The Influence of Tidal Channels on the Distribution of
Salt Marsh Plant Species in Petaluma Marsh, CA, USA

Eric W. Sanderson1, Susan L. Ustin1 & Theodore C. Foin2

1Department of Land, Air and Water Resources
University of California, Davis, CA 95616, USA

2Division of Environmental Studies
University of California, Davis, CA 95616, USA

Key words: California salt marsh, channel order, vegetation patterns, wetland restoration.

Abstract

Tidal channels influence the distribution and composition of salt marsh vegetation in a San Francisco Bay salt marsh. Two channel networks in the Petaluma Marsh, Sonoma County, CA, were mapped and characterized using global positioning and geographic information systems. Plant species abundance was sampled on transects placed perpendicular to and extending away from the channel banks. The vegetation showed significant increases in species richness along channel banks and larger areas of effect which increased approximately linearly with channel size. Composition of species assemblages varies with distance from the channel bank and channel size. These results demonstrate that salt marsh plant assemblages, comprised of both major and minor species, are distributed with respect to the channel network in Petaluma Marsh.

Introduction

Tidal channels are nearly ubiquitous features of salt marsh ecosystems. In the marsh interior, channels control tidal flow into the marsh and determine how much flow reaches various locations in the marsh (Pestrong, 1972; Chapman, 1960) These channels depend on tides for their origin and maintenance and are little influenced by runoff from terrestrial sources (Atwater et al, 1980; Leopold et al, 1993). Channels within the tidal network vary in size, but are typically larger near the tidal source and much smaller near their headward end (Redfield, 1972; Collins et al, 1986).

Many earlier studies have noted that certain plants are associated with the sides of channels or growing on the small levees along the channels (De Groot, 1927; Purer, 1942; Marshall, 1948; Hinde, 1954; Chapman, 1960; Vogl, 1966; Redfield, 1972; Balling & Resh, 1983; Collins & Resh, 1985; Adam, 1990); more recently these species have been investigated as the "peripheral halophyte zone" (Wayne & Parker, 1994; Wayne & Parker, 1996). A recent remote sensing study revealed that the tidal channels had an important influence on the patterns of canopy water content across the marsh landscape, even for very small channels, which, though less than one meter across, influenced the canopy reflectance of pixels 400 m2 in area (Sanderson et al, 1998).

This paper describes field studies which quantified the relationship between the geography of the tidal channel network and plant species distributions in a San Francisco Bay salt marsh. We mapped two typical channel networks at interior and exterior locations in Petaluma Marsh, CA, a large marsh with an extensive channel network. We placed transects with respect to channel order (a surrogate for channel size) and channel origin (naturally formed vs. constructed for mosquito control) to quantify the distribution of plant species at these two sites.

Methods

Site Description

The observations for this study were made in Petaluma Marsh, a 515 ha tidal salt marsh along the Petaluma River, about 8 km from the mouth of the river in San Pablo Bay, Sonoma County, CA (38° 12’ 20" N, 122° 32’ 32" W). This marsh is part of a semi-contiguous network of tidal wetlands that stretch along the northern shore of San Pablo Bay and part of the greater San Francisco Bay Estuary (Atwater et al. 1980). Petaluma Marsh is constrained by levees along the Petaluma River to the east and south, and by rising topography on its western margin.

The vegetation of Petaluma Marsh is dominated by Salicornia virginica, a succulent halophyte, which covers approximately 77% of the marsh surface (Sanderson et al. 1998). Two other species grow in zones parallel to the Petaluma River: Spartina foliosa, a halophytic cordgrass that grows in a continuous belt below mean high tide, (Mahall & Park, 1976a), and Scirpus robustus, a meso-halophytic bulrush, which grows in large roughly elliptical patches at higher elevations above Spartina (Pearcy & Ustin, 1984; Ustin et al. 1982; Whitlow, 1982; Zhang et al. 1997). A suite of other species grow along the banks of natural channels within the marsh, some of which are restricted to the salt marsh, while others are found elsewhere. Typically halophytic species include Frankenia salina I.M. Johnston, Jaumea carnosa A. Gray, Grindelia stricta DC., Atriplex patula L., Cuscuta salina Engelm. and Distichlis spicata E. Greene. Baccharis pilularis DC and Achillea millefolium L. and two invasive species, Lepidium latifolium L. and Rumex crispus L., were also observed in the marsh, but are not restricted to salt marsh habitats. Through the remainder of the paper, each species will be referred to by its generic name.

