How do I meaningfully display Kruskal Wallis Data when I have a lot of zeroes?

In general, when counts are too low, one groups the data. For example, you could group by region or year, depending on what works best for you. 

As it is, the graphs don't seem very informative to me, since they only seem to  be comparing only the right tails of the distributions at best. You could condition your analysis to this case though: "when mits are present, we see more mites of type M than F". In that case, the regions, years and lineage considered would be different, though.

I wouldn't know what to display since that depends on the research question you are trying to answer from the very beggining.

Edited answer:

* Do I just scrap this graph and use the table?

Don't use the graphs in the word document because they are not informative. They barely show and compare the distribution of the right tails.

There's only one table, which is in the Excel file, and it contains all the data. Naturally you do need to use the data/table.

* Would changing the numbers to ranges help? that way the ranking process is easier?

This is one strategy to group the data, and like I mentioned, grouping counts is a useful and common strategy. How to group the data depends on the research question you are trying to answer. Commonly, data is grouped by regions but there are many other choices. I would suggesst combining by region and lineage, for example. Then you can repeat your previous analysis with the same tests.

I also suggested comparing only in the presence of mites (conditioning on the presence of mites). That means discarding all observations with zero mites and repeating your previous analysis.

Please show a better display if you think there is one. Looking for a suggestion of what alternative to use as far as testing/display. The answer doesn't quite address the question of how I would display the data. i.e., switch from a whisker plot to a ____.

The issue here is not that you chose a whisker plot over another graph. The issue is the overabundance of zero counts in your data. I have suggested two alternatives for this excess of zeros, in terms of test and display. I will summarize them here:

1. Find a meaningful way to group the data, according to the context in which the data was sampled and your research objectives. As an example, consider grouping counts of mites by a combination of "Lineage" and "Region". Then repeat your analysis in terms of tests and boxplots comparison.

2. Condition your analysis to the presence of "Mites". Most of the time there doesn't seem to be mites. But when there are, there seems to be differences by region and source. This is what is relevant to compare, in my opinion. This means discarding all observations with zero mites and repeating your analysis in terms of tests and boxplots with the rest of the data.