| Constancea 83, 2002 University and Jepson Herbaria P.C. Silva Festschrift |
|---|
On November 20, 1962 when Paul Silva was editor of the then new journal, Phycologia, the junior author and R. L. Hilton, both from the University of Connecticut, submitted a paper entitled “Pleomorphism in Scenedesmus longus”. By return mail, no e-mail at that time, Paul wrote: “The manuscript has been sent to reviewers, and although I could be mistaken, from a brief glance I should judge that their general reaction will be favorable.” We were certainly encouraged by this support.In just a few weeks we learned that three reviewers took a rather different view, for on January 8, 1963 Paul wrote a second time: “I stand overruled.” Of course the next step was to revise the manuscript and resubmit, but the paper was eventually withdrawn.
We provide this bit of history to point out that when later published in the Bulletin of the Torrey Botanical Club (1963) as “Culture of Scenedesmus longus”, this article became one of the first of many papers from our laboratory dealing with the complex phenotypic expression in a genus, which has recently been divided into Scenedesmus and Desmodesmus (An et al. 1999). Yes, we were appreciative of Paul's initial support, as well as the professional and sympathetic treatment we received from the editorial office. Again, thank you! In retrospect, this was valuable training for later drafting publications dealing with a controversial subject, phenotypic expression.
Ironically, the study of phenotypic plasticity within phycology seems to predate contributions dealing with any other group of organisms. For example, the green algal genus Scenedesmus has been recognized as plastic since the mid 1800's. But the contributions of Grunow (1859), Wolle (1887), Chodat and Malinesko (1893), and Grintzesco (1902) were enveloped in controversy, criticism, and doubt (Smith 1916; see Schlichting and Bruton 1970, Trainor et al. 1971, and Shubert 1975, for further discussion on this topic). In addition Beijerinck (1890), Senn (1899), and Livingston (1900) reported similar phenomena in green algae related to Scenedesmus. Even though phycological research in this area has continued for decades to provide additional support, literature dealing with an evolutionary perspective has been sparse. Only recently, examples including green algae and cyanobacteria have been taken into account (Bell 1991, 1992, Graham and Wilcox 2000, Mazel and Marliere 1989).
In the last few years, experimental evidence for plasticity has increased dramatically. This information has been mainly produced by ecologists and evolutionary biologists interested in unraveling the processes by which organisms are able to cope with sudden changes in the habitats in which they thrive. The majority of this literature concentrates on the study of reaction norms, the set of potential phenotypes that a particular genotype can express in different habitat conditions. Note that interest in reaction norms has had a long tradition within evolutionary biology since Woltereck (1909) coined the term based on his experiments on daphnids. Woltereck's work, coupled with growing interest in the concepts of both genotype and phenotype (Weismann 1885, Johannsen 1911), led to the understanding of the phenotype as the ultimate consequence of gene expression along with the dependence of gene activity on environmental factors. Later, these concepts were fine tuned theoretically (Baldwin 1896, Goldschmidt 1940, Schmalhausen 1949) and confirmed experimentally (Waddington 1942, Clausen et al. 1948).
Before there was consideration of the implications of plasticity, in an ecological and evolutionary context, the neodarwinian evolutionary synthesis (Huxley 1942, Mayr and Provine 1980) attempted to explain the adaptation of organisms over long periods of time and at small spatial scales through the production of ecotypes (Turesson 1942). Ecotypes are genetically distinct sectors of a population adapted to local conditions with the potential of becoming a separate species provided some type of barrier impeding exchange of genes with the original population is present. Thus, the ecotypic adaptation theory disregards both the effect of environmental fluctuations on populations over short periods of time and the fact that plasticity can also be a mechanism for speciation (Pigliucci 1996, Schlichting and Pigliucci 1998). Neodarwinists lacked satisfactory explanations for the remarkable ability of certain organisms to disperse very rapidly (as happens with the majority of fresh-water algae), for why some species are spread over a wide geographical area and yet remain a cohesive group of interbreeding organisms, and most importantly, for the mechanisms determining the ability of some individuals to withstand marked shifts in habitat conditions (e.g., those occurring in coastal systems and temperate lakes).
