Genetic Characterization of Philippine Saba Germplasm ...philjournalsci.dost.gov.ph/.../vol149no1/...saba_.pdf · collection of Saba accessions maintained at the National Plant Genetic

Keywords: bananas, genetic diversity, Musa acuminata, Musa balbisiana, SSR markers

*Corresponding Author: [email protected] [email protected]

Genetic Characterization of Philippine Saba Germplasm Collection Using Microsatellite Markers

Philippine Journal of Science149 (1): 169-197, March 2020ISSN 0031 - 7683Date Received: 26 July 2019

Fides Marie R. dela Cruz1, Rita P. Laude1*, Carlo Miguel C. Sandoval2,3, Lavernee S. Gueco2, Visitacion C. Huelgas2, Roberta N. Garcia2,3,

and Evelyn Mae T. Mendoza2,3

1Institute of Biological Sciences, College of Arts and Sciences University of the Philippines Los Baños, Los Baños, Laguna 4031 Philippines

2Institute of Plant Breeding, College of Agriculture and Food Science University of the Philippines Los Baños, Los Baños, Laguna 4031 Philippines

3Institute of Crop Science, College of Agriculture and Food Science University of the Philippines Los Baños, Los Baños, Laguna 4031 Philippines

The genetic diversity among 75 Philippine Saba gene bank collections and four non-Saba accessions were analyzed using microsatellite markers or simple sequence repeats (SSRs). Eleven primer pairs were found to be polymorphic across all genotypes and amplified a total of 28 alleles. The discriminatory power of the primers was moderately high with mean polymorphism information content (PIC) value of 0.38. Genetic diversity was measured in terms of heterozygosity, fixation index, and genetic distance. Average gene diversity or expected heterozygosities ranged from 0.09–0.58, with an average of 0.48. Observed heterozygosities ranged from 0.09–0.95, with a mean of 0.80. Fixation indices (mean = –0.61) were negative, which suggests an excess of heterozygotes and shows that diversity has been maintained within the gene bank collection. Cluster analysis revealed that the greatest genetic distance (GD = 0.69) was observed between Lakatan (AAA) and Cardaba 6 (ABB/BBB) bananas. Therefore, the SSR markers used were able to separate the Saba cultivars from the non-Saba cultivars. Furthermore, the markers were able to separate the seedy diploid cultivar Pacol from the rest of the triploid Saba cultivars. Future studies can survey other potential SSRs for better discrimination among Saba samples.

INTRODUCTIONBananas (Musa species) are grouped according to the number of their chromosome sets and the relative proportion of Musa acuminata and Musa balbisiana, with genomes denoted as AA and BB, respectively. Modern-

day bananas are products of hybridization among M. acuminata subspecies, or between M. acuminata and M. balbisiana, and are normally polyploids (triploids or tetraploids). Triploids are classified as AAA, AAB, and ABB and tetraploid as AAAA, AAAB, AABB, or ABBB. However, bananas of the species M. balbisiana (BB and BBB) are consumed in the Philippines (Ploetz et al. 2007).

169

The Philippines lies within the area of the world’s greatest diversity in Musa. Thus, it holds the germplasm of various banana genotypes. According to Valmayor et al. (2002), there are 91 banana cultivars of local origin – including one M. fehi cultivar, two unclassified species, 45 M. acuminata cultivars, 29 M. acuminata x M. balbisiana cultivars, and 14 M. balbisiana cultivars. Four cultivars of commercial importance are known locally as Saba. The seedless, edible, triploid Saba cultivars that evolved through chromosome restitution must also be recognized under the same species as their ancestors since the addition of one set of chromosomes through autopolyploidy did not introduce anything new to the genetic constitution of the clones (Valmayor et al. 2002). The Saba varieties are grown all over the Philippine archipelago, covering 182,000 hectares in 2015 with a production of 2.63 million MT and 28.90% share of total banana production in the country (PSA 2017).

M. balbisiana is fast-growing and tolerant to environmental stressors, making the crop easy to grow. However, the existence of pure BBB cultivars in the Philippines has been questioned. A previous study using flow cytometry and molecular markers reports that Saba subgroup is triploid and BBB (Sales et al. 2011); however, another study using the same techniques reports Saba (S14 ITC1745) genome as ABB (Christelová et al. 2017). In 1987, Silayoi and Chochalow identified 15 diagnostic characters used to differentiate M. acuminata clones from M. balbisiana cultivars and their hybrids. One of the characters is bract curling, where the M. acuminata exhibits bract reflex and rolling back after opening while M. balbisiana have bract lifting but not rolling. The typical Saba/Cardaba cultivars have bracts lifting and rolling which gives a hint that the cultivars may have some A genomic compositions as well. Another characteristic is pseudostem color – some Saba/Cardaba cultivars have some blotches, which is typical of M. acuminata species.

Due to the multitude of clones being propagated, various local names, synonyms, and variants for bananas exist all over the Philippines – making it difficult to assess the genetic resources available for Saba in the country. However, in recent years, molecular markers have been useful in establishing the extent of genetic diversity, determining levels of heterozygosity, and serving as molecular fingerprints for hybrid and cultivar identification. The most widely used marker for Musa has been SSRs due to the simplicity of the protocol and its effectiveness in determining genetic diversity within species. These markers consist of tandem repeating nucleotide units about two to six bases long and are valuable as molecular markers due to their abundance in the genome, co-dominant transmission, and the requirement for only a small amount of starting DNA (Creste et al. 2003). SSRs effectively assess the genetic

diversity within species or populations, specifically for different Musa germplasm collections (Christelová et al. 2017; de Carvalho Santos et al. 2019) because the allelic patterns at each SSR locus can be used to determine various measures of genetic diversity – namely, heterozygosity (H), genetic distance (GD), and fixation indices (F). Heterozygosity (H) can be used as a parameter to assess the existing genetic variation within a population. The observed heterozygosity (Ho) is often compared to the expected (He) under the Hardy-Weinberg equilibrium, and any discrepancy is attributed to either inbreeding or selective forces. The analysis of GD, on the other hand, estimates the degree of genetic difference by determining the average number of nucleotide differences across various loci for the generation of phylogenetic trees. Fixation indices measure the amount of subdivision that exists within populations (Sbordoni et al. 2012).

This study aimed to determine the extent of genetic diversity among the existing Philippine Saba accessions in the banana field gene bank collection from the three major groups of islands in the Philippines. The present study also aimed to identify molecular markers that can discriminate the genetic elements of the A and B genomes in Musa varieties.

MATERIALS AND METHODS

Collection of Philippine Banana Leaf SamplesPhilippine banana leaf samples were collected from existing collections maintained at the National Plant Genetic Resources Laboratory (NPGRL), Institute of Plant Breeding (IPB), College of Agriculture and Food Science (CAFS), University of the Philippines Los Baños (UPLB) in Los Baños, Laguna from Nov–Dec 2012. The NPGRL conserves cultivars and wild types of Saba collected from various parts of the country. Each accession has passport data and characterization of some morphological descriptors of banana (Musa spp.) and M. balbisiana genotypes.

Samples included 75 Saba accessions obtained from various regions in the Philippines as well as four non-Saba cultivars (Appendix I Table 1, Figure 1) to serve as control or outliers for genetic analysis. Fresh, young leaves were obtained from each accession and stored at 4 °C. A total of 79 specimens were collected and processed at the Biochemistry Laboratory, IPB, CAFS, UPLB for DNA extraction.

DNA Extraction, Amplification, and DetectionTotal genomic DNA was extracted following a modified cetyltrimethylammonium bromide protocol described by Doyle and Doyle (1987). DNA quality and quantity were

dela Cruz et al.: Genetic Characterization of Philippine Saba

Philippine Journal of ScienceVol. 149 No. 1, March 2020

170

determined using absorbance ratios (A260/280) obtained using SmartSpecTM3000 spectrophotometer (Bio-Rad Laboratories Inc., CA, USA).

SSR amplification was performed using primer pairs that have been previously screened for the presence of polymorphic markers across Philippine banana cultivars

(Doloiras-Laraño et al. 2018). Each polymerase chain reaction (PCR) tube contained 50 ng DNA template, 0.4 µM forward primer, 0.4 µM reverse primer, 0.2 mM dNTPs (Invitrogen, CA, USA), 0.5 U Taq polymerase (Vivantis Technologies Sdn. Bhd., Malaysia), 2 mM MgCl2, and 1X PCR Buffer A with a final volume of 12.5 μL. The PCR conditions followed were those of Kaemmer et al. (1997) with a few modifications. The PCR amplification using Musa-based primers was performed at 94 °C for 4 min for the initial denaturation, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at varying temperatures [optimized for each primer pair as described by Doloiras-Laraño et al. (2018)] for 30 s, and 1 min at 72 °C extension with a final extension at 72 °C for 10 min. The present profile had 1 min at 72 °C extension while Kaemmer et al. (1997) used 30 s at 72 °C. The primer profiles are summarized in Table 1. PCR products were resolved in 2–4% agarose gels using the Bio-Rad Sub-Cell Model 192 electrophoresis system set at 70V. Gels were stained and fixed using ethidium bromide solution. The SSR bands were viewed under UV light using the GelDocTM system (Bio-Rad Laboratories Inc.).

Data AnalysisPolymorphism information content. To determine the degree of variability at each SSR locus, the PIC was computed as follows:

(1)

where p = frequency of the ith allele out of the total number of alleles at an SSR locus

n = total number of different alleles for that locus.

Figure 1. Sites (yellow dots) in different parts of the Philippines of collection of Saba accessions maintained at the National Plant Genetic Resources Laboratory, Institute of Plant Breeding, College of Agriculture and Food Science, University of the Philippines Los Baños.

Table 1. The profile of ten Musa-based and one papaya-based SSR primers used for diversity analysis of Philippine Saba gene bank collection.

SSR ID Motifa Primer sequence (5′–3′) Ta (°C)

Expected

size (bp)

GenBank accession

mMaCIR 13 (GA)16N76(GA)8

F: TCCCAACCCCTGCAACCACT

R: ATGACCTGTCGAACATCCTTTT53 279 X90745b

mMaCIR 152 (CTT)18

F: CCACTTTGAGTTCTCTCC

R: TTTCCCTCTTCGATTCTGT54 175 AM950442c

mMaCIR 260 (TG)8

F: GATGTTTGGGCTGTTTCTT

R: AAGCAGGTCAGATTGTCCC55 230 AM950515c

mMaCIR 25 –F: GTGGTTTGGCAGTGGAATGGAA

R: TGACCCTCCGACACCTATTTGG55 359 –

MA 3-90 (TC)7

F: GCACGAAGAGGCATCAC

R: GGCCAAATTTGATGGAT53 157 Z85972b



171

PIC values greater than 0.50 refer to highly polymorphic SSR loci, while values less than 0.5 have a lower degree of locus polymorphism. Loci that are highly polymorphic also provide the greatest discriminatory power for subsequent DNA fingerprinting, cultivar identification, and determination of genetic relationships.

