Acot and Drouin (1997) expose the progressive ingression of American theories of ecology within the European institution. At that times, a conceptual dispute divided the scientific community between Frederic Clements’ position (e.g., Clements 1916, 1936), and Henry Gleason’s one (e.g., Gleason 1926, 1939). To summarize roughly, for the first, the plant formation can be seen as a whole body, a complex 'super-creature' made of each species working with the other ones; for the second, plants are randomly distributed on the globe as individuals fighting for their own survival. Tansley (1935), who had cultivated a long friendship with Clements, thought, like him (and his fiery disciple Phillips), that some peculiar relationships exist between different species in the same place. Reluctant to consider the biome as an autonomous whole, which was far too transcendent to him, he sought and found a worthy exit out of the debate, and coined the word 'ecosystem'. This is not the place to review the whole debate, but in order to see how this opposition has been understood, and how its reception changed over years, let’s recall e.g. Westhoff (1951), Poore (1955abc, 1956, 1962), Whittaker (1962, 1973, 1978), Westhoff & van der Maarel (1973), Werger (1974), Barkman (1978), Beard (1978), McIntosh (1985), Nicolson (1990), Hagen (1988, 1992), Jourrnet (1991), Mirjin (1994), Palmer and White (1994), Golley (1996), Jax (1998), Nicolson and McIntosh (2002), Eliot (2007), Egerton (2009), van der Valk (2013), Guarino et al. (2018), Mucina (2019).

When Tansley (1935), proposed the new concept, in a way, and put an end to a long discussion, he sent Clements and Gleason back to back and, in doing so, he seemed to leave apart the concept of biome (created by Clements, like the one of climax), or more precisely, he sent it back to its origins: as a legacy, imagined by Clements, of the vague concept of formation, of ancient phytogeography (Grisebach 1838).

It seems that the ecosystem replaced the biome; it is defined by the association of a biotope and a biocenosis. As we can see, the word cenosis is part of the concept, but it is relegated at the same level as the biotope – a term that appears to be very difficult to define today.

In fact, what I wish to emphasize when I argue that the ecosystem studies have taken over in ecological science, is that studies on the biotope, considered alone (often confused with ecosystem itself) largely dominate ecology: the observation, the description, the classification of cenosis appear today to be accessory, except for what concerns plants (phytosociology), and practically not considered at all for what concerns animals or other organisms (with few exceptions, like littoral bionomics or rivers cenosis).

It should be noted, moreover, that the biome, which may have become synonymous of formation, i.e. the great bioarchitectures (essentially vegetal, but not only) recurrent on the planetary scale, is no longer directly linked to the communities (vegetal or other) that compose it.

In fact, Braun-Blanquet and his successors would have full opportunity to imagine a classification of plant communities independently of it. The result, at the end of the 1950s, was a complex phytosociological system (which did not call itself phytocenological yet), separated from both ecosystem studies and biomes studies.

The first ambition of cenology should therefore be, on the one hand, to extend the achievements of phytosociology to other living organisms (to re-equilibrate the biotope with its whole biocenosis), and on the other hand, to try to define the articulation between synsystems (already existing with plants, still to come for other organisms) and the classification of biomes: in other words, to articulate biocenology with biogeography ⁴.

Between the two classifying systems, phytosociology synsystem and phytogeographical classification, thus, some kind of gap exists - which I described as 'quantum jump' (Vincent 2021a) - and we must keep this in mind in order to quietly consider the cenosis/cenology statement. It should be noted, however, that when Braun-Blanquet (1932, 1964) or Passarge (1966) were seeking for a maximum superior unit, they inexorably needed a 'physiognomic' (a phytogeographical) concept: the first one spoke of 'vegetation circle', the second, expressly, of 'formation'.

To achieve these objectives we mentioned above, it is first necessary to define the different concepts of groupings of living beings that are currently used.

