Agriculture is the economic backbone of the developing countries in Southeast Asia. The Association of Southeast Asian Nations, ASEAN, envisioned economic prosperity in its member nations through the created integrated economy that allows the free movement of goods and services, skilled labor, and capital among member nations (Ishikawa, 2021). ASEAN’s major concern to its member nations is food security due to their agricultural dependency and growing population. As of 2022, ASEAN has a combined population among its 10 member nations of approximately 673 million (Statista, 2024). ASEAN’s vision paved the way for the economic growth of the region by supporting livelihood and the protection of social stability (Sundram, 2023).

The ASEAN integration created an economic shift from an agrarian economy to a highly commercialized agriculture. The shift is strengthened by policy initiatives and played an important role in the economic growth of largely agriculture-dependent countries (Hurni et al., 2017). For the last decade, the vision brought economic growth through agricultural expansion in the Southeast Asia region.

Among ASEAN member nations, Vietnam shows impressive economic growth through agriculture expansion. The dramatic economic growth experienced by Vietnam has been attributed to a series of economic reforms known as “doi moi” and its membership in ASEAN in 1995 (Emmers and Le Thu, 2021; Van Lam, 2019). It showed transformation in the agricultural sectors that benefitted the marginalized and reduced poverty. The latest decrease was attributed to the improvement of income of the Vietnamese farmers in the highland by encouraging them to grow profitable agro-industrial crops (The World Bank, 2018).

In the Philippines, the agribusiness shift, which transpired from the ASEAN integrated economic goal implemented in early 2000, converted approximately two (2) million hectares of forest and agricultural lands, or the so-called “idle lands,” to agribusiness areas within 5 years from 2005 to 2010 (Pulhin and Ramirez, 2013). The agribusiness shift was under the directives of the 2004–2010 Philippine Medium Term Development Plan, adopted in the 2011–2016 Philippine Development Plan (PDP), further enhanced by PDP 2016–2022, and now the PDP 2023–2028, with the current administration, further enhanced as the government’s strategy to improve the economy and alleviate poverty (Briones et al., 2023; Custodio and Sombilla, 2020).

In both countries, Vietnam and the Philippines, aside from agricultural expansion, intensive agriculture using pesticides is practiced (Tran et al., 2018). According to the latest data, Vietnam has a total land area of 331,690 sq. km; 39.4% is agricultural land, and 47% is forest area. Vietnam’s agricultural value added in USD is 51.40B, with a crop production index of 109.6 index points. But has high fertilizer usage of 427.6 kg per hectare of arable land (TheGlobalEconomy.com, 2024b). The Philippines has a total land area of 343,450 sq. km, of which 42.5% is agricultural land with only 24.2% forest area. The agricultural value added in USD is 41.08B, with a crop production index of 104.0 index points. The fertilizer usage is 232.0 kg per hectare of arable land, which is almost 50% less than Vietnam’s (TheGlobalEconomy.com, 2024a).

In agriculture, the benefits of pesticides are well recognized and understood, as they are developed with strict guidelines for very reasonable purposes. However, if necessary, agricultural intensification uses more pesticides to increase productivity per land area. An increase in the chemical application will affect the crops and soil quality, which will then affect the water quality of the receiving water through agricultural runoff. Pesticide application in farms contaminates air, soil, water, and other organisms. This study assesses the residual pesticide contamination on the receiving surface water brought about by agricultural runoff.

The residual pesticide contamination in the surface water validates the presence of contamination of the pesticides farmers extensively use. Assessing residual pesticides in water resources is challenging due to the very low solubility of ppb (parts per billion) concentrations in water (10⁻⁹) or even lower. One major problem in residual pesticide assessment is the fluctuating and irregular emission of pesticides in surface water. Hence, using the traditional grab water sampling method to obtain water samples taken in a specified time can give a wide range of values that often give unreliable results.

To address the problem of monitoring residual pesticides in water, it has been substituted by monitoring pesticides on sediments and tissues. These substitution methods have drawbacks. Despite direct water exposure, the tissue samples failed to consider other chemical pollutants, low bioaccumulation, avoidance, metabolism, and organism death (Alvarez, 2013). Aside from the low solubility of pesticides, sporadic emission is also a significant problem in which the commonly used water sampling method is limited. Passive samplers (PS) are an alternative sampling technique that enables the accumulation of pollutants in the samplers for more reliable monitoring of residual pesticides in water (Valenzuela et al., 2020).

The sampling method must be addressed to attain reliable results. Thus, this study uses passive samplers (PS) instead of the traditional water grab sampling method. Passive sampling methods were first developed in the early 70s for air pollutants and were used to determine occupational exposure to volatile organic chemicals in the air (Carmichael et al., 2003). Byrne and Aylott (1980) made the first patented PS in water, which used a nonpolar solvent in a nonporous membrane. Early PS was developed to effectively measure traces of POPs in water using a semipermeable membrane device, SPMD and is still currently used (Booij et al., 1998; Gonçalves et al., 2021).

