Integrating Physicochemical and Benthic Macroinvertebrate Functional Traits to Assess Ecological Status of Rivers Draining into a Tropical Volcanic Lake

3.1 Water quality

Table 4 summarizes the physicochemical characteristics of the river water observed during the study period. All measured parameters complied with the Indonesian river water quality standards (Class 2), which apply to water used for recreational purposes, aquaculture, livestock, and agricultural irrigation. Higher classification levels (Classes 3–4) indicate progressively lower water quality and more restricted applications, with Class 4 suitable only for agricultural irrigation. Certain parameters, such as EC and turbidity, are not included in the Indonesian standards. Nevertheless, based on the WHO guidelines [27], EC values in this study remained within the acceptable limit of 1.5 mS/cm, while turbidity values were consistently below 5 NTU, except at three sites (U4, U6, and U8).

Table 4. Physicochemical parameters observed in the rivers surrounding Lake Maninjau

Note: Class 2 river water quality standards as stipulated in Government Regulation of the Republic of Indonesia No. 22 of 2021.

Conventionally, water quality has also been assessed using the Pollution Index (PI). Figure 2 illustrates the variation in PI values during the observation period. All sampling stations exhibited PI values below 1, indicating that the sites were unpolluted and complied with the Class 2 water quality standard. In general, higher PI values correspond to lower water quality.

Figure 2. Values of the pollution index in the upstream rivers surrounding Lake Maninjau

The variation in PI values in this study was primarily driven by pH levels. Figure 3 shows a significant positive correlation between PI and pH levels (r = 0.95; p < 0.01), indicating that higher pH levels contributed to an increase in PI values. pH levels in aquatic ecosystems can be affected by various factors, including geological formations, biological processes, precipitation, and anthropogenic inputs [28]. Generally, pH values below 5 or above 9 are considered detrimental to aquatic organisms. Acidic conditions, in particular, tend to lower the diversity of benthic fauna and enhance the mobilization of heavy metals, which can be toxic to benthic macroinvertebrates [29].

Figure 3. Spearman correlations between biotic indices, water quality parameters, and biodiversity indices

3.2 Macroinvertebrate density and diversity

A total of 1,509 macroinvertebrate individuals, representing 59 taxa from three classes, 11 orders, and 36 families, were collected. Insecta comprised the majority of taxa (57), while Crustacea and Clitellata were represented by one taxon each. Insecta Coleoptera accounted for the highest number of taxa (17), with Elmidae as the most represented family. Site U8 exhibited the greatest number of taxa (24), mainly composed of Coleoptera, whereas U5 and U9 showed the lowest (8 taxa each), primarily consisting of Diptera and Ephemeroptera, respectively. Chironominae (Diptera) and Caenis (Ephemeroptera) were the only taxa recorded at all sites (Table 5).

Table 5. Dominant macroinvertebrate taxa: their abundance, tolerance scores, and FFG classifications (T1 = Biotilik; T2 = ASPT^THAI; T3 = SIGNAL2)

Figure 4. Variation in density and the number of taxonomic groupings in the rivers surrounding Lake Maninjau

Figure 4 illustrates the variation in macroinvertebrate abundance, ranging from 50 ind/m² at site U9 to 703 ind/m² at site U7. Macroinvertebrate communities at five sites (U2, U3, U4, U5, and U7) were dominated by Diptera, whereas Ephemeroptera predominated at sites U1 and U9. In contrast, Trichoptera and Coleoptera were the dominant orders at U6 and U8, respectively. Although site U7 supported the same number of taxa as U6 and fewer than U8 (16 taxa), its higher abundance was mainly attributed to Diptera, particularly Antocha (Limoniidae), which reached 272 ind/m² (Table 5). Moreover, Hydropsychidae (Trichoptera) also contributed to the elevated macroinvertebrate abundance observed at this site. Site U9 exhibited the lowest abundance, with the lowest number of taxa recorded among all sites.