Genetic diversity analysis. For each SSR locus, only the alleles with molecular sizes within the expected sizes were considered and scored to eliminate non-specific amplicons that were not removed during optimization. PowerMarker V3.25 software (Liu and Muse 2005) was used to compute Nei’s genetic distance (Nei and Takazaki 1983) and subsequently assess the degree of genetic diversity among the Saba and non-Saba accessions.

Genetic relationships were determined by cluster analysis using the Unweighted Pair Group Method Average (UPGMA). The same software was used to generate 1,000 trees for bootstrap analysis. A final consensus tree was generated using the Phylip V3.695 software (Felsenstein 2005).

Statistical analysis. Genetic parameters were computed to estimate the level of genetic diversity. Gene diversity or expected Heterozygosity (He) and observed heterozygosity (Ho) were calculated using the PowerMarker V3.25 software. These values were used to manually compute fixation indices (F) of the Philippine banana field gene bank collection using the formula:

(2)

According to Peakall and Smouse (2012), F values close to zero are expected under random mating. Positive values

indicate the presence of inbreeding while negative values mean there is an excess of heterozygosity.

RESULTS AND DISCUSSION

Characterization of SSR MarkersA total of 28 distinct and reproducible alleles were amplified across the 79 Philippine banana accessions (Appendix II Figure 1). Two to three alleles were amplified for each primer. PIC was computed to determine the informativeness of each SSR marker. Table 2 shows that PIC values ranged from 0.09–0.48, with an average value of 0.38. The values indicate a relatively good discriminatory power of the markers used. The highest PIC value was observed with primer mMaCIR 22 where three alleles were amplified. The lowest observed value was observed in BGAL, with only two alleles. Due to the relatively high PIC values observed, the markers can be considered good determinants of genetic similarity, distance, and variation among the different Philippine banana accessions. The PIC values were slightly lower than the observed 0.68–0.98 by Dacumos et al. (2011) among Philippine bananas but were more or less consistent with values (0.23–0.80, mean = 0.55) observed by Doloiras-Laraño et al. (2018) in Philippine M. balbisiana cultivars.

Genetic Diversity Analysis Heterozygosity and fixation index. Gene diversity, also known as expected heterozygosity (He), is considered the probability that two randomly chosen alleles from the population are different (Liu and Muse 2005). The average gene diversity or He of the 79 Philippine banana accessions from Luzon, Visayas, and Mindanao was

mMaCIR 01 (GA)20

F: TGCTGCCTTCATCGCTACTA

R: ACCGCACCTCCACCTCCTG55 252 X87262b

pMaCIR 348b –F: ACAGAATCGCTAACCCTAATCCTCA

R: CCCTTTGCGTGCCCCTAA53 181 –

pMaCIR 108 –F: TTTGATGTCACAATGGTGTTCC

R: TTAAAGGTGGGTTAGCATTAGG55 248 –

mMaCIR 22 –F: AAGTTAGGTCAAGATAGTGGGATT

R: CTTTTGCACCAGTTGTTAGG50 381–424 –

mMaCIR 39 (CA)5GATA(GA)5

F: AACACCGTACAGGAGTCAC

R: GATACATAAGGGTCACATTG53 350 Z85970b

BGAL –F: CCGCGGCAAGACTATCATGG

R: TTGACTCCCGTTCTCCATCTC55 250–290 –

aBlanks (–) indicate that the SSR motif was not provided in the reference.



172

He = 0.48 and ranged from 0.09–0.58 (Table 3). High gene diversity was also previously reported by Ravishankar et al. (2013) using 63 SSR markers on 30 accessions of M. acuminata, M. balbisiana, and their natural hybrids, with a range of 0.14–0.93 (mean = 0.87). Further, Amorim and colleagues (2012), observed similar He values ranging from 0.09–0.90 with an average of 0.65 in 22 Musa spp. genotypes using 34 SSR markers. However, the average He was higher than what has been reported in Philippine abaca (Musa textilis) accessions (mean = 0.40; Yllano et al., pers. comm.) and South Indian Musa cultivars

(mean = 0.42; Resmi et al. 2011).

In this study, observed heterozygosities (Ho), were found to be significantly high, with values ranging from 0.09–0.95 at an average value of 0.80. The high Ho values indicate that a large proportion of the population is a heterozygote. This, in turn, leads to a high degree of allelic polymorphism in the population. In comparison, Ho values were significantly higher than He values indicating an excess of heterozygosity in the population, as reflected by the negative mean fixation index (F) computed for the population (F = -0.61). Excess of heterozygosity was also reported among populations of M. balbisiana in China (Ge et al. 2005). Yllano et al. (pers.comm.) also observed negative fixation indices in Philippine abaca (M. textilis) populations.

According to Stevens et al. (2007), excess of heterozygosity may be due to the polyploid nature of the species. Heterozygote advantage or overdominance of the marker loci or other loci linked to marker loci may also lead to an excess of heterozygosity. In theory, the excess increases in the population over time as selective forces favor the deletion of non-advantageous recessive alleles (Stoeckel et al. 2006). However, overdominance may not always explain the phenomenon since SSRs are generally located within noncoding regions (Kashi and King 2006). It is more likely that excess of heterozygosity is due to obligate cross-pollination among Musa species (Valmayor et al. 2002). Furthermore, the collection and conservation of strains that are morphologically distinct from the rest may have also contributed to this result. The excess of heterozygosity reflects the considerable diversity within the Philippine banana field and tissue culture gene bank collections maintained at NPGRL, IPB, CAFS, UPLB.

Genetic distance. The average genetic distance (GD) among all Saba and non-Saba accessions was 0.17 (Appendix I Table 2). The value is indicative of low genetic distance within M. acuminata and M. balbisiana. This is consistent with the cross-compatible nature of species under the genus Musa. Some modern edible bananas are interspecific hybrids (AB, AAB, ABB, and ABBB) between wild ancestors of the two species (AA or AAA and BB or BBB), with several identical genetic elements (Valmayor et al. 2002). The greatest genetic distance (GD = 0.69) was observed between Lakatan (AAA) and Cardaba 6 (ABB/BBB), followed by the genetic distance (GD = 0.62) between Lakatan and Tiparot (ABBB). The GD values were greater than 0.50 showing that despite the inherent genetic similarities between M. acuminata and M. balbisiana, there is significant genetic variability between the two species. This was also observed in other non-Saba cultivars such as Latundan (AAB) which was most distant (GD = 0.65) to Butuhan (BBw). A GD = 0.55 was observed between Saging Pula (AAA) and Cardaba

Table 2. Summary of alleles observed among Saba and non-Saba accessions from NPGRL gene bank collection using 11 SSR primers and their respective polymorphism information content values.

SSR ID No. of alleles Relative band size (bp) PIC

mMaCIR 13 3 270–320 0.45

mMaCIR 152 3 190–230 0.44

mMaCIR 260 2 190–210 0.37

mMaCIR 25 3 230–480 0.36

Ma 3-90 2 240–260 0.37

mMaCIR 01 3 220–300 0.46

pMA 348b 2 200–240 0.37

pMA 108 3 230–310 0.42

mMaCIR 22 3 390–450 0.48

mMaCIR 39 2 310–330 0.37

BGAL 2 290–310 0.09

Total 28 – Mean: 0.38

Table 3. Observed heterozygosity (Ho), expected heterozygosity (He), and fixation index (F) values among 79 Philippine Saba and non-Saba accessions from the National Plant Genetic Resources Laboratory gene bank collection.

Locus Ho He F

mMaCIR 13 0.95 0.55 –0.72

mMaCIR 152 0.88 0.55 –0.62

mMaCIR 260 0.90 0.49 –0.82

mMaCIR 25 0.70 0.48 –0.46

Ma 3-90 0.85 0.49 –0.74

mMaCIR 01 0.86 0.56 –0.54

pMA 348b 0.94 0.50 –0.88

pMA 108 0.88 0.53 –0.66

mMaCIR 22 0.83 0.58 –0.43

mMaCIR 39 0.89 0.50 –0.79

BGAL 0.09 0.09 –0.05

Mean 0.80 0.48 –0.61



173

6 (ABB/BBB) and GD = 0.47 between Sulay Baguio (AAA) and Sabang Binong (ABB/BBB). On the other hand, a GD = 0.44 was observed between Pacol (BBw), a wild seeded M. balbisiana, and Banana 1.

Cluster analysis. Bootstrap analysis of 1,000 trees divided the Philippine banana samples into statistically significant and informative clusters (P > 0.50). The highest bootstrap value (P = 88) was observed in a cluster including a Bohol accession of Saba (Bohol 2) and Cadisnon/Cardaba 1 (Figure 2). Two accessions, namely, Saba Nuang/Datu and Saba Aurora 6, also grouped together at P = 74, indicating a high degree of similarity between the two species. Cluster analysis also did not segregate the genotypes according to the geographic place where the different accessions originated. The lack of clustering into distinct geographical populations can be attributed to the existing clonal propagation practices in the country. Transfer of planting material from one geographical area to the next is easily facilitated and populations tend to overlap.

Cluster analysis revealed three sub-groups – Saba/Cardaba (ABB/BBB), Gubao/Bluggoe, and non-Saba. The first sub-group (denoted as cluster 1) was composed of accessions with triploid ABB/BBB genomic constitution – including all Saba, Cardaba, Dippig, Cadisnon, Kalimpos, Luyluy, and Dalian samples. These are their local names as given by the farmers during collecting missions and written in the passport data (Appendix I Table 1). Saba cultivars can be distinguished morphologically by the rolled bracts of the male bud and the pigmented pseudofruit. Our molecular data corroborate the results of Pillay et al. (2000) who reported that clones of Saba are genetically ABB or BBB. The second sub-group (Cluster 2) was composed of Gubao/Bluggoe accessions – including Pondol, Sabang Binong, Saba Nuang/Datu, and Susong Dalaga. In addition, Banana 1, Banana 2, Sabang Pulpol 1, and Saba Aurora 6 might be the local names given by farmers during collecting trips. The Bluggoe subgroup consists of starchy cultivars used primarily for cooking but can also be eaten raw. The Saba is usually placed under this subgroup; however, these group seems to be morphologically different from the typical Saba. The cultivars found in the Philippines under the Gubao/Bluggoe subgroups are usually dark green in fruit peel color, are shorter in plant stature, and have fruits that are sour and watery when ripe. However, the diploid BB wild Pacol clustered with the remaining non-Saba accessions. This shows that the markers used in the study are unable to distinguish copy numbers of the genome and more markers may be needed to separate Pacol from the rest of the triploid non-Saba accessions.

