Division of Statistics, University of California, Davis CA
Download the full report (Adobe Acrobat PDF - 935 KB)
Note: PDF replaced 8/6/18 with updated version that corrects fomatting issues with several figures and formulas.
Additional key words: Quercus douglasii, fire, grazing, mortality, canopy gap
Blue oak (Quercus douglasii) is an endemic California oak that occurs throughout the foothills of the Coast Ranges and the western slope of the Sierra Nevada. There is a widespread belief that blue oak may not be regenerating well over much of its extensive range. Past research projects have demonstrated that a number of factors can affect the growth and survival of blue oak seedlings and saplings, but have not led to a consensus about the overall status of blue oak regeneration.
This study was undertaken to assess the status of blue oak regeneration at the stand level and to determine how environmental and management factors influence blue oak sapling recruitment. A major objective of this research was to determine how management practices (such as grazing and clearing), stand characteristics (such as tree canopy and understory vegetation), and site factors (such as slope, aspect, soil type, and precipitation) affect the likelihood of blue oak sapling recruitment. This information might then be used to help develop management methods that would favor natural regeneration by blue oak.
We assessed blue oak sapling recruitment, tree mortality, and site variables in plots located at 15 study locations distributed throughout the range of blue oak. Study locations were selected on the basis of having blue oak as the dominant canopy species, and having available information on the history of grazing, fire, clearing, and other management practices over the preceding 30 years. The selection of locations and plots was conducted without prior knowledge of the amount of sapling recruitment present. A random-origin systematic grid was used to locate 100 sample plots at each location over an area of about 61 ha (150 acres). Circular plots with a 16 m radius (0.08 ha = 0.2 acre) were established at approximately 80 m 100 m centers. For this study, saplings were defined to have a basal diameter of 1 cm or greater, and a diameter at 1.4 m (dbh) of 3 cm or less. We defined three size subclasses within the sapling size class, and noted the size subclass and position relative to overstory canopy of each sapling.
The effects of history and environmental variables on sapling recruitment were analyzed using logistic regression. Only six locations had high enough frequencies of recruitment to permit us to construct logistic regression models for factors associated with recruitment within locations. Due to inhomogeneity between locations, only summarized location data could be used to construct models to compare the effects of factors between the 15 study locations. All of the logistic regression analyses were complicated by the fact that many of the environmental and management variables were highly correlated with each other. Ecological observations recorded at each location were used to help interpret the results of the statistical models.
Overall, 15.3% of the plots contained saplings. We found moderate numbers of saplings at four locations, no saplings at all at another four locations, and few to very few saplings at the remaining seven locations. The majority of all saplings were shorter than browse line (1.4 m). Most of the saplings we observed arose from seedlings rather than as sprouts from cut stumps, but sprout-origin saplings outnumbered seedling-origin saplings at one location. All locations had some sprout origin trees, but the incidence of sprout-origin trees varied widely between locations.
We observed natural mortality of mature blue oak trees at all locations, but estimated mortality rates varied between locations. Based on the balance between tree mortality and sapling recruitment at the plot level, 13 of the 15 study locations appear to be experiencing a net loss in blue oak density and canopy cover. Only two locations had more plots which were likely to gain blue oak density and canopy cover, due to sapling recruitment, than plots which had lost density and canopy cover due to mortality.
Saplings were more likely to be found in the open than under canopy, but saplings were rarely found in plots lacking blue oak canopy cover. High levels and low levels of tree canopy cover were generally less favorable for sapling recruitment than intermediate levels. Seedling-origin saplings were more likely to be found in plots with recent (30-42 years) tree cutting or other types of canopy gaps than in plots with no recent gaps. However, saplings rarely occurred in old fields and other very old clearings.
We observed that most locations with little or no blue oak sapling recruitment also had little or no regeneration of other woody overstory and understory species. Shrub presence was positively correlated with blue oak sapling recruitment. It appeared that the occurrence of other woody understory plants in plots with blue oak saplings was due to the fact that these are related outcomes which are favored by the same conditions.
Across all locations, intense browsing was negatively associated with sapling recruitment. At locations that had been grazed by livestock, saplings were more likely to occur in areas that were less heavily used by livestock, such as on steep slopes or among rock outcrops. At the one location that had both moderate levels of recruitment and variation in grazing history, sapling recruitment was significantly more likely to be found in a field that had been nongrazed for 20 years than in adjacent grazed fields.
Infrequent fires appear to have either no effect or a slight positive effect on sapling recruitment and growth. At the one study location that had both recruitment and many fires in the past 30 years, portions of the study area which burned repeatedly had fewer saplings than areas which had burned only once or had not burned.
In general, recruitment tended to be more common at more mesic locations. At xeric locations, recruitment tended to occur in more mesic plots. However, at relatively mesic locations, canopy species other than blue oak often dominated the most mesic plots, and blue oak saplings were more likely to occur in somewhat xeric plots.
We believe that most of the blue oak sapling recruitment we observed developed from seedling advance regeneration in the form of small persistent seedlings. Gaps in the overstory tend to favor the recruitment of saplings from seedling advance regeneration. Pioneer colonization of open areas by blue oak is rare under current range conditions, and saplings are seldom recruited under a dense canopy.
Regeneration can be inhibited by factors that deplete the reserve of persistent seedlings in the understory, inhibit the transition from seedling to sapling, or prevent saplings from advancing to the tree stage. Since the process of sapling recruitment can be arrested at different stages, variables related to the recent past history of a site are often better predictors of recruitment than are current site conditions. Sapling recruitment may be constrained by a number of different factors at a location, so that relieving a single constraining factor may have little or no impact on the rate of sapling recruitment.
Funding for this project was provided by the California Department of Forestry and Fire Protection through contract 8CA17358.