Agreed. However does incidental consumption count?
The teasel, Dipsacus fullonum is known to catch invertebrates in its water filled leaf bases, but experimental testing of reproductive benefits of this have been lacking. We report the effects of insect supplementation/removal and water removal during spring/summer on Dipsacus in two field...
journals.plos.org
I would argue, we need to tweak the working definition of carnivory in plants. It needs to include an evolutionary component to exclude the myriad "non-standard" ways plants take up nutrients besides from labile minerals in soil water through roots hairs, which are not primarily selected to capture nutrients from animals. I've been thinking this through as examples like
Philcoxia and teasel have come out in the literature. If we continue to use the same standards, we will find that many plants are "carnivorous", which I suspect from the stand point of natural selection is not true.
I believe we need to move toward a working definition of carnivory that is more like this one:
Carnivory in plants is occurs when plants have significant and costly adaptations which are primarily under selection to capture nutrients directly from non soil mineral sources in the kingdom Animalia.
Even this I argue needs work. It leaves out the old clade of protists which might be actively captured, but I think it is a start. It would exclude things like teasel, and
Philcoxia (which may well be coating its leaves primarily as a form of protection from light in its environment, with "prey" capture being incidental), but include
Heliamphora,
Darlingtonia, and the
Nepenthes which are, shall we say, directly manured. It also ends what I believe to be a tiresome debate about whether or not to include
Roridula, which this definition includes. It does not include the autonomy in carnivory, but I have been thinking about those as well. I've not yet though about to present them in a way that avoids the pitfalls of the
scala naturae in the non-directional nature of evolution. Plus, most people probably think I'm off my rocker thinking about the definition this much, let alone how to create catagories within the new definition.