The influence of water quality on the variation patters of benthic macroin-vertebrate communities in the lakes in the central highlands of Peru was eva-luated. Samples of water and sediments were collected in 23 different sam-pling sites last 2017. The physiochemical variables of water quality deter-mined on site were: DO, TDS, EC, temperature and pH. The results obtained revealed that the physiochemical indicators are within the environmental quality standards for water, except COD and BOD 5. Regarding the benthic macroinvertebrates, four phyla were identified wherein the most common is the phylum Arthropoda having the abundance and richness of taxa. The PCA reduced the variables to a few significant components that caused variation in water quality between lakes. The cluster analysis in relation to the relative abundance of benthic macroinvertibrates grouped the 22 sampling sites into three groups with the similar characteristics. The PCoA analysis of the ben-thic macroinvertebrate communities showed a clear separation of sites. The SIMPER analysis at the family-level showed the distribution of the most common species. Therefore, at a significance level of 0.01 it demonstrates that there are significant differences between the number of species and abun-dance of the areas that were evaluated.
The growing shortage of water resources and their impact of the functioning of ecosystems, make water one of the main objects of environmental protection [
In many developing countries, the use of water resources faces socio-environmental conflicts due to the growing deterioration that aquatic ecosystems are experiencing [
Currently, new tools are being generated to evaluate the state in which continental aquatic ecosystems are found, in order to reduce the intrinsic uncertainties and subjectivities of the environmental problems that will be experiencing [
The watershed of the Mantaro River is located in the central highlands of Peru, between, 10˚39' and 13˚30' South latitude and between 74˚00' and 76˚30' West longitude, with altitudes ranging between 500 to 5350 meters above sea level. It covers an area of approximately 34550.08 km2 and forms part of the slope of the Atlantic Ocean, covering the provinces of Pasco, Junin, Huancavelica and Ayacucho. The Mantaro River originates in Lake Junin and flows 735 kilometers to its confluence with the Apurimac River. As a result, the study area has a very frigid climate with an average annual temperature above 0˚C and below 7˚C, a maximum temperature of 15˚C and a minimum temperature of 9˚C. It has an annual rainfall of 920 mm with violent thunderstorms. Its flow depends on rainfall throughout the watershed, the level of the Lake Junin and the lakes located at the foot of the mountains in the western mountain ranges and the snowy Huaytapallana. The lakes that are included in this study are: Pomacocha (Pc), Tragadero (Tg), Cuncancocha (Cc), Incacocha (Ic) and Ñahuinpuquio (Ñn) (
Water sampling was carried out in five lakes of the central highlands of Peru, on 2017. In each lake, 23 sampling sites were defined and each of them were determined in situ: dissolved oxygen (mg/L), total dissolved solids (mg/L), conductivity (µS/cm), temperature (˚C) and pH, using Hanna Instruments multiparameter probes (HI 991301 Microprocessor pH/temperature, HI 9835 Microprocessor Conductivity/TDS and HI 9146 Microprocessor dissolved oxygen), previously, the equipment was calibrated at the respective sampling site. Also, 1L of water from a depth of 20cm was also collected from each sampling site for bacteriological analysis and analysis of nitrates, total phosphorus, COD and BOD5, in containers previously mixed with a hydrochloric acid solution in proportion 1:1 and rinsed with distilled water. The measurements of these parameters were carried out according to the standard methods [
The collection of sediments to obtain benthic macroinvertebrates was carried out at the sites defined by Ekman-Birge dredger from Hydro-Bios. At each sampling site, four sediment samples were taken. Then, 5% formaldehyde was added to each of the samples for preservation and subsequent observation and identification in the laboratory. The taxonomic identification of the benthic macroinvertebrates was carried out at the family level through a trinocular stereomicroscope, as they occur in the bioevaluation studies of freshwater ecosystems [
The analysis of the physicochemical variables of the water and the benthic macro invertebrate communities were analyzed using the software PRIMER―E v7 and MINITAB v 18. Because of the anthropogenic intervention in the evaluated lakes can affect the distribution of benthic macroinvertebrate communities, the physicochemical variables of the water were analyzed according to the standard of principal components analysis (PCA) in order to generate two-dimensional
management maps [
In the cluster analysis of benthic macro invertebrate communities, a hierarchical and agglomerative classification was performed, generating a similarity matrix with Bray-Curtis indices base on an abundance matrix of transformed species by square root in order to produce a dendrogram [
- Shannon diversity index (H'), as follow: H ′ = − ∑ i = 1 S p i log 2 p i , where pi is the proportion of characters belonging to the ith type of letter in the string of interest. In ecology, pi is often the proportion of individuals belonging to the ith species in the dataset of interest.
