Chelsie Guild
Bio 1615 Paper
Judy Bunkall
Merged Trees in Second-Growth, Fire-Origin Forests in Patagonia, Chile: Positive Spatial
Association Patterns and Their Ecological Implications
Introduction
It is stated that as far as the authors know, no one else has tested this experiment. The
location was in Central Patagonia, Chile, an area that had been covered by a forests mostly
consisting Nothofacus Pumilio, a plant or shrub known for rejuvenating easily after fires. The
history behind this area and the N. pumilio is that the pioneers that had resided in this region
burnt down a great portion of this forest so their livestock had open land to live. What resulted
was multi-stemmed trees of N. pumilio second-growth occurred more often at the outer edge of
the forest. Alex Fajardo and Eliot J. G. McIntire established that some communities living in
these environments are often characterized by increased resource availability or a high degree of
abiotic stress.
Some unknown variables that surfaced with the experiment are that the accounts of
merged trees occurring at the forest edge is not assured; another is whether the occurrence at the
forest edge is less common in other situations. It is unknown whether the density dynamics in
these environments fit the more traditional dynamic models described for temperate forests.
Taking from what Fajardo and McIntire do know, “that during the stem-exclusion phase the
negative density-dependence process acts with distance, making initial clustering disappear with
time”, contributed to the question to ask, ‘how spatial patterns of merged trees fit spatial patterns
expected to emerge in the stem-exclusion stage’.
Materials and Methods
N. pumilio is a deciduous species. It illustrates mast seeding of 3-8 year intervals. The
wind is most likely the way these seeds spread. This species is very thin barked, which makes it
easy for them to be killed, for example, by fire. Samples for this experiment of N. pumilio were
taken from the outer edge and the interior of the second-growth forest. Samples were also taken
from the mature forest. N. pumilio is not known for sprouting; instead just the touching of multistemmed trees may start stages of fusion. In order to tell if the plants were multi-stemmed, the
scientists did a cross-section slice of the root collar to see if there were clusters of individuals or
if there were just one single individual.
The first objective was establishing the sample size to be tested. The sample size used for
the experiment was 10 individuals. The second objective was to provide summary statistics.
Taken from the article, some important descriptions of what was found was the stand density
(number of stems ha -1), tree age and basal area (BA; m2ha-1). Where the individuals were
found (n=2 for interior and n=4 for the edge). Relative importance values were found by (RIV=
relative density of merged trees + relative basal area of merged trees, all divided by 2). The third
objective was to assess changes in growth and competition rate. Testing began with the edge
individual to see if being on the edge promoted merging-rate growth. This was done by using
log(DBH) with two fixed effects- the site type and merged status (merged or single). Secondly in
this third objective was to compare the merged and single trees. The fourth objective was to
quantify the small-scale spatial-correlation structure of patterns. This was done by using the
nearest-neighbor (NN) and the pair correlation function. The comparison was done again
between the merged and single trees.
The authors hypothesized that positive tree-tree interactions would be best represented by
a ‘monotonically decreasing distribution’- that is, the most common nearest neighbor occurrence
is in trees very close in distance. If competition had been a main process in a specific area the
distance from tree-tree would increase or get larger. Now to finalize the experiment, the
determination whether spatial patterns differ from standard expectations for temperate forests
was addressed. The univariate pair-correlation, g(r) was used to determine if major second-order
effects in the univariate pattern, the comparison of the observed pattern with 99% Monte Carlos
simulation envelope. Finally, they used a goodness-of-fit (GOF) test, “that summarized scaledependent information contained in the pair-correlation function into a single test statistic that
represents the total squared deviation between the observed pattern and the theoretical result
across the distances of interest”. This GOF test was accomplished with ‘Progamita software’.
Results
The multi-stemmed trees that the scientists studied, the sample of 10, did have a cluster
of individuals within the root collar. There were 2 to 6 individuals within this multi-stemmed
tree. This showed that there had been the fusion of individuals instead of each individual
sprouting on its own. The site demography showed that the dominant species in both the mature
and second-growth forest was the N. pumilio. Merging occurrence composed mostly of 2 to 6
merged stems were found mostly at the edge of the second-growth forest. There was only one
merged stem found in the interior of the forest. The spatial patterns did differ at the edge of the
forest the individual stems were clumped at distances, but in the interior forest had more random
distribution of individual stems was found at all distances. Overall, the nearest-neighbor
distances for each tree differed depending on if it was a multi- or single- stemmed tree.
Discussion
The merging of stems does not occur randomly in space of the second-growth forests of
N. pumilio. This may suggest a link of characteristics at the forest edge. In a companion study,
they found that the seedlings of N. pumilio establishes in groups. Abiotic stress may be the
reason for merged trees at the edge of forests. Positive interactions overcame competitive
interactions due to the increased growth of merged stems. Fajardo and McIntire state “that
resource heterogeneity, poor seed dispersal and low growth rates, and vegetative reproduction or
facilitation are several mechanisms that may generate slumped spatial patterns”. It is concluded
that the ultimate mechanism for merging must have occurred before this study and it had to have
been involved with positive interactions. Environmental conditions contribute to major
competition for the multi-stemmed plants and often reduce density-dependent effects. When
environment conditions are limited, it contributes to random mortality and independent densities.
It was found that once the tree stems become merged, the competition factor was no longer an
issue. This means that many individual tree stems, not merged stems, are more likely to be
affected by environmental competition.
Conclusion
After the study ended, it is concluded that multi-stemmed trees most often occurred at the
forest edge. It was also known that the multi-stemmed trees have multiple origins. The pattern
was pretty consistent in that the positive reactions most likely contributed to rough environments.
The scientists hypothesis overall was correct, in that the merged trees occurred at the forest edge
and the spatial aggregation of stems was consistently random.
Works Cited
Fajardo, Alex and Eliot J. B. McIntire. Merged Trees in Second-Growth, Fire-Origin Forests in
Patagonia, Chile: Positive Spatial Association Patterns and Their Ecological Implications.
American Journal of Botany 97(9): 1424-1430. 2010.