Talk:Hypericum swinkianum

Range map

@Fritzmann2002: regarding Distribution of Hypericum swinkianum.png, this is a new (2016) and poorly studied species. To my knowledge it's not known from anywhere in eastern Michigan, Wisconsin, or Ohio; the original description lists a handful of counties within Illinois, Indiana, and Michigan.

I think in general it would be helpful if you are more precise with these Hypericum range maps and include a list of sources with each one on Commons. To me, range maps like this should never be based on synthesizing vague text-based descriptions, they need to look at finer scale data, like plotted coordinates of individual occurrences or at least county-level presence (typical in the US, e.g. USDA PLANTS, BONAP). There's a difference between coloring in an entire state or country, with obvious political boundaries, and coloring in by estimation/guesswork an approximate natural range boundary by hand. The latter seems like original research and the former isn't all that helpful for species like this with a small, sporadic range. In the case of H. swinkianum, I think it would be best to have no map at all, or one that only includes the pinpoints of known occurrences from published, reliable references.

I'm sure this has come up before—maybe @Peter coxhead, Plantdrew, and Pagliaccious: have thoughts and/or know of a related discussion or guidelines? Thanks! —Hyperik ⌜talk⌟ 00:30, 10 December 2018 (UTC)

@Hyperik:, you bring up good points, I have just been making these range maps en masse for my grand project of List of Hypericum species for a few different reasons, but I am definitely not as experienced as others in this field, so I'm open to changing my methods. I just have a few general things to throw out there that might be useful for discussion:

1. An image can have better data about the distribution of the species shown than raw text, and can store that information in a smaller, easier to discern packet for the reader. This was my biggest motivation.
2. An image is more aesthetically pleasing to the reader, this was my second motivation
3. OK in regards to the specificity:
   1. When I am making the maps in the future, I will be sure to reference where I get the distribution info on the Commons. Most of them come from Norman Robson's assorted works, most of which are accessible on Hypericum MySpecies.
   2. For most of the species, the best distribution info is vague ranges (from general area x to general area y) and translating that into a map is, at best, a very rough interpretation
   3. I'm speaking with no experience here, but in my opinion, a vague-ish map is not necessarily "misleading" to the reader. Most, if not all, readers can get better info out of a vague map than a distribution listing, even if the map boundaries are just approximations. They can look an see "ok, so this species is native to the southern regions of Africa in sparse populations" or "oh, this species is only found on one mountain in Turkey" even if the southern region of Africa isn't exact or the Turkish species is a few degrees off in longitude.
   4. If vague text-based info is the only info out there, I don't see why range maps can't be made from it, and updated as new info comes out. If there is detailed info on distribution for species, then I agree those maps should be much more precise, and I will personally try to do better on making those
4. Until some consensus is reached, I'll put a pause on making the maps, but I do think they are improvements to text-based distributions in list articles, and would like to keep making them for the Hypericum species list article

Thanks for starting some good discussion, hopefully some meaningful consensus can be reached! Best regards and deepest respect, Fritzmann2002 ^{T, c, s, t} 01:51, 10 December 2018 (UTC)

@Hyperik: unfortunately I am not familiar with the creation of range maps nor am I familiar with any past discussions on the matter. My only suggestion would be perhaps to dig through this search to try to find prior discussions on range maps. On the matter of range maps, I am of the opinion that range maps, even if a degree vague, benefit the article by making the range more accessible and visual. For certain cases I could see the counterargument, especially when the range is small and scattered, that pinpoints are more beneficial than a spread. One thing I am certain of is that political boundaries should be ignored and a general geographical spread should be considered, unless a source explicitly states that the range encompasses an entire state or other region. Best regards, Pagliaccious (talk) 02:10, 10 December 2018 (UTC)

@Pagliaccious: for plants, the smallest distribution units in the WGSRPD (used in most reliable sources) are political in nature. See e.g. the map for Hypericum perforatum at PoWO. Peter coxhead (talk) 10:04, 10 December 2018 (UTC)

@Peter coxhead: Thank you for pointing that out, I was completely unaware. I've revised my comment. Pagliaccious (talk) 17:13, 10 December 2018 (UTC)

Older map (left, deleted) overreaches in most areas, including states where it doesn't occur.

Thanks Fritzmann/all, I totally agree a range map is worth a thousand words (or a few dozen at least), and a worthy endeavor to make and add to plant pages. I just want to caution against uploading hand-drawn (non-political) distributions unless they're paired with and match referenced sources. If, based on limited available data or descriptions, we need to back up to using political boundaries like county, province, state, or even country, that is better than drawing in "a part of southern Turkey" or "a zone around Lake Michigan" based on a line of text. I don't think it's our place as editors to make maps that are off multiple degrees of longitude (hundreds of km) or several orders of magnitude in area, though some level of imprecision is of course necessary when we don't have exact coordinates. As always, it depends, and while more care should be taken for species that are of conservation concern and/or with small distributions, less precision is needed with broad-ranging species. As a comparison, here are two maps side-by-side, the older one originally uploaded and one I just added which plots the locations of known specimens of this plant (centered for each county). —Hyperik ⌜talk⌟ 02:52, 12 December 2018 (UTC)

You make a very good argument, and a good map, so I am won over. So, just for clarity's sake, consensus was:

• Uploaded distribution ranges should have references on Commons
• For species with limited distribution, specific collected localities based on exact or as close to exact as possible data from sources
• For species with far-spread distribution, more general range maps can be accepted, but using political boundaries is preferred if at all given in the references

I will certainly be focusing more on quality over quantity as I make new range maps for Hypericum, so thanks for helping me realize that on Wikipedia, it is important that the usefulness and accuracy of information is far more valuable than sheer bulk. Best regards, Fritzmann2002 ^{T, c, s, t} 12:18, 12 December 2018 (UTC)