![Fire Management Today (69[4] Winter 2009) Will Feature Articles on Cooperative Efforts to Manage Fire on Inter- National, National, and Local Levels](http://data.docslib.org/img/6bcbf758d1f0bb9f1b90e0d75b19cf8a-1.webp)
Fire today ManagementVolume 69 • No. 3 • 2009
Looking Ahead, Working Together
Disaster Response Partnerships Coping with Change Fire Ecology in Public Schools Federal-Local Cooperation
United States Department of Agriculture Forest Service Coming Next…
Fire response seldom involves only one agency or affects only one group of stakeholders. The next issue of Fire Management Today (69[4] Winter 2009) will feature articles on cooperative efforts to manage fire on inter- national, national, and local levels. Examples include U.S. contributions to fire recovery efforts in Australia, the logistics of U.S. military support of firefighting missions in the United States, the Forest Service’s loan of wildland firefighting equipment to a local fire department, and how vol- unteers organized by the Nebraska Forest Service are reaching out into rural communities to educate landowners about fire prevention.
Other articles in the issue address on-the-ground aspects of fire manage- ment: modeling smoke distribution in anticipation of prescribed fires, measuring fuel moisture for its affect on fire intensity, and training fire- fighters in an intensive academy setting.
Erratum In Fire Management Today, vol. 69(2), the article Assessing Changes in Canopy Fuels and Potential Fire Behavior Following Ponderosa Pine Restoration, the crowning index maps on page 50, Figure 4, were inadvertently switched. The map under “1996/97” illustrates conditions in 2003 and the map under “2003” illustrates conditions in 1996/97. View the corrected figure in the online version of 69(2), located at
Fire Management Today is published by the Forest Service of the U.S. Department of Agriculture, Washington, DC. The Secretary of Agriculture has determined that the publication of this periodical is necessary in the transaction of the public business required by law of this Department.
Fire Management Today is for sale by the Superintendent of Documents, U.S. Government Printing Office, at: Internet: bookstore.gpo.gov Phone: 202-512-1800 Fax: 202-512-2250 Mail: Stop SSOP, Washington, DC 20402-0001
Fire Management Today is available on the World Wide Web at
Tom Vilsack, Secretary Melissa Frey U.S. Department of Agriculture General Manager
Thomas L. Tidwell, Chief Monique LaPerriere Forest Service Managing Editor
Tom Harbour, Director Editors Fire and Aviation Management EMC Publishing Arts
The U.S. Department of Agriculture (USDA) prohibits discrimination in all its programs and activities on the basis of race, color, national origin, age, disability, and where applicable, sex, marital status, familial status, parental status, religion, sexual orientation, genetic information, political beliefs, reprisal, or because all or part of an individual’s income is derived from any public assistance program. (Not all prohibited bases apply to all programs.) Persons with disabilities who require alternative means for communication of program information (Braille, large print, audio- tape, etc.) should contact USDA’s TARGET Center at (202) 720-2600 (voice and TDD). To file a complaint of discrimi- nation, write USDA, Director, Office of Civil Rights, 1400 Independence Avenue, S.W., Washington, D.C. 20250-9410, or call (800) 795-3272 (voice) or (202) 720-6382 (TDD). USDA is an equal opportunity provider and employer.
November 2009
Trade Names (FMT) The use of trade, firm, or corporation names in this publication is for the information and convenience of the reader. Such use does not constitute an official endorsement of any product or service by the U.S. Department of Agriculture. Individual authors are responsible for the technical accuracy of the material presented in Fire Management Today.
Fire Management Today 2 Fire
Management today Volume 69 • No. 3 • 2009 On the Cover: Contents Anchor Point: Fire Management Into the Future ...... 4 � Tom Harbour Hired for Fire: Wildland Fire Management Projects � Putting People To Work ...... 6 � Mary Carr Response Partnerships During Disasters: Emergency Support � Function 4 ...... 8 � Gordon Sachs Coping With Change ...... 13 � Shawna Legarza
Mapping plow line disturbance with After-Action Reviews—Who Conducts Them? ...... 15 � GPS in a pine flatwoods community Anne E. Black, Kathleen Sutcliffe, and Michelle Barton at Savannas Preserve State Park, Mitigation on Alabama’s Gulf Coast: Bon Secour National � Florida. See the article “Plow Wildlife Refuge ...... 18 � Line Disturbance From Wildfire Jeremy A. Keller Suppression in Two Florida State Parks.” Photo: Jeffrey T. Hutchinson. An Answer to a Burning Question: What Will the Forest Service Spend on Fire Suppression This Summer? ...... 26 � Karen Abt, Jeffrey Prestemon, and Krista Gebert The USDA Forest Service’s Fire and Aviation Management Staff has adopted a logo Teaching Fire Ecology in Public Schools ...... 29 � reflecting three central principles of wildland Traci Weaver fire management: • Innovation: We will respect and value Plow-Line Disturbance From Wildfire Suppression in Two Florida � thinking minds, voices, and thoughts of State Parks ...... 32 � those that challenge the status quo while Jeffrey T. Hutchinson and Richard E. Roberts focusing on the greater good. Lessons From the Hayman Fire: Forest Understory Responses to the • Execution: We will do what we say we Scarify-and-Seed Postfire Rehabilitation Treatment ...... 38 � will do. Achieving program objectives, improving diversity, and accomplishing Paula J. Fornwalt targets are essential to our credibility. Wildland Fire Behavior and “The Course of Science” Flowchart: � • Discipline: What we do, we will do well. Is There a Connection? ...... 44 � Fiscal, managerial, and operational Martin E. Alexander discipline are at the core of our ability to fulfill our mission. Fuels Treatments and Fire Models: Errors and Corrections . . . . 47 � J. Morgan Varner and Christopher R. Keyes
Firefighter and public safety is our first priority.
Volume 69 • No. 3 • 2009 3 by Tom Harbour Anchor Director, Fire and Aviation Management Point Forest Service, Washington, DC
Fire ManageMent into the Future �
oday, as I formulate this article, lished. In it, I see the success of our wildfires are burning in Arizona, We need to use better- future fire management efforts in TCalifornia, New Mexico, and defined protocols for promoting: Texas. Earlier in the year, we’ve • Fire-adapted communities, assisted in or managed fire suppres- managing wildfires and • Fire-adapted ecosystems, and sion efforts in Oklahoma, Florida, the tools that help us • Fire-adapted business practices— and South Carolina. Another fire evaluate risk and make how we operate. season is certainly upon us. So, better risk-informed what’s different about this fire sea- What’s the Same? son? What’s the same? And what decisions during fire What’s the same is our ongoing does fire management look like as incidents commitment to long-term coop- we look into the future? eration, management, and safety. We need to continue to work with existing partners, develop new part- What’s Different? dent-level support to fire manag- ners, and make those partnerships This year, we’ve received new guid- ers under the “two kinds of fire” as effective as possible. We need to ance for the implementation of the framework. WFDSS is an array of continue our work to accomplish Federal Wildland Fire Policy. While decision support applications that much-needed fuels projects to the policy itself has not changed, calculate risk and probability and enhance the safety of communities the implementation guidance has predict what may happen on a and our firefighters when wildland been updated. That guidance estab- fire—providing fire managers with fires happen. We need to promote lished two categories of fire: pre- tools to determine the safest, most the creation of fire-adapted com- scribed fire and wildfire. Prescribed efficient, effective management munities and elevate a widespread fires are defined as those with options within applicable land and understanding of fire—how to live planned ignitions, while wildfires resource management plans. with fire and our responsibilities are those started from unplanned prior to, during, and after fire inci- ignitions. Unplanned ignitions What’s in the Future? dents. include the naturally caused igni- What changes do I see in our tions formerly referred to as wild- future? To address the subject, I We need to use better-defined pro- land fire use. The new guidance consulted the Quadrennial Fire tocols for managing wildfires and allows for management of any nat- Review (QFR). The QFR is an inter- the tools that help us evaluate risk urally caused wildfire for resource agency assessment of current and and make better risk-informed benefit so long as that manage- future strategies and capabilities decisions during fire incidents. The ment response is supported by the and identifies core mission points, WFDSS tools will help managers respective unit’s land and resource future trends, and forces driving identify and focus on high-value management plans. All human- those trends. It is not a policy doc- objectives where success is likely, caused fires, however, will continue ument and does not make policy making the best use of available to be suppressed using the safest, recommendations. firefighting resources. most efficient, effective means. The first QFR was found to be an extremely accurate reflection of We have a fire doctrine that pro- The Forest Service created the developments that have brought us motes an informed, shared-learning Wildland Fire Decision Support to our current position. In January culture in which firefighters avoid System (WFDSS) to provide inci- 2009, the second QFR was pub- unnecessary risk and that encour-
Fire Management Today 4 ages keen awareness and observa- We work in a hazardous environ- Further Information tion, knowing the leaders’ intent at ment—it’s a fact. The Federal InciWeb may be accessed at all times, and assisting with adapt- Wildland Fire Management Policy,
Successes and In May of 2009, we mourned the Challenges loss of fellow firefighters Tom Risk, Mike Flynn, and Brian Bliss when The continued success of our Neptune Aviation Services Tanker efforts depends greatly on our #42 crashed while responding to a commitment to risk management wildfire in New Mexico. We must and safety within our workforce always remember these firefighters and work practices. This requires and those we have lost before them, continued investment in that work- dedicating ourselves daily to safety force to meet and maintain the and managing the risks we take in highest standards in the vital work every action. we perform.
Volume 69 • No. 3 • 2009 5 hired For Fire: Wildland Fire ManageMent ProjeCts Putting PeoPle to Work Mary Carr
ildland fire management is out front in economic Wrecovery efforts as the Forest Service helps put people back to work through the American Recovery and Reinvestment Act of 2009 (ARRA, colloquially coined the “stimulus bill”).
The ARRA provides $1.15 billion to the Forest Service for work on the Nation’s forests with a focus on providing and retaining jobs. Congress appropriated $650 mil- lion for Capital Improvement and Maintenance and $500 million for Wildland Fire Management. Of the $500 million for Wildland Fire Management, Congress directed Funds for wildland fire management will be used to thin dense stands of trees such as this, not only reducing hazardous fuels but also resulting in biomass that can be used as $250 million to be used on Federal a source of clean energy. Photo: Tom Iraci, Forest Service, Pacific Northwest Research lands and $250 million on State Station. and private lands. and engineering) gained in the Fire management projects funded More than half the initial Forest Service to help facilitate through AARA include removal of projects proposed for and coordinate the program. Some dense underbrush and other haz- projects also incorporate assistance ardous vegetation from crowded the Forest Service’s from rural community partners and forest lands. The intent of such billion dollars of funding other cooperators. work is to help protect communi- through the ARRA ties from large, unnaturally severe were for wildland fire In addition to creating or saving fires and to contribute to the resto- some 10,000 jobs with Wildland ration of fire-adapted ecosystems by management. � Fire Management Funds, Forest diminishing the rate, severity, and Service economic recovery efforts size of wildland fires. duction. A biomass utilization team overall also address Capitol is working to improve estimates of Improvement and Maintenance Related projects promote the use woody biomass and to develop inte- Projects that will repair and of biomass generated from fuels grated systems and technologies for improve roads, facilities, trails, reduction projects as a way to con- biobased products and bioenergy. administrative sites, and the tribute to renewable energy pro- ecosystems and watersheds affected Most of the jobs created by these by them. Mary Carr is a technical publications edi- projects are in the private sector, tor with the Forest Service, Ecosystem As of September 2009, the U.S. Management Coordination, Publishing with a small proportion of special- Arts, in Olympia, WA. ist positions (such as contracting Secretary of Agriculture had
Fire Management Today 6 Most jobs created by wildland fire management economic recovery projects will be in the private sector. �
released more than $429 million of Wildland Fire Management funds for hazardous fuels projects, inva- sive species projects, wood-to-ener- gy grants, and forest health proj- ects. These projects are on Federal, Small-diameter materials removed during hazardous fuels reduction work will be piled State, private, and tribal lands. and chipped for use as fuel for boilers under funding from the American Recovery and Reinvestment Act of 2009. Photo: Forest Service Eastern Region. Targeted projects are already under- way, and most will be completed within 2 years.
For updated lists of stimulus-fund- ed wildland fire management proj- ects and for other fire management news, visit: • The Forest Service economic recovery Web site at
Volume 69 • No. 3 • 2009 7 resPonse PartnershiPs during � disasters: eMergenCy suPPort � FunCtion 4 Gordon Sachs
ildland fire agencies respond to more types of emergen- The purpose of ESF4 is to provide Federal support Wcies than only wildland fire. for the detection and suppression of wildland, They also respond to hurricanes, tornadoes, floods, earthquakes, ter- rural, and urban fires resulting from, or occurring rorist attacks—any type of natural coincidentally with, an incident requiring a or manmade disaster or emergency coordinated Federal response for assistance. could result in a request for Federal wildland fire resources. Thus, the Forest Service, U.S. Department of Agriculture, may be called on to There are six Departments or agen- these resources are mobilized to support response to and manage- cies identified as Support Agencies support Federal or State entities ment of situations that involve to ESF4 to provide technical sup- with situations that do not involve multiple agencies. port, assistance, and expertise in wildland fire. specific areas related to firefight- A structure exists to guide such ing operations. In addition to the Forest Service Washington Office, multiagency interactions. When U.S. Department of the Interior Fire and Aviation Management, an event results in a Presidential and the U.S. Fire Administration, Disaster and Emergency Operations declaration of emergency or major these support agencies are the Branch, is the day-to-day link to disaster, the response is coordi- National Weather Service, U.S. FEMA and provides the national nated under the National Response Environmental Protection Agency, ESF4 coordinator. Each Forest Framework. Under the National the U.S. Department of Defense, Service region and area has a des- Response Framework, the Forest and the U.S. Department of State. ignated ESF4 coordinator to work Service serves as the Coordinator Their roles related to ESF4 are with their FEMA region(s). As sup- and Primary Agency for Emergency identified in the ESF4 Annex of the port agencies, U.S. Department of Support Function 4 (ESF4). During National Response Framework. the Interior bureaus and the U.S. all types of disasters and major Fire Administration have ESF4 emergencies, ESF4 is the primary Emergency Support coordinators to maintain a close link between the wildland fire com- Function 4 working relationship with the national ESF4 coordinator. munity and the U.S. Department The purpose of ESF4 is to provide of Homeland Security, Federal Federal support for the detection Emergency Management Agency During a disaster or emergency, and suppression of wildland, rural, ESF4 may be staffed at all levels of (FEMA). The Forest Service coor- and urban fires resulting from, or dinates and staffs ESF4 with the FEMA operations (figs. 1 and 2): occurring coincidentally with, an • The National Response support of the U.S. Department incident requiring a coordinated of the Interior and the U.S. Fire Coordination Center (NRCC) Federal response and assistance. at FEMA headquarters in Administration and serves as the Under the National Response face of wildland and structural Washington, DC; Framework, ESF4 manages and • A regional response coordination firefighting resources to FEMA and coordinates Federal firefighting other involved agencies. center (RRCC) in any of the 10 activities by mobilizing firefight- FEMA regions; Gordon Sachs is a disaster and emer- ing resources in support of State, • A joint field office (JFO) estab- gency response specialist for Fire and tribal, and local wildland, rural, and lished in any State affected by a Aviation Management in the Forest Service urban firefighting agencies. Often, Washington Office. disaster or major emergency;
Fire Management Today 8 • A geographic area coordination center (GACC); • The National Interagency Coordination Center (NICC): • A State emergency operations center (EOC); or • A FEMA emergency response team (ERT) or incident manage- ment assistance team (IMAT).
