Prescribed fire and cutting as tools for reducing woody plant succession in a created salt marsh

Summary

In 1993 a salt marsh was created within the Sabine National Wildlife Refuge when a mixture of dredged sediment and water was pumped into an area of approx. 200 ha. Water is only able to flow in and out through the unleveed northern and western sides, where the marsh abuts natural, lower elevation marsh areas; thus, flooding occurs far less frequently than in these natural tidal salt marshes. After initial colonization by herbaceous salt marsh plants, colonization by woody species including eastern baccharis Baccharis halimifolia and Jesuit's bark Iva frutescens began 3–4 years after site creation. At the time of this present study, I.frutescens dominated much of the high-mid marsh, at a density averaging 12 shrubs per m², and was continuing to spread. This paper reports on efforts to reduce I. frutescens. Treatments were a winter burn, cutting individual shrubs at ground level, and a combination of the two.

Study site: The study was conducted in a created salt marsh within the U.S. Fish & Wildlife Service Sabine National Wildlife Refuge (NWR) (29º53'N, 93º21'W) in Louisiana, southeast USA.

Plots and replication: Ten 30 x 30 m plots (corners marked with metal poles) were established in the mid-high elevation marsh zone dominated by I.frutescens. The burn treatment was randomly assigned to five plots with five serving as unburned controls. From 60 pre-identified plants in spread through each plot, 30 were randomly selected in each burned plot and 20 in each unburned plot for sampling, each marked with an aluminum tree tag, and pretreatment measurements were taken: number of primary stems per individual; maximum stem length; and diameter of the longest stem at 20 cm above the ground. A non-destructive estimation of fuel load was also made.

Burn treatment: Three plots were burned on January 8 and two on 14 January, 2003; weather conditions were similar on both days: mostly clear skies, maximum daily temperature 14–20˚C, 11–21 km/h west to northwest winds, and 47–59% relative humidity.

Prior to ignition, fire-monitoring devices (''Fireloggers'') were placed near the base of five shrubs through each plot to measure fire temperature, fire duration and total heat index. Fuel moisture was determined by collecting leaf samples (just prior to plot ignition) of I.frutescens and understory vegetation from each burn-treatment plot. In the laboratory, the samples were weighed, dried and reweighed to calculate percent moisture. Fire severity was evaluated on the day or the day after the burn to determine maximum scorch height (cm), percent woody material browned or burned, and percent leaves and understorey vegetation consumed by the fire.

Cutting treatment: The cutting treatment was applied to a random subset of 10 tagged individuals in each plot; each was cut at ground level with a saw. Shrubs in unburned plots were cut during the same week that the burns were conducted; however, those in burned plots were not cut until 1 month post-fire, at which time any original stems or post-fire resprouts were removed.

Regrowth: Regrowth was evaluated in May 2003 and September 2003 (at the end of one growing season). An individual was classified as 'surviving' if it produced new growth; each surviving shrub was remeasured for the number of stems, maximum stem length and diameter of the longest stem at 20 cm aboveground. In February 2003, an assessment of how well the non-destructive measures of response predicted actual aboveground biomass of undisturbed I.frutescens, was also made.

Abiotic site characteristics: In February 2003, ground water monitoring commenced when a PVC water well was installed in the centre of each plot. Data on the depth to and salinity of, the ground water was gathered from February 23 to September 2, 2003.

In April 2003, approximately 3 months post-burn, three randomly collected sediment samples from the 0–5 cm layer were taken from each plot. Samples were analyzed in the laboratory for total carbon, total nitrogen and total phosphorus concentrations. The samples from the unburned plots served as baseline nutrient characterizations, those from the burned plots allowed evaluation of any differences in sediment nutrient composition after burns.

The root zone (<20 cm below ground) was flooded for 56% of the growing season. Salinity varied across plots (range 10.8 to 28.7 g/l over 7 months), but within each plot it varied little throughout the duration of the experiment. Three months after burns, sediment from burned plots did not differ in nutrient content from unburned plots: average carbon 14.11 mg/g; total nitrogen 0.24 mg/g; total phosphorus 0.45 mg/g. Lack of differences may be due in part to little combustible organic matter within the sediment, as the sediment was young, undeveloped and had a high mineral content.

There was no mortality of I.frutescens following any of the treatments; each shrub resprouted over the course of one growing season and similar levels of regrowth were observed for all treatments. The 100% survival may be attributed to the time of year that the treatments were applied; in winter many perennials are dormant/semidormant, storing a large portion of their carbohydrate reserves in underground storage structures. Although the treatments resulted in a loss of all or most live aboveground biomass, it is possible that carbohydrate reserves were protected in underground structures. High sediment moisture at the time of the burns may have also contributed to the lack of mortality; burning under drier conditions may allow heat to penetrate deeper and for a longer period into the soil.

Recovery was negatively affected by ground water salinity, i.e. fewer new stems produced and smaller stem diameter. Recovery of stem length was affected by water level i.e. as depth to the ground water decreased, individuals produced taller stems. Individuals with larger initial stem diameters tended to produce a greater number of stems and taller individuals tended to produce taller stems, upon recovery.

Although fire severity varied greatly across the plots, such that some individuals were burned almost completely to the ground while others were barely singed, all survived and resprouted prolifically. Total heat index and duration of heating, however, did have a stimulatory effect on the number of new stems.

Conclusions: In this study, the single winter burn or cutting treatments, or a combination of both, did not shift the plant community structure away from I.frutescens-dominated back to herbaceous-dominated species. No I.frutescens mortality occurred in response to treatments.
Individual shrub recovery was related to initial plant size, ground water level and salinity, and two fire characteristics (total heating >60°C and total heat index >60°C). Fire severity, sediment nutrient concentrations, and other abiotic factors had no observable effects.


Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at:

http://www.springerlink.com/content/w6118631148u6784/fulltext.pdf