There are many way to classify species, and many groups can be proposed. They can be more or less precise, more or less imaginary, more or less arbitrary. We can separate for instance herbivorous from carnivorous animals, or perennials from annuals plants, or black from white animals, or edible from non-edible mushrooms, and even animals that belong to the Emperor from those that broke the pitcher… Clades, communities, assemblages, populations, guilds, associations, synusia, etc.: all these terms exist, are frequently used and are more or less familiar to us, but they are also often mistaken for each others. Thus, many researchers regret the lack of harmony and precision. I shall consider a few examples to try to retrieve consensual definitions and measure why this confusion is still active: Wilson (1991, 1994), Keddy (1993), Fauth et al. (1996), Looijen & van Andel (1999), Lockwood (2011). All of them contain a huge bibliography on the topic.

One of these papers (Fauth et al. 1996), regretting the confusion observed between these different terms, where usage often associates certain terms to certain groups, denoting the conceptual 'weakness' of ecological science, attempted to synthesise 'once and for all' the different types of groupings. The authors argue that groups can emerge from different conceptual approaches, depending on the type of relationship or link the members of the group studied have with each other. They define three 'sets' of relationships, 'delineated by phylogeny, geography, and resources', which would be the three main types one can find in the literature: phylogenetic studies are concerned with the specific relationships between the members, geographical studies with the portion of space where members are gathered, and resource studies with the grouping from the point of view of feeding, that is to say, for biological reasons, concerning metabolism - breathing is probably another side of the problem, however it is not mentioned (Fauth et al. 1996). They propose to represent those relationships in the form of a Venn diagram, which represents each of them individually, in a two-by-two overlap, and the central overlap of the three of them. I reproduce it in Figure 2.

The independent units, the first grade sets, are, respectively: taxon, community, guild. For the second grade sets: the taxon-community association (A+B) defines an 'assemblage', the community-guild one (B+C) a 'local guild', and the taxon-guild one (A+C) does not have a particular name, but can characterised by the description of its components. Finally, the set composed of the three sets (A+B+C) is, for the authors, the 'ensemble'.

As a synthetic review, it is obvious that several questions need to be asked:

1. Is it true, or sufficient, that only three ‘sets’ are useful, and not less (or more) than three? In what follows, I shall consider that is the case (there is no place here to discuss to a higher detail).

2. In calling the synthesis of the three groups an ‘ensemble’, the authors use a term which, in addition to being vague, is nothing less than rare in ecology, not to say very original.

3. Finally, it seems to me disturbing, on the one hand to designate a set by a void (A+C), and on the other hand to designate another set (B+C) by a name already used elsewhere.

4. Moreover, no account is taken of society, population, community in the sense of ‘settlement’, even crowd, etc., other common terms in ecology, and of course association, cenosis or synusia. How can these problems be solved?

First of all, I propose to keep the term community (what stands for ‘population’ for the authors) for any biological group of organisms, as biological societies, in the Taylor's (1935) or Phillips’ (1931) meaning, ‘irrespective of its nature’ (whether phylogenetic, geographical or resourcial), as stated by Mörzer Bruyns (1950). A synusia, an alliance, a life-form, or a clade should be considered as such (yet different) types of communities, and we’ll see below how those communities can be articulated with each others. I should exclude this term from the diagram.

Let’s review the other terms. The two-dimensions community defined by the overlapping of geographic and phylogenetic sets: as far as 'assemblage' is a random grouping of different species, and a vague term as showed by Looijen & van Andel (1999), I think it is not at its right place: the species maybe part of the same clade (as in paleomalacological sense, like in ‘quaternary tufa molluscs assemblage’ studies) but it is not necessary (let’s remind classic intertidal, or underground assemblages). The overlapping of geographic and phylogenetic community is clearly to me what is generally called a 'population' as in ‘a population of hippopotamus’, or ‘of edelweiss’. Therefore, the term 'assemblage' should be reserved for the mere first grade geographical group: there is no phylogenetic and scarce trophic links (in sociology it may refer to crowd or mass, cf. Grassé 1951).

The second two-dimensions group, defined by both geographical and resourcial sets, is not easy to resume; but there is a French term that fits: 'peuplement', in the English sense of 'settlement' (as in an human sense it refers to a mix of various groups gathered together with an economic link). 'Peuplement' should be literally translated in 'community', which is not usable a second time. It is not very satisfying, but I keep the French term in the diagram.