Cluster analysis also separated Tiparot (ABB) as the most distant cultivar among all Saba accessions. In Tiparot, male buds are unstable; they are either present or absent.

Figure 2. Consensus dendrogram (UPGMA) showing genetic relationships among Philippine Saba and non-Saba accessions from the National Plant Genetic Resources Laboratory gene bank collection using Nei’s genetic distance and significant bootstrap (P) values. Cluster/sub-group: I – Saba/Cardaba, II – Gubao/Bluggoe, and III – outliers.



174

In some cases, Tiparot produces two bunches in one plant. It clustered together with Cardaba 6. It is possible that Cardaba 6 is also a Tiparot, but the farmer only named it as Cardaba. However, the separation of this cultivar from the rest of the Saba accessions is attributed to the presence of a single copy of the A genome. The out-grouping of non-Saba cultivars Lakatan (AAA) and Sulay Baguio (AAA) further demonstrates the ability of the SSR markers to amplify alleles exclusive to the A genome. Saging Pula, also known as Murado is a triploid M. acuminata cultivar (AAA), which was separated in the group.

The Saba sub-group used in the study came from three geographically separated gene pools – namely Luzon (N = 36), Visayas (N = 6), and Mindanao (N = 18) – with two accessions of unknown sources. Figure 2 shows that Sab-a Manila is the most distant among all Luzon accessions, Sab-a 2 is the most distant among Visayas accessions, and Dwarf Cardaba is the most distant for the Mindanao samples. The greater number of Luzon accessions in this cluster is due to its greater sample size. In the Philippines, Mindanao is the center of large-scale banana farming, including cooking type bananas. Thus, the most number of Saba cultivars are also found here, which is indicative of the great diversity in both gene pools.

The Saba/Cardaba sub-group separated from the Gubao/Bluggoe type, which includes four cultivars from Mindanao (Figure 2). The last sub-group is composed of Lakatan (AAA), Pacol (BBw), Sulay Baguio (AAA), and Latundan (AAB). Lakatan and Sulay Baguio are expected to cluster together since they are both AAA, while Latundan (AAB) has 2 A genomes and 1 B genome may possibly have a closer affinity with Lakatan and Sulay Baguio. Surprisingly, Pacol should not have been in this cluster since this is a wild diploid M. balbisiana. These outliers may be explained by the observed low PIC values of some SSRs used in this study (Table 2). Hence, future studies can survey other potential SSRs for better discrimination among Saba samples.

CONCLUSIONThe 62 Philippine Saba accessions maintained at the NPGRL, IPB, CAFS, UPLB were grouped into one cluster based on the 11 SSR markers used. The observed heterozygosities (Ho) were significantly greater than the expected heterozygosities (He). Furthermore, fixation indices were generally negative in value, indicating an excess of heterozygosity in the collection. This is attributed to the collection of samples from different geographic locations with different climate and growth conditions (e.g. soil type, altitude, etc.), long cultivation with possible spontaneous mutation, and conservation of

unique genotypes that may have led to the accumulation of more heterozygous alleles in the gene bank compared to natural populations of Saba. Therefore, the collection can be considered a valuable genetic resource for future conservation and breeding programs. The SSR markers used were able to separate the Saba cultivars from the non-Saba cultivars. Furthermore, the markers were also able to separate the seedy diploid cultivar Pacol from the rest of the triploid Saba cultivars. Future studies can survey other potential SSRs for better discrimination among Saba samples.

ACKNOWLEDGMENTSThe authors gratefully acknowledge the grant (DABIOTECH-R1132) from the Department of Agriculture Biotechnology Program to EM Tecson-Mendoza and the support of the IPB, CAFS, UPLB. Fides Marie R. dela Cruz sincerely thanks Dr. Ma. Genaleen Diaz and Dr. Antonio Laurena for their mentorship and invaluable inputs towards the completion of this study. The invaluable inputs of Asst. Professor Aprill P. Manalang, Ms. Genevieve Mae B. Aquino, and Ms. Cecille Ann L. Osio – as well as the technical assistance of Ms. Arnelyn Doloiras-Laraño, Ms. Cleotilde A. Caldo, and Ms. Flordeliza C. Mendoza – are likewise acknowledged with thanks.

STATEMENT ON CONFLICT OF INTERESTThe authors declare no conflict of interest with any financial, personal, or other relationships with people or organizations related to the material discussed in the manuscript.

NOTES ON APPENDICESThe complete appendices section of the study is accessible at http://philjournsci.dost.gov.ph

REFERENCESAMORIM EP, SILVA PH, FERREIRA CF, AMORIM

VBO, SANTOS VJ, VILARINHOS AD, SANTOS CMR, SOUZA JUNIOR MT, MILLER RNG. 2012. New microsatellite markers for bananas (Musa spp). Gen Mol Res 41(2): 1093–1098.



175

CHRISTELOVÁ P, DE LANGHE E, HŘIBOVÁ E, ČÍŽKOVÁ J, SARDOS J, HUŠÁKOVÁ M, SU-TANTO A, KEPLER AK, SWENNEN R, ROUX N, DOLEŽEL J. 2017. Molecular and cytological char-acterization of the global Musa germplasm collection provides insights into the treasure of banana diversity. Biodivers Conserv 26(4): 801–824.

CRESTE S, NETO AT, SILVA SO, FIGUEIRA A. 2003. Genetic characterization of banana cultivars (Musa sp.) from Brazil. Euphytica 132: 259–268.

DACUMOS CN, LALUSIN AG, NAMUCO LO, PAT-ENA LF, BARBA RC. 2011. Diversity analysis of Philippine bananas using simple sequence repeats markers. Philipp J Crop Sci 36 (11): 1–10.

DE CARVALHO SANTOS TT, DE OLIVEIRA AMORIM VB, DOS SANTOS-SEREJO JA, DA SILVA LEDO CA, HADDAD F, FERREIRA CF, AMORIM EP. 2019. Genetic variability among autotetraploid populations of banana plants derived from wild diploids through chromosome doubling using SSR and molecular mark-ers based on retrotransposons. Mol Breed 39(7): 95.

DOLOIRAS-LARAÑO AD, GARCIA RN, SANDOVAL CMC, LALUSIN AG, GUECO LS, HUELGAS VM, TECSON-MENDOZA EM. 2018. DNA fingerprinting and genetic diversity analysis of Philippine Saba and other cultivars of Musa balbisiana Colla using simple sequence repeat markers. Philipp J Crop Sci 43: 1–11.

DOYLE JJ, DOYLE JL. 1987. Isolation of plant DNA from fresh tissue. Phytochemical Bulletin 19: 11–15.

FELSENSTEIN J. 2005. PHYLIP (Phylogeny Inference Package) version 3.6. Distributed by the author. Depart-ment of Genome Sciences, University of Washington, Seattle.

GE XJ, LIU MH, WANG WK, SCHAAL BA, CHIANG TY. 2005. Population structure of wild bananas, Musa balbisiana, in China determined by SSR fingerprinting and cpDNA PCR-RFLP. Mol Ecol 14: 933–944.

KAEMMER D, FISCHER D, JARRET RL, BAURENS FC, GRAPIN A, DAMBIER D, NOYER JL, LA-NAUD C, KAHL G, LAGODA P. 1997. Molecular breeding in the genus Musa: a strong case for STMS marker technology. Euphytica 96: 49–63.

KASHI Y, KING DG. 2006. Simple sequence repeats as advantageous mutators in evolution. Trends Genet 22(5): 253–258.

LAGODA PJ, NOYER JL, DAMBIER D, BAURENS FC, GRAPIN A, LANAUD C. 1998. Sequence tagged microsatellite site (STMS) markers in the Musaceae. Mol Ecol 7(5): 659–63.

LIU K, MUSE SV. 2005. PowerMarker: integrated analy-sis environment for genetic marker data. Bioinformat-ics 21(9): 2128–2129.

NEI M, TAKEZAKI N. 1983. Estimation of genetic dis-tances and phylogenetic trees from DNA analysis. Proc 5th World Cong Genet Appl Livest Prod 21: 405–412.

PEAKALL R, SMOUSE PE. 2012. GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research – an update. Bioinformatics 28: 2537–2539.

PILLAY M, NWAKANMA DC, TENKOUANO A. 2000. Identification of RAPD markers linked to A and B genome sequences in Musa L. genome 43(5): 763–767.

PLOETZ RC, KEPLER AK, DANIELS J, NELSON SC. 2007. Banana and plantain—an overview with empha-sis on Pacific island cultivars. In: Species Profiles for Pacific Island Agroforestry. Elevitch CR ed. Holualoa, HI: Permanent Agriculture Resources (PAR). Retrieved on 18 Jun 2019 from http://www.traditionaltree.org/

[PSA] Philippine Statistics Authority. 2017. Selected Sta-tistics on Agriculture. Retrieved on 28 Dec 2018 from https://psa.gov.ph/sites/default/files/SSA2017%20%281%29.pdf/

RAVISHANKAR KV, RAGHAVENDRA KP, ATHANI V, REKHA A, SUDEEPA K, BHAVYA D, SRINIVAS V, ANANAD L. 2013. Development and characteriza-tion of microsatellite markers for wild banana (Musa balbisiana). J Hort Sci Biotech 88(5): 605–609.

RESMI L, KUMARI R, BHAT KV, NAIR AS. 2011. Molecular Characterization of genetic diversity and structure of South Indian Musa cultivars. Int J Botany 7(4): 274–282.

SALES EK, BUTARDO NG, PANIAGUA HG, JANSEN H, DOLOZEL J. 2011. Assessment of ploidy and ge-nome constitution of some Musa balbisiana cultivars using DArT markers. Philipp J Crop Sci 36: 11–18.

SBORDONI V, ALLEGRUCCI G, CESARONI D. 2012. Encyclopedia of Caves, 2nd ed. Academic Press. p. 608–618.

SILAYOI B, CHOMCHALOW N. 1987. Cytotaxonomic and morphological studies of Thai banana cultivars. In: Proc Banana and Plantain Breeding Strategies. Persley GJ, De Langhe EA eds. Canberra, Australia.

