| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Shannon Point Marine Center, Western Washington University, 1700 Shannon Point Rd., Anacortes, Washington 98221-4042
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Abbreviations: b, slope of the log-log plot • OLS, ordinary least squares regression estimate of line fit • RMA, reduced major axis estimate of line fit • 95% CI, 95% confidence intervals (for slope and intercept estimates) • %SEE, percent standard error of estimate
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
In spite of many quantitative studies of cnida size (Stephenson, 1929; Chintiroglou, 1996; Chintiroglou and Simsiridou, 1997; Chintiroglou et al., 1996, Chintiroglou et al., 1997; Chintiroglou and Karalis, 2000; Zamponi and Acuña, 1991; Acuña and Zamponi, 1997; Williams, 1996, Williams, 1998, Williams, 2000), changes in cnida size with body size appear to have been overlooked or discounted; and no studies have been conducted on cnida scaling per se. In the taxonomic literature, tissue-specific size ranges for each cnida population have been treated as species-specific characters (discussed in Fautin, 1988), although a possible relationship between developmental stage or polyp size and cnida size has occasionally been noted (Cutress, 1955; Sebens and Laakso, 1978; Fautin et al., 1989; Ryland et al., 2004).
Sea anemones (order Actiniaria) make good subjects for a scaling study for four reasons. First, body size, rather than age, appears to be the most critical determinant of the life history in this group. Second, growth is reversible: under favorable conditions anemones grow and under unfavorable conditions they shrink, so size and age are essentially decoupled (review in Shick, 1991). Third, in clonal species that aggregate (e.g., Anthopleura elegantissima; Francis, 1973a), a full range of stages may occur simultaneously, providing a morphologically diverse array of individuals cloned from the same zygote (Francis, 1976). Finally, among anthozoans, sea anemones have evolved the greatest variety both of cnidae and of polyps, with the highest level of regional specialization (i.e., different sizes and types of cnidae in different parts of the body), resulting in substantial between-species variability (Schmidt, 1974; Weill, 1934a, Weill, 1934b). This variation within and among species allows strong comparative tests relating cnida scaling to species habitats and tissue functions.
I chose spirocysts for this first study of cnida microscaling for several reasons. First, spirocysts are easy to recognize. Second, they are an abundant, widespread, unique, and probably monophyletic cnida type, showing little evolutionary divergence in form within the Hexacorallia (Schmidt, 1974) and no major morphological differences among the actiniarian anemones (Rifkin, 1991). Third, in members of the genus Anthopleura, spirocysts occur in two tissues that have quite different functions: tentacles that are used mainly in feeding, and acrorhagi that are specialized, inducible structures (Francis, 1976) used only in territorial battles (Francis, 1973b), most often with other anemones (Francis, 1985). Finally, Williams (1996, 1998, 2000), who developed a standardized protocol for sampling, reporting, and testing cnida size variation, has confirmed that spirocysts from Metridium showed higher than usual coefficients of variation (Williams, 2000), validating a general impression that within-sample variability is higher for spirocysts than for other cnidae.
Here I describe scaling of one adhesive cnida type (spirocysts) from two different tissues (ectoderm of the feeding tentacles and acrorhagi) of two sympatric anemones that have contrasting diets, growth forms, and social structures. Anthopleura elegantissima (Brandt, 1835) dominates large areas of the mid-intertidal on rocky shores by replicating asexually to form dense, segregated clonal groups (Francis, 1973a). It uses the inducible acrorhagi to attack and repel all other anemones except clonemates (Francis, 1973b). While A. elegantissima eats primarily plankton, invertebrate larvae, and smaller intertidal invertebrates, Anthopleura xanthogrammica (Brandt, 1835) commonly eats larger intertidal invertebrates, including dislodged mussels and barnacles (Sebens, 1981a). A. xanthogrammica develops larger polyps than A. elegantissima (Sebens, 1981b), does not replicate asexually, and exhibits less aggression against conspecifics (Sebens, 1984). Data for the tentacle spirocysts of Tealia crassicornis (Mueller, 1776) (a closely related aclonal species without acrorhagi), are included for contrast, because this species captures larger and more active prey—including moderate-size crabs, snails, and mussels (Sebens and Laakso, 1978)—than is typical for species of Anthopleura.