Let’s consider the third two-dimensions grouping, which is an empty space; this one is defined by both phylogenetic and resourcial sets, and is what I understand exactly as a 'guild'.

So here again, the replacement of terms calls for another word (for the resourcial set alone). But if we want to be very accurate, this metabolic, biologic function reminds of some kind of autoecological setting, as life-form. If acceptable, it would represent the best conceptual bridge between growth-form and life-form sciences (sensu Raunkiaer 1934) and applied ecology.

The new three sets should then be: resourcial > life-form; geographical > assemblage; and phylogenetic > taxon; even if the latter term is correct, I would prefer to use ‘clade’ (and keep 'taxon' in a linguistic sense: the name of a clade, whatever the rank be). Their crossing brings to population, peuplement, and guild.

The last step leads to qualify the set made up of all these elements. A community of organisms linked by phylogeny, sharing the same resources in a given space and time, isn't what we simply call... a cenosis?

To sum up, if we apply the arguments made so far to the Venn diagram of Figure 2, we obtain Figure 3.

One should mull things over, but it is hard to see how any community could be separated from one of the three dimensions: they are interconnected at the very root of both individual ‘organism’ and living ‘creature’; it is the matter of their own being.

Communities can be studied at least from three different points of view: the phylogenetic one (cladistic > taxonomy), the spatio-temporal one (chorology and phenology: what we should call territory, considered in a dynamic sense, whose studies could be called 'choriology', on the model of Figure 1's 'choriotope': "science of territory", from gr. 'χῶρος', but distinguished from the prefix of 'choreology') and the resources one (metabolic > biological or autoecological; see in Figure 4 Barkman's three kinds of merocenosis).

But I retain that we must consider the cenosis as the real basic-unit of any community of living beings. In clear, there is little chance that any creature is not at least partly linked either to a family, a species, a space and time, or any kind of metabolical function as breathing and feeding. What I mean is that there is little chance that, in ecological studies, we easily escape a cenological implication, where ecological elements (synspecies) share the same ‘household’, and therefore the same clades, the same spatio-temporal dimensions and the same resources.

Science can obviously speak about one of the second-grade sets, and that is what exactly happens, and that is very useful for the plain knowledge of life beings. But it is also true that the constant difficulty of defining and naming these different communities may indicate certain problems of conceptual permeability. First, it is possible that ‘groupings’ overlap is due to organisms overlaps too or, to put it better, that individuals share different groups contemporary. This is the great problem of heterogeneity and boundaries pointed out by Looijen & van Andel (1999). That is: there is a distance between concrete group of individuals and abstract communities (and that is why these authors prefer to speak about ‘community individuals’, which is only part of the solution). Since groups, as any classification types, are a self-oriented, mind-constructed reality, confusion must be admitted as part of the problem: and the articulation between concrete and abstract realities is part of the question phytosociology fronts since its debuts.

Maybe the problem stands in the conceptual lack of rigour shown by scientists and researchers, as asked by the philosopher Lockwood (2011). It is possible, but it is still an incomplete reason, many researchers demonstrating a strong epistemological stature.

What is much less emphasised, however, and which is spectacularly obvious in the diagram (apart from the game of musical chairs between notions) is, above all, the total absence of one of them, the loser of the game: cenosis.

Once we consider the cenosis as a logical minimal unit of ecology (rather than ecosystem), indeed, it’s still quite surprising, and extremely revealing, to note that the field of cenology has been gradually evacuated to the point where the term cenosis has become practically a rare and endangered one - whereas it seems to me to be the very foundation of ecology itself! Lockwood (2011) argues that it may be ‘equivalent to ‘community’ but the ambiguity in use is such that the term is probably not a clear expression of a particular perspective’, which is a clear statement of the actual situation (but not of the historical process). A simple research on any online library browser shows very significant results: on Google Scholar, the search of ‘ecological guild’ reaches 145,000 results, that of ‘ecological assemblage’ 475,000, ‘ecological communities’ peaks at 2,990,000, while biocenosis and cenosis respectively get 28,500 and… 4320.

