Let me start with the origins of the first paper the students discovered, that on Resilience (Holling, C.S. 1973. Resilience and stability of ecological systems. ARES 4: 1-23.). Since that paper really opened my eyes to the ecosystem scale, I’ll then spend a bit more time referring to it, and how it originated.

That paper came from a series of earlier experimental studies and papers analyzing a particular process, predation. The goal was to see how far one could go by being precise, realistic, general and integrative. These are goals that normally are dealt with independently in at least partial isolation from each other in order to achieve useful and useable simplification. (The key, classic references are (Holling 1965 & Holling 1966).

Those studies did well, and eventually led to a way to classify categories of predation into four types of functional response (how much they eat) and three types of numerical responses (how many there are). The categories and resulting simplified models seemed to apply to everything from bacteria foraging for food to submarines hunting ships! But none of that was ecosystem research. It was all traditionally experimental and analytical; but at least it was synthetic, non-linear and had great generality.

The key conclusion relevant for ecosystem science, was that it was possible to develop small suites of well tested realistic models and define a small number of general classes of responses for key population processes. The marvelous dean of ecology at that time, Bob MacArthur, wrote me at the time of the publication of the first Functional Response paper, arguing the work was too detailed and complex to be very useful for theory in ecology. That is true in a narrow sense, but he did not know that the paper was a planned step in a process that finally did yield less complex equations, but ones more complex than was traditional for the theory of the time. The “somewhat more complex”, however, led to a world of differences in the behavior of systems, because of the non-linearities in the processes. And, most important, the equations representing the various classes of processes, were sufficiently realistic, something I thought then, and now know, was a central need for further development of theory for ecosystems. That was the first hint of the “Rule of Hand” – not too simple, not too complex- that was highlighted in the conclusions to the book Panarchy (Gunderson and Holling, 2002). That is, all that is needed is a handful of key variables. The classic “disc equation experiments” and paper launched the whole sequence that led, finally, to simpler mathematical representations that captured the essential reality that I thought was needed (Holling, 1959).

The same simple equation and experiments also became the foundation for the development of optimal foraging theory, when Eric Charnov joined my laboratory as a visiting student at the University of British Columbia. He accepted the basic construct of the disc equation, that the time available for a predator was divided into time spent in various categories of search, prey handling and digestive pauses. And I gave him the wonderful data I had collected from experiments with praying mantids, that he then used to show that optimality emerged in prey choices by predators. I never went in that direction myself, but many others have, and so a well tested theory of optimal foraging developed, launched from the infamous disc experiments and equation.

My interests were more focused on describing and integrating the components of behavior to add generality. That is what ended up in the Functional Response papers, where the effects of hunger, learning and avoidance were shown experimentally in a way that permitted expansion of the disc equation. Truly the work began to be applicable to predation by insects, birds, mammals and fish. As an example, one of my students even enjoyed himself in eastern Africa observing the distances of stalk and attack of lions attacking gazelles and wildebeests, He was in one vehicle filming the action as his wife did the same in a protective cage, some distance away. Binocular perception allows calculation of distances between predator and prey! It was a fine piece of work with as much a consequence for understanding the co-evolution of attack and escape strategies as for behavior. And there were a number of other such generalizing examples and tests.

It ended up being truly general, leading, ultimately, to the four basic types of functional responses and equations for them. It also became the point in the early 1960’s, where I discovered the tremendous value of simulation models. The expression of the experimental results into a generalized model of predation, showed me how significant the new programming languages and computers were in explosively expanding our power of understanding. They made it more natural to represent non-linearities of various kinds. And projections of the results were dramatically easier.

But it became completely clear that some rigor had to be applied – don’t try everything; just expand slowly on the basis of what we know. Then slowly add the partially known and unknown guesses, testing against the reality of whole systems behavior along the way. The work therefore avoided the tendencies that exploded in the International Biology Program of the time, where often more and more was expressed about more and more, in a way that smothered the work in over-complexity. A simple thread of modeling and investigation became much more powerful. Again, that was the discovery of the “Rule of Hand”—complex enough, but not too complex.

At this point two paths opened. One was marvelous work with various beasts to expand the behavioral discoveries. That was done with Larry Dill, a well known behavioral ecologist, who, in my view, is simply the best whole animal experimentalist in the world! He has developed elegant and insightful explorations of salmon and killer whales on the west coast, of archer fish and their aerial prey, of dolphins, dugongs and sharks in Shark Bay, Australia. We developed experiments with small fish reacting to barracuda and model predators, and to mock situations with an endless patterned belt that showed that their movement, once they were located on the edge of their “zone of fear”, was dictated by the appearance of a corner- no corner on a belt, therefore no movement. Of mahi-mahi attacking prey, and of schooling and solitary fish in a Hawaiian oceanarium disturbing and reacting to potential predators. And on to ducks off the coast of British Columbia reacting to boats; and of school children in a field reacting to a runner as an aircraft filmed the interactions. And to flocking ibis in Florida sketching V’s, W’S and Y’s in the air as they flowed, almost with magic, from foraging grounds to nesting islands. Beautiful situations – not work at all, but full of the joys of understanding the patterns of life.

All showed the foundations we earlier had discovered. That is, there were general laws, expressible in general equations of fairly simple form that explained all the variants we observed and filmed, each as a limiting condition of a general equation. Moreover, we discovered that the rules were not precise and accurate, but rather were simple and just sufficient. In short, they were “quick and dirty” and were adaptive. Adaptive options are retained to correct a response if a mistake is made.

So we concluded that nature does not optimize for the “best” based on assumptions of complete knowledge, in the traditions of simple decision theory. Nor are its responses efficient. The actions were based on just sufficient information to assure adequately the object’s fundamental nature and provide options for reversal- likely small enough to attack safely vs. likely big enough to avoid. That is all strictly the consequence of evolved responses. It has the same features that later characterized the tactics and goals of Adaptive Ecosystem Management that Carl Walters and I later developed for designing policies and responses for managing resources in ecosystems. The mistakes become possibilities for learning, not routes to failure. Larry and I began to write a book, but that book, partially done, still waits completion.

That was because another path began to swallow my attention. This path moved me into very new territory, that was truly ecological. By that I mean I began to recognize that the way organisms are affected by their environment is only half the story. The way that effect feeds back to affect the environment itself is the other half. That interaction creates new structures, some at one scale, some at others, and those create new options for evolutionary change. At what scales was that significant? Were all species interacting in important ways? Or were there a few that developed relations among themselves and their environment that created new entities upon which evolution and human management acted? All that was launched by my discovery, or invention of resilience.

References:
Holling, C.S. 1959. The components of predation as revealed by a study of small mammal predation of the European pine sawfly. Canadian Entomologist. 91:293-320.

Holling, C.S. 1965. The functional response of predators to prey density and its role in mimicry and population regulation. Mem. Ent. Soc. Can. 45: 1-60.

Holling, C.S. 1966. The functional response of invertebrate predators to prey density. Mem. Ent. Soc. Can. 48: 1-86