- Margalef diversity index (d) using for calculate the species richness; d = ( S − 1 ) / L o g ( N ) ; where s is the number of species present, and N is the total number of individuals found (belonging to all species). The notation Ln denotes the Neperian logarithm of a number.
- Simpson index (1 − λ) for measure the degree of concentration when individuals are classified into types, 1 − λ = 1 − ∑ i = 1 R p i 2 ; where R is richness (the total number of types in the dataset). The equation is also equal to the weighted arithmetic mean of the proportional abundances pi of the types of interest, with the proportional abundances themselves being used as the weights.
The main indicator species and the associated percentage of indication were determined for each significant set of species using the similarity percentage (SIMPER) [
The BIOENV analysis [
In
Indicator | Pc | Tg | Cc | Ic | Ñn |
---|---|---|---|---|---|
pH | 6.84 ± 0.13 | 7.56 ± 0.42 | 7.16 ± 0.25 | 8.61 ± 0.19 | 7.63 ± 0.06 |
EC (µS/cm) | 254.24 ± 3.03 | 269.81 ± 7.50 | 260.35 ± 4.39 | 262.53 ± 4.02 | 258.77 ± 4.04 |
COD (mg/L) | 31.09 ± 0.92 | 59.64 ± 0.23 | 46.88 ± 1.25 | 38.83 ± 0.31 | 34.52 ± 0.40 |
BOD5 (mg/L) | 18.07 ± 0.39 | 28.81 ± 0.68 | 18.68 ± 0.87 | 18.51 ± 0.37 | 20.80 ± 0.40 |
DO (mg/L) | 6.68 ± 0.11 | 6.12 ± 0.07 | 6.49 ± 0.16 | 6.43 ± 0.05 | 7.12 ± 0.08 |
Temperature (˚C) | 11.74 ± 0.78 | 17.70 ± 0.16 | 11.38 ± 0.35 | 10.51 ± 0.41 | 10.61 ± 0.37 |
TDS (mg/L) | 3.09 ± 0.07 | 9.20 ± 1.39 | 2.58 ± 0.07 | 2.63 ± 0.0.19 | 2.75 ± 0.06 |
Total phosphorus (mg/L) | 0.042 ± 0.006 | 1.519 ± 0.036 | 0.802 ± 0.019 | 0.500 ± 0.038 | 0.740 ± 0.008 |
Nitrates (mg/L) | 0.396 ± 0.023 | 0.768 ± 0.029 | 0.466 ± 0.003 | 1.506 ± 0.023 | 1.555 ± 0.037 |
Total iron (mg/L) | 0.034 ± 0.001 | 0.044 ± 0.000 | 0.033 ± 0.000 | 0.017 ± 0.000 | 0.038 ± 0.000 |
Total copper (mg/L) | 0.002 ± 0.000 | 0.005 ± 0.001 | 0.002 ± 0.000 | 0.002 ± 0.000 | 0.002 ± 0.001 |
Total lead (mg/L) | 0.000 ± 0.000 | 0.001 ± 0.000 | 0.000 ± 0.000 | 0.001 ± 0.000 | 0.000 ± 0.000 |
Total zinc (mg/L) | 0.000 ± 0.000 | 0.001 ± 0.001 | 0.000 ± 0.000 | 0.000 ± 0.000 | 0.000 ± 0.000 |
(60.1%) correlated positively and significantly with the temperature, the concentration of total dissolved solids, lead, copper, BOD, COD and total phosphorus, and negatively with dissolved oxygen and nitrates. On the other hand, the second component (17.4%) correlated positively and significantly with total iron and negatively with nitrate concentration and its pH. Therefore, the first component would be indicating an anthropogenic pressure gradient, especially in Tg Lake, while the second component could be related to the seasonal variability of water characteristics. For example, at a level of significance of 0.05, the pH showed a high positive correlation with nitrates (0.80 Pearson rank) and negative correlation with total iron (−0.58), the COD correlated positively with BOD (0.87), temperature (0.78), total dissolved solids (0.78), total phosphorus (0.92), total copper (0.81), lead (0.84), and zinc (0.68) and, it was negatively correlated with the DO (−0.73).
