In biology the "species" has occupied its particular position in the theory and practice of systematic and evolutionary biology. "The species problem" is in fact double-layered: species taxon and species category. Many biologists and philosophers of biologists, including Ernst Mayr, Michael Ghiselin, David Hull and others, have extensively discussed the species problem during the last half century. However, until now there is no single definition of species; we have more than twenty alternative species concepts (morphological, biological, phylogenetic, evolutionary, ecological, etc.). All these species concepts are species categories which are defined by particular criteria (reproductive isolation, monophyletic relationship, morphological similarity, etc.). Any species category can be explicitly defined as a set with its members, that is, species taxa. From the ontological viewpoint "the species taxon problem" and "the species category problem" must be discussed separately because species taxa are much more difficult to be defined than species categories. Recent developments in large-scale biodiversity studies over the Earth call once again our attention to the definitions and meanings of species taxa and species categories in biological systematics. For example, conservation programs for engendered species depend on which species concept is adopted. Here we must recognize that the species problem has its conceptual roots not only in biology but also in metaphysics and cognitive psychology. Phylogenetic reconstruction in general aims at estimating the most plausible tree based on character data (molecular or morphological). Such historical reasoning is based on "abduction", that is, a form of non-deductive inference to the best hypothesis for a given data. Phylogenetic abduction does not require the species concept, but the unique existence of The Tree of Life. Taxonomy is not the only tool for ordering our knowledge of fauna and flora. Still, "The species problem" is always with human being whether it will be resolved in the future or not. To classify is human.
One of the most fundamental questions in biology is the reason why organisms in any habitat fall into discrete, nonoverlapping groups. This was one of the key questions that Charles Darwin asked in the Origin of Species, but his theory, of course extraordinarily important in changing our view of life, was unsuccessful in explaining how speciation occurs and how continuum of potential genetic variations breaks into the distinct clusters called species. More simplified but essentially the same question asked in ecology was how closely can species be packed in a natural environment. The conventional wisdom in theoretical ecology is that there is no limit for the number of species to be packed in (MacArthur 1970, Roughgarden 1972), unless there are random extinctions by stochastic environmental disturbance (May 1973). Here I reexamine the MacArthur (1970)'s model of species packing in one dimensional niche space, by slightly relaxing the special assumptions of the previous models in the functional forms with which the resources the competing species utilizes distribute themselves over a niche space. I find that the "evolutionarily stable" community that is reached after repeated events of invasions and extinctions of species, generically consists of a discrete set of species, contrary to the MacArthur (1970)'s conjecture. The discreteness arises purely as a consequence of competition for common resources between phenotypically close species, and does not rely on any form of disadvantageous hybridization that Ernst Mayr proposed, or stochastic fluctuation of environment that Robert May emphasized.
The systemic concept of species is proposed. This is an integrated concept and resolve the so-called species problem. An elucidation for our taxonomic activities is also presented. "A species" has several meanings or are used in several ways: (1) a class, (2) a collection of organisms, (3) a (material) system composed of organisms. I define that a species is (4) a (mechanismic) system producing the organisms of one kind. When used in biological classification, a species taxon functions as a class. Due to the atemporal identity of a class, we can measure the biological world, especially the basic biological entities: organisms. In the identification at the species level, species taxa (e.g., _Homo sapiens_) are values of nominal (=categorical) variable "species". So, species as a taxonomic category (rank) is a measuring unit. Grouping of organisms is possible, because we observe that organisms have a varied numbers of characters (features) in common. Further, the latter is possible, because a mechanismic system produces the organisms which are considered the members of a species taxon (identification). The discretenesss of species systems make the between-species discreteness of character distribution of organisms. The mechanismic system of a species here defined does not composed of organisms, but of mechanisms. Such mechanisms are universified through a (multilevel) control system. By the deeper knowledge of developmental systems, we will be able to represent mechanismic models of species (at least diagnostic indicators for species systems). Organism-producing mechanisms indicate to construct equivalent classes: species taxa. They are the objective and natural classification at the species level.
Although the modern concept of recapitulation stems from the Biogenetic Law by Haeckel, it is generally understood now as an integration of von Baer's and Haeckel's ideas. Namely, the embryogenesis of an animal, in an idealized condition, is expected to follow a sequence, in which each stage represents a pattern shared by progressively and hierarchically specified taxa, such as class, order, family and genus, not the patterns of adult ancestral animals. Unlike von Baer, who did not believed evolution and assumed independent archetypal patterns for four major animal groups, modern Evo-Devo supports the hypothetical ancestral pattern of bilaterians (urbilateria), consisting of their synplesiomorphic traits. Further basally in the phylogenetic tree, Haeckel has assumed a hypothetical animal, Gastraea, possessing only two germ layers, as an ancestral morphological pattern for all the metazoans. We are more familiar with the concepts of 'phylotypes', represented by a conservative embryonic morphology among a phylum, explaining the origin of global morphological homologies within that taxon. Thus, the morphological homology is tightly linked with the phylogenetic evolution of animals as well as archetypes, in the patterns of developing embryos. The dogmatic and transcendental nature of the recapitulation, however, should be eliminated in modern Evo-Devo studies. Typically, the morphological novelty such as jaws in gnathostomes and turtle shells are not necessarily patterned in the developmental stages expected from their evolutionary appearance, but are rather patterned through radical modification of developmental programs that eliminates the conservative pattern recognized as the archetype for that taxon (vertebrates, arcchosaurians). Especially the turtle shell is known to exhibit reversed topographical pattern between scapula and ribs, eliminating the morphological homology defined by Geoffroy's rule of connectivity. Taking the turtle shell as an example, it will be shown that turtle-specific development takes place after the stage of vertebrate phylotype but long before the stage of archosaurian-archetype; some generalized amniotic anatomical patterns are conserved throughout turtle development, but others are eliminated by their embryonic remodelling. It will also be shown that the latter difference corresponds to the specific developmental program at cellular and molecular levels. Evolutionary remodelling of developmental program thus appears to be biased through the genetic and developmental network, rather than simply following the hierarchical course of phylogenetic evolution.