The Minitab 17.0 edition's general linear model was used to analyze the variance in the acquired data25. Data collected on the main and interactive effects of charcoal production sites, depth, and location in measured and selected soil parameters were examined. The PCAs were conducted using the scikit-learn library in Python26. The HCA was performed using the scipy.cluster.hierarchy module in Python27. Dendrograms were generated to visualize the hierarchical structure and identify optimal clusters. The results were presented in tables and figures, to demonstrate the substantial differences of means obtained at a 5% probability level of confidence, Tukey HSD was used as an experimental post-hoc test. Microsoft Excel 2016 was used to calculate the treatment means and standard errors.
The predominant soil type is Alfisol, comprising sand (60.74, 58.95, and 57.42%), clay (28.26, 27.93, and 27.25%), and silt (11.00, 13.11, and 15.33%), with sand being the most abundant, followed by clay and silt. At Ìrèle and Ìpaò, the soil pH was strongly acidic, while it was moderate for Òkè-Àkò. Below the critical value of 10 mg/kg, Ìrèle exhibited the lowest available phosphorus level, whereas Òkè-Àkò and Ìpaò showed significant increases. Nitrogen and potassium (exchangeable) levels exceeded the critical levels of 1 mg/kg and 0.2 cmol/kg, respectively, although calcium and magnesium levels were lower, with the highest values recorded in Òkè-Àkò, significantly differing from the other two locations. The cation exchange capacity (CEC) was lower than the recommended range of between 10-20 cmol/kg, and exchangeable acidity (EA) was higher than the critical limit of 1.0 cmol/kg. At all locations, the base saturation level was slightly above the critical limit of 50% and there were low levels of trace elements including copper, zinc, iron as well as manganese. Our previous work, further presents the detailed physical and chemical properties of soil samples taken from three locations.
The presentation of precipitation and temperature patterns in the study area is outlined in. The cumulative rainfall, measured in millimeters, during the period from April to August 2018 and the recent data in 2023 at the research site is detailed as follows: 120 mm in April, 152 mm in May, 168 mm in June, 170 mm in July, and 131 mm in August. Concurrently, the average temperatures, expressed in degrees Celsius, were recorded as 27.4 °C in April, 26.9 °C in May, 26.2 °C in June, 25.8 °C in July, and 24.9 °C in August.
Significant interaction effects between site and location were observed for the activities of phosphorus and sulfur cycling enzymes (Table 1). The recorded interaction effects were significant and varied both in size and direction of their response. Notably, phosphatase (Pho) levels were significantly higher at NPS in Òkè-Àkò and Ìpaò than sulfur enzymes. However, the peak value (2.94 mg/ml/min) was observed at NPS in Òkè-Àkò. Thiosulfate dehydrogenase (Tsd) exhibited significant site-by-location interaction effects, with the highest value recorded at NPS in Ìrèle and the lowest activity level at NPS in Òkè-Àkò, although the variation was insignificant. The recorded interaction effects were both significant and diverse, involving variations in magnitude and direction of response. Additionally, Dimethyl sulfoxide reductase (Dsr) was found to be significantly higher at CPS in Ìpaò, while low activity levels of (0.80 and 0.68 µg/ml/min) were observed at NPS in Òkè-Àkò and Ìpaò, respectively. Lastly, phytase (Phy) activity remained consistent across both sites and depths but showed significant variations among means.
Significant interaction effects between site and location were observed regarding the soil nutrient status, as indicated in Table 1. Copper and cobalt were found to be significantly highest at CPS in Òkè-Àkò, however similar trends occurred for Iron in Ìpaò. In contrast, significantly lower values were recorded for cobalt, and zinc in Ìpaò at NPS.