After introducing SPMD, sampling devices for nonpolar organic contaminants have been developed. One is the polar organic integrated sampler (POCIS), tested in hydrophilic compounds like atrazine, diazinon, and 17-α-ethynylestradiol, which were not readily measured in the other devices (Kolpin et al., 2002). Several passive samplers have been developed, including an Empore disk, a Chemcatcher, a membrane-enclosed silicon collector, and a polyoxymethylene sheet (Becker et al., 2021; Gravell et al., 2020; Odetayo et al., 2020).

Some may argue that residual pesticide assessment using PS involves considerable investments in analytical instruments, which could be challenging in developing countries. But this is also true for the grab water sampling method. Generally, the PS method is cost-effective, consistent, practical, and suitable for routine monitoring (Deltares, 2016; Górecki and Namieśnik, 2002; Regan et al., 2018). It can detect low concentrations of organic compounds in pg./L to ng/L, which is difficult to attain in the grab sampling method (Hawthorne et al., 2011; Moschet et al., 2014).

For Vietnam and the Philippines, pesticides are used for practical purposes, such as preserving crops and controlling infestations for continuous production to cater to the client’s needs. What is overlooked is that when a pesticide application is mismanaged or overused, it has the potential to cause harm to humans and the environment. These residual pesticides in farms are potential runoff pollutants in surface water that can affect communities living along rivers dependent on these contaminated surface waters for domestic use. Worst, if residual pesticides are measured in communities where these pesticides were never applied.

Thus, in this study, the assessment of residual pesticides of two river systems in two developing Southeast Asian countries, namely, Can Tho River, Vietnam, and Cagayan de Oro River Basin (CdO), Philippines, the samplers developed by Deltares for more than a decade will be used (Deltares, 2019a). This set of PS gives good estimates of residual organic pesticides for the following reasons: it can detect and measure fluctuating emission of organic pollutants; it can detect more organic substances on water samples taken at the same time compared to the grab sampling method (Booij and Smedes, 2010; Deltares, 2019a; Smedes and Booij, 2012).

Moreover, the PS used in this study is one of the few deployed in the abovementioned advanced sampling devices to analyze and evaluate the residual pesticides that are unintentionally disposed of and transported through surface runoff from upstream farms to receiving waters. The water sampling using PS will be conducted uninterruptedly for at least 30 days for 2 consecutive years. In addition to the residual pesticides, traces of persistent organic pollutants (POPs) banned two decades ago will also be assessed. These POPs include organochlorine pesticides (OCP), polychlorinated biphenyls (PCB), and polycyclic aromatic hydrocarbons (PAH).

Both the SR and SD samplers were capable of detecting multiple pesticides. While certain pesticides were consistently measured throughout the 2-year monitoring period, others were only detected in the first year and not in the second year, and vice versa. The residual pesticide concentrations are detailed in Tables 1, 2, while values for other organic contaminants can be found in Tables 3, 4. It is important to note that “not detected (nd)” values do not necessarily indicate the absence of a particular pesticide. If a pesticide was detected at least once in a sampling site, the lack of detection in subsequent samples may be attributed to extremely low concentrations, making quantification unfeasible, or potentially due to farmers not utilizing such pesticides during the sampling period.

In Vietnam, 2,4-D was measured in only in SD samplers for the 2-year monitoring. It was above the applied threshold level of 10 ng/L only at VS1 and VS2 of year 1. Recently, 2,4-D was banned in Vietnam as proven one of the components of Agent Orange (PAN Asia Pacific, 2017). Atrazine and atrazine-desethyl were measured in all sampling sites for 2 years. Their values were higher in SR than in SD samplers. Atrazine and atrazine-desethyl were proven to be endocrine disruptors that were banned in the European Union.

Carbendazim, dimethoate, ethoprophos, fipronil, propiconazole, and triadimenol have values above the applied threshold level in at least one sampler at a specific sampling site. These compounds are classified as “possible human carcinogen,” except ethoprophos, which is classified as “likely to be carcinogenic to humans” by USEPA (2018).

The values of 27 prevalent pesticides per sampling site within the year of monitoring were statistically compared. The pesticide values on the same sampling sites monitored for 2 years were also statistically compared. Results show no significant differences between the values of these prevalent pesticides from VS1 to VS3 measured within the year for the 2-year monitoring period. No significant differences existed between the pesticides measured from the same sampling stations for a 2-year monitoring period. The ANOVA two-factor with replication result is shown in Supplementary Table 2.

