Depends on your definition of "keystone", but no.

A top predator could have little impact, especially if it is a minor player in a guild of top predators, or if it is limited by something other than food.

I agree with Dr. Truman P Young answer. Just as a follow up, I think would be very hard to indentify a keystone species just with the food- web topology. You can argue that there is a single species that may serve as resources to many species, which are resources for many others. In this case, you can say that this single species is very important, with a great "impact" on the food web. But what about it's density? If you use the most accepeted definition for a keystone species, it's effect must be disproportional to it's density. Thefore, just with the topology, you may not get all the information you need to apply the definition.

I think a top predator is very likely to be 'un-substitutable' and thus a keystone species.

I'm not so sure that the theory of keystone species really works in the real world. I would suggest for a top predator, their existence is dependent on the availability of suitable habitat for the species and for their prey species. One cannot exist without the other. If you wish to compartmentalise a subsets of ecosystem to allow simplified analysis of processes then the concept of keystone species may be of use. But, even then, the core concept would need to focus on keystone species - plural NOT keystone species - singular.

Just after I posted this, I came across this article looking at corridor selection systems. It is not about predators but it does discuss the need to take in the complexity of ecosystems in order to understand processes. Article Does one size fit all? A multispecies approach to regional l...

I like the question, and I largely agree with Turman and Rafael. While a top predator may not be a "keystone" species in some, if not many, cases. Top predators may not have the direct or indirect effect on communities, depending on their numbers, foraging habits, and other factors. However, top predators may frequently be "umbrella" species. They tend to have large ranges that require a diversity of habitats. As a result, the protection of top predators has indirect conservation benefits to the other species that use those habitats.

Top predators? No, absolutely not. Termites, ants, bees and many mushrooms ans other microorganismes are on top of the keystone function — if there is any.

If one or more species in a community/habitat plays an important role in protecting and sustaining that community/habitat, it or they are keystone species. Thus, being a top predator is not sufficient quality alone for being identified as a keysone species!

What's changed in our understanding of the keystone species concept? Of course top level predators can be keystone species! Paine's original example was the sea star, Pisaster ochraceus, and sea otters, wolves and jaguars have all, also, been described as keystone species. The latter two are clearly apex predators. I won't comment on jaguars, but the evidence for wolves as keystone predators seems quite strong. So, I'd argue that high level or apex predators CAN be but are not always keystone species.

I wonder if interaction strength could be used as a (partial) measure or indictor of a keystone species. See for example

Ripple et al. Status and Ecological Effects of the World’s Largest Carnivores. 2014 Science 343 (6167): 1241484 DOI: 10.1126/science.1241484

Sala, E. & PKD Dayton. 2011. Predicting strong community impacts using experimental estimates of per capita interaction strength: benthic herbivores and giant kelp recruitment. Marine Ecology 32: 300-312 DOI: 10.1111/j.1439-0485.2011.00471.x

Sala, E. and M.H. Graham. 2002. Community-wide distribution of predator-prey interaction strength in kelp forests. Proceedings of the National Academy of Sciences 99: 3678-3683.

Estes, J.A.; Tinker, M.T.; Williams, T.M.; Doak, D.F. (1998-10-16). "Killer whale predation on sea otters linking oceanic and nearshore ecosystems". Science. 282(5388): 473–476. doi:10.1126/science.282.5388.473.

I'm also intrigued by the possibility that the "interaction" need not be a direct, trophic interaction:

Gómez, José M.; González-Megías, Adela (2002). "Asymmetrical interactions between ungulates and phytophagous insects: Being different matters". Ecology. 83(1): 203–11. doi:10.1890/0012-9658(2002)083[0203:AIBUAP]2.0.CO;2.

Top predators always need not be the keystone species, as a keystone species in a food web is always determined by their importance in that ecosystem. Usually a keystone species is one which had direct effect on communities, and in many cases those organisms which occupy the lower trophic positions can serve as keystone species. It depends on the role played by that species in a particular ecosystem.

According to "The concept of keystone species in the ecology of the restoration: key ... or impasse?" by Robert Barbault (Institut Fédératif d'Écologie Fondamentale et Appliquée), CNRS, Paris (France):

Whether we try to understand how it works or whether we intend to manage or restore it, ecological systems oppose us, by the multitude of species that compose them, a major difficulty. To hope to overcome the obstacle by reducing such hyper-diversified sets to a few trophic levels appears as unsatisfactory as the indefinite accumulation of species monographs.

Born in the sixties, the concept of keystone species is now a new craze: in the current debate on the conservation of biodiversity it appears as a privileged access to the complexity mentioned above . Because it postulates that some species are more important than others, playing a pivotal role or crossroads in food webs, this concept suggests that this could be a focus of attention and effort on a small number of species, structuring species, neglecting all others that depend directly or indirectly on them. In fact, it seems that this concept raises more difficulties than it solves. Although it does not by itself constitute a miracle tool that can be used directly in the ecology of restoration, it nonetheless appears as structuring by the underlying ideas that it conveys:

- the idea of ​​an ecological system (which opposes the first perception of a multitude of inconsistent species);

- the idea of ​​interactions between species (logical consequence of the first point);

- the concept of a population-environment system that integrates all the benefits of population biology.

Also, without constituting a miraculous key but far from being an impasse, the concept of keystone species finally leads to interesting prospects for the ecology of restoration.

Oum Kalthoum Ben Hassine very well said. Would it be as simple to say the sum of the ecosystem characteristics and species are what define keystone species. In my work, I could define keystone species on temporal and spatial scale and interactions with other elements witin ecosystems and come up witha different keystone species at almost every turn.

Adding to the answer of Andreas Speich...Some of our most important keystone species are not top predators or predators at all...think beaver.

Thanks for all the answers! I do not know that the concept has been growing that much since I first learned it five years ago. With such complicated position of the concept, how could I test this concept to undergrads?

Identifying the key species or species in a trophic network is a complex process. You can not always assume that this species is the top predator. In these articles you will find methods for the identification of key species in trophic networks.

Libralato, S., Christensen, V., & Pauly, D. (2006). A method for identifying keystone species in food web models. ecological modelling, 195(3-4), 153-171.

Resetarits, E. J., Cathey, S. E., & Leibold, M. A. (2018). Testing the keystone community concept: effects of landscape, patch removal, and environment on metacommunity structure. Ecology, 99(1), 57-67.

Ortiz, M., Hermosillo-Nuñez, B., González, J., Rodríguez-Zaragoza, F., Gómez, I., & Jordán, F. (2017). Quantifying keystone species complexes: ecosystem-based conservation management in the King George Island (Antarctic Peninsula). Ecological Indicators, 81, 453-460.