Measurements of physical variables revealed that the highest temperature at the breeding sites was recorded at the buffalo footprints site (39.4 °C), whereas the lowest temperature was found in freshwater puddles (29.0 °C). Light intensity was highest in rice fields (421 lux) and lowest at the buffalo footprints breeding site (103 lux). Generally, the water flow conditions at the twelve breeding sites were stagnant, except for the estuary habitats, which exhibited slow-moving to flowing water. The largest habitat areas were recorded at the estuaries, ranging from 250 to 500 m2, whereas the smallest habitat area was recorded at the water spring, measuring 1 m2 Table 2. Additionally, the chemical variables revealed that the water pH at the breeding sites ranged from 7.3 to 9.2, salinity levels ranged from 0‰ to 20‰, and dissolved oxygen levels ranged from 1 to 8 mg/l. The highest pH and salinity levels were recorded in coastal puddle 1, while the highest dissolved oxygen level was recorded in the water spring Table 2.

The multivariate structure of physicochemical characteristics among multiple aquatic habitat types was determined via principal component analysis (PCA). PC1 and PC2 explained 34.3% and 27.0% of the total variance, respectively, with a cumulative contribution of 61.3% Figure 2. These cumulative variances indicate that the first two axes represent major environmental gradients explaining most of the variation distinguishing the habitats.

The major contributors to PC1 were dissolved oxygen (DO), light intensity and habitat area with positive loadings on this axis, which constituted 56.0 of the total variance. Conversely, temperature was negatively associated with PC1. This means that PC1 indicates a gradient from larger, more or less well illuminated, oxygen rich habitats to relatively smaller and hotter water bodies. Habitats like water springs, freshwater puddles, estuary 1 and rice fields were located on the positive side of PC1 showing their association with greater dissolved oxygen concentrations, higher light exposure and larger surfaces areas. On the contrary were buffalo footprints, buffalo wallows and numerous coastal puddles which appeared on the negative side of PC1 at relatively higher temperatures and smaller dissolved oxygen levels.

The variation along PC2 was largely driven by pH and salinity (both with positive loadings) indicating that this axis represents a chemical gradient. Coastal puddle 1 clustered strongly separably along PC2, and was significantly positively associated with high salinity, distinguishing it from the mostly fresh water habitats. Conversely, river pool 2 and estuary 1 represented the moderate positions along PC2, indicating intermediate chemical properties.

Strong clustering patterns are observed ecologically in the biplot. We distinguished coastal habitats for which salinity was measured from inland freshwater habitats. Regardless, anthropogenic or semi-managed habitat types (e.g., rice fields) reflected tighter associations with light intensity and dissolved oxygen. Small, surficial and organic-rich habitats (i.e., buffalo footprints and wallows in aggregations) yielded similar physicochemical profiles at or near the surface with high temperatures and limited dissolved oxygen.

In conclusion, the PCA shows that habitat differentiation is largely structured by (1) a certain oxygen–light–area gradient along PC1 and (2) a salinity–pH gradient along PC2. Our results indicate that these environmental gradients likely underlie important measures of ecological suitability and the distribution of species in the studied aquatic systems.

Turbidity is higher in buffalo wallows and coastal puddles than other types of puddles. At midday exposure to the sun, these puddles may be thermally stratified, with high surface temperatures due to absorbed solar radiation(Paaijmans et al., 2008). The ecological factors controlling turbidity in the lakes and reservoirs were soil type and rainfall in these water bodies. Tidal influence contributes to a unique sediment profile for each of the coastal areas in West Sumba. The sediment along the coast is dominated by medium and coarse sand (Nugroho & Putra, 2019). Such temperature observations (29–39.4 °C) were similar to those of previously described sun-exposed, turbid water bodies found near buffalo footprints, puddles and rice fields in East Sumba matching previous reports with pH of 7–9 generally supporting high densities of larvae (Willa & Kazwaini, 2015). Other coastal situations described in Motaain, NTT indicated that Anopheles larvae are concentrated in swamps, rice field pools and estuaries demonstrating the relevance of stagnant partially saline surface water for vector breeding (Darma et al., 2022).

