Traditional grass identification keys will typically focus on choosing between two options to gradually filter down options. This helps to quite accurately determine the relevant grass species.
Sometimes the features included at certain stages of the keys can be difficult to judge by an inexperienced but keen user who goes down a false path and then has to back track. This can be off putting for some.
To help make the process easier to follow and introduce some additional criteria at certain stages I’m developing an online interactive decision tree model which can complement the more traditional approaches and maybe help to better engage people in identification skills. This might not be the most botanically correct, sequential, way of identifying grasses but it’s another angle that I thought I’d explore using web technology.
I’ve used vegetative features plus also included habitats where the probability is this is only where these grasses would be found. This approach helps filter out some potential options; however, the grasses will also be included within the more traditional options as well.
It is a reasonable assumption to remove Lyme grass, Sea Couch and Hybrid Sea Couch from the select options of ‘Smooth blades’ and ‘Lightly hairy blade side/s’, as it is unlikely (but not impossible) that they would be found elsewhere in the wild, except by the coast.
I will probably add them into those options as well, but for now I just want to develop it and trial it as a web app.
Each selected option/s sits within the higher level select option, like a Russian doll effect. A small part of a data object is included above as an illustration.
Once it’s completed I’ll provide a diagram which identifies the options to show how the grasses are identified. A pop-up, or similar, help option will be added to illustrate the different features which are given in the text, e.g. ligule, auricle, light hair density on sheath etc.
Chris Gray, 25th June 2017