Wildland Fuel Management Options for the Central Plains of Martha’s Vineyard: Impacts on Fuel Loads, Fire Behavior and Rare Plant and Insect Species William A. Patterson III, Gretel L. Clarke, Sarah A. Haggerty, Paul R. Sievert, and Matthew J. Kelty Department of Natural Resources Conservation University of Massachusetts, Amherst 01003 Final report submitted to the Massachusetts Department of Conservation and Recreation RFR# DEM705 March, 2005 Front Cover: Intense surface fire in unmanaged barrens fuels (left; photo by A. Woolsey). Harrowed and mowed firelane (right; photo by G. Clarke). Massachusetts Department of Conservation and Recreation state forest lands - southeastern Massachusetts. For additional information and photographs see: www.umass.edu/nrc/nebarrensfuels/ ACKNOWLEDGMENTS This work was supported by a grant from the Massachusetts Department of Conservation and Recreation (to W. A. Patterson, P. R. Sievert and M. J. Kelty), and by a grant from the Joint Fire Science Program (to W. A. Patterson and D. W. Crary) which facilitated our outreach efforts. The creation of the SW Experimental Fuels Break area was accomplished with the support of a US Forest Service funded National Fire Plan grant to the Massachusetts Department of Conservation and Recreation (Wildland Urban Interface Fire Grant). We would like to thank the staff of DCR, especially Bill Rivers, for his efforts to initiate the project, and Jim DiMaio and Jim Rassaman for ongoing support. Thanks especially to John Varkonda for his assistance with many aspects of this project, including finding and coordinating contractors, and generally facilitating our work on the Forest. Prescribed burns would not have been possible without the assistance of Joel Carlson and his crew from the Massachusetts Chapter of The Nature Conservancy and Dave Crary and his crew from Cape Cod National Seashore. Many individuals, representing many conservation organizations on and off the Island, assisted with prescribed burns. Becky Brown provided sheep, trained us in the basics of shepherding, and took an interest in our research. Other contractors patiently cooperated with us in our experiment in adaptive management including Mickey Callahan and Bruce Gurney. Matt Duveneck, Kimberly Iwamoto, and John Norton-Jensen assisted with the field tour conducted in 2004 and with analysis of data as the report was finalized. This research would not have been possible without field assistance from many including Dana Brennan, who was tireless in her efforts over several field seasons, and Christopher Wood, Ben Cotton, Kim Iwamoto, John Norton-Jensen, Timothy O’Riordan, Matt Duveneck, and Mike Ohman. Carol Knapp provided invaluable assistance in rare plant identification, as did Roberta Lombardi, Karen Searcy and Pam Polloni at the University of Massachusetts Herbarium. Jeff Boettner enthusiastically and willingly answered any and all Buckmoth questions. Michael Nelson, Tim Simmons, and Paul Goldstein provided input on study design and rare insect natural history. Glenn Motzkin, in particular, was generous in sharing his knowledge of the Martha’s Vineyard sandplain ecosystem. EXECUTIVE SUMMARY One of the largest undeveloped sandplains in Massachusetts exists in the Manuel F. Correllus State Forest (MFCSF) on the island of Martha’s Vineyard. Nearly 4000 acres (1670 ha) of barrens vegetation remain in an area currently recognized as critical habitat for a number of plant and animal species that are rare in the Commonwealth of Massachusetts. MFCSF has the state’s highest concentration of state-listed terrestrial animals, some of which have been extirpated from mainland New England. Unique coppice-oak stands on the central plains are comprised of individual oaks that almost certainly predate the time of first European settlement in the early 17th century. These plants may be several hundred years old and collectively resemble British coppice woodlands that have been managed for centuries. Sandplain vegetation can be highly flammable, and under dry, windy conditions it can support extreme fire behavior. In addition to its flammability, high fuel loading contributes to fire hazard in barrens vegetation. Scrub Oak stands, in particular, are highly flammable and support high litter and shrub fuel loads (1-hr plus 10-hr fuels = 14.3 t/acre (32 mt/ha) and fuel depths of 4-5 ft (1.3-1.5 m). Pitch Pine and Oak Woodland stands support lower surface fuel loads (10.9 and 10.2 t/acre; 24 and 23 mt/ha, respectively) and fuel depths of only 1.6 and 1.9 ft (0.5 and 0.6 m). Scrub Oak and Pitch Pine stands can support canopy fires with extreme fire behavior, whereas Oak Woodlands are inherently less flammable. Most rare plant species on MFCSF occur in culturally maintained grasslands. Previous work suggests that Scrub Oak is most important for rare Lepidoptera species. MFCSF is currently surrounded by private land which would be threatened by wildfires spreading from the Forest. As an initial response to the inherent fire danger, a series of fuelbreaks was