Enhancing agricultural production in the Philippines by sustainable use of shallow groundwater

Work on groundwater management within the Philippines has focussed on two pilot sites within neighbouring municipalities of the province of Ilocos Norte, namely Burgos and Pasuquin, which is located on the north-western tip of Luzon, Philippines. For year 1, activities within these two sites focused on assessment of groundwater supply and demand. This has involved assessment of groundwater condition and installation of monitoring systems that will be used to assess groundwater recharge and sustainable levels of groundwater use, and determination of the current level of groundwater use, and the current level of production of various crops within the region that might benefit from irrigation with groundwater. Topographic surveys of the study area have been completed, and the locations of all existing shallow tube wells and dug wells have been mapped. Observation wells have been identified, their locations surveyed and monitoring has commenced. Water table and salinity maps have been prepared. Some preliminary groundwater sampling and water chemistry analysis has taken place, and further sampling is currently being carried out. Rain gauges have also been installed for collection of rainwater samples for chemical analysis. Comparison of rainfall and groundwater chemistry will permit assessment of the rate of recharge to the aquifers. A tendering process is currently in place for the installation of deep bores and multilevel piezometers. These will enable determination of vertical variation in salinity within the aquifer, and will help define the current extent of saline intrusion. They will also help determine the thickness of the aquifers, and the total volume of fresh water that they contain. Geophysical surveys have been carried out to assist mapping of salinity distribution and stratigraphy.
The total number of wells currently used for groundwater within the region has been determined, and rate of extraction from each well as been estimated, and the existing use of the water determined. A socioeconomic survey has been completed, involving a total of 121 respondents. Areas currently planted to various crops have been determined on the basis of this survey. A soil survey of the region has also been completed, which will assist with determining land suitability for expansion of current areas of cropping.
A three-day workshop on Groundwater Investigation and Management was conducted in Ilocos Norte from February 22-24, and attended by 25 staff from BSWM and other agencies involved with water management in the Philippines. The workshop concentrated on basic aquifer hydraulics (calculation of groundwater flow rates and flow directions from measurements of hydraulic head and hydraulic conductivity) and calculation of groundwater balances.
Work is also underway within Australia, to design artificial recharge systems for use in sandy aquifers that predominate in lowland areas of the Philippines.

Work on groundwater management within the Philippines has focused on two pilot sites within neighbouring municipalities of the province of Ilocos Norte, namely Burgos and Pasuquin, which is located on the northwestern tip of Luzon. For year 2, activities within these two sites focused on characterisation of the groundwater resource, and collection of data necessary to determine sustainable yield and to set up a groundwater model. The groundwater model will enable future impacts of groundwater extraction to be determined, as well as impacts of changed management. Pumping tests have been conducted to determine aquifer transmissivities, and pressure loggers have been installed in a number of wells to record fluctuations of water levels.
An economic assessment of agricultural productivity within the two field sites has been completed. Rice and garlic are the two main crops, with rice grown during the wet season, and garlic supplemented by groundwater grown during the dry season. Labour and fertilizer costs are the largest input costs for both rice and garlic production, although the cost of fuel for pumps is significant for garlic. Growing one hectare of garlic generates a total return of three to four times that of rice. In terms of gross margin, garlic provides a return five to six times greater than rice. Rice, however, remains a traditional crop for two reasons: it provides a staple food for household consumption, and it provides rice straw as mulching material for garlic.
Preparatory to the formulation and implementation of water use strategies, a two-day farmers' training was conducted in Barangay Susugaen, Pasuquin, Ilocos Norte last May 4-5, 2006. Attended by 26 farmers and 19 agricultural technicians from the study sites, lectures on water saving technologies (i.e., alternate wetting and drying or AWD for rice production and the use of aerobic rice) were presented. Field practicum on the installation of piezometers for monitoring available water in the field formed part of the training. Aerobic rice seeds (i.e., at 2 kilos per participants) were distributed to participants for them to reproduce seeds for bigger areas.
A three-day workshop on Groundwater Modelling was conducted in Manila from June 8-10. The first day was attended by twenty-six staff from BSWM and other agencies involved with water management in the Philippines. The second day involved hands-on demonstration of computer modelling, and involved eight staff working on four computers. The workshop discussed the principles of groundwater modelling, and the demonstrated the practical aspects of modelling with the MODFLOW software.
Work on groundwater modelling of the Burgos site is now in progress. When completed, the model will allow the hydrological impacts of a range of changed land and water management scenarios to be evaluated. Economic analyses of different management scenarios will also be conducted.
Work is also underway within Australia, to design artificial recharge systems for use in sandy aquifers that predominate in lowland areas of the Philippines. An evaluation of the performance of roughing filters in removal of colloidal kaolin as a pre-treatment for biofiltration (slow sand filtration) was completed. A review of biofiltration is nearing completion and a testing facility has been established adjacent the Urrbrae Agricultural High School wetland to enable evaluation of biofiltration design variables. Studies are in preparation to characterise labile nutrients in stormwater as a predictor of biofilm growth and clogging in injection wells. This will be used to design a water sampling and evaluation program in Ilocos Norte. A review of well completion techniques enabled design of wells for construction at the Urrbrae site for an ASR pilot study. Unfortunately the thin alluvial aquifer was not encountered in either well and it is proposed that a deeper well be installed once water quality testing has established that biofiltration is likely to be effective for an ASR trial. This would give confidence before application in Ilocos Norte.