Except for the 30-45 cm soil depth, which displayed lower values (1.78 & 0.84 µg/ml/min) at NPS for Tsd and Dsr, respectively, Table 2 highlights significant interaction effects between site and depth for the selected phosphorus and sulfur enzymes. Particularly noteworthy are the interaction effects observed, such as Tsd showing marginally higher values at the 0-15 and 30-45 cm soil depths at NPS and CPS, respectively, with the highest value recorded at the 0-15 cm soil depth at NPS. Additionally, Pho exhibited significant variation, with the highest activity at the 15-30 cm soil depth at CPS and the lowest activity at the 15-30 cm depth at NPS. Furthermore, at CPS, Cobalt, Iron, and Zinc exhibited their highest concentrations at the 15-30 cm soil depth, while the lowest amount of copper was observed at the 30-45 cm depth. Conversely, at NPS, zinc showed its significantly lowest concentration at the 30-45 cm soil depth. Interestingly, copper was found to be significantly highest at the 30-45 cm soil depth at CPS.
Significant interaction effects between location and soil depth were noticed in soil P and S enzymes (Table 3). Particularly, at the 0-15 cm soil depth, significantly higher activity levels (3.32 mg/ml/min & 9.41 µg/ml/min) were observed for Pho in Òkè-Àkò and Tsd in Ìrèle, respectively. A similar pattern was observed at the 30-45 cm soil depth, where Ìpaò recorded the highest activity levels for Dsr (1.78 µg/ml/min) and at Ìrèle for Tsd (12.51 µg/ml/min) at 15-30 cm depth. However, at the 30-45 cm depth, notably lower values were observed at Ìrèle for phosphatase and Òkè-Àkò for both Tsd and Dsr activity. Phytase activity remained consistent across the locations and depths but exhibited significant variations among means. In addition, location and depth showed significant interaction effects on the soil micronutrients. For instance, In Ìpaò, Iron, and manganese were significantly highest at the 0-15 cm depth, whereas in Òkè-Àkò, both elements were found in lower amounts at the same depth. At the 15-30 cm depth, copper and cobalt were significantly highest, but they were located in different locations (Ìpaò & Ìrèle) respectively. Zinc exhibited the highest amount at the 0-15 cm soil depth in Ìrèle and the lowest amount at the 30-45 cm depth in both Ìrèle and Òkè-Àkò.
At CPS, Pho, Tsd, and Dsr activity increased across all locations with increasing soil depth, except in Òkè-Àkò where it decreased (Table 4). Phy showed a consistent trend across all locations and depths at both CPS and NPS. For soil nutrients, Fe decreased across locations and depths. However, Co and Zn followed similar trends, except in Òkè-Àkò. Cu and Mn activity increased across locations with increasing soil depth, but Mn showed no significant difference. At NPS, Cu, Co, and Fe increased across two locations with increasing depth, except for Ìpaò. Mn and Zn decreased down the soil profile at all locations, except in Òkè-Àkò. Pho, Tsd, and Dsr decreased with increasing soil depth, except Pho, which increased at Ìpaò across soil depth.
Three-way analysis of variance presented in [Table 5], indicates the effect of charcoal production on some selected P and S enzymes at the CPS and NPS. Regardless of location and soil depth, there were no significant differences (P > 0.05) indicated in Phy and Tsd between CPS and NPS, but higher and significant differences (P < 0.05) were recorded in CPS with Pho and Dsr. The same table also showed the effect on some selected P and S enzymes at different locations. Irrespective of the production sites and soil depth, there were significant differences in soil P and S status amongst the three different locations. Phosphatase was significantly (P < 0.05) higher in Oke-Ako than in Irele and Ipao, higher significant differences were also recorded in Tsd and Dsr at Irele. Although no statistically detectable differences were indicated amongst the three locations. There were significant differences in soil P and S enzyme status as affected by soil depth (Table 4). Phytase and Dsr decreased in the order of increasing soil depth with the lowest value recorded in 30 - 45 cm depth, however, Tsd and Pho with no significant differences recorded higher values in 0-15 and 15 - 30 cm depths respectively. There were significant interaction effects of site by location by soil depth recorded in P and S enzymes. The effect of charcoal production on some soil nutrient status at the CPS and NPS [Table 5], showed that Cu, Co, Fe, Mg, and Zn were higher in CPS than NPS. In terms of location, Cu, Co, and Mn were found to be higher in Ìpaò compared to the other two locations. However, Fe and Zn were found to be higher at Ìrèle with significant differences (P > 0.05). Furthermore, Cu and Co increased with increasing depth while Fe and Mn followed the opposite trend. Finally, Zn remained consistent across the soil profile showing a significant difference (P > 0.05).
