Some of the following content first appeared on my travel blog – grabetravels.blogspot.com. Here I have appended additional content explaining the educational potential of using identification keys and apps for identification of plants and animals.
———–
For anyone from the midwest, photographing the flowers of Hawaii becomes a constant activity. The flowers are everywhere even this time of year and they are truly beautiful. At some point, you exhaust yourself on the artistic value of your growing photo collection and you start to think beyond the vibrant colors and the huge blossoms. You begin with simple questions. I wonder what this plant is called? You may wonder about other things as well. How does this flower grow in the crook of this tree without having roots in the soil?
I started mixing tech with photography a long time ago. My biology background even plays a role in both my plant photography and my interest in technology. I began exploring apps on my phone that claimed they could be helpful in plant identification. Point your camera at a plant and the app would tell you what that plant was and direct you to additional information about the plant. I have read about face recognition and the role artificial intelligence plays in making this possible. It made some sense that the same technology might be applied to plant identification.
During last year’s visit to Kauai, I explored an app called PlantSnap. This year I added iNaturalist to my app collection. Both work in a similar way. You take a photo and the app makes a guess as to what appears in the photo. The app also offers some other possibilities and relies on you to evaluate the first choice and the alternatives. If nothing makes sense, the app farms out your photo to the public that follows the app to see if anyone has an idea. Selecting a choice offers a way to get additional information.
Evaluating these apps presents a problem. You must know what you are looking at in order to determine if the app made a correct identification. Here are three photos I took walking home from the coffee shop and the top choice offered by each app.
[PlantSnap]
[iNaturalist]
[PlantSnap]
[iNaturalist]
[iNaturalist]
I know both apps were correct with the genus. I am not capable of evaluating the accuracy at the species level.
There are other ways to do this. You can purchase a plant identification book. There are even online keys that offer an identification approach based on color. I have decided the flowers are probably easy, but I wonder what happens when the app is asked about a plain green plant. I don’t have any examples and I am not certain I would be able to check the responses, but I have a few more days here and plenty of specimens I can photograph.
There may be no post tomorrow. We have to move out of our condo and into a hotel. Two of our kids and their families did not have a Spring break option during the two months we originally booked so we had to find a way to extend our stay. I hear there is still snow and flooding in the midwest so this has worked out well.
I started my career as a biology major and the preparation to teach biology in high school. My interest in science education and learning (and the army) led me another direction. Many of my original interests remain and pop up in some of the things I write.
Some of my zoology, entomology, and botany courses required that I identify specimens in the field or lab. To make these identifications, we were taught to rely on dichotic keys. A key works something like a “choose your own adventure” story. Instead of a segment of text that ends with a decision point offering options and the option selected determines the next segment of text read. In an identification key, the key asks the user to make an observation to answer a dichotomous question – do smaller branches appear along a larger branch opposite each other or from alternating sides of the larger branch? Depending on the option that applies to the specimen to be identified, the user is sent to different follow-up questions. Eventually, the process arrives at what the key assumes is the species to be identified. Creating such a process would obviously be a tremendous challenge, but for the user a key offers a practical way to make an identification. Note that a key works in a different way from the identification resources you might purchase to identify birds, wildflowers, insects, etc. A key is supposed to be an algorithmic as opposed to a heuristic approach to solving a problem.
Identification of an unknown is a skill one learns in several sciences (e.g., geology). Exploring the process of identification using several approaches would make a great biology lab. Bring in several “mystery” specimens and have students use a key, an identification book, and maybe an app to see what they can accomplish (here is a simple online botany key). A good related exercise is to consider what the might be the advantages and disadvantages of the approaches I have identified.
