Printer Friendly

Genetic variation and polymorphism in the inducible spines of a marine bryozoan.

The problem of how genotypes evolve a broad phenotypic range or extreme "plasticity" has been of interest to biologists since Woltereck (1909) conducted the first norm-of-reaction experiments with genotypes of Daphnia that produced a range of helmet types. The problem attracted the interest of experimentalists such as Schmaulhausen (1949), Waddington (1959), and Bradshaw (1965). Recently the evolution of plasticity has received both empirical and theoretical attention (Via and Lande 1985; Schlichting 1986, 1989; Stearns 1989; Scheiner and Lyman 1991; Via et al. 1995). Of particular value to understanding the evolution of plasticity are the few clear examples of adaptive plasticity, such as inducible structural defenses of animals (reviewed in Havel 1986; Dodson 1989; Harvell 1990). The inducible structural defenses of animals are particularly dramatic and include such examples of changed morphology as enlarged helmets and spines in cladocerans (Woltereck 1909; Havel 1986), enlarged spines in rotifers (Gilbert 1966; Gilbert and Stemberger 1984), enlarged spines in marine bryozoans (Yoshioka 1982; Harvell 1984), asymmetrical growth forms in barnacles (Lively 1986), thickened shells in mollusks (Appleton and Palmer 1988), and deeper body form in fish (Bronmark and Miner 1992). These phenotypically variable morphologies should provide an opportunity to examine natural selection in action, but first the frequencies of different phenotypes must be mapped in field populations and the genetic basis of the inducible characters established. Although some of these different phenotypes have typically been viewed as abrupt developmental shifts (Stearns 1989), there is also evidence that the induced spines of the bryozoans (Harvell 1990) and the neckteeth and helmets of cladocerans (Vuorinen et al. 1989; Parejko and Dodson 1991; Spitze 1992; Tollrian 1995; Spitze and Sadler 1996) are amplifiable characters with a continuous trait distribution once a threshold of inducer has been exceeded.

What is unknown for most of these examples is the extent to which field populations are a mix of inducible and constitutive phenotypes and what the frequencies of different types are in the field. And indeed, only for Daphnia are there estimates of the heritability of the inducible response (Parejko and Dodson 1991; Spitze 1992) and information on the variation in the inducible response among individuals (Spitze and Sadler 1996). Because inducible and constitutive individuals are not directly separable in morphology or even by simple behavioral assays, determining the frequency of inducible and constitutive phenotypes from field populations requires experiments in which newly metamorphosed individuals are grown in a common environment and then induced. The norm of reaction is a standard way to express the variation within genotypes over an environmental range (Woltereck 1909; Via and Lande 1985; Stearns and Koella 1986; Dodson, 1989; Stearns 1989; Parejko and Dodson 1991; Spitze 1992; Rollo 1995), although it is not always possible to actually divide genotypes among environments. However, many of the inducibly defended invertebrates, such as the cladocerans, rotifers, and bryozoans, are clonal (Harvell 1990). This permits direct description of within-clone variation as a norm of reaction, which is defined as the difference in mean phenotypic values that are expressed in several environments by clonal replicates (Via 1987).

The inducible spines of bryozoans are an unusually good study system for analysis of the genetic and environmental components of variation in an inducible, plastic response. The spines are produced rapidly (within two days), appear to vary with the level of inducer, and can be nondestructively assayed. Because they consist of a single genotype, colonies can be subdivided to permit direct measurement of clonal heritability and norms of reaction. Using Membranipora membranacea, I estimated three major parameters affecting the potential for evolution in field populations: (1) the frequency of the inducible and constitutive phenotypes in the population at Friday Harbor laboratories (FHL); (2) the clonal heritability of the inducible trait; and (3) the range of phenotypic responses of genotypes from two field environments and across a gradient of inducing factors.