The correlation analysis revealed significant associations among various parameters in the study area as presented in [Table 5]. The pairwise Pearson correlations indicate the significance of these associations. A weak negative correlations was observed between Mn and Zn, as well as Fe and Zn, indicating that fluctuations in Mn and Fe levels are not consistently linked to changes in Zn levels. Similarly, there is a weak positive correlation between Co and Zn, and Cu and Zn, suggesting that variations in Co and Cu levels are not significantly tied to changes in Zn levels. Furthermore, there are weak negative correlations between Nematode (10) and Zn, and Biomass P and Zn, indicating that variations in Nematode (10) and Biomass P levels do not consistently correspond to changes in Zn levels. Conversely, a strong positive correlation is detected between Dsr and Zn, and Tsd and Zn, both of which are statistically significant. This implies that as levels of Dsr and Tsd increase, Zn levels tend to increase.
Transitioning to correlations between different elements, a very weak positive correlation between Fe and Mn was observed, while weak negative correlations exist between Co and Mn, Cu and Mn, Nematode (10) and Mn, Fungi (10) and Mn, Bacteria (10) and Mn, Biomass p and Mn, Dsr and Mn, Tsd and Mn, Phy and Mn, Pho and Mn. None of these correlations are statistically significant, indicating that variations in these elements are not reliably associated with changes in Mn levels. Also, a moderate positive correlation was identified between Co and Fe, and strong negative and moderate negative correlations are noted between Cu and Fe, and Nematode (10) and Fe, respectively. These correlations are statistically significant, suggesting that as Co levels increase, Fe levels tend to increase, while an increase in Cu and a change in Nematode (10) levels are linked to a decrease in Fe levels.
The correlations between Fe and Fungi (10), Bacteria (10), and Biomass P are very weak and not statistically significant. However, a moderate positive correlation is found between Dsr and Fe, and a strong negative correlation is observed between Pho and Fe, both of which are statistically significant. These results indicate that as Dsr levels increase, Fe levels tend to increase, while an increase in Phosphatase (Pho) levels is associated with a decrease in Fe levels. The analysis extends to correlations between elements such as Co, Cu, Nematode (10), Fungi (10), Bacteria (10), and Biomass P, revealing various weak and very weak correlations, none of which are statistically significant. Shifting to Tsd and Fe, a moderate positive correlation is noted, indicating that as Tsd levels increase, Fe levels tend to increase. When exploring relationships with Cu, there is a moderate negative correlation with Co, a very weak negative correlation with Nematode (10), Fungi (10), and Biomass p, and a weak positive correlation with Bacteria (10). While the correlation with Nematode (10) is not statistically significant, correlations with Co and Fungi (10) are, suggesting that as Cu levels increase, Co levels tend to decrease, and an increase in Cu levels is associated with a decrease in Fungi (10) levels. Bacteria (10) exhibits very weak correlations with Nematode (10), Fungi (10), and Biomass P, none of which are statistically significant. Moving on to Phy, there is a weak positive correlation with Co, a very weak positive correlation with Fungi (10), and a very weak negative correlation with Nematode (10). None of these correlations are statistically significant, suggesting that variations in Phy levels are not reliably associated with changes in Co, Fungi (10), or Nematode (10) levels. Phosphatase displays strong positive correlations with Cu, a strong negative correlation with Fe, and a weak positive correlation with Fungi (10), all of which are statistically significant. These findings indicate that as Pho levels increase, Cu levels tend to increase, Fe levels tend to decrease, and there is a weak positive association with Fungi (10) levels.