The current study aligns with these patterns, especially in identifying An. sundaicus and An. subpictus as the dominant species in estuarine and animal-impacted sites such as buffalo wallows and footprints. In addition, previous studies conducted in North Kalimantan and Makassar also reported high frequences of these species, which were mainly found in muddy coastal habitats characterized by vegetation and aquatic biota such as grass, tadpoles and dragonfly nymphs (Sugiarto et al., 2016);(Nurmalasari et al., 2019);(Sindhania, 2020). Buffalo presence is strongly associated with maintaining Anopheles populations by promoting beneficial breeding conditions. Buffalo deposit dung in water bodies and create new habitats for aquatic organisms through the generation of habitat by contributing to creating more water-holding depressions (hoofprints) while compacting the soil and reducing water infiltration (Buxton et al., 2022);(Östman et al., 2015). Grazing also creates small, dirty water pools that are ideal sites for mosquito breeding (Östman et al., 2015). The affinity for warm, shallow and biologically productive niches demonstrated by these coastal vector species underlines the commitment to monitoring transitional zones where estuarine, agricultural and livestock land uses converge. Cumulatively, this evidence reinforces the interpretation that buffalo-affected coastal habitats are consistently and substantially supporting areas of high-potential malaria vector populations across Nusa Tenggara and similar Southeast Asian environments.

Light intensity at these breeding sites sites were ranging from 103 to 421 lux, and even though trees and mangroves exists here, it was more correlated to their closeness along the coastline. Habitats exposed to direct sunlight are ideal for An. sundaicus to thrive. Direct sunlight increases water temperature, accelerates evaporation, and subsequently increases salinity–conditions under which euryhaline mosquitoes like An. sundaicus and An. subpictus competitively thrive (Surendran et al., 2011). Sunlight also promotes algal growth, providing abundant food and higher dissolved oxygen for An. sundaicus larvae. While previous study report optimal developmental temperature of 20 °C to 27°C (Sugiarti et al., 2020), mosquitoes can survive at more extreme temperatures, though metabolic processes slow or cease below 10 °C or above 40 °C (Christiansen-Jucht et al., 2014). These findings align with (Ernawati et al., 2012), which revealed that temperatures ranging from 26 °C to 33 °C remain suitable for Anopheles spp. larval development. Furthermore, (Kermelita et al., 2024) reported that the optimal light intensity for larval growth ranges from 224 to 674 lux, which aligns perfectly with the peak light intensities recorded in our sun-exposed breeding sites.

The size of the breeding sites differed widely, from 1 m² to 500 m². (Mulyadi, 2010), also found that Anopheles spp. larval habitats have been recorded from 0.5 m² to >100m², displaying some spatial plasticity in these early life stages. This enormous breeding range (from small footprints of buffalo to extensive estuaries) allows for stable vector populations facilitating year-round malaria transmission in the village of Gaura, especially when these sites are located in close proximity to human dwellings so that they can remain undetected and thus unaddressed potentially targeting them for control. The physical, chemical, and biological characteristics contribute to the significant differences in Anopheles breeding sites size in Gaura Village. This difference in size is associated with physical factors (temperature, light, and water flow) that impose habitat stability, chemical variables (pH, salinity, DO) that dicatate developmnetal suitability. These differences are amplified by biological conditions, whereby aquatic vegetation serves as both habitat and food source. In the end, these preferences convey Anopheles mosquitoes adaptive behaviors of exploiting habitats which providing the greatest opportunity for survival.

