Our data suggest that P. violaceum might contain cryptic species. There are 6 groups (A, B, C, D, E, and F) with the most basal split being between the groups A/B and C/D/E/F. Though the genetic distances between these two main groups are relatively small (0.010-0.021), they are larger than the differences between accepted species in the dictyostelids using the same part of the 17S sequence, for example between D. citrinum (OH494) and D. dimigraformum (AR5b), 0.009; D. clavatum (TNS-C-189) and D. longosporum (TNS-C-109), 0.003; D. mucoroides (TNS-C-114) and D. sphaerocephalum (GR11), 0.001 or between D. brunneum (WS700) and D. giganteum (WS589) 0.004 ). The 17S to 5.8S sequence is relatively conserved compared to other genes in a wide variety of organisms (for example [42–44]). This sequence has enough resolution to distinguish between sister taxa in the dictyostelids [16, 27].
Furthermore, the variation between geographical populations accounts for less than 10% of the total variation. Many haplotypes are found in more than one geographic population. In addition, group B is cosmopolitan, with individuals from almost every geographic location belonging to this group. This suggests that the population structure in P. violaceum is not due to geographic constraints alone. The major differences that we see in the 17S seem for the most part consistent with species-level differences, with the species often occurring in the same areas.
By and large, the division between the two groups was reinforced by our mating experiments. Like Clark [5, 11], we found two groups of non-interbreeding individuals; however, we observed a few instances of mating between the groups. This leaves open the possibility for some gene exchange between different groups should those macrocysts actually be able to germinate. Unlike the model organism, D. discoideum, P. violaceum reported germination rates have been upwards of 50% . Examining the germination of our between-group macrocysts, which would require prolonged ageing of macrocysts , would be a fruitful line of research for the future.
The split between the two groups is also apparent when looking at cooperation during fruiting body formation. We have found that only clones from group B exhibit strong mixing and cooperation with other group B clones in forming the fruiting body. When the clones were from different phylogenetic groups sorting was more complete. Both the phylogenetic diversity and the behavioral changes suggest that there may be at least two different morphologically identical sister species in P. violaceum. Both lines of evidence are consistent with the same division and the variation that we observed in phylogenetic structure affects the behavior that we observe in the social stage.
Relatedness allows altruism to be beneficial if the altruistic acts are directed towards relatives. Because clones from two different species are not related, there should be selection for species discrimination. Because we are unsure of the exact nature of the relationship between individual clones, we use the term kin discrimination rather than species discrimination. In D. discoideum, the further the genetic distance between clones, the greater the propensity for kin discrimination to occur . In our study, a few clones cooperated with each other to form chimeric fruiting bodies, but most clones tested sorted out to form mostly clonal fruiting bodies. All the clones that cooperate with each other were in the same group (B). This fits with the idea of kin selection, with only closely related clones cooperating, though we do not have information on exact values for the other half of kin selection: the relative costs and benefits of cooperation. Benefits of larger groups are likely to include lower proportions of cells destined for stalk relative to spore, and ability to move greater distances, while costs center on becoming a sterile stalk cell.
Previous studies on kin discrimination in the dictyostelids have given differing results depending on the species used. Kin discrimination has also been investigated in Dictyostelium discoideum, D. purpureum and D. giganteum. Clones of D. discoideum exhibit kin discrimination with more distantly related clones sorting more than clones that are more closely related . D. purpureum shows kin discrimination as well . In D. giganteum, some genetically distinct clones exhibit kin discrimination while others do not . The question of whether D. giganteum is one species worldwide with varying levels of kin discrimination or multiple cryptic species has not been resolved, but North American clones show no differentiation . Our results show that P. violaceum exhibits kin discrimination; like the other dictyostelids, clones from different cryptic groups within P. violaceum sort to form clonal fruiting bodies while closely related clones sometimes cooperate to form chimeric fruiting bodies.
Most of the dictyostelids have been identified and distinguished on the basis of morphology. An exception is recent work on Polysphondylium pallidum and its sister species P. album  as well as D. ibericum . Romeralo, Baldauf, and Cavender  used morphology to identify a new species, and molecular phylogenetics to place that species within the dictyostelids. Kawakami and Hagiwara  use a combination of mating type and morphological characters to redefine these two species. They show that there are three groups, one that matches the P. pallidum type specimen and mates with P. pallidum strains, one that matches the P. album type specimen and mates with P. album strains, and one that matches neither exactly and mates with neither. The relationship of the third group to P. pallidum and P. album remains unclear. These recent studies make it clear that relying on morphology alone to dictate species boundaries is not sufficient, and mating type analysis and molecular work is needed to correctly identify species boundaries and the relationships between species.
Improperly identifying cryptic species also affects biodiversity metrics as well as estimates of geographical distributions. By identifying all members of a cryptic species complex as the same species, biodiversity is underestimated and geographic distributions are overestimated. In the identification of protists this can be especially difficult because of a lack of distinguishing morphological characteristics . The difficulty of correctly identifying cryptic species has contributed to debate on protist biogeography. Finlay  suggests that there is something fundamentally different about microorganisms, including protists, such as higher rates of migration and lower rates of speciation that causes them to be more cosmopolitan than larger organisms. Foissner  suggests that more endemic species are present in part because of molecularly distinct but morphologically similar species that are endemic but are classified as a single cosmopolitan species.