A Factor Analysis Study of the Functional Significance of Angiosperm Pollen
1978; American Society of Plant Taxonomists; Volume: 3; Issue: 1 Linguagem: Inglês
10.2307/2418528
ISSN1548-2324
Autores Tópico(s)Tree Root and Stability Studies
ResumoInvestigations on the adaptive significance of pollen morphology were conducted on 17 variables of pollen, floral and anthecological characters in 100 species of angiosperms. Three close interrelationships among the characters were established by factor analysis: (1) positive correlation between grain size, exine thickness, and aperture number; (2) strong association of grain shapes with aperture types, i.e. the prolate grain with colpus and the spherical and the oblate shapes with pore; and (3) a positive correlation among the pollinator's body size, the floral size, the style length, and the pollen grain size. Based on the above results, the following suggestions on the adaptive aspects of pollen can be made: (1) exine thickness is important for mechanical support with respect to grain size and apertur-e number; (2) larger grains require more apertures for rapid water absorption from the stigmatic surface; (3) the association of grain shape and aperture type plays a role in accommodation of volume change when the grains absorb or lose water; and (4) grain size is related to the capacity of stored nutrients which determine the potential length of a pollen tube. Diversity of the angiosperm pollen grain has been recognized since the last century (Wodehouse, 1935; Davis & Heywood, 1963; Faegri & Iversen, 1964; Erdtman, 1966), but relatively little attention has been paid to the functional and adaptive aspects of the highly durable exine. Kerner appreciated the functional role of the grooves (previously folded inwards) when the grain is inflated by water uptake prior to germination (in Payne, 1972) and Wodehouse (1936) speculated on the evolution of harmomegathy in the accommodation of volumetric changes. More recently numerous anecdotal observations on various functional aspects of pollen have been noted (e.g. Larson & Lewis, 1961; Henrickson, 1967; Muller, 1969, 1973; Dulberger, 1975; Taylor & Levin, 1975; Kress, Stone & Sellers, 1978) and a few workers (e.g. Payne, 1972; Chaloner, 1976) have met with limited success in showing correlative features that link pollen structure to function. Perhaps the most intriguing new findings are those of J. Heslop-Harrison and colleagues (Heslop-Harrison, 1976). The chambered exine that is characteristic of so many angiosperms serves as the repository for tapetum-derived incompatibility proteins. The functional corollary to these findings implies that pollen lacking chambered exine (e.g. Cana, Helicornia, Musa, parasitic Gentianaceae, etc.) would not be capable of sporophytic incompatibility. Confirmation of corollary has yet to be established. In spite of these provocative new findings, the adaptive significance of most pollen features is not known (Cronquist, 1968; Ehrendorfer, 1973; Lewis, 1977). As Stebbins (1974, I Botany, Duke University, Durham, NC 27706. Pr-esent address: Biology, jeonbuk National University, Jeonju, Korea. This content downloaded from 157.55.39.116 on Sun, 18 Sep 2016 06:02:05 UTC All use subject to http://about.jstor.org/terms 2 SYSTEMATIC BOTANY [Volume 3 p. 292) has pointed out, this lack of understanding is due largely to the complexity of the problem and the paucity of information on a comparative basis. The complex variables of pollen diversity and pollination systems might usefully be analyzed by multivariate statistics. Factor analysis is one such statistical technique for reducing a large number of variables to a small number of hypothetical factors. These factors are often biologically interpretable because they summarize the predominant trends of correlated character variation and thus identify subgroups of functionally related characters (Seal, 1968). Employment of factor analysis seems appropriate for study in which I have attempted to isolate subgroups of inter-related pollen and anthecological characters and to discover possible causal factors responsible for these correlations (Sokal & Rohlf, 1969). MATERIALS AND METHODS Pollen and anthecological characters from a total of 100 species from various angiosperm families (Table 1) were observed and scored. Sample size for each species was generally 3-10 pollen grains or flowers. Selection of the species was based on the availability of limited data for a given species from pollen slides, herbarium specimens, and information about the pollination mechanism. In some cases where no slides of pollen or flowers were available, I was able to substitute data from Stuchlik's (1967) pollen descriptions and the floral descriptions by Grant and Grant (1965). The observed correlations among the various characters associated with the floral biology have led to some very positive statements. I wish to stress, however, that sample sizes are small and fewer than 0.04% of the angiosperms have been examined. Pollen from the collection of Dr. D. A. Livingstone in the Department of Zoology, Duke University, was examined with a brightfield microscope; flowers were measured under a stereo-microscope from the specimens housed in the Duke University Herbarium. The raw data of the taxa examined are presented in Table 1. Characters Examined.