The chemical features of the breeding sites in Gaura Village, especially their alkaline pH, low dissolved oxygen content and both fresh and brackish salinity greatly affect composition of Anopheles species. An. vagus, An. subpictus, and An. sundaicus are known to be consistently associated with coastal puddles with alkaline conditions (pH up to 9.2) and with moderate salinity (up to 20.0‰), all of which are also known to tolerate or prefer brackish, sun-exposed habitats. Estuarine zones and buffalo-impacted areas such as wallows and hoofprints, which are influenced by tidal intrusion and livestock activity, also supported An. sundaicus and An. subpictus, suggesting that both natural tidal dynamics and anthropogenic ecological features contribute to their presence. Meanwhile, although adaptable species like An. vagus and An. subpictus were also present, freshwater habitats like river pools and rice fields, characterized by zero salinity and neutral to slightly alkaline pH, were more frequently associated with species such as An. barbirostris, An. annularis, and An. flavirostris, which prefer non-saline environments. Together these results show the unique physicochemical profiles of larval habitats in Gaura Village that provide microhabitats to diverse Anopheles species, each associated with specific environmental conditions.

A previous study by(Taher et al., 2021) reported that An. sundaicus develops optimally in brackish water with salinity levels ranging from 12‰ to 18‰ but cannot survive at salinity levels above 40‰. Ecologically, coastal brackish habitats offer abundant microbial food and reduced competition that An. sundaicus to exploit. Physiologically, An. sundaicus adaption is likely supported by reduced cuticular permeability and specialized osmoregulatory responses in the rectum, permitting effective ionic balance in highly saline environments(Ramasamy & Surendran, 2011). In Kondamaloba Village, Central Sumba, An. sundaicus breeding sites have a salinity of 12‰ and a pH of 8.8(Wayan & Adnyana, 2011). In the current study, An. sundaicus demonstrated extreme flexibility; it was found in high-salinity coastal puddles (20.0‰), extremely low-salinity buffalo wallows and buffalo footprints (0.1‰), and even entirely freshwater river pools (0.0‰).

Additionally, previous studies also revealed that Anopheles spp. larvae tolerate pH values ranging from 7.9–8.9. Water pH influences aquatic organisms and water fertility, with optimal microbial activity generally occurring at pH values between 6.5 and 8.3. Extreme pH levels can deactivate or kill these microorganisms, creating an intolerable environment by reducing the availability of food for Anopheles larvae. Larvae predominantly feed on microorganisms–including algae, bacteria, and protozoa–as well as organic detritus suspended in water. When unsuitable pH conditions suppress microbial growth and survival, then resulting decline in food availability negatively impacts larval survival and abundance. Furthermore, dissolved oxygen is crucial for larval survival and is inversely related to water temperature(Mading & Kazwaini, 2014);(Febriani et al., 2019).

Anopheles subpictus was found to thrive in both saline and non-saline aquatic habitats. Its presence in rice fields with zero salinity demonstrates its high adaptability to various breeding conditions, further highlighting its role as a widespread malaria vector. A previous study by (Laumalay, 2022) reported that An. subpictus is broadly distributed and successfully inhabits both freshwater and high-salinity environments.

Anopheles barbirostris was found exclusively in freshwater puddles, rice fields, and river pools, with no salinity detected. As a strict freshwater mosquito, An. barbirostris lacks of the osmoregulatory adaptation found in euryhaline mosquitoes. In particular, it does not possess specialized salt-excreting structures, such as modified rectal segments, rendering its larvae unable to effectively eliminate excess salts. As a result, exposure to saline water leads to osmotic imbalance and larval mortality(White et al., 2013). This aligns with(Jastal, 2005), who reported that An. barbirostris larvae inhabit lowlands, hills, and mountains, predominantly utilizing rice fields, freshwater pools, springs, fishponds, and swamps.

Anopheles vagus exhibited broad ecological tolerance, being found in river pools, buffalo footprints, buffalo wallows, and coastal puddles–even in high-salinity environments reaching 20‰.(Laumalay, 2022) identified An. barbirostris, An. subpictus, An. vagus, An. vagus var. limosus, and An. indefinitus in Lifuleo Village. This confirms that coastal ecosystems, with their mosaic of fresh and brackish micro-habitats, can support a highly diverse vector community with varying degrees of salinity adaptability.