STEVENS L, SALOMON B, SUN G. 2007. Microsatel-lite variability and heterozygote excess in Elymus-trachycaulus populations from British Columbia in Canada. Biochem Syst Ecol 35: 725–736.

STOECKEL S, GRANGE J, FERNANDEZ-MAN-JARRES JF, BILGER I, FRASCARIA-LACOSTE



176

N, MARIETTE S. 2006. Heterozygote excess in a self-incompatible and partially clonal forest tree spe-cies – Prunus avium L. Mol Ecol 15(8): 2109–2118.

VALMAYOR RV, ESPINO RRC, PASCUA OC. 2002. The Wild and Cultivated Bananas of the Philippines. Los Baños, Laguna: Philippine Agriculture and Re-sources Research Foundation, Inc. and Bureau of Agricultural Research. 242p.



177

Keywords: composting, heavy metals, helminths, sludge, total coliform, vermicomposting

*Corresponding Author: [email protected]

Efficiency of Combined Co-composting, Vermicomposting, and Drying in the Treatment of Cadmium, Mercury, Helminths,

and Coliforms in Sludge from Wastewater Facilities for Potential Agricultural Applications

Philippine Journal of Science149 (1): 179-188, March 2020ISSN 0031 - 7683Date Received: 13 Aug 2019

Maria Aileen Leah G. Guzman1*, May Ann A. Udtojan1, Marylle F. Del Castillo1, Emilyn Q. Espiritu1, Jude Anthony N. Estiva2, Jewel Racquel S. Unson1,

Joan Ruby E. Dumo1, and Jay Roy E. Espinas1

1Department of Environmental Science, Ateneo de Manila University, Loyola Heights 1108 Quezon City, Philippines

2Aparri Engineering LLC, 131 Main St., Suite 180, Hackensack NJ 07601 USA

Sludge generated from wastewater treatment facilities has been applied in agriculture as soil conditioners. However, the incomplete and/or inappropriate treatment of wastewater may result in sludge that may still contain heavy metals, helminth ova, and coliforms posing a risk to both humans and the environment. This study assessed various pretreatment techniques such as co-composting, vermicomposting, and a combination of these on sludge samples to remove heavy metals (cadmium and mercury), helminth ova, and coliforms. Physico-chemical and biological analyses were used to compare untreated (i.e. raw) and treated sludge samples. The results showed that for the raw sludge, mercury (4.02 +/– 0.17 mg/kg) and cadmium (6.30 +/– 0.48 mg/kg) exceeded the limits specified under the Philippine National Standard (PNS) for Organic Soil Amendments of 2 mg/kg and 5 mg/kg, respectively. Laboratory examinations also revealed the presence of helminth ova (5 ova/g) and coliforms (10 CFU/g) in the samples. Sludge samples subjected to a combination of co-composting and vermicomposting resulted in the elimination of mercury and a significant reduction in cadmium concentration from 6.30 mg/kg to 1.12 mg/kg. No helminth ova were observed in the samples after further drying. However, both treated and untreated sludge samples had low nutrient content. The study highlights the need for raising public awareness and educating farmers on the potential risks associated with the use of raw sludge for agriculture.

INTRODUCTIONThe management of sludge produced from wastewater treatment facilities is one of the most expensive challenges faced by the wastewater industry that engineers and regulators are trying to solve (Metcalf and Eddy 2003;

Sinha et al. 2009). The production of massive quantities of sludge has led to the development of various disposal methods including the use of landfills, incinerators, and land application. Among these, the use of sludge for land application is proposed to be the most resourceful and economical alternative method for disposal. Sludge is a potential source of nutrients that can be used as a soil

179

conditioner or fertilizer (Stadelmann et al. 2001; Metcalf and Eddy 2003). Several studies have shown that sludge contains compounds that provide a rich source of organic matter, nitrogen, phosphorus, potassium, and other plant nutrients that may be of agricultural value (EC 2001; Usman et al. 2012). The high organic matter in sludge can improve the physical characteristics of soils such as its structure, water retention, and porosity (Hossain et al. 2017). Hence, it is environmentally friendly and also an economically efficient method of solid waste disposal.

However, the land application of sludge has been restricted due to the reported presence of toxic contaminants, such as heavy metals and microbial pathogens of which pose environmental and human health risks (USEPA 2000). Heavy metals are of major concern due to their non-biodegradable (Duruibe et al. 2007) and accumulative nature that can lead to a decline in soil fertility (Singh and Kalamdhad 2011). For instance, some crops such as tomatoes or lettuce grown in soils with high metal concentrations could directly transfer the heavy metals to humans if the crops are eaten raw (Cappon 1981). Similarly, pathogenic organisms such as helminths and coliforms can persist in soils and in plants for months to years (USEPA 1993, 2003). As such, these contaminants may expose humans to various diseases (e.g. long-term cadmium exposure can cause kidney damage while helminth ova can cause gastrointestinal infections) as they enter the food chain through crop consumption (WHO 1989; Schonning and Stenstrom 2004; Lesmana et al. 2009). These issues highlight the need for a more thorough understanding of the nature of sludge and to search for possible remediation measures to ensure its safe application in agriculture.

Various treatments have been suggested from the United States of America, the United Kingdom, Netherlands, France, etc. (USEPA 1994; EC 2001). However, when considering an economical and sustainable approach, composting technology is the most practical option (Nair et al. 2006). Most studies on sludge treatments were conducted in Pakistan, Ghana, and Malaysia – specifically on co-composting (Bazrafshan et al. 2006; Cofie et al. 2009; Hock et al. 2009). On the other hand, vermicomposting studies have been conducted in Malaysia, Florida, and Australia (VermiCo 2013; Eastman et al. 2001; Sinha et al. 2009) while integrated co-composting and vermicomposting studies have been conducted in Iran (Ndegwa and Thompson 2001; Alidadi et al. 2007). In the Philippines, there has been little to no studies about the characterization and toxic effects of sludge and its treatment. Only the study by Manguiat (1997) was found wherein the author studied sewage sludge from agro-industrial wastewater treatment plants and showed that the sludge passed both US and

German standards on heavy metals. Furthermore, the author used irradiation, specifically gamma radiation, for pathogen control. The study also suggests that biological composting and alkaline stabilization using a combination of heat, high pH (12), and drying can be used as an alternative procedure for treating sludge.

In response to this need, this study aimed to investigate the potential of sludge as a safe and cost-effective alternative soil conditioner. Specifically, the research aimed to: (a) determine the physicochemical characteristics and composition of the sludge obtained from various sources; and (b) describe the effectiveness of various sludge treatment processes for heavy metal, helminth ova, and coliform removal.


Description of SitesThe sludge samples used in the study were obtained from two wastewater treatment facilities on the island of Luzon in the Philippines. One is a septage treatment plant while the other is a sewage treatment plant.

The septage treatment plant is a local government-owned and -controlled corporation with a 30 m3 daily septage capacity. It only provides septage service to its domestic water customers. The desludging service cycle is every five years on a regular basis (BWD 2012). The septage treatment process separates the solid (sludge) and the liquid (effluent) components of the raw septage. The sludge is disposed of as earth fill and soil enhancer whereas the effluent is used in irrigation and re-used within the wastewater treatment facility (e.g. for cleaning, landscaping, etc.) according to the treatment facility staff.

The sewage treatment plant is originally designed as an oxidation ditch system with a treatment volume capacity of 8,600 m3 of wastewater per day. Only 20% of its design capacity, however, was being treated (JICA 1991). It was initially constructed to service domestic areas connected to the sewage system but later included commercial areas that were constructed within the vicinity. The plant also accepts septage from areas that are not connected to the sewage line. Sewage is directly loaded into the oxidation ditches. Part of its full function is the inclusion of sludge thickeners and sludge drying beds to process sludge as a by-product of the wastewater treatment process. The dewatered sludge is dried in drying beds and then sold as an alternative agricultural fertilizer/ soil conditioner (Robinson 2003; CEPMO 2014).

Sample Collection and PreparationThe sample size and the method of analysis were determined using the Publicly-Owned Treatment Works

Guzman et al.: Efficiency of Combined Co-composting, Vermicomposting, and Drying


180

Sludge Sampling and Analysis Guidance Document (USEPA 1989). The sampling of sludge from the two wastewater facilities was conducted once during the wet season. The sludge was dewatered using a dewatering machine and drying beds resulting in 72% and 10% moisture content, respectively. Composite sludge samples were randomly collected from each treatment facility and placed in bags. Seven (7) bags each weighing approximately 10 kg were obtained from the septage treatment plant and one bag weighing approximately 20 kg of sludge samples from the sewage treatment plant was collected. These were immediately transported to the laboratory for processing and analyses.

Physico-Chemical Characterization of the Sludge SamplesThe raw sludge samples (i.e. with 72% and 10% moisture content) were analyzed for various parameters using standard methods. The heavy metals arsenic, cadmium, chromium, and lead were analyzed using USEPA Method 3050B; while mercury was analyzed using the cold-vapor technique. These were quantified using atomic absorption spectrophotometry. In compliance with the PNS for Organic Soil Amendments (PNS/BAFS 183:2016), the carbon-nitrogen ratio was computed, total nitrogen was analyzed using the Kjeldahl method, total phosphorus using USEPA Method 265.3, organic matter and moisture content using the loss-on-ignition method, pH using a portable soil pH meter, temperature using a thermometer, and color consistency and odor using sensory observations.

Microbiological Characterization of the Sludge SamplesThe raw sludge samples were subjected to microbiological characterization based on the total helminths (expressed as ova/g) following the Standard Methods for the Recovery and Enumeration of Helminth Ova in Wastewater, Sludge, Compost, and Urine-diversion Waste in South Africa (Moodley et al. 2008). Total coliform was determined using the pour plate method of the Philippine Coconut Authority laboratory.

Optimization of the Sludge Treatment In order to comply with local standards for the use of sludge in agriculture (PNS/BAFS 183:2016), prior treatment is necessary to remove pathogens and heavy metals that may be present in sludge. Several treatment strategies were employed to determine which of these would produce the desired results. Each treatment was prepared in triplicates along with a control set-up of pure sludge. The sludge samples with 10% moisture content were hydrated to raise the moisture level to 60% using

distilled water (here referred to as “reconstituted” sludge or RS). The other sludge samples with 72% moisture were used as is in the succeeding stage.