Work on groundwater management within the Philippines has focused on two pilot sites within neighbouring municipalities of the province of Ilocos Norte, namely Burgos and Pasuquin, which are located on the northwestern tip of Luzon, Philippines. For year 3, activities within these two sites focused on construction and calibration of groundwater models, and scenario testing using these models. Also, preliminary water use and management strategies were formulated based on the results of the modelling under different scenarios and on field observations. Demand side water management (i.e., to reduce water consumption), aerobic rice seed was distributed to selected farmers of both Burgos and Pasuquin. Aerobic rice was introduced to reduce water consumption in rice production. During the Nov-Feb. 2006 cropping of garlic, a techno-demo farm on water management aspects for garlic production was established in both sites in which farmer-cooperators were taught how to monitor soil moisture in the root zone during the cropping period. They recorded soil tensiometer readings on a daily basis, and static and pump water levels at every irrigation application.

In June 2006, a training course in groundwater modelling was held in BSWM offices in Manila. A one day introductory seminar was attended by 26 staff from BSWM and other agencies concerned with water management. Eight of these staff then received hands-on training using the MODFLOW software. Following this course, Engineers Jane Fantilanan and Oscar Carpio took part in a three week advanced training program in Adelaide in November 2006, where they began construction of a groundwater model for the Burgos site, under the guidance of Dr Peter Cook (CSIRO) and Mr Paul Howe (REM). Paul Howe then spent three days at BSWM in Manila in June 2007, to assist Engrs Fantilanan and Carpio set up the model for the Pasuquin field site.

In the modelling exercise, the scenarios they simulated include the possibility of increasing the areas for garlic production during dry season cropping, assuming the current level of recharge in both basins continues. The possible effect of reduced rainfall due to drought and or dry spell (i.e., usually associated with El Nio phenomenon) was also studied. In the process, the number of wells and the possible expansion areas were determined such that future groundwater extraction will not exceed the rate of recharge. The initial results of the groundwater modelling will be presented to farmers and local policy and decision makers during the scheduled consultation-workshop in the third week of July.