Positions Within ESF4 ESF4 is staffed by qualified person- nel from the Forest Service, U.S. Department of the Interior, and the U.S. Fire Administration. There are four ESF4 positions: ESF4 Primary Leader, ESF4 Support–Wildland, ESF4 Support–Structure, and ESF4 Support-Administrative (fig.3). U.S. Figure 1—Coordination of resources can be complicated during a presidential declaration Fire Administration employees fill of emergency or major disaster. The National Response Framework identifies the roles the role of ESF4 Support–Structure and structures of Federal agencies to provide support to States or other agencies through emergency support functions (ESFs). The National Response Framework identifies ESF4 to provide expertise on structural as the coordinator for wildland, rural, urban, and suburban firefighting support. EMAC: and urban firefighting. Persons Emergency Management Assistance Compact; ERT-A: Emergency Response Team-Advance qualified in each of the positions Element; DRG: Disaster Readiness Group; NOC: National Operations Center. must complete an intense, 3-day ESF4 course and must demonstrate proficiency through trainee assign- ments and completion of a position task book.
When activated, ESF4 staff at the NRRC serve as the link between Forest Service and U.S. Department of the Interior leadership and FEMA leadership. Similarly, ESF4 staff at the regional response coordination centers or joint field offices serve as the link between Forest Service and Department of the Interior regions or State offices and FEMA regions. In addition to being the conduit linking the resource ordering process to FEMA funding mecha- nisms, ESF4 staff gather ongoing intelligence for the ESF4 agencies and FEMA about emerging details regarding the incident. Figure 2—ESF4 is the link between the wildland fire resource ordering process and FEMA funding mechanisms during a disaster response. Upon request for support from a State, FEMA provides ESF4 with a mission assignment, and if accepted, ESF4 notifies the geographic area coordination center. As with wildland fires, the geographic area coordination center mobilizes and coordinates firefighting resources at the geographic area level.
Volume 69 • No. 3 • 2009 9 support of State or Federal agen- cies. Any mission assignments outside of the primary or support missions identified in the National Response Framework are closely scrutinized by ESF4 staff. When possible, ESF4 staff provide adviso- ry services, guidance, and training to the requesting agency to help them build capacity rather than performing the task for them.
Incidents Involving the Emergency Support Function Figure 3—ESF4 positions are staffed with qualified individuals from the identified agencies. The “ESF4 desk” is typically staffed in this basic configuration at the National Federal disaster response was tested Response Coordination Center and the regional response coordination centers and joint heavily during fiscal year (FY) field offices and sometimes at the geographic area coordination center, depending on the type of emergency. NMAC: National Multi-Agency Coordination Group; GMAC: Geographic 2008 during both exercises and Multi-Agency Coordination Group. declared emergencies and disasters. Exercises involved many Federal Tactical and Logistical and medical teams; Departments and agencies, and Support • Cache equipment and supplies ESF4—with staff from the Forest to support Federal urban search- Service, the U.S. Department of In addition to the primary ESF4 and-rescue task forces; the Interior, and the U.S. Fire mission, the Forest Service and the • Technical assistance and logisti- Administration—participated at U.S. Department of the Interior are cal support at oil and hazardous both the national and regional also identified as support agencies materials spills; and levels. There were also nine ESF4 to other emergency support func- • Law enforcement and investiga- activations under the National tions. These agencies may provide tion personnel. Response Framework: three for fire secondary support identified in emergencies and six for all-hazard the National Response Framework Additionally, FEMA can issue a emergencies. These activations when resources are available. direct mission assignment to were: For example, under these support the Forest Service or the U.S. Southern California Fire Siege— missions, the Forest Service may Department of the Interior through Fires driven by the Santa Ana be asked to provide: ESF4 to provide support outside winds in September and October • Transportation assets, such as of that specifically identified in the 2007 resulted in a fire situation aircraft; National Response Framework. that prompted a Presidential • Radio communications systems A mission assignment authorizes declaration of major disaster and and support personnel; FEMA funding for resources in the activation of ESF4 nationally • Engineering, contracting, and procurement personnel and equipment to assist in emergency removal of debris; When activated, ESF4 staff at the NRCC serve • Planning support to multiagency as the link between Forest Service and U.S. coordination centers; Department of the Interior leadership and FEMA • Resources and supplies for evacu- leadership. Similarly, ESF4 staff at the RRCC or ation shelters; • Staff for establishing mobilization joint field office serve as the link between Forest centers; Service and U.S. Department of the Interior • Personnel, equipment, and sup- regions or State offices and FEMA regions. plies to support Federal health
Fire Management Today 10 and regionally. ESF4 received a Acronyms mission assignment to “provide wildland firefighting resources As with any administrative function, FEMA has its own shorthand to include structure protec- that all ESF staff must learn and use. The following are acronyms that tion urban interface resources speed communication—once you’ve learned them. through the wildland fire mobi- lization system to Federal, State, DHS U.S. Department of Homeland Security and local agencies in support of DOI U.S. Department of the Interior the 2007 Southern California Fire DRG Disaster Readiness Group Siege.” More than 125 engine EOC (State) Emergency Operations Center strike teams and more than 300 ERT-A Emergency Response Team-Advance Element miscellaneous overhead person- ESF Emergency Support Function nel were requested on a single FEMA Federal Emergency Management Agency mission assignment—the largest FS Forest Service such request in history. GACC Geographic Area Coordination Center IMAT Incident Management Assistance Team Texas Winter Fires—Unusually JFO Joint Field Office dry conditions throughout the MA Mission Assignment winter and spring of 2007–2008 NICC National Interagency Coordination Center resulted in a presidential decla- NOC National Operations Center ration of emergency for many NRCC National Response Coordination Center counties in west Texas. FEMA NRF National Response Framework activated ESF4 regionally to sup- RRCC Regional Response Coordination Center port the State of Texas with direct USDA U.S. Department of Agriculture firefighting assistance. The Forest USFA U.S. Fire Administration Service supplied more than 15,000 personnel days of fire sup- pression assistance over a period of more than 6 months. local responders. Forest Service ally, and, for over a month, personnel assisted with the over- provided coordination between Micronesia Drought/Flood— sight of base camp establishment, FEMA and the wildland fire com- Drought conditions followed by selection of feeding sites, incident munity through operations at salt water intrusion resulted in planning, and development of the California State Emergency an emergency situation in the emergency purchasing processes Operations Center, two geograph- Federated States of Micronesia, and protocols. ic area coordination centers, a former U.S. Trust Territory. and the National Interagency Under the National Response Northern California Fire Siege— Coordination Center. Framework, FEMA activated In late June, an unprecedented ESF4 regionally to provide a dry lightning storm started more Hurricane Dolly—In advance forester to serve as part of a pre- than 1,000 fires in northern of Hurricane Dolly in late July, liminary damage assistance team California. These and other fire FEMA activated ESF4 nation- to evaluate damage to local water outbreaks resulted in a presi- ally and regionally. This storm supplies and food sources. dential declaration of emergency made landfall near South Padre for several counties in northern, Island, Texas, as a category 2 hur- Midwestern Floods—In June, central, and southern California. ricane. Forest Service personnel heavy rains resulted in significant More than 20,000 firefighters maintained situational awareness flooding in many Midwestern were deployed to California dur- and remained ready to evalu- States. FEMA activated ESF4 ing July, including firefighters ate requests for assistance made regionally to assist the State of from Canada, Australia, New by FEMA or the State of Texas. Iowa with logistical support for Zealand, and Greece. ESF4 was Because of heavy wildland fire hundreds of Federal, State, and activated nationally and region- activity, few resources were avail-
Volume 69 • No. 3 • 2009 11 able to assist with this all-hazard There are six Departments or agencies � response; fortunately, there were no requests that could not be identified as support agencies to ESF4 to � filled. provide technical support, assistance, and expertise in specific areas related to firefighting Hurricane Gustav—This category operations: the U.S. Department of the Interior, 3 hurricane made landfall on September 2 near Cocodrie, LA, the U.S. Fire Administration, National Weather not far from New Orleans. FEMA Service, U.S. Environmental Protection Agency, � activated ESF4 nationally and the U.S. Department of Defense, and the � regionally prior to landfall. ESF4 U.S. Department of State. received a mission assignment for 5 incident management teams, 14 hand crews, and numerous other resources. Forest Service person- to strike in early September dur- and numerous overhead person- nel, coordinated through ESF4, ing the height of the response nel and other resources. These were instrumental in supporting to Hurricane Gustav and during resources ultimately supported the aviation portion of the mass preparations for Hurricane Ike. the State of Texas with logistics evacuation from the Gulf Coast coordination, emergency debris region. Hurricane Ike—This strong cat- clearing, damage assessments, egory 2 hurricane made landfall and aviation coordination. Hurricane Hanna—FEMA acti- on September 13 near Galveston, vated ESF4 in two regions to TX. FEMA activated ESF4 nation- More information about ESF4 is assist with planning for the ally and regionally prior to land- available through a short (20-min- possible impact of Hurricane fall. ESF4 received a mission ute) independent study overview Hanna on the East Coast. The assignment for 5 incident man- course at
Fire Management Today 12 CoPing With Change � Shawna Legarza
n recent years, we have wanted and seen both organizational Iand systematic changes in Forest Service, Fire and Aviation Management. Although they under- stand that change is constant and transformation takes time, even the most experienced fire managers may become overwhelmed and be left feeling confused and uncertain within themselves and with regard to the decisions of fire manage- ment agencies. While we continue to work more diligently to keep up with changes and obtain good results, we often feel our lives spin- The San Juan Hotshots in Lake Chelan, WA, 2005. Photo: San Juan Hotshots. ning out of control, and, thus, the cycle continues. It seems we have this said, understanding the process less and less free time to commu- Is it plausible to relax and the effects of change is a chal- nicate effectively with our families lenge for any leader. and take care of our inner selves. and communicate in Where have our fundamental pri- moments of uncertainty During your career, the connec- orities gone? and confusion? tion between challenges in your personal life and organizational As we balance our fate and ideology also our employees, overwhelmed changes in your professional career as an agency, we need to allow time may evolve into a complex web of for personal reflection and remem- with the stress and implications of change. Employees who become confusion that could become highly ber our fundamental priorities. I stressful and profoundly unhealthy. believe we need to be the leaders overwhelmed have difficulty man- aging stress and, thus, lessen the You may become addicted to the of organizational and systematic risk of change and the speed of the change; we need to continue to connections with their cowork- ers and families. So we must ask: emergency, while not truly know- develop the wildland firefighting ing the long-term physiological culture within our doctrine to take How does one become receptive to the effects that change has upon and psychological effects. With the better care of ourselves, our fami- competitively based subculture lies, and our employees. oneself, coworkers, subordinates, and families? Is it plausible to relax of firefighting, we often forget to and communicate in moments of take care of ourselves and oth- The Ultimate Challenge uncertainty and confusion? ers because we, too, have become While enduring political, global, obsessed with trying to understand and ecological challenges and the Historically, it is clear that change and keep up with all the changes structured flows of technology, we can raise doubts concerning per- that make us uneasy and, some- often find not only ourselves, but sonal beliefs and existing order. times, outright unmanageable. Statistical comparisons show that During this negative reinforcing Shawna Legarza is a district fire manage- the inability to deal with change ment officer on the San Juan National closely correlates to a negative sub- loop, communication and per- Forest in Bayford, CO, and founder of liminal behavior exhibited briefly sonal reflection become even more the Wildland Firefighters Life Challenge during times of undue stress. With important. Personal reflection will Program. allow you to evaluate yourself.
Volume 69 • No. 3 • 2009 13 With time, you may find a relation- the stress-related effects of sense- ship between your inner self and making—the ability to see things the observed chaos in the outside how they are and to be open to world. By understanding this inner other frameworks during times of reflection, it may become easier to change, and the ability to find your embrace change and differentiate passion and connect this passion between the fundamental variables with leadership. To understand and defining your personal and profes- communicate effectively within sional life styles. Take time to slow yourself is self-leadership; to under- down. stand yourself is to understand life. All the while, understanding the Individual Leadership success of your life means to under- Traits stand the variables of incremental Shawna Legarza, superintendent of the change over time. I believe we all have unique leader- San Juan Hotshots, 2006. Photo: San Juan Hotshots. ship traits in our internal toolbox. Embracing Change The traits in your toolbox will allow you to manage the transformation peace within and on the outside. As leaders in Fire and Aviation of change on an individual basis. Knowing how to effectively lead Management, we need to continue Some people may add various skills yourself during times of change is to embrace change while educat- to their toolbox and use them to difficult for even the most gifted ing our employees about taking their utmost, others just a little, individuals. effective care of themselves (self- leadership) and their employees (leadership). I believe the essence of You may become addicted to the risk of change leadership is leading your passion and the speed of the emergency, while not within through the challenges in life. To honestly believe what is true truly knowing the long-term physiological and to you is the most reflective self- psychological effects. � leadership discipline.
Dig deep to find your answers. or not at all. It is your choice to The principal difficulty in evalu- Remember, the most consistent accept and understand change, just ating your own success is con- event in your lives will be the as it is your choice to be a leader, a templating your own experiences challenges of coping with change. follower, or a leader of leaders. with trial and tribulation. During With the global challenges we are some leadership opportunities in now facing, we need to continu- Individuals have different leader- the past, maybe you used trial and ously embrace the transformation ship traits, supported by different error to learn lessons for next time. of change by being adaptive and types of incremental change, and If you make substantial changes creative in both our personal and with all levels of leadership come within following your mistakes, professional lives. We will all learn individual challenges and complexi- you can learn to communicate from those who remember the fun- ties. One must continue to lead more effectively about your feel- damental priorities and embrace within during times of uncertainty, ings. You can express the feelings the transformation of change to communicate honestly with of change to yourself, your cowork- within themselves and their oneself and with others, and to find ers, and your family, thus reducing organizations.
