Summary
Measuring the impacts of climate change is becoming increasingly
important as biodiversity in ecosystems continues to decline. This is as true in South America- the
geographic focus of this study- as it is anywhere. The goal of this study is to find the best
way to measure ecosystem resilience against increases in climate stress.
Several
methods have been used to research this problem, which the researchers state,
but they are all qualitative. Unfortunately,
no background was provided to support these qualitative methods- only that they
involve ecological stability theory and ecological niche theory. The article doesn’t specifically state what
these methods entailed, only that they apply to the theories. References are
provided for these methods, making it up to the reader to investigate them.
Their
research proposal is to generate a quantitative method by coupling the two
theories. The idea is to construct
spatial models that measure the resilience of ecosystems “through a metric of
climate suitability, based on the multidimensional niche preferably occupied by
them” (p. 2). The educational
significance of such a method is to provide the only quantitative way of
measuring ecosystem resilience against climate stress. Results indicate that forests are more
vulnerable to climate change in South America than savannas or grasslands.
Critique
I did not find a comprehensive review of the related
literature in this study. Citations
mostly applied to the data being used to delineate ecosystem boundaries and
vegetational cover. While it is stated
that the reference for vegetational cover was a MODIS satellite, a reference
for delineating the ecosystem boundaries was not. The reader needs to look in the reference
section to find data for that.
There is not a specific question to be answered or a
hypothesis to prove or disprove. The
goal of this research is to make the case for a quantitative method of
measuring ecosystem resilience against climate change. As we’ll see later, they don’t really answer
this question so much as what the findings imply for conservation.
The only instrument used was a MODIS satellite to
delineate ecosystem boundaries by examining the vegetation content of each
pixel. Ecosystems were classified based
on the percent of vegetation in them: less than 5% was grassland or desert,
5%-60% was savannah, and over 60% was forest.
MODIS satellite data is a cheap and effective way to estimate ecosystem
boundaries, since these would be difficult to determine by fieldwork alone.
Samples were collected for ecological niche
modeling. Based on the classification
scheme, 53% of the samples were savannah, 38% were forest, and 9% were
grassland. A total of 37,763 samples
were used in the model, indicating an impressive amount of data for the
study. However, it was not mentioned if
the sample locations were randomly generated, systematic, or stratified in any
way, leading the reader to wonder how much room there is for sampling
error. The model also used data from
CHPclim (precipitation) and the WorldClim temperature database. The need for these variables in the model
wasn’t clearly stated in the text; we are to assume they are the key climatic
ones when measuring climate change resistance.
Using the biomed2 package in R software, the model was processed using
the variables above. It was run using 10
different methods! The highest quality
methods were selected based on a True Skill Statistics (TSS) score of .7 or
above. An exponential equation was then
used to fit the models into a curve that would generate an indicator of
ecosystem resistance to climate stress.
The design of this research is stated with great
attention to detail but is difficult for someone without knowledge of R and
statistics to thoroughly understand. Because
there is so much detail, other researchers familiar with these topics should be
able to replicate the method if desired.
Results from the study are clearly explained, not only in
the text but in the maps and graphs provided.
The maps and graphs were appropriately clear and colored to make the
data easy to read. Figure 1 showed a
color-coded map for each of the three bioregions (grassland, savannah, forest)
and their resilience level. Figure 2
showed the relative frequency of resistance to climate change response for the
bioregions. Figure 3 showed patterns
between climatic variation in temperature and precipitation and ecosystem
resilience.
Based on the figures, moisture availability showed the
strongest interaction with climate resilience.
Forest ecosystems in South America have a relatively narrow ecological
niche, and since the models predict there will be less moisture in the south of
the continent, that is where the researchers expect the least resistance to
climate stress. Stronger resistance can
be seen on map B for savannah, having more extensive areas of high resistance
across the continent. A potentially
confusing aspect about map A for grassland is that it shows low resilience
across the continent, yet it is stated in the text that high resistance can be
expected.
Results are discussed in terms of the ways this method
can apply to applications within the science of conservation, whether it is in
evaluating ecosystem resilience or restoration policies. I didn’t find a strong argument supporting
why this quantitative one would be a better option than the qualitative ones briefly
mentioned in the introduction. The
results would have been stronger if they had addressed which method is more
accurate, costly, and beneficial.
Surprisingly, there was no suggestion of further research
involving their method. Much of the
discussion revolved around the findings and what can be done to conserve the
environment in South America. Placing
more attention on forests is an important conclusion since they demand more
moisture and are more important to a balanced global climate. But this seems to be something
conservationists- and perhaps the researchers themselves- already
suspected. If the hypothesis had been
that forests in the south would show less resilience, then the discussion would
have more aligned with it.