Recently, novel biochemical techniques have elucidated the genetic and molecular mechanisms behind plastic responses (Pigliucci 2001, Schlichting and Pigliucci 1993, 1998, and references therein). It is now known, for example, that plasticity is simply a matter of a modification in the ability of a genotype to express a particular character. Such a modification may occur before or after a character has been expressed and it is always the response to a change in either intra or extracellular environments, or both. Exploration of genetic processes such as epistasis (whereby the expression of one character is linked to the expression of others) and pleiotropy (when a particular gene controls the functionality of others) has further contributed to the elucidation of plastic mechanisms at the genetic and molecular levels (Pigliucci 2001, Pigliucci et al. 1996, Schlichting 1989).
Experimental evidence concerning phenotypic plasticity now seems so striking and unambiguous that widespread phenotypic stability can not be taken for granted. Yet some phycologists (e.g., Håkansson and Chepurnov 1999, Hegewald 1997, Round 1996) have reservations about the role of the environment in shaping the organism's phenotype --perhaps for historical reasons (Morales and Trainor 1997, 1999, Trainor and Morales 1999). As we shall elaborate later, two procedural factors seem to have been responsible for the failure to appreciate the role of phenotypic plasticity: the exclusive use of field collected and preserved material, as well as the use of standard conditions for the growth of organisms in the laboratory.
In an evolutionary sense, it has been repeatedly demonstrated that natural selection favors the evolution of plastic genotypes, since the unpredictability of the environment tends to suppress the production of an optimal phenotype (Wright 1931, 1932, Scheiner 1998). Thus, well-adapted organisms must be able to track environmental change and produce highly adapted morphs under a variety of environmental conditions. Indeed, plasticity can be adaptive as demonstrated in the predator-induced morphological changes in a plethora of organisms of diverse taxonomical affiliation, including the algae (Havel 1987, Lürling 1999). More recently, adaptive plasticity has also been discovered to be the response to physical environmental pressures (Gomez et al. 1995, Zirbel et al. 2000).
In the present review we attempt to place available information on algal plasticity in an evolutionary context and discuss the implications for other phycological fields. The merging of phenotypic plasticity data with current phycological thinking completely changes our view of algae as biological entities. It leads to an integrated species concept in which morphology, physiology, ecology, and evolution are intertwined. In addition, it helps us to understand the relationship between organisms and their immediate environment and how individuals can endure in constantly changing and even newly invaded habitats.
Environmental effects upon the genetic machinery are brought about by various intracellular and extracellular factors operating throughout the life history of organisms. These factors elicit genetic responses that ultimately manifest themselves through a modification (physiological and/or morphological) of the phenotype. In other cases, as takes place with the green alga Scenedesmus, morphological and physiological changes are observed from parents to offspring (Morales and Trainor, 1999) (Fig. 1).
Genetic responses to environmental changes are coordinated through an epigenetic system capable of regulating the functioning of its parts (genes) (see reviews in Pigliucci, 2001, Pigliucci et al. 1996, and Schlichting and Pigliucci, 1998). The effect of such functioning is the “calibration” of an organism to the new surrounding conditions. Hence, many phenotypic characteristics are the end result of gene activity, with the amount and direction modulated by environmental variability.
The elucidation of the mechanisms underlying plastic responses is fairly recent, although some of the elements have been embedded in biological knowledge for some years. DNA research has revealed that not all the genetic information of an individual is used at one time, if it is used at all. In humans, for example, about 90% of the DNA has an uncertain function and remains inactive during the life time of an individual. Given the diversity of human cells, we can infer that for each cell type only a small portion of the total DNA will be functional. Genes that are not necessary for the expression of characters needed at a particular developmental stage, environmental condition, or to accomplish a particular function in the organism are “turned off”, but remain a part of the genotype. In more general terms, the environment triggers the expression of those regions in the DNA that provide the most appropriate characters required for survival at any particular time.
The process of turning genes on and off is known as gene regulation, and the resulting phenotypic variation is often marked and drastic (Levins 1968, Pigliucci et al. 1996, Schlichting and Pigliucci 1995, 1998, Stearns 1989). Gene regulation has been clearly illustrated in an environmental context in Calothrix (Mazel and Marliere 1989). When this marine cyanophyte is grown in media with low sulfur concentrations, it is still able to produce phycobilins by expressing a gene that codes for a phycobilin molecule composed of amino acids without sulfur.