We counted 1307 individuals of benthic macroinvertebrates grouped in 12 taxa. According to the point of partition, at 63% similarity, the cluster analysis grouped the 23 sampling sites into three different groups. The sampling sites of Tg Lake showed a greater dissimilarity with respect to the rest of the lakes. In relation to relative abundance, the Pc, Ñn and Ic Lakes generate a cluster with a minimum similarity of 70%, differing from the one generated by the Cc Lake at a level of similarity of 62%, making the cluster analysis allow in generating a discriminating factor (
The non-metric multidimensional scaling analysis shows an average value of stress level of 0.13 that according to the range given by Kruskal indicates an acceptable interpretation in the perceptual map, since the observation would replicate the original distributions fairly. The principal coordinate analysis (PCoA) of the composition of the benthic macroinvertebrate communities (
from the others in a significant way. The groups b and c, have a higher similarity value but still are significantly different in relation to the number of species and abundance of macro invertebrates. The grouping of these groups is categorized by a range of similarity of 63%. In addition, the PCO shows that in group a there is up to 80% similarity and in groups b and c with 63% similarity.
The SIMPER analysis at the family level shows the distribution of the most common species, and confirms that the clustering by the cluster analysis is addressed by the diversity indices, for example, it is observed that group b is the most important in terms of diversity (
The PERMANOVA analysis of the benthic macroinvertebrates community is divided according to the factor of the groups of species, reveals that at a significance level of 0.01, there are significant differences between the number of species and abundances of the evaluated zones; and is that at least one of the gaps
Species Assemblages | Taxon | Simper | S | N | d | H' | 1 − λ |
---|---|---|---|---|---|---|---|
a | Chironomidae | 51.23% | 7 | 468 | 0.98 | 1.134 | 0.478 |
Psychodidae | 21.44% | ||||||
b | Chironomidae | 38.84% | 11 | 635 | 1.55 | 1.768 | 0.231 |
Ceratopogonidae | 25.06% | ||||||
Hydrophilidae | 18.39% | ||||||
c | Chironomidae | 53.50% | 6 | 167 | 0.98 | 1.2 | 0.4 |
Ceratopogonidae | 34.44% |
S, number of species; N, number of individuals; d, Margalef index; H', Shannon index; 1 − λ, Simpson’s index.
differs from the others. In addition, using the zone of study factor in the comparison analysis of pairs, the Tg Lakes shows significant differences in comparison with the other lakes, in relation to the number of species and abundances. This difference is much more significant compared to Pc Lake, but comparing the Pc and Ic Lakes the results reveal that these are statistically similar. However, the zones with a high level of similarities are the areas of Cc, Ic and Ñn.
The BIOENV analysis revealed that the variations in the distribution of the benthic macroinvertebrates of the high Andean lakes can be better explained taking into account the 13 water quality variables considered in the study (
We used the model based on linear distances which analyzes and models the relationship between a multivariate data cloud and one or more predictor variables, with several options for model selection. The coefficient of determination obtained allows us to understand the relationship between these data. By testing all the predictor variables, the marginal test indicates that the BOD, COD, DO, dissolved total solids, total phosphorous, iron, copper, lead and total zinc are predictors that explains the distribution of the benthic macro invertebrates at a level of significance of 0.05. While the variables of conductivity, nitrates and pH, do not show significant differences. The sequential test shows that the most important predictors are the BOD and the total phosphorus, since they explain with a higher proportional index to the distribution of the biological observations at 0.01 significance level. Using these two predictors, the coefficient of determination R2 shows a value of 0.45, which is acceptable to reflect the goodness of fit of a model to the variable that it is intended to explain (
The result of the redundancy-dbRDA analysis of the physicochemical variables of water and the relative abundances of benthic macroinvertebrates is presented in
The results obtained from the physicochemical evaluation of the water quality of high Andean lakes reveal some deterioration, which is probably caused by the increase of anthropogenic activities as a result of the accelerated urban expansion in the region. The highest values of conductivity, COD and BOD recorded in Lake Tg are due to the presences of ions and anions that are concentrated due to evaporation matter in water [
N˚ of variables | Correlation | p-value (0.05) | Selections |
---|---|---|---|
1 | 0.574 | 0.004183 | Total phosphorus (mg/L) |
2 | 0.609 | 0.002041 | BOD5 (mg/L), Total phosphorus (mg/L) |
3 | 0.611 | 0.001954 | COD (mg/L), BOD5 (mg/L), Total phosphorus (mg/L) |
4 | 0.614 | 0.00183 | COD (mg/L), BOD5 (mg/L), Total phosphorus (mg/L), Total copper (mg/L) |
5 | 0.614 | 0.00183 | COD (mg/L), BOD5 (mg/L), Total phosphorus (mg/L), Total copper (mg/L), Total lead (mg/L) |