Fire Management Today 14 aFter-aCtion revieWs— Who ConduCts theM? Anne E. Black, Kathleen Sutcliffe, and Michelle Barton
eflecting on the links between of discrepancies, and is the practice each shift, assignment, incident, or intentions and outcomes is a having a positive impact on perfor- season). Rkey practice of a learning orga- mance? nization (Garvin 2000). The After- We also asked respondents about Action Review (AAR) is a formal We asked 668 randomly selected their perception of their group’s reflection process intended to assist survey participants from across the performance. We are still in the groups in capturing lessons learned country about their AAR practices, process of analyzing the full dataset from a task. AARs typically ask four as part of a larger study of high but want to share answers to the questions regarding fire-response reliability behaviors among perma- first of these questions: who is con- operations: (1) what did we set nent fire staff in the Forest Service, ducting AARs and at what point? out to do, (2) what actually hap- pened, (3) why is there a difference between the first two, and (4) what AARs typically ask four questions regarding fire- should we continue/what should we change? Since the Wildland Fire response operations: what did we set out to do; Lessons Learned Center sponsored what actually happened; why is there a difference training workshops on AARs for between the first two; and what should we the fire community in 2002, the continue, and what should we change? practice seems to have been widely adopted. You can hear the term almost every where you go these the Bureau of Land Management, Data Classification days, from engine bays to incident and the National Park Service (see: Because there were far more sup- command posts. But just how wide- “A Multidisciplinary Approach pression events (373) than pre- ly has the practice been implement- to Fire Management Strategy scribed fires (59) or wildland fire ed? Are all levels and all functions Suppression Costs, Community use (54) events, we grouped the in the fire organization conducting Interaction, and Organizational respondents’ roles first by basic AARs? How good are AARs as cur- Performance,” in this issue). We organizational unit: agency (e.g., rently used at getting at root causes asked each whether their groups agency representative, line offi- (the units with which they worked cer, fire management officer, duty Anne Black is a social science analyst and most closely) had conducted an officer, resource advisor); dispatch ecologist for the Forest Service’s Rocky AAR in association with the last (including dispatch function and Mountain Research Station in Missoula, fire they were on and, if so, at what MT. Kathleen Sutcliffe is associate dean Geographic Area Coordination point the AAR was conducted. To and the Gilbert and Ruth Whitaker Center level); or team (for anyone Professor of Business Administration at the help understand current practices, assigned to the incident). Ross School of Business at the University of we also asked them what general Michigan, Ann Arbor, MI. Michelle Barton type of fire this was (suppression, is a doctoral candidate at the Ross School Because there were significantly of Business at the University of Michigan, prescribed, or wildland fire use); more team (538) than dispatch (73) Ann Arbor, MI. the class of that event (initial or agency (57) respondents (fig. attack, extended attack, home unit, NOTE: Policy terminology and guidance 1a), we broke team respondents have changed since this article was pre- or project-complex); the role they down into their functional roles: pared. “Wildland fire use” is no longer played during this incident (or the considered a category of fire management aviation (helicopter managers, heli- role they spent the most time in but describes one potential management tack, etc.); Incident Commander on that event); and at what point strategy for dealing with natural-ignition (type 5–1, area command); and wildfires. The term has been retained here they conducted their AAR (after to reflect the structure of the study. others according to their Incident
Volume 69 • No. 3 • 2009 15 Figure 1—Percentage of permanent Federal fire staff Command System (ICS) function (information, finance, reporting that their group conducted an After-Action Review logistics, operations, planning, or safety) (fig. 1b). on their last assignment in 2007, with total number of responses per class shown at top of bar, Finally, because operations numbers outnumbered the (a) Basic organizational unit responses from other ICS functions by an order of mag-
100 nitude, we broke out the operations function into general 90 resource type or level: division supervisor, firefighter 1, 538 80 57 firefighter 2, interagency hotshot crew, operations sec- 70 tion chief, task force/strike team leader, crew, engine, and 60 50 a category that includes specialized roles (e.g., sawyers,
40 73 bulldozer operators, firing bosses, and fire use module 30 members) (fig. 1c). 20 10 While these queries and subsequent analysis cannot tell 0 Agency Dispatch Team us about the quality of the AARs, they do indicate some significant differences in current practices. Subsequent (b) Team respondents by Incident Command System analysis will help us associate learning practices with per- function on suppression events, and ceptions of performance. 10 0 90 95 291 Results 80 9 14 70 54 37 According to this study, staff on prescribed fire, fire sup- 20 16 60 pression, and wildland fire use events use AARs in similar 50 40 proportions: about 74 percent of the time. There are, 30 however, significant differences in AAR use by basic orga- 20 nizational unit, class of fire, team function, and opera- 10 tional role (table 1, fig. 2). Frequency of AARs conducted 0 by groups involved in initial attack (64 percent) is signifi-
Plains Safety Finance cantly different than on those on project-complex fires. Aviation Logistics Information Operations Only 37 percent of respondents in dispatch units reported Incident Command their group held an AAR, whereas 86 percent of team- based respondents said their group had done so. Perhaps (c) Operations function by resource type. surprisingly, respondents working in aviation reported that their group conducted AARs less frequently than 22 100 19 38 all but those in an information function. At an overall 90 35 28 59 80 38 79-percent rate, those in safety reported the group they 33 70 19 associated with lagged significantly behind those in opera- 60 tions functions (85 percent). Among the operations func- 50 tions, division supervisors reported their group held AARs 40 30 less frequently than any other category (although not 20 statistically so), while all of those participating in a crew 10 reported conducting AARs. 0 1
Crew Special Engine Most of these AARs are conducted after each shift and/ or after the incident—at least for prescribed fire and sup- pression operations (fig. 3). Respondents who last worked Operations Chief Firefighter, Type Firefighter, Type 2 Taskforce Leader Division Supervisor on a wildland fire use incident reported that their AARs were generally conducted after an assignment. This tim- Interagency Hotshot Crew ing may reflect the different tempo of action on different
Fire Management Today 16 Table 1—Statistical results for assessing differences in after-action review (AAR) practices on wildland fires.
Categories Compared Pearson Chi-Sq test results Type of fire (suppression, prescription, wildland fire use) (χ² (2, N = 653) = 2.71, p = .258) Organizational unit for all classes (χ² (2, N = 668) = 63.08, p = .000*) Class of fire (initial assessment, extended assessment, home, or project/ (χ² (3, N = 499) = 11.15, p = .011) complex) Team function for suppression events (χ² (8, N = 541) = 27.57, p = .001) Operational role on suppression events (χ² (8, N = 291) = 16.63, p = .034)
* Significant results (bold) are those with p-values less than 0.05. Models indicate whether or not there are significant differences among the categories, but not which ones are significantly different.
Figure 2—Percentage of respondents on different types of types of incidents: “boots-in-the-black” forces conducted suppression events reporting that their group conducted an AARs by operational shift; prescribed fire and host units, AAR, with total number of responses from each type at top after the incident; and wildland fire use, after an assign- of bar. ment or in place of a close-out. Quite a few respondents
100 noted that they conducted several types of AARs—after 20 90 shifts, after assignments, and after the fire season. 180 80 186 70 113 More information on AARs, including background infor- 60 50 mation and training materials, as well as fire-related AARs 40 and details on how to share your own AAR, may be found 30 at the Wildland Fire Lessons Learned Web site:
70 Rx 259 Wildland Fire Use 60 Suppression 50 204 55 40 30 20 45 10 11 0 after periodic after after after shift assign incident season
Volume 69 • No. 3 • 2009 17 Mitigation on alabaMa’s gulF Coast: � bon seCour national WildliFe reFuge Jeremy A. Keller
hile southern Alabama is and are now realizing the benefits Fort Morgan Peninsula—a narrow, better known for the city of from our efforts. This article pro- sandy strip bordered on the south WGulf Shores and its vacation vides a case study about our efforts by the waters of the Gulf of Mexico and tourism attractions, it is also to carry out a successful fire man- and to the north by Mobile Bay. The home to the Bon Secour National agement program on this challeng- area is widely known for its stun- Wildlife Refuge (NWR). In an area ing unit of the NWR System. ning white “sugar sand” beaches of intense wildland-urban interface and the vacation condos of Gulf (WUI) development pressure, it was The Fort Morgan Shores. The Fort Morgan Peninsula challenging to successfully man- Peninsula and Bon is undergoing explosive develop- age large areas of fire-dependent Secour NWR ment, much of which is occurring ecosystems without sacrificing our in WUI areas bordering the refuge. core responsibility of maintaining Bon Secour NWR occupies about Residents in the area view the ref- and restoring endangered species 7,000 acres (3000 hectares) on the uge favorably, taking advantage of habitat. Table 1—Historic wildfire activity on Bon Secour National Wildlife Refuge, 1988–2007.
Key to maintaining Number of Total Acres Number Total Acres Year Year the ecological integrity Fires Burned of Fires Burned of the refuge is a 1988 2 0.2 1998 1 3.0 comprehensive fire 1989 0 0.0 1999 1 3.0 management program, 1990 1 0.3 2001 1 455.0 including prescribed fire. � 1991 3 885.2 2002 1 12.7 1992 1 175.0 2003 0 0.0 Despite the challenges, we succes- 1993 0 0.0 2004 2 5.5 sively implemented an aggressive 1994 4 57.6 2005 0 0.0 fire management plan that met our primary objectives: safeguarding 1995 2 21.0 2006 0 0.0 life and property from wildfire and 1996 1 49.0 2007 1 180.0 caring for the natural resources 1997 2 728.1 of Bon Secour NWR. We put aside many traditional ways of doing business, worked hard to develop Table 2—Significant wildfire incidents, Bon Secour National Wildlife Refuge. effective cooperative relationships, Incident Acres Year
Jeremy A. Keller was a wildland-urban Oyster Bay Fire 820 1991 interface fire specialist for the Gulf Coast Refuge Complex, U.S. Fish and Little Dauphin Fire 175 1992 Wildlife Service, Gautier, MS. Presently, Mouse Fire 722 1997 he is the coordinator for the Top of Ohio Resource Conservation and Development Little Point Clear Fire 455 2001 Council, Natural Resources Conservation Service, U.S. Department of Agriculture, Three Rivers Fire 180 2007 Bellefontaine, OH.
Fire Management Today 18 ment in which almost every fire every almost which in ment environ - an to contribute all sions subdivi- encroaching and debris, hurricane terrain, impassable loads, fuel heavy ecosystems, dependent Fire- 2). and 1 (tables suggest alone frequency and size average their than complex more are incidents however,hectares); these (453 acres 112 of size mean a with year a once about incident wildfire a has NWR Secour Bon the average, On communities. local of screen” “radar the off fall to fires wild- of impact the allowing ously simultane- while re-accumulate to loads fuel permit periods quiet ing activity.interven- fire The intense of two or year a between passing typically years several with rence, occur wildfire of pattern cyclic a has Peninsula Morgan Fort The NWR Secour Bon on Problem Wildfire The fire. prescribed ing includ- program, management fire comprehensive a is refuge the of integrity ecological the maintaining to Key birds. migratory for point stopover important an is and tles tur sea for sites nesting and bates) ammo- polionotus (Peromyscus mouse beach Alabama endangered the for habitat critical protects refuge The Coast. Gulf the along all development rapid of because extent in dwindling ecosystems coastal unique protect to 1980 in established was NWR Secour Bon ricanes. hur of threat the on focused is tion atten- more because developments to wildfires of danger potential the to oblivious generally are visitors However,and veloped. residents unde- remain will it that fact the appreciating simply or tunities oppor recreational numerous the Volume 69 • No. 3 • 2009 • 3 No. • Volume69 - - - - a whole, data for the city of Gulf of city the for data whole, a as peninsula the for available not are data census While opment. devel- future for zoned or ment, develop- under developed, either is property private of foot square buildable every nearly means which values, estate real high in resulted has Peninsula Morgan Fort the on development Rapid money.and resources of commitment major a requires Source: U.S. Census Bureau. Census U.S. Source: AL. Shores, Gulf and County Baldwin unicorporated for data permit building in reflected as 1990, since Refuge Wildlife National Secour Bon around growth 2—Population Figure reflected in census data for Gulf Shores, AL. Source: U.S. Census Bureau. Census U.S. Source: AL. Shores, Gulf for data census in reflected as 1990, since Refuge Wildlife National Secour Bon around growth 1—Population Figure Building Permits Issued Total Population 1000 1500 2000 2500 10000 500 2500 5000 7500 0 0 1990 Source:
Source: U.S. Census Bureau 1990
U.S. 1992 1992 iy fGl Shores Gulf of City ufSoe Trendline Shores Gulf Census
adi Cut n ufSoe Alabama Shores Gulf and County Baldwin
1994
Bureau e Bidn emt Ise 19902006 Issued Permits Building New 1994 Gulf Shores Alabama Census Area Population Estimates 19902006
State 1996 1996
of the Year Year Year
increasing dramatically since 1990. since dramatically increasing County Baldwin and Shores Gulf for issued permits of number the story,similar with a tell 2) (fig. period same the for data permit Building 1). (fig. percent 10.5 about of increase annual average an or 1990, since population in increase overall 167-percent a experienced has Shores Gulf area. the in growth rapid the reflect County Baldwin unincorporated for and Shores 1998 Cities adi C. Trendline Co. Baldwin nnoprtdBlwn County Baldwin Unincorporated 1998
Database
2000 Population 2000
2002 2002
Linear 2004 2004
(Population)
19 2006 2006
Table 3—Fire resources normally available for initial attack on Bon Secour National the refuge. Fire management on Wildlife Refuge, with approximate response times. the refuge is the responsibility of our fire staff at Gulf Coast Refuge Response Time to Agency Fire Resources Complex in Gautier, MS, more than Bon Secour NWR 2 hours away. Initial attack for the 2 x Engine, Type 6 refuge is provided by Fort Morgan U.S. Fish and 2 x Engine, Tracked (Type 6) 2–3 hrs Volunteer Fire Department (VFD) Wildlife Service 3 x Tractor-Plow, Type 4 for refuge lands in Baldwin County, and Gulf Shores Fire-Rescue (a Alabama Forestry 1 x Tractor-Plow, Type 4 30–40 min combination career and volunteer Commission 2 x Tractor-Plow, Type 5 department) for refuge lands inside Fort Morgan 2 x Engine, Type 6 the city limits of Gulf Shores. These Volunteer Fire 3 x Engine, Type 1/2/3 10–15 min two fire departments are backed up Department by Alabama Forestry Commission Gulf Shores Fire- 2 x Engine, Type 6 tractor-plow units stationed in 10–15 min Rescue 5 x Engine, Type 1/2/3 Loxley, 30 to 40 minutes away (table 3). Every fire on Bon Secour NWR is In 2004, Hurricane Ivan complicat- a WUI incident. The fires are infre- ed fire management on the refuge Response to wildfires on the Fort quent but of high intensity and in two ways. First, salt water and Morgan Peninsula is complicated by difficult to control. Vegetative cover high winds killed many overstory jurisdictional issues. The U.S. Fish on the Fort Morgan Peninsula is trees, which have since cured and and Wildlife Service has jurisdiction dominated by fire-dependent spe- created a high load of 100- and over all refuge lands, but has no cies. The sand pine (Pinus clausa), 1,000-hour fuels. Secondly, Ivan fire resources on site. The Alabama saw palmetto (Serenoa repens), and deposited large amounts of debris, Forestry Commission has primary scrub oak (Quercus spp.) vegetation most of which remains in the jurisdiction for wildfires on all complex that covers most of the remote parts of the refuge. Some of State and private property, but has refuge presents a challenging fuels this debris contributes additional few resources available. The city of environment for fire managers. The heavy fuel in the form of lumber, Gulf Shores has annexed the entire density and energy content of these fuels require the use of mechanized equipment in the form of tractor- To successfully tackle the wildland fire issues plows; no other firefighting tool facing the refuge and local residents, a fully can counter the intensity and rapid engaged community is required. � rates of spread produced.
Bon Secour NWR terrain is mostly right-of-way along State Highway utility poles, and other ordinary dune and swale topography where 180, which runs the length of the combustibles. Lurking among the ancient beach dunes generally run peninsula. Owners of individual debris are hazardous materials, and east and west, separated by tidally properties bordering the right- although much was cleaned up, influenced sawgrass swales and of-way may petition the city for undetected items still remain on cypress (Taxodium sp.) strands. annexation, which is rarely denied. the refuge as hidden dangers dur- While the sandy ridges are traf- The same owners may likewise peti- ing fire operations. ficable and offer good locations for tion for de-annexation; again, these firelines, the swales and strands requests are rarely denied. This has are a formidable obstacle—difficult Fire Resources on the resulted in an ever-changing patch- to cross on foot, and impossible to Fort Morgan Peninsula work of jurisdictional responsibility cross with tracked equipment even Although the Bon Secour NWR is regarding fire protection between during extreme drought. These the largest single landowner on Fort Morgan VFD and Gulf Shores areas can literally be dry enough to the Fort Morgan Peninsula, there Fire-Rescue. burn intensely, yet too wet and soft are no fire personnel stationed at to allow equipment operation.
Fire Management Today 20 Permanent firelines around The Rookery subdivi- sion. Photo 1 shows area behind homes prior to treatment (Oct. 2006). Photo 2 shows area after Phase I brush-cutting treatment (Nov. 2006). Photo 3 shows area following Phase II removal of all vege- tation and Phase III creation of park-like visual buf- fer (Sept. 2007). The red arrow points to the same house in each photo for visual reference.
