I am not sure we can make the distinction. You never now when one "constructor" change the environment (abiotic or biotic) if it will have a feedback or not on the "constructor". The feedback could be now or later, indirect...

First, thank you for your answer!

That is the main reply from the authors when they receive critics like those, and are a fair one, but something I think that we need consider is the dynamic of the feedback, if it happens immediately or within a short-time number of generations, we'll have response to selection of any kind (genetic, epigenetic, cultural, etc) that will maintain that activity/behavior that affect niches. The most convincing examples that they presents are in that class. Cultural coevolutionary phenomena, animal construction that affect survival, plants changes in soil, ecological coevolutionary scenario, sexual selection: all imply direct short time feedback. Ecological Legacies require that.

Beaver's dam shows that, but oxygenation of atmosphere doesn’t. Photosynthesis was selected for other reasons, and is a very strange say that how it change niches (including of the photosynthetic organisms) was the same that the other cases. But I not saying that these cases are not important, only very different.

Do you think I'm missing something?

I agree that the strength of the feedback need to be incorporated in the model

I agree with your examples, but I am not sure it is always so straigthforwrad.

If you think about the most classical example of bio-cultural interaction in human: lactase persistency. The change that constructor made, was to use dairy product after domestication of cattle. The feedback, (to be able to digest this milk, thanks to the LP mutation) was only possible after the increase in frequency that took several generations. At the begining of the niche construction almost no one was carrying the LP mutation...

I will not be able to follow this topic next week...

I think this scenario too involves a strong feedback, if the lactose absorption gene existed in low frequencies, and there was complications that would be deleterious and/or lethal to individual lacking that mutation, selection operate since the beginning in changing its frequencies due non-random differential survival and reproduction of those who had the lactose absorption genes, and the spread for whole populations took time, but that’s like usually happens in all cases of gradual selection.

The definition of Odling-Smee, Laland and Feldman is somewhat flawed, as they mix up niche with the habitat/environment conditions. One should make clear that modifying the surroundings abiotic/biotic factors to construct the conditions required for the species is one thing, the other is the effective change in the requirements (niche evolution). There is a very good recent review on the several concepts related to niche construction:

Linder, H. Peter, Olga Bykova, James Dyke, Rampal S. Etienne, Thomas Hickler, Ingolf Kühn, Glenn Marion, Ralf Ohlemüller, Stanislaus J. Schymanski, and Alexander Singer. "Biotic modifiers, environmental modulation and species distribution models." Journal of Biogeography 39, no. 12 (2012): 2179-2190.

The distinctions you and other researchers and philosophers of science make regarding niche construction are valid, however, I wonder if making any distinction regarding a particular type of niche construction as opposed to using the umbrella term itself isn't driven by function. Specifically, if one's research were within the domain of direct or "true" niche construction, then certainly due to the need to communicate clearly that distinction should intentionally be made.....but actualy might occur by default anyway. Further, this question you have appears to fall under the heading of "splitters versus lumpers", terms originally used to describe disparate taxonomic approaches, but which might be appropriate in this discussion. From my perspective your question seems to suggest that for your own purposes a clearly delineated and narrow definition of niche construction is appropriate, if so, isn't your question already addressed?

Antony,

Thanks for your answer

My question is not for a specific research, but to search for a semantic precision when dealing with some given phenomena. When we dealing with evolutionary feedbacks, this address the inheritance requisite that constitute Darwinian evolution. The ecological inheritance presupposes feedbacks, direct or indirect. More confusing is that in some papers and some examples they only consider the "true" niche construction. All formal models use feedbacks too.

Would be helpful for they organize more precisely their examples and terminology. And I agree is a matter of limper/splitter until some point, but with objective practical consequences for evolutionary biology.

I agree that a priori, we can't know that if something is or not subject to feedback. But when dealing with a posteriori examples that was not subject to feedback, like atmosphere oxygenation, the effects in OTHER species, WITHOUT feedback, are very different, and I think really misguided consider those thinks like one process, like the author describe it. The way that cases like those influence evolution are very different from when we have feedbacks, so why call two different things by the same name?