This study represents a preliminary analysis of variation in size and shape in a large and diverse molluscan clade, the Strombina group, over the past 20 million years. Restored height and width were measured on 5,099 individuals of 72 species and 5 genera. Size was estimated by adding height and width, and shape was estimated by dividing height by width. Patterns of variation were analyzed quantitatively among species, subgenera, and genera using univariate statistical tests comparing means and variances and linear regression.
Results of univariate tests show that both shape and size vary among species within subgenera (the specific level) and among genera (the generic level). However, only shape varies among subgenera within genera (the subgeneric level). Regression analyses for each species show that the relationship between height and width is linear, indicating that growth is never allometric with respect to these characters. Because of this nonallometric growth, rates of shape relative to size change can never vary, imposing a severe constraint on shape change and, in turn, shape evolution. Regression models for species within subgenera have equal slopes but differ slightly in intercept. Subgeneric models differ more in intercept. Generic level models differ in slope. These results suggest that formation of species within subgenera primarily involves extension or contraction of trajectories between height and width within species (=static vectors), resulting in size change without shape change. Shape change is more important in the evolution of higher categories.
To examine overall morphologic change in the clade through time, mean sizes and shapes of species were analyzed using nonparametric statistics. Only a slight tendency exists within the clade for increase in species size, and this tendency is best expressed within two species-rich subgenera having long stratigraphic ranges. No directional trends exist for change in species shape. No relationship is found between species size and probability of speciation or extinction, or between species size and species duration, thus negating the role of species selection. Nevertheless, species are smaller in size in the northern Caribbean, a region characterized by extreme fluctuations in temperature. Larger species occurred only after a taxonomic radiation in the eastern Pacific, a more restricted region possibly characterized by reduced environmental disturbance. During this radiation, normal patterns of morphologic change associated with speciation appear to have been disrupted.