Two sites were established in locations with known differences in nutrient cycling and species distribution (Zhang et al. 1997) The first site (hereafter the "River Site") lies along the Petaluma River, on the eastern margin of the marsh, approximately 0.5 km south of Lakeville, CA (Figure 1). It contains a large, central channel network and parts of three other channel networks, which were used to delineate the site boundaries. The Petaluma River (the sixth order channel) borders the eastern side of the site, and Tule Slough (a fifth order channel) borders the southern edge.

The second site (hereafter the "Pond Site") is located near the geographic center of the marsh where several large ponds, or salt pans, are found. A central channel system of natural and constructed channels which winds between the ponds was chosen for study. Adjacent channels and the pond boundaries were mapped. The site is bordered by Tule Slough on its northern margin and by arbitrarily chosen channels on the eastern, southern and western margins. The River and Pond Sites are similar in overall area.

Constructed channels were dug in the 1960s and 1970s and to control production of mosquitoes in Petaluma Marsh (Marin/Sonoma Mosquito and Vector Control District, 1997; Wolfe, 1996) These channels are long and straight, branching at right angles, in contrast to the sinuous character of the naturally formed channels. Most constructed channels were placed near the headward end of the drainage systems as first and second order channels. Several constructed channels cross natural drainage divides between otherwise unconnected channel networks. These channels have not been maintained since the early 1980s and tend to carry little to no tidal flow (J. Collins, personal communication).
 

Channel Mapping

The channel networks and pond boundaries were mapped using Global Positioning Systems (GPS) (Trimble ProXL System, Trimble Navigation Limited, Sunnyvale, CA) (Figure 2). Position information was collected 1 position/second as we walked along the channels (approximately 1 position/meter). Concurrently another researcher mapped the channels by hand to insure that all junctions were accounted for and to aid the digitizing process.

After collection, the GPS data were differentially corrected with base station data from Sacramento, CA (90 km from the site) or Sunnyvale, CA (60 km away) to an estimated accuracy of approximately 5 m. The corrected positions were entered as points into geographic information system (GIS) coverages in ARC/INFO (Environmental Systems Research Institute, Inc., Redlands, CA). Arcs representing the channels were digitized by connecting points, guided by maps made in the field. All geographic data was projected to the NAD-27 CONUS datum in the Universal Transverse Mercator (UTM) coordinate system (Zone 10 South).

Channel order and channel type was assigned to each channel length by examination. Channel order was determined according to the method of Strahler (1964) once the channel networks were mapped. Channel type, whether constructed or natural, was determined by examining the sinuosity of the channels and supplemented by field notes. Ponds were arbitrarily given the designation, zero order.

Two types of first order natural channel were recognized at the River Site: channels which flow directly into the Petaluma River along the eastern edge of the site ("exterior" first order channels), and channels which flow into second or third order channels in the interior of the marsh ("interior" first order channels). The exterior channels are created by erosion due to tidal action causing short, straight channels through predominantly inorganic sediments. Interior first order channels are typically longer and more sinuous, dissecting sediments which are mostly organic in origin (J. Collins, personal communication).

The physical size of channels was measured by channel order and type at both sites. For each channel order and type, three randomly placed cross-sections were determined by measuring the width of the channel and depth at five to ten evenly spaced points along the cross-section. The area of these cross-sections was calculated using trapezoids to approximate the areas between each depth measurement. For the fifth and sixth order channels, the channel width was measured by GPS and the depth with a marked and weighted line. The uncertainty of the fifth and sixth order depth measurements was approximately 0.5 m; the uncertainty of fifth and sixth order width measurements was approximately 5 m. All other depth and width measurements had an uncertainty of approximately 5 cm.

