Blog Author: Devra Hock
Blurb Authors:
Ethan P. White, Department of Wildlife Ecology and Conservation, University of Florida
Field of Study/Research: Data-Intensive Ecology, Ecological Forecasting, Computational Ecology, Quantitative Ecology, Macroecology.
Ethan’s studies data-intensive problems in ecology including ecological forecasting and using high resolution remote sensing to understand individual level patterns in ecological systems at large scales. He is actively involved in computational training efforts as a Data Carpentry founder, The Carpentries Executive Committee member, and lead developer of the semester long Data Carpentry course.
Paper Authors:
Olaf Arrhenius.
Summary/Main Points
1. Main Question:
Background—Expansion on previous paper analyzing the relationship between geographic area and the number of species
-Understanding the distribution of species across areas of unequal size.
Assumptions—Calculating species across a transect is an accurate representation of total species in the area
-Mathematical equation still applies to areas of various and unequal sizes, different elevations, and different habitats (i.e. islands, lowlands, mountains).
Main Question—Determining the distributional relationship of species in various geographical areas
2. Methods:
Data—Species counts from collection data
Methods—Synecological transect-examination method, i.e. the collection of vegetation along a transect is the same as the collection over the whole surface through which the transect is laid
-Measure the length of associations crossed by the line and expresses the length of every association in percentage of the whole line
-1 line per 150 meters provides accurate results.
3. Results
For the equation x=k/y^n, n=0.55 from observations. From calculation1, values correspond to n=0.55, from calculation2, n=0.5 The calculated values agree nearly equally with observed values.
-The relation between the linelength and the mean deviation is what is expected according to probability. Mean deviation is inversely proportional to the square root of the number of observations
4. Summary
-Increase of species within an area can be expressed by a mathematical approximation; holds true for areas with large distances between them, and areas of unequal size.
-The mean deviation between transect lines can be calculated according to the law of probability
5. Questions
-I would have liked to see an expanded results section that better talked about the numbers shown in the tables. What species were found, where certain species found with more abundance, etc. Better explanation would have made the mathematical equations and numbers easier to understand.
-I would have also liked to see a discussion on the species’ distribution. There was almost no discussion of the numbers calculated and it hard to comprehend the full importance of the research.
Paper 2- Arrhenius (1920)
Blog author: Sebastian Botero
Distribution of the species over the area
· Blurb author: Ethan P. White.
o PhD Biology (with distinction), University of New Mexico
o BA Biology (magna cum laude), Colorado College
o When his commentary on Arrhenius (1920) was published, he was a professor at the Dept. of Biology and Ecology Center, Utah State University (from 2007 – 2017).
o Currently, associate Professor, Dept. Wildlife Ecology & Conservation, University of Florida.
o His research focus on “Data-intensive problems in ecology including ecological forecasting and using high resolution remote sensing to understand individual level patterns in ecological systems at large scales” – Sounds like macroecology.
o He has an interesting lab webpage with the projects and publications as well as data management software. http://www.weecology.org
· Paper author: Olaf Arrhenius
o Born in Stockholm, Sweden, on 2 November 1895 and died in 1977.
o The son of the Nobel laureated chemist Svante Arrhenius.
o Got a D.Sc in agricultural chemistry in 1920 (at the time he was publishing this paper) and then moved to work at an agricultural experimental station in Stockholm.
o Received several awards during his career.
· Paper summary
1. Main question:
Background:the author doesn´t give much background, he just mentions that knowledge on the species-area relations is important for plant geography and then continues to discuss the empirical findings on the species-area relations.
Assumptions:There are no explicit assumptions, but I guess that the main one would be that differences in species numbers from the compared sites is not very much influenced by sampling effort.
Main question:What is the mathematical relationship between the area of a site and the number of species present on it? And how well this relation fits the data and is consistent across scales and regions?
The last part of the paper also covers the question on the precision of a transect method for quantifying vegetation types covering an area (only the species-area section of this paper will be covered below).
2. Methods
Data:Arrhenius used plant species number and area for several sites at each of five regions representing different sampling scales (from dm to Ha).
Method:The author used the following equation to describe the relation between area size and the number of species present on it:
y/y1=(x/x1)n
where y represents the area and x the species richness of one site and y1 and x1 represents the area and richness of a second site. n is a constant and is the parameter of interest for the author.
The data was used to derive an empirical value of n, which was compared among regions and sizes to determine how general the pattern was.
3. Results
o Four of the studied regions presented the same scaling parameter (3), with a good match between the number of species predicted by the equation and the observed number. It was significant that the prediction appeared to work well for disparate geographical areas and sampled area sizes.
o The data from the Aland regions showed an exponent of 5.6 in discordance with the other sites. Arrhenius attribute this to the small range of sizes sampled there.
4. Conclusion:The relation between the area of a site and its species richness can be described by the above equation, generally using an exponent of 3. This pattern holds true for different spatial scales of analysis and different areas.
5. Comments:
o To compare these results, take into account that with the way the equation is formulated by Arrhenius, the values for n are equivalent to z values from 0.2 to 0.3.
o I find interesting how a paper that just describes a pattern, with a concrete result, but with further discussion of its implications or underlying mechanisms had such an important impact.
The age of this paper definitely shows. It is obviously written for other ecologists and not the public, which was typically how it was. There is no real introduction to the topic and no link to how this can be applied or used in ecology or why his work was beneficial.
ReplyDeleteI am surprised that all landscapes demonstrated the same pattern of increase in species over the area. I guess because I would think some communities would have fewer or more species interactions, which would impact the increase in species over an area.
I would have liked this paper to give more background. It was at times quite difficult to follow (not surprisingly, since it was published in 1920). However, I found it impressive that Arrhenius could look at some ecological data, catch a glimpse of a pattern, and come up with an equation to explain that pattern.
ReplyDeleteI felt somewhat lost in some areas, such as the synecological transect-examination method. Google gave me a bit to work with, but I'd like to learn more detail about this. It was interesting that this paper has such historical weight.
ReplyDeleteAs much as every grad student groans at reading stacks of papers, we seem to have been spoiled by the predictable and organized format of modern journals. Reading this, I very much wished for a distinct background, methods, discussion. I would have been interested to read what he thought were the implications of his find. I also thought it remarkable that, as E.P.White said, "he didn't have any theoretical reason for expecting that the power-function would describe the SAR." I would be blown away if an "accidental" part of my research was still being discussed in depth a century later.
ReplyDeleteI believe this is one of the earliest data categorized areal tables that visualizes the patterns more easily. In the 1920s, the idea of processing and analyzing data was imaginably difficult since it was in the earliest stage of science. The attempt of explaining the sampling and analysis was adequately done and the author has provided some details considerably well for readers to follow the analytic process. Adding a graph presentation would be a great way to show the results instead of another table. It needs more interpretation details of the results.
ReplyDeleteI feel like this paper did a great job of explaining the statistic portion of the relationship between the area and the amount species that are found in the area. I enjoyed reading about the synecological transect-examination method. It calculates the results using both assumptions and actual measurements and I found this the clearest math break down. I would have liked to read more of background on the paper. This would have probably made this an easier paper for me to understand.
ReplyDeleteI admire how at that time, he had such a visionary insight and was able to prove it in a relatively simple way. He didn't use the computing tools we have right now. It makes me reflect on how spoiled we are now by the technological resources, that even if they are a big help in extending our knowledge in ecological and other process that could not be studied otherwise, we often forget to actually observe by ourselves and think without the computer analyses.
ReplyDeleteI would have also liked to know how we calculated that sampling one transect every 150 my was enough to determine the species composition of an area.