Barriers to Making Order out of Chaos: species with fuzzy boundaries

Chaos is a force that humans have been fighting from antiquity. This conflict is probably best embodied in the early Greek myths of the struggle between Apollo and Python, where Apollo, god of reason and logic, battles the serpent offspring of the goddess of Chaos. Sometimes Chaos loses, sometimes Apollo. We humans will never destroy Chaos, finally dotting the last i and crossing the last t of the ultimate and final scientific grand treatise of everything, but we must try to hold it at bay. Chaos in the basement of the UNM biology department exists as fuzzy species boundaries. When you are down there identifying organisms, sometimes it is unclear just where one species starts and the other ends, as one species seems to grade into the next. This lack of clear distinction is what I call fuzzy species boundaries.

I study the identification and classification of plants. I am a taxonomist. There is an implied assumption that a good taxonomist doing good work with good characters will find clean boundaries between species. It is said that if you fail to find these clean boundaries, the characters, the study, and the personal integrity of the taxonomist is open to question. I labored under that illusion for 4 years doing my masters thesis on a very messy genus. Finally, instead of seeking psychotherapy, I decided that the lack of good species boundaries was not a personal failure, but instead a phenomena of interest in itself. I have been studying this issue for the past 2 years and I am working towards a model for the evolution of species complexes. This presentation is a summary of my understandings and my research goals.

"... fuzziness reflects the way the word is, rather than a lack of conceptual clarity." Baum & Donoghue, 1995.

The purpose of this webpage is to give you a sense that there is a complex interaction of genes, time, and environment that causes unclear, or fuzzy species boundaries. You may end up agreeing with me that there is no apriori reason to assume species boundaries should be discrete. This page is also, in effect, a cathartic expression of the frustration that comes with being required to pretend that biology is simple. Unfortunately nature is rarely tidy.

PRELIMINARY AND USEFUL DEFINITIONS:

Don't call a clade a "species complex"!

• COENOSPECIES: n. a conceptual entity that encompasses the total ecological potential of the species.
• ECOGENESIS: n. differentiation by selection in sister groups.
• ECOPHENE: n. a phenotypic response to a particular environment.
• ECOSPECIES: n. the ecological potential of the species which is realized in nature, a subset of the coenospecies.
• ECOTYPE: n. a race within an ecospecies.
• LINEAGE SHATTERING: n. non-cladogenic speciation occurring when isolated populations become fixed independently.
• ORDINATION: n. a graphic technique to explore patterns in data matrices.
• PHYLOGENESIS: n. differentiation between similar populations for reasons of separate ancestry.
• SPECIES COMPLEX: n. an array of populations which seem sufficiently distinct to recognize as species, but which, because of incongruence between variation in multiple characters, cannot be efficiently categorized into taxa.
• For other useful phylogentic terms see UBC Useful Term Page.

Who has fuzzy boundaries?

There are many taxa that fall within the limits of being a species complex. The whole list is quite long, but here is a short representative list, showing it occurs across the whole biological kingdom.

Recognizing and Classifying Fuzziness

As scientists we must always promote our own research interests. Promoting a outright confrontation with chaos brings up a basic psychological issue, wherein anything messy, ugly, and seemingly intractable to research is best denied existence, or at least relegated to be insignificant. Species complexes are undeniably common. Evolutionary theory has to eventually be able to explain these kinds of complex species as well as ones with tidy boundaries. As long as scientists assume that tidy boundaries are the modus operandi of nature, we will only have a partial understanding of evolution.

There are reasons that thinking about variability is an inherently messy problem. The first is that variation is tricky, and its statistics are not well developed. Lets check some our own assumptions about variability; formulate and answer then click on the question.

• If things are morphologically distinct can we assume they are genetically distinct?
• And is the reverse true, if something is morphologically similar is it genetically similar?
• If something is very variable morphologically, is it variable genetically?

Well...so does fuzziness in phenotypes mean fuzziness in genotypes??

SYSTEMS ANALYSIS APPROACH:

According to my conceptualization of species complexes there are there interelated factors that determine genome boundary sharpness.

• Germplasm
• Environment
• Historical Click here to see why I say it is a matter of "could, should and did".

• Models leading to a fuzzy pattern:
  Here are five models of the processes leading from the system to fuzziness. There is no body of literature on species complexes per se, so I developed this list from my readings of studies of individual species complexes. Each of these models requires certain relative contributions from each of the system elements. Each model is presented on it's own page.

Early stage divergence

Adaptation by maintenance of variation

Environmentally driven convergence

Lineage Shattering

Variable Refugia or The Melting Snowpatch Model

Peripheral Isolation

• Summary
  A pattern of variation in taxonomic groups is designated as a species complex when ever the geographic variation suggests that populations differ in multiple characteristics, but the patterns of differentiation in different characters are not well correlated. This lack of correlation between patterns of variation causes ambiguity as to the appropriate boundaries between taxa within the species complex. The result is that the circumscription of species within a complex poses insoluable taxonomic difficulty for the monographer.
  There has historically been an assumption that the diversifying processes would lead to correlated patterns of variation given that good characters with strong genetic components are studied and that the groups under study were indeed separate evolutionary units. However, direct access to quanitification of genetic variation by molecular methods has shown that evolutionary processes do not always lead to dichotomous phylogenies of a genome as a whole. As a result of the tokogenetic processes within taxa, different genes within the same organism may have very different phylogenies. The result of this knowledge is that taxonomists have recognized that many taxonomic groups are inherently fuzzy in the sense of not having clear cut boundaries between taxa simply because of their genetic organization.
  Why are some taxa more predisposed to show these complex patterns? The historical most parsimonious assumption is that this pattern neccessarily indicated the recency of divergence between lineages. A species complex was neccessarily a youthful entity which simply had not had time to undergo a full differentiation between lineages. To challenge this assumption of simple early stage phylogenesis, I have presented four alternative models of how species complexes may be formed.
  The first alternative model suspends phylogenesis indefinitely by maintaining variation. Species complexes may be closer to the point of bifurcation within a phylogeny, but phylogenesis does not occur at a particular rate or even neccessarily occur at all, the amount of divergence between lineages is not strictly a function of time. At the phylogenetic bifuraction point, the populations can be assumed to be transitioning from tokogenetic relationships where reticulation binds the individual alleles together into a common gene pool and phylogenetic relationships where the gene pools are separate. Species complexes might be transiently at this stage of evolutionary divergence, or, alternatively, might have adaptive strategies to keep near that point. Reticulation by hybridization is a process which would keep the genome at a point of maximum genetic variation. Compilospecies are an example of groups where that strategy is evidently favored. One expectation of this model is that it would be favored where the environment is highly variable or under some kind of destabilizing dynamics. Outcrossing and lack of reproductive barriers will also predispose a taxon to this scenario.
  The second major alternative explanation of why some groups form species complexes calls for ancestral separation of the genomes with an overlaying superficial mask of convergent characters. This would be most likely in a morphologically defined species complex and would arise from processes that counter diversification between lineages without the capability of hybridization. If the selective forces are such that some optimal phenotype or adaptive peak among many lineages exists then any centrifugal diversifying trends will be balanced by centripedal convergent selection. This model would be favored under highly selective situations. In my own research, the Cirsium arizonicum species complex may be convergent upon characters favored by hummingbird pollination. Homoplasy can be recognized by referencing patterns in selectively neutral characters.
  The third scenario under which species complexes are more than early stage bifurcation products is when a large variable parental taxon gives rise to multiple variable descendant taxa through a process of lineage shattering producing a polytomy. This model might be the result of the derivitives being in fragmented populations, more or less simultaneously, or might be the result multiple derivation events from the same parental taxon over the course of time. This model is probably more realistic than a simplistic model of symmetrical division of the ancestral taxon into sister taxa.
  My contention is that biology is not simple, and because of the temporality of human life we must work within the world as it exists here and now. Our theories of evolution must address all patterns resulting from evolutionary processes, not just the simplest and most computationally tractable. The study of species complexes forces the issue of addressing the variation inherent to the evolutionary process itself and may facilitate an integration of evolutionary theory.
  Although we can generally correlate patterns and processes, species complexes give us only a few hints as to the mechanics of their origins. Click here to read about these patterns.
  How this pertains to my research is a paragraph yet to be written. Check back in June! Meanwhile I have some advice for those who work with messy taxa.

Keys to the Gates of Chaos

Keeping your sense of the Cosmos

• Annotated Bibliography
  List of References Please take a moment to send me your reaction to this website. Several changes are already in the works, but some feedback would be very useful.

Cite this website as follows:

Barlow-Irick, Patricia. 1997. Barriers to making order out of chaos: fuzzy species boundaries. Version date, Website http://www.edu.unm/~pbarlow/spcomp.html.

If you can cite a personal communication, why not a website?