Natural channels of first through fifth order, and constructed channels of first and second order, were observed at both sites. Sixth order channel observations were only made at the River Site, along the Petaluma River. One third order constructed channel was observed at the Pond Site.
 

Vegetation Sampling

Using the GIS coverage of the channel network, random locations along the channels were determined for vegetation sampling, stratified by channel order and type. A random number table was used to determine which side of the channel to sample. Transects were oriented perpendicularly to the channel direction at that point and no restriction was made to find isolated channel points, though in general channel density was low enough that there was little multiple channel influence on any given transect. For each channel type and order, three transects were measured (a total of 27 transects each site).

For each transect, thirty 0.25 m2 square quadrats were measured, the first quadrat laid immediately alongside the channel, the next nineteen quadrats laid every 0.5 meter (i.e. adjacent to the previous quadrat), and the remaining ten quadrats laid every one meter apart (resulting in thirty observations over twenty meters). For each quadrat, plant cover was estimated according to Braun-Blanquet cover class system (Mueller-Dombois & Ellenberg, 1974) for each species present, the standing height was measured to the nearest 5 cm, and the tallest species noted. Plant species identifications were based on Hickman (1993) and Mason (1957). All plant names follow the nomenclature of Hickman (1993). Driftwood or other tidally deposited materials present in the quadrat were also noted as wrack. A total of 900 quadrats were measured at the River Site , 810 quadrats at the Pond Site. All vegetation observations were made between August 23 and September 11, 1996.

The positions of the starting and end quadrats were measured by GPS. For each starting and ending point, at least 180 GPS positions were collected to provide a precision of approximately 1 m for the average positions. The center points of the intervening quadrats were calculated using trigonometry based on the starting and ending locations. Slight discrepancies between the channel map and the transect locations were accommodated by making adjustments to the transect locations; all adjustments were less than 2 meters. The final estimated precision of the quadrat locations was 3 meters or less.

Data AnalysisLength and drainage density of the channel networks were analyzed by examination of the attribute tables of the GIS coverages after establishing topology. These data and the cross-sectional areas, widths and depths were summarized by channel type and order and study site in Table 1.

Quadrat cover values were translated from their cover class to the midpoint of each class (Barbour et al. 1987). Plant cover values by species were plotted against distance for each transect. Species richness measures were calculated for each transect based on the species data (not clusters) and subjected to analysis of variance by channel order and type.

Species richness is the number of species observed along the transect, without regard to abundance. Two-tailed t tests were used to compare the mean transect species richness between sites for each channel order.

Species richness near channels was compared to species richness away from channels by taking the difference between the number of species observed in the first 10 meters and the number of species observed in the last 10 m. Observations in the first 10 meters were subsampled to match the density of sampling in the last 10 meters (1 observation/meter). These differences were compared to a null model of no difference using Student’s t-test.

Initial examination of species data plotted against distance indicated that certain groups of species with defined amounts of cover re-occurred in the transects. To capture these patterns, we performed cluster analyses on the vegetation data. Field notes and examination of the transect data indicated sufficient differences in species associations at the two sites to warrant separate clustering. A dissimilarity matrix was calculated for all the quadrats at each site using the chord distance method (Ludwig & Reynolds, 1988) These dissimilarity data were used to construct clusters of quadrats based on their species cover values, using the average linkage method (Ludwig & Reynolds, 1988). Seven clusters represented 99% of the quadrats at each site; several clusters with low representation among the quadrats were dropped from the analysis.