The presence of An. sundaicus, An. subpictus, An. barbirostris, and An. vagus—confirmed malaria vectors in East Nusa Tenggara—highlights the urgent need for a multifaceted vector control strategy to prevent malaria transmission in the coastal areas of Gaura Village(Jastal, 2005);(Laumalay, 2022). Larval source management, through the targeted filling, draining, or modification of coastal puddles and stagnant water bodies can reduce breeding habitat availability. In localized areas, establishing biological control agent, such as predator reservoirs, may offer supplementary suppression. Furthermore, selective larviciding should be deployed in habitats that are persistent or unsuitable for physical modification. Simultaneously, standard adult-focused interventions, including the use insecticide-treated nets and indoor residual spraying, are critical to block transmission from anthropophilic species such as An. vagus and An. barbirostris. Finally, community education aimed at reducing outdoor activities during peak biting periods and minimizing exposure to breeding sites will further bolster these integrated interventions (Ndoen et al., 2010).

The biological variables of the 12 breeding sites Table 3 confirmed the important ecological functions performed by aquatic flora and fauna affecting Anopheles spp. larval habitats. Aquatic vegetation is an important oviposition site and offers shelter for mosquito larvae to resist predation whereas aquatic fauna act as predators and natural competitors(Huang et al., 2006);(Culler & Lamp, 2009). For instance,(Febriani et al., 2019) reported that the presence of larvivorous fish inversely affects larval density, acting as a direct biological control–higher fish populations lead to lower larval densities, and vice versa. In addition, aquatic plants affect dissolved oxygen concentrations which in turn determines the extent to which habitats are suitable for certain other aquatic organisms.

Previous research shows that these ecological drivers exert influences on larval development but also have carry-over effects with adult traits–size, fecundity and longevity–that shape the population dynamics of mosquitoes (Roux & Robert, 2019). While this study did not directly address the emergence of adults, the long-term occurrence of supportive aquatic flora and fauna suggests that these sites function as multilayered developmental sites providing critical ovipositioning, shelter, foraging resources, and refuge from predation (Depkes, 2004). The aquatic vegetation forms a niche and also provides adult mosquitoes with much more microhabitats for rookery fattening in order to complete gonotrophic cycles.

The highest larval densities were recorded in animal-impacted habitats, specifically buffalo wallows (3.5 larvae per dip) and buffalo footprints (3.1 larvae per dip), both of which predominantly hosted An. vagus, An. subpictus, and An. sundaicus. Coastal puddles similarly hosted these three species with moderate densities ranging from 1 to 2 larvae per dip. In contrast, agricultural and natural freshwater habitats exhibited lower productivity; rice fields had a density of 1.1 larvae per dip (hosting An. vagus, An. annularis, An. barbirostris, and An. sundaicus), estuaries recorded 1 larva per dip, and river pools recorded the lowest density at 0.7 larvae per dip. These findings reinforce that animal-impacted, spatially confined habitats are the most highly productive breeding sources in the village.

Spatial analysis found that the 12 breeding habitats were found located between 6 and 159 m from the coastline, and between 108 and 512 m to residential areas. The close association of these breeding places with human habitation greatly increases transmission risk of malaria, especially with the presence of confirmed malaria vectors (An. subpictus, An. sundaicus, and An. barbirostris) and suspected vectors (An. vagus, An. annularis, and An. flavirostris). Based on these results, coastal areas in Gaura Village are considered to be at high risk for malaria transmission as all potential breeding site recorded were well within the flight range of Anopheles spp. This aligns with the findings of(Heriyanto et al., 2011), that these mosquitoes usually fly as much as aerodynamically would paint out approximately 1–2 km from their breeding grounds, following the fact that each of the residential buildings are at a most distance of only 512 meters.

Over the coastal areas, 12 potential breeding habitats were recorded in total during both survey rounds. Of the 12 potential breeding sites surveyed, 10 were larvae positive giving an overall habitat index of 83.3%. The potential from this risk is also reflected by the majority of malaria vector confirmed in East Nusa Tenggara (An. subpictus, An. sundaicus, and An. barbirostris), alongside suspected vectors (An. vagus, An. annularis, and An. flavirostris) (Kemenkes, 2017);(Heriyanto et al., 2011).

