|
Soil Salinity Can Predict Invasion Rates of Phragmites australis in Salt Marshes
David M. Burdick
Jackson Estuarine Laboratory and Department of Natural Resources
University of New Hampshire, Durham, NH
dburdick@christa.unh.edu
Robert Buchsbaum and Eric Holt
Massachusetts Audubon Society, Endicott Regional Center, Wenham, MA
rbuchsbaum@massaudubon.org
A presentation given by the first author at the 2nd Annual Coastal Science Conference held on March 4th, 1999 at the Rupert A. Nock Middle School, Newburyport, Massachusetts. We gratefully acknowledge assistance from the students, paid professionals and especially volunteers who have helped us over the past three years.
Introduction:
Can we predict invasion rates of Phragmites australis using salinity information?
After their extensive use by the colonists and early Americans for cattle and sheep, salt marshes have been neglected over the past century. We have begun to recognize that healthy salt marshes are important part of our coastal ecosystem that provides us with fish and shellfish as well as recreation.
As you know, some salt marshes have been degrading due to direct impacts, such as filling for development and also from indirect impacts associated with road and railway bed restrictions that eliminate or reduce tidal flow, and increased freshwater runoff due to nearby development. The direct impacts from soil disturbance coupled with the changes in hydrology from indirect impacts have been accompanied by rapid increases in the distribution of Phragmites australis, the common reed.
Phragmites australis is able to out compete most salt marsh plants because it grows very tall and overshadows these marsh species, leading to inadequate light to support vigorous growth. Because Phragmites have a greater potential for deeper rooting than other salt marsh plants, we also wondered if Phragmites australis could avoid salinity stress by tapping fresher water resources at greater soil depths.
Although this reed occurs naturally in New England, its proliferation has awakened natural resource managers to take action to stop and reverse its recent spread. Why? Not only does Phragmites have negative values of being poor wildlife habitat and a fire hazard, its rapid spread is seen as a symptom of poor coastal resource management.
Several years ago, Robert Buchsbaum and I set out to see if we could use a few simple measures to help us understand why Phragmites was spreading rapidly.
Goal
Our goal is to understand the physical and biological conditions associated with rapid spread of Phragmites australis.
Objectives
We decided to use small wells placed at three depths in patches of Phragmites. We developed the following three objectives:
1) Examine soil water salinity in shallow, moderate and deep soils of Phragmites patches over the growing season
2) Determine the spread rates of Phragmites colonies
3) Relate patterns in soil salinity to the spread rates of Phragmites colonies
Hypotheses
1) Does salinity vary seasonally, with fresher periods in the spring and early summer (due to runoff and spring rains) and saltier periods in the late summer and early fall?
2) Can Phragmites australis expansion be predicted from average well salinity and amount of tidal flow?
3) Can we predict whether the adjacent station, #5, is likely to support rapidly expanding colonies of Phragmites australis based on our salinity data?
Study Sites and Methods
We established several sites in and near the Rough Meadows sanctuary. For this presentation, I will focus on three sites: Oak Knoll #1, a Phragmites patch on an oak island with unrestricted tidal access; Mud Creek, a patch in a degrading tidal marsh where Route 1A has restricted tidal flow; and Railroad, a Phragmites stand found in a tidally-restricted marsh where tidal exchange was recently increased with the installation of a new larger culvert under the upgraded MBTA rail line.
Our sample design is shown in Figure 2, below. At each marsh site, five stations were established in the following arrangement. Three of the stations were placed along a transect that was established from the center of a Phragmites stand to a salt marsh creek. The first station was placed in the center of the stand, the second in the transition zone at the edge of the stand where Phragmites graded into Spartina patens, and the third at the center of the Spartina patens stand, between the transition zone and the creek. Station #4 was placed in the transition zone some distance from the transect and station #5 was placed along the marsh-upland interface where Phragmites did not occur, but where one might find the transition zone if Phragmites was present. At each station a set of three wells drawing soil water from three depth intervals: 5-20 cm, 35-50 cm, and 65-80 cm, were installed.
