A ecological process is something undergoing change (e.g., growth); whereas a mechanism is the thing that caused the change (e.g. nitrogen allocation), either directly or indirectly. In Ecology, we strive to find causal links between processes and the mechanisms that drive them. That said, I see these two terms used rather interchangeably in the literature.

In my view, the term "function" is more meaningful and relevant than "role". A function is defined in mathematical terms as an ability to change a factor or quantity (i.e. y is a function of x). A role is defined in terms of a performance or literary work (i.e. Pierre played a part in a story). In Ecology, we strive to describe a function using mathematical models.

A ecological process is something undergoing change (e.g., growth); whereas a mechanism is the thing that caused the change (e.g. nitrogen allocation), either directly or indirectly. In Ecology, we strive to find causal links between processes and the mechanisms that drive them. That said, I see these two terms used rather interchangeably in the literature.

As a follow-up to Georgianne's reply, we typically refer to something with a rate as a process in stream ecology. In this case primary production and leaf decomposition are the primary processes, and light, nutrients, disturbance, and food web structure are some of the primary mechanisms that control them.

To me (personal opinion, contingent on experience and open to revision) process and mechanism are both forms of causal relationship. Of the two, mechanism is more generic in the sense that some of the mechanisms by which things happen in ecological systems would only at a stretch be referred to as processes. For example, being eaten by a predator is a possible mechanism causing mortality in Zebra, but one would hardly refer to this as a process at the individual level. I think the key thing here is that some mechanisms can be essentially instantaneous at an ecologically meaningful time scale, in which case it doesn't make sense to think of them as processes, whereas others take measurable amounts of time at such time scales and are meaningfully considered as processes. What constitutes a meaningful time scale is contingent on the system in question and the temporal granularity of the interest of the observer.

role and function seems fairly close terms. Btw we speak of "functional role" of ecosystems. Ecosystem approach in the 2 key term function and service. "Ecological functions" are biological processes operating and maintaining the ecosystem, causing the maintenance of ecosystem services to society.

In my personal opinion, process is more fundamental than mechanism. Ecosystems consist of configurations of interacting processes. Autocatalytic dynamics are inherent in most such configurations, so that facultative selection can occur among component mechanisms. The result is that mechanisms come and go according to how well they contribute to the overall configuration (their functions). The entire scenario is fraught with (necessary) contingencies, so that prediction over the long-term is infeasible. It is my belief that ecology is largely a process-based science that is ill-served by the conventional "objects moving according to laws" metaphysics. I have tendered an alternative "ecological metaphysics" grounded in process thought. (See: http://people.biology.ufl.edu/ulan/pubs/3rdWindow.htm). Best wishes, Bob Ulanowicz

Thank you all for your explainations.

About process vs mechanism, I appreciated answers from Georgianne and Neil. Something still unclear however.

Georgianne, you said a process is something undergoing change and suggest growth as exemple.Fish body size or tree heigth for instances, are actually something undergoing changes, and growth is one of the process/mechanism responsible of this change, grazing may be another process/mechanism (for the tree exemple).

So I agree more with Neil, process and mechanisms are both forms of causal relationship, whith process that is more general, let's say at an ecosystem scale perspective (at least, not at an individual scale), despite this may appear subjective.

This sounds like a hierarchy, therefore. Could we consider that a process is composed by several mechanisms?

For instance, population size is affected by several "processes" including predation-induced mortality. Several "mechanisms" affect predation-induced mortality, as competition for shelter. Several "factors" affect competition for shelter: shelter availability and population size. (density dependance).

Telling this, is 'density-dependent mechanism' sounds as nice as 'density dependent process' ?

Thank you all

Best regards

@ Neil and Pierre......"This sounds like a hierarchy, therefore. Could we consider that a process is composed by several mechanisms?" This is self explanatory.........but pls, kindly assist with Ecological mechanisms determining the plant species

(a) composition

(b) Abundance

(c) Richness

in a Forest community?

Thank you.

Hello Pierre and Zubairu,

Its great to break it down and think of specific examples. I also agree with Neil's comments and Zebra example. But there are clear differences between these terms that are perhaps becoming more blurred. Allow me to try again to delineate them.

Use the term "mechanism" to refer to HOW the change or event happened. As Neil pointed out, it could be an isolated event or something more fundamental. In plant ecophysiology, we reserve the term mechanism for phenomenon that recur in some type of repeatable/predictable way. The mechanism explains the pattern! So a 'density-dependent mechanism', as mentioned before, would refer to instances where we explain HOW something happened differently, using density to differentiate the multiple mechanisms.

Use the term "process" to refer to WHAT is happening. As mentioned before by Thomas, processes tend to have rates (growth rates, mortality rates, rates of consumption by predators/herbivores). Therefore, I disagree that size is a process, it isn't. It is an attribute (or factor) we use as a point of reference to determine a process (e.g., size at time 1 compared to size at time 2 determines growth), but alone we cannot use size as a proxy for process. Another example: vegetation productivity at the landscape scale is measured with NDVI (infers greener parts of the landscape are more productive), but until we know how it changed, such predictions are likely wrong! We therefore must understand the link between greenness and biomass, and then how much it changed over time to predict growth.

Here is the distinction between process and mechanism...we can measure something repeatedly to monitor growth. But if we knew the mechanisms for growth well enough (solar radiation, soil moisture), we can model growth fairly well. Such is called a "mechanistic model", because it uses a priori knowledge of HOW growth occurs to predict growth. "Process-based models" refer to a similar set of models, but they may require measures of growth as inputs, while mechanistic models only require, for example, light and moisture.

I do not see these as hierarchical, but could be convinced otherwise. 

Looking forward to hearing what others think of this!

--Georgianne Moore

I would reiterate the subordinate position of mechanisms, which undergo selection by configurations of processes. My preferred metaphysics for ecology has moved to (http://people.clas.ufl.edu/ulan/publications/philosophy/3rdwindow/).