Finally, I found differences in spirocyst size among tissues and species to be consistent with apparent differences in the selective regimes; however, increase in spirocyst size with body size was similar for all populations, with scaling exponents similar to those reported for cell size variation within and among animal species (Munro, 1969; Munro and Gray, 1969; Maldonado et al., 1974; Peters, 1983; Calder, 1984; Stevenson et al., 1995). To account for these unusually small scaling exponents, I introduce a possible mechanical explanation for microscaling at the cellular level. Cnida scaling (and the scaling of cellular secretions, generally) may typically reflect the underlying phenomenon of cell scaling.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Within 2 weeks of collection, most animals were either examined alive, or preserved in 10% formalin after first anesthetizing them in seawater by adding an equal volume of magnesium sulfate solution (7.5% by weight of common Epsom salts).
Anemone selection
To determine whether average size of the tentacle spirocysts is a function of body size for the clonal aggregating anemone, A. elegantissima, I collected individuals spanning the range of sizes found within a clonal aggregation at Eagle Cove, Washington. Recently divided animals with obvious fission scars were always excluded. To compare the scaling of acrorhagial and tentacle spirocysts, I collected one or two individuals with well-developed acrorhagi from seven separate clusters of A. elegantissima in densely populated tidepools at Eagle Point, Washington.
To compare scaling patterns for aclonal species, I also collected individuals of Anthopleura xanthogrammica and Tealia crassicornis (another sympatric, aclonal species). Specimens of A. xanthogrammica were selected to span the size range from within a single population that included very small juveniles. Specimens of T. crassicornis included freshly collected individuals from Anacortes, Washington (Shannon Point Marine Center [SPMC]) and a few very large individuals that had been kept in the laboratory for some years (at SPMC, Anacortes, WA, and the University of Washington Laboratories, Friday Harbor, WA).
Tissue selection
One medium-to-large tentacle was taken from each individual anemone. To reduce variability (Robson, 1988), I consistently examined only the proximal one-third of the largest tentacle for the size-graded series of anemones from a single clone. Acrorhagial tissue from the Anthopleura species was obtained from the white tip that contains specialized cnidae used only in aggressive interactions with other anthozoans.
Spirocyst identification and measurement
Unfired spirocysts are easy to recognize in squash preparations of either fresh or preserved tissue. The capsule wall is unusually transparent, and the crystalline material inside the inverted tubule produces very strong birefringence. Under interference contrast illumination, spirocysts look like glowing tapered springs that have been squashed slightly in their transparent capsule sacks (Fig. 1). I prepared smears by macerating a sample of tissue mechanically in a drop of seawater and pressing this suspension into a thin layer between a slide and coverslip. Spirocysts were identified using a light microscope at 1000x (oil immersion) and measured to the nearest 0.1 µm using a computer-linked video camera (image analysis) system. Data on spirocyst sizes were collected systematically by scanning each slide (as described in Williams, 1996) and measuring maximum lengths and widths (method of Hand, 1954) of the first 20 clearly visible, intact, and unfired capsules with their long axes parallel to the plane of the slide. Duplicated measurements using 20 video images indicated that spirocyst dimensions were reproducible within ± 1% (lengths) and ± 4% (widths).
|
Mean capsule dimensions and their standard errors were calculated for systematic samples of 20 measurements from each tissue, which provided a good estimate of the true means (Williams, 1998). Subsequent analysis of variation in these means, rather than the raw data, offers better resolution for detecting real, between-sample differences, especially where levels of within-sample variability are very high, as they are here.
The nonparametric Spearman rank sum correlation (one-tailed test) was used to determine P-values for correlations between mean capsule dimensions and polyp wet weights (Zar, 1984), with groupwise error (asterisks) determined using the sequential Bonferroni adjustment (Jaccard and Wan, 1996).
Graphs were constructed using a linear scale on the Y-axis and log scale (for wet weight) on the X-axis (Longley, 1984) to eliminate the perceptual bias of the traditional log-log plot typically used in scaling studies (Smith, 1984). To compute scaling exponents, anemone wet weights and mean dimensions of the spirocyst capsules were transformed to common logarithms (base 10) for analysis.
Model II regression—slopes (scaling exponents) and intercepts for log-log plots.
