Riparian areas are three-dimensional zones of vegetation that lie at the interface of terrestrial and aquatic ecosystems, extending longitudinally along the stream, laterally from the stream to the uplands, and vertically from the groundwater to the overstory canopy. Riparian habitat acts as a transfer zone between the stream and uplands, whereby organisms, energy, and materials (organic and non organic) flow from one to the other reciprocally. The relationship between streams and riparian habitat should not be considered in a single direction where the stream influences the structure and composition of a riparian habitat. Riparian habitats have significant influences to the stream; structurally and compositionally—where the nutrient and sediment inputs to a stream from the riparian zone can directly affect in-stream species composition. Vegetation diversity is an important characteristic of riparian habitats because higher diversity creates a more resilient ecosystem that can rapidly and effectively respond to disturbances such as flooding. A riparian area with low diversity may have only a few response trajectories which could require significant time to regenerate critical habitat. Alternatively, a riparian area with high species (and structure) diversity will be able to respond to disturbances along numerous trajectories that can regenerate critical habitat much more rapidly. In order to initiate and maintain riparian areas with high species diversity, we must first understand the drivers contributing to diversity. The purpose of this paper is to identify what patterns of diversity occur along a stream network. I will identify the general trend of diversity along the stream network continuum, statistical differences between fundamental channel morphologies, patterns within these channel morphologies, and the “depth of effect” that each channel morphology has on riparian diversity.
The South Fork Trinity River, in northern California, was selected as the study site. It has a watershed size of approximately 2,428 square kilometers, and is the largest tributary of the Trinity River. The watershed terrain is predominately mountainous and forested, with only about 15 percent of the basin currently used as farmland, most of which occurs in the Hayfork Valley near the towns of Hayfork and Hyampom. Elevations within the watershed range from over 2,400 meters to its confluence with the main branch of the Trinity River at 135 meters. Mean gradients of along the network range from 0.0001 to just over 1.06. Valley widths along the stream network range from less than one meter in the headwaters to over two kilometers near Salyer, California. Geology in the western portion of the watershed consists of more erodible and unstable South Fork Mountain schist, Galice metasediments and Franciscan terrane, while the eastern part is more stable, consisting of Rattlesnake Creek and Hayfork terranes. The region was settled in the late 19th century and the local economy has been historically dominated by logging. By 1977, 52 percent of the watershed had been logged and 5,480 kilometers of road had been built.
Measuring the Stream Network
This study is designed to illuminate the patterns of species diversity along a stream network. In order to quantify these patterns a few definitions and measurements must be clarified. Firstly, the stream network continuum implies that there is a generally linear relationship between diversity and location within the watershed—where the highest diversities will occur in the lower reaches and the lowest diversity will occur in the highest reaches of the watershed. Secondly, this study provides statistical evidence supporting the hypothesis that there are significant differences in species diversity between fundamental channel morphologies—straight, bend and tributary junctions. Each of these channel morphologies was designated using standard hydrologic measurements. Thirdly, the influence of these channel morphologies varies in distance from the stream. In order to quantify this distance, the sampling design included 10 meter wide belts of vegetation plots at 0, 10, and 20 meters from the stream, named Zones A, B, and C, respectively.
The primary findings of this study include support of the generalized stream network continuum concept; the statistical differences in species diversity between channel morphologies; the inference of patterns of species diversity near tributary junctions; and the statistically delineated “depth of effect” that each channel morphology has on riparian diversity.
Species Diversity along the Stream Network Continuum
The first step for assessing the influence of the stream network continuum is to compare the species diversity to ranked locations within the watershed. Figure 3 through Figure 8 are scatter plots that represent this relationship within each of the three channel morphologies for tree and shrub diversities. The r², or square of the residuals after being fit with a linear trendline, identify any divergence from the generalized stream network continuum concept. In this case, the r² values quantify the ratio of data (species diversity) that can be explained by the watershed location.
To further assess the relationship between species diversity and watershed location, a Pearson Product-Moment correlation analysis was conducted for each channel morphology, in each zone and for trees and shrubs separately. Here you can see that within Zone A, for both trees and shrubs, there is sufficient correlation between species diversity and watershed location to reject the null hypothesis. Additionally, there is a correlation in Zone C in the Bend reaches for tree diversity. This relationship was unexpected and there is currently insufficient data to justify why it exists.
