The basic spatial entity in this analysis is the management unit, or for fire planning purposes, the burn unit. A unit is defined as a single parcel of land which receives management treatments according to a particular regime. To give a simple example, a preserve may be divided into two parts, each receiving a fall burn every three to five years, but in no instance are both parts burned in the same year. Each of the two parts in this example is a separate burn unit.

A single unit may be divided into sub-units if the sub-units are contiguous and treated in close sequence. Another term for divisions within a contiguous fire unit is compartments. In the example just given, suppose each of the two units is larger than could be burned safely at one time. For logistical purposes, each unit could be broken into sub-units to be treated sequentially within a short period of time, where the order of treatment is irrelevant to the desired treatment effects. In most cases, however, all of a unit is exposed to a management treatment in one time frame.

A general guideline is to delineate two or more management units which encompass each priority species and community to be treated with a distinct treatment and schedule. The delineation would be done for each site, group of sites, or the portion of one site included in a Fire Regime Proposal. In this way, no one entire habitat or population would be burned in a given treatment.

If, for instance, fall and spring burns are two treatments to be applied in a site with only one community element, the guideline would call for dividing the site into at least four units, two for each treatment, with each unit including representation of the community element. If two different treatments are to be used in a site with two community elements, four units would be the minimum but more might be necessary depending upon how the occurrences of these communities are spatially distributed. This approach provides some protection against temporary negative fire effects, since unburned areas may function as refugia. It also supplies comparison areas to help managers and others discern fire effects (keeping in mind that management fire plans seldom have the design needed to ascribe causation with a known degree of confidence).

Apart from the habitat needs of high priority species, the scale and configuration of units has a large influence over management efficiency. In general, efficiency is increased if units are as large as possible without exceeding containment capabilities. This is because the cost of fire management tends to be proportional to the length of unit perimeters while benefits are proportional to unit areas. Units with approximately square or rectangular shapes are easier and more efficient to burn than irregularly shaped units. Straight line unit boundaries are easier to work than curving boundary lines. However, highly regular boundaries may be aesthetically offensive or ecologically inappropriate.

Ideal unit boundaries do not traverse slopes, boggy areas, or topographic depressions. Keep in mind what might happen if the fire should cross a unit boundary and escape. Would the fire make a run upslope? Would wet areas impede control efforts? Although it is tempting to employ stream channels as firebreaks, if the stream is at the bottom of a canyon or gully, fire can race upslope should it jump the unit boundary.

Often a good approach is to use watershed boundaries to delineate units, with firebreaks placed on the high ground dividing watersheds. Be aware, however, that significant species and communities may be narrowly restricted to ridge tops and hence be severely impacted by firebreak preparation. Other consequences may include the threat of erosion. A firebreak may be placed downslope from the ridge top, just outside of the watershed containing the burn unit. Potential problems with this are the difficulty of working on slopes and the potential for burning material to roll downslope, spreading the fire outside of the unit. There are no perfect solutions to these problems, but if the important factors are understood, an acceptable alternative can usually be found.

There are different views about whether to have permanently located fire unit boundaries or to relocate them for each burn. Permanent locations concentrate impacts from firebreak preparation. This may be better or worse than distributing and diluting impacts over more area. Much depends on the roads, trails, natural firebreaks which already exist, and the type of vegetation you are dealing with.

Natural or pre-existing fire barriers such as rock slides, roads, trails, ponds, fuel type changes, etc. obviate certain cost, management and impact problems. In placing unit boundaries along pre-existing fire barriers, and in certain relationships to topography and barriers to access, always consider the contingency of fire escaping. Natural barriers may present greater problems from escapes than manager-produced firebreaks.

Similar types of considerations exist in formulating treatment schedules. A very regular schedule, such as, "burn in April once every five years," facilitates planning and makes it easier to interpret fire effects, but the lack of variability may be ecologically inappropriate. Temporal and seasonal variability may be essential to maintain community and landscape diversity. Regular schedules also place burdens on the manager if conditions do not permit effective burning. Flexibility in schedules makes it more likely that treatments will be accomplished when conditions are suitable. There are several approaches to developing a burn schedule. You will have to decide which is best for your situation.

One approach to defining a burn schedule is to state the quantitative parameters that will be used to trigger the need for a subsequent burn. This allows the manager latitude in selecting the time for the next burn based on previous burn results and the environmental or fuel conditions that produced those results. For example, suppose a previous burn did not kill encroaching hardwoods to the degree expected. This might warrant a re-burn at a relatively short interval, or under drier or windier conditions. On the other hand, a better than expected kill may allow for a longer interval before the next burn. This trigger approach is best applied when a site goal involves maintaining the habitat of a single species or when the site is in a restoration phase.

Another approach is to focus on known fire regime attributes (i.e. burn frequency, severity, seasonality, scale) that will maintain the desired community or landscape mosaic at the site. A range of fire regime variables might be derived from available knowledge; e.g. a consensus of opinion supports a fire regime for a particular savanna type that includes a return interval averaging 4 years, within a range of 1 to 8 years. Exceed this range or repeatedly deviate too far from the average and the community will likely change and species might be lost. The manager can then track time since last burn for the community types within each burn unit. In planning for a burn season, treatment priority may be directed toward those units where the fire regime will soon deviate beyond its normal range and mean. Variation in frequency or seasonality can be incorporated by developing a weighted or constrained random numbers generator or table. Prescription parameters should be broad enough to allow variation in fire intensity and burn severity. Keep in mind that the variation in fire regime components may be more important ecologically than average properties.

What about the situation where there are several different communities within a burn unit, each maintained by a different fire regime? For example, one community is characterized by frequent surface fires occurring throughout the year, the other by infrequent fires that normally occur only under seasonally limited severe burning conditions. In this situation, burns must be designed so that the community requiring infrequent fire does not burn too frequently, at the same time allowing for frequent burning of the other community such that a dynamic ecotone is maintained between the two. Flexible burn units, or burning subunits within fixed boundaries, may facilitate this need. Developing a burn schedule for a mosaic of communities usually requires mapping burned and unburned areas each time fire has been applied, so that fire histories can be analyzed for community occurrences across the landscape. GIS-based decision making tools are becoming available for this analysis which will use manager-defined fire regime parameters to flag areas needing fire or areas that are in danger of burning too frequently.

1. Identify management units and delineate them on a map. Name or number the management units. Overlay on a community or vegetation map. Provide rationale for unit delineation. Describe unit boundaries, such as types of firebreaks.

2. List or display on a map the conservation targets or other significant species/communities for each unit.

3. State unit-specific objectives to accomplish conservation goals for targets.

4. Identify factors that might constrain fire management on each unit.

5. If in a restoration phase, give initial treatment schedule and explain rationale in terms of desired fire effects.

6. If in a maintenance phase for a single species, describe desired habitat structure and triggers or indicators for subsequent burn treatments.

7. If maintaining a single community or mosaic of communities across a landscape, describe the fire regime that characterizes each and employ or develop a decision-making tool that incorporates variability in frequency, intensity, and timing. Large or complex sites may require GIS-based technologies.