Here's a puzzling one...

We have added to the Measurement Type term list, a number of categories of observed evidence of the presence of various species and groups. This seemed to be the most appropriate way to record such data consistently. However we are now having a rather odd problem which is preventing import of some data.

The data in question is records of birds, with which we are using the BTO's codes for breeding status evidence (listed below). To contain the data I have added a new category to the term list which I have called Observation (it has been through a variety of incarnations but that wording seems to make sense when reporting on the data), which contains a single Measurement Qualifier called Evidence and a number of measurement units, each with lists of allowed values. One of these measurement units is Bird breeding which has allowed values as follows (in bold text) Of course only the letters in bold appear in the list of allowed values, the key below is copied from the text description in the term list.

X = Breeding behaviour not indicated or recorded by observer

Non-breeding = Nature of evidence not specified
F = Flying over
M = Migrating
U = SUmmering non-breeder

Possible breeding = Recorder did not supply evidence.
H = Observed in suitable Habitat
S = Singing male

Probable breeding = Recorder did not supply evidence
P = Pair in suitable nesting habitat
T = Permanent Territory (defended over at least 1 week)
D = Courtship and Display
N = Visiting probable Nest site
A = Agitated behaviour
I = Brood patch of Incubating bird (from bird in hand)
B = Nest Building or excavating nest-hole

Confirmed breeding = Recorder did not supply evidence
DD = Distraction Display or injury feigning
UN = Used Nest or eggshells found from this season
FL = Recently FLedged young or downy young
ON = Adults entering or leaving nest-site in circumstances which indicate an Occupied Nest.
FF = Adult carrying Faecal sac or Food for young
NE = Nest containing Eggs
NY = Nest with Young seen or heard

The data we are trying to import has just one of these codes for every record in a field called Breeding evidence. At the column selection stage we set this column as Taxon occurrence data/Observation of Evidence (Bird breeding)

After column selection, when the Validate records button is pressed a number of records are highlighted red in the breeding evidence column. Oddly however (and this is a behaviour I have witnessed before at this stage of imports, but cannot explain) as the cursor passes over each of the highlighted records, it disappears, until all are gone and there are no errors left. Even after the errors have gone though, the Next button is still unavailable and the text on the left of the window still indicates that there are errors to be corrected, but does not say what they are, so we are unable to procede with the import.

There is an interesting clue as to what the problem might be, because the items highlighted as errors all contain either D or F and you will notice that the list of allowed values also includes DD and FF.  Could it be that the importer is getting confused somehow?

I could try modifying the list of allowed values, but since these are the ones BTO recorders use I would rather not.

Any suggestions?