Life History of Membranipora membranacea

Membranipora membranacea is a circumglobally abundant cheilostome bryozoan that is epiphytic on laminarian kelps. In the San Juan Archipelago, Washington, the planktonic larvae of Membranipora first recruit onto kelps during May. Larvae settle and metamorphose in high numbers, creating dense populations of the bryozoan on kelp blades. Recruitment continues until September, when the adult colonies die (Harvell et al. 1989).

Like some other cheilostome bryozoans, M. membranacea is a protandrous hermaphrodite: individual colonies ontogenetically progress from a prereproductive stage through a male stage to a transitional stage, when both sperm and oocytes are produced. This sequence of reproductive stages appears to be developmentally set; however, the time spent in any one stage is not well correlated with colony size and varies with crowding (Harvell and Grosberg 1988; Harvell and Helling 1993). Colones release long distance dispersing larvae that live and feed in the plankton for two to four weeks (Yoshioka 1982).

Frequency of Spined Phenotypes

In May 1995, 210 colonies newly settled on lucite panels (less than one day postmetamorphosis; panels = 2.5 x 5.0 cm) were transferred from underneath the breakwater at FHL to laboratory culture. Newly metamorphosed colonies can be aged, because it takes approximately one day for metamorphosis to proceed from initial settlement of the larvae to development of feeding lophophores in the twinned initial zooids of the colony. Colonies were transferred to laboratory culture as a continuous cohort every day from 15-24 May, resulting in seven replicate vessels each containing about 30 colonies. Colonies were clipped onto racks and each 4-L culture stirred with an oscillating paddle. Colonies were fed Rhodomonas (approximately 10,000 cells/mL) once daily. This is an adequate feeding schedule for small colonies ([less than] 100 zooids); larger colonies require more feedings or continuous culture because their clearance rates are so high. The laboratory colonies grew at a rate comparable to the field cohort (Harvell, pers. obs.) and 85% of colonies survived the entire experiment. This is excellent survivorship, since we were transplanting newly metamorphosed colonies, which are a delicate part of the life cycle, from field to laboratory conditions. On 15 and 16 June, colonies were exposed to nudibranch water (from Doridella steinbergae) and spine lengths were censused on 20 June. Spines can be induced by exposing colonies in culture to filtered sea water that has contained nudibranchs actively feeding on bryozoans (Spine-inducing substance: SIS). Typically spines are apparent within 48 hours, but continued monitoring for five days gives slower colonies time to respond. On each colony, the five longest spines were measured as an estimate of average maximal response. Data were analyzed with ANOVA as the change in spine length from preexposure (15 June) to postexposure (20 June). Colonies with spines recorded at the preexposure time were classified as constitutively spined and those with no spines by five days postexposure were classified as unspined. The colonies that increased the length of spines over the duration of the experiment were classified as inducible.

Clonal Heritability

Clonal heritability was assessed for 16 colonies (genotypes) that were collected from eight algal blades collected from Rocky Cove. Blades were haphazardly collected by scuba divers, but were later screened to ensure undamaged, homogeneously sized colonies (mean diameter of 19.34 [+ or -] 0.68 mm) that were uncrowded by conspecifics and unspined. Each circular colony was divided into four equal-sized wedges, allowed to heal for 24 hours and then tested at a single concentration of SIS to reveal variation in response of individual genotypes, all of which were exposed to a common environment. The experiment was run as a blocked design, with two groups of eight genotypes to allow independent estimation of clonal heritability. Each group of eight genotypes was tested in four vessels such that each vessel contained a section from each of the eight genotypes. The four vessels were intended to be as identical as possible, but there was still the possibility of a vessel effect due to slight variations in light level among vessels or in timing of dosing. For instance, because the bryozoans are fed algae, variation in light could affect resource availability among vessels. To mitigate these effects, experiments were conducted in an environmental chamber and colonies were moved among vessels weekly. Each vessel received a dose (1 L per 4-L vessel) of the same inducing factor given to all other vessels. To minimize any variation in SIS among vessels, the dosage SIS was prepared by collecting filtered water in which 100 actively feeding nudibranchs had been cultured overnight in two 4-L vessels, and subdividing the pooled samples of SIS among the eight vessels of the experiment. Vessels were dosed on day 1 and day 2; the entire experiment ran for approximately 72 hours.

