Conner, E.F. and E.D. McCoy. 1979. The
statistics and biology of the species-area relationship. American Naturalist
113:791-833.
Blog by Rebbeca Kiat
Paper
Authors: Edward F. Connor, Earl D. McCoy
(Commentary
by Brian A. Maurer)
Brian
A. Maurer
· Ph.D. from
University of Arizona
· Associate Professor at Michigan State
University
“Ecology
is understood to be the study of phenomena that occur on a wide range of
spatial and temporal scales. My interest focuses on the largest spatial and
longest temporal scales studied by ecologists.
To understand the importance of processes at geographical scales, it is
important to understand how they connect to local scale processes. Thus, I am interested in modeling population
and community dynamics in a geographical context. I am working with a variety of vertebrate
organisms to model how population dynamics and abundance vary from one place to
the next within their geographical ranges.
These focal questions, however, give rise to a number of different
ancillary questions regarding such things as the ecological and evolutionary
importance of body size, geographic patterns of species diversity, and resource
use behavior. Modeling ecological
systems in space and time is a major tool that I use to answer questions about
geographic scale ecological processes.
To this end, my lab uses a variety of quantitative and computational
technologies. We intensively use geographical information systems (GIS) to
analyze and model spatial processes.
Spatial statistics and related techniques from geostatistics provide the
analytical framework for many of our statistical analyses. Finally, use of nonlinear and spatially
explicit mathematical modeling techniques allows us to develop theoretical
approaches to large scale ecological systems.”
Edward
F. Connor
· Ph.D. from
Florida State University
· Professor at San Francisco State University
“I am broadly interested in population and
community ecology, statistical ecology, biogeography, insect-plant
interactions, and sampling and monitoring design in ecology and conservation.
My field research employs insects communities and insect-host plant systems to
answer basic questions about the population dynamics of pest insects and the
structure of insect communities. My interests in statistical ecology and
biogeography lie in developing probabilitistic statistical procedures for the
analysis of biogeographic data and in developing a rigorous framework for
hypothesis testing and staitsical inference using both experimental and
non-experimental evidence in ecology. My
work on sampling design focuses on questions concerning the optimal design and
allocation of effort in monitoring plant, bird, and bee populations and
communities.
My recent research has focused on trying to
determine the mechanism by which insects induce pall galls (tumors) and its
evolution, developing hierarchical models for abundance data, and automating
avian population monitoring.”
Earl
D. McCoy
· Ph.D. from Florida State University
· Professor and Associate Chair at University of
South Florida
“My students and I study a broad
range of ecological and biogeographical problems. Many of our projects relate
in some way to conservation biology, either in theory or in practice. Most of
our current research deals with conservation and restoration of severely
threatened upland habitats, particularly sandhill and scrub, in Florida. Within
this framework, my students have focused their projects on a variety of topics:
structure of gopher tortoise populations, demography and autecology of sand
skinks, restoration of Florida mouse populations on lands mined for phosphate,
and comparative biology of common and rare frogs, for example. Other students
have focused their projects on topics such as methods of ecological analysis
and the composition of species' assemblages. My own research encompasses
additional topics in the areas of disturbance ecology, particularly fire
ecology; biogeographical theory; and the philosophical basis of ecology.
Virtually all of
the research being conducted by my students is aimed at solving particular
problems and, therefore, probably would be labeled "applied research"
by many persons.”
Paper
- Summary/Main points:
· Concerned with use and interpretation of
species-area curves.
· Power function model vs. alternative models –
is it the best model? Theoretically unique in explaining observations?
· Biological explanations for these models?
(1) “Does the
equilibrium model provide a unique theoretical basis for the species-area
relationship?”
· Habitat diversity hypothesis - with increased
area, more variety of habitats?
· Area-per se hypothesis - equilibrium theory of
island biogeography?
· Sampling hypothesis – passive sampling;
increased sampling for larger areas?
(2) “Is the
power function model (log/log), derived from equilibrium theory, the best model
of the species-area relationship?”
· Is there a model that statistically fits the
data the best?
· Danger in applying a model without comparison
to other models/ignoring important assumption made when applying a model.
· Analysis of 100 data sets of species-area
curves from literature and re-analyzing using various models – found no
best-fit model.
· Note: did not look at curvilinear models?
· The power function and the untransformed
models provide good fits most frequently.
(3) “Can the
parameters of the power function or other species-area models be interpreted
biologically?”
· Closer look at parameter values for models
(e.g. z , k, log k)
· Slope parameters – predictions and limitations
and what this means
· Published predictions and interpretations at
the time were not supported by available evidence – needed more data/due to the
nature of the claim could not be tested
Questions:
· In this paper, the authors also discuss
Arrhenius’s 1920 & 1921 papers and the equations from them (note previous
reading). Were there limitations to these equations or were there underlying
assumptions?
· Do we have new evidence that might better
explain some of these models now, biologically?
· Why did the authors not look not look at
curvilinear models? Is this relevant?
The best I could say to summarize the paper is that the authors argue the constants in Arrhenius's model are only relevant as parameters in said model and have no biological significance, and that scientists in fields that utilize Arrhenius's equation or its derivatives should be skeptical when applying it to their data sets. Honestly though, this paper felt like it went in so many directions I don't know what to make of it.
ReplyDeleteI feel this paper had many angles and objectives, making it challenging to follow and 'get into'. I'm not sure I understand the significance of the power function or why the authors recommend its continued use if, as the authors suggest, the results obtained when using it should only be viewed as constants and not as biologically significant. If that's the case, that seems to imply that the power function serves as an unnecessary statistical step if its results are generally considered insignificant.
ReplyDelete