A second type of genetic control of plasticity is allelic sensitivity, a process that produces gradual changes in the phenotype of an organism through a change in the amount and/or activity of a gene product (Schlichting and Pigliucci 1998). This mechanism is commonly known as acclimation, and it will be referred to as such in this text. For example, the process of chromatic adaptation, the capability of pigmented organisms to maximize photosynthesis in a fluctuating light environment, has been demonstrated to be under this type of genetic control. This is the case of the cyanophyte Fremyella diplosiphon which is able to maximize photosynthesis by production of different amounts of phycoerythrins and phycocyanins following changes in light quality (Kehoe and Grossman 1996). Recently, the molecular mechanism for chromatic adaptation has been connected with light sensing molecules closely related to phytochromes of higher plants (Yeh et al. 1997). Sensing molecules track changes in environmental light quality and then send the information (via a process still under study) to genes encoding the production of phycobiliproteins (Yeh et al. 1997). A similar mechanism has been found in Synechocystis sp. (Wilde et al. 1997).
On the other hand and as discussed previously, phenotypic changes due to environmental effects do not imply alterations in genetic structure and composition (mutations), but rather signal that different portions of the genetic code are expressed. The effects of this turning on and off of genes may, in some ways, be comparable to the appearance of novel traits by mutation (Goldschmidt, 1940), and may also produce adaptive phenotypes (Schaal and Leverich 1987, Via 1993, Via et al. 1995, and others).
Since in some cases the expression of characters is the mere product of gene action (i.e., without participation of the environment exterior to the cell), changes triggered by environmental cues have been traditionally referred to as “non-genetic”. This has contributed to the erroneous notion (sometimes termed Lamarckian) that the phenotype per se is able to respond to shifts in the external environment or medium (for a culture). It should be clear, however, that environmental factors operate over the genetic structure of an organism, a new physiological state is produced, and then a variant phenotype arises (Fig. 1). That is, there is always an underlying genetic action behind any phenotypic change.
Considering the rare occurrence of mutations and the small probability of their fixation by natural selection, phenotypic plasticity may be an important mechanism for adaptation and survival until genetic novelties appear. Asexual populations are considered to have less genetic diversity to counteract environmental changes. However, if plasticity is a property of their genetic complement, other alternative phenotypes may be “used” to cope with these changes (Levin 1988). Sexually reproducing organisms are also plastic, and genetic diversity within a species results in genetic variation for plasticity in the population (Scheiner 1998, Schlichting 1986, Via and Lande 1985). Therefore, each new recombinant may differ in its plastic response, providing the population with more opportunities (than in asexual populations) to withstand environmental demands.
For example, Sorokin and Krauss (1962), studied the effects of light intensity and temperature on cellular growth of Chlorella pyrenoidosa in synchronized cultures. Data were collected before cell division occurred, so the recorded changes in cellular material correspond to variations within a single generation of individuals. Temperatures above 15°C produced a wide range of doublings of cell material per day. At 40°C for example, the number of divisions varied from 3, at less than 125 µmol m2s-1 illumination, up to 13 doublings at approximately 750 µmol m2s-1. In contrast, temperatures below 15°C limited cellular growth due to increased decomposition of chlorophyll molecules.
Another example has been provided by Ben-Amotz (1975) from experimentation with Dunaliella parva, an alga that is able to persist in highly saline waters. Again in this case, variations in intracellular osmotic water content were recorded before the organism was able to divide. Even just 5 min after the exposure of cells to increased salinity, a dramatic decrease in the osmotic water content was observed. Under these “non-metabolic” conditions (photosynthesis is negligible in such short time periods), glycerol content in the cell remained constant enabling the cell to maintain an osmotic equilibrium with the external medium. In the long term and when photosynthesis was operating, osmoregulation depended on the synthesis (through photosynthesis or metabolic degradation of starch) or degradation of glycerol. Cells incubated for periods of about 90 min at 25°C showed an increased glycerol content and a constant content of osmotic water throughout the period.