In the Philippines, some pesticides measured were banned or restricted in the EU and the US due to their adverse effect on human health. Carbendazim was banned in the EU (PPDB, 2019); dichlorvos was classified as a probable human carcinogen and was proposed in 1995 by US EPA to cancel its usage (PubChem, 2019a); diuron is known/likely to be a human carcinogen (Huovinen et al., 2015); monuron in the US is not registered for use but manufactured for export (PubChem, 2019d); fenamiphos was listed restricted in the Philippines since 1989 but needed for banana and pineapple plantations (Fertilizer and Pesticide Authority, 2023); pyrimethanil was categorized as a possible human carcinogen (PubChem, 2019b); and simazine was reclassified as not genotoxic in 2005 after being classified as a carcinogen in 1989 (PubChem, 2019c; USEPA/Office of Pesticides Programs, 2006).

The values of 13 prevalent pesticides per sampling site within the year of monitoring were statistically compared, and so with the pesticide values on the same sampling sites monitored for 2 years. Results show significant differences between the values of these prevalent pesticides from PS1 to PS2 to PS3 measured within the year for the 2-year monitoring period. Significant differences existed between the pesticides measured from the same sampling stations for a 2-year monitoring period. The ANOVA two-factor with replication result is shown in Supplementary Table 3.

The statistical results can be further explained by the nature of the two river systems under study. The water flow of Can Tho River, Vietnam, comes from one river system. Meanwhile, the CDO River Basin, Philippines, considered in this study, consists of two river systems that flow toward PS3 downstream.

OCPs, PCBs, and PAHs were measured in trace concentrations in both countries, even banned or restricted by both governments. This shows that after two decades, these organic compounds have persisted in the environment for years. Stockholm Convention of 2004 initially banned or restricted most of these organic compounds (Lallas, 2001) while some compounds were under review and eventually added to the Rotterdam Convention in 2016 (United Nations, 2016). These compounds have been proven to persist, bioaccumulate in the food chain, and adversely affect the environment and humans.

While the residual pesticides measured in this study fall within the tolerable levels of contaminants in river water as established by governmental standards, their presence cannot be disregarded when considering the goal of ensuring sustainable access to clean water for all, as well as ensuring the safety of water for future generations. Prevention and mitigation should be in place to avoid potential health damage. Instances of irreversible consequences, such as the case of over 4,000 workers/farmers in banana plantations in Davao, Philippines, exposed to the pesticide dibromochloropropane (DBCP), serve as stark reminders. Their prolonged exposure to DBCP left 80% of the complainants infertile, with some even testing with zero sperm counts (Chen, 2012; Gomez, 1997). It is an example of the chronic effect of prolonged exposure to pesticides by farm workers.

The extensive presence of multiple residual pesticide contaminations in surface water, spanning rural communities and urban areas downstream, stresses the pervasive impact of agricultural runoff on water quality along these water bodies. The expansion and intensification of agriculture in Vietnam and the Philippines show that agricultural runoff contaminates surface water particularly during the rainy season, when agricultural runoff is expected to rapidly increase.

Despite the residual pesticide levels being within “acceptable” limits according to the government guidelines, the imperative for regular monitoring and evaluation remains paramount. It is crucial for concerned government agencies to establish baseline data to preempt pesticide-related disputes. The inefficacy of current pesticide monitoring methodologies impedes efficient implementation in both countries.

To address this challenge, adoption of an alternative methodology is imperative. Passive samplers, as demonstrated in this study, offers a promising way forward. Passive samplers captured 83 and 69 residual pesticides at concentrations of ng/L in Vietnam and the Philippines, respectively, a result which is unattainable through conventional grab sampling methods. Additionally, trace concentrations of banned compounds such as OCPs, PCBs, and PAHs were detected at pg./L levels. This success can be attributed to the combination of passive samplers tailored for both nonpolar (water-insoluble) and polar (water-soluble) organic contaminants, coupled with sensitive analytical instruments including GCMSMS and LCMSMS.

In light of the demonstrated effectiveness of passive samplers in detecting a wide array of residual pesticides, it is recommended to adopt these innovative sampling techniques into routine water quality monitoring programs in developing countries that are agriculture dependent. This would provide a more comprehensive and robust assessment of these pesticide contamination levels, enabling to substantiate our findings and provide early warnings for the enforcement of effective mitigation strategies. For example, given the detection of trace concentrations of banned compounds such as OCPs, PCBs, and PAHs at pg./L levels, there is a pressing need for stringent regulation and enforcement measures to stop the use and release of these hazardous substances.

Additionally, efforts should be directed toward enhancing pesticide education and training programs for farmers, emphasizing the importance of adopting sustainable agricultural practices and alternative pest management strategies to reduce reliance on chemical pesticides. Collaborative initiatives between government agencies, research institutions, and agricultural stakeholders can facilitate knowledge exchange and capacity-building efforts aimed at promoting environmentally friendly farming practices.

Overall, the adoption of passive sampling techniques coupled with regulatory measures and educational interventions will be required to mitigate pesticide contamination in surface waters and safeguarding environmental and public health of the two countries of this study and beyond. Continued research and investment in innovative monitoring technologies and sustainable agricultural practices seem to be crucial for achieving long-term water quality goals and ensuring the well-being of both ecosystems and communities reliant on these vital water resources.