Since specimens were deliberately selected to span the existing size range of the anemones, the wet weights (and their logs) are not normally distributed. Furthermore, all of the variables are measured with error; and measurement error for the X and Y variables is similar (± 3% for wet weights and ±1%–4% for cnida dimensions). Thus these data do not meet the assumptions for Model I regression analysis (McArdle, 1988), nor for associated parametric methods such as ANOVA, ANCOVA, and multivariate analysis. For R values less than 0.9, the ordinary least squares (OLS) method typically underestimates the scaling exponent; and for R greater than 0.9, predictions of the Model I and Model II regressions converge (McArdle, 1988; LaBarbera, 1989).
Model II regression is appropriate for describing "functional relationships" such as these (Sokal and Rohlf, 1981; Rayner, 1985). Hence, to determine scaling exponents (slope of the log-log plot), lines were fitted to the log-transformed data using the reduced major axis (RMA) method, also known as the geometric mean regression, probably the most robust method for determining the slope for morphometric data of this kind (McArdle, 1988; LaBarbera, 1989). In addition to being scale-independent, the RMA estimate of the slope (bRMA) is rotation-invariant (Smith, 1984; Rayner, 1985) and easily calculated from the OLS estimate of the slope (bOLS) and R, the Pearson product moment correlation (bRMA = bOLS/R). The asymmetrical 95% confidence intervals (95% CIs) for the slopes were estimated using bootstrapping (10,000 iterations; SYSTAT, 1998, version 9; SPSS Inc., Evanston, II). None of the conclusions in this study were sensitive to the choice of regression models.
Goodness of fit and standard error of estimate (SEE).
The Pearson product moment correlation (R) is affected by the slope of the line and the range of values of both variables, and thus is a biased measure of the goodness of fit (Smith, 1984). Since R decreases as the slope of the line decreases, this is particularly problematic where the slopes are quite small, as they are here. As a measure of the strength of association, Smith (1984) recommended reporting the standard error of estimate (the standard deviation of residuals which is sometimes called "standard error of regression"), expressed as a percent of the mean Y value, for ease of interpretation. For log data, percent standard error of estimate can be calculated as follows:
 |
Significance of between-sample differences.
Comparisons between tissues and species were a priori tests of initial hypotheses, and were based on paired samples from the same individuals or from weight-paired individuals of different species. Paired student’s t-tests (Sokal and Rohlf, 1981) were used to test for significant differences in the dimensions of tentacle and acrorhagial spirocysts from the same specimens. Between-species differences in capsule size were tested similarly by pairing polyps by weight (Zar, 1984) and testing for significant differences in spirocyst capsule dimensions and in polyp wet weights (weight differences never significant, here). The paired t-test assumes only that the differences (and not the variables themselves) are normally distributed (Zar, 1984). This method provides several advantages over simply comparing Y-intercepts using estimated 95% confidence intervals. (1) Conclusions are based on direct comparison of the data and are thus independent of assumptions used in calculating the best fit line and its confidence intervals. (2) Assessment is based on consistency across the overlapping range of body sizes, rather than on predicted differences for one arbitrary body size. (3) Providing P-values allows correction for groupwise error.
Scaling exponents were considered to be significantly different if one lay outside the 95% confidence interval for the other. Where scaling exponents do not differ significantly, line elevations may be compared similarly using the 95% CIs for the Y-intercepts (Hess, 1993).
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
|
|
|
|
|
Scaling of acrorhagial spirocysts from Anthopleura
For the genetically diverse sample of A. elegantissima, both capsule width and capsule length of the acrorhagial spirocysts increased significantly with increasing body weight (Tables 1, 2; Fig. 3). In contrast, for the non-clonal anemone A. xanthogrammica, neither width nor length of the acrorhagial spirocysts was significantly correlated with body size (Tables 1, 2). However, the consistent absence of acrorhagi in very small individuals reduced the size range of this sample (n = 8 individuals with acrorhagi; Fig. 4).
|
|
This shape variation was also apparent within the A. elegantissima clone: the scaling exponent for log individual capsule width as a function of log capsule length (Table 2; Fig. 2C) was 0.66, which was significantly smaller than the theoretical value of 1.0 predicted for proportional (isometric) scaling (95% confidence intervals; Table 2).
