Petals connate (at least two of them) ?

[BubikolRamios1]

As I understand that it means they are fused together.
All members of compositae should have that according to this: http://www.colby.edu/info.tech/BI211/PlantFamilyID.html
Wrong, right ?

[Anagallis]

All members of compositae should have what?

I think you take this family characteristics stuff too literally. It's not a definition but a listing of characteristics. While Asteraceae (always?) have fused petals, that does not automatically mean that there are no exceptions. Even if there are no exceptions now, one might evolve in the future.

[BubikolRamios1]

http://www.colby.edu/info.tech/BI211/PlantFamilyID.html

95. Petals all free from one another
96. Petals connate (at least two of them)

If you select 95 and see listing , it contains compositae
If you select 96 and see listing , it does not contain compositae

I think that combination directly "litteraly automatically mean that there are no exceptions" = Petals connate (at least two of them) (95),
for Compositae.

I know, It might change in future, some species moved somewhewre else (or already did because listing is outdated), ... But in general, for my purpose (cross checking) accurate enough.

[Anagallis]

95. Petals all free from one another
96. Petals connate (at least two of them)

If you select 95 and see listing , it contains compositae
If you select 96 and see listing , it does not contain compositae

Vice versa.

[Anagallis]

I find this "punch card" system a bit pointless. To understand that the petals of Asteraceae are fused, you need solid knowledge about the antomy of their inflorescences anyway, and if you have that knowledge, you already know the family.

[BubikolRamios1]

Not completely pointless, at least data behind can be extracted/analysed to some use. Someone spend a lot of time gathering/checking that data so we don't have to.

[Marlies]

According to http://www.colby.edu/info.tech/BI211/PlantFamilyID.html species in compositae can have connate petals or valvate petals. (Don't ask me, which of them.) I'm afraid, this distinguishing feature is too difficult for unexperienced botanists. This seems to be the general problem with the key, as Anagalis already mentioned above.

[Anagallis]

Actually, "valvate petals" (= touching each other but not overlapping) probably means "valvate ray florets". But in Asteraceae, all petals of a floret are fused either to a tube or a "ray" anyway. The "valvate petals" check box seems to be some help for people wo don't know the general anatomy of Asteraceae florets.

[BubikolRamios1]

That is:

79. Sepals all free from one another
80. Sepals connate (at least two of them)

This two are opposite.
Valvate is another separate thing. I think.

[Anagallis]

But petals cannot be fused and valvate at the same time.

[BubikolRamios1]

To define:
https://en.wikipedia.org/wiki/Aestivation_(botany)

valvate – margins of adjacent petals or sepals touch each other without overlapping.

Right, they are not fused.

Hence I think "95. Petals all free from one another" contains also valvate.