Home » bplant.org Blog » Range Map & Taxonomic Update Progress

Range Map & Taxonomic Update Progress

January 31st, 2024 by Alex Zorach

Back in September of 2022 we announced that we had completely retired our first-generation range maps, replacing them with new and improved maps. However, many aspects of our maps remained incomplete. Many species still lacked maps, and most maps were strictly limited to the lower 48 US states, not extending into Canada, Alaska, or Mexico.

We have since made great strides in map completion:

• 16838 plants now have range maps. This is over 81.5% of the plants listed on our site.
• 5061 maps have been completed for all of North America, including Canada, Alaska, and Mexico for species that occur there. This is now over 30% of our published maps.
• 5653 plants (over 27% of plants) have been interlinked with Plants of the World Online (POWO) and had their taxonomy updated to reflect the information from POWO.
  
• We began marking plants that we wanted to list on our site, but that do not occur in the wild in North America and thus have no map or an empty map. We currently list 27 such plants and this number will probably increase quickly now that we have this category.

What is perhaps most exciting, however, is not reflected in the numbers alone: we have developed techniques and frameworks for building range maps outside the lower 48 states. This means it is just a matter of time and hard work before the remaining maps are completed.

Why has the rate of map completion slowed down?

If you have been following us since the beginning, you will notice that our rate of completion of range maps has slowed down considerably. We have been putting more time and effort into range maps recently though. The slowdown is both because we completed the easier maps first, and because we are developing higher standards as we further refine existing maps.

In many cases, constructing maps was straightforward because there was complete, accurate data available from the various sources we consult, and the different authorities agreed on plant ranges and establishment means (i.e. native vs. introduced.)

In other cases, things get messy. Sometimes data is missing or incomplete. Often, different sources disagree, and we need to research and make tough judgment calls about which material to accept as correct. Common issues include misidentifications, and different standards in different sources about when to consider a record a cultivated plant, vs. a waif (a transient introduction into the wild that does not persist), vs. solidly established in the wild. Many plants also have taxonomic complexities, such as splits, merges, and other reclassifications. Taxonomy can also interact with identification issues, such as when a record was identified as a particular variety or subspecies, but that taxon then later up getting moved into another species.

We have been building our range maps by conducting multiple sweeps through the data. On each sweep, we construct maps using the tools and processes we have available, and put a certain portion of the remaining map in a "messy" bin, to do later. At the same time, we also work to improve and streamline our map building process. Over time, the plants left in the "messy" bin become progressively more difficult to deal with, so even with better tools to facilitate our work, our maps become slower to build with each pass.

Our standards are also higher with our second-generation range maps; early on we just wanted to get something out, whereas now we have more different categories on the map legend and are taking care to notate uncertainty in the map itself. Related to these high standards, if you see something that you think is inaccurate or could be improved, please contact us. You do not need to be an expert to help us refine our maps. Seeing a species in the wild outside its reported range, or just seeing something on a map that you think doesn't make sense, are both signs that you may have something to contribute.

Interlinking with Plants of The World Online (POWO)

POWO is a taxonomic backbone run by the Kew Botanical Garden in London. In contrast to our routine, mostly-automated interlinking with other databases, our interlinking with POWO involves examining POWO's treatment of each taxon and comparing it to the schemes used by other authorities. Records integrated with POWO have mostly been brought up-to-date with modern taxonomic changes. In the overwhelming majority of cases, we adopt POWO's treatment. In the few cases where we adopt a different treatment (this usually happens in the rare cases where POWO is not up-to-date on newer work and another source we consult is), we still carefully examine POWO's listing. In a few cases we have enacted splits that POWO has not, in line with Flora of the Southeastern U.S. (FSUS) by Alan Weakley.

Differences in classification are a primary reason for discrepancies in plant range maps between different sources. There is often not a 1-to-1 relationship between records. You can read about how we resolved these discrepancies in the notes fields on each plant page. Taxonomy notes are located above the table-of-contents, right under the plant's names. This placement ensures people will have a clearer idea of what taxon the page refers to before reading the article or seeing the map, in the case that there have been confusing reclassifications or inconsistencies in naming. The range map notes are found directly under the map, and explain any judgment calls we made in constructing the map as well as any clarifying info beyond that conveyed by the legend.

Although the notes are brief on most articles, and many articles lack notes entirely, the total extent of notes is massive. We have already published over 16,000 words in the taxonomy notes, and over 58,000 words in the range map notes. This amounts to about 250 full pages of text. These counts do not include automatically-generated legends nd other verbal notes listed under many maps. The volume of these notes explains why our progress on the more difficult plants is progressing more slowly. We want to make sure we both document and explain our choices, especially in the cases where different sources disagree on taxonomy or ranges.

No Maps: Plants Not Occurring in The Wild in North America

Although we focus on plants occuring in the wild in North America, we list a number of plants with empty range maps, meaning that these species are not and have never been found in the wild here. These plants broadly fit into one of two categories: plants that are widely cultivated, but have never established or been recorded in the wild, or plants that were reported somewhere in the wild but where those records were later deemed to be invalid. We retain both of these types of records for several reasons.