The analysis concludes with correlations involving Biomass P, showing weak positive correlations with Co and Tsd, and a marginal positive correlation with Phy. None of these correlations are statistically significant, suggesting that variations in Biomass P levels are not reliably associated with changes in Co, Thiosulfate dehydrogenase, or Phy levels. Finally, the correlations between various enzymes Dsr, Tsd, Phy, and Pho are explored, revealing weak and very weak correlations, with only the correlation between Tsd and Dsr being marginally statistically significant.
The correlation coefficients and their empirical coupling regression equations between the amounts of micronutrients including enzymes and increasing soil depths at the charcoal production sites (CPS) and non-charcoal production sites (NPS) are summarized in Table 6. Copper and cobalt were found to be positively correlated with increasing soil depths at both sites in contradistinction to iron and manganese which were negatively correlated with increasing soil depths. Zinc was negatively and positively correlated with increasing soil depths at the charcoal production site and non-charcoal production site, respectively. There was a highly significant correlation between the amounts of Pho and increasing soil depth at both sites. The correlation between Phy and soil depth at both sites was practically zero. Both Tsd and Dsr were positively correlated with increasing soil depth at the charcoal production site, while the reverse is the case at the non-charcoal production site. On the contrary, no significant relationship was found for Phytase (Phy) for both CPS and NPS. A simple linear correlation and regression analysis between enzymes, micronutrients (Y), and increasing rates of soil depth (X) showed significant positive and negative relationships (Table 7). In Ìrèle, regressing enzyme and micronutrient parameters (Y) against increasing soil depth (X) of Ìrèle indicated highly significant (P ≤ 0.001) negative relationships for Tsd, Dsr, Fe, and Zn. In addition, a highly significant positive relationship was found for Pho, Co, and Mn. Moreover, Cu showed a negative relationship with a correlation coefficient of 0. Furthermore, the regressing enzyme and micronutrient parameters (Y) against increasing soil depth (X) of Òkè-Àkò indicated highly significant (P ≤ 0.001) positive relationships for Phy, Dsr, Co, Fe, and Mn whereas a negative relationship found for Pho, Tsd, Cu, and Zn. Lastly, for Ìpaò, all four enzymes showed a highly significant positive relationship likewise all but Co, Fe, and Mn showed a negative linear correlation and regression relationship against increasing soil depth (Table 7).
The biomass phosphorus content in the designated areas exhibited its highest levels at CPS in the 0-15 cm soil depth in Ìrèle, Òkè-Àkò, and Ìpaò Fig. 1. Subsequently, it gradually declined as the soil depth increased across the profile. Conversely, at NPS, the biomass phosphorus content demonstrated higher patterns at the 0-15 cm depth and reached its peak at the 15-30 cm soil depth in Òkè-Àkò. Then it further decreased with increasing depth across the soil profile. In the study area, the bacteria activity displayed a significant increase in the 0-15 cm soil depth, followed by a gradual decline across the three locations Fig. 2. Specifically within the NPS, Òkè-Àkò and Ìpaò exhibited significantly higher bacteria activity at a depth of 15-30 cm. At the 30-45 cm soil depth, Òkè-Àkò registered the highest bacteria activity, while Ìpaò showed the lowest. Conversely, at CPS, bacteria were relatively abundant at the 0-15 cm depth in all three locations and at the 15-30 cm depth in Ìrèle; thereafter, it progressively decreased with an increase in depth. Fungal activity was found to be abundant at CPS across various soil depths and in all three locations Fig. 2. Furthermore, it was observed to intensify with increasing soil depth at CPS. It was significantly higher at 0-15 cm in Ìpaò, at 30-45 cm in Ìrèle, and reached its peak at 0-15 cm soil depth in Òkè-Àkò. Conversely, at NPS, Fungi exhibited a different pattern, with a significant increase at the 15-30 cm soil depth across the three locations. Additionally, at the 30-45 cm depth, fungal activity was notably lower at Ìrèle and Ìpaò. 4Nematode abundance was found to be significantly higher at natural production sites (NPS) compared to charcoal production sites (CPS) in soil layers of 0-15 cm, 15-30 cm, and 30-45 cm across all three locations Fig. 2 . However, there was less variation in nematode abundance at the 0-15 cm soil depth, unlike Ìrèle, where the highest abundance was observed at the 15-30 cm soil depth at NPS. Similar patterns were observed in Òkè-Àkò and Ìpaò.
