For This Given Assignment, Two Land Use and Land Cover (LULC) Files Were Used, One From

Bradley Lynn

GEOG 491

November 6, 2008

Lab #4

For this given assignment, two land use and land cover (LULC) files were used, one from 1986 and the other from 1996. Thus, these two data sets can be compared to provide an idea of the changes sustained in this area over a ten year period. This was done by taking the four categories in which the datasets were separated into and multiplying the 1986 dataset by ten, and adding the 1996 dataset. Thus, one can determine the 1986 and 1996 status of a given area of land by looking at the tens and ones place of the value provided. Thus, if forested land has a value of 1, and land without vegetation a value of 3, a composite value of 13 provides that the land type was originally forested land in 1986, but has changed to barren land in 1996. A figure showing the composite values for this area is provided as Figure A.

From this, the areas were reclassified (simplified, really) to show areas of overall change within the study area. Within Figure B, unchanged areas are shown in red, and changed areas are shown in blue. Below this are figures for overall change and change within the farms under study. Please note that there is a marked increase in LULC change** within farms when compared to the greater area. Un-generalized figures for both the overall area of study and areas only within farms are provided in Figure B-2. Please note that the level of LULC change is higher within farm areas, as well as the overall level of deforestation.

The percentage of LULC change within farms (as determined as change from forest to any other Land Cover, over total area of a farm, times 100) was analyzed with access from a farm to roads and other transportation venues. This was determined by taking the order of roads bordering a given farm (1st order being a multi-lane unpaved path, 2nd order being a single-lane unpaved path, and 3rd order being a smaller path) and multiplying this order by the number of changes in road order taken to reach either a paved two-lane road (as featured in the southeast of the area) or a large river (prominent in the northern reaches of the area). Thus, a low number is reflective of relative ease of travel from a farm to a dependable travel venue, and a high value can be seen as indicative of increased difficulty of access to such roads and rivers. An output including values assigned to roads, farms, and rivers can be seen in Figure C. Also included in this figure is a scatter plot featuring the association between ease of travel and level of LULC Change in each farm. One would expect a negatively inverse association between the two, as change in land within a farm could be assisted through input of tools and materials from outside the farm, as well as an increased ease in which unwanted material such as trees can be moved off the farm. However, no such relationships between travel ease and increased LULC change appear. Instead, a very slight (and most likely statistically insignificant) parabolic relation is a better fit. Thus, within the scale and scope of this analysis, ease of transport to a farm is not a reliable indicator of the percentage which has undergone LULC change.

The linear distance from the center point of each farm under study to the closest identified community was measured and put in a function against percentage LULC change per farm. The result of this analysis can be seen at the bottom of Figure D. Again one might expect an inverse relationship between distance from community centers and LULC change, but again this appears to not be the case. Instead, a very slight positive relationship exists between distance and amount of LULC change present in a given farm. LagoAgrio is a larger village in the northwest of the study area. A similar relationship in distance to this community was found using the same process. For both of these circumstances, a possible explanation is thus: as areas near to communities are typically more developed, this increased stability of human environment may lead to an increased stability in land use type. (e.g. vegetation used for agriculture tends to remain in use for agriculture, barrens and common areas tend to remain as such) Lands further out from these community centers have yet to be developed and may undergo increased LULC change as they are developed, when compared to areas already developed.

What else then could be used as a determining factor of LULC change within a farm? Looking at the population of the given farm could possibly provide insight. Higher population farms within the study seem to be concentrated toward the south, generally towards the community in the center of the overall study area. Can a higher population farm be associated with a higher level of deforestation** within the ten year study period? The trend, as shown as the first chart in Figure E is encouraging. There seems to be a slight positive correlation between population of a farm and the percentage of land deforested in the ten years under scrutiny. The second chart in Figure E depicts the correlation between hectares of farm product reported for every given farm in a year, and the percent deforestation calculated from the LULC change dataset. An ever slightly more significant positive correlation is found between these two factors than that of farm population. The question of crop diversity on level of deforestation, however, is irrelevant, a very low correlation was found between number of different crops grown and percent of land deforested within a given farm. This is featured as the third graph of Figure E. The fourth graph features access to technology as the determining factor, and comes away with a slight negative correlation between number of available technologies and percent deforestation. The fifth graph, perhaps the most interesting, with the highest level of statistical significance among these assessments, features education level as a function of deforestation percentage. As the level of the Jefe’s education increases, the destructive potential of a farm decreases by no small amount. This is perhaps because of techniques acquired in higher levels of education that teach of the benefits of a natural environment upon farm settings.

Of course, none of the analyses have a desirably high level of statistical significance or relation, therefore the processes of LULC change and deforestation examined within this lab cannot singularly explain the phenomenon. However, when compounded and looked at comprehensively, with extensive knowledge of both the data used and of a GIS, a more sound explanation may be reached. It is certainly not out of the question to assume that other factors than those recorded play a large role in LULC change and deforestation. Alternatively, looking at simple patterns, such as those presented in Figure F, the process of change seems to be concentrated in areas of higher population communities with access to more reliable roads and waterways. The relation is spatial, as it often is, and this can be said with certainty.

**Please note: Until it is directly specified within this report, rather than using a measure of “deforestation” (that is, change from an area previously forest to another land cover) when the term “LULC Change” is used, it refers to all change in LULC found between 1986 and 1996, regardless of type. This consideration is put into execution because it can be assumed that all change, regardless of type is detrimental to the environment and reflects an area of chaos in which biodiversity can be lost. Additionally, the documentation and instructions provided for this lab do not specify the term “LULC Change” as being explicitly “deforestation”. Thus, for the sake of inclusion, all LULC change was considered in figures A through D.