Biotic and abiotic controls on evolutionary rates
Biotic and abiotic factors are widely believed to
influence extinction risk and to have influenced the
diversity dynamics of clades. I use the marine fossil
record to test models relating such factors to
evolutionary rates over a range of spatial and temporal
scales.
In my postdoctoral research at Stanford University I have
been primarily focusing on the relative influences of
different forms of rarity on the macroevolution of marine
animals in collaboration with Jonathan Payne (Stanford)
and Carl Simpson (Humboldt University Berlin). Our work
(Harnik et al. in prep.) has shown that extinction
is consistently biased against narrowly-ranging genera.
This pervasive effect did not produce a secular trend
toward broader geographic range size, however, because of
the antagonistic effects of origination.
I have also recently investigated extinction dynamics in
marine bivalves in collaboration with Rowan Lockwood
(College of William & Mary). In our review we
presented new estimates of extinction and origination
rates for marine bivalves through the Phanerozoic that
account for temporal variation in the quality of the
fossil record (Harnik & Lockwood in press). We
then assessed the role of diversity-dependence in
mediating these evolutionary rates and and reviewed the
literature on biological predictors of extinction risk.
Marine ecosystems today are changing rapidly due to a
variety of factors and relatively little is known about
the ecological and evolutionary consequences these changes
will have. With Seth Finnegan (CalTech) and Rowan Lockwood
(William & Mary), I am co-leading an interdisciplinary
working group at the National Evolutionary Synthesis
Center focused on determinants of extinction in anicent
and modern seas.
Macroecological drivers of extinction risk in the early
Cenzoic
For my dissertation research at the University of Chicago
I developed and tested a set of multivariate statistical
models that relate the ecology of bivalve species to their
persistence over the Paleogene (65-28 MYA) in the U.S.
Gulf and Atlantic Coastal Plains. Contrary to the general
assumptions of independence implicit in most models of
extinction selectivity, these analyses show that broadly
distributed species tend to be abundant and larger bodied.
I have used structural equation modeling to examine
simultaneously the direct and indirect contributions of
these factors to species duration in three major lineages
of marine bivalves (the Carditoidea, Pectinoidea, and
Veneroidea). Species-level data for these analyses were
gathered through quantitative field sampling and use of
existing museum collections and literature records. These
analyses indicate that geographic range is the primary
explanatory variable in predicting species duration with
abundance having little direct effect and body size having
opposing direct effects among clades (Harnik
2011).
Abundance and extinction rates
To examine the influence of abundance on extinction rates
globally over the post-Paleozoic, Carl Simpson and I
analyzed data for marine bivalve genera using the
Paleobiology Database. These analyses show that abundance
was an important factor in bivalve extinction rates over
the last 250 million years (Simpson
and Harnik 2009). Yet, surprisingly our results
reveal a persistent non-linear relationship between
abundance and extinction rates which only in part
corroborates general predictions. With increasing
abundance extinction rate declines yet the most abundant
taxa exhibit elevated rates.
Structure of diversity
The distribution of species among genera is markedly
uneven, with most genera species-poor and few
species-rich. This structure to taxonomic diversity may
arise through differential rates of speciation and/or
extinction. However, ecological factors such as
competition and geographic range expansion/contraction
also likely contribute. In collaboration with David
Jablonski, Andrew Krug, and James Valentine, I have used a
global database of extant bivalves to characterize the
taxonomic structure of marine biomes and provinces and
assess the contributions of taxon age and provincial area
to these large-scale diversity distributions (Harnik
et al. 2010).
In this research I have used ecological models, developed
to describe the distribution of individuals among species,
to reveal a general form to the distribution of species
among genera in modern marine faunas, and null models to
show that this spatial variation in taxonomic structure is
not explained solely by variation in species richness.
These diversity distributions in combination with
age-of-first-occurrence data suggest that the taxonomic
structure of regional faunas is shaped by differential
speciation potentials among clades, as constrained by
spatial variation in diversity accommodation space.
Rarity and sampling
One of the principal challenges in assessing extinction
risk today and in the geologic past is that rarity is
believed to influence extinction but is also known to
affect sampling. Rare taxa are, by definition, encountered
infrequently and their observed occurrences strongly
controlled by sampling effort. To minimize this bias, I
have developed new methods for sampling rare species
through the integration of historical data from museum
collections and the published literature and estimates of
species abundance gathered from quantitative field samples
(Harnik
2009).
Combining these two sources of data can provide a more
comprehensive estimate of abundance and taxonomic
diversity without substantial increase in current sampling
effort, thereby expanding the scale of abundance and the
sample size of species that can be included in
paleoecological and evolutionary analyses. Applying these
methods to data I have compiled from the literature for
Paleogene mollusks underscores the magnitude of veiled
diversity in marine fossil assemblages and the potential
of existing sources of data to unveil rare taxa, allowing
them to be incorporated into quantitative diversity
studies.
Student-Scientist Partnerships
Engaging students in research is among the most effective
and compelling ways to teach science. Student-Scientist
Partnerships (SSP) using local paleontological materials
solve the pedagogical problem of how to get students in
touch with real science that is interesting to them,
connects to their lives and prior knowledge, but requires
little background or training to make a contribution to
data acquisition and analysis. While at the
Paleontological Research Institution, my colleagues and I
developed the Devonian Seas Project, an SSP which engaged
upper elementary through high school students and teachers
in paleoecology research using the Devonian marine fossil
record of New York State (Harnik
and Ross 2003, 2004).
Participants were involved in classroom-based research
experiences and field-based professional development
workshops. The goals of the project included specific
educational and scientific results and also the creation
of a model by which other institutions might institute
similar partnerships in the geosciences.