Work is also underway within Australia, to design water pre-treatments for managed aquifer recharge systems for use in unconsolidated sandy aquifers that predominate in lowland areas of the Philippines. An evaluation of the performance of roughing filters in removal of colloidal kaolin as a pre-treatment for biofiltration (slow sand filtration) was completed. The method was found to be highly effective and required minimal maintenance. A review of biofiltration was undertaken and a testing facility established adjacent the stormwater harvesting wetland at Urrbrae Agricultural High School, South Australia into enable evaluation of biofiltration design variables. Stormwater from the local catchment and from Melbourne were used to test different biofiltration media. Removal of colloidal material was effective but removal of dissolved biodegradable organic carbon was only partially effective. This suggests that these treatments will be adequate for consolidated carbonate aquifers, such as at Burgos and parts of Pasuquin, but would only be effective for unconsolidated sand aquifers if the labile organic carbon content of water is low. A report is in preparation on methods to characterise labile nutrients in stormwater as a predictor of biofilm growth and clogging in injection wells. More direct measures of clogging potential are also in evaluation with the aim of producing a simpler and more direct assessment method that is relevant to local aquifers.

Work on groundwater management in the Philipppines has focused on two pilot sites within neighbouring municipalities of Pasuquin and Burgos, Ilocos Norte in the north western tip of Luzon, Philippines. Initial activities (Year 1) included the conduct of training on "Groundwater investigation and management" (Feb 22-24, 2004), topographic survey and mapping of wells, soil and land use surveys, geologic investigation (i.e., including geo-resistivity survey) and installation and establishment of measuring devices. In the process, water table maps and salinity maps were prepared and water chemistry analysis was done. In Pasuquin project site, 362 wells composed of 106 irrigation wells and 256 wells for household/domestic purposes were mapped. These wells tapped sandy to gravelly aquifers generally located within 5 -12 meters below ground surface. On the other hand, in Burgos project site, 184 wells were mapped with 85 irrigation wells and 99 individual household wells that tap limestone aquifers (i.e. within 2 mbgs to 30 mbgs). Elevated salinity was observed in the midstream section (or that portion being cultivated for crop production) particularly in Burgos project site. Average water level fluctuation during the observation period was 2.52 m in Burgos and 3.57 m in Pasuquin. With the monthly observation of water level fluctuation for the period 2005-2006, the probable groundwater recharge was calculated (i.e. rise in water level in response to a significant rainfall event). As percentage of rainfall, it was estimated that about 10% of rainfall recharged the aquifers in Pasuquin while 13%-17% of rainfall recharged the aquifers in Burgos during the rainy months of Jul to Oct.
In Year 2, the characterization of the groundwater resource, and collection of data necessary to determine sustainable yield and to set up a groundwater model in both sites were continued. Piezometers were installed at both sites to collect information on lithology, groundwater levels and water quality. Pumping tests were conducted to determine aquifer hydraulic properties (i.e. transmissivity and storativity), and pressure loggers were installed in piezometers to frequently record fluctuations of water level. During the same period, sub-surface lithology from upstream to the coastal zone was described, depth and thickness of aquifers were observed and correlated with the results of the geo-resistivity survey. An economic assessment of agricultural productivity within the two field sites was also undertaken. In essence, the results of the study showed that shallow tube wells in both study sites have strong positive contributions to improving farmer's income. Shallow groundwater is the most available water resource in the area and therefore needs proper protection and management.

