Dear Mark,

in my humble opinion of a young investigator I think in pratical terms on of the hypothesis for measuring the niche (for salt marsh vegetation for example) could be throughout the space where some physico-chemical conditions prevail and are maintained (in a dynamic and seasonal mode) allowing a determined specie or group of species to colonize and survive completing their life cycle.

Thanks Bernardo, but what if a sympatric species that is a facilitator dramatically expands the physical space and range limits a species can liive in because of habitat amelioration? Habitat space must typically be an important dimension of the n-dimmensional  hyperspace, no?

Yes I agree completely. Probably the best definition should be a compromise between both a physico-chemical definition and a dimensional one, considering that within this space the characteristics are probably the same if they are increasing.

Very interesting question Mark, I never though that such a simple and commonly accepted definition could generate such an interesting discussion!

Could you rephrase your question in terms of niche only (without bringing in habitat) or vice versa? Maybe rephrasing could lead to an answer.

No - That is the question!

Dear  Dr. Bertness.  As you know,  although the "niche concept" may seem clear, its management is often very complex. For instance, I work with a cold water fish at the edge of its natural distribution. When we talk about the physiological limits of a species (such as temperature) it is often found that fish suffer other constraints (e.g. Interspecific competitive relationships) reducing the amplitude of its thermal niche. Thus, we might say that the realized niche is characteristic of a given population (it being inserted in given community) more than of the species itself.

I think that when you study an extinted species you are finding tracks of its realized niche and only a comparative study of different communities may help you.

I'm sorry for my awful english. I hope explained well, and I wait expectant the results of your interesting work.

Mark;

     A teacher's characterization of the niche is something like "the niche is the creature's job description in the habitat". So, qualities like thermal tolerances and thermal limits are about niches. Dietary inventories and proportions, diel and seasonal utilization are also niche parameters that can be readily measured. Adoption of the n-dimensional niche idea creates a visualization of what a niche might be like and suggests appropriate statistical analyses.

     The Functional Groups idea used by many ant ecologists emphasizes the notion of a trophic niche and when a species list for a place is assembled the functional groups seem to fall out nicely along generic lines...that is species in a genus share a Functional Group. The differences in habitat utilization among the species in a genus evidently satisfy the Competitive Exclusion principle. If this idea turns out to accurately describe the interactions among related species, that would be lovely because then the taxonomists would have provided a handy first approximation of the niche for any taxon.

     This problem is most lovely one!

The realized niche should be equal to habitat utilization.  The fundamental niche would only ever be equal to habitat utilization when neither competition nor facilitation nor predation/herbivory or disease are defining habitat utilization. 

I think most ecologists are interested in the fundamental niche, so I would say yes that defining the niche based on habitat occupancy in relation to chemical and climatic attributes are bogus. 

Some researchers have suggested that limitations in our understanding of the niche are driven by a lack of detailed look at n-dimensional hypervolume, suggesting that we just don't measure enough, but this doesn't address the problem you're highlighting, that we can only ever really measure the realized niche very well.  Even if we use experiments to address what conditions a species may exist in when biotic influences are removed, it's not clear. For example, many trees can be planted and grown successfully way outside of their ranges and normal conditions.  One only has to go to a university with an arboretum to see examples of this.  So dispersal limitations are also troublesome in this regard. 

I'm not sure this is helpful.  Good luck with the paper

Given some additional thought, I'd say the other issue is that proponents of the view that we need to measure more variables to define the n-dimensional niche are probably incorrect.  I think the nature of abiotic variability is that species generally can't specialize too finely on narrow portions of abiotic gradients because the narrower you specialize, the more spatially rare the fundamental niche should become.  Also, a narrow portion of the gradient will likely be more temporally transient.  In my mind, this means smaller species can probably more narrowly specialize on a given gradient. I have a few papers that look at this for plants, if you're interested.

Anyway, what I'm trying to say is that if you look at species responses to gradients, they tend to respond strongly to a few gradients, so I doubt understanding the niche better is a matter of measuring more things.  That said, I suspect that it could be the case increasingly as species size decreases; i.e., more true for a grass than a tree. 

Years ago Dan Janzen mused about species "fitting" in a community, (Janzen, 1985. On Ecological Fitting. Oikos 45: 308-310). In various parts of its range each species - excepting narrow endemics- interacts with different species assemblages. The species may be rare at one site and quite common at another. If these differing abundances are due to differences in the realized niche of our species but the interacting species aren't the same ones, then explanations for differing utilization of one niche parameter or another become pretty tenuous.

Common garden studies can at least address the abiotic contributors but they entirely  leave out the biological interactions suggested above. Following the cottonwood literature makes these problems quite clear.