Maximum spine length was estimated by measuring the five longest spines from each colony portion, and using mean maximum spine length as the response variable. To examine the repeatability of sampling only the five longest spines from each colony, spine length was measured for 10 haphazardly chosen zooids in each of four regions of the colony. A pooled average for each colony was calculated from the averages of each region. This estimate of spine length was compared to the predictions of spine length based on only sampling the five longest spines from each region which yielded an [r.sup.2] of 0.80 (n = 13 colonies), indicating that sampling the five longest spines is a reasonable predictor of the spination state of a colony.

The spine lengths were analyzed with a one-way ANOVA to assess the magnitude of the within- and between-vessel variation. A nonsignificant vessel effect indicates that the variation among clones in the different vessels was negligible and also that within-clone variation in response is low. A significant among-clone effect indicates a heritable response. The clonal heritability was estimated as the broad-sense heritability: genotypic variance divided by the total environmental or phenotypic variance (Falconer 1985; Via 1991). The variances were obtained with SAS procedure VARCOMP to determine the between- and within-clone components of variation. The within-clone variance will be overestimated in this case because it also contains the variance due to vessel effects; thus the heritabilities were measured conservatively.

Norms of Reaction

Norm-of-reaction experiments were conducted by dividing each colony into four, equal-sized portions and exposing each portion to one of a four-level concentration series (1, 2, 10, and 25 nudibranchs/L). Colonies were collected from two different sites. Turn Island is a typical high-current site in the San Juan Islands, exposed to open channel water and high currents. The abundance of nudibranchs is lower at Turn Island than at many other sites we usually sample (Harvell 1985). The FHL dock is located inside the town harbor and is protected from the high currents and waters of the open channel. It is consistently a site with the highest densities and largest nudibranchs (Harvell 1985), and a good site to compare with the open channel to investigate whether the response of colonies varied with environment. Variation between two such close sites is not expected because of the highly dispersive nature of the nudibranchs.


Frequency of the Inducible Phenotype

Eighty-five percent of the newly metamorphosed colonies (n = 210) survived the approximately 30 days in the laboratory to be assayed for spine type. Although this is excellent survivorship, the loss of 15% could bias the estimate of the frequency of the different types in nature in the unlikely event that mortality was differential. Because the mean increase in spine length did not vary among the seven replicate vessels (ANOVA F = 1.79, P = 0.10), data from the vessels were pooled for subsequent analysis. Spine length increased after exposure to SIS in both the inducible and constitutively spined colonies [ILLUSTRATION FOR FIGURE 1 OMITTED]. Before exposure to SIS, the constitutively spined colonies already had spines that differed from the other colony types (ANOVA F = 59.39, P = 0.0001; Scheffe P [less than] 0.05 for comparison of constitutive with both unspined and inducible). Spine length increased similarly in both the constitutively spined and inducibly spined colonies (ANOVA F = 113.3, P = 0.0001; Scheffe P [greater than] 0.05), and both had significantly longer spines than the colonies classified as unspined (Scheffe P [less than] 0.05). The constitutive colonies were 6.2% of the population (total n = 178), the unspined colonies were 13.4% of the population, and the inducible colonies were 80.3% of the population. In 1993, a preliminary trial of this experiment with 121 colonies yielded similar frequencies of the three types, including a percentage of 88.4% for inducible colonies, 6.6% for constitutive, and 4.9% for unspined. Spine length was not measured in 1993.

