The very etymology of Ecology, from the greek 'Qikos', 'the household', implies that ecologists should devote some attention to the 'house' or habitat of the population or community they are studying. However, as Charles Elton (1966) has so forcibly pointed out, 'definition of habitats, or rather lack of it, is one of the chief blind spots in Zoology'. Elton himself has provided us with a qualitative classification of habitats, while another past President, Alex Watt (1947) highlighted the dynamic nature of habitats by his phrase, 'pattern and process'. Elton referred to the need to quantify habitat characteristics. In this Address I will attempt some quantification;however, you will all be aware that in doing this I will not be able to emulate those former Presidents who have been able to provide a definitative synthesis of a field or of their own studies, my offering can be but a small beginning, an indication of the type of characteristics we should quantify. In considering ecosystem patterns and environment R. M. May (1974) writes 'it is to be emphasized that although patterns may underlie the rich and varied tapestry of the natural world, there is no single simple pattern. Theories must be pluralistic'. Indeed, the complexity of the subject is daunting and in any attempt to formulate some type of general framework, one is continually beset with exceptions. In stressing the need for a framework I am echoing a plea of my predecessor Amyan Macfadyen (1975) who cited K. E. F. Watt's (1971) vivid image 'if we do not develop a strong theoretical core that will bring all parts of ecology back together we shall all be washed out to sea in an immense tide of unrelated information'. In some ways I think we may see ourselves at a similar point to the inorganic chemist before the development of the periodic table; then he could not predict, for example, how soluble a particular sulphate would be, or what was the likelihood of a particular reaction occurring. Each fact had to be discovered for itself and each must be remembered in isolation. It is noteworthy that from Dobereiner's early efforts in 1816 it took more than fifty years before Mendeleeff ormulated his Periodic Law (1869) and even after this there were various attempts at rearrangement. Another parallel may be drawn with astronomy before the development of the Hertzsprung-Russell diagram that relates the evolution and the properties of stars. Again in our own subject biology, the situation is somewhat analagous to that before the formulation of the Linnean system of classification; but now from this system of classification, we are able to organize our knowledge of, for example, the functional morphology of organisms and we can even make assumptions, with a high probability

Habitat, the template for ecological strategies?

In many insects dispersal by flight occurs prior to reproduction, and the ability to fly is often lost once reproduction begins (37, 92). Thus, dispersal is viewed as an "evolved adaptation" characteristic of a particular stage in the ontogeny of the insect (91-93, 98) rather than a spontaneous response to current adversity. In certain groups of insects, species exhibit polymorphisms that affect flight ability. Variations in wing length and flight muscle development are the most obvious examples. In some species flight may be primarily an adaptation for dispersal, and the proportion of morphs capable of flight may be a reliable measure of level of dispersal. Adaptations for dispersal in animals are for the most part difficult to identify because locomotory move­ ments potentially serve many important functions (or produce many impor­ tant effects). Flight polymorphisms in insects are thus of particular interest in understanding the adaptive significance of dispersal in natural popula­ tions. This review focuses on two aspects of insect flight polymorphisms: how morphs are determined and how polymorphisms are maintained. I do not attempt to document all (or even most) examples of flight polymorphisms but instead consider possible answers to the following questions: Do differ­ ent morphs represent different genotypes, or do they represent alternative developmental patterns produced by a single genotype in response to differ­ ent sets of environmental inputs? What do studies of flight polymorphisms reveal about the evolutionary significance of dispersal? What are the selec­ tive advantages and disadvantages of the normal (winged, not necessarily most common) and flightless morphs in an insect population?

Dispersal Polymorphisms in Insects

Migration of terrestrial arthropods in relation to habitat

A biology of dragonflies

The evolutionary fitness of an animal depends significantly upon an optimal diet in both quantity and quality. Foraging strategies are therefore rigorously shaped by natural selection and should be considered in terms of the degree to which they maximize the net nutrient gain from feeding, and to which they minimize the risks to survival. Any discussion offoraging behavior is complicated by the forager's perceiving the environment at several hierarchical levels. We endeavor to categorize these, being fully aware that any such framework is bound to be plagued with exceptions and examples of blurred boundaries. Our classification includes three such levels: the habitat, the patch, and the food item. Of these, the food items are generally the easiest to define. They are the prey of predators, the hosts of parasitoids, the leaves for caterpillars, the feeding sites for mosquitoes, the nectaries and anthers for bees, and so on. Such items are almost invariably heterogeneous in their spatial distribution, which makes it appropriate for us to identify an aggregation of food items-a "patch"-as the next level in the hierarchy. The definitions of what should and should not constitute a patch have been various (174, 179). We look upon patches as spatial subunits of the foraging area in which aggregations of food items occur. A patch is most readily identified when the food items are distributed among discrete natural units, as are, for exam­ ple, prey on leaves. But we must beware of identifying a patch solely by what we perceive or consider reasonable. The forager itself defines the patch, and we should look to changes in the forager's behavior to identify patch boundaries. A patch in these terms is therefore an area containing a stimulus or stimuli that at the proper intensity elicit a characteristic foraging activity in a r esponsive forag er (174). Such a definition, rather than one based on the forager's response itself, avoids attributing more patches to the environment of a responsive forager than to that of an unrespon­ sive one.

J. S. Kennedy

Papers

A Turning Point in the Study of Insect Migration