Plots of the presence of each cluster by distance and channel order were prepared to show the distribution of clusters with distance from the channel. Clusters at each site that formed the majority of sampled units were dominated by Salicornia and were considered the "background" vegetation type (Clusters I and I’ at the River and Pond Sites, respectively, Table 2). Salicornia is generally agreed to be the salt marsh dominant in salt marshes of San Francisco Bay based on its high cover (Josselyn, 1983). The distance to background was estimated for each transect as the length of transect from the beginning to a continuous sampling over 3 m of background vegetation. The total area of marsh influenced at each site was estimated by multiplying the mean distance to background by channel length for each channel type and order and summing (Table 3.)

Analysis of variance was conducted on the distance to background by channel type and order. Comparisons were conducted to test whether the type of channels (natural vs. constructed), the site and channel order among natural channels significantly influenced the distance to background. Comparisons of the means were conducted with two-tailed t tests.

Mean distances to background by site, channel type and order were plotted against the logarithm of the corresponding channel cross-sectional areas and trendlines were drawn separately for small (first and second) order and large (third through sixth) order channels. Insufficient data points (3 observations for each channel order and type) prevented statistically significant regression coefficients of these trends.

All statistical tests were conducted using SAS (SAS Institute, Cary, NC).

Results

Morphological characteristics of the channel networksAt the River Site, 4,607 meters of tidal channel were mapped within an area of 109,611 m2 (10.96 ha), for an overall drainage density of 0.042 m channel m-2 marsh area (Table 1). The Pond Site was larger in area (18.18 ha), with more channels overall (5,327 m), but with a lower drainage density (0.029 m channel m-2 area). Constructed channels comprised 26.6% of the total channel length at the River Site, compared to 37.2% at the Pond Site. The smallest channel observed was a first order natural channel at the River Site, 15 cm across and 6 cm deep. The largest channel was the Petaluma River, 117 meters wide and 4 meters deep.

Average width, depth and cross-sectional area of channel varied with channel type, order and site. Constructed channels at both sites were similar in size, but smaller than natural channels of the same order, with areas ranging from 0.05-0.11 m2. First order "exterior" channels were narrower but deeper than first order "interior" channels at the River Site, leading to a larger cross-sectional area. First, second and fifth order natural channels at both sites were roughly the same size; however third and fourth order channels at the Pond Site were much wider and somewhat deeper than the corresponding channels at the River Site. At the River Site, cross-sectional area increased approximately one order of magnitude for every increase in channel order. At the Pond Site, channel area varied as a step function with channel order: first and second order channels are similar, third and fourth channels are similar and much larger, fifth order channels were larger still. These site differences probably indicate differences in the geomorphological dynamics at different positions in the overall channel network.

Seventeen salt pans or ponds were mapped at the Pond Site, varying in area from 1 m2 to 3752 m2. They occupied 7% of the total area at the Pond Site. Several small potholes or depressions, which may be incipient ponds, were noted at both sites, but not mapped.
 

Influence of tidal channels on species richness

Natural channels had a significant influence on the vegetation within 20 meters of the channel. More plant species are found within the first 10 meters from natural channels than the second 10 meters (one-tailed t test, p < 0.001) (Figure 3). These effects were also reflected in patterns of species richness. There is significant variation in species richness across channel orders (ANOVA, p < 0.001) and between sites (ANOVA, p < 0.008), though two-tailed t tests of the means indicate significant differences only for first and fifth order channels between sites at the 0.05 level. (Figure 3). Species richness increased from first through fifth order channels at the River Site, then dropped for the sixth order channel. Species richness was similar for second through fifth order channels at the Pond Site.
 

Cluster Analysis

Twelve plant species were observed at the River Site and fourteen at the Pond Site even though less area was sampled (830 quadrats at the Pond Site; 900 quadrats at the River Site). Clustering of the quadrats using the dissimilarity matrices indicated that these species occur in regular assemblages at each site. Seven species clusters were found for each site; each set of seven clusters represented more than 99% of the quadrats at their respective sites.