Co-composting. For the RS, co-composting (CC) experiments were performed in plastic bins containing a 2:1:1 ratio of vegetable scraps, RS, and soil – with a final weight of 2 kg. Holes were drilled at the bottom of the bins to allow adequate drainage. The interior of the bins was lined with aluminum foil so as to reach an internal temperature of 35–65 °C. The internal temperature was checked daily using a thermometer. The daily moisture content of the co-composting materials was monitored and maintained at 60%. The manual turning of co-composting materials was done once a week to aerate and homogenize the mixture. The sludge samples with 72% moisture were mixed with vegetable scraps in a 1:2 sludge-to-vegetable-scraps ratio, with a total weight of 3 kg. The mixtures were composted for 30 d, similar to the procedure followed with the RS.

Vermicomposting. For the RS, vermicomposting (VC) experiments were performed using a set-up similar to that of CC as described previously, with a final weight of 2 kg before the addition of 50 g of Eudrilus eugeniae (African Nightcrawler) earthworms to each bin. To prevent the earthworms from escaping, a plastic net was placed over each bin. At the end of the treatment, earthworms were sieved to separate them from the vermicompost. After weighing the worms, the worms were subjected to heavy metal analysis.

For the sludge with a 72% moisture level, the final weight was 3 kg. Vermicomposting was performed following the same procedures done to the RS with the addition of 60 g African Nightcrawler earthworms to each bin instead of 50. The difference in the number of worms used was based on the weight of the composting materials. Vermicastings were manually harvested after 30 d.

Combined co-composting and vermicomposting. For the RS, combined co-composting-vermicomposting (CV) experiments were performed by first co-composting the sludge with the vegetable scraps as described previously for 28 d, followed by vermicomposting as described previously. For the sludge with 72% moisture, vermicomposting of sludge was mixed with market waste on a 1:2 sludge-to-scraps ratio. The mixtures were composted for 20 d in bins with holes and mixed for proper aeration. After 20 d of co-composting, 60 g of earthworms were added and the vermicasts were manually harvested after 10 d.

Combined co-composting and vermicomposting with subsequent oven-drying. For the sludge with 72% moisture, co-composting and vermicomposting



181

Table 2. Summary of the characteristics of sludge subjected to various treatment strategies.

Parameter

Control Co-composting(CC)

Vermi-composting(VC)

Co-composting and vermi-composting (CV) PNS-BAFS

standards (2016)Sewage treatment

plantSeptage

treatment plantSewage

treatment plantSeptage

treatment plant

Sewage treatment plant

Septage treatment plant

Sewage treatment

plant

Septage treatment

plant

Color Brown Black Brown Black Brown Black Brown Black Brown to black

Odor No foul odor No foul odor No foul odor No foul odor No foul odor No foul odor No foul odor No foul odor

No foul odor

Consistency Not friable Slimy Friable Friable Friable Friable Friable Friable Friable

Temperature 30 °C 28 °C 37 °C – 30 °C 41 °C – 28 °C 28 °C – 25 °C 28 °C 33 °C – 29 °C 40 °C –28 °C –

pH 6.07 ± 0.12 6.61 ± 0.04 6.80 ± 0.07 6.91 ± 0.13 7.11 ± 0.10 6.66 ± 0.4 7.02 ± 0.01 6.80 ± 0.11 Not regulated

Moisture content (100%) 7.93 ± 0.29 22.3 ± 21.6 8.85 ± 0.19 20.6 ± 12 4.20 ± 0.14 27.9 ± 0.01 4.21 ± 0.37 26.7 ± 6 10–35

Organic matter (%) 24.70 ± 0.79a 47.2 ± 10 31.50 ± 1.69a 13.3 ± 3.8 17.68 ± 0.57b 23.8 ± 0.03 17.13 ± 0.68b 13.9 ± 1.6 > 20

Total N-P2O5-K2O

1.29 ± 0.15a 0.04 ± 0.01 1.12 ± 0.13a 0.016 ± 0.004 0.51 ± 0.007b 0.025 ± 0.001 0.49 ± 0.05b 0.0001 ± 0.00

2.5 – <5%0.057 ± 0.02ac 5.2E06 ± 0.0 0.086 ± 0.043cd 0.000047 ± 4.2E06 0.0056 ± 0.003b 0.000057 ±

7.8E050.017 ± 0.005ab

0.000028 ± 3.9E05

0.354 ± 0.007b 0.001 ± 0.6 0.356 ± 0.004b 0.003 ± 0.004 0.37 ± 0.003a 0.003 ± 0.01 0.36 ± 0.009b Not detected

Total coliform < 10 ± 0 < 10 ± 0 4.3E3 ± 2.5E02 < 10 ± 0 1.5E04 ± 5.0E03 < 10 ± 0 < 10 ± 0 < 10 ± 0 < 10cfu/g

Total helminths Not detected 13 ± 3 1 ± 0 18 ± 5 1 ± 0 19 ± 4 Not detected 26 ± 16 Not regulated

Each value represents the mean ± SD (n = 3). Results sharing the same letter are not significantly different at p < 0.05. C – control (untreated soil); N – nitrogen; P – phosphorus; K – potassium; Sewage treatment plant – sludge samples with 10% moisture content Septage treatment plant – sludge samples with 72% moisture content

Table 1. Summary of initial sludge characteristics obtained from the two wastewater treatment facilities.

Physico-chemical and microbiological properties

Sewage treatment plant Septage treatment plant PNS-BAFS standards (2016)

Color Brown Black Brown to black

Odor No foul odor No foul odor No foul odor

Consistency Not friable Friable Friable

pH at 30 °C 6.84 ± 0.06 6.61 ± 0.04 Not regulated

Moisture content 10.62 ± 0.04 72.3 ± 0.01 10–35%

Organic matter 25.52 ± 0.60 83.5 ± 0.10 > 20%

Total N-P2O5-K2O 1.28 ± 0.08 (N)0.04 ± 3.01E–03 (P2O5)

0.35 ± 0.03 (K2O)

0.04 ± 0.01 (N)5.2 ± 0.01E06 (P2O5)

0.001 ± 0.6 (K2O)

2.5 – < 5%

Arsenic 1.18 ± 0.17 0.118 ± 0.10 20 (mg/kg)

Cadmium 6.30 ± 0.48 0.0163 ± 0.40 5 (mg/kg)

Chromium 20.19 ± 1.09 Not analyzed 150 (mg/kg)

Lead 23.37 ± 2.16 0.0204 ± 0.01 50 (mg/kg)

Mercury 4.02 ± 0.17 Not analyzed 2 (mg/kg)

Total coliform < 10 ± 0 < 10 ± 0 < 10 (cfu/g)

Total helminths 1 ± 0 5 ± 1 Not regulated (ova/g)

Each value represents the mean ± SD (n = 3). Sewage treatment plant – sludge samples with 10% moisture content Septage treatment plant – sludge samples with 72% moisture content PNS/BAFS 183:2016 – Philippine National Standard / Bureau of Agriculture and Fisheries Standards for Organic Soil Amendments



182

were performed as previously described for 30 d. The vermicasts were then subjected to further oven-drying at 57 °C for 1 h (USEPA 1994).

At the end of each treatment, physicochemical and microbiological analyses were conducted to compare the sludge characteristics before and after the treatments, as well as the effectivity of each treatment in removing the helminths and heavy metals. Furthermore, after the experiments, the earthworms that were used in the treatments were disposed of following the guidelines described in the Department of Environment and Natural Resources (DENR) Administrative Order 36, Series of 2004 (DAO 2004-36) for disposing of hazardous waste.

RESULTS AND DISCUSSIONS

Physico-chemical and Microbiological Characterization of the SludgeThe initial visual examination of the sludge collected from the two wastewater treatment facilities during the wet season showed that both samples met the required color and odor (Table 1). However, the sludge sample from the sewage treatment plant was not friable while the moisture content of the sludge sample from the septage treatment plant was too high (72.3 ± 0.01% moisture). Hence, both sludge samples failed to meet the requirement for consistency and moisture content, respectively.

The results of the initial chemical analyses, expressed as mean + standard deviation (SD), showed that the average concentrations of the organic matter (sewage treatment plant, 25.52 ± 0.60%; septage treatment plant, 83.5 ± 0.10%) of both samples were within the required level set by the PNS for Organic Soil Amendments (i.e. > 20%). However, the nitrogen and potassium level of both sludge samples failed to meet the established limit for soil conditioner (2.5 – <5%), while the total concentration of Hg (4.02 ± 0.17 mg/kg and Cd (6.30 ± 0.48 mg/kg) from the sewage treatment plant exceeded the prescribed standard. Microbiological results showed that the total coliform count of both samples met the standard, but helminth ova were detected in both sludge samples. The results of this study highlight the need for additional treatment to remove heavy metals, and helminths.

After the 8-wk treatment process, changes in the physicochemical properties of the initial samples were observed. The end product of each treatment was observed to be much darker in color, odor-free and visually homogenous as compared to their initial condition. The resulting pH values (6.02–7.11) were found to be slightly acidic to neutral. This could be attributed to the

production of organic acids from microbial metabolism as well as the production of humic and fulvic acids during decomposition (Ndegwa and Thompson 2000; Dominguez and Edwards 2004; Suthar and Singh 2008).

For the sludge samples from the sewage treatment plant,

Figure 1. Average heavy metal concentration (mg/kg), expressed as mean ± SD, in the treated sludge samples after treatment. Each value represents the mean ± SD (n = 3). Results sharing the same letter are not significantly different at p < 0.05. C – control (untreated soil); CC – co-composting; VC – vermicomposting; CV – co-composting + vermicomposting.

a 27.5% significant increase (p < 0.05) in organic matter content was found after CC (31.50 ± 1.69%) whereas a 28.4% and a 17.13% significant loss (p < 0.05) was detected after VC (17.68 ± 0.57%) and CV (17.13 ± 0.68%), respectively (Table 2). Results for the sludge samples from the septage treatment plant showed a significant loss (p < 0.05) in organic matter content of 67.5%, 41.8%, and 66% after CC (13.3 ± 3.8%), VC (23.8 ± 0.03%), and CV (13.9% ± 1.6), respectively (Table 2). Partial mineralization and humification are chemical changes that occur because of organic matter bio-oxidation (Fornes et al. 2012). Greater reduction in organic matter content indicates better degradation and mineralization, giving more stable products (Ndegwa and Thompson 2001). Hence, VC and CV treatments can be used to efficiently stabilize sludge.