Site U8 recorded the highest Shannon–Wiener diversity index (H′) value (2.70), while Site U5 showed the lowest (1.38) (Figure 5). These values were inversely related to the dominance index, with U8 exhibiting the lowest dominance (0.10) and U5 the highest (0.34). Overall, the studied sites indicated low dominance and moderate diversity. The dominance index ranges from 0 to 1, with values near 0 indicating a highly diverse community without dominant species, whereas values approaching 1 suggest that one or a few species dominate the community, resulting in low diversity [30].

Figure 5. Biodiversity index values of macroinvertebrates in the rivers surrounding Lake Maninjau

H′ = Shannon-Wiener diversity index, D = Simpson dominance index, J = Pielou evenness index, R = Margalef richness index

Pielou’s evenness (J) values varied across sites, ranging from the lowest at Site U7 (0.69) to the highest at Site U9 (0.97) (Figure 5). An evenness index value near 1 indicates that species are more evenly distributed, with similar numbers of individuals across taxa. Conversely, values close to 0 suggest that a few species dominate the assemblage [31]. Three sites (U2, U6, and U7) exhibited characteristics of a semi-stable community (J between 0.5 and 0.75). In contrast, the remaining six sites represented stable communities (J > 0.75), where dominance index values tended to be lower [32]. The Margalef index (R) values ranged from 1.37 (U5) to 3.76 (U8) (Figure 5). Sites U6 and U8 showed medium taxa richness (2.5 < R < 4), while the other sites were characterized by low taxa richness (R < 2.5) [33, 34]. A higher Margalef index value indicates greater species richness relative to the total number of individuals within the assemblage [35].

3.3 Ecological water quality based on biotic indices

Table 6 presents the biotic index scores for all sites along with their corresponding water quality classifications. Although the Pollution index indicated good water quality or compliance with Class 2 standards across all sites, the ecological water quality showed considerable variation. This discrepancy highlights that chemical compliance does not necessarily reflect ecological health. Such findings underscore the limitations of water quality standards, which are often designed for human use, and emphasize the importance of biomonitoring to more accurately assess the integrity of aquatic ecosystems.

Table 6. Scores and water quality classifications based on three biotic indices in the rivers surrounding Lake Maninjau

Based on the Biotilik index values, the water quality in the studied rivers ranged from Very Clean or Very Mildly Polluted to Rather Heavily Polluted. Meanwhile, the ASPT^THAI and SIGNAL2 indices indicated conditions ranging from clean water or healthy habitat to moderate pollution. Higher scores from these three indices generally reflect better water quality, indicating the presence of numerous sensitive, high-scoring taxa.

The three biotic indices generally provided comparable assessments of water quality, particularly at sites U1 and U8, which were classified as Mild Pollution and Excellent, respectively. Nevertheless, the Biotilik index indicated lower water quality at U2, U5, and U9 compared with the other indices. The ASPT^THAI index reflected higher water quality only at U4 and U7, while the SIGNAL2 index suggested lower quality at U3 but higher quality at U6. These differences likely stem from incomplete tolerance data for some taxa at the studied sites within the selected indices, as well as variations in tolerance scoring across indices. Notably, several taxa, including Psychomyiidae, Elmidae, Dryopidae, and Limoniidae, lacked tolerance scores for the Biotilik calculation. Among the three indices, SIGNAL2 included a more comprehensive set of scored taxa and also incorporated abundance weighting in its calculation, whereas Biotilik and ASPT^THAI considered only the number of taxa, without accounting for their abundances.

Among all the sites studied, U8 and U9, which are non-inflow rivers of Lake Maninjau, exhibited the highest ecological water quality, as indicated by at least two of the three biotic indices used. Four other sites (U1, U2, U3, and U6) were classified as lightly polluted, whereas the remaining sites (U4, U5, and U7) were moderately polluted. The decline in water quality was associated with the loss or reduced abundance of sensitive taxa, coupled with an increase in tolerant taxa. In the SIGNAL2 index, water quality was also influenced by the relative abundances of both tolerant and sensitive taxa. Overall, in this study, sensitive taxa were represented by EPT groups such as Leptophlebidae, Heptageniidae, Perlidae, Philopotamidae, and Glossosomatidae, whereas tolerant taxa included Gerridae, Simuliidae, Chironomidae, and Naididae.