Principal Component Analysis (PCA) analysed the relationship between a number of biological, chemical and enzymatic variables at charcoal production sites (Fig. 3). The PCA variables include biomass, bacterial and fungal activities, nematode population, phosphatase and phytase activities, sulfur cycle enzyme activities including thiosulfate dehydrogenase and dimethyl sulfoxide reductase as well as heavy metals such as Cu, Co, Fe, Mn and Zn. The first principal component, PC1, which accounted for 38.28% of the total variance, was mainly related to variables which are related to general biological activity for instance microbial biomass P and bacterial activity. This component may therefore be a proxy for the site nutrient status and general biological activity and thus areas with higher values of these variables will be more biologically active and richer in nutrients.
The second principal component (PC2) for 22.09% of the variance represented various ecological factors mainly affected by nematode activity and fungal distribution. This shows that there is a combination of different factors that work together or in contrast with each other to determine the soil ecosystem, as represented by the directions of their loadings. This can be to show balances or conflicts that are occurring within the ecosystem for instance competition for resources (Fig. 3a). The first principal component was mainly related to biological production and the dynamics of nutrients. The areas with high biomass and bacterial activity for instance Irele had high PC1 scores indicating that they had nutrient rich and biologically active soils. The second principal component was mostly affected by biophysical parameters including nematodes and fungi.
Places like Ipao showed opposite scores for PC2, which might be related to the soil's microbial balance or other environmental stresses. Oke-Ako had lower scores on PC1 and PC2, reflecting degraded or nutrient-poor conditions, probably as an influence of lower microbial and enzymatic activities (Fig. 3b). The same plot also showed clear clustering associated with shallower depths of 0-15 cm (Fig. 3c). The depths positioned high with PC1 scores are strongly interrelated with higher biomass P, bacterial activity, and enzymatic contribution, referring to nutrient-rich and biologically active soils. The intermediate zone between the extremes in both factors PC1 and PC2 was dominated by depths within the 15-30 cm range; it expresses moderate biological productivity and enzymatic activity. While the higher soil depth between 30-45 cm showed lower scores on PC1 and a moderate score on PC2, these depth levels recorded reduced biological activities and nutrient cycling, probably as a result of a reduction in microbial populations and availability of organic matter.
Clusters are colour-coded, with samples from Cluster One marked in green, Cluster Two in blue, and Cluster Three in red. The red dotted line at a distance of 5 represents the threshold line, which is the level at which the dendrogram has been cut to define these three clusters. This display gives an overview of the grouping of the samples and how each cluster is defined based on the hierarchical analysis. Samples index 1, 2, 4, 7, 16, and 17 in Table 8 belong to Cluster One; 3, 5, 6, 8, 9, 10, 11, 12, 13, 14, and 15 to Cluster Two; and Cluster Three contains sample 18 only (Fig. 4a). This representation spreads the samples over three clusters and hence provides a more orderly overview of their falling in a group based on the analysis. Cluster One is dominated by high microbial biomass P and bacterial counts, indicating a biologically active environment with probably a good level of nutrient supply. High fungal activities further support ecological functions such as decomposition and nutrient cycling in the cluster. Although moderate, nematode activities point toward a healthy soil ecosystem which could be potentially beneficial for plant growth and soil structure. Cluster Two shows a more moderate level of microbial biomass P and bacterial count, probably reflecting less nutrient-rich or disturbed soils than in Cluster One. The fungal activities still support certain ecological roles, but the general ambient conditions may be less than perfect for strong microbial activities to take place. The rate of nematode activities remains similar to that of Cluster One, showing balanced ecological interactions. Cluster Three has only one sample with notably lower microbial biomass P and fungal activities, which may indicate a probably stressed environment with limited ecological functions. Higher nematode activity could be indicative of stress responses or dominance of certain nematode groups that may be detrimental or indicative of conditions that need further investigation. Thus, Cluster One is the richest and most biologically active, hosting thriving ecosystems; Cluster Two represents more moderate ecological conditions, possibly requiring some management to enhance biological activity. Cluster Three may need detailed investigation on account of its potential environmental stress or unique conditions.