established around and throughout MFCSF in the early 20th century. These fuelbreaks can both slow the spread of fire and provide access for fire suppression personnel. Fuels management along the boundary of the Forest in conjunction with the maintenance of existing breaks could help reduce the intensity of fires and the threat to adjacent properties. In the past, fuelbreaks have been created and widened using harrows to clear away native vegetation, a procedure which has created “fire lanes” dominated by grasses and forbs. Although these lanes currently support populations of state-listed plants, the vegetation represents an artificially created state not natural to the central plain. It has been argued that alternative management practices could be used to both reduce fuels and maintain open sandplain habitats more typical of natural conditions. This study examined the effectiveness of alternative methods for fuelbreak establishment and maintenance including thinning, mowing, grazing, and prescribed burning. Our goal was to evaluate ways to prevent wildfire escaping the boundary of MFCSF while at the same time conserving or enhancing the habitats of five rare plant species and 22 rare insect species. Here we report the results of our work and provide recommendations for future fire break creation and management. It is important to note that this report does not provide a “fire management plan” for the Forest as a whole; nor 3 does it assume that management techniques appropriate for the small portions of the Forest that might be managed as fire breaks would be appropriate if applied more widely as general vegetation management procedures. What we have learned could, however, serve as a guide to future efforts to develop management practices and plans for the Forest as a whole. Treatments we evaluated reduced slash and shrub loads and heights in the first growing season after treatments. Scrub Oak plots showed the most pronounced change from pretreatment conditions (shrub heights were less than a quarter and loads were well under half their pretreatment values). Mowing in Pitch Pine and Scrub Oak reduced shrub loads to well under 50% of pretreatment values. Sheep will graze new woody shoots following mowing, effectively reducing shrub loads however the expense of this treatment (which is more than four times that of mowing) may be prohibitive. In mow/graze Scrub Oak plots post-treatment loads were <10% of pretreatment values. Effects of treatments in Oak Woodlands were less pronounced as they were more heterogeneously applied. Pile burning and mowing are comparable in cost, and both effectively remove slash in thinned Pitch Pine stands. Creating lanes using alternative techniques can be comparable in cost to harrowing (although grazing is more expensive). However, long-term maintenance costs may be more expensive in the Experimental Fuel Break (relative to the cost of mowing harrowed lanes). Prior to treatments, average flame lengths [under moderate fire weather conditions of 3.5 mph (5.8 km/hr) midflame wind speeds and humidities of 65%] were more than 6.5 ft (2 m) in Pitch Pine and Scrub Oak and 3.5 ft (1.1 m) in Oak Woodlands. Untreated fuels supported rates of spread greater than 15 ft/min (4.6 m/min) in all three fuel types. Fire behavior in treated plots was greatly reduced with little difference between treatments and fuel types: average flame lengths were reduced to less than 2.1 ft (0.6 m) and rates of spread to less than 7 ft/min (2.1 m/min). Fire behavior in these treated plots is comparable to fire behavior expected in firelanes; predictions suggests that had prescribed burns been conducted on the driest windiest day of burning, flame lengths would have been 2.9 ft (0.9 m) and rates of spread of 59.4 feet/min (18.1 m/min) in firelanes. Custom fuel models performed well in predicting observed fire behavior and appear to be widely applicable across MFCSF. The longevity of treatment effects is unclear. Fuel loads will probably take somewhat longer to recover in Scrub Oak plots. However, it may take less than five years without further treatments. Growing season treatments could deplete shrub root reserves and slow fuel load recovery. Patterson (unpublished data) found that five annual treatments virtually eliminated Huckleberry from Oak Woodlands on Cape Cod that are similar to those on MFCSF. Litter layer compaction may have a more lasting effect, especially in Scrub Oak and Pitch Pine mow plots where increases in litter load were greatest. Thinning of Pitch Pine stands increased the estimated wind speed at which crowning would occur from 21 to 62 mph, and this change will be long-lasting, especially where seedling recruitment is likely to be dominated by Oaks. 4 Harrowing, and thinning with a fellerbuncher, increased soil compaction significantly relative to controls, however the magnitude of the change was relatively small (maximum values in harrowed and thinned areas were 180 PSI (1241 KPa) whereas values of 150-160 KPa were not uncommon in control stands).