Brown, J.H. 1984 - On the Relationship between Abundance and Distribution of Species

Author: James H. Brown

Author of Macroecology

Member of the National Academy of Sciences

Professor Emeritus, University of New México

Professor, University of Arizona - 1975-87

Associate Professor, University of Utah - 1971-75

Assistant Professor, UCLA - 1968-1971

Postdoc at UCLA, 1967-68

Ph.D. at University of Michigan, 1967 - Emmet T. Hooper (Student of J. Grinnell)

Research interests: Scaling biology and macroecology, experimental desert ecology 

Blurb author: Christy M. McCain (Same blurb author from S. Anderson 1977)

Definitions:

Abundance: Local population density. Terms for extremes of abundance: rare and common.

Distribution/range:  spatial distribution. Terms: restricted (local) and widespread

Brown proposes a theoretical model to make predictions about the ecological and
evolutionary consequences of niche dimensionality that can be tested by evaluating
further the three assumptions:  relationships among environmental limiting factors,
patterns of abundance and distribution, and degree of evolutionary differentiation in taxa
for which adequate data are available.

The Theory: Brown addresses two main topics or patterns in this paper:

• Spatial Variation in Abundance Within Species 
• Correlation Between Abundance and Distribution Among Species 

- Based on 3 assumptions (the first two relating principally to the first pattern and all three
addressing the second pattern):

1)  “Variations in population density of a species over space is assumed to reflect
the probability density distribution of the required combinations of environmental
variables” → Changes in population density are tied to the likelihood that their
biotic and abiotic niche requirements are available in a certain area. (Hutchinson's
(1957) multidimensional niche for each species)

2)  “ Environmental variation also is autocorrelated” → Tobler's first law of
geography: everything is related to everything else, but near things are more
related than distant things. If this is true, Brown argued, the center of the species
distribution will offer the most optimum environmental conditions, explaining why
the highest abundance of that species is there.

3)  “... closely related, ecologically similar species differ substantially in only one or
a very small number of niche dimensions. This limited differentiation reflects
evolutionary constraints on morphology, physiology, and behavior as a result of
relatively recent descent from a common ancestor.”

Spatial Variation in Abundance Within Species:

Using previous work and data sets of songbirds and plants, Brown shows how population
density across species ranges is going to have its highest frequency in the center of its
range and that frequency is going to decrease towards the edges or boundaries of the
range (Fig 2).

From figure 2. The bottom graph shows mean population density along 4 transects from the center to the edge of the geographic range (the black cross in the top map).

This form of abundance peaks near the geographic center of their distribution and
declines toward the edges, forming a normal curve (Fig 3).

Environmental independent variables added together create different niche dimensions
and will determine different population densities according to their requirements,
generating normal distributions of density. “As a consequence of their multiple niche
requirements, almost all species have restricted distributions; Often this is a response to
a single variable, which may be abiotic or biotic (predator/prey/competitor density).
This explains why some species distributions change abruptly or are patchy, which are
exceptions to the unimodal distribution of population density (fig. 8).

Is the pattern in figure 7 an artifact of sampling (are abundant species just more likely to be

found throughout their range)? Brown takes subsamples of datasets and finds that the answer

is no - increased sample size does not affect predicted range size significantly.

Correlation Between Abundance and Distribution Among Species 

• For closely related and ecologically similar species, the area of the geographic

range is positively correlated with maximum density, average density over the area

where the species occurs, the number of sites inhabited within a local region. 

• This is explained by the multidimensionality of niches. If a species is able to

survive and reach high densities in a place, then the species will be able to

occur on many different sites, but if a species has narrow requirements and

cannot attain high abundances, the species will be restricted to the few sites

(niches) that are able to satisfy its needs. 

• As with Hanski (1982), this model can only compare closely related species

(with assumed-to-be similar niches)

• Mosaic nature of the environment: larger organisms tend to have less dense, more

widely distributed populations, whereas smaller ones attain higher densities and

exhibit more local and patchy distributions

Why he does not agree with Hanski:

Hanski assumes that all sites are equally suitable for all species and that species distributions
among those sites vary randomly over time.

1. it is highly unlikely that all sites are equally favorable for all species (species different

requirements and tolerances = difference in distributions.

2. Hanski assumes that rare species rely on colonization in order for populations to persist.

Brown argues that this is unrealistic. Not all rare species have the capability for the

dispersal that would be needed to maintain small, isolated populations, and the more

distant populations are the less likely they are to have suitable habitat between them.

3. “... bimodal distribution of frequency of occurrence is almost certainly an artifact of
sampling a small number of sites with a local region.” The bimodal distribution of species

abundance becomes unimodal if the scale is expanded beyond a small subset of the

species’ total range.

General Discussion

Brown viewed this as a qualitative theory that would serve as the framework for future
statistical analyses of organismal diversity. He recognizes that a criticism of this model
may be that it does not incorporate any environmental mechanisms that may limit range
or abundance, but points out that a strength of the model is that there are simply too
many species and too many relationships to adequately consider all of them (enter
macroecology!). 

Brown saw implications for community ecology in this range-wide pattern. It implies that
widespread species are found in communities of varying composition across their range,
and that niches do not evolve in a tight-knit community but rather in many different
settings across a species range. 

Based on his findings, Brown rejects the idea that there is a trade-off between a
generalist or specialist lifestyle. Although this idea predicts that specialists should be
more common within their limited ranges, Brown finds that they are both rarer and less
widely distributed than closely related generalists. He extends this idea to evolution and
speciation, adding that it implies that there is a broad spectrum of evolutionary success
across organisms. 

So, in accordance with biogeographic theory, rarer species are more likely to go extinct.
This begs the question, “...whence come the species that replace them?” Brown uses
some of his own thoughts on diversity in his answer, including centrifugal speciation,
stating that central (large and widespread) populations are resistant to directional
selection, peripheral (small and isolated) populations are more likely to undergo
speciation. These new species have two paths to increasing their distribution: a) the
environment changes or b) they expand their niche by evolving to compete more
effectively with closely related species. 

Brown concludes by exploring how this model of speciation fits four macroevolutionary
trends: stasis in the fossil record, morphological differentiation in the speciation process,
differences in extinction and speciation rates between taxonomic groups, and the
tendency of evolutionary innovations to occur in marginal habitats. 

Questions

1. What environmental variables and population attributes determine the abundance and
limit the distribution of particular local populations? Could we define general determinants for

the distribution of species?

2.  Brown says that “the general relationships between abundance and distribution developed

here eventually should contribute to our understanding of the biogeography, population genetics,

and evolution of species as well as the ecological attributes of populations and communities.”

What does the model that Brown presents here tell you about your own research? Did this paper

help you think about your own research? New questions or possible explanations to findings you

have had

3. What are your thoughts on Brown’s treatment of Hanski’s model? Brown dismisses its conclusions about bimodal distribution as unrealistic for larger scales. If this is the case, does Hanski’s model make any other useful predictions?