Another noticeable clue is given by scientists' bibliographies: if specialists of the topic still refer to Clements, Gleason and Tansley battles, the whole school of the 1930s, the enormous continental European production of the 1950s is almost absent from the radar and nobody talks about great thinkers like Rioux, Rabeler or Mörzer Bruyns.

So, why have ecologists gradually abandoned the terms 'cenology', 'cenotics' and 'cenosis' (or have left them to botanists, which is the same thing)? Is it simply because of the confusion between roughly equivalent notions that they have gradually been marginalised? Probably not, since the others remain, on the contrary, prominently alive. It cannot neither be because the term itself is too complex. I therefore assume that one reason may be epistemological. What displeases or frightens the scientist is the part of subjectivity and interpretation in the observation, description and classification of life which formed the core of natural sciences in the past (until the 1970s) (Vincent and Catteau 2021; Cramer & van den Deale 1985). The development of statistics in ecology studies expresses the ever-growing positivism that nourishes science but also society, obsessed by numbers and evaluation. By the way, the ever increasing use of statistics in phytosociology shows what could be called a notable epistemological bias, since phytosociology retrieves its own matter from statistical methods! To improve the synsystem viability and validity on the ground, as it’s often the case with statistics tests, should not be the only task of phytosociology; it should be one of the methodological steps, but statistics must be considered as a tool, and never as a finality, as it seems to be nowadays. Statistical studies on various traits, but also gradients, patterns, and factors are much more about autoecology, and even biology, than synecology (another word for cenotics) and ecology.

Following again Lockwood (2011), the solution of such ecological and epistemological issues lies precisely in a perspectivist approach, balancing objectivity with subjectivity, calculation with interpretation (and I’d add: laboratory studies with ground studies). This predominance of numbers in ecological studies, the loss of inspired descriptions and theoretical interpretations are maybe the reasons why cenology increasingly declined. Factors, patterns gradients, traits, are all made of measurable quantitative information, but phytosociology, and biocenotics in general, as a morphophysics (de Foucault 1986) must point on qualitative information first. And this would resolve the other problems mentioned before, the ones of heterogeneity (the researcher decides which boundaries he states, in order to find a homogeneous milieu).

And the process is not over (as we see with old debates about phytosociology). In a recent issue of the journal Naturae, of the Paris Natural History Museum (MNHN), Yves Meinard and Gilles Thébaud even advised researchers, managers and naturalists, and I quote, ‘avoid using syntaxonomic identifications’ (Meinard and Thébaud 2019)! One cannot simply regret the lack of precision of the system or the difficulty of access to phytosociological resources (because this is not true). On the contrary, this strong stance tends to suggest a willingness to get rid of phytosociology's so-called stranglehold on current applied ecology.

Cenosis is a grouping within a clade. It is up to the researcher decide the clade in which he operates; phytosociology operates within the Viridiplantae clade; generally it leaves apart algae, and sometimes also ferns. We sometimes find cenosis associating lichens and annual plants: as usual, this is not problematic since this choice is well defined beforehand but, technically, it seems as risky as associating frogs with fish or mammals (even if they are all tetrapods) (see Berg et al. 2020).

This is no small problem because, even within the same clade, molluscs for example, further subdivisions can be found: Cephalopods can be easily distinguished from Gastropods; but in the latter, is it necessary to ask whether the pulmonates should be separated from the operculates? A priori no, but from another point of view, their way of life is totally different as concerns water resources! If we are interested in fishes, should we separate Chondrichthyes from Osteichthyes? A priori yes, but don't they form fairly a mixed group? And what about Cetaceans?