Clonal Heritability

Four fragments from each of 16 initially unspined genotypes were exposed to the same level of SIS to measure the between- and within-clone variance in length of spines produced [ILLUSTRATION FOR FIGURE 2 OMITTED]. The experiment was blocked into two replicates to allow independent estimation of the clonal heritability (Falconer 1985; Via 1991) for each of eight genotypes. The broad-sense heritability is the proportion of the total phenotypic variance that is explained by the genetic variance. The estimates of the within- and between-colony variance components are presented in Table 1. Not only is the between-colony variance highly significant relative to the within-colony variance, but the two estimates of heritability are both high. The mean heritability for the two blocks is 0.692. As expected given the high heritability and small within-colony standard errors, spine length varied significantly between colonies (nested ANOVA F = 265.8, P [less than] 0.0001; colony F = 12.878, P [less than] 0.0001).

Norm of Reaction

Figure 3 shows the mean responses of colonies that originated in two locations (FHL dock and Turn Island) to a concentration gradient of spine inducer. Because each genotype was measured over the four concentrations to obtain a genotype-specific response, the repetition of sampling the same genotype needed to be accounted for in the analysis; therefore, a nested ANOVA was used. Extract concentration significantly affected spine length at both sites (Table 2). At Turn Island, increasing extract concentration resulted in slightly longer spines, although the response appeared to rapidly saturate [ILLUSTRATION FOR FIGURE 3 OMITTED]. At the FHL dock, the effect of increasing SIS was more complex, with an odd decrease in spine length at higher concentrations, possibly due to an inhibitory effect of SIS at higher concentrations. At both sites, spine type also changed with level of inducer [ILLUSTRATION FOR FIGURE 4 OMITTED]; membranous spines were not induced at the lowest concentration, but were produced in combination with straight corner spines at all the higher concentrations (dock [[Chi].sup.2] = 26.88, P = 0.0001; Turn [[Chi].sup.2] = 13.43, P = 0.0038).


The biggest gaps in our understanding of the potential for evolution of inducible defenses are knowing the variability and heritability of the inducible defense and the composition of phenotypes in natural populations. Because many inducibly defended animals are clonal (Harvell 1990), a norm-of-reaction experiment is a good method for both assessing clonal heritability and examining the range of genotypic responses across environments. Using M. membranacea as a study organism, I found variation between colonies and high clonal heritabilities for the magnitude of the inducible response. Clonal heritabilities are also called broad-sense heritabilities, because they include maternal and epistatic effects. Narrow-sense heritabilities, typically obtained from parent/sexual offspring regressions, partition out maternal effects and estimate only additive genetic variance. Most heritability estimates published for clonal animals such as aphids (Via 1994), cladocerans (Parejko and Dodson 1991; Spitze 1992), and hydrozoans (Yund 1991) are broad-sense heritabilities. If genetic effects were perfectly transmissible and unaffected by environment, we would expect the clonal heritability to be one. However, even if the genetic effects are completely transmissible, character values are often strongly influenced by environment. Within the spatial scale of a colony, the environmental effects should be uniform and thus contribute to a high heritability. Even in colonies of uniform genotype, there is the potential for different parts of a colony to grow in different environments (Buss 1983; Harvell 1991) and therefore to have a changed response due to environmental history. Once the clone is subdivided, different environments should act on each ramet to disrupt the heritability. An added complexity of working with colonial invertebrates is that they have the potential to exist as a chimera and thus may not be uniform genetically either because of somatic mutation or fusion of similar genotypes (Whitham and Slobodchikoff 1981; Buss 1983). The results of the heritability experiments with M. membranacea indicate high clonal heritability, suggesting that none of these genetic and environmental effects are interfering with the expected uniformity of a clonal response. Furthermore, the high heritability in length of spines and low variability across clones indicates the replication was without artifact. Since all the colonies in the heritability experiment came from a similar field environment, it also seems likely that the differences in spine length among colonies is due largely to genetic differences and not environmental differences in their ontogeny. It is possible that this experiment overestimates clonal heritability due to initiating the experiment only one to two asexual generations after the clonal isolates were made. This is considered an adequate interval to rid cladocerans (Lynch and Ennis 1983) or aphids (Via 1994) of maternal environmental effects, but may be less adequate in a clonal isolate with recently severed links to the original base colony.
TABLE 1. The within- and between-colony variance components
(of spine length) for 16 colonies exposed to SIS. The experiment was
run in two blocks with eight colonies per block; each colony was
divided into four equal parts.