The three biotic indices were correlated with water quality and biodiversity parameters (Figure 3). Biotilik and ASPT^THAI were negatively correlated with nitrite (r = –0.75 and –0.76, respectively; p < 0.05), indicating that higher nitrite levels corresponded to lower index values, reflecting reduced water quality. These results suggest that both indices are sensitive to nutrient-related stress.

The three biotic indices also showed significant positive correlations with the Shannon–Wiener diversity index, indicating that increased macrozoobenthos diversity was associated with higher biotic index values, reflecting improved water quality. In this study, sites with high H′ values (e.g., U8) also exhibited high values for all three biotic indices, indicative of clean water and healthy habitats, whereas sites with the lowest H′ (e.g., U5) showed correspondingly low biotic index values. The Biotilik index further showed strong and significant correlations with the Margalef index (r = 0.75, p < 0.05) and the Dominance index (r = –0.87, p < 0.01). Increased taxa richness was associated with higher Biotilik values, reflecting better water quality, while increased dominance of tolerant taxa, such as Chironomidae, significantly contributed to lower water quality as indicated by the Biotilik index. The presence of Chironomidae larvae has been reported to reflect anthropogenic impacts [36].

3.4 FFG analysis

In total, 59 macroinvertebrate taxa collected during the study period were assigned to six FFGs: collector-gatherers (18 taxa), predators (16 taxa), scrapers (13 taxa), collector-filterers (8 taxa), shredders (3 taxa), and omnivores (1 taxa). Collector-gatherers contributed the highest numbers of taxa, whereas omnivores were the least represented.

Overall, collector-gatherers were the most abundant FFG at almost all sites (40–82.35%), except at site U6, which was dominated by scrapers representing 45.83% of the assemblage (Figure 6). The highest proportion of collector–gatherers (82.35%) was recorded at site U2, a forested upstream area, with an abundance of 245 ind/m². Across all sites, their abundance ranged from 20 ind/m² at U9 to 390 ind/m² at U1 (Figure 7). Collector–gatherers were mainly represented by Caenis (Ephemeroptera) at U1 and U3, Antocha (Diptera) at U7, and Chironominae at the remaining sites.

Figure 6. Relative FFG density during observations

CF = collector-filterers, CG = collector-gatherers, SC = scrapers, PR = predators, SR = shredders, OM = omnivore

Figure 7. Density of FFG of macroinvertebrates during observations

CF = collector-filterers, CG = collector-gatherers, SC = scrapers, PR = predators, SR = shredders, OM = omnivore

The high proportion of collector-gatherers, particularly at the upstream sites, aligns with findings from the Yellow River [37], the Blyde River [38], and the Steelpoort River [39]. In contrast, this pattern differs from that reported for the Ciliwung River, where collector-filterers are predominant [40], and the Cisadane River, where scrapers dominate [41]. The high proportion of collector-gatherers but scarcity of shredders in the headwater sites (U2 and U9) contradicts the predictions of the River Continuum Concept (RCC). However, a low proportion of shredders is commonly reported in tropical streams [42, 43]. In this study, the relative density of shredders was low across sites, ranging from 0% (U1, U2, U8, and U9) to 3.39% (U5). The genus Pelonomus (Coleoptera) was the only representative at U3, U4, and U5. Additionally, Pelonomus co-occurred with Amphinemura (Plecoptera) at U6.

The dominance of collector-gatherers indicates a greater contribution of fine particulate organic matter (FPOM) compared to CPOM. In tropical regions, CPOM input to streams may be reduced because high temperature and humidity accelerate leaf decomposition on land [44, 45]. The low quality of leaf litter, due to the high content of secondary compounds in tropical vegetation, is also thought to contribute to the low abundance of shredders [46]. Shredders were generally scarce, comprising 0–3.39% of the community with abundances of 0–10 ind/m² (Figures 6 and 7). Despite U2 being located upstream within a forested area without adjacent rice fields (Table 2), it lacked shredders (0%), a lower value than other upstream forest sites that were near rice fields (U3, U4, and U6). In this study, shredder abundance showed a significant positive correlation with TSS and TN (Figure 8). Elevated TSS levels may indicate an increased particulate organic matter that supports the presence of shredders, while TN can stimulate microbial conditioning that enhances leaf-litter quality.