Figure 4b Provides HCA of enzyme activities. Colour-coding of the samples is identical as in Fig. 4a. Cluster One includes samples 6, 10 and 11. These samples are grouped together because of their similar level of at least one of the assayed enzymatic activities and may thus perform similar biochemical functions in similar environmental conditions. Samples 1, 2, 3, 4, 5, 7, 8, 9, 12, 13, 14, 15, 16, 17 were grouped in Cluster Two. This is the biggest cluster, indicating general similarity across a wide range of samples, possibly reflecting a common environmental background or similar metabolic activity level. At the same time, Cluster Three contained only one sample of 18. It stands alone in its cluster, reflecting unique enzymatic activity profiles that are considerably different from all other samples. Such could indicate either an outlier condition or a different environmental or physiological state. These clusters divide the samples according to their enzymatic activities and allow drawing some conclusions about ecological or biological similarities and differences. Cluster One had moderate phosphatase but very high thiosulfate dehydrogenase activity, suggesting a strongly sulfur-cycling-capable environment that may point toward enrichment in either sulfur or organic matter. The high levels of dimethyl sulfoxide reductase possibly suggest adaptation to oxidative stresses or environments with a high organic sulfur compound (Fig. 4b). The largest cluster, Cluster Two, exhibits high phosphatase activity, suggesting nutrient-rich conditions that are favourable for phosphorus cycling, important to biological growth and soil health. Lower levels of thiosulfate dehydrogenase and dimethyl sulfoxide reductase suggest less specialization in sulfur compound metabolism and may represent a more typical terrestrial or soil environment with high moisture content. Cluster Three contains one outlier sample that has very low phosphatase and thiosulfate dehydrogenase activities but very high phytase activities. Such a profile may indicate a special ecological niche or physiological state where phosphorus is cycled in a peculiar way, possibly in waterlogged soil or in a specialized microbial community focused on organic phosphorus mobilization.
For the micronutrients, Cluster One represents samples 3, 7, 9, 14, 15, and 16 (Fig. 4c). Samples which could potentially show similarities regarding micronutrient concentration within the studied areas fall within this cluster. Cluster Two consists of samples 1, 2, 4, 5, 6, 8, 10, 11, 12, 13, and 17. Since it is the largest cluster, these samples share more similarities across the measured micronutrients concentrations and may indicate that these samples have a common environmental or geological background. In Cluster Three, only one sample, 18, was found to stand alone in its cluster, indicating a uniqueness of micronutrient concentration profiles which are significantly different from other samples. This could imply another kind of environmental or contamination scenario. The cluster also showed a high and moderate level of the following metals: copper, iron, and manganese. It might suggest samples pertaining to areas influenced either by natural mineral deposits or those places with activities relating to charcoal production in moderation. For example, iron and manganese are common in both natural deposits and as industrial byproducts through combustion and mining. Cluster Two is the largest cluster and tends to have generally lower values of Cobalt and Zinc compared to the other clusters. These samples are probably representative of typical background levels from less contaminated situations. The low metal concentration suggests a minimal influence due to charcoal production or less exposure to pollution sources. Cluster Three is dominated by extremely high levels of Cobalt and Manganese, far above the other clusters. This cluster indicates a special environmental setting, possibly near mining sites or areas with geological anomalies. High levels of Cobalt and Manganese can be linked to specific types of mineral deposits or industrial contamination.