Variance                          Block 1           Block 2
component                         Estimate          Estimate

Var (vessel)                     0.00002331        0.00008002
Var (colony)                     0.00084145        0.00141186
Var (error)                      0.00024422        0.00040255

Total phenotypic variance        0.0013069         0.0018943
Clonal heritability              0.64              0.74

Levels of within- and between-colony plasticity have also been investigated in tropical bryozoans to evaluate the systematic value of these characters and to assess the scope for evolution of plasticity. Cheetham et al. (1995) found that for two species of cheilostome bryozoans, within-colony variation accounted for two-thirds of the phenotypic variance. They also detected significant heritability for plasticity itself, supporting the notion that plasticity is a significant trait among cheilostomes. More surprising was a low between-site variation in plasticity, since inducible characters would likely vary among sites. This suggests that none of the plasticity in the zooid morphologies that they detected was induced by biotic agents.

One of the next issues to resolve in linking observed inducible types with their genetic potential is understanding how the genotype-specific response varies with concentration of the inducer. Norm-of-reaction experiments are designed to reveal a range of responses expressed by individuals encountering different environments. I used a nested sampling design (clones within sites, predator cue treatments within clones) to estimate the pattern of variation in spine length. When individual colonies of M. membranacea were partitioned and simultaneously exposed to a concentration series of inducer, the observed response indicated a wide norm of reaction and considerable variation among individuals in the shape of the response. Some colonies from Turn Island showed an amplifiable response - an increase in spine length with increasing concentration of cue - although the response was not linear. For colonies from the dock, the amplifiable response was obscured because the spines were actually smaller at higher levels of inducer, suggesting an inhibitory effect of inducer in some colonies. Most interesting was the result that individual colonies produced different spine types at different concentrations of inducer. Colonies from both sites produced small corner spines at a low concentration. At most higher concentrations, the colonies produced membranous spines as well as the corner spines. No information is yet available about the functional significance of the different spine types. Most earlier studies with inducible spines of Membranipora have focused on the membranous spines (Harvell 1984, 1991, 1992), since this is the more definitive response. The deterrence of corner spines alone has never been tested, nor have costs been partitioned for the different spine types, since the usual response to nudibranch predation is the production of both corner and membranous spines. Overall, the experiments revealed large clone-specific variation in amplitude of spine length from both sites, confirming the result of significant between-clone differences in the heritability experiment.
TABLE 2. Analysis of corner spine length in norm-of-reaction
experiment (mixed-model nested ANOVA).

                  Sum of
Source     df     squares     Mean square     F-ratio     P-value

Const       1     2.89425       2.89425       221.16      0.0007
Tmt         3     0.03926       0.01309         4.11      0.0156
Colony     28     0.08923       0.00319         0.65      0.8746
Site        1     0.01789       0.01789         3.64      0.0656
Error      31     0.15226       0.00491

Total      63     0.29864

A future issue will be to determine if the selective regime (measured as choices by predators) is responsive to the observed variation in spine length and type. The norm-of-reaction experiment was conducted with colonies from two field sites that vary in predator level. As would be expected with a highly dispersive species like M. membranacea (larvae spend four weeks in the plankton), there are no consistent differences in the reaction norms between sites. No conclusion about causes of variation between the sites is possible since they were not replicated, but the differences between these two sites does indicate the potential for a between-site analysis to reveal meaningful patterns in response, such as mapping heightened responsiveness of colonies from high predation sites.