Even though there are numerous other examples among the microalgae, we believe that several are best discussed as typical of adaptation to “coarse grained” variation in the external environment (see later). For investigational studies of acclimation in the algae, macroalgae, especially seaweeds, have been more useful experimental organisms than microalgae (Dawes et al. 1976, Fain and Murray 1982, Mathieson and Burns 1971, Ogata 1966, among many others).
Heterophylly in aquatic emergent angiosperms constitutes an example of how individuals exhibiting continuous growth (also called modular organisms) are able to change their appearance during their lifespan. It has been proposed that heterophylly may be the outcome of at least two mechanisms (Winn 1996). First, the production of different phenotypes is the result of a developmental pathway that is already programmed in the genotype. This process, known as heteroblastic development, does not depend on external environmental influences, and has been demonstrated for some aquatic plants (Arber 1919, 1972, Kaplan 1980). The second mechanism, and the one that concerns us in the present paper, is plasticity itself (Young et al. 1995), the response to environmental variation leading to morphological changes in successively produced leaves. Experiments have demonstrated that the switch from an aquatic to aerial location, or from aquatic to terrestrial habitats, can lead to impressive changes in leaf physiology and appearance (Chabot and Chabot 1977, Cook 1968, Cook and Johnson 1968, Ku and Hunt 1973, among others). Such differences prove in many cases to be adaptive, improving the fitness of the organism to the new habitat state (see discussion presented by Schlichting and Pigliucci 1998).
Examples of heterophylly-like phenomena among the macroalgae have been studied for a number of years (Norton et al. 1981). Working in Pterocladia caerulescens, Santelices (1978) showed that the thallus of an individual alga exhibited morphological variation related to seasonal variations. Thus, while the summer thallus has an elongated and thin appearance, the winter thallus is broader and more robust. In other cases, shape modification can be more dramatic as it occurs in Fucus gardneri (Blanchette 1997), Hedophyllum sessile (Armstrong 1989), and Griffithsia pacifica (Waaland and Cleland 1972).
Conspicuous morphological changes can also be seen in microorganisms. In the case of the dinoflagellate Ceratocorys horrida (Zirbel et al. 2000), phenotypic alterations are triggered by a physical effect, e.g., turbulence. Cultures containing long-spined morphs, subjected to agitation in the laboratory, yielded populations composed of cells with much shorter spines. In a few of the cultured cells that particular modification in morphology occurred within an hour after agitation was initiated, and by 12 days ca. 70% of the individuals had developed short spines. Since growth rates were negligible during the morphological modifications and these occurred prior to the 16-day doubling time, morphological changes occurred within individuals. Zirbel et al. (2000) also demonstrated the ability to reverse this change of spine length, for cultures of short-spined cells, returned to static conditions, exhibited an almost complete phenotypic reversal in a period of 30 days. Although a small percentage of cells in all cultures remained spineless, further studies demonstrated their capability of quickly producing either the short or long-spined phenotypes.
A critical body of evidence concerning plasticity has been gathered for a variety of cyanobacteria, and through this evidence, taxonomic implications of plasticity have become very clear. The historical development of the classification of this group depicts very well a classic confrontation between “splitter” and “lumper” approaches. Early publications (Geitler 1932) presented extensive lists of organisms that were later complemented with additional descriptions of cyanophytes from other parts of the world (Cocke 1967, Desikachary 1959). The number of “species” was staggering, but it was not clearly stated whether the names were designed to represent different genotypes or strikingly different phenotypes. Subsequent collection of laboratory data on the plasticity of individual blue green algal genotypes led different authors to reduce drastically the number of taxa to be accepted.
Drouet and Daily (1956) for example, reduced the number of accepted names of coccoid forms from 2800 to just 32! Although most researchers criticize Drouet's conclusion, since it was based on examination of preserved material rather than on experimentation, later work with other cyanobacterial groups under controlled laboratory conditions has demonstrated extensive plasticity. Many of the characters used to separate species were found to be highly dependent on the conditions to which the organisms were subjected (e.g., Stein 1963, Foerster 1964, Robinson and Miller 1970, Sinclair and Whitton 1977). For the genus Spirulina for example, Jeeji-Bai and Sheshadri (1980) concluded that the degree of coiling of the trichomes (used to separate this genus from Arthrospira) varies with light intensity and nutrient concentration in the culture medium. Additionally, other “species” showed a wide spectrum of morphologies depending on both nutrient availability and the age of the material. In some cases, a single clone produced morphologies that could be placed within four different genera! Such was the case of Calothrix fuellerbornii, a species able to produce its own typical morphology, as well as Plectonema-like, Lyngbya-like, and Anabaena-like trichomes (Jeeji-Bai 1977). Recent molecular studies confirm that different cyanobacterial morphologies can be found in a single clade (Baurain et al. 2002).