Differences in scaling exponents between tissues and species
Scaling exponents for capsule length did not differ significantly between the two Anthopleura species, for either tentacle or acrorhagial spirocysts (slopes all lay within each others’ 95% confidence intervals, Table 1). Although the tentacle spirocysts from T. crassicornis showed a slightly higher scaling exponent than tentacle spirocysts from the genetically diverse samples of A. elegantissima and A. xanthogrammica, this exponent was not significantly higher than that of the A. elegantissima clone (Table 1). Furthermore, the higher scaling exponent of T. crassicornis was heavily influenced by two of the largest individuals (see Fig. 5A).
By contrast, the scaling exponents for capsule width were significantly lower for spirocysts from A. xanthogrammica tentacles and acrorhagi than for the other two anemones, whose capsule width exponents were indistinguishable from each other.
Spirocyst size: significant differences between tissues and species
Consistent differences in mean capsule size are clearly apparent from the combined semilog plot of polyp means (Fig. 5), and from photographs (Fig. 1) of average-sized spirocytes taken from specimens of the same size (10 g wet weight). Tentacle spirocysts from the genetically diverse sample of A. elegantissima were significantly shorter (paired t-test, P < 0.001*), but not narrower (P < 0.2), than their own acrorhagial spirocysts, and were both shorter and narrower than the tentacle spirocysts from T. crassicornis (P < 0.05*). Both tentacle and acrorhagial spirocysts from A. xanthogrammica were significantly shorter and narrower (P < 0.05*) than the intermediate-sized tentacle spirocysts of A. elegantissima.
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
Intraspecific variation in cnida size
First and most importantly, cnida scaling is real: larger Anthopleura and Tealia do produce larger spirocysts. Furthermore, a reexamination of published data suggests that cnida scaling is actually widespread among anemones (Table 3). Of the 39 studied populations, 25 show positive correlations between cnida length and polyp size, and 3 of 13 show positive correlations between cnida width and polyp size. Curiously, rather than characterizing and explaining between-sample variability, previous studies have focused exclusively on defining invariant characters for species description and phylogeny (Stephenson, 1929; Weill, 1934a, Weill, 1934b; Carlgren, 1949; Williams, 1996, Williams, 1998, Williams, 2000; Chintiroglou, 1996; Chintiroglou et al., 1996, Chintiroglou et al., 1997; Chintiroglou and Simsiridou, 1997; Chintiroglou and Karalis, 2000). In this search for stable taxonomic characters, cnida scaling and other examples of intraspecific variability have typically been dismissed as inconvenient or unworkable (but see Ryland et al., 2004). Although cnida scaling appears widespread among anemones, I am not aware of comparable information for other cnidarians.
Second, cnida size varies continuously. No size classes exist within individual samples (Figs. 2–4), or within the pooled sample from a single clone (Fig. 2C).
Variation in cnida shape with increasing cnida size
Multiple lines of evidence indicate that cnida shape changes with cnida size. This is particularly apparent from the combined sample of 200 tentacle spirocysts from 10 clonemates of A. elegantissima (Fig. 2C, Table 2). The width of individual capsules did increase as a function of capsule length; but the scaling exponent of 0.66 was significantly less than the scaling exponent of 1.0 expected for isometry (P < 0.001). In all tissues, all species, and all cnida types examined, cnida length increased more dramatically with increasing anemone size than did capsule width; the single exception being the acontial b-mastigophores from Sagartia elegans (data from Stephenson, 1929; Kramer and Francis, 2004; this study, Fig. 5, Tables 1–3). Size-range data in species descriptions show the same pattern. For 39 distinctive cnida populations from A. elegantissima, T. crassicornis, and A. xanthogrammica, the average percent difference between reported minimum and maximum capsule dimensions was 56% for capsule widths and 102% for capsule lengths (data from Hand, 1955). Larger cnidae typically become disproportionately longer, regardless of cnida type, tissue, or species.