MARGINAL TESTS | ||||
---|---|---|---|---|
Variable | SS (trace) | Pseudo-F | p-value | Decreasing Prop. |
BOD5 (mg/L) | 5671.3 | 10.78 | 0.001 | 0.339 |
Phosphorus total (mg/L) | 5531 | 10.382 | 0.001 | 0.331 |
Total copper (mg(L) | 5200 | 9.4801 | 0.001 | 0.311 |
Total lead (mg/L) | 5203.6 | 9.4897 | 0.001 | 0.311 |
COD (mg/L) | 4821.1 | 8.51 | 0.001 | 0.288 |
Total dissolved solids (mg/L) | 4524.8 | 7.792 | 0.001 | 0.2706 |
Temperature (°C) | 4296.4 | 7.263 | 0.001 | 0.2569 |
Total zinc (mg(L) | 4001.9 | 6.6085 | 0.001 | 0.239 |
Total iron (mg/L) | 2762.6 | 4.1569 | 0.004 | 0.165 |
Dissolved oxygen (mg/L) | 2142.2 | 3.086 | 0.017 | 0.1281 |
Conductivity (S/cm) | 1219.4 | 1.652 | 0.145 | 0.07 |
Nitrates (mg/L) | 761.49 | 1.002 | 0.395 | 0.045 |
pH (Unity) | 742.87 | 0.976 | 0.425 | 0.04 |
The results obtained through the PCA reveals that the lakes which are under study have been experiencing a process of worsening water quality due to the strong anthropogenic pressure that they have been supporting. Being the Lake Tg that displays a quality of bad water according to COD, BOD and DO. The low DO concentration would be due to the consumption of this gas in the biodegradation processes, as revealed by the high concentrations of BOD registered in this lake. These results are supported by Shi et al. (2017) [
Temperature is one of the factors with a great importance in the development of the different processes that are carried out in water, such as the dissolution of oxygen, since it determines the tendency of its physical properties, and the richness and distribution of biological communities [
Phosphorus in aquatic environments limits the growth of algae and plants, because their determination can detect problems of eutrophication [
Aquatic ecosystems have been experiencing a progressive decline in their quality, affecting the aquatic life they harbor and the services they provide. In Peru, the monitoring of water quality in these ecosystems is limited to the use of physical and chemical determinations. The entities responsible for water management have not yet considered monitoring the quality of water in an integral manner, taking into account the composition and structure of the biological communities [
The results obtained show that the most abundant benthic macroinvertebrates correspond to individuals of the Insecta Class, order Diptera. They also show significant differences between the macroinvertebrate communities of the evaluated lagoons, being the Chironomidae family the most common with a wide range of distribution [
The results also indicate that each of the taxa responded differently on the environmental conditions of each lake. For example, group b comprised of the families Chironomidae, Ceratopogonidae and Hydrophilidae reached the greatest abundance and diversity. However, these families are considered resistant and resilient taxa due to the anthropogenic pressure that impoverish biological communities and consequently alter aquatic ecosystems [
This study used several multivariate statistical techniques to evaluate the influence of water quality on the variation patterns of the benthic macroinvertebrate communities in the central highlands of Peruvian lakes. The PCA reduced the variables to a few significant components that caused variation in water quality between the lakes. The cluster analysis in relation to the relative abundance of benthic macroinvertebrates grouped the 23 sampling sites into three groups with similar characteristics. The analysis of the main coordinates of the composition of the benthic macroinvertebrate communities showed a clear separation of sites, mainly between Tg and the other lakes. The SIMPER analysis at the family level shows the distribution of the most representative species, and confirms that grouping by cluster analysis is addressed by diversity indices.
The PERMANOVA analysis of the community of benthic macroinvertebrates divided according to the factor of groups of species reveals that at a level of significance of 0.01 there are significant differences between the number of species and abundances of the evaluated areas. The marginal test indicates that the BOD, COD, DO, total dissolved solids, total phosphorus, iron, copper, lead and total zinc are predictors that explain the distribution of the benthic macroinvertebrates at a level of significance 0.05. The DistLm analysis according to the sequential test reveals that BOD and total phosphorus are the most important physical variables with a coefficient of determination of 0.45. Therefore, these results are useful to establish a baseline for the detection of changes in ecological status and anthropogenic impacts.
The authors express their gratitude to the General Research Institute of the Universidad Nacional del Centro del Peru for the financing of the study, to the Water Research Laboratory for allowing us to make use of the equipment and materials for this study.
The authors declare that they have no conflict of interest.
Custodio, M. and Peñaloza, R. (2019) Influence of Water Quality on the Variation Patterns of the Communities of Benthic Macroinvertebrates in the Lakes of the Central Highlands of Peru. Open Journal of Marine Science, 9, 1-17. https://doi.org/10.4236/ojms.2019.91001