A classic study on cyanobacterial plasticity was presented by Lazaroff and Vishniac (1961, 1964) and Lazaroff (1966), illustrating plastic responses of a strain of Nostoc muscorum. Morphological changes in this cyanophyte were triggered by differences in light intensity. A culture maintained in dark conditions always produced “aseriate” colonies, probably due to either a lack of essential substances for the formation of filaments, or by accumulation of inhibitory material. Under dim light conditions (0.2-0.9 µmol m2s-1), a transition from aseriate to filamentous forms was observed. In the range 16-90 µmol m2s-1 (still considerably less than full sun) typical filamentous forms were yielded. Furthermore, a culture containing aseriate colonies could be induced to produce filaments by adding an extract of a light-grown culture (Lazaroff and Vishniac 1961).
Spectral quality was also found to induce formation of different morphs in N. muscorum; cultures of aseriate forms maintained in red light produced filaments, but, green light triggered the production of aseriates from filamentous colonies. Allophycocyanin and one or various phycoerythrins were determined to be the photoreceptors for red and green light respectively (Lazaroff 1966). Although no attempts have been made to relate all these results to natural conditions, it would not be difficult to imagine changes in light intensity and spectral quality in bodies of water related to both position in the water column and seasonal changes. Broadly, we may hypothesize that during winter less light would be available to organisms because either day-length is short or the ice layer (in the temperate zone) would not allow deep penetration of light into the water column. Under these conditions one can expect to find large populations of aseriate colonies. During early summer, on the other hand, longer days and more red light are available to the organisms, resulting in production of filaments.
Salinity also plays a role in plastic responses of other cyanophytes. Stam and Holleman (1975) demonstrated that what were previously considered to be different freshwater species of Phormidium, in fact corresponded to morphs of a single genotype. These strains had the capability of changing their appearance back and forth when they were transferred to different media with variations in salinity. However, when such strains were cultured together under the same conditions, similar morphotypes were produced. Similarly, salinity has been demonstrated to trigger morphological changes in other organisms such as diatoms (e.g., Cox 1995, Schultz 1971).
On the other hand, the fate of plastic invading genotypes is far more complex. If the new environment is unstable, but the invading genotypes are sufficiently plastic to cope with these changes, then natural selection will favor them without any alteration in their genetic material. However, if the new habitat is substantially different from the original one, some of the plastic potential of these genotypes may be lost over time through genetic drift and/or modification of the epigenetic system (Schlichting and Pigliucci 1998). If the latter two processes are coupled with reproductive isolation from the main population, they might lead to speciation.
The fact that algal genotypes are able to adapt to their habitat through plasticity has been under-explored in evolutionary and phycological research. It is not difficult to conceive, however, that plasticity may be an important factor imparting fitness under changing environmental circumstances, and thus, confer on organisms at least a temporary resistance to evolutionary change (Bradshaw 1965, Sultan 1987). Many authors propose that it is the reaction norm itself which is the subject of natural selection. Genotypes that are able to produce a wider range of phenotypes and to cope more rapidly with changes in habitat conditions will be more apt to survive (Baldwin 1896, Bradshaw 1965, de Jong 1995, Gause 1947, Pigliucci et al. 1996, Scheiner and Lyman 1989, Scheiner 1993, Schlichting 1986, Stearns 1982, Via and Lande 1985).