Addressing the Fire directions especially difficult to As development has intensified, Problem on the Refuge predict. They also deny the pen- many of these old firelines are insula of precipitation that falls inadequate because of their insuf- A chief goal of our fire staff is to on the nearby mainland, mean- ficient width and tendency to reintroduce prescribed fire to Bon ing that the peninsula can have meander along terrain features Secour NWR. Prescribed fire is drought effects far in excess of rather than follow actual property recognized as a critical tool for areas just a few miles away. lines. meeting the ecological manage- ment goals of the refuge, while • Intense development hinders the To address these issues, a multiyear simultaneously keeping fuel loads use of prescribed fire. Refuge plan was initiated (table 4). Notable manageable. Our use of prescribed areas that were bordered by phases included the following: fire is especially challenging on this undeveloped and unused wild- • To use prescribed fire on a more refuge for several reasons: lands just a few years ago are regular basis with the least • Because Bon Secour NWR is now crowded by subdivisions amount of risk to surrounding located on a peninsula separating on all sides. This restricts burn communities, we implemented the Gulf of Mexico from Mobile windows because of the risk of an integrated program of new, Bay, unique coastal weather phe- escaped fires and smoke impacts. permanent fireline construc- nomena, such as sea and land In the past, simple two-track tion coupled with fuel reduction breezes, complicate fire behavior. roads served as adequate fire- buffer zones. We established 6.2 These cape-effect dynamics must breaks between refuge lands and miles (10 kilometers) of 16-foot be factored into all fire planning adjacent undeveloped properties. (4.9-meter) minimum width, and operations; they make wind
Volume 69 • No. 3 • 2009 21 October 2006 September 2007 Buffer zones around The Rookery subdivision. Photo on left shows pretreatment condition with heavy shrub- layer fuels (Oct. 2006). Photo on right shows buffer area after removal of shrub layer and creation of buf- fer zone (Sept. 2007), preserving aesthetic values for local residents while providing a park-like fuel break between refuge and private property.
bare mineral soil firelines along for fire equipment and improved problem. To successfully tackle the refuge boundaries. These fire- access to existing refuge road and wildfire issues facing the refuge and lines are offset from current and fireline networks. Gravel lanes local residents, we actively engaged soon to be developed residential were strategically located along the community through a concert- areas by 50 to 100 feet (15 to 30 the new permanent firelines to ed and long-term program of coop- meters) onto the refuge, creating allow full access for local fire erative relations with key partners 25 acres (10 hectares) of buffer department equipment around in the area. zones. Within these buffer zones, subdivisions and other vulnerable shrub-layer fuels were reduced residential areas. Prominent among these coopera- mechanically while retaining all tive relationships are those we have large overstory trees. The result With all these measures in place, fostered with local fire depart- is a park-like appearance that implementation of further pre- ments. We enjoy a particularly close serves as both an extension of scribed burning is planned for fiscal partnership with the Fort Morgan the fireline and a visual buffer for year 2009 or fiscal year 2010. This VFD, an organization that has his- residents. (See the sidebars for will allow us to better manage the torically served as the initial attack images of the fireline and buffer fuel problem on Bon Secour NWR force for the majority of wildfires zones.) while simultaneously achieving the on the refuge. The strength of our ecological objectives for which the partnership with Fort Morgan VFD • To support prescribed fire and refuge was established. is exemplified by the location of its wildfire response, we initiated a main fire station on refuge prop- series of infrastructure improve- Addressing the Fire erty. ments. For many areas, heavy Problem Through fire equipment previously was Partnerships Our cooperative relationship unloaded by stopping traffic on with Gulf Shores Fire-Rescue has busy Highway 180, which pre- Fire management problems on strengthened in recent years as sented obvious safety hazards to the Fort Morgan Peninsula do not well. While this organization’s main both fire crews and motorists. We stop at the refuge boundary. While focus is oriented toward urban fire constructed a permanent stag- the measures described above are issues, there are still significant ing area that provides a 100- by designed to mitigate fire manage- blocks of refuge land for which it 150-foot (30- by 45-meter) gravel ment issues on the Bon Secour provides first-line fire protection. parking and unloading apron NWR, they address only part of the And as annexation rapidly contin-
Fire Management Today 22 ues, its interest in wildland issues increases accordingly.
Because these two fire departments are key to the success of our fire management objectives on Bon Secour NWR, we have taken several measures to bolster their wildland capabilities. Between 2001 and 2008, we secured $11,000 in funds from the U.S. Department of the Interior Rural Fire Assistance (RFA) grants program for Fort Morgan VFD and $2,000 for Gulf Shores Fire-Rescue, money that primar- ily purchased personal protective equipment and hand tools for the departments. Where RFA funds were not adequate, we made direct equipment loans to the depart- ments to meet specific wildfire- Figure 3—Map exercise from the Fire Operations in the Wildland-Urban Interface (S-215) fighting capability shortfalls. The course taught at Fort Morgan Volunteer Fire Department in January 2007. centerpiece of these efforts was Fort Morgan VFD’s purchase of a new brush truck (type 6 engine), which Table 4—Integrated fireline, fuel reduction, and fire apparatus access program for Bon involved a cost-sharing arrange- Secour National Wildlife Refuge. ment with the State of Alabama; the Gulf Coast Refuge Complex Phase Activities Fiscal Year provided the necessary hoses, appli- ances, and hand tools to bring it Initial planning into service. Endangered species, wetlands, and archaeo- 0 2005 and 2006 logical consultations In addition to improving our Marking of fireline corridors cooperator’s equipment status, Initial brush cutting and tree removal to I 2006 we have also worked to improve open 20-foot (6-meter) fireline corridors its human capital by offer- Blading of 16-foot (5-meter) fireline corri- ing training courses, such as II dors to mineral soil on areas accessible to 2007 Basic Wildfirefighter Training equipment (S-130/S-190), Fire Operations in the Wildland-Urban Interface Creation of 50- to 100-foot (15- to (S-215), and Advanced Firefighter III 30-meter) park-like buffer areas between 2007 Training (S-131). In 2008, funding firelines and vulnerable residential areas from the Ready Reserve Program Construction of gravel staging area for was secured for additional train- IV unloading of heavy equipment and park- 2007 ing in 2009 through 2012. This ing of transports training will include Wildland Hand clearing of 16-foot (5-meter) fireline Chainsaws (S-212), a Wildland V corridors in areas unsuitable for equip- 2009 Engine Academy, and a session ment use of Basic Wildfirefighter Training built around the National Wildfire Creation of gravel access lanes for fire Coordinating Group/National Fire VI department apparatus on firelines around 2010 Protection Association Crosswalk vulnerable residential areas
Volume 69 • No. 3 • 2009 23 Agreement and geared toward the the Baldwin County Emergency enhanced—an important consider- career firefighters of Gulf Shores Management Agency (EMA) to pro- ation in an area where hurricanes, Fire-Rescue (fig. 3). Eventually, we vide emergency planning expertise hazardous material spills, and hope to work with our coopera- and share our substantial com- search-and-rescue incidents are all tors to develop a cadre of qualified munications capability. Based on real possibilities. engine bosses and initial attack discussions with this agency, the incident commanders as a way of standard Annual Operating Plan An Early Return on expanding local capabilities to safe- for Bon Secour NWR was modified Investments: The ly and effectively address the local to more closely align with exist- Three Rivers Fire fire problem. ing county emergency plans. The current version follows the Federal On August 26, 2007, Fort Morgan We have not reached out exclu- Emergency Management Agency VFD informed us they were en sively to fire departments in our planning format as much as pos- route to a wildfire on a remote efforts to stay ahead of the WUI sible. An additional benefit of our part of Bon Secour NWR. As rela- fire threat on the Fort Morgan strengthened relationship with the tive humidity was between 80 and Peninsula. Because managing fire Baldwin County EMA is that our 90 percent on the peninsula that hazard is not a task for firefight- preparation for all-hazard incidents day, the initial expectation was the ers alone, we actively engaged on Bon Secour NWR has been fire would not amount to much.
b
a c
Figure 4—Three Rivers Fire: (a) intense fire behavior in local fuel types; (b) tractorplow lines used to control the fire; and (c) morning briefing with representatives from local fire departments in attendance (Aug 2007).
Fire Management Today 24 Because of extreme drought condi- incident, it showcased the success in managing the wildfire problem tions, however, the fire was able of all our cooperative efforts over on the Fort Morgan Peninsula to burn intensely in heavy fuels the years. Both Fort Morgan VFD had begun to take place. The close and posed a potential threat to a and Gulf Shores Fire-Rescue com- relationships we developed with subdivision as it continued to grow mitted resources to the fire and our cooperators were evident when through the night (fig. 4a). were fully involved in managing these agencies quickly responded to the incident (fig. 4c). All agencies the initial fire, seamlessly handed Because the fire was located on a involved worked hard to bring the off to our resources, and contin- remote and roadless part of the incident to a successful conclu- ued to provide meaningful support refuge, Fort Morgan VFD’s engines sion by taking advantage of their through the successful conclusion were unable to reach it; even their strengths and supporting each of the incident. all-terrain vehicles became hope- other as appropriate. While the lessly bogged down. Firefighters department’s primary firefighting Conclusion eventually made their way in on resources could not directly engage Our approach to the fire problem foot, but quickly realized that the fire, they were able to stand by on the Bon Secour NWR is a clear hand tools were ineffective against for structural protection. example of how an aggressive wild- the fire. When Gulf Coast Refuge fire management program can be Complex resources arrived on Fort Morgan VFD jumped in to pro- successfully pursued even without scene, an attempt to reach the vide water supply support for the dedicated on-site fire resources in fire with a U.S. Fish and Wildlife FWS tracked engine and to provide the face of extreme WUI develop- Service (FWS) low-ground-pressure facility support. Gulf Shores Fire- ment pressure. The key to our suc- tractor-plow unit was made, but Rescue provided an amphibious cess has been our close and mutu- even this machine became bogged all-terrain vehicle to re-supply FWS ally beneficial partnerships with down. Eventually, the fire was con- units and provide emergency medi- other agencies in the community. tained at 180 acres (73 hectares) cal support on the fireline. Baldwin Only by accepting the reality that using helicopters and indirect lines, County EMA stood ready to support we cannot “go it alone,” and that along with the assistance of a thun- our efforts with their radio cache we need to address wildfire hazards derstorm that dropped rain almost and reverse 911 system, although as a team, have we been able to directly on the burned area (fig. 4b). ultimately these were not needed. carry out an effective and progres- sive fire management program. While the Three Rivers Fire was not The Three Rivers fire showed how especially remarkable as a wildfire our efforts to engage local partners
Volume 69 • No. 3 • 2009 25 an ansWer to a burning Question: What Will the Forest serviCe sPend on Fire suPPression this suMMer? Karen L. Abt, Jeffrey P. Prestemon, and Krista M. Gebert
ildfire management has here, we also make forecasts for the become an ever-larger 2 fiscal years beyond that (2 and 3 Our models show that Wpart of Forest Service, U.S. years ahead). The table shows the Department of Agriculture, and report for our current-year fore- suppression costs can other land management agency casts, made in November 2008 for be statistically estimated appropriations and expenditures. the FY 2009 fire season. All the largely from previous In fiscal year (FY) 2008, the wild- dollar values reported for 2009 and years’ suppression fire program budget was nearly 44 shown in the table and accompa- percent of initial Forest Service nying figures are in estimated FY costs, climate, � discretionary appropriations (U.S. 2009 dollars to allow consistent drought conditions, � Congress 2008). Total expenditures comparisons across years. and a time trend. � for suppression eventually exceeded the initial appropriations by more Our models show that suppression than $500 million, resulting in costs can be statistically estimated on our preferred current model, additional appropriations from largely from previous years’ sup- Benchmark 2, and uses the best Congress and internal transfers pression costs, climate, drought available data and forecasting meth- from Forest Service programs. conditions, and a time trend. ods at our disposal. The Benchmark Clearly, wildfire suppression has Hazardous fuels are not directly 2 model is slightly different from become a dominant part of Forest included in our model, largely those reported in Prestemon and Service budgeting, planning, and because data are not available for others (2008) and Abt and others activities. all regions and all years (1977 to (2009). 2008), but the effects of climate Modeling the Past and weather on fuels and the time- We model the Forest Service trend effect of increasing fuel loads In an effort to provide early warn- regions separately, regressing real are captured in part by the other ings to wildfire managers and to (discounted) suppression costs on variables. Other influences on provide additional information the independent variables noted costs that are not directly included for the Forest Service budgeting above, and then we estimate all the are input price trends (energy, process, we forecast Forest Service regions together using statistical labor, capital, etc.) and manage- suppression costs in collabora- techniques to account for cross- ment changes (such as Appropriate tion with Forest Service Fire and region correlations. We then devel- Management Response and the Aviation Management by using op “jackknife” forecasts to test the National Fire Plan). Even with the computer models that use weather accuracy of our forecast models. best available data and statistical trends and suppression costs as These jackknife forecasts estimate methods, a portion of suppression inputs. We make our forecasts in the model coefficients with all but costs is unpredictable. As a result, November for the current fiscal 1 year of the data; then the coeffi- our best models can only account year’s fire season (the current-year cients and the independent variable for between 59 to 89 percent of the forecast) and, while not discussed data from the year not included annual variation in costs. are used to forecast the costs for Karen Abt is a research economist and the missing year. This process gen- Jeffrey Prestemon is a research forester In the process of developing a fore- erates a time series of historical at the Forest Service’s Southern Research cast, we test several models, and “backcasts” (fig. 1). Comparisons Station in Research Triangle Park, NC; we develop new models each year. Krista Gebert is a research economist of the backcasts and observed costs at the Forest Service’s Rocky Mountain The forecast in the table is based produce estimates of the fore- Research Station in Missoula, MT.
Fire Management Today 26 Table 1—Wildfire suppression cost forecast results for FY 2009 (current year forecast) for Forest Service regions, in 2009 dollars.
Regions 1-6 Rest of (Western Region 8 Region 9 Region 10 the Forest Total Forest Regions) (Southern) (Eastern) (Alaska) Service* Service
Millions of 2009 dollars
Point estimate 1,067 73 25 8 125 1,298
Mean 1,113 86 37 8 131 1,375
Median 1,113 75 26 8 131 1,368 95-percent confidence interval Lower bound 815 27 5 3 37 990 Upper bound 1,411 209 130 13 225 1,795 90-percent confidence interval Lower bound 862 32 7 4 52 1,048 Upper bound 1,364 177 101 13 209 1,720
*The “Rest of the Forest Service” includes emergency suppression related expenditures by national offices not tied to the regions and by the agency’s research stations. cast accuracy (root mean squared the outcome from the simulation in While the Forest Service FY 2009 error). Our forecasting methods the middle of the probability densi- budget has not been finalized at are explained in greater detail in ty function. The root mean squared the time of this writing—as of Prestemon and others (2008) and error associated with this forecast December 2008, the Forest Service Abt and others (in press). is $189 million (for the entire data was operating under a continuing period from FY 1982 to FY 2008). resolution for October 1, 2008, to For the current-year forecast, we use the forecast models and the error distribution for all of the input data to simulate a probability density function for each forecast year (fig. 2). From the forecast model we get the point forecast, while from the simulation we get the mean, median, and the 90- and 95-percent confidence intervals, as shown in the table and figure 2.
