Isozymes and the Analysis of Genetic Structure in Plant Populations

Abstract

87 Ecologists and plant evolutionary biologists have long recognized that plants are not distributed at random within communities but, ra~her, are clustered in distinct patches. Environmental heterogeneity is usually cited as playing a critical role but colonization patterns and stochastic events affecting establishment and mortality are also important. More recently plant evolutionary biologists have demonstrated that genetic variation in plant populations is also distributed nonrandomly (Antonovics, 1971; Allard et a1. 1972; Hamrick and Allard, 1972; Turkington and Harper, 1979). Rather, like the plants themselves, genes and genotypes tend to be clumped, with marked genetic differences occurring over short distances. This nonrandom distribution of genetic variation is often referred to as the genetic structure of a population (Loveless and Hamrick, 1984). Plant biologists were somewhat slow to Il.ecognize population genetic structure because spatial genetic variation in characteristics sucfu as height, shape, and size is confounded by environmental influences on the phenotype. THus, for many traits, particularly those that are most likely to be adaptive, it has been necessary to grow plants from different habitats in "common gardens" or to make reciprocal transplants into different habitats. It is not surprising that such studies involve populations that occur in strikingly different habitats. The classic work of Bradshaw and his colleagues (Antonovics et aI., 1971; Bradshaw, 1971) on the heavy metal contaminated soils of old mine spills and the elegant work ·of Snaydon and Davies (1972, 1976) on the Park-Grass Experiment illustrate researcH of this type. Studies of more subtle habitat differences , although less abundant in the literature, have also demonstrated significant genetic differences at a local spatial scale (Antonovics, 1!971; Warwick and Briggs, 1978; Turkington and Harper, 1979). Thus, whereas our understanding of the ability of plants to adapt to local environmental conditions has been greatly advanced, for logistic reasons such studies are usually limited to small, herbaceous annuals or perennials. Furthermore, quantitatively inherited traits do not lend themselves to studies of the evolutionary processes that influence the development of genetic structure. For studies of evolutionarily important factors such as gene flow and the breeding system, plant evolutionary biologists prefer to use traits controlled by single Mendelian loci. For many years single-gene morphological traits were used to obtain quantitative estimates of mating systems, gene flow, and occasionally selection. These traits, although providing meaningful data, have at least two practical drawbacks: (1) there are usually very few loci available for any plant species; (2) their expression is often dominant-recessive. As a result, progeny testing is necessary for accurate estimates of genotype frequencies. Biochemical techniques, most notably starch gel electrophoresis, provided plant biologists with additional single-gene markers with which to study evolutionary processes. Isozyme loci have several advantages over single-gene morphological traits: (1) genetic inheritance of electrophoretic ally detectable traits can be easily demonstrated; most loci have discrete Mendelian inheritance; (2) most are codominant and allele frequencies can be calculated directly; (3) estimates of levels and distributio of genetic variation can be compared directly CHAPTER 4 87 Ecologists and plant evolutionary biologists have long recognized that plants are not distributed at random within communities but, rather, are clustered in distinct patches. Environmental heterogeneity is usually cited as playing a critical role but colonization patterns and stochastic events affecting establishment and mortality are also important. More recently plant evolutionary biologists have demonstrated that genetic variation in plant populations is also distributed nonrandomly (Antonovics, 1971; Allard et al. 1972; Hamrick and Allard, 1972; Turkington and Harper, 1979). Rather, like the plants themselves, genes and genotypes tend to be clumped, with marked genetic differences occurring over short distances. This nonrandom distribution of genetic variation is often referred to as the genetic structure of a population (Loveless and Hamrick, 1984). Plant biologists were somewhat slow to r.ecognize population genetic structure because spatial genetic variation in characteristics such as height, shape, and size is confounded by environmental influences on the phenotype. Thus, for many traits, particularly those that are most likely to be adaptive, it has been necessary to grow plants from different habitats in "common gardens" or to make reciprocal transplants into different habitats. It is not surprising that such studies involve populations that occur in strikingly different habitats. The classic work of Bradshaw and his colleagues (Antonovics et al., 1971; Bradshaw, 1971) on the heavy metal contaminated soils of old mine spills and the elegant work ·of Snaydon and Davies (1972, 1976) on the Park-Grass Experiment illustrate research of this type. Studies of more subtle habitat differences , although less abundant in the literature, have also demonstrated significant genetic differences at a local spatial scale (Antonovics, 1971; Warwick and Briggs, 1978; Turkington and Harper, 1979). Thus, whereas our understanding of the ability of plants to adapt to local environmental conditions has been greatly advanced, for logistic reasons such studies are usually limited to small, herbaceous annuals or perennials. Furthermore, quantitatively inherited traits do not lend themselves to studies of the evolutionary processes that influence the development of genetic structure. For studies of evolutionarily important factors such as gene flow and the breeding system, plant evolutionary biologists prefer to use traits controlled by single Mendelian loci. For many years single-gene morphological traits were used to obtain quantitative estimates of mating systems, gene flow, and occasionally selection. These traits, although providing meaningful data, have at least two practical drawbacks: (1) there are usually very few loci available for any plant species; (2) their expression is often dominant-recessive. As a result, progeny testing is necessary for accurate estimates of genotype frequencies. Biochemical techniques, most notably starch gel electrophoresis, provided plant biologists with additional single-gene markers with which to study evolutionary processes. Isozyme loci have several advantages over single-gene morphological traits: (1) genetic inheritance of electrophoretic ally detectable traits can be easily demonstrated; most loci have discrete Mendelian inheritance; (2) most are codominant and allele frequencies can be calculated directly; (3) estimates of levels and distribution of genetic variation can be compared directly D. E. Soltis et al. (eds.), Isozymes in Plant Biology