SPSS analysis and interpretation

Table 5 displays the statistical analysis conducted for each dumpsite where soil sampling method 1 was applied. This way, differences in average pH and EC between the dumpsites and in the control areas can be assessed. Below the table, the results for each dumpsite are shortly discussed.

UAUCUStudent Research Exchange Collected Papers 2018

160

Table 5:

Statistical findings of soil parameters within waste dumping sites and their control areas

* Significant difference at 0.05 level (2-tailed).

** Significant difference at 0.01 level (2-tailed).

Babijn

For Babijn, the mean pH and EC nearby the waste are 6.61 and 1105.788 µs, respectively, while in the control samples, average values are 7.6525 for pH and 16083.477 µs. The difference in pH between the two sites is seen to be significantly different (p = 0.00).

Lagabena

Despite the small difference in average pH within the dumpsite in comparison to the control area with values being 7.6 and 7.98, respectively, it was found to be significant (p = 0.03). Similarly, the conductivity levels are significantly different (p = 0.01) with increased mean EC within the dumpsite compared to away from it as the levels are seen to be 96.194 µs and 5.0175 µs respectively.

MokoNo significant difference in mean was found for the soil parameters within Moko with pH and EC levels averaging to 7.9110 and 22.2964 µs, respectively, within the dumpsite and 8.0805 and 54.0776 µs in the control area. In this dumpsite, the conductivity was actually found to be higher away from the waste than within.

Parish Hill

In Parish Hill, comparable pH levels were found near the waste and away from it with values attaining 7.027 and 7.8425, respectively. This difference in means is found to be significant (p = 0.00). On the contrary, the EC levels are not significant, but are also found to be higher within the sites than away from it as the conductivity is 250.4305 µs and 33.686 µs, respectively. The fact that the difference is not significant may be explained by an outlier, which is reflected in the high variation in conductivity levels detected in the within dump soil samples.

Tamarijn

Significant results can be observed for Tamarijn at 0.05 level as the pH and EC values are 7.315 and 514.04 µs, respectively, within the site and 6.941 and 110.13065 µs in the control, with significance values of (p = 0.014) and (p = 0.015).The pH in the inactive site appears to be more neutral than the slightly acidic values found in the control area.

Matadera

Significant differences in average pH (p = 0.00) are also obtained in Matadera with values seen to be 6.9245 within the dumpsites compared to 7.522 at the control site.

Meiveld

Significant pH differences (p = 0.00) are seen in Meiveld when comparing the area at the top of the cliff near the waste to the control site as well as the bottom of the cliff. Indeed, the values were found to be 6.87 at the top and 7.19 at the bottom compared to 7.89 in the control area. The difference in pH between the top and the control is seen to be the greatest out of all dumpsites.

UAUCUStudent Research Exchange Collected Papers 2018

162

For the transect method, correlation coefficients, specifically Spearman’s and Pearson’s coefficients were obtained for the relationship between distance from the dumpsite and the soil parameters as shown in table 6 and 7 below. Below the tables, the results for each dumpsite are shortly discussed.

Table 6:

Correlational findings between the distance and the soil parameters of the dumpsites

Table 7:

Correlational findings between the distance and the soil parameters of the dumpsites

to the distance from the dumpsite, but these were not statistically significant.

Figure 4: pH gradient at Mon Pos Shaba

The pH data for Shaba showed a positive correlation between pH values and distance from the dumpsite where the Pearson’s rho coefficient was most notable at 0,622 and statistically significant (p = 0,00). This is shown in a gradual increase in pH as samples are collected further away from the dumpsite. The first few meters show a clear trend, as shown in figure 5. The EC also mostly increase as samples are collected further away from the site, but these coefficients were smaller and not significant.

Figure 5: pH gradient at Shaba.

Alto Vista

Relatively small positive coefficients were found for the pH at Alto Vista. Although a small increase between the dumpsite and transect samples can be seen, the correlation between distance and pH levels is not supported by the SPSS output as they were not significant. The EC measurements, on the other hand has one of the strongest correlations measured for the Spearman’s rho coefficient of 0.880, which was also highly significant (p = 0.00).

Figure 3: EC gradient at Alto Vista.

MokoDespite the apparent inconsistencies within the transect data, the correlation was positive at 0.690 and 0.651 for linear and monochromatic regression, respectively. Both coefficients were highly significant (p = 0.00). However, the EC measurements along the transect were not significant.

Mon Pos

Mon Pos showed positive correlation coefficients for both the linear and monochromatic approach, especially when two outliers were removed from the data, the coefficients became 0,631 and 0,832, respectively. Both were highly significant at (p = 0,00). For this dumpsite a monochromatic model fit the data better than the linear model, compared to Moko where this was the other way around. The EC measurements showed a negative correlation in relation

UAUCUStudent Research Exchange Collected Papers 2018

164

the only exception to this, where chromium levels were higher in control than dumpsite soil. Matadera had the highest chromium and copper concentrations compared to all tested sites. Additionally, this site also possessed the greatest number of waste items per plot. This is in agreement with the results of a study that has previously found that the quantity of waste dumped greatly influences the pH, conductivity and heavy metals available (Ali et al., 2014). The soil parameters suggest that Meiveld and Matadera are the most environmentally impactful out of the 10 assessed dumpsites. As both sites are still highly operational, this finding would be in line with the idea that active dumpsites are (at least slightly) more damaging to the environment than inactive ones as decomposition processes tend to stabilize over time.

When looking into the transect approach, a leachate gradient was observed with lower pH close to the waste and higher values in the samples collected progressively away from the site. These increasing values with increasing distance from the site are not in accordance with results found in a previous study by Akinbile (2012) and the results found for the dumpsite-control comparisons. It was found that leachate spread was measurable up to 20 meters from the dumpsite, especially when the transect was set out to follow the water flow direction. An opposite trend is seen in the conductivity measurement of Moko and Mon Pos where values get increasingly lower away from the dump. Despite the general trends that were found for EC measurements, there were also many differences between the sites. At Alto Vista for example, the EC measurements seemed to get lower close to the dumpsite compared to most other sites where an opposite trend was found. Evidently, dumpsites and their effects are heavily dependent on the waste dumped and the characteristics of their location.

Most of the dumpsites were similar in waste composition with regards to industrial wastes. The large amounts of building debris, concrete blocks and metal scraps