S. Anderson 1977
Commentary Author– Christy M. McCain
- Ph.D. University of Kansas 2003
- University of Colorado Boulder- Department of Ecology and Evolutionary Biology
- Associate Professor & CU Museum Curator of Vertebrates
- Research
- Mechanisms producing and maintaining patterns of species distribution, abundance, and diversity.
- Small mammal range dynamics, abundance patterns across altitudinal ranges and species richness patterns along latitudinal and elevational gradients. Mountains
- Publications
- Grytnes, J. A. and C. M. McCain. In press (invited article). Elevational patterns in species richness. Encyclopedia of Biodiversity (S. Levin, editor), Elsevier, Inc.
- McCain, C. M., R. M. Timm, and M. Weksler. 2007. Sigmodontomys aphrastus: Redescription, taxonomic comparison, and natural history. Proceedings of the Biological Society of Washington 120:117
Dr. Sydney Anderson
- Biologist
- Curator of Mammals, American Museum of Natural History (1961–1992)
- Dr. Anderson led an expedition to Bolivia where the first three specimens of Oecomys sydandersoni were collected
- American Society Mammalogists (Director, President 1974–1976)
- Associate of the Museum of Southwestern Biology Division of Mammals
- First to describe the frequency distribution of range sizes in mammalian faunas
Extended these analyses to faunas of birds, fishes, amphibians, and reptiles - Described pattern of range size-frequency distributions
- Most species have small ranges and large ranges are rare
- J.C. Willis (1949) previously described this pattern in plants and termed the decreasing frequency distribution of range sizes the “hollow curve”
- Anderson noted mammalian range sizes vary over six orders of magnitude
- New model and contrast of other groups necessary
- Posited gap in existing theory on geographic ranges occupied by species
- Examined the ranges of North American terrestrial mammals and found a regular decline in the number of species having ranges in successively larger size-classes of ranges
- Pattern or frequency distribution based on sizes of geographic range occupied by different species in a fauna
- Explored existing theories
- "hollow curve" (Willis, 1949)
- Theory of island biogeography (MacArthur and Wilson 1967)
𝜶-area = smallest biologically meaningful area occupied by a species is the area physically occupied by one individual at one time
- Among mammals, 𝜶-areas range from 4 cm² (small shrew) to 200 m² (blue whale)
𝜷-area = home range, the area traversed in time by an individual
- Among North American mammals, 𝜷-areas range from less than .1 ha for some small rodents to 100 km² for a cougar, with many more species having ranges near the lower end of the range
𝜸-area = composite of the 𝜷-areas of all the individuals in a contiguous population
- Concept of a deme in population genetics is equivalent to the concept of 𝜸-area
- Among mammals, 𝜸-areas range from about 1 km² for the entire range of an insular species like Microtus breweri to 10⁷km² for the original range of the lynx, subspecies Felis lynx canadensis
𝜹-area = total species range
- Largest biologically meaningful unit of area for a species
- 𝜹-area changes during the species lifetime
- Among mammals, range from 1 km² for Microtus breweri to 5.1 x 10⁸km² for Homo sapiens
ϵ- areas = biomes, life-zones, and other faunal areas
- Composite of different species
It is necessary to use maps of different scales to depict areas at different levels from 𝜶 to ϵ. Different magnitudes of sampling and measurement techniques are necessary to obtain useful data.
Scale– as faunal extent decreases, the linearity of the range-size distribution increases
- Compared range-size distributions of all mammals, insular mammals, and several mammalian clades to show pattern variation using a logarithmic scale
- Contrasted these with latitudinal diversity of each clade
- Little evidence except in bats.
