The concentrations of Pb, Cd, and Zn in the blood of 50 house sparrows from the five sites studied in the city of Meknes show significant variations depending on the activity of each site. In general, the distribution of heavy metals follows distinct trends for each metal. The highest concentrations of lead were observed in the industrial zone (IZ) and the Town centre (TC), with average values of 336.02 µg/L and 152.96 µg/L, respectively. In contrast, in the rural area (Ref), lead levels were on average ten times lower than those in the industrial zone. A similar pattern was observed for Cd, where industrial and urban areas showed high concentrations, with averages of 12.28 µg/L for IZ and 19.51 µg/L for TC, while the Ref area had the lowest levels, with an average of 1.83 µg/L. Regarding Zn, the highest concentrations were found in the industrial and rural areas, with averages of 1736.09 µg/L for IZ and 1386.53 µg/L for Ref. Medium traffic areas, such as the Fes-Meknes main road (MR) and the Sidi Said bus station (SS), showed intermediate concentrations, reflecting moderate zinc contamination due to traffic and human activities. Figure 2
The statistical analysis of the data for the three variables (Pb, Cd, and Zn) and 50 individuals (blood of house sparrows) was conducted using Principal Component Analysis (PCA). The correlation matrix of the different elements studied shows a close relationship between the various parameters (Table 1). This relationship is demonstrated by the correlation coefficients.
Table 1
Pearson Correlation Matrix
| Zn (µg/L) | Pb (µg/L) | Cd (µg/L) |
Zn (µg/L) | 1.000 | 0.292 | 0.020 |
Pb (µg/L) | 0.292 | 1.000 | 0.418 |
Cd (µg/L) | 0.020 | 0.418 | 1.000 |
The results indicate: |
-
A weak positive correlation between Zn and Pb (0.292), suggesting that these metals may have some common sources but their concentrations are not strongly linked.
-
An almost negligible correlation between Zn and Cd (0.020), indicating that their sources or mechanisms of accumulation are likely distinct.
-
A moderate positive correlation between Pb and Cd (0.418), implying that these two metals often co-occur, possibly due to shared sources of pollution such as industrial emissions or traffic.
The results of the PCA show that the first axis (PC1) explains 50.66% of the total inertia of the data, and the second axis (PC2) explains an additional 32.72% (Table 2 and Fig. 3). The projection plane PC1-PC2 shows that the three variables are well represented on the correlation circle (Fig. 4).
Table 2
Correlation between Variables and Principal Components.
| Eeigenvalue | Inertia |
PC1 | 1.520 | 50.66% |
PC2 | 0.981 | 32.72% |
PC3 | 0.498 | 16.62% |
The correlation circle shows the relationships between the variables (Zn, Cd, Pb) and the first two principal components (PC1 and PC2) of the principal component analysis (PCA). PC1 explains about 50.7% of the total variance and is mainly influenced by lead (Pb) and cadmium (Cd). PC2 explains about 32.7% of the total variance and is strongly influenced by zinc (Zn). Lead (Pb) is moderately correlated with PC1 and positively with PC2, while cadmium (Cd) is positively correlated with PC1 and negatively with PC2. Zinc (Zn) is strongly correlated with PC2 and very weakly with PC1. This representation shows that Pb moderately influences the first two components, while Zn and Cd have distinct influences on PC1 and PC2, respectively.
The main axis PC1 is strongly correlated with two variables, Pb and Cd (located on the positive side of the component), while the axis PC2 is correlated with Zn (located on the positive side of the component) (Table 3).
Table 3
Correlation between Variables and Principal Components.
| PC1 | PC2 | PC3 |
Zn (µg/L) | 0.513 | 0.812 | 0.278 |
Pb (µg/L) | 0.864 | -0.013 | -0.504 |
Cd (µg/L) | 0.715 | -0.567 | 0.410 |
The analysis of the projection of sparrows on the plane formed by the first two principal components (PC1 and PC2), as well as the results of the K-means clustering applied to the same data, is presented in Fig. 5.
The left plot shows each point representing a sampling site, differentiated by colors. Blood samples from sparrows at IZ and TC sites are mainly on the right side of PC1 and the upper side of PC2, indicating higher contamination levels. MR points are in the middle, SS points are in the lower left, and Ref points are on the left side of PC1 and the middle of PC2, indicating lower Pb and Cd but higher Zn levels..
The right plot shows the results of the K-means clustering. The optimal number of clusters was determined using the elbow method, where the sum of squared distances (SSE) was plotted against the number of clusters. The optimal number was chosen where the SSE starts to level off, forming an "elbow" shape, indicating three clusters as the best balance K = 3.
-
Cluster 0 (red): Sparrows mainly from rural areas (Ref), SS, and MR, with high levels of Zn and low levels of Pb and Cd.
-
Cluster 1 (blue): Sparrows mainly from industrial zones (IZ) and the city center (TC), as well as some from SS, with low to moderate levels of Zn, Pb, and Cd.
-
Cluster 2 (green): Sparrows mainly from industrial zones (IZ) and the city center (TC), with high levels of Zn, Pb, and Cd.
Samples from the industrial zone (IZ) and the city center (TC) tend to group in clusters 1 and 2, suggesting similar contamination profiles for lead, cadmium, and zinc. In contrast, samples from the rural site (Ref) stand out, forming a distinct cluster (Cluster 0) with high levels of zinc and low levels of lead and cadmium. These results provide a clear perspective on the spatial distribution of heavy metal pollution in the studied region.