The total N content of the final product of each treatment found in the sludge samples from the sewage treatment plant ranged from 0.49–1.12%. Significant loss (p < 0.05) was found after VC (0.51 ± 0.01%) and CV (0.49 ± 0.05%) but no significant reduction was measured after CC (1.12 ± 0.13%). On the other hand, the total N found in the sludge samples from the septage treatment plant ranged from 0.01–0.02%. Significant loss was found after CC (0.016 ± 0.004%), VC (0.025 ± 0.001%), and CV (0.0001 ± 0.004%). The significant reduction in total N content after treatment could be due to mineral N leaching because of frequent application of water (Cogger et al.



183

2000) during vermicomposting to maintain optimum moisture content. A similar result was reported by Fornes et al. (2012) in which heavy irrigation resulted to greater leaching of the total N content and most macronutrients in vermicomposting and integrated composting systems as compared to composting alone.

An increase in available P concentration was detected after CC (0.086 ± 0.043%) of sludge from the sewage treatment plant and after VC (0.000057 ± 7.8E05%) of sludge from the septage treatment plant. While decreases in the available P were noticed after VC (0.0056 ± 0.003%) and CV (0.017 ± 0.005%) from the samples taken from the sewage treatment plant and after CC (0.000047 ± 4.2E06%) and CV (0.000028 ± 3.9E05%) from the samples taken from the septage treatment plant. The result of the VC treatment showed a significant decrease (p < 0.05) in P concentration. This may be linked to phosphate leaching brought about by frequent watering during VC. The decrease in available P was consistent with the findings of Ndegwa and Thompson (2001).

The total K concentration following all composting treatments generally increased as compared to the control, with the VC treatment showing a statistically significant increase (p < 0.05) in K concentration. This could be attributed to the addition of soil, which was found to contain 0.37% K. The increase in the total K concentration seen after the VC (sewage treatment plant, 0.37 ± 0.003%; septage treatment plant, 0.003 ± 0.01%) treatment might have been due to the acid production by the microorganisms present in the gut of earthworms which could solubilize the insoluble potassium (Kaviraj and Sharma 2003). Similar findings were reported by Tripathi and Bhardwaj (2004), Gupta and Garg (2008), and Suthar and Singh (2008).

Results of the Heavy Metal AnalysesHeavy metal contents of the final products showed an increase of 41.6% in Cd concentration after CC (mean ± SD = 5.39 ± 1.22 mg/kg) but this result did not differ significantly from the Control (mean ± SD = 3.81 ± 0.74 mg/kg). A 70.5% significant reduction (p < 0.05) on Cd concentration was found after CV (mean ± SD = 1.12 ± 0.02 mg/kg) (see Figure 1). The results for the total

Hg level showed a decrease after CC. The combined composting also demonstrated a significant (p < 0.05) and 100% removal of Hg among all the treatments (see Figure 1).

Several studies have shown that earthworms can accumulate heavy metals from organic waste through their skin and gut (Suthar and Singh 2008; Sinha et al. 2009; Suthar et al. 2014; Soobhany et al. 2015). However, only a few studies focused on the use of Eudrilus eugeniae as vermiworms for the removal of toxic heavy metals from specific waste (Graft 1982; Iwai et al. 2013; Soobhany et al. 2015) due to their narrow temperature tolerance compared to Eisenia fetida (Reinecke et al. 1992). In this study, a large increase in the concentration of heavy metals was measured from the body tissues of Eudrilus eugeniae relative to the initial concentration. For Cd, the concentration increased by 1.1 mg/kg after VC and 2.21 mg/kg after CV. For Hg, the concentration increased by 1.29 mg/kg after VC and 2.92 mg/kg after CV (Table 3). Specifically, a significant increase (p < 0.05) in heavy metal concentrations was measured from the earthworm tissues after the CV treatment. Earthworms subjected to the CV treatment accumulated 40.84% and 86.98% of the total Cd and Hg in their system, respectively. This could be due to the palatability of waste mixtures in the combined treatment as evidenced by the remarkable increase in earthworm biomass. It is worth mentioning that the CV system was the treatment that also exhibited a significant decrease in organic matter decomposition.

Table 3. Average change in accumulated metal concentration in earthworm tissues, expressed as mean ± SD (n = 3).

Heavy metal (mg/kg) Initial concentration After VC % accumulated metals

after VC After CV % accumulated metals after VC

Cd 0.79 ± 0.08a 1.89 ± 0.87ab 20.40 3.0 ± 0.24b 40.84

Hg 0.02 ± 0a 1.31 ± 0.80ab 39.31 2.94 ± 0.80b 86.98

Results sharing the same letter are not significantly different at p < 0.5.VC – vermicomposting; CV – co-composting + vermicomposting



184

Percent accumulation of metal in the earthworm tissues was calculated from the equation:

Percent removal efficiency is calculated from the equation:

Therefore, the combined treatment of co-composting and vermicomposting (i.e. CV) can be used to effectively remove heavy metals – specifically Cd and Hg from sludge – as shown in Table 4. These results are consistent with findings from other studies using different metals, organic wastes, and earthworm species (Suthar and Singh 2008; Sinha et al. 2009; Iwai et al. 2013).

Table 4. Percent removal efficiency of co-composting, vermicomposting, and combined co-composting and vermicomposting.

Heavy metal % removal efficiency

CC VC CV

Cd 0.20 29.82 79.07

Hg 75.00 89.14 100.00

In terms of the helminth ova and coliform levels found in the sludge from the sewage treatment plant, total coliform after the CV treatment was below 10 CFU/g and no helminth eggs were detected. This is in contrast with the other treatments (CC and VC) in which helminth eggs were detected and a large increase in the concentration of total coliforms was measured after CC and VC. Moreover, sludge from septage treatment plant after CC, VC, and CC exhibited below 10 CFU/g total coliform count but the total helminth count exceeded the allowable limits (Table 2) – hence the need to further dry the treated sludge to significantly reduce the helminth ova present (USEPA 1994).

For CC, the temperature only reached 37 °C on its 5th day and decreased subsequently while the VC temperature stayed below 30 °C. According to USEPA (2003), composting can significantly reduce helminth ova and coliforms provided that the temperature of sewage sludge is raised to 40 °C or higher for 5 d and the compost piles exceeds 55 °C within 4 h during the 5-d period. This suggests that the temperature in both treatments (CC and VC) was not high enough to kill the helminth ova and coliforms; thus, raising the possibility of coliform multiplication or re-activation.

However, a more detailed temperature specification for the composting system was noted by Jenkins (1999). Based on his study, complete pathogen reduction in a composting system can be assured if the temperature is attained at 62 °C for 1 h, 50 °C for 1 d, 46 °C for 1 wk, or 43 °C for 1 mo. Unlike the other treatments, the temperature for the CV treatment reached 50 °C on its first day, thereby resulting in low coliform count and elimination of helminth ova. These findings indicate that the CV treatment is more effective in terms of reducing heavy metal concentrations (Cd and Hg), helminth ova and coliforms than a single composting treatment in sludge with 10% moisture. While a combination of co-composting or vermicomposting followed by further drying is the best treatment to completely eliminate helminth ova in sludge with 72% moisture.

CONCLUSIONThere is a need to characterize sludge prior to agricultural application. This study shows that the raw sludge samples from either type of treatment plant did not completely meet the minimum requirements set in PNS/BAFS 183:2016. High concentrations of Cd and Hg in sludge from the sewage treatment plant were found to exceed the maximum allowable level. Moreover, the presence of helminth ova in sludge from the septage treatment plant indicates that raw sludge cannot be recommended for land application as a soil conditioner. Thus, there is a need for further treatment prior to its use in land application to prevent risks to both the environment and human health. The findings of this study showed that a combination of co-composting and vermicomposting was significantly more effective than a single composting treatment in terms of eliminating Hg and significantly reducing the concentration of Cd. For the 72% moisture sludge, further drying was necessary after the combined treatment in order to eliminate the pathogens (coliforms and helminths). Moreover, caution needs to be observed with regards to the proper disposal of earthworms used in the treatments as stated in DAO 2004-36 to avoid contamination.

% Metal Accumulation = (Metal in ET After Treatment – Metal in ET Before Treatment) x 100 Metal in the Initial Wastes (1)

% Removal Efficiency = (Metal in the Initial Wastes – Metal in the Compost) x 100 Metal in the Initial Wastes

(2)



185

Further and more intensive work is necessary to investigate the beneficial effect/s of vermicompost to plants. As this study produced vermicomposts with low levels of nutrients, composting of sludge with nutrient-rich organic waste materials is highly suggested to maximize its full potential for soil amendment. It is also suggested to investigate seasonal comparison and the tolerance of earthworms to heavy metals such as those found in this study to determine their maximum efficiency in bioremediation.

ACKNOWLEDGMENTSThe authors would like to acknowledge the Asian Development Bank, the Accelerated Science and Technology Human Resource Development Program of the Philippine Department of Science and Technology, and the Ateneo de Manila University for their generous funding support. Special thanks to Stella Tansengco Shapero and Kevin Paolo Bartolome for their valuable assistance in the conduct of this research.

REFERENCESALIDADI H, PARVARESH AR, SHAHMANSOURI

MR, POURMOGHADAS H, NAJAFPOOR AA. 2007. Combined compost and vermicomposting process in the treatment and bioconversion of sludge. Pak J Biol Sci 10(21): 3944–3947.

BAZRAFSHAN E, ZAZOULI MA, BAZRAFSHAN J, BANPEI AM. 2006. Co-composting of Dewatered Sewage Sludge with Sawdust. Pak J Biol Sci 9(8): 1580–1583.

[BWD] Baliuag Water District. 2012. Septage Collec-tion Scheme. Retrieved on 22 Apr 2017 from http://baliwagwd.com/contents/septage-collection-scheme/

CAPPON CJ. 1981. Mercury and Selenium Content and Chemical Form in Vegetable Crops Grown on Sludge-Amended Soil. Arch. Environ. Contam Toxicol 10: 673–689.

[CEPMO] Baguio City Environment and Parks Manage-ment Office. 2014. A Situationer: Wastewater Manage-ment in the City of Baguio. Retrieved on 24 Apr 2017 from http://www.unescap.org/sites/default/files/10-Waste%20water%20management% 20in%20the%20city%20of%20Baguio,%2 0Philippines%20.pdf

COFIE O, KONĔ D, ROTHENBERGER S, MOSER D, ZUBRUEGG C. 2009. Co-composting of faecal sludge and organic solid waste for agriculture: process dynamics. Water Res 43(18): 4665–4675.

COGGER CG, SULLIVAN DM, HENRY CL, DORSEY KP. 2000. Biosolids Management Guidelines for Wash-ington State. Washington State Department of Ecology.

