Photo/Visual Plant Identification has It's Limits
This page last modified on 04 June, 2009 ==================
Most of the following was written in May 2004 and applied to the Photo-Key on this website. Since then I've been working on a more effective plant identification method using 3 main search stages:
See the first main topic at top of S.D.Wildflowers overview page for details.
This note is accompanied by a small collection of illustrations that I've placed in the Pseudo-Family PhotoIDConfused.
Summary: Plants I can only identify partially will be marked as [genus or species name] TBD (To Be Determined) from now on. If you can correct and/or improve my identification, please let me know!! ... klb.
Subdivision of flowering plants into Families originated from the idea that plants with similar visually identifiable characteristics could be presumed to be related to each other. Within a Family it has more recently been assumed that plants belonging to different Genera (plural of Genus) have inherited common characteristics from ancient common ancestors in the same family tree. Modern botanists have learned that the visually derived family groupings are often misleading, and may not imply common ancestry. With the advent of DNA sequencing, the evolutionary ancestry is becoming progressively clearer - and resulting in various re-assignments of the accepted Family and Genus groupings. Within a Genus, distinctions among the several to many Species are often more subtle. In the identification Keys published by professional botanists, the plant characteristics used to distinguish one grouping from another have often become very dependent on items one cannot observe visually in the field (i.e. without using specialized laboratory equipment).
Conclusion: Those of us interested to get a reasonably good visual fix on the identity of plants we see in the wild must almost always be satisfied to get an approximate species identification. In the tree (hierarchy) of recognized plant groupings, visual clues may sometimes allow us to identify a plant all the way to the Species or Sub-Species end of a branch. But quite often, non-visual clues will be needed to sort out differences not only at the Species level, but at the Genus or even Family levels. The extent to which the non-visual clues are important varies considerably from one Family to another, also depending on the extent that certain Families occur commonly or only rarely in localized regions or ecologic conditions, on year to year variations in weather patterns or following wildfires, etc. If you need to have the best currently available information on identity of some group of plants, you need the help of an experienced and well equipped botanical taxonomist who has a physical sample of the plant in question, and the time to sort through the necessary details.
So what's the role of a photo-based plant key designed for people who are not experienced plant taxonomists, and don't have plant specimen samples or the necessary laboratory equipment?? We live in an era when the issue of continued plant diversity has reached a crisis point for human society as a whole. Rapid destruction of natural plant habitat areas by urban development, and by the search for energy resources, is leading to the extinction of many species - with long term adverse consequences for humans that are still not well understood. But there are far too few experienced plant taxonomists with the budgets, time, and authority needed to keep up with the pace of this destruction so as to slow it down. Many of us who are not plant taxonomists have the time, and more than an idle interest, to help in practical ways that will at least assist in slowing down the destruction. We need photo-based plant keys as tools with which to make our plant observations good enough that the few plant taxonomists can avoid wasting time (searching for too few needles in too many haystacks!!).
And what's the role of a computer-based photographic plant key as compared with a printed publication?? In fact there are severe practical limits to the amount of detail - necessary for effective identification of most plants - that can be packed into a printed book. Those limits involve cost of publication, the large number of pages needed to carry the information, and the impracticality of carrying large books into field locations where the plants grow. Regional botanic centers accessible to the public (e.g. the San Diego Natural History Museum) provide a great resource. But the individual non-specialist wanting to use such a center must cope with a huge time overhead to make effective use of such a resource. Today's computers, with Internet access, can be made effective tools to mitigate or overcome most of these limits. With huge amounts of memory, and fast Internet access, both available at low cost, computers can be made a lot more effective for searching than books. Most print-based visual plant keys are based on grouping plants according to just one or two visual characteristics - such as color, flower shape and petal count, inflorescence shape, etc. But effective visual plant identification (even within it's limits) requires many views of flowers, inflorescence, leaves, and other characteristics. A computer-based key can now be built to provide any such grouping schemes simultaneously, permitting a user to search with less effort depending on circumstances. But it's still unlikely that the necessary computers will soon become sufficiently portable (with network access) to be used effectively while hiking in hilly terrain and in bright sunshine. So the serious observer (not equipped with license to dig up and take home a whole-plant sample) needs to take very good notes, or to take sample photographs to be brought home for comparison with a computer based key.
Finally: With a physicist/engineer mentality, I've come to realize that the detailed floras (published by and for professional botanists) are only approximations that describe the characteristics of known plant species. The many Species, Genera, and even Families that we see in the wild today have come about as a result of evolution from their ancestors. And the evolution continues today. So the Jepson Manual frequently mentions the tendency of some Species, and even Genera, that have a known tendency to hybridize (i.e. mix characteristics) with their cousins. And, while it is the current best detailed resource for California plant descriptions, the Jepson Manual, is the product of writings by many different individual professional botanists whose preferences obviously differ on what characteristics to emphasize, or leave out depending on print space limitations. I'm finding that many of the needed identification details are simply missing from the Jepson descriptions - one needs to follow through to the referenced research literature to get those details that are currently known.
Bottom Line: It seems that this computer-based photographic key can be of greatest utility by making detailed sets of photos available for as many (obviously different) plant species as possible - with identifications carried as far as possible. Where the identification cannot be carried beyond the Genus or even Family level, it seems better to mark the photos as TBD (To Be Determined) and ask for help from anyone with sufficient expertise to improve the partial identification. That will release me, as initial compiler of the key, to spend most of my time compiling the key rather than fruitlessly wasting most of my time trying to get my identifications correct beyond the limited level of my competence. It will also release more time for me to make corrections to my own flawed previous identifications - when E-mail advice arrives from folks who know better. In my Notes appended to each species, I will gratefully attribute any identification assistance to it's author (unless requested not to do so).
-- Ken Bowles
Useful References: