shows the optimum range of day and night temperatures for flower and seed production of a number of ornamentals. Lowry (1969) provides a compendium of what is known about specific physiological requirements of plants in a microclimatic context. For example, Lowry discusses in detail the "heat units" concept, which relates the development of a plant to the thermal environment to which it is exposed. Briefly, the rate of development of a plant increases linearly with temperature above a certain threshold. It would be expected, therefore, that any increase in the temperature climate to which a plant is exposed would result in its more rapid development, other factors being equal. A relevant question in this context is what contribution a secular increase in temperature due to an augmented greenhouse effect would make to the development of a particular plant or plant association.
Plants are exposed to variations in climate and microclimate that span time scales ranging from a few seconds to years or decades. They survive this variation in a number of ways, ranging from inherent flexibility and adaptation to dispersal to more favorable microclimates. It is not my purpose to create a dictionary of strategies that plants adopt. Rather, it is to look at the range of microclimates in which plants, especially, exist. In a rough way, we can compare the temperature variability of their microclimates with the increase in temperature we presume will result from an enhanced greenhouse effect. This predicted increase is of the order of 0.02°C to 0.03°C per year in the global average, and perhaps 50 percent to 100 percent greater than the global average in the high northern latitudes (IPCC-I). A location might thus experience a secular increase of 0.02°C to 0.06°C per year.
Such increases would clearly have some impact on specific plants; indeed, some would not be able to survive. Maximum microclimatic temperatures might increase to the point where a particular plant could not reproduce or would suffer fatal heat injury. An assembly of plants with similar microclimatic requirements would either die out or, alternatively, invade an area with a more suitable microclimate. It would thus be of interest to see what plants are actually exposed to in terms of climatic variability, against a background of annual increases that might be the result of global warming.
A wealth of climatic and microclimatic data is available in various compendia, notably those of Geiger (1965), Yoshino (1975), and Landsberg (1958), and in many standard texts on climatology. Data from these and other sources are plotted in Figure 4. (Again, this is a log-log plot to encompass the many orders of magnitude involved.) Data were selected not only to represent "maximum" microclimatic fluctuations, but also to be representative of various climates and microclimates. In general, these are air temperatures measured at various heights from near-surface to screen height. Unless otherwise indicated on the figure, measurements are presumed to be at heights of 1.25 to 2.0 meters above ground level, the standard specified by the World Meteorological Organization (WMO, 1983).8
The annual range of mean daily temperatures varies from about 1°C in the tropics to about 67° in the extreme continentality of Siberia. Most mid-latitude values range from about 20° to 30°. Diurnal temperature ranges are somewhat smaller. The largest I could find for an agricultural surface was about 30°, at ground level beneath an orchard. (I have ignored diurnal fluctuations on bare ground surfaces as being less relevant to plants. The greatest ground-surface diurnal range I could find was about 50°, in the Sahara (Geiger, 1965).)
Information on shorter-period fluctuations is scarce in the scientific literature; a few representative values are given in Figure 4. It is apparent that many plants and plant surfaces are subjected to short-period fluctuations on the order of 1°C to 30°C.
Does this variability have any significance for how well plants and ecosystems would survive a secular warming on the order of 0.06°C per year? My own experience leads me to believe that such warming would, in general, lead to slow ecological changes of the same order of magnitude as those that have always been a part of ecosystem behavior. In terms of human perceptibility, these changes are usually scarcely noticeable. Thus, the forest-prairie boundary may move back and forth in response to climatic change, but the rate may be measured in centimeters or meters per year. Looking back in geologic time tends to produce a foreshortened view. The thousand-year-long warming at the end of the last ice age may appear to us to have had ecological consequences more drastic than any observer on the scene at the time would have noticed.
As I noted previously, a plant that is experiencing a secular change in the microclimate, slow though it may be, may find that it can no longer exist or reproduce and reestablish itself in that changed microclimate. It may be subject to ravages of disease or pests that did not exist previously. However, its reproductive processes may permit the plant to become established in a nearby microclimate that is more favorable to its survival. Indeed, this is one of the ways that species distribution changes in response to