[DENR] Department of Environment and Natural Re-sources. 2004. DENR Administrative Order 36, Series of 2004: Revised Procedural Manual on Hazardous Waste Management.

DOMINGUEZ J, EDWARDS CA. 2004. Vermicompost-ing organic wastes: a review. In: Soil Zoology for Sustainable Development in the 21st Century. Hanna SH, Mikhail WZ ed. Cairo: Safwat H Shakir Hanna. p. 369–395.

DURUIBE JO, OGWUEGBU MO, EGWURUGWU JN. 2007. Heavy Metal Pollution and Human Biotoxic Ef-fects. Int J Phys Sci 2(5): 112–118.

EASTMAN BR, KANE PN, EDWARDS CA, TRYTEK L, GUNADI B, STERMER AL, MOBLEY JR. 2001. The effectiveness of vermiculture in human pathogen reduction for USEPA biosolids stabilization. Compost Sci Util 9(1): 38–41.

[EC] European Commission. 2001. Disposal and recycling routes for sewage sludge Part 3. Retrieved on 22 Jul 2016 from http://ec.europa.eu/environment/archives/waste/sludge/pdf/ sludge _ disposal3.pdf

FORNES F, MENDOZA-HERNANDES D, GARCIA DE LA FUENTE R, ABAD M, BELDA RM. 2012. Composting versus Vermicomposting: A Comparative Study of Organic Matter Evolution Through Straight and Combined Processes. Bioresour Technol 118: 296–305.

GRAFT O. 1982. Vergleich der Regenswurmaten Eise-nia foetida und Eudrilus eugeniae hinsichlich ihrer Eignung zur Proteinwinnung aus Abfallstoffen. Pedo-biologia 23: 277–282.

GUPTA R, GARG VK. 2008. Stabilization of Primary Sewage Sludge During Vermicomposting. J Hazard Mater 153(3): 1023–1030.

HOCK LS, BAHARUDDIN AS, AHMAD MN, SHAH UKM, RAHMAN NAA, ABD-AZIZ S, HASSAN MA, SHIRAI Y. 2009. Physicochemical Changes in Wind-row Co-Composting Process of Oil Palm Mesocarp Fiber and Palm Oil Mill Effluent Anaerobic Sludge. Aust J Basic & Appl Sci 3(3): 2809–2816.

HOSSAIN MZ, FRAGSTEIN PV, NIEMSDORFF PV, HESS J. 2017. Effect of Different Organic Wastes on Soil Properties and Plant Growth and Yield: A Re-view. Sci Agric Bohem 48(4): 224–237.

IWAI CB, TA-OUN M, CHUASAYATEE T, BOON-YOTHA P. 2013. Management of Municipal Sewage



186

Sludge by Vermicomposting Technology: Converting a Waste into a Bio Fertilizer for Agriculture. Int J Environ Rural Dev 4(1): 169–174.

JENKINS J. 1999. The Humanure Handbook. A Guide to Composting Human Manure, Second Edition. Grove City, PA: Joseph Jenkins, Inc. Retrieved from https://humanurehandbook.com/downloads/H2.pdf

[JICA] Japan International Cooperation Agency. 1991. Basic Design Study Report on the Baguio Septage System Rehabilitation Project in the Republic of the Philippines [Supplementary Study].

KAVIRAJ S, SHARMA S. 2003. Municipal Solid Waste Management Through Vermicomposting Employing Exotic and Local Species of Earthworms. Bioresour Technol 90(2): 169–173.

LESMANA SO, FEBRIANA N, SOETAREDJO FE, SUNARSO J, ISMADJI S. 2009. Studies on Potential Applications of Biomass for the Separation of Heavy Metals from Water and Wastewater. Biochem Eng J 44: 19–41.

MANGUIAT IJ. 1997. Sewage sludge: turning an en-vironmental pollutant into an agricultural resource. Retrieved on 28 Oct 2019 from http://agris.fao.org/agris-search/search.do?recordID=PH1998010123

METCALF AND EDDY, INC. 2003. Wastewater Engi-neering: Treatment, Disposal and Reuse, 4th ed. New York: McGraw-Hill Publishing Company Ltd.

MOODLEY P, ARCHER C, HAWKSWORTH D, LEI-BACH L. 2008. Standard Methods for the Recovery and Enumeration of Helminth Ova in Wastewater, Sludge, Compost and Urine-Diversion Waste in South Africa. Pretoria, South Africa: Water Research Commission.

NAIR J, SEKIOZIOC V, ANDA M. 2006. Effect of Pre-Composting on Vermicomposting of Kitchen Waste. Bioresource Technology 97: 2091–2095.

NDEGWA PM, THOMPSON SA. 2000. Effects of C-to-N Ratio on Vermicomposting of Biosolids. Bioresour Technol 75: 7–12.

NDEGWA PM, THOMPSON SA. 2001. Integrating Composting and Vermicomposting in the Treatment and Bioconversion of Biosolids. Bioresour Technol 76: 107–112.

[PNS] Philippine National Standards, [BAFS] Bureau of Agriculture and Fisheries Standards. 2016. PNS/BAFS 183:2016: Organic Soil Amendments.

REINECKE AJ, VILJOEN SA, SAAYMAN RJ. 1992. The Suitability of Eudrilus eugeniae, Perionyx excava-tus and Eisenia fetida (Oligochaeta) for Vermicompost-ing in Southern Africa in Terms of Their Temperature

Requirement. Soil Biol Biochem 24(12): 1295–1307.

ROBINSON A. 2003. Urban sewarage and sanitation: lesson learned from case studies in the Philippines. World Bank Water and Sanitation Program – East Asia and the Pacific.

SCHONNING C, STENSTROM TA. 2004. Guidelines for the Safe Use of Urine and Faeces in Ecological Sanitation. Sweden: Ecosanres, SEI.

SINGH J, KALAMDHAD AS. 2011. Effects of Heavy Metals on Soil, Plants, Human Health and Aquatic Life. Int J Res Chem Environ 1(2): 15–21.

SINHA RK, HEART S, BHARAMBE G, BHARAMB-HATT A. 2009. Vermistabilization of Sewage Sludge (BIosolids) by Earthworms: Converting a Potential Biohazard Destined for Landfill Disposal into a Pathogen-free, Nutritive and Safe Biofertilizer for Farms. Waste Manag Res 28(10): 872–881.

SOOBHANY N, MOHEE R, GARG VK. 2015. Compara-tive Assessment of Heavy Metal Content During the Composting and Vermicomposting of Municipal Solid Waste Employing Eudrilus eugeniae. Waste Manage 39: 130–145.

STADELMANN FX, KULLING D, HERTER U. 2001. Sewage Sludge: Fertilizer or Waste? Retrieved on 24 Apr 2017 from http://library.eawag.ch/EAWAGPublications/openaccess/ Without_EAWAG_number/eawagnews/www_en53/en53e_printer/en53e_stadelm_p.pdf

SUTHAR S, SINGH S. 2008. Feasibility of Vermicom-posting in Biostabilization of Sludge from a Distillery Industry. Sci Total Environ 394: 237–243.

SUTHAR S, SAJWAN P, KUMAR K. 2014. Vermireme-diation of Heavy Metals in Wastewater Sludge from Paper and Pulp Industry Using Earthworm Eisenia Fetida. Ecotox Enviro Safe 109: 177–184.

TRIPATHI G, BHARDWAJ P. 2004. Decompostion of Kitchen Waste Amended with Cow Manure Using an Epigeic Species (Eisenia fetida) and an Anecic Species (Lampito mauritti). Bioresour Technol 92(2): 215–218.

[USEPA] United States Environmental Protection Agency. 1989. Publicly Owned Water Treatment Works Sludge Sampling and Analysis Guidance Document. Retrieved from http://greencriminology.org/glossary/publicly-owned-water-treatment-works-powts-us/ on 08 May 2019.

[USEPA] United States Environmental Protection Agency. 1993. Methods for Microbiological Analyses of Sew-age Sludge. Washington, DC: Dynamic Corporation.



187

[USEPA] United States Environmental Protection Agency. 1994. Biosolids Management Handbook EPA Region VIII. Retrieved on 24 Apr 2017 from https://www.epa.gov/sites/production/ files/documents/handbook1.pdf

[USEPA] United States Environmental Protection Agency. 2000. Biosolids Technology Fact Sheet. Land Appli-cation of Biosolids. Washington, DC: Environmental Protection Agency.

[USEPA] United States Environmental Protection Agency. 2003. Control of Pathogens and Vector Attraction in Sewage Sludge. Cincinnati, OH: US Environmental Protection Agency.

USMAN K, KHAN S, GHULAM S, KHEN MU, KHAN N. 2012. Sewage Sludge: An Important Biological Resource for Sustainable Agriculture and Its Environ-mental Implications. Am J Plant Sci 3: 1708–1721.

VERMICO. 2013. Journal Publishes Study of Worms Reducing Human Pathogens in Biosolids. Retrieved on 17 Jan 2017 from http://www.vermico.com/journal-publishes-study-of-worms-reducing-humanpathogens-in-biosolids/

[WHO] World Health Organization. 1989. Health Guide-lines for the Use of Wastewater in Agriculture and Aquaculture [Technical Report Series 778]. Geneva.



188

Keywords: groundwater age, isotope hydrology, radioactive isotopes, recharge rate, stable isotopes

*Corresponding Author: [email protected]

Isotopic Data for Inferring Groundwater Dynamics in Cagayan De Oro City, Philippines

Philippine Journal of Science149 (1): 189-199, March 2020ISSN 0031 - 7683Date Received: 11 Jun 2019

Charles Darwin T. Racadio1,2*, Soledad S. Castañeda1, Flerida A. Cariño2,3, and Norman DS. Mendoza1

1Philippine Nuclear Research Institute, Department of Science and Technology, Commonwealth Ave., Diliman 1101 Quezon City, Philippines

2Institute of Environmental Science and Meteorology, College of Science, University of the Philippines, Diliman 1101 Quezon City, Philippines

3Institute of Chemistry, College of Science, University of the Philippines, Diliman 1101 Quezon City, Philippines

A groundwater study was conducted in Cagayan de Oro City (CDO) located in the north-central part of Mindanao, the Philippines using isotopic techniques. The study identifies the recharge sources of the groundwater in the city and estimates the groundwater age and groundwater recharge rate. Monthly integrated samplings of rainfall were conducted in three locations of varying altitudes from October 2012 to March 2015. Groundwater samples from production wells and shallow wells were also collected within the same period at least twice, during the dry and the wet seasons. The samples were analyzed for their stable isotopic compositions and groundwater in selected deep wells was dated using tritium-helium dating. A meteoric water line of δD = (8.26 ± 0.21) δ18O + (11.56 ± 1.88) was calculated using the precipitation-weighted reduced major axis (PWRMA) method. Isotopic compositions of groundwater show that local precipitation recharges the shallow aquifer. An interaction exists between the shallow and deep aquifer possibly due to the absence of a well-defined multilayer aquifer based on available lithological profiles. Coastal shallow wells appear to be recharged from 35–80 meters above sea level (masl), while deep wells appear to be recharged from 200–300 masl. A recharge rate of 380 mm/yr was estimated, which is more than 20% of the average annual rainfall and twice the estimated recharge using the precipitation-based method.