For garden and landscaping plants not recorded establishing in the wild, it is always possible that these plants will establish in the wild at some point in the future. Being aware of these plants and knowing how to identify them can ensure they are more quickly detected if they ever do escape and naturalize.

Some plants were widely planted for years or even decades without naturalizing, only to eventually escape into the wild, sometimes even becoming invasive. For example, kudzu (Pueraria montana) was brought to the US in 1876, but the first herbarium specimen we could find, collected from the wild in North America, was from 1901, and kudzu was not recognized as invasive until decades after that.

It is also possible that plants are already present in the wild in North America, but that people have not reported them because they are hard to identify and/or they do not know that they need to check against them. For eaxmple, we have found Japanese honewort (Cryptotaenia japonica) in the wild in two ecoregions, in Delaware and Pennsylvania, and that species is not reported occurring in the wild by BONAP or POWO. Excepting the purple-leaf cultivar, that species may be underreported because it is visually similar to the native honewort (Cryptotaenia canadensis).

A completely different scenario is when older sources report a particular plant somewhere in North America, but the record has been deemed invalid. In these cases, we retain a record specifically because there is other information out there that needs to be corrected. Even when the original source has been updated to correct the record, it is common, especially with sites like USDA PLANTS, that numerous other websites and print sources repeated the inaccurate information before it was corrected, and these other sources are often not corrected. Sometimes misinformation can propagate in secondary sources for decades after the original source corrected it because no one goes back to check the primary source. In the plant world, this phenomenon is especially prevalent in sources relating to horticulture, gardening, and the commercial nursery industry, where the standards of scholarship are much lower than in the sort of scientific sources we rely on.

Examples of such plants include mountain soursop (Annona montana), Syrian cephalaria (Cephalaria syriaca), braceletwood (Jacquinia armillaris), and willowleaf frostweed (Helianthemum salicifolium). We have noticed such errors are especially common in south Florida where there is a huge volume of introduced tropcial species that cannot survive in most of the rest of the continental US.

The First Range Maps and Lists for Canada and Mexico Finer-Tuned Than Provinces/States

One thing that has frustrated me for years is the lack of fine-tuned range maps and plant lists for regions outside the US, particularly, for Canada and Mexico. Apart from specific sources such as the Silvics of North America, which only covers about 200 tree species, and some print field guides, which similarly do not exhaustively cover the plants of North America, there are no sources with range data as fine-tuned as that presented by BONAP and USDA Plants.

Although Canada and Mexico both have administrative divisions finer than States and Provinces, there is no equivalent of USDA PLANTS and BONAP for these countries. Canadensys, the closest to a Canadian analogue of these sources, only has data at the level of individual provinces, and the existence of finer-tuned regional data, when it is available at all, is spotty and not accessible in a single place. In Mexico, there is no such website at all.

Although the lack of data has made it much harder for us to build range maps in Canada and Mexico, it also makes our maps much more valuable and groundbreaking in the cases in which we have been able to build them. In these cases, our maps offer a finer level of detail than anything else out there.

The process of constructing these maps is slow and also error-prone. In many cases, it involves examining individual herbarium records. Although there are good tools (the most important of which is GBIF, the Global Biodiversity Information Facility), there are numerous limitations of the data available. Some herbarium specimens have yet to be digitized, thus making it hard to assess the completeness of some range maps (i.e. it being harder to conclude that a species does not occur somewhere, relative to knowing that it does occur somewhere.) Where records exist, many records have geolocation errors, and in many cases there is insufficient verbal description of the location to catch such errors. Even when the locations are accurate, other aspects of the records can be erroneous, such as plants reported as wild that were actually cultivated, or specimens that were misidentified. And on top of this, we have all the same reclassification issues discussed above, but they are uglier because different herbaria (and even different individuals) may follow different schemes, and reclassifications can thus play out differently from one record to the next.

The process of building these maps thus involves delving into individual herbarium records in a way that rarely is necessary when constructing maps for the lower 48 US. As such, it is much slower. However the results can be rewarding.

As we complete these maps, we move towards plant lists for ecoregions outside the continental US. Back when we launced ecoregion-based plant lists in March of 2022, a major deficiency of our lists was that we had no lists for Canada, Alaska, and Mexico. Although we have not completed them, we are seeing the beginnings of these lists for all three of these regions, and it's just a matter of time before they are complete enough that we can publish tentative lists. Our work is also improving the accuracy and completeness of the plant lists for ecoregions that are split between the lower 48 US and either Canada or Mexico, and there are many such regions. Canada and Alaska will likely be completed long before Mexico, both because there is much less plant biodiversity in the Arctic and more in the Tropics, and also because there are more sources available for Canada and Alaska, and lastly because there is less of a language barrier.

Moving Forward: Things To Come

You can expect a continued focus on range map completion over the next few months. We are still in a phase where we are getting a lot done and we hope to keep improving our tools and techniques so that we don't end up getting completely stuck on the difficult maps. So far, so good!

I also want to thank everyone too for ongoing financial support that has allowed me to keep working on this project. Although we still have not met our next financial goal of $20,000, we are getting a lot closer to it. The vision of expanding into a larger organization with staff may remain far off, but for now we are able to keep the site going and keep it ad-free for the forseeable future. I have not yet published the year-end finances but this is another thing to expect in the next several weeks.