(site numbers refer to Figure 2)
click to enlarge
Soil salinity was measured at each depth 15 times during the 1996 and 1997 growing seasons, with a temperature corrected optical refractometer. In fall of 1996, following the first growing season after the wells were established, and again in the fall of 1998, we assessed the vegetation at each site. The distance from well set #2 to the inner (where only Phragmites occurs) and outer (where Phragmites ends) edges of the transition zone were measured. The cover of each plant species was assessed within a 1 meter square area on the right and left sides of each well set, on the basis of 100% maximum cover. At the center of each sampling area, marked by a temporary flag, the height of five vegetative and reproductive shoots that were closest to the center of the area were measured.
Results and Discussion
Last year at this meeting, I reported that the salinity at the sites differed in that it was fresher in the spring, saltier and more stressful for Phragmites in the summer and fall (see Figure 2). This information led us to reject our null hypothesis from last year: Ho: Salinity is consistent seasonally; and accept our alternative:
Ha: Salinity varies seasonally, with fresher periods in the spring and early summer (due to runoff and spring rains) and saltier periods in the late summer and early fall).
We also ended our talk last year with the null hypothesis:
Ho: Phragmites australis expansion cannot be predicted from average well salinity.
Today I want to expand that hypothesis and examine it using information from three of our sites. Well salinity at each station is averaged over depth and 6 to 8 sampling dates for spring/early summer and compared to late summer/fall for each of our three sites in Figure 3. At all sites, salinity was greater later in the growing season. It is also clear that salinity was greater at the Oak Knoll site than at either of the other two sites where tides are restricted. Furthermore, when compared to the Railroad site, Mud Creek appears to have the lowest salinities, especially at stations 1, 2 and 4 in the Phragmites patch.
Figure 3_2
click to enlarge
The differences in salinities at the three sites were reflected by changes in the population characteristics as well as differences in the expansion rates of the patches of Phragmites (Figure 4 ). At sites with greater salinity (unrestricted tides), average plant height declined (an indicator of salinity stress in the species) and no expansion was measured. At the lowest salinity site, Mud Creek (tidally restricted), Phragmites expanded rapidly (about 1 meter per year over two years), while the cover and average height of the plants increased. At the site with intermediate salinity, Railroad, expansion was slow (0.1 meters per year at the leading, outer edge), and although plant cover increased, heights declined (Figure 4). The decline in height may be related to the restoration work that was performed at this site at the beginning of the 1998 growing season which increased the culvert size and partially restored tidal exchange at this site.
Figure 4_2
click to enlarge
This year I also want to pose a final hypothesis (3) regarding potential Phragmites expansion at station #5 for our three sites. Figure 5 shows salinity averaged by station at our three marsh sites. It is clear that salinity levels at station #5 in Oak Knoll is saline enough to restrict Phragmites establishment, if it were to be able to colonize the area. At the other two sites, Mud Creek and Railroad, it appears that if Phragmites were to colonize these adjacent areas, it would probably be able to become established and expand rapidly. This may be due to the human influence that caused the tidal restriction at these two sites. At the Mud Creek site, salinity data and rapid expansion rates indicate that this site is degrading and should receive strong consideration for tidal restoration. Because tides at the Railroad site have been restored recently, it would be valuable to collect further data to determine if the restoration is restricting or reversing the spread of Phragmites.
Figure 5_2
click to enlarge
Results of Hypothesis Tests
H1o: Salinity is consistent seasonally - REJECT
H1a: Salinity varies seasonally, with fresher periods in the spring and early summer (due to runoff and spring rains) and saltier periods in the late summer and early fall) - ACCEPT
H2o: Phragmites australis expansion cannot be predicted from average well salinity
H2a: Phragmites australis expansion can be predicted from average well salinity
Preliminary data suggest H2a is better than H2o, but this hypothesis needs to be tested using more sites.
H2o: If salinity at station #5 is similar to salinity at stations #1 or #2, we cannot predict whether Phragmites can expand rapidly if it colonizes the site.
H2a: If salinity at station #5 is similar to salinity at stations #1 or #2, we can predict that Phragmites will expand rapidly if it colonizes the site.
An experiment using potted Phragmites collected from the original patch and planted near station #5 could be used to help answer this question.
Conclusions
1) Salinity varies seasonally - fresher in the spring and early summer, saltier in the late summer and fall.
2) Salinity is greater at sites where tides are unrestricted and lower at the sites where tides are restricted.
3) Phragmites australis appears to spread more rapidly where tides are restricted and low salinity water is available to the plant at some depth.
|