The population of M. membranacea at Friday Harbor in 1993 and 1995 was composed of a mix of inducible and constitutively spined colonies. Inducible colonies represented the majority of phenotypes; 80-88% of colonies were inducible. However, there was a significant fraction of undefended and constitutively spined colonies in both years. This experiment highlights the difficulty of working with phenotypically plastic morphologies: separating inducible from noninducible morphologies requires a laboratory assay. I could not have distinguished the constitutively spined individuals without growing them in a predator-free environment for their entire ontogeny. The experiment clearly indicates the presence of three different types in the population. Is this a genetic polymorphism?

The discovery of considerable variation in the inducible spine response in field populations provides an unprecedented opportunity to investigate the factors maintaining variability in inducible defenses. Particularly promising is the indication that there is a polymorphism in response, with distinct morphs, such as inducible, constitutive, and unspined, present in a single population. Some caution is necessary before concluding that this is a true genetic polymorphism of discontinuous morphological types. It is also possible that the population is actually a continuous distribution of colonies with variable thresholds to the inducer. Since the common-garden experiment was conducted with a single level of inducer, it is possible that the "unspined" types are actually inducible colonies with a high threshold for induction. However, the design was biased to detect this by using a high level of inducer (the equivalent of 15 nudibranchs/L), exposing colonies on two consecutive days, and allowing five days for a response. Similarly, the "constitutive" colonies could have been responding to some other nonnudibranch culture condition and in the field might not have produced constitutive spines. Nonetheless, they produced spines in cultures before any individuals had been exposed to SIS. Taken together with the previous experiments showing a high heritability of the inducible response, the three spine types seem to represent a genetic polymorphism in inducible defenses.


The evolution of phenotypically plastic characters is a notoriously difficult problem, in part because of the need to describe a single genotype across several environments. The problem is compounded by the potentially large effects of previous environment and maternal environment (Rollo 1995; Tollrian 1995). The inducible defensive spines of the marine bryozoan, M. membranacea, are a character of clear functional significance, and provide a good opportunity to examine the type of variation in plasticity in natural populations and its causes. This work reveals the complexity of the inducible response that is characterized by two different spine types, varying with concentration of inducer; a threshold of activation; a slope of amplification; and an upper threshold for size of spine (Harvell 1990), all of which vary among different colonies in a population. Thus the inducible response is actually composed of several correlated characters, any of which may be the actual target of selection. I used spine length at a fixed time interval from exposure as an aggregate measure of this complex response and detected high variation in this character among colonies and a high clonal heritability. The norm-of-reaction experiments show the range of character values from two field environments and sensitivity to concentration of inducer. Most interestingly, a common-garden experiment revealed a polymorphism in inducible spine type.

The next steps in detecting the evolution of inducible spines are to assay more geographically widespread populations to evaluate if there is correlation between level of plasticity and predator load, and to conduct selection experiments to demonstrate the predicted effects of selection on the range of variation present (i.e., high predator loads should cause local extinction of undefended types and the inducible type should dominate in a variable environment). However, it is likely that the variation in these populations is maintained by several factors. There is a cost to producing spines for the inducible type that could balance the benefits of defense (Harvell 1986, 1992; Grunbaum 1997) and thus maintain the polymorphism in defense type. Nothing is known yet about the relative growth rates and reproductive outputs of the three distinct types in the population. An obvious question is whether the constitutive morph grows more slowly than the inducible and undefended morphs. Even more interesting is the question of whether the undefended morph grows more rapidly than the inducibly defended one. Finally, these bryozoans do exist in a complex biotic regime where the threats to survival and growth from competition are as great as those from predation (Harvell et al. 1989). Colonies within these same populations also produce an inducible defense against competitors - large colonies produce stolons against smaller competitors (Harvell and Padilla 1990). Negative or positive correlations between these two inducible defenses could also maintain the polymorphism and variation in inducible defenses.


