Oecologia (Berlin) (1987) 71:601 607 Oecologia 9 Springer-Verlag 1987 The measurement of small-scale environmental heterogeneity using clonal transplants of Anthoxanthum odoratum and Danthonia spicata J. Antonovies, K. Clay*, and J. Schmitt** Botany Department, Duke University, Durham, NC 27706, USA Summary. To assess the scale of micro-environmental heter- ogeneity perceived by two co-occurring grass species, An- thoxanthum odoratum and Danthonia spicata, cloned tillers of each species were planted into the natural habitat at a range of spacings (from 2 cm to more than 2 m apart) and measured for survival and fecundity over three years. A. odoratum responded to heterogeneity at a scale of 4-8 cm and at a scale of 2-8 m but not to intermediate scales. D. spicata did not respond significantly to heterogeneity. How- ever one genotype infected with the systemic fungus Atkin- sonella hypoxylon showed a large response to heterogeneity at both small and large spatial scales. The results showed that the scale and level of environmental heterogeneity as measured by its fitness impact depends on the species and genotype involved. The results indicate that small scale en- vironmental heterogeneity could play a role in the mainte- nance of sexual reproduction in A. odoratum. Key words: Microclimate - Fecundity - Anthoxanthum - Danthonia- Atkinsonella Numerous authors have pointed out that the scale of envi- ronmental heterogeneity is a crucial factor in many evolu- tionary and ecological processes (e.g. Bradshaw 1965; Levin 1968; Antonovics 1971; Hedrick et al. 1976; Stearns 1976). In particular, environmental heterogeneity has been in- voked in several theoretical discussions of the evolutionary maintenance of sexual reproduction (Maynard Smith 1978; Williams 1975; Bulmer 1980). Clearly, any rigorous test of these theoretical predictions requires knowledge of the scale and magnitude of such heterogeneity. Such measure- ments are difficult. In many cases, the environmental factor to which the organism is responding is either not evident or is so complex as to be beyond easy measurement. Even if the factor is recognized, multiple measurements are neces- sary to assess the scale of heterogeneity, so it is likely that it is only possible to study variables that do not require complex and expensive instrumentation for their measure- ment. Moreover, different species or even different geno- types of one species may differ in sensitivity to various Offprint requests to: J. Antonovics * Present address: Department of Biology, University of Indiana, Bloomington, IN 47405, USA ** Present address: Program in Ecology and Evolutionary Biolo- gy, Division of Biology and Medicine, Brown University, Provi- dence, RI 02912, USA environmental factors, and consequently experience differ- ent scales or degrees of heterogeneity within the same habi- tat. Over fifty years ago, Clements and Goldsmith (1924) suggested the use of the organism itself to measure the envi- ronment. They coined the term "phytometer" to refer to a plant used as a measuring instrument. They pointed out two major advantages of the phytometer technique: first, the environment is measured as perceived by the individual plant; second, phytometers can be deployed relatively cheaply and on a large scale. Recently, there has been a revival of the phytometer method, particularly because the measurement of the survival and fecundity of the phyto- meter provides a practical approximation to "the fitness" of the organism and the results can therefore be applied in many ecological and evolutionary contexts. This method has been used to demonstrate adaptation to micro-hetero- geneity in pasture populations (Turkington et al. 1979), to demonstrate spatial heterogeneity in demographic proper- ties (Fowler and Antonovics 1981 b) and to study life-cycle components of selection (Antonovics and Primack 1982; McGraw and Antonovics 1984; Roach 1984). Here we re- port on a study using "phytometer" transplants of clones of the grasses Anthoxanthum odoratum and Danthonia spi- cata to determine whether these co-occuring species experi- ence different scales of heterogeneity within the same habi- tat. Genotypes ofAnthoxanthum odoratum show differential response to large scale spatial heterogeneity (Antonovics et al. 1987), but such large scale heterogeneity was insuffi- cient to provide an explanation for the evolutionary mainte- nance of sexual reproduction within this population. How- ever, small scale heterogeneity could permit the evolution and maintenance of sexual reproduction within a popula- tion, if sib-competition or genotype-specific habitat selec- tion existed (Maynard Smith 1978; Williams 1975). The present study therefore sought to assess environmental het- erogeneity on a finer scale relative to the seed dispersal distances of the study species. In particular, we ask: What is the scale of microenvironmental heterogeneity? Do two co-occurring species perceive heterogeneity differently? Is there variation within species in their response to hetero- geneity? Methods The sampling of genotypes and the experimental studies were all carried out in a regularly mown, but otherwise relatively undisturbed field dominated by perennial grasses