Separation of polymorphic and plastic responses in natural populations is rather difficult (despite claims to the contrary [Mann 1999]), for a mixture of both processes often occurs in members of the population. Francke and Ten Cate (1980) working on populations of the chlorophyte Stigeoclonium tenue illustrated this point. Clones isolated from several sites in the Netherlands were subjected to varying concentrations of phosphorus, ammonia, and chloride. The clones exhibited dissimilar growth responses (Fig. 2). Not only did clones from different geographical areas respond differently (plasticity), but there were also differences between clones from the same locality (local polymorphism) . These data suggest that the species S. tenue is both plastic and polymorphic, and that over its range of distribution, some of its genotypes are locally adapted as shown by differences in growth rates between allopatric clones grown at similar nutrient concentrations. Phenotypic plasticity in S. tenue is not restricted to growth; conspicuous morphological changes also occur, and these have had taxonomic implications (Francke and Simons 1984, Campbell and Sarafis 1972).
Accumulation of these costs can affect both the ecological performance of plastic organisms and their evolution. At the ecological level, an exaggerated cost of an adaptive plastic response may affect the competitive ability of an organism, favoring less plastic individuals that allocate more resources directly to reproductive fitness. Thus, even though plasticity may lead to the production of the most adequate phenotype in a given environment, such plasticity may not be selectively advantageous (DeWitt et al. 1998).
Perhaps, even in the face of evidence, some researchers may still insist that the common “rule” in nature is the tendency for a genome to produce similar phenotypes when it is exposed to different environments (e.g., Håkansson and Chepurnov 1999, Round 1997). The maintenance of a phenotype under changing environmental conditions (canalization) is the result of stabilizing selection (Schmalhausen 1949, Waddington 1942). In general terms, canalization arises when a polymorphic population is subjected to a shift in the environment (e.g., seasonal changes in a lake) and a subsequent selection of genotypes producing optimal phenotypes under those conditions. Upon alteration in the sequence of habitat change (e.g., a prolonged drought in the lake), the capability of producing a particular morph will tend to be selected against, restricting a plastic genotype to the point of no plasticity. The observation of the occurrence of canalization in nature may have contributed to the establishment of the dogma on stability of form. However, we see canalization simply as an additional outcome of the interaction between the genotype and the environment. Unexposed portions of the reaction norm will be manifested in newly modified environments, and this reaction norm will be again diversified by sexual reproduction. Plasticity can not be merely “erased” from genotypes simply because genes within a genotype are always interacting with each other (see Schlichting and Pigliucci 1998 and references therein), and because adaptive plasticity is a character that will be maintained by natural selection. A second possibility for the pervasive use of a canalization argument in phycological literature could be related to the nature of sampling utilized in taxonomic studies. The majority of algal floras have been based on single collections for a locality at one or only a few points in time. This has contributed to the observation of only a small window in the life history of taxa (Trainor 1998).
With development of transmission electron microscopy (TEM) and molecular techniques, it becomes apparent that light microscopy (LM) and scanning electron microscopy (SEM) are not the only alternatives for description of new taxa. Modern approaches to classification, as shown by bacteriologists, could also include biochemical and physiological characteristics of organisms. Combination laboratory and field approaches would no doubt yield more accurate data regarding organismal relationships (Pringsheim 1967). The use of DNA and RNA sequencing in strains of Scenedesmus, for example, seems to be a key step in solving affinities within this genus (Hegewald 1997). More natural classification schemes would be produced by integration of multidisciplinary studies. LM and SEM become useful again when they are combined with TEM for the study of life cycles (reproduction, mitosis, cytokinesis), and morphological studies -including plasticity- on vegetative and reproductive cells (e.g., examining plastids, cell walls, flagella, etc.). These data paired with DNA, RNA, protein sequencing and allozyme analysis would no doubt yield a more reliable and objective algal taxonomy and systematics (Morales and Trainor 1997, Trainor and Morales 1999).
When we use the concept of a layered cake for the genus, a typical cake slice (Fig. 3), incorporating several morphs such as A, B, C, D, E, is the species; many of the older 'species' or varieties from the literature simply become morphs. But in time, with the accumulation of more data, we would recognize that two adjacent cake slices, which at first sight appear morphologically identical, are often physiologically and reproductively very, very different. In some groups in which numerous varieties and forms have traditionally been recognized (Scenedesmus/Desmodesmus), the actual species diversity might remain low, but in these species we would have to recognize and incorporate numerous morphs.
In another example, a particular A B C D E slice might better be depicted as A B C D E, simply to point out the increase in thickness of one layer. In nature this readily identifiable morph might persist in the environment far longer than the similar morph for the rest of the species. Is this alteration due solely to environmental influences or because of a mutation? In time and with more data available we might recognize taxa for either option, both of which would be genetic, one because of gene regulation, the second due to mutation.