‘Rapoport’s rule’ - Areas of high diversity contain species with smaller ranges
- Data on the 𝜸-area or habitat level should be used along with data on the 𝜶-area (localities of individual specimens) in order to determine ranges on the 𝜹-level
Methods and basic data–
- Measured geographic ranges of North American mammals on previously published maps (Hall and Kelson, 1959) including new information on the ranges and taxonomy
- 15 species that occur in both North America and Eurasia
- Lemmus sibiricus (nigripes and trimucronatus), Dicrostonyx torquatus (groenlandicus), Canis lupus, Vulpes vulpes (fulva), Ursus arctos (a plethora of names), Ursus maritimus, Mustela nivalis (rixosa), Felis lynx (canadensis), Cervus elaphus (canadensis, merriami, and nannodes), Alces alces, Rangifer tarandus, Gulo gulo (luscus), Castor fiber (canadensis), Microtus gregalis, and Microtus oeconomus
- The broken line shows Anderson’s hypothesis as to the distribution of species with small ranges that will be approached as more data becomes available
- Anderson plotted the centers of ranges of species against latitudes with different symbols for different orders– the resulting graph was “too complex for convenient reproduction”
- Figure 3 shows differences in both sizes of geographic ranges and latitudes of the centers of ranges for the species of major orders
- Insular species occupy smaller ranges
- Artiodactyls and carnivores occupy larger than average ranges
- Chiroptera and insular species were noticeably more southern
- Presence of more diverse fauna and more islands in tropical part of the continent
- From 10³ to 10⁶km² in range size, bimodality as to latitude is evident
- Tropical and temperate species with centers of their ranges averaging (for ranges of different sizes) from 15 to 20 degrees of latitude for tropical species and near 35 degrees for temperate species.
- At ranges larger than 10⁶ the tropical-temperate distinction does not exist and the average range is centered at progressively more northern latitudes as ranges become larger
- Species with the largest ranges (2 x 10⁷km²) are centered at about 49 degrees
- Bimodality reflects faunal distinction
- Ranges of species in a more diverse fauna do not necessarily have smaller average ranges
- Although this is true for North American bats
Discussion–
Are the species spread evenly throughout the possible ranges?
- In Figure 4 the numbers of species present in each 100 km² size class, averaged over each order of magnitude
- The species are not spread evenly, but are ~an order of magnitude (10 times) less concentrated in each successively larger order of magnitude range
- Anderson suggested the best estimate of the actual distribution in Figure 4 would be curvilinear
- The Central Limit Theorem of statistics
- The means of samples drawn from a population of any distribution will approach the normal distribution as sample size increases, or, essentially all additive statistical distributions are asymptotically gaussian, or ‘normal’
Discussion questions–
- These ranges were determined using historical data, and considered environmental factors independent of a species range. Thinking about conservation and management, what do you think are the implications of using historical ranges as a foundation?
- Looking at Figure 3
- Artiodactyla vs Carnivora, can anyone speak to the differences in distribution? Due to their predator prey relationship, one would expect Carnivora to follow their distribution a little more closely.
- In looking at the range and latitude of Insectivora, why is it that the peak of these distributions are so similar ? Are there other instances of this? Even when looking at the hollow curve of all species range and latitude, the curves are mirrored?
- This paper mentions the ranges of plants following these hollow curves, likely from age. In your research, do your organism(s) of study follow these same patterns? Does anyone find discrepancies in their own research?
Comments
I am having a little trouble understanding figure 3 in the context of the rest of the paper. I am surprised the distribution of the range graphs are left skewed for most of the faunal groups and for the graph showing all of the species. It seems to me that the rest of the results of the paper imply the distribution should be right skewed, with more species having small ranges and very few having large ranges. Am I interpreting this wrong? Is anyone else having trouble with this?
Figure 3 is something I had a hard time interpreting as well. It seems to me that Anderson used the bimodality of figure 3 to show that mammalian clades (apart from North American bats) don't support Rapoport's rule. He says "ranges of species in a more diverse fauna do not necessarily have smaller average ranges". I did think it was somewhat contradictory of the overall results of the paper but it might have something to do with his use of scale?
I was surprised that there are mammal species with a "delta area", or species range, of only 1km squared. It's such a small range.
Figure 3 -- When you look at the pattern of range size within groups, with artiodactyls and carnivores it's easy to see how those groups would have a high percentage of species with large range sizes. I agree that the "all species" graph is confusing. Based on the description of Fig. 3 in the text, I think Anderson is trying to get at patterns of range sizes across latitudes but Fig. 3 doesn't necessarily do a good job of tying together these two datasets.
And for the second question, does this difference in distribution range between Artiodactyla and Carnivora and have to do with the number of species or number of individuals in each order?
Also, how is it that neotropical species have a smaller area than mainland species in higher latitudes? (figure 2)
Manuela asked about neotropical species sometimes having smaller geographic ranges than mainland species at higher latitudes. We talked a little about this in the Zoom call. Animals in the tropics experience reduced seasonality and greater habitat heterogeneity, so they may be adapted to narrower niches (i.e. following the pattern observed in Janzen's "Why Mountain Passes are Higher in the Tropics." It may also relate to recent, repeated glaciations in the Pleistocene: animals that have been able to disperse back to regions formerly covered by glaciers are likely biased towards those with larger ranges anyway.