Variably colored sea urchins Lytechinus variegatus are common throughout the western Atlantic and Caribbean, from Beaufort, North Carolina to Brazil. Urchin spine color is white, green, purple, red, and pink and is often a combination of two or more colors. Test color too is highly variable and can be one or more colors and distinctly patterned. Sampling across the range reveals geographic differences in coloration with local phenotypes.  

Here I present data from laboratory crosses designed to examine the inheritance of color phenotype. I created a series of crosses with urchins from Beaufort, NC and Tavernier Key in the Florida Keys. Urchins of similar and differing coloration were crossed. Three visible color traits were examined in the F₁ offspring: spine phenotype, test phenotype and patterning. Since parental urchins were of unknown genotype, I assumed a simple one-locus Mendelian inheritance for all three traits. The observed phenotypic frequencies in the F₁ generation were tested against expected Mendelian frequencies for one-locus traits (i.e.  0:1, 1:1 or 3:1 phenotypic ratios for each of the three traits). Deviations of the observed color ratios were tested with a chi-square test. From these ratios, parental P₁ genotypes were deduced.

Results/Conclusions

Spine color, test color and patterning are all heritable traits. There is no evidence for linkage among these three traits. Most of the crosses fit within expected Mendelian ratios based on observed spine and test color frequencies. White is dominant to purple and green. In many individuals the white parent is a heterozygote. Purple is recessive. Purple and green are co-expressed in offspring and demonstrate strict spatial positioning – green is always basal to purple on the spines. In these crosses there are three spine phenotypes but there could well be others. Dark tests are dominant over light tests.

Color variability in L. variegatus is a heritable trait and potentially subject to natural selection. The nonrandom geographic distribution of phenotypes seen across the range suggest several things: long distance dispersal may not be as extensive as it is assumed given the pelagic larval life stage; long distance dispersal does occur but there may be selection on local phenotypes such that novel phenotypes are eliminated and do not contribute to the population; or there is selection on trait/s unrelated to color differentiation.

The ambiguity in the mechanism responsible for the geographic variability needs further examination. The ecological significance of color variation is also poorly understood.