To stimulate more discussion:

The relative concept of the 'niche'

Some people claim that different species do not share the same 'niche', for instance defined as an X-dimensional or endless-dimensional space where individual species live. Some would even claim that there is no overlap in the niche across species because of natural forces such as inter-specific competition. However, the level of niche overlap between individuals or species will probably depend on scales of analysis. For instance, unicellular or multi-cellular individuals or species should express niche overlap when the niche is defined with a small number of variables. Each unicellular or multi-cellular individual probably have their unique niche when the niche is defined with X-dimensions or endless dimensions.

Is the size of the niche perception-dependent or place-dependent?

The definition of the niche will evidently depend on what variables are empirically studied. If only one dimension is used to define a niche, such as the place where an individual organism lives and moves, does an individual vertebrate and an endoparasite (e.g. ancient mitochondria) living inside an individual vertebrate share exactly the same niche or not? Both the vertebrate and the organism living permanently in or on the vertebrate move together and therefore may share the same niche. However, the average physiological environment of the endoparasite will not be the same than the average physiological environment of the vertebrate, so individual endoparasites and individual vertebrates would not share the same niche from a physiological point of view.

If information mentally perceived are defined as resources exploited by living beings, does this imply that the size of the individual niche depends on the spatiotemporal scale of perception mechanisms involved? The perception niche of physical contact will probably be substantially smaller than the perception niche of a visual system or the perception niche in dream expressions. How to identify dynamic perception niches in wildlife, accepting scientists can never obtain detailed information provided by wildlife not been able to explain what they see, what they smell, what they hear, what they dream, etc.?

Different species should express significant overlap in niches when a given species explores niches of other species, for instance to find energy or material for living (soil, wood, plants, animals). If humans explore any environment in the world, for instance because of curiosity or to get resources for living, what fraction of the so-called human niche is shared with niches from so-called non-human organisms or species? Does this imply that both the individual-specific or species-specific niches are probably difficult to define with precision from an empirical point of view?

Could a niche be defined as a probabilistic place of living or a probabilistic space of perception occupied by an organism with a probabilistic design? Accepting that Earth as a niche of living is permanently moving either around its axis, is permanently moving around the sun, and is following moving trajectories in the universe, would this imply that the definition of the niche changes with research domain (e.g. astronomy versus earth ecology), also to provide additional arguments that individual niches are unique in time and place for at least one scale of analysis?

Accepting nothing is fixed an individual will always express continuous spatiotemporal variation in physical design for at least one level of analysis, and therefore could be defined as a probabilistic unit.

Niche differentiation across species exploiting artificial winter feeders

Artificial food such as fat balls hanging from a wire attract small passerines like great or blue tits or chickadees, but also attract other species such as robins or house sparrows having niches different from that of tits/chickadees. When a fat ball is attached on a wire 10-20cm above the ground, tits will use the fat ball as a perch to get food of which some is falling on the ground. Sparrows or robins will never touch the fat ball, but just wait under it until the food extracted by tits is falling on the ground. Why robins or house sparrows did not evolve simple niche shifts to exploit the artificial fat balls? Why is species-specific plasticity in habitat exploitation reduced, even for very simple food searching tasks, like extracting food from fat balls hanging from a wire? Do some species through their foraging activities provide food to other species, as illustrated with the case described above?

What is the size of a 3-dimensional garden, property, or territory?

Properties, gardens, or so-called territories are most often defined using 2-dimensional measurements expressed in length and width, but how to define a property, a garden, or a territory using 3 dimensional measurements? For instance, is the 3-D of a garden limited by the height of the tallest trees in the garden and the deepest roots in the garden, or not? Does this imply that 3-D of a garden continuously changes? And what if someone can build a ladder of more than 1 km high? Are the people that are climbing the ladder still in the same garden? What is the highest acceptable height defining the 3 dimensions of a garden? If an airplane is flying over a garden, is it flying in the garden defined with 3-D, or not?

Other questions

+ What are consequences for estimates of intraspecific population densities or interspecific densities influencing life-history expressions within species (e.g. lay-date, clutch size)? For instance, tropical forest research indicates that different species are observed at different heights in the same trees, which implies that different species occupying these trees may have the same territory or niche size at a two-dimensional level, but not at a three-dimensional level.

+ How to define a territory or niche when every individual movement is unique in time and space?

+ How far should we go in the past to define a wildlife territory of an individual? If history is not important in the definition of a functional territory, can the functional wildlife territory only be defined through current position and perhaps probabilistic future movements?

+ In an ever dynamic world, the individual functional territory probably equals the individual space of movement best represented by a network of traces on a map.

A niche is one-dimensional. The influences or controlling factors are n-dimensional. Define the niche. Identify the influences or controlling factors. Account for the influences and controlling factors. Marcel's excellent answer shows how to bound the influences and controlling factors.