In the case of Scenedesmus/Desmodesmus (An et al. 1999), our hypothesis is that although at one time there were approximately 330 described species, with the recognition of plastic organisms the number will first be lowered considerably, perhaps even to 20 (Table 1). But then over time it will begin to increase, for, with new interpretations coupled with correction of obvious past mistakes, along with the use of plasticity data, the actual number might eventually approach 250, or more. How is this possible?
As we were alerted to, recognized, and made use of plasticity data with its sequence of development, the number was first lowered. But with additional characters, such as molecular and/or non-morphological data, we recognize that morphologically similar slices are indeed different taxa. In the end, considering our objectives thoroughly and employing a battery of stable characters, the number could steadily rise. But our understanding of each of these taxa, aided by emended protologues for each, will be very different. If we consider that adjacent cake slices might represent species that in reality are not closely related (determined by molecular information), the cake becomes a phylogenetic representation of members of a genus.
We might first want to consider two adjacent cake slices in Figure 4. These may initially look identical, but understand that they exhibit polymorphism in their physiological, molecular, and perhaps even, in a limited sense, in some morphological characters. We might learn that they are sufficiently different so that interbreeding fails, or without reverting to the biological species concept, sufficiently different to express markedly different reaction norms. For Sellaphora pupula (=Navicula pupula) Mann (1989) pointed out that morphologically similar populations were actually reproductively isolated species, but we could always use more supporting data.
In the case of the diatom Fragilaria construens, several morphs of diverse morphology have been described and illustrated (Krammer and Lange-Bertalot 1991), and they have been assigned a variety or form status in some recent taxonomical reviews, while some have even been elevated to species! Although many diatomists have felt that studying the life cycle of diatoms may introduce new and more reliable criteria useful in diatom taxonomy (Cox 1996, Geitler 1968, among others), the more traditional morphologically based taxonomy still persists. The F. construens group includes a diversity of morphs that range from cruciate to triradiate to linear and elliptical. Yet it has not been determined whether there is a purely genetic or an environmental basis for such a diversity. Although it is premature to present a firm conclusion at this point, it is often the case that some of the morphs tend to be restricted to certain environmental conditions, whereas others are associated together under similar environmental circumstances. Whether these are genetically distinct and locally adapted segment of a species (varieties) or morphological variations induced by environmental change (forms) is simply not known.
Nevertheless one can hypothesize that a mixture of both processes is occurring simultaneously within the F. construens complex. Such a supposition can be based on both observations of field collected material and experimental work on closely related species. A well-developed population of F. construens was found in Avery Pond, a small eutrophic system located in southeastern Connecticut, USA (Morales 2001). Several morphological series were found in a surface sediment sample collected from this locality (Fig. 5). In each series, the nominal variety and the variety venter were found to exhibit similar characteristics, except for valve shape.
The first series (first column in Fig. 5) represents what is illustrated in the literature as the typical F. construens var. construens and F. construens var. venter. This series is characterized by a more or less narrow sternum, slightly elongated areolae composing the striae, spines located on the costae, and a few apical pores each surrounded by a rim located on both valve poles.
The second column (Fig. 5) represents a series with wider areolae, which are arranged in much shorter striae, delimiting a wide sternum on the valve face; the spines are also located on the costae. Some specimens in this series have well-developed apical pore fields on both poles of the valves, but again, a siliceous rim surrounds each individual pore.
Finally, the third series (third column in Fig. 5) is characterized by wide areolae arranged in striae that delimit a narrow sternum along the longitudinal axes of the valves. The spines lie on the costae between neighboring striae. Apical pore fields are also developed and located at both valve poles; a siliceous rim surrounds each individual pore.
Since the three series above seem to be closely related to each other, at least morphologically, they might represent genotypes within a single population. It is difficult to justify a variety status for the elliptical morphs for there seems to be a continuum of morphological characteristics with the cruciate morph. There is a distinct possibility that both morphs are expressions of even a single genotype subjected to microenvironmental differences in the same habitat. Hence, a mixture of polymorphism and plasticity could occur in this population.