-The total of 17 characters examined consists of 8 pollen and 9 anthecological features. Characters exhibiting continuous variation were measured in conventional metric units. Discrete states for characters 4, 6, 9, 10, 11, 12, 13 were numerically coded on criteria ordinal scales representing each trend of morphological variation. The characters and their states (noted in parentheses) or units, in cases of measured characters, are as follows. Pollen characters 1) Grain length, the length of polar axis in gm. 2) Grain width, the equatorial diameter in ,um. 3) Equatorial shape (width/length ratio). 4) Aperture number (0-100). This content downloaded from 157.55.39.116 on Sun, 18 Sep 2016 06:02:05 UTC All use subject to http://about.jstor.org/terms 1978] LEE: FUNCTIONAL SIGNIFICANCE OF ANGIOSPERM POLLEN 3 5) Exine thickness in ,gm. 6) Sculpture pattern [psilate, scabrate, ver-rucate, gemmate, clavate, psilate, baculate, echinate (1); rugulate, striate (3); reticulate, areolate, insulate (5)]. 7) Pore width, diameter perpendicular to the polar axis in ,um. 8) Colpus length in gum. Anthecological characters The anthecological characters and the reasons for including each character are described below. 9) Vector size [abiotic (1); insects (3); birds, bats (5)]. The relationships between plants and their pollinators are maintained by the balance of a mutual need for energy, i.e. pollinators maximizing food intake and plants minimizing energy devoted to providing food for pollinators while maximizing seed set (Stiles, 1971; Lyon & Chadek, 1971; Heinrich & Raven, 1972). The value of pollen as a food source may be expressed not only by the quantity of pollen, which is a function of size, and the amount of pollen per anther, flower, or plant, but by the pollinator's specific nutritional requirements (e.g. bat-flower pollen which contains higher concentrations of certain amino acids, Howell, 1974). Since study did not assess related size features such as nutrition, data for pollen size per se may be expected to show only weak correlations with data for pollinator size. However, the inclusion of character is worthwhile in order to see how well it correlates with flowersize characters (Taylor & Levin, 1975). 10) Floral type [primitive, simple (1); actinomorphic (2); reduced, protected (3); zygomorphic (4)]. Leppik (1957, 1958, 1968) classified floral shape into the above six types and discussed the evolutionary trends and significance in relation to pollinators. The floral types as well as other morphological aspects of the flower would have resulted from the adaptation to the pollinators and seed dispersal vectors (Grant, 1963). It is not known, however, what relationships are involved between the evolutionary trends in the various floral characters and between floral and vector characters. Although some relationships of the floral types to the sensory capability and food searching activity of pollinating insects have been revealed by Leppik's efforts, it may be difficult to relate the floral characters to the pollen characters. It is interesting, however, to look for relationships between pollen and flower morphology and to investigate the functional significance of such relationships. The primitive and simple and reduced and protected types were lumped in the same categories due to the difficulty of separating them. 11) Temperature parameter [arctic (1); temperate (3); tropical (5)]. Temperature is an important environmental variable (Kurtz & Liverman, 1958). Most organisms flourish within a limited temperature range and the tolerance for flowering is even more restricted. This content downloaded from 157.55.39.116 on Sun, 18 Sep 2016 06:02:05 UTC All use subject to http://about.jstor.org/terms 4 ~~~~~~~~~~~~~~~~~SYSTEMATIC BOTANY [Volume 3 (Ll) AS1X0 O O o O O O0O O n O) 00 O1 Oco 0 0 G11 x O GO OO (91) -1-d O vb n O O0 dHn tG = OC) r0' 0' t0 tu00 GQ O C) O) C t 00 t MG4roG sH g, .ANIVMD N~~c n c c c 't n cq cq cq LC N4 N4 n 4tc\1 n c c Q0 c n n C I cVM n cq cq N n t-N N qn n n c Lc c 0 c GN c ; a=;t W ;= 2
Referência(s)