I am grateful to B. Helmuth for field assistance and to K. Wirtz, C. Griggs, A. Dettelbach, and J. Shaffer for help with laboratory experiments. J. West and K. Kim improved the manuscript and L. Buttel assisted with statistical analysis. The research was funded by National Science Foundation grants OCE-8817498 and IBN-9408228 to CDH. As ever, I am very grateful to A. O. D. Willows and R. R. Strathmann for providing tremendous facilities at the Friday Harbor Laboratories.


APPLETON, R., AND A. R. PALMER. 1988. Waterborne stimuli released by predatory crabs and damaged prey induce more predator-resistant shells in a marine gastropod. Proc. Nat. Acad. Sci. 85:4387-4391.

BRADSHAW, A. D. 1965. Evolutionary significance of phenotypic plasticity in plants. Adv. Genet. 13:115-155.

BRONMARK, C., AND J. G. MINER. 1992. Predator-induced phenotypic change in body morphology in crucian carp. Science 258:1348-1350.

BUSS, L. W. 1983. Evolution, development, and the units of selection. Proc. Nat. Acad. Sci. 80:1387-1391.

CHEETHAM, A. H., J. B. C. JACKSON, AND L. A. HAYEK. 1995. Quantitative genetics of bryozoan phenotypic evolution. III. Phenotypic plasticity and the maintenance of genetic variation. Evolution 49:290-299.

DODSON, S. 1989. Predator-induced reaction norms. Bioscience 39: 447-453.

FALCONER, D. S. 1985. Introduction to quantitative genetics. Longman, London.

GILBERT, J. J. 1966. Rotifer ecology and embryological induction. Science 151:1234-1237.

GILBERT, J. J., AND R. S. STEMBERGER. 1984. Spine development in the rotifer Keratella cochlearis: induction by cyclopoid copepods and Asplanchna. Freshwater Biol. 14:639-647.

GRUNBAUM, D. 1997. Hydromechanical interactions within an encrusting bryozoan colony: mechanisms of colony organization and cost of defense in Membranipora membranacea. Limnol. Oceanogr. 42:741-752.

HARVELL, C. D. 1984. Predator-induced defense in a marine bryozoan. Science 224:1357-1359.

-----. 1985. Partial predation, inducible defense, and the population biology of a marine bryozoan. Ph.D. diss. Univ. of Washington, Seattle.

-----. 1986. The evolution of inducible defenses in a marine bryozoan: cues, costs and consequences. Am. Nat. 128:810-823.

-----. 1990. The ecology and evolution of inducible defenses. Rev. Biol. 65:323-340.

-----. 1991. Coloniality and inducible polymorphism: the interaction of zooid and colony level control. Am. Nat. 138:1-14.

-----. 1992. Inducible defenses and allocation shifts in a marine bryozoan. Ecology 73:1567-1576.

HARVELL, C. D., AND R. K. GROSBERG. 1988. The timing of sexual maturity in clonal animals. Ecology 69:1855-1864.

HARVELL, C. D., AND R. H. HELLING. 1993. Experimental induction of localized reproduction in a marine bryozoan. Biol. Bull. 184:286-295.

HARVELL, C. D., AND D. K. PADILLA. 1990. Inducible morphology, heterochrony, and size hierarchies in a colonial monoculture. Proc. Nat. Acad. Sci. USA 87:508-512.

HARVELL, C. D., H. CASWELL, AND P. SIMPSON. 1989. Density effects in a colonial monoculture: experimental studies with a marine bryozoan (Membranipora membranacea). Oecologia 82:227-237.

HAVEL, J. 1986. Predator-induced defenses: a review. Pp. 263-278 in W. C. Kerfoot and A. Sih, eds. Predation: direct and indirect impacts on aquatic communities. Univ. Press of New England, Hanover.

LIVELY, C. M. 1986. Competition, comparative life histories, and maintenance of shell dimorphism in a barnacle. Ecology 67:858-864.

LYNCH, M., AND R. ENNIS. 1983. Resource availability, maternal effects, and longevity. Exp. Gerontol. 18:147-165.