For year 3, activities within the two sites focused on the construction and calibration of groundwater models. The models were used to determine future impacts of changes in groundwater extraction on the resource and demonstrate the level of uncertainty in the forecasted impacts. At the onset, a training course in groundwater modelling was held at BSWM, Manila on June 2006. The course consisted of one-day introductory seminar with 26 participants from BSWM and other agencies and a hands-on training using the MODFLOW software that was attended by eight project staff. Afterwards, 2 BSWM engineers, Ms. Jane Fantilanan and Mr. Oscar Carpio took part in a three-week advanced training program in Adelaide on November 2006, where they began developing a groundwater model for the Burgos site. The groundwater model for Pasuquin was set up in Manila with the assistance of Mr. Paul Howe who visited the Philippines in June 2007. They explored the possibility of increasing the areas for garlic production during dry season cropping considering the current level of recharge in both basins. The possible effect of reduced rainfall due to drought and or dry spell (i.e., usually associated with El Nio phenomenon) was also studied. The number of wells that could still be installed and the possible expansion areas (i.e. groundwater extraction is not greater than recharge with sufficient groundwater discharge at the coast to prevent saline water intrusion) were determined. Preliminary water use and management strategies were also formulated based on the result of modelling under different scenario and field observations. Relative to the demand side water management (i.e., to reduce water consumption) aerobic rice seeds were distributed to selected farmers of both Burgos and Pasuquin. Aerobic rice was introduced to reduce water consumption in rice production. During the Nov-Feb. 2006 cropping, a techno-demo farm on water management for garlic production was established in both sites in which farmer-cooperators were taught how to monitor the soil moisture in the root zone during the cropping period. On a daily basis, they recorded soil tensiometer readings and static and pumped water levels every irrigation application.

It was found in 2004/05 that groundwater use of 135,000m3 in Burgos was responsible for 66% of the PhP7.2 M total value of agricultural production, and in Pasuquin use of 286,000 m3 groundwater accounted for 69% of the PhP16.1 M total value of production. In addition 2 to 3 % of the rice production could be attributed to irrigation with groundwater before harvest.

In the groundwater modelling exercise, three separate models with different aquifer parameters were used to evaluate sustainable levels of sustainable groundwater pumping at each site. Each model fits equally well the field data on groundwater levels during the short period of model calibration, but results in a different estimate of sustainable yield of the groundwater system. Four future scenarios, as specific combinations of increased groundwater extraction, reduced recharge due to climate change and recharge augmentation were developed for each of the three models for each of the two sites. Further monitoring and evaluation will be needed to refine the estimate of sustainable yield and this can be done during a period of gradual expansion of groundwater use.

From among the three candidate models of each site, only one model was selected as the basis for economic modelling each site. The economic results therefore serve only as an example of the economic opportunities of increased or more efficient groundwater extraction, and are not regarded as reliable forecasts. This economic modelling can be repeated when the groundwater models are refined. It does serve as an indicator of what is likely and the process, now established, is easily repeatable.

The result of one of the simulations for Pasuquin indicated that sustainable extraction of groundwater is conservatively estimated to be of the order of 280,000 m3 which is equivalent to about 90 ha for garlic production intensification in highly suitable areas during the dry season. This area is similar to the current level of irrigation utilization and, by relocating some production to soils that are highly suitable, this can be translated into improved production of 810 metric tons per year amounting to an increase in production value (i.e. through shallow groundwater utilization) of PhP17.8 M. As yet there is no evidence of saline intrusion or of brackish groundwater at the base of any of the wells in Pasuquin so it is possible that the volume of extraction could be increased. However the area of soil suitable for growing high valued irrigated crops during the dry season is likely to constrain further irrigation development.

In Burgos, one of the selected groundwater models gave a mid-range prediction of sustainable yield of 190,000 m3. This is equivalent to 60 ha for garlic production intensification in highly and moderately suitable areas during the dry season. The current level of groundwater extraction is estimated at about 43 ha. Crop diversification with garlic and onion as major crops is also possible for an optimum service area of about 58 ha (i.e. since onion has higher water requirements than garlic). Garlic production intensification using shallow groundwater could be translated into increase in production of 186 metric tons per year for an increase in total production value of PhP6.5 M in highly and moderately suitable areas. In parts of Burgos there is evidence of brackish groundwater at the base of a number of wells and increasing salinity during pumping. Hence it is possible that the volume of extraction could only be increased in selected areas and that model predictions are optimistic. As in Pasuquin, the area of soil suitable for growing high valued irrigated crops during the dry season is likely to constrain further irrigation development to about the levels forecast in this model.

