|
|
Seedling Dynamics
Background
Forests are dominated by long-lived organisms that often exhibit inertia in their demographic response to change. Thus, there is a clear need to consider indicators that are likely to be sensitive to change, such as reproductive biology and growth rates. Of particular note here is the tantalizing suggestion from the first phase of our research, that, in agreement with recent findings from eastern deciduous forests (Pacala et al. 1996), recruitment and death rates play a much greater role than growth rates in driving forest dynamics. This contradicts some of the basic assumptions of many forest dynamics models, and suggests that reproductive life history stages may be most sensitive to climatic change, and may ultimately drive forest change. While we have an established network of permanent plots for determining sapling and mature tree response to climatic variation, we lack critical data on seed production and seedling recruitment, and how they vary with microclimate and mature tree demographics. As noted by Bennett (1998), seed dispersal and subsequent seedling establishment may be the most critical determinants of the rate of forest response to climatic change. Our permanent demography plots provide an opportunity for quantifying dispersal curves for the dominant species under a range of physical settings and biotic backgrounds.
To provide an integrative framework for understanding current forest conditions, investigate future climate scenarios, and evaluate potential management options, we will incorporate these new results of field studies into our forest dynamics model, Zelig.
Current results
Our data show that Abies spp. (Abies concolor, A. magnifica) account for the vast majority of the species in the seedling plots. Pinus spp. (Pinus albicaulis, P. balfouriana, P. contorta, P. jeffreyi, P. lambertiana, P. monticola, P. ponderosa) are found at low frequencies, typically at the mid- and high-elevation plots. Other common species, particularly at low elevations, are Calocedrus decurrens and Quercus kelloggii.
Elevation is an important predictor of the total number of seedlings found in each plot. The results show elevation to be a good predictor of the average number of seedlings found at each plot and the model explained large amounts of variance in the data (Figure 1). Including the burned plots in the analysis reduces the strength of this relationship. Fir and pine seedling distributions considered separately also followed a negative exponential function. There was a greater amount of variance in the fir species reducing the fit of the model. Pine species closely followed an exponential decay pattern with elevation.
As expected, the seedling height class distributions are highly positively skewed and kurtotic (i.e., there are high frequencies of small seedlings with very few large-sized seedlings). This pattern holds when the plots are broken out by elevation classes (Figure 2). There are small, but statistically significant differences in the seedling height class distributions when the plots are grouped elevation (Pearson c 2 = 94.5, df = 10, p<0.0001). The differences among the elevation classes are primarily driven by differences in the relative number of seedlings in the smallest two height classes (0-10cm, and 11-25 cm). Compared to the low and mid elevation plots, the high elevation plots have relatively fewer seedlings in the smallest two height classes. However, the low elevation plots also have fewer seedlings in the smallest height class compared to the mid elevation plots, so there was no consistent trend towards greater relative numbers of larger seedlings with increasing elevation. Including the two burned plots (Upper and Lower Tharp) in this analysis did not substantially change these results.
To assess the reproductive output of the parent trees found in the plots, we calculated parent/offspring ratios (p/o). This analysis considers the number of seedlings observed versus the number of potential parent trees found in the plot. We did not consider the burned plots in this analysis. Parent/offspring ratios increase noticeably with increasing elevation, indicating that reproductive rates are lower in our high elevation plots (Figure 3).
To determine the influence of fire on seedling establishment rates we compared the number of seedlings found in two burned plots (Upper and Lower Tharps) versus two unburned plots roughly matched for elevation and community composition (Upper and Lower Log). The 13 ha headwater drainage containing the Tharp plots was prescribed burned by a combination of strip headfires and backing fires in October of 1990. There is a striking, but statistically non-significant increase in the number of seedlings in the burned plots (t = 2.79, df = 2, p = 0.11) (Figure 4), mostly in the form of additional fir establishment. It seems likely that these differences would become significant with a larger sample size. The seedlings found in the burned plots tend to be larger than in the unburned plots (Pearson c 2 = 392.1, df = 5, p<0.0001). It appears that there has been sufficient time for seedlings in the Tharp plots to grow into the larger height classes. These findings suggest that post-fire environments not only encourage greater seedling establishment rates, but also provide favorable growing conditions. A more extensive survey of seedling establishment and growth following fire will be needed to determine if these are general trends in the Sierra Nevada.
 |
Jon Keeley |
Comments
to: webmaster@werc.usgs.gov
URL: http://www.werc.usgs.gov/sngc/seedling_dyanmics.htm
This page last updated: Thursday, March 22, 2007