PAREJKO, K., AND S. I. DODSON. 1991. The evolutionary ecology of an anti-predator reaction norm: Daphnia pulex and Chaoborus americanus. Evolution 62:1311-1324.

ROLLO, C. D. 1995. Phenotypes: Their epigenetics, ecology and evolution. Chapman and Hall, London.

SCHEINER, S., AND R. F. LYMAN. 1991. The genetics of phenotypic plasticity II. Response to selection. Evol. Biol. 4:23-50.

SCHLICHT1NG, C. D. 1986. The evolution of phenotypic plasticity in plans. Annu. Rev. Ecol. Syst. 17:667-693.

-----. 1989. Phenotypic plasticity in Phlox .II. Plasticity of character correlations. Oecologia 78:496-501.

SCHMALHAUSEN, I. I. 1949. Factors of evolution: the theory of stabilizing selection. Blakiston, Philadelphia, PA.

SPITZE, K. 1992. Predator-mediated plasticity of prey life history and morphology: Chaoborus americanus predation on Daphnia pulex. Am. Nat. 139:229-247.

STEARNS, S. C. 1989. The evolutionary significance of phenotypic plasticity. BioScience 39:436-447.

STEARNS, S. C., AND J. C. KOELLA. 1986. the evolution of phenotypic plasticity in life-history traits: predictions of reaction norms for age and size at maturity. Evolution 40:893-913.

TOLLRIAN, R. 1995. Predator-induced morphological defenses: costs, life history shifts and maternal effects in Daphnia pulex. Ecology 76:1691-1705.

VIA, S. 1987. Genetic constraints on the evolution of phenotypic plasticity. Pp. 47-71 in V. Loeschcke, Genetics constraints on adaptive evolution. Springer-Verlag, New York.

-----. 1991. The genetic structure of host plant adaptation in a spatial patchwork: demographic variability among reciprocally transplanted pea aphid clones. Evolution 45:827-853.

-----. 1994. Population structure and local adaptation in a clonal herbivore. Pp. 58-85 in L. A. Real, ed. Ecological genetics. Princeton Univ. Press, Princeton, NJ.

VIA, S., AND R. LANDE. 1985. Genotype environment interaction and the evolution of phenotypic plasticity. Evolution 39:505-523.

VIA, S., R. GOMULKIEWlCZ, G. DE JONG, S. SCHEINER, C. D. SCHLICHTING, AND P. H. VAN TIENDEREN. 1995. Adaptive phenotypic plasticity: consenses and controversy. Trends Ecol. Evol. 5:211-217.

VUORINEN, I., M. KETOLA, AND M. WALLS. 1988. Defensive spine formation in Daphnia pulex and induction by Chaoborus crystallinus. Limnol. Oceanogr. 34:245-248.

WADDINGTON, C. H. 1959. Canalization of development and genetic assimilation of an acquired character. Nature 183:1654-1655.

WHITHAM, T., AND S. SLOBODCHIKOFF. 1981. Somatic mutation in plants. Oecologia 49:287-292.

WOLTERECK, R. 1909. Weitere experimentelle Untersuchungen uber Artveranderung, speziell uber das Wesen quantitaver Artunterschiede bei Daphniden. Verh. Dtsch. Zool. Ges. 1909:110-172.

YOSHIOKA, P. M. 1982. Predator-induced polymorphism in the bryozoan Membranipora membranacea. Exp. Mar. Biol. Ecol. 61:233-242.

YUND, P. O. 1991. Natural selection on hydroid colony morphology by intraspecific competition. Evolution 45:1564-1574.
COPYRIGHT 1998 Society for the Study of Evolution
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Harvell, C. Drew
Date:Feb 1, 1998
Previous Article:Evidence from the fossil record of an antipredatory exaptation: conchiolin layers in corbulid bivalves.
Next Article:Evidence of a biogeographic break between populations of a high dispersal starfish: congruent regions within the Indo-west Pacific defined by color...

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |