Lecture by Keith Jackson
BIOL 606 Session, University of Alberta, March 10, 1999.
Rapporteur: Grant McIntyre
The lecture concerned 'heritable novelties which spring from phylogenetic and developmental constraints and serve as vehicles for carrying taxa across valleys in adaptive landscapes (Patterson 1988). Evolutionary novelties under discussion are those leading, or thought to lead, to adaptive radiation of the taxa in which they arise. Two reasons for studying innovations, or evolutionary novelties, is that they offer potential explanations for the design of organisms and they may be related to evolutionary success. The lecture emphasized the latter aspect and was primarily concerned with issues surrounding the identification of true adaptive novelties.
Early debates about novelties concerned the teleology of novelties (whether a drive for innovation exists) and the rate of novelty evolution, whether it is gradual of saltatory. The teleological aspect has been generally dismissed and the rate of novelty evolution remains as one part of the existing debate about gradualism. Current debates include how novelties might arise when the evolutionary trend towards simplification is so strong, what the best way is to study novelties, and how novelties should be identified and treated. It was stated that combining several diverse approaches is necessary for an understanding of novelties.
A discussion of the identification of novelties began with the statement that novelties might only be as novel as their explanation and that apparent novelties are sometimes really just exaggeration of existing traits and patterns. Apparent novelties are often combinations of traits that develop sequentially in a lineage, e.g., flight and associated traits in birds. The identification of novelties is further confounded by a tendency to arbitrarily select groups of different taxonomic rank and assume equivalence. Selecting diverse taxa and looking for novelties to explain their diversity is not the best way to identify novelties.
Cracraft (1990) developed 4 criteria for identifying novelties. 1) Clear differentiation of smallest taxonomic division of interest. 2) Exhaustive sampling of evolutionary innovations. 3) Working phylogenetic hypothesis. 4) Ability to account for all taxa within a clade (i.e. no extinction). These criteria were used to define a novel coloration pattern within the Pionopsitta parrots, but they are probably only appropriate in a limited number of situations.
Another method for identifying novelties is to look for multiple occurrences of similar traits. The development of specialized pharyngeal teeth in two groups of teleost, the Labroidei and the Exocoetoidei is one such example. In both cases the groups with this trait were more speciose than sister groups lacking the trait, suggesting that the development of specialized pharyngeal teeth is associated with adaptive radiation.
Methods for testing for the effect of a trait on species abundance within a clade have been developed where the trait has developed in multiple clades. Mitter et al. (1988) compared species number between primitively phytophagous insect taxa and their sister taxa. In 11 out of 13 comparisons the phytophagous group had greater diversity: this difference is significant using a one-tailed sign test, supporting the notion that phytophagy is an adaptive novelty. In another example Lydeard (1993) compared viviparous actinopterygian fish with their oviparous sister taxa. In most cases the viviparous taxa contained more species. An unbiased proportion of members in the oviparous category was calculated as (#oviparous/(# in both taxa minus1)) and tested for departure from the null hypothesis using Fishers exact test. The null hypothesis was not clearly defined during the lecture.
The idea of internal consistency was raised. If a specific novelty leads to adaptive radiation then all taxa possessing the novelty should be relatively speciose. Departure from this ideal was illustrated with the Halecostomi fish. All groups possess a putative adaptive novelty, the interoperculum, but the taxa within the group contain from 1 to > 12,000 species.
The lecture concluded with introduction of the idea that a novelty either increases rate of evolution of decreases the rate of extinction.
Discussant: Chris Cameron
At two separate points the question was posed: is number of species in a taxon a character in itself or merely reflective of traits contributing to speciation rate? No consensus was reached but it was suggested that taxon number might be a useful or at least interesting character for constructing phylogenies.
Cameron suggested that inconsistency of species richness among taxa possessing a novelty might be explained by other differences in the biology of the groupsm possibly reproductive rates or generation times. Wilson pointed out that these traits did not differ in the fish examples presented during the lecture.
As a further example of a possible novelty, the various auditory structures of fish were described. It was suggested that most groups with the Weberian apparatus were species rich.
As discussant, Cameron presented ample background material on the Platyhelminthes suggesting that relationships within the phylum are not well established, and the phylum may be polyphyletic. He stated that the flatworms are far more interesting than the focal paper and they are a very poor choice of group for investigating evolutionary novelties.
A discussion concerning the nature of homology occurred. Surprisingly, no universally satisfactory definition was arrived at, although several of the major issues were raised. The consensus was that the criteria in the focal paper were quite good for assessing homology and it was later pointed out that the paper had more to do with homology and less to do with novelties than the title and abstract led one to believe.
McIntyre suggested that the loss of novelties should be included in the multiple testing methods for assessing the adaptiveness of novelties. This could be incorporated into the insect phytophagy example from the lecture. Cameron pointed out that gain or loss of a trait in different or even the same taxa was not a true replication of the event since it occurred in a different time and situation.
Strobeck raised the point that species number may not adequately describe success and that speciation rates are related to the method of reproduction. Some very successful groups have few species containing large numbers of individuals (e.g., pines are widespread, not speciose and wind pollinate over large distances). Other successful groups have many species that are relatively less abundant (e.g., orchids have a method for rapid speciation&endash;insect pollination&endash; but are not a dominant member of most ecosystems).
Palmer suggested that a plot of # of descendant taxa versus the ratio (ratio only after discussion with Strobeck) of the number of synapomorphies should reveal whether novelties are more often single characters or suites of characters or if there is no clear relationship between the two variables.