3.2. Occurrence of Glyphosate and AMPA
Levels of the herbicide glyphosate and its metabolite AMPA were measured in fresh water, sediment, and SPM samples from the Suquía River basin (Córdoba, Argentina). This monitoring program was carried out over 1 year (between 2010 and 2011), with sampling in five sites located along the river (
Figure 1), making a total of 60 samples. From all the analyzed samples, 35% contained glyphosate, AMPA, or both compounds (
Table 3). Glyphosate was detected at least once in samples collected in all the studied sites, whereas AMPA was detected at least once in four sites. In the LC site, the AMPA concentration was always below the LOD. The presence of the herbicide and/or its metabolite in all five sites is proof of its use all along the basin. This detection frequency is similar to that of the rural and urban freshwater surface in Ontario [
38]. With regard to the concentration of glyphosate and AMPA spread in the different environmental matrices studied, 61% was found in sediment, with glyphosate levels ranging from <LOD to 1882.3 µg kg
−1 and AMPA levels ranging from <LOD to 266.1 µg kg
−1. In total, 36% of glyphosate and AMPA was measured in SPM, with the range for glyphosate being between <LOD and 1570.7 µg kg
−1 and for AMPA from <LOD to 684.9 µg kg
−1. Only 3% of the total measured glyphosate and AMPA was in the water compartment, with levels ranging from <LOD to 125.0 µg L
−1 for glyphosate and from <LOD to 4.8 µg L
−1 for AMPA. Despite the high solubility of glyphosate and AMPA in water, our results show a high affinity of these compounds for SPM and sediment in contrast to water (12 and 20 times higher, respectively) [
39]. In order to explain the partition of pesticide between both matrices, a pesticide sediment-runoff partition coefficient (Kp) expressed as L kg
−1 was calculated. In the case of CM and RP, the Kp was 287 and 24 L kg
−1, respectively, denoting a high adsorption on sediments. Lupi et al. ([
40], pp. 687–694) reported similar results for glyphosate in Quequén Grande river (26 L kg
−1). The higher concentrations of glyphosate and AMPA found in sediment when compared to water could be also explained by the fact that these pollutants have lower possibilities for microbial decomposition when they are adsorbed to particulate matter [
41].
Glyphosate was detected in water samples at a maximum of 125.0 µg L
−1, which is equivalent to 166.7 µg L
−1 glyphosate isopropylamine salt. Thus measured levels did not exceed the limits established for freshwater aquatic protection (240 µg L
−1) according to the Argentinean Environmental Water Quality Guidelines [
42].
The tree diagram obtained by cluster analysis of glyphosate and AMPA concentrations in fresh water, sediment, and SPM, identified differences between the sampling sites (
Figure 2). It shows two clusters at 50% of the total distance, wherein the lower cluster represents only the CM site. At 25% of the total distance, three clusters are observed. The top clusters represent SR, RP, and LP, while LC follows, and lastly, there is CM. Thus, a high similarity among the three sites with extensive agricultural models can be observed, with less similarity to LC, the site without agricultural activities, and a clear difference to CM, the location with intensive crops.
The concentrations found at each site reflect that the most polluted sample location is CM (
Figure 3; Kruskall–Wallis
p < 0.05), with a maximum level of glyphosate plus AMPA of 127.2 µg L
−1 in water, 2148.4 µg kg
−1 in sediment, and 1570.7 µg kg
−1 in SPM, displaying a 50% detection frequency. The measured concentrations could be the result of the intensive use of this pesticide on the neighboring vegetable gardens. Even without significant differences, the sites that showed an intermediate contamination rate were RP, SR, and LP. The former had a maximum level of glyphosate plus AMPA of 4.8 µg L
−1 in water, 390.9 µg kg
−1 in sediment, and 684.9 µg kg
−1 in SPM, with a highest detection frequency result of 58%. In SR there was an absence of glyphosate and AMPA in water, with a maximum value of glyphosate plus AMPA of 335.4 µg kg
−1 in sediment and 473.5 µg kg
−1 in SPM, with a detection frequency of 42%. In LP neither glyphosate nor AMPA was found in water and SPM, but there was a maximum value of glyphosate plus AMPA of 472.1 µg kg
−1 in sediment, with a detection frequency of 17%. These three sites are situated in the main agriculture area of Córdoba state, where the cultivated sectors reach the riverbank.