Important similarities were found between species assemblages at the two sites (Table 2). Both sites had Salicornia-dominated clusters (I and I') which comprised the majority of quadrats observed at each site (68.4% River; 70.1% Pond). These clusters were defined as the "background" at their respective sites and used to estimate the spatial extent of channel influence (below). Both sites also had clusters dominated by Frankenia (H and H'), Spartina (A and A') and Jaumea (J and J'). The site specific differences between these common clusters (except for A and A') depended largely on the presence of Cuscuta, a parasitic plant which was commonly observed at the Pond Site, but observed in only one quadrat at the River Site.

Several other clusters differed in composition between sites, but appeared to have similar physiognomic characteristics. Clusters C and D at the River Site and clusters E, F and G at the Pond Site were all dominated by one or a combination of herbaceous perennials, particularly Grindelia, Baccharis and Lepidium, with an understory of shorter perennials, including Frankenia, Rumex and Salicornia.
 

Area of effect of tidal channels

Except for transects from the sixth order channel, all transects ended in the background vegetation, defined by clusters I and I’ (at the River and Pond Sites, respectively). The distance to background varied significantly with channel order and type (Figure 4, ANOVA, p < 0.001), with natural channels influencing from 1.8 to 15.7 meters away from the banks depending on order and site (ANOVA, p < 0.001), though no significant differences were found between the means at each site across channel order. At the River Site, the distance to background increased with channel order among natural channels; the effect at the Pond Site was variable, but distance to background was 57-87% greater than at the River Site for second, third and fourth order natural channels.

The variation in distance to background between sites was related to channel size by plotting the mean area of effect by order and site against the logarithm of the corresponding channel cross-sectional areas (Figure 5). This plot shows three relationships between area of effect and channel size which depend on channel type and order. First the constructed channels, with evenly sized, small channels show almost no channel influence, cluster in the lower left corner of the plot, undifferentiated by channel order. Second the distance to background of larger channels (third through sixth order) increases approximately linearly with the logarithm of channel size. Third, small sized natural channels (first and second order) scatter along a line with much steeper slope than the slope observed for the larger order channels. The smaller channels appear to have a disproportionately larger area of influence compared to the larger channels.

The total area of marsh influenced at each site was estimated by multiplying the mean distance to background by channel length for each channel type and order and summing (Table 3.) The total area of channel of influence at the River Site was 20,706 m2, representing 18.89% of the total area of the site. The area of channel influence at the Pond Site was greater at 27,218 m2, but a lower percentage of total site area, 14.97%. At both sites, the natural channel network accounted for nearly all the area influenced by channels.
 

Sequences of vegetation clusters with channel size and distance from channel

For natural channels, there was a semi-regular progression of species clusters from first to sixth order and with distance from the channel (Figures 6 & 7). In general there is indication of substitution of the nearest channel cluster from smallest channels to largest channels across channel order. With each increase in channel order, the cluster that previously occupied the near channel position is shifted farther away from the channel or drops out (e.g. cluster H at the River Site and cluster G at the Pond Site). Simultaneously the dominance of the background clusters (I and I’) in the near channel environment declines with increases in channel order.
 

The influence of constructed channels

In general constructed channels had no significant effect on the vegetation based comparison of the first 10 meters to the second 10 meters (one-tailed t test, p = 0.23.) Transect species richness was significantly greater for natural channels than constructed channels of the same order (ANOVA, p < 0.001). The distribution of species clusters varied with channel type (natural vs. constructed channels); the background clusters (I and I’) dominated constructed channel transects (Figures 6 and 7). As a result the distance to background measurements were near zero and significantly different from the natural channels (ANOVA, p < 0.001) These results indicate that the constructed channels have not structured the vegetation in the same way as the natural channels.

Discussion

Taken as a whole, we believe this study demonstrates the importance of tidal channels in structuring the salt marsh vegetation of Petaluma Marsh. These patterns are complicated, resulting from a combination of factors including channel size, channel origin, and distance from channel. The location of the channel network, whether in the interior of the marsh (e.g. Pond Site) or exterior of the marsh (e.g. River Site), also seems to influence the details of species distribution. Nevertheless this study makes the first attempt to quantitatively evaluate how salt marsh plant distributions respond to the influences of the channel network.