If one observes material from additional ecosystems from North America and other parts of the world, many more morphological series can be found for the construens group (e.g., Rumrich et al. 2000). This means that the slice of the cake that represents F. construens is very generous, very thick! Each of the morphs within a series (cruciate or elliptical) will occupy a single layer in the cake. But are there more layers within a slice of F. construens?
There are at least two additional morphs exhibiting characteristics similar to those of the cruciate and elliptical types (Fig. 4). One is biundulate and has been described as a distinct variety, i.e., F. construens var. binodis, while the triradiate morph has been depicted as a separate species, namely Fragilaria exigua. The features of the areolae, spines and apical pore fields are similar in these four morphs, although apical pore fields might be extremely reduced or absent in the triradiate morph. It is very common to find a mixture of these morphs in field samples, although the triradiate morph tends to be more restricted in distribution. With these facts in mind, one might conclude that all morphs are related to each other and that they represent environmental variations of the same species. Thus a single genotype could produce all these morphs when subjected to a suite of environmental conditions. Therefore, each morph would be represented in but one layer in the cake (Fig. 4).
With the exploration of wider geographical areas, and intensive use of SEM we learn that several other fragilarioid 'species' also exhibit the above range of morphological variation (Krammer and Lange-Bertalot 1991, Rumrich et al. 2000). The F. brevistriata group, for example, also presents cruciate, elliptical, biundulate (even triundulate), and triradiate examples (Fig. 4, triundulate morph not shown), while those frustules associated with F. pinnata produce ellipsoid and triradiate morphs. Lastly, the F. leptostauron group includes cruciate, ellipsoid, and tetraradiate morphs (they would be positioned on the far right in Figure 4; only the elliptical morph is shown). With the experimental evidence presented by Schmid (1997) for F. crotonensis and Centronella reicheltii, we recognize that the cake analogy for fragilarioid taxa might be perfectly viable.
Schmid (1997) working with unialgal clones under controlled laboratory conditions showed that the change from linear to triradiate valves was reversible and depended on environmental conditions at the time of auxosporulation. This experiment remains one of the most elegant demonstrations of developmental plasticity, for we learn that the phenotype developed depends on the conditions under which the filial individuals are generated.
2. At times algal evolutionary, taxonomic, and ecological thinking does not incorporate sufficiently the concept of phenotypic plasticity. Perhaps we should no longer continue to think of many algal species as morphologically stable entities, since all genotypes may potentially be plastic and/or polymorphic. Our task is to ascertain the degree of plasticity and polymorphism in each one of them.
3. The implications of plasticity in other branches of phycological research (e.g., taxonomy and systematics) are obvious. Classification and identification should be based on analyses of as many types of traits as possible - the traditional morphological aspects as well as information on physiology, life histories, and even ultrastructural features, could provide important clues regarding organismal affinities. Considerations of phenotypic plasticity and polymorphism may significantly affect our view of phylogenetic relationships and classification schemes. In a number of cases, the use of laboratory techniques (e.g., culturing) is of utmost importance in the elucidation of plasticity, since they provide the opportunity to work with genetically homogeneous populations and environmentally homogeneous conditions. However, field and laboratory approaches must not be divorced from each other, for both may contribute to the establishment of true links among organisms. The adoption of either of these approaches to the exclusion of the other may lead to serious error. The development of more natural classification systems should now incorporate multidisciplinary considerations, including phenotypic plasticity and polymorphism.
| Taxa | |
|---|---|
| Initially, an abundance of taxa. | 330 |
| With recognition of an ordered developmental sequence, based solely on morphological features, the number of taxa decreases. The organisms could be displayed in a layered cake; 20 slices equal 20 species. |
20 |
| Over time, with more data, the species number slowly increases, with light microscopy level morphological similarity coupled with SEM differences. Much, much thinner slices of the cake. | 70 |
| Some morphologically similar thin slices are revealed to be physiologically different, i.e., some of these 70 (above) species are polyphyletic. We can no longer use the slice of the layered cake example. | 200 |
| Two isolates, with identical morphs, differ in the duration of display of each morph. . | 250 |
| Transitions between individual morphs is more gradual with considerable overlap. | 300 |
| Combinations of two or more of the above characters, coupled with DNA sequence data. | 400 |