Inferring the causes for change in the fossil record has been a persistent problem in evolutionary biology. Three independent lines of evidence indicate that a lineage of the fossil stickleback fish Gasterosteus doryssus experienced directional natural selection for reduction of armor. Nonetheless, application to this lineage of three methods to infer natural selection in the fossil record could not exclude random process as the cause for armor change. Excluding stabilizing selection and genetic drift as the mechanisms for biostratigraphic patterns in the fossil record when directional natural selection was the actual cause is very difficult. Biostratigraphic sequences with extremely fine temporal resolution among samples and other favorable properties must be used to infer directional selection in the fossil record.

A stickleback with brilliant white dorsal breeding colours is widely distributed in north-eastern Nova Scotia, Canada, where it often breeds sympatrically with the threespine stickleback, Gasterosteus aculeatus. Breeding males are highly conspicuous and visible at distances of 20 m or more whereas sympatric G. aculeatus are cryptic and difficult to detect even at 2 m. The white stickleback nests only above the substrate in filamentous algae, whereas G. aculeatus nests only on the substrate. The white stickleback is smaller in size and more terete than G. aculeatus, but it is morphologically similar in having a complete row of lateral plates and similar lateral plate and gill raker numbers. The white stickleback occurs only in environments where there are filamentous algae (which appears to be an obligatory nesting substrate) and where the water is clear, saline and relatively still. Female choice tests in the laboratory show that the white stickleback is reproductively isolated from G. aculeatus, and field observations on natural spawning support this conclusion. We suggest that the bright breeding colouration may have evolved through sexual selection and/or to advertise unprofitability to predators.

Widely distributed freshwater populations of Gasterosteus aculeatus exhibit reduction in the bones of the pelvis and the numbers of lateral plates and dorsal spines. We investigated the relationship between skeletal reduction and the characteristics of the environments in which it occurs. A combination of environmental factors, including environmental ion composition – lake productivity, geographic position, opportunity for gene flow, and the presence of other fishes, is associated with skeletal reduction. The principal correlates of reduction in all skeletal traits are low concentrations of calcium, magnesium, silicon, H+, and reactive phosphorus. We hypothesize that the ion composition of lake water is the main selective agent promoting skeletal reduction in sticklebacks in the Cook Inlet area, Alaska, and that other fishes and local gene flow may modify its extent. Our results show that a suite of skeletal traits responds in common to the interacting effects of at least three environmental factors (ion composition of lake water, presence of other fishes, local gene flow), and they emphasize that unitary explanations of the evolution of skeletal reduction are unlikely to be adequate.

The pelvis of Pungitius pungitius (ninespine stickleback) is normally a robust, bilaterally symmetrical structure with stout spines. In some populations it is reduced in size, sometimes to complete absence. The first sign of reduction is loss of the spines, which is accompanied by a reduction in the size of the pelvic bones (vestiges) and an increase in bilateral asymmetry. Further reduction in the size of the vestiges leads to their eventual loss. In a pelvis-reduced population the propensity to spine asymmetry is highly heritable (h2 = 0.85 0.14) but asymmetry of the bony vestiges in fish without spines is not heritable. The absence of spines is heritable, as is size of the bony vestiges (h2 0.49 0.12). Crosses of spineless fish from a reduced population with spined fish from either of two normal populations produce only spined progeny. The inter-population heritability of pelvis size is h2 =0.26 0.10, but that of bilateral asymmetry is zero. The expression of spines and vestiges is influenced relatively little by variation in pH, calcium, and salinity. We propose a polygenic model of inheritance with two phenotypic thresholds, the upper for the presence of spines and the lower for complete absence of the pelvis. Canalization breaks down between the thresholds. This genetic system probably predates divergence of the stickleback genera. We discuss its relevance to understanding the dynamics of pelvis reduction in extant and fossil stickleback populations.

SUMMARYApeltes quadracus is polymorphic for the number of dorsal spines, the variation is heritable, and it is subject to natural selection. Here we investigate selective predation favouring the higher-spined morphs. We predict that populations with greater mean spine number will have longer spines, and we so find. However, the morphs do not differ for spine length within populations, and we conclude that spine number and spine length are selected independently. Populations with greater mean spine number also have longer and deeper bodies, thus having a greater defensive circumference. Again, the morphs do not differ within populations.Experimental tests demonstrate selective predation on the morphs for four fish species in the laboratory, and for a fish and a bird species in the field.However the experiments also show that: (a) the higher-spined morphs are not always favoured (b) vegetation plays a role in selective predation (c) factors other than spine number are involved (d) these factors operate before Apeltes is captured (behavioural/ecological) (e) the factors are correlated with spine number. We conclude that predators are selective agents on the polymorphism, but that the relationship between fitness of the morphs and predators is complex.