While the stream network continuum hypothesis is supported by this study, there are some additional relationships within the data that require deeper investigation. Below are scatter plots of the relationship between species diversity and watershed location for all channel morphologies in Zone A. While they all show a positive relationship, there are obvious differences in species diversity between the channel morphologies. The next stage in this study requires a statistical test of difference in species diversity between these channel morphologies.
Patterns of Species Diversity between Channel Morphologies
This second stage in the project is an assessment of whether the different stream morphologies have statistically significant differences in species diversity. Below are bar graphs representing the species diversity in the three channel morphologies. While it looks obvious that junction reaches have significantly higher diversities, a Z-test can be used to statistically test this assumption.
Upon conducting a Z-test, these differences were statistically verified. The figures below summarize the Z-test results. For tree species, there is a very high confidence interval for statistical differences between the junction reaches and both the bend and straight reaches. This occurs in all three Zones, except in Zone C when comparing the junction and straight reaches.
For shrub species, statistically, there are no differences in diversity between channel morphologies. However, while they are not significantly different, the shrub diversity is “more” different in Zone A than the other Zones.
Patterns of Species Diversity near Tributary Junctions
While it has now been shown that there is a statistically significant increase in species diversity in the areas near tributary junctions, it is important to also identify what patterns specifically occur in the areas surrounding these junctions. The first step in this process is to graph the range of diversities that occur in a gradient emanating from the junction location, along the stream corridor. The figures below represent mean diversities as a function of distance from the tributary junctions. These figures also show the maximum and minimum diversities to represent the overall range of situations.
While a graph of the diversity can be a qualitative descriptor of the situation, a more quantitative method for delineating the area of influence a tributary junction has on species diversity along the main channel. By adding to the graph, regression lines for species diversity along individual sections of the channel, the area of influence can become more definitive. The figure below shows a regression line for the entire span of diversity values and two additional regression lines for more refined areas nearest to the tributary junction. The approximate areas where the three regression lines and the mean diversity line cross could be considered the area of negative influence. This area is defined by being the region where mean diversity is less than the overall trend of diversity. In both upstream and downstream directions, the decrease in tree diversity reaches to approximately 40 meters from the tributary junction.
The figure above shows a regression line for the entire span of diversity values and two additional regression lines for more refined areas nearest to the tributary junction. The approximate areas where the three regression lines and the mean diversity line cross could be considered the area of negative influence. In the case of shrubs, the situation is more complicated. The area upstream shows a significant spike in diversity, reaching to approximately 50 meters from the tributary junction. The area downstream shows a slight drop in diversity, reaching to approximately 80 meters from the tributary junction.
Species Diversity “Depth of effect”
The data used in this study can also be used to help identify how far the stream has an effect on species diversity. By initially graphing the data, it can be seen that tree diversity in junction reaches is significantly higher in Zone A than in Zone B. While less significant, this relationship can also be seen for shrub species diversity.
In order to identify this relationship statistically, a Z-test was conducted to assess the significance. For tree species, only in the junction reaches is there a statistically significant difference in species diversity between Zones A and B. While there is not a statistically significant difference between Zones B and C, it is larger than any zones in the straight and bend reaches.
For shrub species, it is again only in the junction reaches where there exists a statistically significant difference in species diversity between Zones A and B (Figure 24). While not a statistically significant, there is a small difference between Zones A and B for bend reaches.
An additional method used to identify the “depth effect” is by conducting regression analyses that compare Zone A to Zone B and Zone B to Zone C for the entire stream corridor. Figures 25 through 30 represent scatter plots and regression lines comparing these zones. When comparing the equation for each regression line, the steepest lines imply the greatest correlation between Zones. When comparing Zones A and C for both trees and shrubs, the lines have the lowest slope, representing a lack of relationship between species diversity in Zone A and Zone C. For tree species diversity, the steepest slopes, and therefore the strongest relationships are when comparing Zones B and C (y=0.3716), and Zones A and B (y=0.2444). For shrub species diversity, the steepest slope, and therefore the strongest relationship is when comparing Zones B and C (y=0.293). With these results, it can be concluded that the stream most influences tree and shrub diversity within a 20-meter depth.
The evidence shown in this report provides statistically-supported patterns of riparian species diversity within stream networks. The data identify (1) a strong trend of decreasing species diversity along the stream network continuum from the lower reaches to the top of the watershed; (2) statistically significant increases in species diversity in reaches with tributary junctions; (3) patterns of species diversity in the areas surrounding tributary junctions; and (4) two potential methods for quantitatively identifying the “depth of effect” that each channel morphology has on riparian diversity.