Taxonomic implications of intraspecific variation in cnida size and shape
Because cnida size and shape commonly change with body size within anemone species, their value as taxonomic characters is much reduced (Williams, 1996, Williams, 1998, Williams, 2000). Although taxonomists have acknowledged that intraspecific variability can be problematic (review by Fautin, 1988), cnida size ranges have typically been used in species descriptions, rather than mean cnida dimensions (± SD) for each specimen, because this saves time and space. I strongly recommend including specimen wet weights in all future work, and wherever possible, also including mean capsule dimensions (± SD) for a 1-g specimen, as a basis for quantitative comparisons between populations and species. For more detailed comparative work, cnida scaling exponents and the Y-intercepts (predicted capsule dimensions for a 1-g specimen) provide distinctive characters that may be particularly valuable for separating cryptic species (e.g., Metridium senile and M. farcimen, Kramer and Francis, 2004).
Evolutionary significance of larger acrorhagial spirocysts in A. elegantissima
Increasing localization and specialization of cnidae is a common evolutionary trend among sea anemones (Schmidt, 1974). Specialization through enlargement of the acrorhagial spirocysts in A. elegantissima (this study) and its sibling species, A. sola (Pearse and Francis, 2000), appears to be an example of this pattern. Acrorhagial and tentacle spirocysts are the same size (and relatively small) in A. xanthogrammica (this study) and also in the more distantly related A. artemisia (Hand, 1965), which belongs to a different branch of the genus (Geller and Walton, 2001). Larger acrorhagial spirocysts may be a shared derived character linking the sibling species pair A. elegantissima (this study) and A. sola, and distinguishing them from the closely related A. xanthogrammica.
Functional significance of cnida size and shape
Within the functional limits of a given design, cnida effectiveness may commonly increase with cnida size (see Purcell, 1984). The cnidae are extraordinarily complex collagenous, intracellular secretions serving a wide range of general and specialized functions. They are numerous, costly, and discarded after a single use, puncturing the enclosing cell in the process. Thus their production and use must entail considerable investment by the animals. Larger capsules can carry larger payloads of tubules, spines, barbs, venoms, or glues. Furthermore, fewer cnidocytes are sacrificed during the firing of a few large spirocytes than by firing many small ones. Consequently larger cnidae probably contribute to other advantages of large body size in anemones, including advantage in competition, escape in size from predation, and ability to handle larger prey (Francis, 1979; Sebens, 1983, Sebens, 1987; Kramer and Francis, 2004).
Variation in cnida shape may also impact function. Based on mechanical analysis of pressure vessel geometry, larger cnidae may typically become more elongate, rather than stouter, because of the impact of capsule width on wall stress. Local tension in the walls of a pressurized cylinder increases in proportion to the diameter and is unaffected by the length (Alexander, 1983, p. 153). For the cnidae, this means that wider capsules with the same internal pressure will experience more distortion in the walls, and presumably also more distortion of the catch/latch mechanism, and therefore greater risk of wasteful spontaneous firing. Increasing cnida size beyond the point of spontaneous firing would require a change in cnida design (e.g., thicker cnida walls or a different latch design). This may be one reason that capsule widths increase relatively slowly as cnida size increases with anemone size.
The more elongate shape of larger cnidae may also be important for rapid influx of water during cnida firing (Thomason, 1988). Increase in cnida size without shape change (capsule isometry) would reduce the surface-to-volume ratio (S/V), causing more delayed firing in larger cnidae. The short, narrow spirocysts of A. xanthogrammica should be the fastest in this group, both because of their relatively high S/V ratios and because the relatively shorter tubule in these small spirocysts (Fig. 1A, D) should complete eversion more quickly than the longer tubules in the larger spirocysts of A. elegantissima and T. crassicornis (Fig. 1B, C, E).
Adaptive significance of cnida size, shape, and scaling
Differences in cnida size are likely to be adaptive. For example, relatively higher investment in competitive interference was predicted for clonal than for non-clonal anemones (Francis, 1988), which is reflected in the relatively larger acrorhagial spirocysts of A. elegantissima (this study). As another example, larger defensive nematocysts in the acontia of Metridium farcimen reflect the relatively higher predator densities in its subtidal habitat, by comparison with the smaller acontial nematocysts associated with the shallower habitat of M. senile (Kramer and Francis, 2004).
