Pollen Morphology of the Genus Erythrina (Leguminosae: Papilionoideae) in Relation to Floral Structure and Pollinators

Abstract

The unacetolysed pollen of 99 species of Erythrina was studied with scanning electron microscopy. The exine sculpturing is basically reticulate but displays considerable variation in several features. Pollen types may be divided into those species with tubular, hummingbird-pollinated flowers and those with gaping flowers, pollinated by a variety of passerine birds. Hummingbird type pollen may be defined as medium-sized, with medium-sized lumina and sparse or no sexinous granules. Pollen from passerine-pollinated sections is more heterogeneous but may be distinguished generally by moderate to dense granularity or small lumina size. Some overlap occurs between these groups. Pollen from some passerine-pollinated sections frequently showed an abundance of pollenkitt, which was sparse or lacking in hummingbird groups. These observations are discussed and possible explanations suggested. Palynological contributions to taxonomy, in disclosing and suggesting relationships between taxa, have emerged hitherto principally from comparative study of structure and ornamentation among pollen. Recently, however, attention has become focused increasingly on functional aspects of the structural details observed. The maintenance of pollen in a viable state, its transfer to a stigma, and successful germination on a compatible stigma are essential to normal sexual reproduction in flowering plants. Under constant environmental pressures many features of pollen structure and physiology may lead to failure, or conversely promote success, at any stage. Thus, it seems reasonable to expect that pollen, subject to natural selection just as much as the whole plant, has evolved to meet the demands imposed upon it. However, the situation is unlikely to be simple. A plant is constrained within certain limitations by its basic genetic inheritance. Compromises often must be made between conflicting factors. Nor is it necessarily true that a plant is optimally adapted with respect to all (or any) of its morphological features: for example, an inefficient pollination system may be compensated for by great longevity. Evolutionary processes and environmental conditions are both in a state of continuous chance but not all features are necessarily of any immediately obvious adaptive significance. In a relatively close-knit genus, such as Erythrina, it seems reasonable to assume that the species have radiated relatively recently from an ancestral stock and have preserved a basic similarity in pollen structure and potential for evolutionary change. However, Erythrina is also a large genus, widely distributed in both the Old and New Worlds. Although all species are primarily pollinated by birds of one sort or another (Toledo, 1974; Raven, 1979), there seems to be distinct syndromes in floral biology within the genus, apparently associated with pollination strategy. There is considerable variation among species in floral structure and presentation (Krukoff& Barneby, 1974; Lucas & Theobald, 1982) and in behavior and approach of pollinators. As different modes of survival and efficient transference of the pollen develop, morphological modifications of the pollen itself might be ex1 We are greatly indebted to C. H. Stirton (K) for suggesting the topic and for much initial guidance and to both him and R. M. Polhill (K) for invaluable discussion throughout the investigation. We are most grateful to R. Barneby (NY) for his encouragement and valuable comments on the manuscript and for his help in providing pollen material from NY. P. H. Raven and D. Neill (MO) provided much appreciated comment and discussion as well as material of E. batolobium and we extend our thanks to them both. We thank our colleagues T. Harwood for printing the micrographs and Christine Grey-Wilson for drawing Figure 37. Part of the work was carried out while AJH was in receipt of a Ministry of Agriculture Fisheries and Food vacation studentship at the Royal Botanic Gardens, Kew, and we especially thank the Bentham-Moxon Trust for support to AJH to allow completion of the survey. 2 The Herbarium, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, England. ANN. MISSOURI BOT. GARD. 72: 570-590. 1985. This content downloaded from 207.46.13.115 on Sat, 08 Oct 2016 04:17:05 UTC All use subject to http://about.jstor.org/terms 1985] HEMSLEY & FERGUSON-ERYTHRINA POLLEN MORPHOLOGY 571 pected. However, climatic, ecological, or other factors may also influence evolution, and random variation, whether genetically or environmentally induced, may also occur. Ferguson and Skvarla (1982) have shown that in the subfamily Papilionoideae there is a close association between pollen ornamentation and stratification with flower structure and pollination system, notably in some species with flowers adapted for pollination by birds or bats. This association between pollen morphology and floral structure occurs in species and genera from widely separated tribes, and it is suggested that it is the result of convergent evolution and a secondary adaptation to pollination. The pollen morphology of Erythrina was not included in their investigation. Graham and Tomb (1974, 1977) made a detailed study of the pollen of Erythrina in relation to taxonomy but without consideration of the approach discussed above. These workers gave an extensive bibliography of the pollen morphology of the genus, which can be supplemented by reference to Thanikaimoni (1976, 1980). The aim of the present study, using herbarium material, is to examine the links between pollen morphology, flower structure, and pollination ecology. MATERIALS AND METHODS Pollen was obtained primarily from herbarium specimens at the Royal Botanic Gardens, Kew (K). Material was also obtained from the herbarium of the New York Botanical Garden (NY) and the Missouri Botanical Garden (MO). Details of the specimens examined are shown in Table 1. The data presented in Table 1 are based on scanning electron microscope (SEM) micrographs of untreated pollen removed directly from herbarium sheets. This procedure was followed because, although acetolysis (Erdtman, 1969) in most cases somewhat improved the clarity of the ornamentation and increased the size of the pollen grains, it removed the residual pollenkitt found on many unacetolysed pollen. Differences in the occurrence and volume of pollenkitt present in pollen samples were found to be of some significance. This will be discussed later. The pollen was prepared by removing mature anthers from herbarium sheets and shaking or brushing the pollen onto double-sided sticky tape on SEM specimen stubs. The samples were sputter coated with platinum and examined with a Jeol T20 SEM. POLLEN MORPHOLOGY From the 99 species of Erythrina examined in this study, and from the results of Graham and Tomb (1974, 1977), the pollen may be described as follows: oblate; rounded-triangular in polar view; ranging in size from 22 to 55 Am diam.; triporate, occasionally tetraporate, pores more or less equally spaced, equatorially arranged at the corners; surface ornamentation basically reticulate, but with considerable variation in the size and shape of the lumina, the height and width of the muri, and the occurrence and prominence of sexinous granules in the lumina. In occasional samples the muri are discontinuous on either the polar or equatorial faces (e.g., E. dominguezii, E. lysistemon). Rarely is there little or no organized reticulum, and the ornamentation consists of discontinuous muri or sexinous granules. Each of the main features of variation in the pollen morphology is discussed briefly in turn and the data for each species are summarized in