Petaluma Marsh may be somewhat unusual in evincing such easily observed patterns, because of its landscape position in the San Francisco Bay. In the marshes surrounding the northern end of San Francisco Bay in California, the higher parts of the marsh often experience extraordinarily high salinities, due to concentrations of soil salts from evaporation after flooding, but less anoxia due to less frequent flooding. The lower parts of the marsh have lower salinities due to more frequent flushing, but more anoxic soils (Zhang et al, 1997). In addition the large sediment loads coming from San Pablo Bay and its position 8 km up the Petaluma River may contribute to a relatively high marsh plain characterized by extreme salinity levels.

Nevertheless, we believe this study is valuable in demonstrating quantitatively the extent to which even small channels, 50 cm wide and 75 cm deep, can affect overall species distribution in a salt marsh environment. The effects of small channels do not exist in isolation, moreover, but can be understood in the context of the influence of the entire channel network, based on the interactions of tidal flow and marsh soils. These effects are shown to include not only major dominant species that have been studied previously, like Salicornia, Spartina and Scirpus, but also minor species, like Achillea, Cuscuta, and Jaumea. Channels make an important contribution to the overall plant diversity of Petaluma Marsh.

Although we did not measure edaphic factors in this study, we hypothesize that the influence of tidal channels on the vegetation can most likely be explained by interactions of soil and water along the channel margins and proceeding away from the channels. Several studies have incidentally shown that salt marsh soils near to tidal channels are lower in salinity (Balling & Resh, 1983; Snow & Vince, 1984), have higher nitrogen levels (Brown & Bledsoe, 1996; Valiela et al, 1978), and reduced iron and sulfide toxin levels (King et al, 1982; Wiegert et al, 1983). There is a long history of relating plant species distributions to environmental gradients in the salt marsh literature, both in California (Mahall and Park, 1976abc; Ustin et al, 1982; Callaway et al, 1990) and elsewhere (Chapman, 1960; Adam, 1990), though it is clear from studies in the last decade that simple abiotic gradients are not be the only factors influencing plant distributions (Bertness, 1987; Bertness, 1991; Pennings & Callaway, 1992; Bertness et al, 1992; Bertness & Shumway, 1993; Pennings and Callaway, 1996.)

Given this modern view, we believe that the geography of tidal channels can be understood as setting up the environmental conditions over which all the biotic interactions of competition, mutualism, parasitism and response to disturbance play out to determine the observed species distributions. Though these interactions are important to understand, as a first approximation, the distribution of channels is of primary importance in determining the distribution of the vegetation.

Because of their importance in defining the vegetative structure of the marsh, we believe that tidal channels and channel structure are critical elements in any salt marsh restoration plan. After the correct hydrology is restored, the next important consideration is the distribution of tidal channels, since in turn these tidal channels will establish the other physical conditions important for the existence of healthy salt marsh plant communities. These channels need to be carefully planned and constructed, since poorly constructed channels will not be tidally active, and as a result, have little influence on the vegetation.

Acknowledgements

We thank Steven D. Culberson, David F. Dyson, Han-Yu Hung, E. Wayne Sanderson, Kelley L. Thompson and A. Gail Wheeler for their patient and able assistance in the field. We would also like to further acknowledge many helpful discussions with Eliska Rejmankova, Steven D. Culberson and E. Wayne Sanderson regarding this study. During this research EWS received funding under an Earth System Science Research Fellowship granted by NASA, 1995-GlobalCh00404. We thank Quinn Hart and George Scheer for computer support and the Digital Equipment Corporation for providing DEC Alpha computers under the Sequoia 2000 grant to SLU (Cooperative Research Agreement #1243).

References

Adam, P. 1990. Saltmarsh ecology. Cambridge University Press: Cambridge.