Differences in cnida shape (Fig. 1) are also likely to be advantageous. For example, having short, narrow tentacle spirocysts that complete their firing very rapidly may be particularly advantageous for A. xanthogrammica (Fig. 1A), which captures loose mussels in wave-exposed areas (Sebens, 1981a, Sebens, 1983) of high velocities and rapid accelerations (Denny et al., 1985). A more delayed response by the wider and longer tentacle spirocysts of A. elegantissima (Fig. 1B) may be adequate for capturing smaller, lighter prey (Sebens, 1981a), and should be more economical in terms of cell (spirocyte) loss during firing. On the lower shore, T. crassicornis has very long and wide tentacle spirocysts (Fig. 1C) whose higher payload volumes are perhaps more important than a very rapid response time for capturing and holding crabs (Sebens and Laakso, 1978), which although powerful, are not fast-moving.
Cnida scaling may also be adaptive. For example, if larger spirocysts with their larger payloads permit greater adhesion to prey, then larger tentacle cnidae may yield increased capture success with larger prey. Both cnida size (this study) and prey size increase with body size for A. xanthogrammica (Sebens, 1981a) and for A. elegantissima (Spearman rank correlation for anemone size class vs. average prey size, two-tailed test, rs = 0.762, P < 0.05; calculated using data from Sebens, 1981a). This connection between prey and cnida scaling is also supported by contrasting data for two planktivorous Metridium species, where prey size does not increase with body size (Sebens, 1981a). For M. farcimen and M. senile, the length of tentacle cnidae increases more slowly with body size (scaling exponents, 0.008–0.03, Kramer and Francis, 2004) than for macrophagous Anthopleura and Tealia species (scaling exponents, 0.052–0.086, this study). Thus cnida scaling patterns can be treated as features of cnidarian life histories.
Structural implications of cnida size and shape
From a design standpoint, providing space and support for needle-like cnidae and assuring the stability of surrounding soft tissues could be problematic, especially during cnida firing. Anemones have the greatest variety of cnidae in the class Anthozoa (Schmidt, 1974), which differs from Scyphozoa and Hydrozoa in having elongate, rather than ovoid or spherical, cnidae (Mariscal, 1984). Like the cnidae that they support, the epithelial cells in anemones are unusually tall and thin, secreting and supported by an unusually fibrous and structured mesoglea which almost qualifies as connective tissue (Hyman, 1940; Chapman, 1966; Gosline, 1971; Koehl, 1977).
At the outer limits of stability, very dense arrays of very long cnidae in defensive structures (e.g., capsule lengths 
90 µm for Metridium acontial amastigophores and Anthopleura acrorhagial holotrichs; Hand, 1955) actually do cause local tissue disintegration during firing (Äbel, 1954; Francis, 1973b, Kramer and Francis, 2004). Maximum size of these cnidae may be limited by tissue thickness, or by any tendency to disrupt tissues during regular cycles of extreme body elongation and contraction (illustrations in Shick, 1991).
In contrast, cnidae in the feeding tentacles are less densely arrayed and smaller (capsule lengths 20–30 µm for Tealia and Anthopleura; Hand, 1955), and are not likely to disrupt the tissue during ordinary body movements. However, since tentacle cnidae are used frequently in numerous prey capture events, any tissue disruption during cnida firing will interfere with further prey capture, thereby favoring tissue stability in the feeding tentacles also. In both situations, cnida size is likely to be constrained more narrowly in smaller specimens with thinner tissues and less structural support.
Possible causes of cnida scaling
Cnida scaling could arise through natural selection where size-limiting factors apply more strongly to small animals. Limiting factors that may become progressively less restrictive with increasing body size include physiological and ecological factors such as the food and energy flux (Sebens, 1981a), as well as morphological factors such as tissue and mesoglea strength and stability (Shick, 1991), and physical limitations of space such as tissue thickness (Shick, 1991) and cell size (Peters, 1983; Calder, 1984; Stevenson et al., 1995).
Since the thickness of both the mesoglea and the inner and outer epithelial layers of the anemone body typically does increase with body size (Shick, 1991), larger anemones should be able to accommodate and support larger cnidae. While this may explain scaling of exceptionally long cnidae that are at or near the limits of tissue tolerance (e.g., capsule length of acrorhagial holotrichs 90 µm; Hand, 1955), it would not explain scaling of the much smaller acrorhagial spirocysts interspersed among them (capsule lengths 30µ, this study). Clearly, then, not all instances of cnida scaling are due to gradual release of the more rigorous spatial and support constraints on smaller individuals.
