The study site will be divided into four blocks. Each block will serve as a replicate and receive one complete set of 12 treatments described below. This will result in a randomized complete block experimental design. Fuel loads will be sampled and photographed using standardized methods. The cost and nature of each of the fuels treatments will be recorded. Following treatments fuel loads will be monitored over a two year period and statistically analyzed against pretreatment loads. Resultant fuel loads will be matched as closely as possible to existing fuel models (Anderson 1982), and when no good match is found, custom fuel models will be developed for each of the resulting fuel types. These fuel models will be used as input for fire behavior simulations and run under a range of potential weather scenarios. Output from fire behavior simulations will include fire size, shape, fireline intensity, rate of spread, probability of ignition, and other parameters. These outputs will then be overlayed in GIS and used to model expected fire regimes under each roadside fuel treatment scenario. In addition, native species regeneration will be monitored over the two year period following treatment.

Study components

1) Complete region-wide GIS-based fire history map and database and use data to characterize current fire regimes (fire frequency and size) within the North Kona District.

Fuel Treatments

Each of the three groups of treatments will occur sequentially. Burns will proceed grazing treatments which will proceed spray treatments. Prescribed burns will be of low intensity and will be conducted by a qualified Burn Boss to follow a pre-approved burn plan. Grazing treatments will follow construction of paddocks Grazing prescriptions will be developed for 4" stubble height and 8" stubble height by Natural Resources Conservation Service (NRCS) Range Conservationist. Herbicide treatments will be applied once desired grazing prescriptions are met. Herbicide will be applied aerially in accordance with its label following state and federal laws [Rate: Glyphosate at 5.3 lbs/Ac (5.9 kg/ha) over 237 Ac].

Design

The fuels treatment experiment will be conducted utilizing a 2x2x3 factorial statistical design (Table 1). This design allows the investigation of various treatments alone and in combination with other treatments (Mason et al. 1989). This design will apply burning and grazing as whole plot treatments and herbicide as a split plot treatment randomly applied to half of each plot. This will yield 12 treatments total for each of the four replicates (blocks). Whole plot treatments will each be 4 ha (200m x 200m) in size. When split for herbicide application, this will yield 12 two ha plots per block. Total costs will be tracked for each treatment and summarized on a per hectare/acre basis.