**3.1. Water quality index**

For the Chalma‐Tembembe River, mean values of WQI fluctuated from 52 to 74 (**Figure 2**), from slightly polluted to acceptable for human consumption; however, no significant differences between study sites (*p* > 0.05) were observed. In Apatlaco River, the study sites Las Truchas and El Texcal achieved excellence in water quality with mean values of 96 and 84, respectively (**Figure 2**) and were significantly different from other study sites (*p* < 0.05). The other study sites showed mean WQI values from 57 to 60 with no statistical differences (*p* > 0.05) and water quality fluctuated from polluted to mildly polluted for human consumption. Additionally, the WQI values decreased during the dry season (*p* < 0.05), where August‐ September differed from February‐March and June seasons; and December was statistically different from February‐March and June seasons (**Figure 2**).

**Figure 2.** Water quality index.

### **3.2. Physicochemical quality and bioindication values**

The factor analysis showed a total of 60.32% of explained variance for the first two axes. The parameters that showed a significant correlation, either positive or negative, in the first two axes of the factor analysis and which were considered as qualifying variables for this study, *Ci***2**, were NO3, NO2, NO3, TN, TP, BOD5, DO, DO saturation (%), SO4 2−, color, alkalinity, chlorides and conductivity (**Table 1**). Using these variables to assess the physicochemical quality, we found that five study sites of the Chalma‐Tembembe River obtained a *Pcq*, corresponding to classes 6–7 and one site in classes 7–8. The Apatlaco River sites were distributed across a broad spectrum of physicochemical quality classes: one site in classes 0– 1, another in classes 2–3, four sites in classes 3–4, six sites in classes 4–5, one in classes 6–7 and finally one in classes 9–10. A total of 66 taxa were taxonomically determined, distributed in 5 phyla, 7 classes, 21 orders and 63 families: Oligochaeta, Hirudinea and Turbellaria were identified only to class level. "Las Truchas" was the site that reached the maximum taxa (35).


The bioindication values for each aquatic macroinvertebrates family (obtaining the fifth percentile of the abundance class distributions along the *Pcq* intervals) are shown in **Table 2**.

Numbers in bold represent significant variables from the factor analyses.

other study sites showed mean WQI values from 57 to 60 with no statistical differences (*p* > 0.05) and water quality fluctuated from polluted to mildly polluted for human consumption. Additionally, the WQI values decreased during the dry season (*p* < 0.05), where August‐ September differed from February‐March and June seasons; and December was statistically

The factor analysis showed a total of 60.32% of explained variance for the first two axes. The parameters that showed a significant correlation, either positive or negative, in the first two axes of the factor analysis and which were considered as qualifying variables for this study,

chlorides and conductivity (**Table 1**). Using these variables to assess the physicochemical quality, we found that five study sites of the Chalma‐Tembembe River obtained a *Pcq*, corresponding to classes 6–7 and one site in classes 7–8. The Apatlaco River sites were distributed across a broad spectrum of physicochemical quality classes: one site in classes 0– 1, another in classes 2–3, four sites in classes 3–4, six sites in classes 4–5, one in classes 6–7 and finally one in classes 9–10. A total of 66 taxa were taxonomically determined, distributed in 5 phyla, 7 classes, 21 orders and 63 families: Oligochaeta, Hirudinea and Turbellaria were identified only to class level. "Las Truchas" was the site that reached the maximum taxa (35).

2−, color, alkalinity,

*Ci***2**, were NO3, NO2, NO3, TN, TP, BOD5, DO, DO saturation (%), SO4

different from February‐March and June seasons (**Figure 2**).

**Figure 2.** Water quality index.

46 Water Quality

**3.2. Physicochemical quality and bioindication values**

**Table 1.** Variables used for the calculation of *Pcq* and their correlations with the first two factors F1 and F2 (explained variance).

### **3.3. BMWP index and BMWP water quality classes**

The BMWP values assessed with the calibrated bioindication values fluctuated from 2 to 109, but showed no statistical differences (*p* > 0.05) between seasons. The study sites "Las Truchas" obtained the highest BMWP values (*p* < 0.05) during the four study periods. It was selected as the reference site and their scores were averaged. The BMWP scores above this median value will correspond to the "Excellent" quality category, from which range values for all water quality classes were assigned (**Figure 3a**). Seasonal fluctuations showed that this site fluctuate from "Regular" in August, "Good" in June to "Excellent" in December and February. The study site "El Encanto" had the lowest BMWP index value during its four study periods, from 2 to 6 points and thus qualifying as "Very bad, extremely polluted"; this was followed by La Gachupina, Buenavista 1, Buenavista 2, El Texcal and El Pollo, with values between 2 and 43 points and thus reaching the "Bad, polluted" and "Bad, very polluted" categories. The remaining sites qualified as "Regular, medium pollution" (**Figure 3a**). A comparison of BMWP and WQI scores for each study site is shown in **Figure 3b**.