Forecasting the Future The current-year point forecast for FY 2009 is $1,298 million. Using the simulation analysis, the esti- mated mean is $1,375 million and the median forecast is $1,368 mil- lion. If forced to choose a single Figure 1—Wildfire suppression cost forecasts (point estimates) and actual wildfire number for the forecast, we recom- suppression expenditures from FY 1982 to FY 2008, and the FY 2009 forecast for the mend using the median forecast, Forest Service, in 2009 dollars.
Volume 69 • No. 3 • 2009 27 March 6, 2009—for the purpose of this analysis, we assume that the FY 2009 budget is similar to FY 2008. Using an estimated bud- get allocation for FY 2009 of $854 million in conjunction with the simulation results, we conclude that there is a 99-percent chance that the estimated budget will be exceeded in FY 2009. Of course, that implies a 1-percent probability that the budgeted amount will not be exceeded.
References Abt, K.L.; Prestemon, J.P.; Gebert, K. 2009. Wildfire suppression cost forecasts for the U.S. Forest Service. Journal of Figure 2—Simulation results for the FY 2009 wildfire suppression cost forecast showing Forestry. 107(4): 173-178. the mean and median forecasts as well as indicators of the 9-percent confidence interval Prestemon, J.P.; Abt, K.L.; Gebert, K. 2008. bounds, in 2009 dollars. Suppression cost forecasts in advance of wildfire seasons. Forest Science. 54(4): 381–396. U.S. Congress. 2008. P.L. 110-161 Consolidated Appropriations Act. Available at
Fire Management Today 28 teaChing Fire eCology in PubliC sChools Traci Weaver
hen children describe a the GYA. The remaining grant on the site of a prescribed burn wildfire, they often reflect money will support education conducted by interagency fire Wwhat they have seen and majors from the Environmental crews in 2001 to reduce potential heard from the news media: fire is Interpretation and Education fire fuels around the Teton Science scary and it destroys everything. Department of the University of Schools campus. The fifth-graders Fire education specialists from the Montana Western working with collected data for the park’s fire U.S. Department of the Interior, Montana schools. National Park effects program. National Park Service, and U.S. Service and Forest Service fire edu- Department of Agriculture, Forest cation specialists provide training Each group measured sagebrush Service, in the Greater Yellowstone to the university students, coordi- and bitterbrush in burned and area (GYA) are working to revise nate visits to the national forests unburned areas and then had those perceptions and explore the and national parks, and provide its own research focus, such as positive effects of fire with elemen- specific information for research comparing soil moisture and tem- tary and middle-school students locations in burned areas. perature in each site or counting through hands-on fire ecology the scat of different wildlife spe- research. cies. Each group drew comparison Fire education graphs and presented its findings to After a successful pilot program the other students. in Yellowstone and Grand Teton specialists are joining National Parks in 2008, the with college graduate “The students’ research was Greater Yellowstone Coordinating students and education authentic and useful for Grand Committee (GYCC) granted majors to present a Teton National Park,” said Josh $12,000 to support the fire ecology Kleyman, a lead graduate faculty education field trips throughout the 3-day fire ecology and member at Teton Science Schools. GYA, which includes six national management program “Students were inspired to know forests and two national parks. Fire that involves both that their data collection and aca- education specialists are joining field and classroom demic work had real-world implica- with university graduate students exercises to fourth- tions. Moreover, the teachers were and education majors to present a excited about this new opportu- 3-day fire ecology and management through eighth-grade nity.” program that involves both field students. and classroom exercises to fourth- According to Diane Abendroth, the through eighth-grade students. Grand Teton’s fire ecologist, “Teton In fall 2008, more than 300 stu- Science students are contribut- Fire Ecology in Schools dents from three school districts ing field data that is valuable to Part of the grant provides sup- explored burned areas while learn- the park’s fire management pro- port to Teton Science Schools, a ing about fire’s role in the GYA gram. This partnership gives us program for university graduate ecosystem. In September, 200 fifth- an opportunity to explore new and students in science and environ- grade students in Teton County, important questions about local fire mental education, in working with WY, visited the Teton Science ecology.” Wyoming and Idaho schools near Schools campus in Grand Teton National Park for a 3-day program. Research Goes Back Traci Weaver is a fire communication and The students learned basic fire education specialist serving national parks to the Classroom… in Montana and Wyoming. She is based at ecology and then conducted field Teton Science Schools returned Grand Teton National Park in Moose, WY. research. The research plots were to the fifth graders’ classrooms in
Volume 69 • No. 3 • 2009 29 A standing carpet of 3- to 12-foot Three hundred (1- to 4-meter) tall lodgepole pines, Douglas-fir, and spruce now covers students learned much of the burned acreage. basic fire ecology and then conducted field The students visited Yellowstone research, monitoring through a partnership between the National Park Service and Teton vegetation in burned Science Schools. After receiv- and unburned plots of ing training about Yellowstone’s sagebrush in the park. � fire management program, Teton Science graduate students con- ducted a 3-day fire ecology and research. In the park, the students fire management session for Chief broke into groups and collected data near Undine Falls, one site within the more than 500,000-acre (202,190-hectare) North Fork Fire area, the largest of Yellowstone’s 1988 blazes. The research site Counting trees—Montana sixth-grade contained areas that were severely students identify and count trees that have regenerated after the 1988 North Fork burned, moderately burned, and Fire in Yellowstone. Photo: Traci Weaver, unburned, allowing the students to National Park Service. make comparisons. They learned the spring for follow-up with the how to identify lodgepole pine, students, who created more graphs Douglas-fir, and spruce and how to and analyzed the information they record the number of each species had gathered in the previous fall, in a 10-meter (33-foot)-diameter incorporating math and science plot. standards emphasized by the Teton County School District.
“With an outreach group of instruc- tors visiting gateway communities to explore fire ecology and the follow-up work in the Teton County Holes—Students study holes left behind by insect-seeking birds. Photo: Traci Weaver, fifth-grade classrooms, the nature National Park Service. of the collaboration has truly been profound,” Kleyman said. “We look Joseph Middle School sixth grad- forward to more opportunities for ers in Bozeman, MT. Because authentic research within our parks those children have “Fire Fridays” and forests.” with education specialists from the Gallatin National Forest, they …And the Classroom already had a basic understanding Goes to the Field of wildfire fuels, fire behavior, fire prevention, and the dynamics of the In October 2008, students from wildland-urban interface. two Montana schools visited burn sites 20 years after the famous The graduate students spent the Measuring—Teton Science Schools 1988 Yellowstone fires. They dis- first classroom day teaching the graduate student Sarah Fuller (right) helps covered that, while the fires’ scars Teton County, WY, fifth-grade students students about fire’s role in the measure sage and bitterbrush regrowth, are still there, so is an amazing Greater Yellowstone area and 7 years post-prescribed fire. Photo: Traci display of natural regeneration. explaining how to conduct field Weaver, National Park Service.
Fire Management Today 30 “For many students this was either their first time to the park or, at least, their first time off the road and boardwalks,” said sixth-grade teacher Wendy Pierce. “I have always been unsure how to collect data with a large group of kids, so it was really interesting to watch the grad students’ teaching tech- niques.”
Real-World Scenarios and Decisionmaking The lessons concluded back in the classroom with role-playing accord- ing to different groups affected by wildfire: biologists, ranchers, town Outdoor classroom—Teton Science Schools graduate student Josh Gold explains research council members, tourists, and procedures to a group of sixth-grade students in an area that burned in the 1988 North fire managers. Each group had to Fork Fire in Yellowstone. Photo: Traci Weaver, National Park Service. decide what measures to recom- mend for a hypothetical fire: full suppression, partial suppression, or allowing the fire to fulfill its natu- ral role.
“The role-playing exercise has kids making the tough decisions that land managers must make when they face a wildfire,” said Marianne Baumberger, fire preven- tion and education technician for the Gallatin National Forest. “They definitely understood that fire plays an important role in the Greater Yellowstone Ecosystem, but it is a delicate balance when you consider safety, esthetics, habitat, impacts to people, and cost.” Posting results—Students talk about a graph showing which type of tree shows the most regeneration since the 1988 Yellowstone fires. Photo: Traci Weaver, National Park Service. Expanding on the Experience the group produced a podcast that more children the opportunity Fourth graders from Gardiner captured one important aspect of to witness first-hand the effects Elementary School near what they’d learned. Their podcast fire has on the ecosystem as well Yellowstone’s north entrance com- can be viewed at
Volume 69 • No. 3 • 2009 31 PloW-line disturbanCe FroM WildFire suPPression in tWo Florida state Parks Jeffrey T. Hutchinson and Richard E. Roberts
istorically, lightning-ignited alterations of hydrological patterns, this study was to assess the rela- fires in Florida spread fre- and sites for potential invasion tive amount of disturbance caused Hquently across large areas by exotic plants. Suppression of by fire suppression techniques in of the landscape, creating habi- wildfires creates enduring environ- small natural areas of southeast tats for plants and animals, and mental impacts that are perhaps Florida. naturally reducing fuel loads more significant than the effects of (Abrahamson and Hartnett 1990). wildfire (Taylor and Gibbons 1985; Florida’s Changing Today, large, contiguous landscapes Ingalsbee 2004). With the excep- Landscape are rare because of the develop- tion of a few studies (Taylor and ment of residential areas and Gibbons 1985, Caling and Adams Coastal southeast Florida con- roadways, and because fires that 1999, Ingalsbee 2004), little is tains one of the highest human burn across Florida landscapes are known about the disturbance or population densities in the United rare. However, remnant natural ecological impacts created by wild- States. Based on models, the popu- areas within the landscape remain fire suppression techniques. The lation is expected to increase at susceptible to lightning-ignited long-term impacts of logging roads a rate greater than 28 of the 50 fires, arson, or accidental fires. In on water quality and fish habitat States by the year 2020 (Burchell these areas, wildfires (defined in are well known, but the impacts and others 1999). The increased this paper as any fire not meeting of plow lines on plant and animal urbanization of southeast Florida management objectives, including communities are mostly unknown has resulted in rapid suppression lightning-ignited fires, human- (Ingalsbee 2004). The objective of of wildfires to save structures. At caused fires, and fire that escapes the same time, there has been a the boundary of a prescribed burn) decrease in prescribed burning due to complaints of smoke, pressure often require quick suppression Although the total techniques to prevent loss of prop- on public officials, and a general erty and potential loss of human number of plow lack of knowledge pertaining to life. lines constructed fire ecology by people moving to and area of plow- Florida from other regions. As The primary method for controlling prescribed burning decreases and line disturbance from fuel loads increase, it is inevitable wildfires in Florida is the use of a wildfire suppression tractor and attached fire plow to that wildfires will be more severe create bare mineral soil fire breaks. techniques during a in southeast Florida, threatening Although highly effective in con- wildfire may appear to and possibly destroying homes and trolling wildfires, plow lines create be insignificant, site- other structures. trenches that result in major soil specific impacts may be disturbances, continuous berms, The State of Florida continues to substantial. Generally, purchase preservation lands amid the damage to natural the rapid urbanization. Remnant Jeffrey Hutchinson is a former district resources caused by natural communities in Florida biologist for the Florida Park Service and are often purchased because they is now a Ph.D. student in weed science at the University of Florida Center for Aquatic fire plows is greater and represent unique and rare habitats, and Invasive Plants in Gainesville, FL. longer lasting than the many of which are fire-dependent Richard Roberts is retired from the Florida and contain multiple rare species of Park Service in Hobe Sound, FL, where he effects of a wildfire. � was a biological scientist. plants and animals. Management of
Fire Management Today 32 these urban natural areas becomes sequent buildup of fuels, and the and major roadways that include more difficult as residential devel- limitations of using prescribed fire Interstate 95 and U.S. Highway 1. opments are built directly adjacent in the close vicinity of residential to conservation lands. Many smaller homes make many natural areas The dominant fire-dependent com- natural areas less than 12,000 in Florida susceptible to lightning- munity at the Savannas Preserve acres (4,800 hectares) with fire- ignitedd, arson, or accidental fires. State Park (SPSP) is mesic pine dependent natural communities are flatwoods. This community has a partially surrounded by residential Study Areas fire-return interval of approximate- communities, major highways, ly 1 to 8 years (FNAI 1990), indi- Data were collected from seven and other smoke-sensitive sites cating that fire could potentially wildfires that occurred in Savannas (i.e., schools, hospitals, and nurs- burn through this habitat every Preserve State Park (4,997 acres ing homes). Because of the altera- 1 to 8 years. The flatwoods plant [2,023 hectares]) in St. Lucie tion of habitat, decrease in surface community of this park is highly County and Jonathan Dickinson hydrology, increased fuel loads, and diverse. Dominant plants are saw State Park (11,466 acres [4,642 intensification of urbanization, it palmetto (Serenoa repens), gall- hectares]) in Martin and Palm is now nearly impossible to achieve berry (Ilex glabra), and slash pine Beach Counties, FL, from February the historical natural fire regime in (Pinus elliottii), with an understory 1999 to February 2001. Both parks many of Florida’s natural commu- of wire grass (Aristida spp.), chalky are located near urban areas in nities through the use of prescribed bluestem (Andropogon virginicus coastal southeast Florida, one of burning, even with the most strict var. glaucus), milkworts (Polygala the most highly populated areas burn prescriptions and smoke spp.), rose gentian (Sabatia spp.), in the State, and are partially sur- management. The lack of naturally meadow beauty (Rhexia spp.), and rounded by residential development occurring landscape fires, the sub- yellow-eyed grass (Xyris spp.).
Quantitative effects of plow-line disturbance during seven wildfires at Savannas Preserve State Park (SPSP) and Jonathan Dickinson State Park (JDSP) in southeast Florida from February 1999 to February 2001.
Ratio Burnt Area, Plow-Line Acres (ha) Area Plow-Line Plow-Line Volume, to Type of Burned, Number of Length, Area, Acres Cubic Yards Plow-Line, Park Fire Acres (ha) Plow Lines Miles (km) (ha) (m3) Miles (km) 4,019 SPSP1 Escape 135.1 (64.7) 48 2.1 (3.4) 1.2 (0.5) 64:1 (16:1) (3,675) SPSP Arson 6.2 (2.5) 8 0.4 (0.6) 0.3 (0.1) 921 (842) 14:1 (3.5:1) 2,751 SPSP Arson 41.3 (16.7) 30 1.6 (2.6) 1.0 (0.4) 26:1 (6.5:1) (2,515) SPSP Arson 5.9 (2.4) 10 0.4 (0.6) 0.5 (0.2) 1,044 (955) 16:1 (4:1) 1,804 SPSP Arson 55.6 (22.5) 41 0.8 (1.3) 0.5 (0.2) 75:1 (19:1) (1,650) 2,140 SPSP Lightning 8.4 (3.4) 15 0.8 (1.3) 0.5 (0.2) 11:1 (2.8:1) (1,957) 7,714 50:1 JDSP2 Lightning 189.9 (76.8) 50 3.8 (6.1) 3.2 (1.3) (7,054) (12.6:1) 442.5 20,454 mean=37:1 TOTALS 202 9.8 (15.8) 7.2 (2.9) (179.1) (1,870) (9.3:1)
1SPSP=Savannas Preserve State Park. 2JDSP=Jonathan Dickinson State Park.
