Bradley Lynn
GEOG 491
November 17, 2008
Lab #5
In this lab correlation was analyzed between human land use patterns and guava coverage on Isabela Island, in the Galapagos. Guava is known to be an invasive species and a pest, invading both farmland and preservation areas outside the farms.
First, guava coverage was looked at inside of farms, and compared against provided statistics such as heads of cattle within each farm (Figure A-1), percent of land on the farm dedicated to pasture (Figure A-2), as well as percent of land on the farm dedicated to cropland (Figure A-3). When put into linear functions, these relations were found to have negative correlations. One can explain that when land is being actively used, it does not succumb to immediate guava growth. A high positive linear correlation was found between percent of land reported as unused within the farms and guava coverage. Land reported as unused was calculated as any percentage within a farm not reported as pasture or cropland. This can be seen in Figure A-4, which features a correlation of over 60%. Looking at individual fields, a strong correlation was also found between length of time since a field’s last clearing and guava cover (Figure A-5). This too reinforces the assessment that fallow fields that are not being used are more prone to guava growth.
Buildings and roads were put into consideration for guava cover patterns within fields. In Figure A-6, average percent guava coverage is shown for fields that include roads and buildings, and fields that do not include roads and buildings. The average guava growth in fields containing buildings is noticeably lower. This may be due to a persistent human presence in areas with buildings, which are more likely to be affected by humans through cutting, planting or animal husbandry.
For the next analysis, farms were classified as active, partially active, and abandoned depending on the percentage guava coverage within. The percentage splits used were 0-40%, 40-65%, and 65-100% respectively. This was decided upon after looking at the land use percentage curve as it relates to guava cover (as featured in Figure A-4). Clustering was found in that farms designated as partially active were invariably found adjacent to farms designated as abandoned. Figure A-7 shows that farms bordering abandoned farms are on average more likely to contain a higher percentage of guava cover than those not bordering abandoned farms.
Looking at the spatial organization of guava cover within the fields themselves (as shown in Figure B), cover seems to be clustered away from roads and especially buildings, which reassures the conclusions reached from Figure A-6. However, the question of whether guava overgrowth or human development came first is a question best left for when historical data is to be viewed.
As for the encroachment of guava upon natural lands on the outer borders of agricultural areas, guava cover in these areas was analyzed for amount of cover present compared to the distance from the farmland area. Rings roughly 700 meters wide were used as collection points, and equal sized squares verified as containing guava cover were counted within these rings. A very strong negative linear correlation was found between distance from an agricultural area and squares of guava coverage enumerated. A bar chart featuring the counts within in each ring is featured as Figure C.
A ring 500 meters wide of guava coverage present outside of the farmland area was taken and linked to the closest farm and fields within this farm. These areas outside the farms were then compared with the attributes of these nearby farms and fields. Linear correlations and their coefficients can be seen in Figure D. The most likely correlation was found to be between unused land within a farm, as calculated in Figure A-4, and the growth in areas just outside of these farms. This can be seen in Figure D-2.
Based on the conclusions reached from the types of analyses done throughout this lab, one can conclude that the cause of guava invasion is largely due to a human presence, as is evidence by guava growth in natural areas being largest when bordering human settlements such as the farmland looked at in the study. However, human activities such as construction, farming and animal husbandry, while possibly spreading the guava into natural and unattended areas, stave off guava growth. Fallow and abandoned fields and bordering natural areas are of the largest concern is anything is to be done about the invasion of the non-native guava species.