The least polluted area was LC. In this site a maximum value of 70.0 µg L
−1 of glyphosate was measured in water, but the herbicide and its metabolite were absent in both SPM and sediments, with detection frequency of 8%, the lowest percentage in the Suquía River basin. However, the presence of glyphosate in the water of LC could be attributed to point sources, such as the cleaning of machinery on the riverbank, discarding of recipients with herbicide leftovers, or sporadic domestic usage, since no agricultural activities can be registered in the area (
Table 1).
Similar concentrations of glyphosate and AMPA were found in other agricultural areas of Argentina, such as in the northwest of the state of Buenos Aires [
20]. However, the concentrations here reported were higher than those measured in the southeast of the same province [
19]. Our results showed also higher concentrations of glyphosate and AMPA in water and SPM than those found in studies carried out in Paraná river and Mesopotamic Pampas agroecosystem, as reported by Ronco et al. ([
22], pp. 771–779) and Primost ([
21], pp. 771–779), respectively.
The concentration of glyphosate in relation to its metabolite AMPA indicates that time had passed from the moment the herbicide was employed [
43]. According to the results found in the Suquía River basin, in the LC site glyphosate was detected only in water and it can be assumed the contamination was recent due to the fact its metabolite was not found. In CM, the pollution was mainly produced by glyphosate and, in addition, it was spread in all the analyzed matrices. This fact would indicate a constant glyphosate runoff into the ecosystem.
In RP and SR, the highest concentration was that of the metabolite AMPA, with a greater percentage found in SPM and sediments. This situation could indicate that certain time had passed since the application of the chemical on those sites. In LP, it was observed that the highest concentration was present in the form of glyphosate, which indicates recent contamination, and besides, every positive measure of both the herbicide and its metabolite was found in the sediment. This outcome can be understood if the texture results of the sediment are taken into account. Of all the studied sites, the higher percentage of clay was found in LP, a component known to have a high affinity with the herbicide [
44]. With regards to the temporal distribution, during the present study no significant differences were found between low and high application periods (data not shown).
The horticultural productivity in the subtropical regions of the world is severely limited by pests and diseases affecting crops. As a result, the losses in the field and the reduction of the commercial values of the products make the horticultural business less profitable than expected. The fact that the quality of the products has become a priority worldwide has led to the generation of a group of quality standards in response to the demands of the consuming market. The main criterion used regarding this issue is related to the visual aspect related to shape, color, and the absence of damage [
45]. The use of agrochemicals is the most common method used for the control of pests, diseases, and the improvement in the quality of the products. Based on our results, the use of glyphosate in intensive crops provokes more pollution to the watercourse than the extensive crops.
3.3. Ecological Risk Assessment
In order to estimate the environmental risk associated to the presence of glyphosate and AMPA in areas with intensive and extensive crops, the MECs at five sampling points located along Suquía River and the PNECs obtained from the most sensitive species approach were used for the calculation of HQ (
Table 4). The HQ values in LC and CM indicate that there is significant risk to aquatic organisms associated with the presence of glyphosate in water. In a similar manner, the HQs obtained considering the herbicide content in sediments and the effects over benthic organisms were of 6.7 and 1.4 in CM and LP respectively, suggesting risk in sediment dwellers. HQs over 1 were calculated by Annett et al. ([
8], pp. 458–479) for fish species, using an MEC in a surface stream of Argentina [
20], and for aquatic microorganisms, invertebrates, and amphibians considering realistic environmental exposure concentrations in forest wetland measured by Thompson et al. ([
46], pp. 843–849). Benthic organisms in RP and SR would not be at risk of toxic effects due to the presence of glyphosate in sediments, since none of the calculated HQs reached the risk significance level of 1.
The HQs obtained for AMPA derived from available toxicity data in aquatic organisms and AMPA levels in fresh water suggest it is not a dangerous metabolite. However, there few toxicity studies of AMPA in aquatic organisms on which to base this assessment. Furthermore, it was not possible to calculate HQs in sediment since no toxicity data was available for this metabolite in benthic organisms.
The results of hazard assessment for glyphosate show that both aquatic and benthic organisms of the Suquía River basin are at risk in areas where intensive agricultural models are applied. Extensive crops seem to be dangerous only to benthic species. The presence of glyphosate in unexpected areas due to point sources of pollution also endangers aquatic species.