Principal component analysis (PCA) showed that environmental differentiation of Anopheles larval habitats in the coastal ecosystem of Gaura Village was mainly explained by two large ecological gradients: habitat area as the main gradient of Component 1 and water pH as the main gradient of Component 2 (Figure 3). The revised PCA revealed that light intensity (LI) and the majority of the physicochemical variables had only a small effect on habitat segregation, in contrast to some preliminary results. In these nursery habitats, species-specific structural and chemical properties– not sun exposure or temperature – are more central to making up larval habitat heterogeneity in this landscape.

Strong positive loadings on habitat area were associated with Component 1, which accounted for 34.3% of the total variance. Rice fields, estuary pools, freshwater puddles and river pools cluster on the right side of this axis, indicating that they share structural traits and larger habitats. Smaller, more localized habitats such as buffalo footprints (BF), buffalo wallows (BW), and coastal puddles (P) clustered at the negative end of Component 1. This gradient is relevant to ecology, because habitat area influences hydrological stability, predator assemblages and resource distribution. Broader water bodies accommodate more complicated food webs and higher densities of natural predators to mosquitoes (Minakawa et al., 1999). In contrast, smaller habitats can be predator-poor and very favorable for larval development. The ecological context from our findings is consistent with more broad evidence for habitat mediated patterns of larval competitive interactions (Juliano, 2009). This size-related pattern is also consistent with observations in coastal environments, where spatially constrained habitats regulate larval production(Mutuku et al., 2006).

Component 2 which explained 27.0% of the total variance was influenced by both water pH and salinity. This axis separated alkaline and saline habits such as coastal puddles from those with lower or neutral pH such as buffalo-associated habitats and water springs. The relationship between larval development and pH has been well established because pH affects microbial community structure, algal productivity, and nutrient availability—all crucial resources for larval foraging. Relationship between some physicochemical characteristics (pH, ionic concentration, etc.) have also been previously shown with larval abundance in field studies (Avramov et al., 2024). While salinity does play a role in this PCA output, its expression is not as strong but remains an important ecological feature of coastal ecosystems subject to tidal exchange. The wide range of salinity tolerance of coastal malaria vectors has been demonstrated by early studies on the An. sundaicus complex (Dusfour et al., 2004). Field observations of larval development in brackish water provided further evidence of significant quantitative differences in salinity tolerance and physiological plasticity among populations of An. sundaicus(Surendran et al., 2011). This greater tolerance to environmental conditions aligns well with the use of habitats by dominant malaria vectors at regional and continental scales as described in large-scale distributional analyses(Sinka et al., 2011).

Notably, results from the ordination indicated that light intensity, temperature and dissolved oxygen were traits with short eigenvectors which contributed very little to the main gradients of habitat differentiation. This indicates that whichever breeding sites were sampled had levels of light, thermal and oxygen conditions consistent enough to not exceed the range habitual tolerances in Anopheles larvae. While in some inland environments, light exposure may be critical to algal productivity and larval food availability the coastal habitats of Gaura Village appear to be more influenced by size and water chemistry than energy from sunlight or heat. Among the habitat types, Coastal puddle 1 was unique and distinctly isolated from all other habitats along the positive Component 2 axis. This implies that its physicochemical conditions, massively high or low in pH and salinity relative to other aquatic habitats, distinguishes it clearly from all others. These outlier habitats may host distinctive larval compositions and/ or have different rates of development, but require more ecological scrutiny.

On the other hand, river pools and water springs are distributed near the higher pH and dissolved oxygen areas where they are less influenced by tidal wave, revealing their relative water quality stability. In wrapping up, the PCA reports that habitat area and water chemistry drive heterogeneity of Anopheles larval habitats in this coastal ecosystem. These results suggest that vector management efforts in Gaura Village should characterize habitats based on habitat size, stability and chemical conditions rather than just openness or sunlight exposure. Conversely, highly saline habitats, particularly coastal puddles and spatially isolated buffalo-associated microhabitats, represent important larval sources and should be prioritized in larval source management strategies. Furthermore, the structural and chemical gradients are critical for the development of more ecologically-based vector suitable control interventions in coastal, malaria-endemic areas.