Volume 69 • No. 3 • 2009 33 Jonathan Dickinson State Park by width of each plow line. Volume roadways. The mean area burned (JDSP) covers approximately 9,200 of soil displaced was calculated by per wildfire was 63 acres (26 hect- acres (3,723 hectares) in southeast multiplying length by width by ares). Wildfire suppression activities Florida, with pine flatwoods and depth. Depth was defined as the resulted in a total of 202 plow lines sand pine scrub as the dominant height from the top of the soil for all seven fires with an average fire-dependent communities. The mound to the bottom of the trench. of 29 plow lines constructed per sandhill community is also impor- Ratio of area burned to plow line wildfire. For all seven wildfires, the tant at JDSP because it is at the length was calculated as an index of estimated total area disturbed from limit of its southern range. The disturbance. wildfire suppression was 7.2 acres wildfire at JDSP burned largely (2.9 hectares). in the sand pine scrub and san- Results dhill communities. The sand pine Plow-line length for all wildfires Seven wildfires were documented scrub fire return interval is 20 to totaled 9.8 miles (16 kilometers). in the two parks over a 24-month 60 years (Roberts and Cox 2000) Plow-line lengths varied from 18 to period from 1999 to 2001, burning and the sandhill community fire 3,459 feet (5 to 1,057 meters), with a total of 442 acres (179 hectares), return interval is 2 to 5 years (FNAI a mean of 254 feet (78 meters). or approximately 2.7 percent of the 1990). The scrub is characterized Forty-eight percent of the plow total area of both parks (see table). by a closed-canopy forest of sand lines were attributed to the two Of the seven wildfires, four were pines (Pinus clausa) with a dense largest fires, one ignited by light- caused by arson, two were ignited understory of oaks (Quercus spp.), ning and one escaped during a by lightning, and one escaped dur- and the sandhills have widely prescribed burn. The average area ing a prescribed burn. The wildfires spaced slash pines and turkey oaks disturbance for each wildfire was occurred primarily in mesic pine (Quercus laevis) with a sparse 1 acre (0.4 hectares) and ranged flatwoods habitat (in the SPSP) and understory of scrub oaks (Roberts from 0.3 to 3.2 acres (0.1 to 1.3 sand pine scrub and sandhill habi- and others 2006). hectares). Plow-line depths ranged tat (in the JDSP). All seven wildfires from 0.5 to 3.5 feet (0.2 to 1.1 required immediate suppression meters), with a mean of 0.6 feet Disturbance Estimates because of the close proximity of (0.2 meters). Width of plow lines Data were collected from wildfires homes and major and secondary varied from 1.5 to 12 feet (0.5 to over a 24-month period. The area of each wildfire was determined by walking the perimeter of the burn area with a global position- ing system (GPS); plow lines cre- ated to suppress wildfires were mapped similarly by walking each line with a GPS (fig. 1). Plow-line depth (estimated from the top of the soil mound to the lowest point in the bottom of the trench) and width were estimated in the field with a yard stick and measuring tape (fig. 2), then transformed using PathFinder Office software and downloaded into ArcView 3.1 geographical information system (GIS). Length of each plow line was obtained from ArcView GIS shape- files.
Area of plow-line disturbance was Figure 1—Mapping plow-line disturbance with GPS in a pine flatwoods community at calculated by multiplying length Savannas Preserve State Park. Photo: Jeffrey T. Hutchinson.
Fire Management Today 34 natural communities by pooling water or directing water along the furrowed path. They prevent the sheet and ground water flow that naturally occur in the low-relief wetland and upland communi- ties of south Florida and that are particularly important in southern Florida where elevation changes of less than 3 feet result in differ- ent vegetation types and habitat. These alterations may also reduce hydroperiods and prolong dry sea- son droughts (Taylor and Gibbons 1985).
Increases in water levels and raised Figure 2—Plow lines created during wildfire suppression in xeric communities of soil levels along plow lines create south Florida create long-lasting disturbances that persist for decades. Photo: Jeffrey T. habitat for invasive plants that can Hutchinson. alter the structure and composition of natural communities, includ- 3.7 meters), with a mean of 1.7 feet with a mean of 37:1 acres per mile ing pine flatwoods, dry prairies, (0.5 meters). In many instances, (9:1 hectares per kilometer), indi- wet prairies, and seasonal ponds. multiple plow lines were construct- cating that on average, for every 37 Southeast Florida is highly sus- ed adjacent to other lines, increas- acres (15 hectares) burned, 1 mile ceptible to invasive plants (Gordon ing the total width. (1.6 kilometers) of plow line was 1998, Langeland and Craddock- put in to control the wildfire. The Burks 1998). Old World climbing The range of area disturbed per highest ratios were recorded on the fern (Lygodium microphyllum), plow line was between 7.4 and three smallest wildfires, while the melaleuca (Melaleuca quinquen- 3,469 square feet (6.2 to 2900 lowest ratios were recorded for the ervia), Brazilian pepper (Schinus square meters), with a mean of 147 two largest wildfires. terebinthifolius), torpedograss square feet (13.6 square meters). (Panicum repens), and cogon Total soil displacement per plow Discussion grass (Imperata cylindrica) are line ranged from 1.2 to 1,734 cubic Although the total number of often the first plants to invade yards (0.9 to 1,326 cubic meters), plow lines constructed and area of recently disturbed areas, and they with a mean of 77 cubic yards (59 plow-line disturbance from wildfire may eventually spread into nearby cubic meters). During the 7 wild- suppression techniques during a undisturbed habitat (Langeland and fires, an estimated total of 20,393 wildfire may appear to be insig- Craddock-Burks 1998). Some may cubic yards (15,591 cubic meters) nificant, site-specific impacts may be spread by the tractor and plow of soil was displaced, a volume be substantial (Ingalsbee 2004). by its dragging rhizomes, seeds, that is approximately equivalent to The plow used to create fire breaks spores and other fragments along 7,800 full-size pickup truckloads repositions the first 1 to 3 feet (0.3 the plow line. In south-central of soil. In many areas, disturbance to 0.9 meters) of the soil, rhizomes, Florida, for example, invasive plants from the tractor (e.g., maneuvering roots, and other vegetative debris to were documented along plow lines and turning) and staging of equip- one side of the plow line, creating a created during wildfire suppres- ment also resulted in severe soil berm and channel. In hydric habi- sion in an area where none were and vegetation disturbance; howev- tat, creating plow lines is analogous previously known (Hutchinson and er, this disturbance was not includ- to ditching and draining wetlands. Menges 2006). Additionally, during ed in the analysis. Ratio of area Trenches created by plow lines, wildfires, personnel and equipment burned to plow-line length ranged regardless of depth, act as chan- staging areas typically occur in from 11:1 to 64:1 acres per mile nels and can alter the hydrology of open, disturbed areas that contain (3:1 to 16:1 hectares per kilometer), a high percentage of invasive plants
Volume 69 • No. 3 • 2009 35 (Ingalsbee 2004) that can easily be agency and a representative of approximately $31 million to treat spread into the interior portion of the natural area, such as the land 277,998 acres (112,550 hectares) of natural areas on the clothing of manager or park biologist, should public conservation lands infested firefighters or equipment. coordinate fire suppression activi- with exotic plants (Drew Leslie, ties. Environmental impacts can be FDEP, personal communication). In SPSP, numerous plow lines minimized by utilizing natural bar- The combined deleterious effects of from past wildfire suppression riers such as wetlands (Backer and plow lines, invasive plants, reduc- were observed following prescribed others 2004), but this requires the tion in prescribed fire, increases burns conducted between 1999 guidance of someone familiar with in wildfires, negative recreational and 2001 (J.T. Hutchinson, per- the site. activities, feral hogs, and residential sonal observation). At one site in development on remnant natu- SPSP, the number of plow lines It is important that fire manage- ral communities in south Florida and berms from past wildfire sup- ment objectives based on protec- result in greater increases in edge pression efforts resembled bedding tion of human life and property effects and an increase in frag- preparation for row crops. Wildfire and those based on maintaining mentation on sites that are already suppression creates trenches that remnant natural communities and insular. Limited data exist on the remain an unsightly scar for many biodiversity both be obtained with effects of plow-line disturbance years (FDEP 1999). The bulldozer minimal conflict (Caling and Adams from wildfire suppression, and lines and fire plow impacts from 1999). However, the results of this more detailed research is needed, suppression of a 1971 wildfire at study indicate that wildfire suppres- not only in Florida, but throughout JDSP are still visible 36 years later, sion activities cause disturbances to fire-dependent communities. particularly in xeric communities natural communities, especially in (R.E. Roberts, personal observa- small remnant sites. The ecological Acknowledgments tion). The Florida Department considerations of small remnant This project was supported with of Environmental Protection has natural areas in southeast Florida funding provided by the Florida guidelines in place for re-contour- are often disregarded (Caling and Park Service. We would like to ing plow lines that result from Adams 1999), although continual thank Pete Scalco and George Jones wildfire (FDEP 1999); however, it is plow-line disturbances during wild- of the Florida Park Service for sug- unknown how strictly these guide- fire suppression over decades will gesting the need to document the lines are followed. eventually lead to major structural, fire-line suppression activities in compositional, and hydrological State parks. We also thank Lorraine During the wildfire reported here changes to remnant natural com- Roberts for reviewing and com- in JDSP, we observed a plow line munities, in particular the pine menting on the manuscript. constructed along the ecotone of flatwoods that are dependent on a flatwoods/cypress strand. The frequent fire. cypress strand was inundated with References several inches of water, which Generally, the damage to natural Abrahamson, W.G.; Hartnett, D.C. 1990. Pine flatwoods and dry prairies. In: would have acted as a natural fire resources caused by fire plows is Myers, R.L.; Ewel, J.J., eds. Ecosystems break; however, the operator of greater and longer lasting than the of Florida. Orlando, FL: University of the tractor and attached plow was effects of a wildfire. Because of the Central Florida Press: 103–149. Backer, D.M.; Jensen, S.E.; McPherson, unfamiliar with the habitat and high maintenance costs and envi- G.R. 2004. Impacts of fire-suppression trenched a plow line where none ronmental damage, it is possible activities on natural communities. was necessary. A potential solution that the effects of wildfire suppres- Conservation Biology. 18: 937–946. to excessive plow-line disturbance sion may be more costly over time Burchell, R.W.; Newman, N.; Zakrewsky, A.; DiPetrillo, S.E. 1999. Eastward Ho! during wildfire suppression is for than the actual fire (Taylor and Development futures: paths to more land managers and fire suppres- Gibbons 1985). This is especially efficient growth in southeast Florida. sion officials to meet at least annu- true when invasive plants invade New Brunswick, NJ: Rutgers University, Center for Urban Policy Research: 276 p. ally and discuss the logistics and the plow lines and spread out into Caling, T.M.; Adams, R. 1999. Ecological operation procedures for wildfire undisturbed areas. From 1997 to impacts of fire suppression operations in suppression in each specific parcel 2005, the Florida Department of a small vegetation remnant. In: Lunt, I.; Green, D.G.; Lord, B., eds. Proceedings of land. During a wildfire, a rep- Environmental Protection (FDEP), of the Australian Bushfire Conference; resentative of the fire suppression Bureau of Invasive Plants, spent Albury, Australia. Available at: Fire Management Today 36 www.csu.edu.au/special/bushfire99/ Hutchinson, J.T.; Menges, E.S. 2006. Roberts, R.E.; Cox, A.C. 2000. Sand pine papers/caling/> (accessed April 2009). Evaluation of the invasiveness of non- scrub vegetation response to two burn- Florida Department of Environmental native plants at Archbold Biological ing and two non-burning treatments. In: Protection [FDEP]. 1999. Firebreaks. Station. Florida Scientist. 69: 62–68. Moser, W.; Moser, F., eds. Fire and forest Unpublished document, DEP Program Ingalsbee, T. 2004. Collateral damage: The ecology: Innovative silviculture and veg- Directive 910. Tallahassee, FL: environmental effects of firefighting (the etation management. Tall Timbers Fire Florida Department of Environmental 2002 Biscuit Fire suppression actions Ecology Conference Proceedings, No. 21. Protection. 5 p. and impacts). Eugene, OR: Western Fire Tallahassee, FL: Tall Timbers Research Florida Natural Areas Inventory [FNAI]. Ecology Center. Available at: Volume 69 • No. 3 • 2009 37 lessons FroM the hayMan Fire: Rocky Mountain Forest understory resPonses Research Station to the sCariFy-and-seed PostFire rehabilitation treatMent Paula J. Fornwalt n unburned forests, organic plant do not establish densely enough to the scarify-and-seed treatment by litter and live vegetation help be effective at controlling runoff comparing understory establish- Istabilize the soil and promote and erosion (Robichaud and others ment at two neighboring sites, one water infiltration. Much of this 2006, Wagenbrenner and others which burned but had no postfire plant material is consumed dur- 2006). rehabilitation treatment and the ing severe wildfires, leaving the other which burned and was subse- bare ground susceptible to elevated Consequently, managers are quently scarified and seeded. postfire water runoff and soil ero- increasingly using other treat- sion (Shakesby and Doerr 2006). ments, either alone or in conjunc- The 2002 Hayman Severe wildfires can also produce tion with seeding; for example, Fire: A Research a water-repellant layer in the soil seeding was combined with soil Opportunity that further decreases water infil- scarification on several large burns tration (DeBano 2000). Even after in the Colorado Front Range. The Colorado Front Range has moderate rain events, runoff and Scarification is a mechanical soil experienced several large and erosion in severely burned areas treatment that aims to increase severe wildfires since the mid- can cause extensive and costly dam- water infiltration by roughening 1990s. These fires likely represent age to roads, buildings, reservoirs, up the soil surface and disturb- a shift in fire regimes—from one and ecosystems (Beyers and others ing the water-repellant soil layer of mixed severity to one of high 1998). Land managers often prescribe Land managers often prescribe seeding seeding treatments immediately treatments immediately after wildfire in an � after wildfire in an attempt to mini- attempt to minimize potential damage. � mize this potential damage. This relatively inexpensive rehabilitation This rehabilitation treatment aims to stabilize treatment aims to stabilize the soil the soil and decrease water runoff by rapidly and decrease water runoff by rap- increasing vegetative ground cover. � idly increasing vegetative ground cover. Exotic grasses, including orchardgrass (Dactylis glomerata), (Robichaud and others 2003). When severity—that is partly a result timothy (Phleum pratense), and combined with seeding, scarifica- of fire suppression and graz- wheat (Triticum aestivum), are tion may also help keep seeds on ing activities since the late 19th frequently used because seeds of site until germination occurs. century (Brown and others 1999, these quick-growing species are Unfortunately, the effects of the Kaufmann and others 2001). All readily available (Robichaud and scarify-and-seed treatment on these fires occurred partly in the others 2000). However, scientists postfire runoff and erosion and wildland-urban interface, where and managers are beginning to on regenerating understory plant human values were at risk from realize that seeded species often communities have not been closely both fires and subsequent flooding analyzed. I seized an opportunity and erosion. To date, the largest Paula Fornwalt is an ecologist with the created by the 2002 Hayman Fire and most severe fire known to burn Forest Service, Rocky Mountain Research Station, Fort Collins, CO. to assess the ecological impacts of in the Front Range was the June Fire Management Today 38 (a) (b) rimpaui, a wheat-rye hybrid) was applied to the scarified area at a tar- get density of 26 seeds per square foot (280 seeds per square meter). Both species are exotic annual cere- al grasses that typically germinate quickly but persist on the landscape for only a few years. In 2004, understory plant data were collected in unrehabilitated and rehabilitated study sites within the Hayman Fire area (fig. 1b). The unrehabilitated site and the twenty 0.25-acre (0.1-hectare) upland plots within it were originally established in 1996 as part of other research Figure 1—Burn severity map of the Hayman Fire (a) and locations of Forest Service activities (Fornwalt and others postfire rehabilitation treatments (b). The burn severity map was derived by the Forest Service from a SPOT4 satellite image and is largely based on overstory tree mortality 2003, 2009; Kaufmann and others (Robichaud and others 2003). 2000). Three of these plots were moderately or severely burned and 2002 Hayman Fire, which burned (5,300 hectares) and cost over $1.5 were incorporated into this study. 137,600 acres (55,800 hectares) of million to apply. Soil scarification I also established a rehabilitated land dominated by ponderosa pine was done along terrain contours, study site 2 miles (1.2 kilometers) (Pinus ponderosa) and Douglas-fir either by all-terrain vehicles pull- to the east of the unrehabilitated (Pseudotsuga menziesii). Over 50 ing harrows (fig. 2) or by hand with site. The rehabilitated site is in an percent of the Hayman burned as a McLeods. After scarification, a cer- area that was moderately to severe- moderate-severity to high-severity tified weed-free mixture of 70 per- ly burned by the Hayman Fire and fire (fig. 1a), with complete or cent barley (Hordeum vulgare) and was subsequently scarified and nearly complete overstory mortal- 30 percent triticale (×Triticosecale seeded (fig. 3); the site contains 14 ity and extensive crown, litter, and duff consumption (Robichaud and others 2003). Most of the Hayman Fire occurred on National Forest System land. The steep, dissected topography and the highly erodible granitic soils common throughout the Hayman Fire area made moderately and severely burned areas extremely susceptible to elevated postfire runoff and erosion. Therefore, the Forest Service rehabili- tated approximately 32,000 acres (12,800 hectares) of moderately and severely burned forest in the months following the fire (fig. 1b), at a cost of more than $16.5 mil- Figure 2—The scarify-and-seed treatment was implemented on 13,200 acres (4,340 lion (Robichaud and others 2003). hectares) of moderately to severely burned Forest Service land in the months following the Hayman Fire. Scarification was done in rows either by all-terrain vehicles pulling The scarify-and-seed treatment harrows (as shown in the photo) or by hand with McLeods. Barley and triticale seeds were was implemented on 13,200 acres spread on the scarified areas. Photo: Pete Robichaud. Volume 69 • No. 3 • 2009 39 upland plots. The two sites are sim- ilar in elevation, topography, and pre-fire overstory structure. In each plot, the presence of all vascular plant species was recorded, and per- centage vegetative cover by species was estimated in ten 11-square-foot (1-square-meter) subplots located within the plot. Only live plants were included in the surveys, with the exception of the seeded grasses; dead seeded grasses were also mea- sured because of their potential to affect understory establishment. We tested for differences in understory response variables between sites using multiresponse permutation procedures, a nonparametric alter- native to analysis of variance. Figure 3—Seeded grass establishment was generally imperceptible in scarified and seeded areas of the Hayman Fire, although in this area, establishment was uncharacteristically Impacts of the Scarify- successful. The photo was taken in the summer of 2004, a year and a half after the and-Seed Rehabilitation treatment was applied. Photo: Paula Fornwalt. Treatment on the Forest Understory Scarification is a mechanical soil treatment that The Rehabilitation Treatment aims to increase water infiltration by roughening Of the two grasses seeded, only triticale was found growing in the up the soil surface and disturbing the water- plots; barley was never encoun- repellant soil layer. When combined with seeding, tered. Triticale was present at both scarification may also help keep seeds on site until sites. At the unrehabilitated site, germination occurs. live and dead triticale were found in 33 percent of the plots, while at the rehabilitated site, live triticale mineral soil) was visible at survey Other researchers monitoring the was found in 64 percent of the plots time. effectiveness of the scarify-and-seed and dead triticale was found in 71 treatment in the Hayman Fire area percent. It is unknown how triticale I suspect that unfavorable weather also found that seeding did not arrived at the unrehabilitated site; conditions during 2002 and 2003 increase vegetative cover (Rough seeds may have been inadvertently were at least partially responsible 2007). In addition, they found that dropped from aircraft while seed- for the poor establishment of the average depth of scarification ing other portions of the burn, or seeded grasses on upland slopes. after the Hayman Fire was less they may have dispersed in from These summers were among the than 1 inch (2.5 centimeters), and, upslope rehabilitated areas. The warmest and driest on record in therefore, scarification alone was cover of both live and dead triticale the Colorado Front Range; high not sufficient to break through the was similar between the two sites temperatures and lack of moisture hydrophobic layer that extended (p = 0.599 and 0.484, respectively). may have killed many seeds and nearly 4 inches (10 centimeters) Live and dead triticale, combined, germinated seedlings. The summer into the soil profile. As a result, averaged less than 0.5 percent and of 2004 then brought several high- the treatment was not effective in never exceeded 3.0 percent in any intensity rainfall events that likely reducing sediment yields imme- single plot. No visible evidence of washed away much of the remain- diately after the fire; indeed, the the postfire scarification treatment ing viable seed. mechanical disturbance caused (for example, recently disturbed by scarification may have even Fire Management Today 40 increased sediment movement surviving underground parts or by 1981, Keeley 2004, Schoennagel immediately after the treatment geminating from seeds in the soil and Waller 1999). was implemented. seedbank (USDA Forest Service 2009). Indeed, native plant richness Impacts on Exotic Invaders Impacts on Native Species and cover at both sites were similar Fourteen exotic species were found A total of 128 native plant species to that observed in the unrehabili- in the plots (excluding triticale). were found within the two sites. tated site before the fire, suggesting In general, the number and cover Native plant richness and cover that understory recovery in this of exotic species per plot were low, were similar at the rehabilitated system can naturally occur in as lit- with an average exotic richness of and unrehabilitated sites (p = 0.256 tle as 2 years (Fornwalt and others four species per plot and an aver- and 0.381, respectively), with native 2003, 2009). Others also have found age exotic cover of only 0.6 percent. richness averaging 42 species per that low levels of seeded grass Neither exotic plant richness nor plot and native cover averaging 19 cover did not affect native species cover differed significantly between percent. in fire-adapted ecosystems (Keeley sites (p = 0.835 and 1.000, respec- and others 1981), but high levels of tively). Many of the native species found in seeded grass cover decreased native the rehabilitated and unrehabilitat- plant establishment and growth Six of the exotic species, cheatgrass ed sites are perennial species that in many cases (Keeley and others (Bromus tectorum), musk thistle are abundant throughout unburned Front Range forests, including the graminoids Ross’ sedge (Carex rossii), prairie Junegrass (Koeleria macrantha), and mountain muhly (Muhlenbergia montana); and the shrub kinnikinnick (Arctostaphylos uva-ursi). The perennial forbs hairy false goldenaster (Heterotheca vil- losa), prairie bluebells (Mertensia lanceolata), and eastern pasque- flower (Pulsatilla patens) are also abundant in the Front Range (fig. 4). Each of these species occurred in more than 75 percent of plots studied here, and together, they constituted nearly 30 percent of the native plant cover. The cover of each species appears to have been unaffected by the scarify-and-seed treatment (p > 0.250 in all cases). These findings suggest that the scarify-and-seed treatment had little to no impact on native plant establishment after the Hayman Fire. This is not surprising, given that there was likely little com- Figure 4—The native species prairie bluebells (top left), kinnikinnick (top center), and petition from the seeded grasses. eastern pasqueflower (top right) are abundant throughout the Colorado Front Range and appear to have been unimpacted by the scarify-and-seed rehabilitation treatment. Furthermore, many of the native Canadian thistle (bottom left), mullein (bottom center), and musk thistle (bottom right) understory species are adapted to were three of the six noxious weed species found in the unrehabilitated and rehabilitated regenerate quickly after distur- sites, though they were uncommon and do not appear to have been stimulated by the bance, either by sprouting from postfire rehabilitation activities. Kinnikinnick photo by Laurie Huckaby; all other photos by Paula Fornwalt. Volume 69 • No. 3 • 2009 41 (Carduus nutans), Canadian thistle by the small number of plots References (Cirsium arvense), Saint Johnswort and lack of study site replication, Beyers, J.L.; Wohlgemuth, P.M.; Wakeman, (Hypericum perforatum), butter- they nonetheless suggest that the C.D.; Conard, S.G. 1998. Does ryegrass and-eggs (Linaria vulgaris), and scarify-and-seed treatment had seeding control postfire erosion in chaparral? Fire Management Notes. 58: mullein (Verbascum thapsus), are little to no effect on understory 30–34. noxious weeds in Colorado (see fig. recovery after the Hayman Fire. Brown, P.M.; Kaufmann, M.R.; Shepperd, 4). All these species had negligible This is likely because the scarify- W.D. 1999. Long-term, landscape cover: butter-and-eggs cover was and-seed treatment effects were patterns of past fire events in a mon- tane ponderosa pine forest of central less than 0.5 percent, while all nearly imperceptible: seeded grass Colorado. Landscape Ecology. 14: other species had less than 0.1 per- establishment was negligible, and 513–532. cent cover each. None of the spe- signs of soil scarification were not DeBano, L.F. 2000. The role of fire and soil heating on water repellency in wildland cies appear to have been affected by visible a year and a half after treat- environments: A review. Journal of the scarify-and-seed treatment (p > ment was applied. Furthermore, Hydrology. 231–232: 195–206. 0.500 in all cases). other researchers working in the Fornwalt, P.J.; Kaufmann, M.R.; Huckaby, L.S.; Stoker, J.M.; Stohlgren, T.J. 2003. Hayman Fire found that the scarify- Non-native plant invasions in managed Exotic species often increase fol- and-seed treatment had no impact and protected ponderosa pine/ Douglas- lowing disturbances by taking on postfire erosion. In light of the fir forests of the Colorado Front Range. advantage of reduced competition considerable cost and increasingly Forest Ecology and Management. 177: 515–527. for resources, and management widespread use of this rehabilita- Fornwalt, P.J.; Kaufmann, M.R.; Huckaby, activities such as firefighting and tion treatment, it is imperative that L.S.; Stohlgren, T.J. 2009. Effects of rehabilitation may also introduce researchers and managers continue land use history on understory plant communities in a montane Colorado exotic species to locations where to learn more about both the effec- forest. Plant Ecology DOI: 10.1007// they did not previously occur tiveness and the ecological impacts s11258-008-9513-z. Abstract online (Keeley and others 2006). Although of scarification and seeding after at Fire Management Today 42 Keeley, S.C.; Keeley, J.E.; Hutchinson, Robichaud, P.; MacDonald, L.; Freeouf, J.; Shakesby, R.A.; Doerr, S.H. 2006. Wildfire S.M.; Johnson, A.W. 1981. Postfire suc- Neary, D.; Martin, D.; Ashmun, L. 2003. as a hydrological and geomorphologi- cession of the herbaceous flora in south- Postfire rehabilitation of the Hayman cal agent. Earth-Science Reviews. 74: ern California chaparral. Ecology. 62: Fire. In: Graham, R.T., ed. Hayman 269–307. 1608–1621. Fire case study. Gen. Tech. Rep. RMRS- U.S. Department of Agriculture [USDA], Robichaud, P.R.; Beyers, J.L.; Neary, D.G. GTR-114. Fort Collins, CO: Forest Forest Service. 2009. Fire Effects 2000. Evaluating the effectiveness of Service, Rocky Mountain Research Information System. Missoula, MT: postfire rehabilitation treatments. Gen. Station. Pp. 293–313. Rocky Mountain Research Station, Fire Tech. Rep. RMRS-GTR-63. Fort Collins, Rough, D. 2007. Effectiveness of rehabilita- Sciences Laboratory. Available at Volume 69 • No. 3 • 2009 43 Wildland Fire behavior and “the Course oF sCienCe” FloWChart: is there a ConneCtion? Martin E. Alexander ’ve been involved in wildland fire since 1972. Except for a couple of Iseasons as a hotshot crew member followed by another season with the Forest Service in wilderness fuel inventory capped off by some slash burning, all that time has been spent in fire research. Even as a wildland fire researcher, I’ve kept actively involved in observing and analyzing free-burning wildfires over the years, and I’ve occasion- ally served as an operational fire behavior specialist on major fires Wildland fires are highly volatile, multidimensional phenomena, not always easily and multifire incidents in north- observed, monitored, explained, and documented. Photo: Martin E. Alexander, Canadian ern Alberta and the Northwest Forest Service, Northern Forestry Centre, Edmonton, AB, 1981. Territories. This focused fire background has helped me under- stand that wildland fires are not The Course of Science always easily observed, monitored, Sparse and Infrequent explained, or documented. Observations Observational Incorrect Interpretation Errors of Observations I came across “The Course of Theoretical Science” flowchart on a coffee Misunderstanding room bulletin board in 1990 dur- Management Directives Oversimplified ing a 3-year (1989–92) sojourn in Models Australia. Over the years, I’ve come Further Refinement of to appreciate the humor and cyni- Computer Models Controversy Unimportant Details cal nature embedded in the “Course of Science” flowchart more and Code Errors Unrealistic more. But perhaps of greater value Assumptions is this flowchart’s ability to remind Crude Diagnostic Confusion us of the traps to which we, in the Tools Further Misunderstanding research and development com- Coincidental Agreement Between munity, and in turn the users of the Theory and Observations Dr. Marty Alexander is an adjunct profes- Publication sor of wildland fire science and manage- ment in the Department of Renewable Resources at the University of Alberta, Edmonton, Alberta, Canada. He was on In this author’s opinion, the relevancy of the “Course of Science” flow- a leave of absence from his position as a chart to wildland fire science, and more specifically to fire behavior, senior fire behavior research officer with is probably a far better fit to the general course of research, devel- the Canadian Forest Service, Northern Forestry Centre, in Edmonton, when he opment, and application in this field than most of us would care to wrote this article. admit. Fire Management Today 44 knowledge and products generated An Example Related to Extreme by fire researchers, can so easily fall victim. As Dr. Mary Omodei (2009), Wildland Fire Behavior a wildland fire psychologist, has pointed out, this flowchart “charac- Williams (2007) reported on an interesting hypothesis regarding terizes not only everyday intuitive extreme fire behavior associated with the wildland conflagrations thinking but also science as well in that descended on Canberra, Australia, on 18 January 2003 ( Volume 69 • No. 3 • 2009 45 Arnold, R.K.; Buck, C.C. 1954 Blow-up Additions to the “Course of fires—silviculture or weather prob- lems? Journal of Forestry. 52(6): Science” Flowchart Over Time 408-411. Available at: Fire Management Today 46 Fuels treatMents and Fire Models: � errors and CorreCtions J. Morgan Varner and Christopher R. Keyes ire behavior and fire effects Crookston 2003). Development of models are arguably among the Errors in modeling both fire behavior and effects mod- Fmost important tools in fire fuels treatment, fire els (e.g., Crown Fire Initiation and and fuels management. Given the Spread [CFIS]; Alexander 2007) power, accessibility, and utility of behavior, and fire continues. these models, fuels planners and effects can often be scientists commonly use them to tied to unsupported McHugh (2006) reviewed the use compare potential fire intensity and assumptions about of fire behavior and effects models severity on planned and unplanned by resource managers and fire sci- wildland fires. actual conditions and entists. Fuels managers use these over-reliance on default behavior and effects models in fuels How well the models are run is values. � treatment planning to compare another matter. Modeling errors the results of proposed treatments. in fuels treatment proposals and Fire and fuels scientists also use scientific papers can exaggerate or fuels treatments according to the these models to compare outcomes mischaracterize the effectiveness of “activity fuels” that these treat- on modeled fire behavior and to potential fuels treatments designed ments leave behind. In the Western understand long-term changes in to abate hazardous fire behavior. United States, fuels treatments are living and dead fuels. Between 1996 Unrealistic outputs can typically typically designed to reduce surface and 2009 alone, at least 19 papers be traced to unrealistic inputs, so fireline intensity, prevent crown were published in fire science and close analysis of common input ignition, and interrupt canopy fire forestry journals and conference errors can suggest best practices to spread; several models have the proceedings that used these mod- minimize modeling problems and ability to predict changes in these els to predict fire behavior and/or maximize treatment effectiveness. behaviors, many based on equations effects following fuels treatment. Beyond this, the revision of old developed by Rothermel (1972) and models and the design of new mod- Van Wagner (1977) for surface and Many proposed fuels treatments els can promote ease of use and crown initiation and spread. Fire and some of the recent scientific more realistic outputs in complex effects models, on the other hand, literature contain examples of the and extreme input conditions. predict the consequences of fires, inexact application of these deci- generating estimates of tree mor- sion support tools—or more simply Modeling Fuels tality, soil heating, erosion, fuels stated: “user errors.” For example, consumption, and emissions. Treatments a recent study by Jolly (2007) ana- lyzed the sensitivity of BehavePlus Fire and fuels managers and plan- In a recent survey of fire and fuels to changes in live fuel moisture ners use modeling software to managers by Miller and Landres content, finding that users may predict changes in fire behavior (2004), among the most used fire commit errors if faulty assumptions resulting from surface and canopy behavior models were BehavePlus, are made. Other errors that may FARSITE, and NEXUS. Among fire occur can be attributed to faulty J. Morgan Varner is an associate profes- effects models, several of the most assumptions regarding fuel mois- sor of wildland fire management and used were First Order Fire Effects director of the Wildland Fire Laboratory, ture, wind adjustment, fuel model Department of Forestry and Wildland Model (FOFEM), Consume, the Fire selection, fuel decomposition rates, Resources, Humboldt State University, Emission Production Simulator fuel load estimates, foliar moisture Arcata, CA. Christopher R. Keyes is a (FEPS), Fuels Management Analyst research associate professor of silviculture content, and the patchiness of fuels. and director, Applied Forest Management Plus, and the Forest Vegetation By examining these factors, future Program, The University of Montana, Simulator (FVS)-Fire and Fuels errors by fire and fuels managers Missoula, MT. Extension (FFE) (Reinhardt and and scientists may be avoided. Volume 69 • No. 3 • 2009 47 Best Practices for Modeling Fuels Treatment Effects 1. � Fully disclose all inputs and assumptions. Users should list all value assignments that were used in modeling (fuel moisture, loading, depths, and data sources). The greater the disclosure, the greater the result’s repeatability and level of trust. 2. � Use field data as much as possible; use published estimates when necessary. Field-derived values for fuel loading, depth, and moisture are always preferred. When these are not available and data collection is not possible, published values can be used but must be identified. 3. � Project results over time. Treatment comparisons made immediately following fuels treatments may poorly characterize mid- or long-term fire behaviors. Comparing fire responses over longer timespans will help characterize treatment effects and guide the timing of retreatment or maintenance return intervals. 4. � Press the science community for decision support. Activity fuels dynamics, for instance, are poorly understood; model users should support research that helps guide the management of these increasingly common fuels. Development of decision support tools that incorporate the dynamics of activity fuels and the effects of canopy and surface fuel patchiness should be encouraged. Common Errors in of silvicultural fuels treatments is Modeling Many proposed fuels to increase the spacing between crowns and increase canopy base By being conscious of input factor treatments and recent height (Agee and Skinner 2005), complexity, managers may recog- scientific literature both of which increase in-stand nize and avoid modeling errors contain examples of the windspeeds, a factor not always and increase the credibility of fuels inexact application of taken into account in fuels treat- treatment model output, while ment comparisons. model developers can improve deci- these decision support sion support tools to assist fuels tools—or more simply Future research efforts should planners and scientists. The follow- stated: “user errors.” clarify the effects of different fuels ing are common errors in assign- treatments on the wind adjust- ing parameter values, their conse- ment factor. In the meantime, most quences in modeled outcomes, and decreased woody fuel moisture con- modeling software allows users to best practices to avoid such errors tent due to greater incoming solar project wind speeds with the aid of in developing realistic and effective radiation, greater wind penetration, look-up tables based on published fuels treatments. and reduced relative humidity— approximations (e.g., Albini and factors requiring adjustment of Baughman 1979) or personal expe- Fuel Moisture Estimation input moisture levels. Assigned fuel rience. Fuel moisture is a fundamental moisture values should be derived factor in the ignition and spread of from on-site measurements, and Fuelbed wildland fires, and errors in mois- future research should examine Characterization ture estimation are common in fire fuel moisture variation in beneath- behavior modeling. These errors canopy fuels. Fuel Loading Estimates occur when investigators assign Fuel mass is a major driver of fire identical fuel moistures to different Wind Adjustment Factor behavior and severity, and errors fuels treatment scenarios despite Windspeed is another fundamental in estimating fuel loading are com- the differences in live and dead fuel factor in fire spread and intensity, mon in published fuels treatment moistures among stands with dif- and the effects of fuels treatment studies. Specific errors include use fering amounts of canopy closure. on those winds is not generally of unexamined default values and appreciated: as stand density and crude estimates of activity fuels. Treatments designed to reduce vertical structure are reduced, so is stand density, canopy bulk density, the impediment to wind (Albini and Both underestimating and overes- or crown base height result in Baughman 1979). A common goal timating fuel loads result in poor Fire Management Today 48 prediction of fire intensity and tors that base their fuel modeling This error is most obvious when a severity, particularly where large on collected field values. treated area is subjected to modeled amounts of woody activity fuels wildfire soon after treatment and remain following a treatment. Foliar Moisture Content before fuelbed recovery. Rather than basing their estimates Many investigators fail to disclose of loading on small sample sizes modeled values of foliar moisture As time elapses, downed fuels cure, or on unsupported default values, or use default values that may decompose, and flatten. Particularly researchers and managers should be poor estimates of actual field after fire and herbicide use—but refer to photo series data (e.g., conditions. Some reports include also with mechanical treatments— Ottmar and Vihnanek 2000), con- assigned foliar moisture content tree and shrub mortality leads duct on-site sampling using line values without justification or use to substantial increases in post- intersect sampling (e.g., Brown values that lie on the extremes of treatment surface fuels. Depending 1974) or biomass equations (Means published data. on the time elapsed, fuel levels et al. 1996), or estimate fuel load- may be higher, lower, or the same ing values with much deliberation. In addition to surface fireline inten- as pretreatment fuels. Brose and sity and canopy base height, foliar Wade (2002) examined contrasting Fuel Model Assignment moisture content is a determinant treatments across time, illustrat- Fuel models are based on values of of torching and crown fire initia- ing how treatment effects may be fuelbed loading, bulk density, sur- tion. Its importance is minor at short-lived and long-term effects face area-to-volume ratio, moisture lower surface fire intensities, but its may be unexpected. Furthermore, of extinction, depth, and heat con- proportional importance increases post-treatment changes in vertical tent (Anderson 1982). Every fuel and becomes operationally sig- and horizontal canopy closure may model carries its own assumptions nificant as predicted surface fire change future windspeeds, affecting about the interactions of these intensities increase (Keyes 2006) or surface woody fuel moisture. parameters, making fuel model when dead, dry foliage is attached selection (or “assignment”) an inte- to standing trees. In a review of With time, fuels generated by treat- gral part of meaningful analysis. published foliar moisture content ment begin to decay, affecting Managers and scientists make fuel values of western conifers, Keyes subsequent potential fire behavior. model assignment errors when they (2006) found that foliar mois- Comparison of fuels treatment violate model assumptions. These ture content in conifers of North effects should thus address an fuel model assignment errors are America varied between 73 and 480 extended period of time following most apparent when modeled fuel- percent, depending upon species, the initial treatment—a factor typi- beds contain activity fuel loadings foliage age, and season. The range cally modeled within the FVS-FFE. that exceed the loads used to build and importance of such values the models; a recent example of require careful consideration when Fuelbed Patchiness this is the difficulty many managers modeling fire behavior after fuels Fire behavior and effects models and scientists have faced in mod- treatment. suffer from a fundamental weak- eling masticated fuelbeds. Users ness that hinders their ability to also commit assignment errors in Time Since Treatment match field observations: models applying the same models to both Model users often maintain con- assume fuelbed uniformity, while untreated and treated stands. stant fuel values when projecting canopy and surface fuels are typi- fuel conditions into the future. cally distributed in patches. While Fuel model selection is an impor- Yet fuel load and the availability model developers are the first to tant disclosure in any fuels analy- of activity fuels changes tremen- point out this limitation, users may sis report. The Scott and Burgan dously following treatment: both fail to recognize it. (2005) models and models created values are affected by the amount in Fuels Management Analysts Plus and arrangement of residual over- Canopy fuel is a collection of indi- (Carlton 2006) offer more input story (Carlton and Pickford 1982). vidual crowns that may be patchy, options for fuelbeds with substan- In addition, the time elapsed since regularly spaced, or distributed tial activity fuels. Users should also treatment is critical to fuelbed at random, and this distribution examine the methods of investiga- decay, recruitment, and recovery. influences the spread of canopy fire Volume 69 • No. 3 • 2009 49 whether or not the canopy bulk and parameter assignments needs 86: 91–93, 107. density among patches or stands revision to better support fuels Jolly, W.M. 2007. Sensitivity of a surface fire spread model and associated fire are equal. Surface and ground fuels, treatment decisionmaking. behaviour fuel models to changes in live too, are rarely arranged uniformly, fuel moisture. International Journal of even in fuelbeds without activity Wildland Fire. 16: 503–509. Acknowledgments Keyes, C.R. 2006. Role of foliar moisture fuels. In stands with recent fuels Previous versions of this work content in the silvicultural manage- treatments, activity fuels are typi- ment of forest fuels. Western Journal of benefitted from comments by J. cally patchy, whether aggregated Applied Forestry. 21(4): 228–231. Kane, J. Kreye, and P. Andrews. in piles, on skid trails, or on bare McHugh, C.W. 2006. Considerations in This work was made possible with the use of models available for fuel soil generated by mechanical equip- support from the U.S. Department treatment analysis. In: Andrews, P.L.; ment. Butler, B.W., comps. Proceeding of the of Agriculture, Cooperative State conference: Fuels Management–How to Research, Education, and Extension Measure Success; 28–30 March 2006; In activity fuelbeds, available fuel Service (CSREES) McIntire- Portland, OR. Proc. RMRS-P-41. Ogden, loading and packing may be poor UT: U.S. Department of Agriculture, Stennis research program and by predictors of surface fire spread and Forest Service, Rocky Mountain Research the Applied Forest Management Station: 81–105. intensity across these patchy fuels. Program at the University of Means, J.E.; Krankina, O.N.; Jiang, H.; In lieu of models that incorporate Li, H. 1996. Estimating live fuels for Montana, a research and outreach patchiness typical of managed shrubs and herbs with BIOPAK. Gen. program of the Montana Forest and fuelbeds, managers and scientists Tech. Rep. PNW-GTR-372. Portland, OR: Conservation Experiment Station. U.S. Department of Agriculture, Forest should acknowledge the irregular- Service, Pacific Northwest Research ity inherent in activity fuelbeds and Station. 28 p. the resulting unpredictability in References Miller, C.; Landres, P. 2004. Exploring information needs for wildland fire fire spread and intensity. Agee, J.K.; Skinner, C.N. 2005. Basic princi- ples of forest fuels reduction treatments. and fuels management. Gen. Tech. Forest Ecology and Management. 211: Rep. RMRS-GTR-127. Fort Collins, CO: Conclusions 83–96. U.S. Department of Agriculture, Forest Albini, F.A; Baughman, R.G. 1979. Service, Rocky Mountain Research Errors in modeling fuels treat- Estimating windspeeds for predict- Station. 36 p. ing wildland fire behavior. Res. Pap. Ottmar, R.D.; Vihnanek, R.E. 2000. Stereo ment, fire behavior, and fire effects photo series for quantifying natural fuels. can often be tied to unsupported INT-221. Ogden, UT: U.S. Department of Agriculture, Forest Service, Volume VI: longleaf pine, pocosin, and assumptions about actual condi- Intermountain Forest and Range marshgrass types in the Southeast United tions and over-reliance on default Experiment Station. 12 p. States. PMS 835. Boise, ID: National Wildfire Coordinating Group, National values. In some cases, the basis Alexander, M.E. 2007. Software can assess fuels treatment effectiveness on crown Interagency Fire Center. 56 p. for assigning fire modeling values fire behavior. Fire Management Today. Reinhardt, E.D.; Crookston, N.L. 2003. are available but have not been 67(3): 30. The fire and fuels extension to the for- est vegetation simulator. Gen. Tech. adequately implemented. In other Anderson, H.A. 1982. Aids in determining fuel models for estimating fire behavior. Rep. RMRS-GTR-116. Fort Collins, CO: cases, model limitations are known Gen. Tech. Rep. INT-122. Ogden UT: U.S. Department of Agriculture, Forest but no compensation is made in U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Service, Intermountain Forest and Range Station. 218 p. assigning values or interpreting Rothermel, R.C. 1972. A mathematical results. Experiment Station. 22 p. Brose, P.: Wade, D. 2002. Potential fire model for predicting fire spread in wild- behavior in pine flatwood forests follow- land fuels. Res. Pap. INT-RP-115. Ogden, ing three different fuels reduction tech- UT: U.S. Department of Agriculture, Ultimately, the complex relation- Forest Service, Intermountain Forest and ships between wildland fire and niques. Forest Ecology and Management. 163: 71–84. Range Experiment Station. 40 p. stand structure must be captured Brown, J.K. 1974. Handbook for inven- Scott, J.H.; Burgan, R.E. 2005. Standard in useful rules and readily acces- torying downed woody material. Gen. fire behavior fuel models: A compre- hensive set for use with Rothermel’s sible forms so that the individuals Tech. Rep . INT-16. Ogden UT: U.S. Department of Agriculture, Forest surface fire spread model. Gen. Tech. responsible for prescribing fuels Service, Intermountain Forest and Range Rep. RMRS-GTR-153. Fort Collins, CO: management operations can base Experiment Station. 24 p. U.S. Department of Agriculture, Forest Carlton, D.W. 2006. Fuels management Service, Rocky Mountain Research their plans on a foundation of best Station. 72 p. available scientific knowledge. In analyst plus software. Estacada, OR: Fire Program Solutions, LLC. Van Wagner, C.E. 1977. Conditions for the general, guidance to assist fuels Carlton, D.W.; Pickford, S.G. 1982. Fuelbed start and spread of crown fire. Canadian Journal of Forest Research. 7: 23–34. managers in model assumptions changes with aging of slash from ponde- rosa pine thinnings. Journal of Forestry. Fire Management Today 50 Contributors Wanted We need your fire-related articles and photographs for Fire Management Today! Feature articles should be up to about 2,000 words in length. We also need short items of up to 200 words. Subjects of articles published in Fire Management Today include: Aviation Incident management Communication Information management (including systems) Cooperation Personnel Ecosystem management Planning (including Equipment/Technology budgeting) Fire behavior Preparedness Fire ecology Prevention/Education Fire effects Safety Fire history Suppression Fire science Training Fire use (including prescribed fire) Weather Fuels management Wildland-urban interface Firefighting experiences Submitting Articles: Send electronic files by e-mail or traditional or express mail to: USDA Forest Service Attn: Monique LaPerriere, Managing Editor 2150 Centre Avenue Building A, Suite 300 Fort Collins, CO 80526 Tel. 970-295-5715 Fax 970-295-5885 e-mail: Contributor guidelines are posted at Volume 69 • No. 3 • 2009 51 To fax your orders: 202-512-2104 To phone your orders: 202-512-1800 or 1-866-512-1800 For subscription cost and to Order on Line: http://bookstore.gpo.gov Mail To: U.S. Government Printing Office - New Orders P.O. Box 979050 St. Louis, MO 63197-9000 Fire Management Today 52