Figure 8. Spearman correlations between FFG abundances and water quality parameters

Collector dominance may reflect limited food availability for other groups or competition for shared resources [47, 48]. Moreover, most collectors are generalist feeders that utilize a wide variety of food sources and can inhabit diverse streambed environments, enhancing their survival and reproductive success. Collectors also play an important ecological role in repackaging fine particulate organic matter (FPOM) into larger particles after ingestion. Their abundance is likely related to their ability to exploit a broad range of food items compared with more specialized groups such as shredders and scrapers [49]. In addition, collectors are generally tolerant to environmental disturbances occurring in aquatic ecosystems [11].

Collector-filterers were the second most dominant FFG at sites U3, U5, and U7, where Cheumatopsyche (Trichoptera) dominated at U3 and U7, and Simulium (Diptera) at U5. These two taxa have also been reported as representative collector-filterers in the Upper Ciliwung River [40]. In this study, collector-filterers showed higher proportions at two downstream sites (U5 and U7), accounting for 35.59% and 34.52%, respectively. The combined abundance of collector-filterers and collector-gatherers reached 91.53% at U5 and 88.26% at U7, consistent with the River Continuum Concept (RCC), which predicts collectors as the predominant group in downstream reaches. Similar findings were reported for the Ranggeh River, where the proportion of collectors increased downstream from areas with intensive agricultural practices [50].

The abundance of collector-filterers showed significant positive correlations with EC, TDS, and TP, and a significant negative correlation with DO (Figure 8). The positive relationship between CF and TDS agrees with Uwadie [51] but contrasts with Doong et al. [52]. Increased TDS due to fine organic matter input likely enhances food availability for collector-filterers. Similarly, Zhang et al. [53] reported that higher conductivity promotes an increase in pollution-tolerant macroinvertebrates, consistent with the presence of Cheumatopsyche and Simulium in this study. The positive CF–TP correlation suggests nutrient enrichment promotes collector-filterer proliferation, indicating a disturbed condition in the study area [37].

Scrapers dominated at site U6, composed mainly of Agapetus (Trichoptera) with a density of 165 individuals/m². At four other sites (U1, U2, U8, and U9), scrapers were the second most abundant FFG, accounting for 14.75%, 14.29%, 28.57%, and 30% of individuals, respectively. Scrapers at U8 were represented by Psephenus (Coleoptera), while at the other three sites, they were represented by Heptagenia (Ephemeroptera). The high density of scrapers at U6 may be related to the open canopy, which increases light intensity and primary production, thereby enhancing algal growth as a food resource for scrapers [54]. Additionally, their abundance was supported by riffle habitat conditions [55].

Predators were present at all study sites, with proportions ranging from 2.52% to 23.23%, corresponding to 7-95 individuals/m². During the observation period, only site U4 showed predators as the second most abundant FFG, represented by Rhagovelia (Hemiptera) with a density of 27 individuals/m². Their abundance depends on prey availability and intra-guild competition for food [56]. Our results also indicated that predator abundance was positively correlated with pH, which may be associated with higher prey availability under slightly alkaline water conditions.

3.5 Attributes of the aquatic ecosystem

The P/R ratio is commonly used to represent the relative energy contribution of primary producers, with values across ecosystems serving as proxies for the proportion of allochthonous versus autochthonous organic matter [57, 58]. In this study, the P/R ratio identified only one site (U6) as autotrophic, while the remaining eight sites were heterotrophic, depending largely on allochthonous riparian organic matter input. Five upstream sites (excluding U6) conformed to the predictions of the River Continuum Concept (RCC) with P/R < 1. Similarly, the downstream sites (U5 and U7) also showed P/R < 1 (Table 7). In contrast, the midstream site (U8) deviated from RCC expectations, which predict P/R > 1. This discrepancy is likely due to elevated turbidity (Table 4), which can reduce algal production and thus limit food availability for scrapers.

Table 7. The value of the river ecosystem attributes