Atwater, B.F., Conrad, S.G., Dowden, J.N., Hedel, C.W., MacDonald, R.L. & Savage, W. 1980. History of landforms and vegetation of the estuary's tidal marshes in San Francisco Bay. pp. 347-384. In: Conomos, T.J. (ed), The urbanized estuary. Pacific Division American Association for the Advancement of Science, San Francisco, CA.

Barbour, M.G., Burk, J.H. & Pitts, W.D. 1987. Terrestrial Plant Ecology. The Benjamin/Cummings Publishing Company, Inc., Menlo Park, CA.

Balling, S. & Resh V. 1983. The influence of mosquito control recirculation ditches on plant biomass, production and composition in two San Francisco Bay salt marshes. Estuarine, Coastal and Shelf Science 16: 151-161.

Bertness, M.D. & Ellison, A.M. 1987. Determinants of pattern in a New England salt marsh plant community. Ecological Monographs 57: 129-147.

Bertness, M.D. 1991. Interspecific interactions among high marsh perennials in a New England salt marsh. Ecology 72: 125-137.

Bertness, M.D., Gough, L. & Shumway, S.W. 1992. Salt tolerances and the distribution of fugitive salt marsh plants. Ecology 73: 1842-1851.

Bertness, M.D. & Shumway, S.W. 1993. Competition and facilitation in marsh plants. American Naturalist 142: 718-724.

Brown, A.M. & Bledsoe, C. 1996. Spatial and temporal dynamics of mycorrhizas in Jaumea carnosa, a tidal saltmarsh halophyte. Journal of Ecology 84: 703-715.

Callaway, R.M., Jones, S., Ferren, W.R. & Parikh, A. 1990. Ecology of a Mediterranean-climate estuarine wetland at Carpinteria, California - plant distributions and soil salinity in the upper marsh. Canadian Journal of Botany 68: 1139-1146.

Chapman, V.J. 1960. Salt marshes and salt deserts of the world. Interscience Publishers, London.

Collins, J. & Resh, V. 1985. Utilization of natural and man-made habitat by the salt marsh song sparrow, Melospiza melodia samuelis (Baird). California Fish and Game 71: 40-52.

Collins, J., Collins, L., Leopold, L. & Resh, V. 1986. The influence of mosquito control ditches on the geomorphology of tidal marshes in the San Francisco Bay area: Evolution of salt marsh mosquito habitats. Proceedings of the 54th Annual Conference of California Mosquito & Vector Control Association : 91-95.

De Groot, D.S. 1927. The California Clapper Rail: its nesting habits, enemies and habitat. The Condor 29(6): 259-270.

Hickman, J.C. 1993. The Jepson Manual: higher plants of California. University of California Press, Berkeley, CA.

Hinde, H. 1954. Vertical distribution of salt marsh phanerogams in relation to tide levels. Ecological Monographs 24: 209-225.

Josselyn, M. 1983. The Ecology of San Francisco Bay Tidal Marshes: a community profile. US Fish and Wildlife Service, Washington, D.C. USA.

King, G. M., Klug, M.T., Wiegert, R.G. & Chalmers, A.G. 1982. Relation of soil water movement and sulfide concentration to Spartina alterniflora production in a Georgia salt marsh. Science 218: 61-63.

Leopold, L.B., Collins, J.N. & Collins, L.M. 1993. Hydrology of some tidal Channels in estuarine marshland near San Francisco. Catena 20: 469-493.

Ludwig, J.A. & Reynolds, J.F. 1988. Statistical Ecology. John Wiley & Sons, New York, NY.

Mahall, B.E., & Park, R.B. 1976a. The ecotone between Spartina foliosa Trin. and Salicornia virginica L. in salt marshes of northern San Francisco Bay I. Biomass and productivity. Journal of Ecology 64: 421-433.

Mahall, B.E. & Park, R.B. 1976b. The ecotone between Spartina foliosa Trin. and Salicornia virginica L. in salt marshes of northern San Francisco Bay II. Soil water and salinity. Journal of Ecology 64: 793-809.

