodland structure and the extent of old-growth woodlands. This includes other factors about which we know relatively little, such as historical influences of Native Americans and early post-settlement tree cutting associated with charcoal production for the mining industry. Research Products Bauer, J.M. 2006. Fire history and stand structure of a central Nevada pinyon-juniper woodland. M.S. thesis, Dept. Natural Resources and Env. Sci., Univ. Nevada, Reno. Available online at: http://www.cabnr.unr.edu/weisberg/john-research.htm. Py, C., J. Bauer, F. Biondi, and P.J. Weisberg. 2006. Radial growth responses of singleleaf pinyon (Pinus monophylla) to wildfire. Dendrochronologia 24: 39-46. Acknowledgement The United States Bureau of Land Management (Nevada State Office) provided funding support through the Great Basin Cooperative Ecosystem Studies Unit (Cooperative Agreement # FAA010017). Additional funding was provided by the Nevada Agricultural Experimental Station. Franco Biondi (UNR) provided tree-ring laboratory support, including reference chronologies that were critical for cross-dating dendrochronological samples. This work would not have been possible without the collaboration of Robin Tausch of the United States Forest Service (Rocky Mountain Research Station), who participated in a cooperative research effort and provided additional field crews. Many other individuals contributed to this effort; see Bauer (2006) for a more complete Acknowledgement. 5
Supporting Figures and Tables Fig. 1. Fire severity classification for all sample plots, n = 90. Plot Classification Intensive Extensive Total High Severity 22 49 71 Pinyon Old Growth - no fire 2 4 6 Mahogany Old Growth 2 1 3 Expansion 2 3 5 Mixed Severity 2 0 2 Low Severity 0 0 0 6
Fig. 2. Univariate Chi-square tests for fire cycle as predicted by topographic variables, using the log-rank test. Analysis period was 1570 1880 A.D., with 44 failures, 43 censored points. A positive effect indicates an increase in the fire cycle estimate. Covariate Chi-sq Pr > Chi-sq Effect Aspect 8.24 0.004 - Slope 5.59 0.018 + Rock Cover 7.20 0.007 + Elevation 3.71 0.054 - Dist. to Valley 8.28 0.004 - Relative Distance 5.23 0.022 - Slope Position 1.88 0.170 - Solar Radiation 4.56 0.033 + Fig. 3. Fire year temporal distribution, showing minimal 20 th Century fires.. n = 94 fire scars. 7
Fig. 4. Averaged Palmer Drought Severity Index, using 50-year moving window average, for point 58, -117.5 E, 40.0 N (Cook et al. 2004), with fire cycle estimates for Barrett Canyon watershed, and generalized late Holocene divisions. Period A was generally dry and cool. Period B is considered the Little Ice Age, with cooler and wetter conditions. Period C, starting around 1850, was warmer, and coincident with Euro- American settlement (from Tausch 1999b). FC = fire cycle FC = 187 yrs FC = 427 yrs FC > 1000 yrs 0.5 Averaged PDSI 0.25 A 0 1400 1500 1600 1700 1800 1900 2000-0.25 B Calendar Year -0.5 C -0.75-1 Fig. 5. Plot stand ages, with size of circle proportional to age of plot. Median age = 461 years, maximum age = 751 years, minimum age = 84 years. n = 90 plots, which includes plots classified as mahogany old growth stands. 8