We derived the habitat cores using a tool within Gnarly Landscape Utilities called Core Mapper (Shirk and McRae 2015). To develop a Habitat Surface for input into Core Mapper, we started by assigning each 30m pixel on the modeled landscape a habitat value equal to its GNN OGSI (range = 0-100). In areas with serpentine soils that support habitat potentially suitable for coastal marten (see report for details), we assigned a minimum habitat value of 31, which is equivalent to the 33rd percentile of OGSI 80 pixels in the marten’s historical range. Pixels with higher OGSI retained their normal habitat value. Our intention was to allow the modified serpentine pixels to be more easily incorporated into habitat cores if there were higher value OGSI pixels in the vicinity, but not to have them form the entire basis of a core. We also excluded pixels with a habitat value <1.0 from inclusion in habitat cores. We then used a moving window to calculate the average habitat value within a 977m radius around each pixel (derived from the estimated average size of a female marten’s home range of 300 ha). Pixels with an average habitat value ≥36.0 were then incorporated into habitat cores.

After conducting a sensitivity analysis by running a set of Core Mapper trials using a broad range of habitat values, we chose ≥36.0 as the average habitat value because it is the median OGSI of pixels within the marten’s historical range classified by the GNN as “OGSI 80” (Davis et al. 2015). It generated a set of habitat cores that were not overly generous (depicting most of the landscape as habitat core) or strict (only mapping cores in a few locations with very high OGSI such as Redwood State and National Parks) (see Appendix 3 of the referenced report for more details, including example maps from our sensitivity analysis). We then set Core Mapper to expand the habitat cores by 977 cost-weighted meters, a step intended to consolidate smaller cores that were probably relatively close together from a marten’s perspective. This was followed by a “trimming” step that removed pixels from the expansion that did not meet the moving window average so the net result was rather small changes in the size of the habitat cores, but filling in many individual isolated pixels with a habitat value of 0.

Once the initial set of pixels that met the minimum habitat value for cores was identified, we removed core areas that would likely be too small to support enough martens to viably sustain long-term occupancy. To do this, we set a minimum habitat core size of 1500ha, which was identified in the “Humboldt Marten Conservation Strategy” (Slauson et al. 2019a) as the amount of suitable habitat needed for a “Population Reestablishment Area” (equivalent to approximately five female home ranges). Due to remaining uncertainty about the exact boundary of the species’ precise historical range, we included all cores located within or intersecting the boundary of the historical range in our final model.

This is an abbreviated and incomplete description of the dataset. Please refer to the spatial metadata for a more thorough description of the methods used to produce this dataset, and a discussion of any assumptions or caveats that should be taken into consideration.