In fact, no one of these factors can explain all the available data. If increase in cnida size were purely a function of higher energy availability in larger specimens, then cnida size should increase similarly with body size in all the tissues of a species. In fact, scaling exponents for the tentacle cnidae are typically significantly smaller than for the acontial cnidae in both Metridium senile and M. farcimen (Kramer and Francis, 2004). Since scaling exponents may vary between tissues as well as between species, cnida size clearly is not controlled uniformly by size-related differences in the energy budget of the whole animal.
Cnida scaling also occurs where there is no obvious selective advantage based on functional differences between larger and smaller individuals. Although prey size does not increase with body size for Metridium (Sebens, 1981a), larger individuals do have longer tentacle cnidae (Kramer and Francis, 2004), although these scaling exponents are unusually small (0.018–0.03, as compared with 0.052–0.072 for Anthopleura tentacle spirocysts, and 0.035–0.051 for Metridium acontial nematocysts). Thus larger tentacle cnidae in larger individuals cannot always be explained as aiding in the capture of larger prey.
Developmental significance of cnida size differences
Cnida scaling may be caused proximally by cell scaling. In Hydra, cnida size is related to cell size, which declines with the number of divisions that an interstitial cell undergoes before differentiation. More divisions result in more and smaller cnidocytes, which produce smaller cnidae (Lehn, 1951). This may also be true for anemones, where otherwise continuous cnida populations often include an occasional double-sized or half-sized capsule (Cutress, 1955; Daphne Fautin, University of Kansas, pers. comm.), perhaps produced by one division more or less than the usual for that interstitial cell line.
A direct relationship between cell size and the size of secreted structures may actually be rather common. In Drosophila, local polyploidy in bristle-secreting cells results in larger cells that secrete larger bristles (Adler et al., 2000); and in the roundworm Caenorhabditis elegans, a mutation that produces adult dwarfing has been traced to miniaturization of a cell line that contributes to the size of the cuticle-secreting and syncytial hypodermis (Knight et al., 2002).
If cell size determines cnida size, then cnida scaling presumably implies cell scaling. That is, the observation that larger anemones have larger cnidae implies that the size of the cnidocytes also varies continuously, and reversibly, as a function of anemone body size.
Implications for cell and organelle scaling
While growth typically involves changes in body proportions, the fact that this also occurs at the cellular and subcellular level may be surprising (Brown and West, 2000). Order-of-magnitude agreement between cnida scaling exponents (0.018–0.086; this study; Stephenson, 1929; Kramer and Francis, 2004) and cell scaling exponents (0.017–0.17; Peters, 1983; Calder, 1984; Stevenson et al., 1995) suggests that cnida scaling may reflect a wider pattern of cell scaling.
Based on these accurate cnida measurements, a power law for cnidae and cells may be closer to 1/10 than to the well-known 2/3–3/4 power law for gross morphological and physiological functions. Measured scaling exponents for the lengths (bL = 0.052–0.086) and widths (bW = 0.021–0.039) of unfired cnidae can be used to estimate scaling exponents for cnida surface areas (bSA = bL + bW = 0.073–0.125) and volumes (bV = bL + 2bW = 0.094–0.164) as functions of body mass. Thus cnida functions that depend on capsule surface area or volume should have exponents between 0.073 and 0.164, quite different from the 0.67–0.75 exponents reported for gross body functions (Peters, 1983; Calder, 1984; Schmidt-Nielsen, 1984; Niklas, 1994).
As a general explanation for the scaling of epithelial secretory cells and their structural products (such as extracellular mesoglea and basement membrane), I suggest that larger, taller epithelial cells can supply more structural material per unit attachment area for building and maintaining thicker sheets of pliant, extracellular support materials to resist the increased wall stresses in larger, stouter pressurized cylinders of all kinds—including worms, blood vessels, and guts, as well as sea anemones (Francis, unpubl. obs.). The scaling of cnidocytes and their intracellular cnidae can be considered a special, restricted case of this more general phenomenon, since cnidae cannot be larger than the cells enclosing them.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
Literature Cited |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Kramer and L. Francis Predation Resistance and Nematocyst Scaling for Metridium senile and M. farcimen Biol. Bull., October 1, 2004; 207(2): 130 - 140. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
