The expand taxonomic groups option appears in the report wizard of Recorder. When creating a taxon report (querying the database for taxon occurrences) it is possible to select one or more taxa from any of the lists. In this instance we have selected Platanthera bifolia. The expand option should then include all other taxa which are listed as hierarchical children of the selected taxon. In the normal run of things that would mean subspecies and/or varieties, if the taxon is a species as in this case, but would also include species & monogeneric hybrids if a genus is selected (and so on).

Clearly a hybrid of two species in Platanthera is in one sense a child of the two parent species, but in terms of the dictionary hierarchy it should occupy the same level as the parents, i.e. a child of Platanthera.

I suspect that this and ferret will be far from the only examples in the Rec 3.3 list. I have encountered others and would not be at all surprised if it is the norm for dealing with hybrids in this list. Might be a big job for someone to pick them all out and correct them. I do consider this to be a vital task though. I imagine it would be easy enough to run a query to select all taxa marked as hybrids in the list which are recorded as children of species-level taxa. Where the genus of both parents is the same as that of the hybrid, it should be possible to do a batch update changing the parent to the genus. It would be necessary to check for transgeneric hybrids though, which would become children of the family, sub-family etc.

I do recall another instance in the Recorder3.3 list though, reported a long time ago, but as yet uncorrected. In the genus Tilia (vascular plant - Lime trees) there are three true species listed, but each of these has a hybrid as a child. This is an important example as Tilia platyphyllos (Broad-leaved Lime) is Nationally Scarce, while the hybrid marked as its child Tilia x europaea is one of the most commonly planted trees in UK parks and gardens. It is really not good enough that a query for T. platyphyllos should return records of the hybrid.

Sure I could find a dozen other examples but that really isn't my job.

Thanks

Thanks Chris

While I agree that they are not actually mistakes, they are the result of a way of doing things which conflicts with the design of Recorder in its current form and cause reports to return erroneous, or at least unexpected results.

For what it is worth, my position is that if left uncorrected they will continue to cause errors in the outputs produced by Recorder users and the knowledge of their existence means that many users will have to continue expending time and resources checking for such errors. The few examples I and others have reported will only be the obvious ones.

If a decision can be made that the Rec3.3 way is the wrong way, then a fairly simple update query could be designed to do most of the fixing in one go. However it would still need someone to go through and pick out those where the hybrid should not have a genus-level parent.

As examples I would draw your attention to Frog Orchid (Coeloglossum viride) in the Rec3.3 dictionary, which has three hybrid offspring, all of which are also children of species in the genus Dactylorhiza. On the face of it these hybrids should not be the 'children' of either Coeloglossum or Dactylorhiza. Yet some authors now call Frog Orchid Dactylorhiza viridis.

Interestingly I note that the preferred Vascular Plants list has the three Coeloglossum X Dactylorhiza hybrids as children of Orchidaceae, but it also lists a number of hybrid genuses, including X Dactyloglossum, which is the correct parent of all three (but has no child taxa at present).

An alternative approach to checking through all 844 examples in the Rec 3.3 dictionary might be to modify recorder so that the 'expand taxonomic groups' option is restricted only to preferred lists (or optionally so restricted).

I guess the question is what is the best way of dealing with the legacy of the Rec3.3 list?

Hi Rob, I absolutely agree with what you say and this last question is really the nub of the problem for me. The Recorder 3.3 list has always had its problems but it is the largest pan-group checklist so it has been the list of choice for many Recorder users who just want to log mixed batches of observations. Just over a year ago Charles started constructing his own master checklist, called the ORGANISM table, and he completed it earlier this year before his retirement. Later this month we will be meeting the developers and stakeholders to explore future uses of this table in Recorder and other recording applications. An obvious use would be to help with this 'master checklist' issue. I will just postpone any decision on the Recorder 3.3 list until we have discussed where the other stakeholders think the best direction is. :)

That's an easy one Chris. The solution (partially) adopted by the vascular plants list as described above is the one. Within (eg) Orchidaceae there should be, in addition to the true genera, a number of pseudogenera such as x Dactyloglossum which should contain the intergeneric hybrids.

If you want to consider a genuinely difficult problem then what should we do with back-crosses between hybrids and one or more of the parents? Fortunately there aren't many people recording simple hybrids, let alone the more complex ones.

One thing that the Rec3.3 list has been especially useful for is recording aggregates and difficult to determine species. It is easy, in the highly technical discussions we tend to have regarding taxonomy, to forget that we are all encouraging public participation in recording and that records of poorly determined  taxa are going to become increasingly common.

It's already been running for quite a few years to my knowledge.

OK, I have sorted out the incorrect linkage between Platanthera chlorantha and Platanthera x hybrida. I think the other links should be OK too. I haven't tried sorting through the other hybrids in other genera but if anyone spots them do let me know. :)