In essence, the levels of utilization of shallow groundwater resources in Pasuquin, and possibly also at Burgos, are considered to still be within their sustainable yields, subject to verification by ongoing monitoring. There are indications of competition in groundwater extraction during dry season (i.e. as revealed by farmers during consultation meetings) but could be corrected through proper pumping schedule. Saline water could be pumped at deeper zones (i.e. particularly in Burgos ) and could be prevented by limiting the depth of new wells. With farmers observations and some spread of uncertainties in the result of the modelling further monitoring is required to validate each model and differentiate which model best fits in each site. In the future, when the groundwater models are validated, the economic evaluation may be updated also.

In July 2007, a two-phase consultation was held. The first phase consisted of a technical forum and workshop with the farmers in both pilot sites. The second phase consisted of workshop output presentations to local policy/decision makers. The farmers emphasized their observations on the threat of deteriorating water quality (e.g. high salinity) and the declining water quantity during dry season. At the end, a Covenant of Support to protect and manage our shallow groundwater resource was passed by the farmers. This covenant was presented to local policy makers who showed their willingness to provide a parallel effort to protect and manage our shallow groundwater resource by espousing related local policies. Furthermore, everybody agreed that this should be referred to the National Water Resources Board (NWRB), the national agency that regulates the utilization, protection and management of our water resources.

Meanwhile, parallel works were also undertaken in Australia, to design artificial recharge systems for use in sandy aquifers that predominate in lowland areas of the Philippines. An evaluation of the performance of roughing filters in removal of colloidal kaolin as a pre-treatment for biofiltration (slow sand filtration) was completed. A review of biofiltration was completed and a testing facility was established at the Urrbrae Agricultural High School to enable evaluation of stormwater biofiltration design variables. A review of well completion techniques enabled design of wells for construction at the Urrbrae site for an ASR pilot study. Unfortunately the thin alluvial aquifer was not encountered in either well and it is proposed that a deeper well be installed once water quality testing has established an effective pre-treatment for an ASR trial. This would give confidence before application in Ilocos Norte. Studies have identified indicators of labile nutrients in stormwater to predict biofilm growth and clogging in injection wells. Testing with combinations of roughing filters and biofiltration in either sand or granular activated carbon found that a suitable quality of water for injection into fine grained aquifers could not be obtained using natural materials alone.

Hence it is concluded that limestone, as found in parts of Pasuquin and much of Burgos could be a suitable target for recharge enhancement using roughing filters as a pre-treatment for settled stormwater but fine-grained alluvial aquifers are not viable as a target due to excessive pre-treatment costs. Alternatives such as pond infiltration using stilling basins and recharge basins may also be considered where aquifers are unconfined and soils are free-draining. No recharge should be considered within reasonable proximity (3-6 months travel time) of wells used as drinking water supplies to avoid potential contamination. Field measurements and modelling suggest that aquifers refill during the wet season and discharge to streams, at least in Burgos. Hence enhancing recharge during the wet season is not considered to be useful unless the aquifer is not fully replenished by natural recharge. If the dominant cause for salinisation of wells is up-coning from deeper saline groundwater during pumping, it is doubtful that additional recharge in the wet season will be effective against salinisation. However restricting rates of pumping, spatial intensity of pumping and well depths will be beneficial.

The project has successfully assessed the potential of shallow groundwater resource in both pilot sites. Through modelling exercises (i.e. groundwater and economic modelling) the limit of future groundwater extraction was projected. Public consultations increased the level of awareness of farmers and decision makers on the current situation through which management strategies were formulated. However, the bottom line is that farmers as groundwater users should also learn how to monitor the impacts of unregulated groundwater extraction to the resource (i.e. in terms of potential groundwater level decline and increasing salinity) so that they could undertake immediate remediation measures themselves. They should be encouraged to contribute in the overall management of groundwater as "resource managers" to ensure a more sustainable use of this very precious resource. A project extension aims to support the farmers in managing their groundwater as a sustainable resource.