**Table 2.** Calibrated bioindication values for the aquatic invertebrates of the Apatlaco and Chalma‐Tembembe Rivers.

The multiple linear regression equation of the quality test for predicting the BMWP scores is presented below (with *r* = 0.935).The adjusted line of BMWPobserved *vs* BMWPexpected indices attained *R*<sup>2</sup> values of 0.874 (**Figure 4a**); furthermore, all points related to the study sites were distributed within the confidence interval (*α* = 0.05). The Nash and Sutcliffe efficiency model was of 0.87, demonstrating a well‐fitted model:

Calibrating and Validating the Biomonitoring Working Party (BMWP) Index for the Bioassessment of Water... http://dx.doi.org/10.5772/66221 49

**Taxon Bioindication value**

Scathophagidae 8

10

7

5

4

2

1

Lepidostomatidae

Philopotamidae Pseudothelphusidae Ptilodactylidae Saldidae Scirtidae

Hydrobiidae Hydroptilidae Leptophlebiidae Polycentropodidae Pyralidae Thiaridae

Leptohyphidae Physidae Planorbidae Psychodidae Sphaeriidae Stratiomyidae Turbellaria

Corydalidae Elmidae 3

Libellulidae Tabanidae Tipulidae

Muscidae Nepidae Notonectidae Oligochaeta Simuliidae Syrphidae

was of 0.87, demonstrating a well‐fitted model:

**Table 2.** Calibrated bioindication values for the aquatic invertebrates of the Apatlaco and Chalma‐Tembembe Rivers.

The multiple linear regression equation of the quality test for predicting the BMWP scores is presented below (with *r* = 0.935).The adjusted line of BMWPobserved *vs* BMWPexpected indices attained *R*<sup>2</sup> values of 0.874 (**Figure 4a**); furthermore, all points related to the study sites were distributed within the confidence interval (*α* = 0.05). The Nash and Sutcliffe efficiency model

Perlidae

Cordulegastridae Heptageniidae

Caenidae Gyrinidae

48 Water Quality

Aeshnidae Blaberidae Cambaridae Helicopsychidae Hydrobiosidae Naucoridae

Corbiculidae Dixidae Dryopidae Ephydridae Glossosomatidae Gordiidae

Ancylidae Asellidae Calopterygidae Gomphidae Hirudinea Hyallelidae Hydropsychidae

Belostomatidae Coenagrionidae Hebridae Staphylinidae

Baetidae Chironomidae Corixidae Culicidae Dytiscidae Hydrophilidae Lestidae

**Figure 3.** Scores of the BMWP. A: Excellent; B: Good, not sensible affected; C: Regular, moderate pollution; D: Bad, pol‐ luted; E: Bad, very polluted and F: Bad, extremely polluted. (a) Scores by period and study sites. (b) Scores of the BMWP and WQI.

**Figure 4.** Validation model. (a) BMWP observed *vs* expected. (b) BMWP observed *vs* expected including validation da‐ ta. (c) BMWP observed and expected values for the index validation. (d) BMWP observed and expected values for the regional extrapolation sites.

=- + - + - + -- + - - expected 5 BMWP [138.854 7.69 Nitrates 123.51 Nitrites 0.18 Ammonia 2.17 Total N 33.11 Total P 0.80 Sulfates 1.18 Color 0.14 Alkalinity 0.28 Chlorides 1.85 BOD 24.96 DO 0.18 Conductivity 2.19 %DO]. (3)

### **3.4. Index validation and regional extrapolation**

The BMWP values for their validation and regional extrapolation (**Figure 5**) span the whole range of water quality classes: from "Very bad, extremely polluted" (Tlatenchi) to "Excellent" (Arriba Chalchihuapan).The observed and expected BMWP values for the nine index validation sites in the Apatlaco River and the seven regional in the neighboring river subbasins (Amacuzac, Cuautla and Yautepec) were calculated using the previously BMWP scores calibrated and the derived multiple linear regression model. Four sites of the BMWP values lay outside the confidence limits (*α* = 0.05) for the BMWP linear regression models (**Figure 4b**). Pearson's correlations were calculated (*α* = 0.05) for the BMWP observed and expected values for the index validation and regional extrapolation sites (**Figures 4c** and **d**). As in the previous analysis, acceptable (*p* = 0.048) values were calculated for the index validation, whereas the regional extrapolation showed weaker correlation (*p* = 0.091). Three out sites of the nine independent sites were outliers for the calculated multiple linear regressions BMWP model (**Figures 4c** and **d**).

**Figure 5.** Scores of the BMWP for the validation and range extension study sites. A: Excellent; B: Good, not sensible affected; C: Regular, moderate pollution; D: Bad, polluted; E: Bad, very polluted and F: Bad, extremely polluted.

The index validation, adding nine independent sites, validated the regression model as a satisfactory indicator for river water quality in the Apatlaco River for the BMWP index (*r* = 0.67, *p* = 0.048). For the combined nine sites, there was a positive significant correlation for the BMWP index (**Figure 4a** and **b**).