Mahall, B.E. & Park, R.B. 1976c. The ecotone between Spartina foliosa Trin. and Salicornia virginica L. in salt marshes of Northern San Francisco Bay III. Soil aeration and tidal immersion. Journal of Ecology 64: 811-819.

Marin/Sonoma Mosquito and Vector Control District (1997) "Mosquito Control." Internet: http://www.sonic.net/~msmvcd/moscontr.html.

Marshall, J.T. 1948. Ecologic races of Song Sparrows in the San Francisco Bay region Part 1. Habitat and abundance. The Condor 50: 193-215.

Mason, H. 1957. A flora of the marshes of California. University of California Press, Berkeley, CA.

Mueller-Dombois, D. & Ellenberg, H. 1974. Aims and methods of vegetation ecology. John Wiley & Sons, New York.

Pearcy, R. & Ustin, S.L. 1984. Effects of salinity on growth and photosynthesis of three California tidal marsh species. Oecologia 62: 68-73.

Pennings, S.C. & Callaway, R.M. 1992. Salt marsh plant zonation - the relative importance of competition and physical factors. Ecology 73: 681-690.

Pennings, S.C. & Callaway, R.M. 1996. Impact of a parasitic plant on the structure and dynamics of salt marsh vegetation. Ecology 77: 1410-1419.

Pestrong, R. 1972. San Francisco Bay tidelands. California Geology 25: 27-40.

Purer, E. 1942. Plant ecology of the coastal salt marshlands in San Diego County, California. Ecological Monographs 12: 81-111.

Redfield, A.C. 1972. Development of a New England salt marsh. Ecological Monographs 42: 201-237.

Sanderson, E. W., Zhang, M., Ustin, S.L. & Rejmankova, E. 1997. Geostatistical scaling of canopy water content in a California salt marsh. Landscape Ecology 13:79-92.

Snow, A.A. & Vince, S.W. 1984. Plant zonation in an Alaskan salt marsh II. An experimental study of the role of edaphic conditions. Journal of Ecology 72: 669-684.

Strahler, A. 1964. Quantitative geomorphology of drainage basins and channel networks. Pages 4-39 - 4-76. In: Chow, V. (ed.) Handbook of Applied Hydrology. McGraw-Hill Book Company, New York.

Ustin, S.L., Pearcy, R.W. & Bayer, D.E. 1982. Plant water relations in a San Francisco Bay salt marsh. Botanical Gazette 143: 368-373.

Valiela, I., Teal, J.M. & Deuser, W.G. 1978. The nature of growth forms in the salt marsh grass Spartina alterniflora. The American Naturalist 112: 461-470.

Vogl, R.J. 1966. Salt-marsh vegetation of Upper Newport Bay, California. Ecology 47: 80-87.

Wayne, L.B. & Parker, V.T. 1994. Regeneration ecology of a peripheral zone halophyte, Grindelia stricta var. angustifolia in a San Francisco Bay tidal salt marsh. Bulletin of Ecological Society of America Supplement 75: 242-243.

Wayne, L. & Parker, V. 1996. High rainfall and fungal pathogens in a San Francisco Bay salt marsh -- a test of the importance of competition and salinity in determining zonation. Bulletin of Ecological Society of America Supplement 77: 470.

Whitlow, T.H. 1982. Plant-soil interactions in a San Francisco Bay salt marsh. Dissertation. University of California, Davis, CA.

Wiegert, R.G., Chalmers, A.G. & Randerson, P.F. 1983. Productivity gradients in salt marshes: the response of Spartina alterniflora to experimentally manipulated soil water movement. Oikos 41: 1-6.

Wolfe, R.J. 1996. Effects of open water management on selected tidal marsh resources: A review. Journal of the American Mosquito Control Association 12: 701-712.

Zhang, M., Ustin, S.L, Rejmankova, E. & Sanderson, E.W. 1997. Monitoring Pacific Coast salt marshes using remote sensing. Ecological Applications 7: 1039-1053.
1999, Center for Spatial Technologies and Remote Sensing (CSTARS)

University of California, Davis