We explored three approaches for combining Linkage Mapper and Omniscape results. The first one turned out to be a stepping stone to the following two, which turned out to be the most promising, and are summarized below. The first one, not shown here is briefed in the “Supplementary Material: Geoprocessing Notes and Lessons Learned”.
Option One: Clipped Omniscape OverlayIn this approach, we clipped the classified Omniscape product (i.e. Normalized Current Flow) to the HCAs, and overlaid that on the various Linkage Mapper products. Here it is overlaid on the evenly weighted sum between Linkage Pathways (i.e. least-cost corridor) and Linkage Priority value.But it was hard to translate this map to conservation priority, this led to the challenge of getting all of these values on the same numerical scale, yielding the different “OmniCore” approaches below.
Option Two: “OmniCore”We made three variations of the process we called “OmniCore”.
OmniCore BasicIn all the diagrams below, the blue circle indicates an evenly weighted sum. Each of the inputs is normalized to range from 0-1, so they are all on the same initial numerical range. The Omniscape product was normalized linearly with the score range normalization discussed elsewhere in the Linkage Mapper documentation, such that the 100 became a 1, the 0 a 0, and all other values scaled accordingly.
We tried the above with cores valued at several levels, and found that 0.5 was the best balance, according to expert opinion (J.Gallo and T. Miewald), between giving high value to cores, but giving connectivity conservation priority to make sure the linkages get conserved. A value of 0.6 was not used, and would be good to examine in the future. Here is how the product (available in Science Base and Data Basin) looks:
OmniCore with Linkage Priority:We then added the nuance of Linkage Priority to the OmniCore concept, thereby giving higher value to the higher priority linkages, using the following logic model and yielding the following result.
OmniCore with Linkage Priority and Pinchpoints:Finally, we explored the possibility of including Pinchpoints to explore if it “double-counted” pinchpoints too much (since both Pinchpoint Mapper and Omniscape draw from the Circuitscape algorithm). We used the following logic model to yield the following figure, and determined that the “double counting” isnot excessive, but that including pinchpoints was not worth the extra work, and did not present this option during the workshop.
Option Three: “Crosswalked OmniCore”For this approach, we assigned relative connectivity conservation values to the five categories of the Omniscape product: “Normalized Current Flow Smoothed”, as per the below table, then combined that with the Linkage Mapper product as per the OmniCore with Linkage Priority approach, above. We chose that OmniCore method based on guidance from the WWHCWG to aim for parsimony when possible (making things no more complicated than they need to be). We chose the numerical values based on informal expert opinion (J.Gallo and T. Miewald). This was challenging, because each class appeared to represent different levels of conservation priority in different parts of the landscape, so a recommendation was made for a more formal expert opinion process if this method were to be adopted.
Of all these options, OmniCore, most notably “OmniCore with Linkage Priority” was deemed the least problematic and most promising.