In fact, we can therefore ask ourselves whether we should not operate on the cenosis based on the notion of biological type (sensu Raunkier e.g. 1934)? This would mean, for example, that there would be groupings roughly comparable to strata or layers (for vegetation), but considered from an ecological point of view. This is where synusia comes into play, carried and defended by the alternating current of synusial phytosociology (Gillet et al. 1991; useful original definition in Gams 1918). The synusia proposed by ‘sinusialists’ are often (but not only) assimilated to biological types, being understood that the ecologies of annuals and perennials, for instance, cannot be similar, and therefore cannot be mixed together. It would then be a matter of determining the major biological types in animals, mushrooms, and else, which is no small task – especially as these probably differ more or less from one clade to another… (see Berg et al. 2020 again).

Many studies about zoocenosis try to link those zoocenosis to phytocenosis, but most of them conclude that it is not that easy, if not impossible, or even foolish. Indeed, 1) there are far fewer animal species in most clades (except arthropods) and it is probably impossible to get the same degree of precision such as in phytosociology; 2) many animal species are either extremely endemic or very common; and 3) big cenosis like certain vegetation types are at the same time inhabitants and habitats (for other cenosis): they are at the same time contents and containers – which is also the case for most organisms in relation to the world of bacteria.

There are two consequences of this: first, that all synsystems should be built independently from the others, at least at the beginning of the research; second, that each system can be linked to each other to reach, at the end, but in another dimension, the biogeographical one.

In order to reconnect with the original definition of plain ecology, we must define our topic. Molinier and Vignes (1971), for instance, after having stated that ecology (which they prefer to call 'ecobiocenotics') is the heir of environment (‘milieu’) studies and autoecological (biological) studies, claim that: 'ecobiocenotics results from the confluence of two great scientific currents [...]. It represents the culmination of the natural sciences known as ‘field survey’ [...] it could not be exceeded one day by a larger discipline. It constitutes a limitation towards which tend asymptotically, voluntarily or not, all the efforts of the naturalists who devote themselves to the study of the environment' (Molinier and Vignes 1971: 7).

'Biocenotics' (also called 'biocenology' or 'synecology') seeks to observe and describe ecological communities called cenosis, based on their recurrent specific composition, on a given scale of a given clade; those cenosis can be structured within the clade in an integrative system ('synsystem'), with specific ranks, from microhabitat and synusia up to the higher ones.

Rioux (1958) rightly points out that 'biocenotics' it is the branch of ecology that is precisely interested in living beings and their relationships (he then speaks of 'ethological' ecology); it is opposed to what he calls 'mesological' ecology, which rather studies the flows of matter and energy (i.e. the ‘ecosystem thought’!).

'Ecobiocenotics' integrates the system of biology, according to Gams, Du Rietz, etc. Gams (1918) stresses that there are four fields in biology: the relationship between species and between species and the environment, classification, distribution, history; these fields are considered either from a static point of view or from a dynamic point of view.

It relies on biogeography, which describes the great formations, or biomes. There is a meeting point (which is a tipping point) between the two, and it usually concerns what many ecologists or botanists call 'formation' and what I call elsewhere (Vincent 2021a) a 'biocen', i.e. the sum of all the biocenosis, within a biome.

It is thus also a synthesis (Gams 1918: 298; Rioux 1958: 129; Mörzer Bruins 1950: 1-3); of the knowledge of cenosis according to their singular features, that is to say, in the fields of systematics, genetics (or phylogeny), choriochronology, ethology, physiology, morphology (Table 1).

I will conclude with a few words about the spirit of the methods, which each cenologist can adapt according to the considered clade. It is inspired by phytosociology, which duly and regularly improved it. It is a structuralist method, that involves different stages: field and laboratory, analysis and synthesis, objective and subjective phases. This last point should not be surprising: the whole part of creativity, but also of imagination, but finally of interpretation requires this flexibility and sharpness that only the human eye and brain are able to provide (Figure 5).

Thus, on one hand, observation, inventory, sorting and data analysis, lead on an interpretation, a sort of a novelization that allows to apprehend the biological dynamics in progress, the interactions between the different cenosis, or between species within the cenosis. On the scale of a natural site, the cenosis provides a robust insight into the issues at stake, as well as the effects of possible human interventions. Then, monitoring is able to enter.