INTRODUCTION

Background and ObjectivesGroundwater provides the bulk of the water supply in the Philippines, especially in areas outside Metro Manila (WB 2004). Over the last decades, the population in the Philippines has increased dramatically leading to an increase in water demand. This has led to increased

exploitation of groundwater, with it being easily available and requiring little treatment. As early as 1991, the Japan International Cooperation Agency (JICA) has listed nine major cities in the Philippines (i.e. Metro Manila, Davao City, Baguio City, Angeles City, Bacolod City, Iloilo City, Zamboanga City, Cebu City, and CDO) as water-critical areas (JICA 1998). In CDO, the demand for water has increased over the last decades due to population expansion and the growth of industrial activities. From 3.3 million m3/mo in 2000, the groundwater extraction

189

rate increased to 4.7 million m3/mo in 2011 (Palanca-Tan 2011). It is suspected that the city is overexploiting its aquifers as manifested in the drop in the groundwater level of some production wells in recent years. Of the 250 listed operating wells within the city in the Local Water Utilities Administration database, only 67 have permits in the records of the National Water Resources Board (NWRB), the country’s water regulating body. Sustainable use of water resources with sufficient understanding of groundwater recharge and flow is, therefore, essential for future management of the city’s groundwater resources.

CDO is located in the north-central part of Mindanao in the southern side of the Philippines (Figure 1). The city supports a population of 675,950 according to the last census of the Philippine Statistics Authority (PSA) in 2015 (PSA 2017). The annual average temperature is about 26.7 °C. Based on the Modified Coronas Climate Classification, CDO falls under the Type III climate. This type is characterized by not very pronounced seasons, which are relatively dry from December–April and wet during the rest of the year. It has an average monthly rainfall of 140 mm based on a 35-yr data of the Philippine Atmospheric, Geophysical and Astronomical Services Administration from 1977–2012 at Lumbia Airport gauging station (NWRB 2016). Annual average rainfall based on the same data is 1699.1 ± 309.1 mm. A radial network of 12 river systems originating from Mt. Kitanglad, an extinct volcanic center in Bukidnon, flows into the city towards the bay. The major river system is the CDO River, which has an average inter-annual discharge rate of 54 m3 s–1.

In order to develop an effective water management policy, it is necessary to understand the behavior and dynamics of groundwater. Published studies conducted on the groundwater systems of CDO have focused mostly on the threats and factors affecting the city’s groundwater resources (Palanca-Tan and Bautista 2003; Palanca-Tan 2011). Most of the data on the city’s water resources are also not available due to issues in data ownership. In this case, the use of isotopic techniques provides a relatively inexpensive tool for the characterization of flow systems where there is a scarcity of hydrogeological information (Yasuhara et al. 1993). Although the use of isotopes of water (oxygen-18, 18O; deuterium, 2H; tritium, 3H) has long been proven to provide substantial information on the behavior of aquifers (Clark and Fritz 1997; IAEA and UNESCO 2003; Charideh and Zakhem 2010), studies in the Philippines using the technique are relatively sparse. Most of these studies have applied the technique in geothermal environs (Alvis-Isidro et al. 1993; Gerardo et al. 1993; Gerardo-Abaya 2005). Some studies in the country using the technique on groundwater assessment remain unpublished as detailed in the International Atomic Energy Agency (IAEA) database (IAEA 2008).

Specifically, the study objectives are: (1) to identify the sources of recharge, and (2) to estimate the recharge rate of groundwater in CDO. In this study, we provide a comprehensive analysis of groundwater resources in CDO through the use of several isotopic tracers (oxygen and hydrogen). This multi-tracer approach has been used to determine the origin and residence time of groundwater (Bouchaou et al. 2009; Cartwright and Morgenstern 2012, 2016; Davisson et al. 1999; Gusyev et al. 2019; Harrington et al. 2002; McGuire et al. 2002; Palmer et al. 2007; Plummer et al. 2001; Ravikumar and Somashekar 2011; Subyani 2004). Tritium-helium dating was also applied for the first time in the Philippines through this study. Tritium-helium dating has been applied in several hydrological studies to determine groundwater age (Hinkle 2009; McMahon et al. 2011; Shapiro et al. 1998; Solomon and Sudicky 1991; Solomon et al. 1995; Szabo et al. 1996). While classical techniques are still important in hydrological studies, we demonstrate in this study that isotope data provide important supplemental information for studying groundwater recharge processes and residence time.

Geological and Hydrogeological SettingsCDO is located in the central coast of northern Mindanao and lies between the latitude 8°14′ to 8°31′5” N and longitude 124°27′ to 124°49′ E (Figure 1). It has an area of about 57,000 ha and is geographically nestled between the central coastline of Macajalar Bay to the north and the plateaus and mountains of the provinces of Bukidnon and Lanao del Norte to the south. It is bounded by the municipalities of Opol on the west and Tagoloan to the east.

The city is located in a small alluvial plain formed by the CDO River. This alluvial plain is extended with a long and narrow belt along the Macajalar Bay. Another alluvial plain with a fairly wide area is located along the Iponan River. The lowland is relatively flat with an elevation of not more than 10 masl. The hinterland surrounding the plain is occupied by a hill-top terrace of vast area with elevation ranging from 100–150 m. The top of the terrace is made of relatively flat land surrounded by steep cliffs. In the eastern part of the city, a vast mountainous area – dissected by many small streams – towers from 400–500 masl. This mountainous area is made of limestone. Hilly terrace and slope areas with high elevation are continuously distributed near the seashore. The city is drained by 12 river systems which originate from the slopes of Mt. Kitanglad in Bukidnon Province. These rivers form collectively five major watersheds – from west to east, these are the Iponan, Cagayan Oro, Culambog, Cugman-Agusan, and Alae watersheds (NWRB 2016). About 28% of the city has a slope of 0–8%. This is concentrated on the narrow coastal plains, the alluvial plains of Cagayan and Iponan Rivers, and in the upland terraces. The remaining land has a slope of more than 8%.


Racadio et al.: Isotopic Data to Infer Groundwater Dynamics

190

Geologically, the area is formed by six different formations. The quaternary alluvium (QAl) is deposited in the narrow areas along the CDO and Iponan Rivers and extending to the east of the city. This formation is made of silt, sand and gravel, and boulders. The Cagayan Terrace Gravel (CT) of Pleistocene to Holocene and Quaternary eras is distributed on both sides of the CDO River, forming hilly areas with elevation ranging from 100–150 m between the top of the terrace and riverbank. This formation has numerous boulders in the terrace basement consisting of gravel, sandstone, and shale. The Bukidnon Formation (BF) of Late Pliocene to Pleistocene eras is mainly distributed in the eastern part of the city and in the hinterland of the CT formation. It forms a hilly area with a high elevation of 250–300 m as well as a fairly flat top area. The formation consists mainly of tuffaceous sandstone, agglomerate, pebbly sandstone, and conglomerate. The Indahag Limestone (IL) of early to late Pliocene era is mainly distributed in the southern part of the city for a wide area. It is made up of corral rubbles and sandy limestone. The Opol Formation (OF) of the Miocene era is mainly distributed on the west side of the Iponan River beyond the city area. It is composed of conglomerate, pebbly sandstone, agglomerate and tuff. The Schist of the Mesozoic era is distributed in the mountainous area to the east of the city. This area has an elevation of about 500 m and is fairly dissected.

The majority of the city’s deep wells are distributed in the narrow alluvial sediment area along the CDO and Iponan Rivers, and a small plain east of the city in Barangay Bugo. Of these well fields, the most productive are those distributed in the alluvial area of CDO River. Based on available lithological logs, the formation layers along this well field consist of sand and gravel with interspersed layers of clay. Known to locals as Macasandig formation, the aquifer yields an average of 7,100 m3/d per well (JICA 1998). Lithological logs also show almost similar formation layers on the well fields distributed in the alluvial plains of Iponan River and on the small plain east of the city in Barangay Bugo. Other minor well fields are distributed along the coastal stretch of Gusa-Cugman-Agusan, and along the coastal area near the mouth of the CDO River. Groundwater flow models made by the NWRB show that recharge of groundwater in the city occurs through the IL and CT formations. The groundwater flows northward towards the bay (NWRB 2016).


Sample CollectionPrecipitation in the study area was sampled on a monthly basis between October 2012 to February 2015 in Barangay

Figure 1. Site map showing boundaries of CDO, the adjacent towns and the sampling sites.



191

Bugo, CDO (labeled S3 in Figure 1) and in two locations in Bukidnon Province (labeled S1 and S2, respectively). The rain collector was fabricated using an enclosed 5-gal water bottle fitted with a big funnel having a diameter of 30.5 cm. Thermal insulation was wrapped around the collector. Paraffin oil was also added inside the bottle to minimize evaporation. At the end of each sampling, the total volume of rainwater collected was recorded. Samples for stable isotopes analysis were collected using double-capped 20-ml polyethylene (PE) bottles that are fully filled with no headspace and immediately sealed tight to prevent exchange with the atmosphere. These were stored in a temperature and humidity-controlled room away from direct sunlight until ready for analysis.

Groundwater samplings were conducted at least twice, once during the dry season and another during the wet season. Samples were collected from production wells and shallow wells (e.g. handpump) within CDO and Bukidnon Province. Sampling locations are shown in Figure 1. In total, 55 samples were collected from production wells and 14 samples from shallow wells.

Groundwater samples were collected from production wells through a faucet directly connected to the well’s discharge pipe or through a bypass discharge pipe. For shallow wells (e.g. handpumps), they are pumped manually by hand. Water was allowed to continuously flow while the conductivity, pH, and temperature were measured. Samples for stable isotopes analysis were collected using double-capped 20-ml PE bottles that are fully filled with no headspace and immediately sealed tight to prevent exchange with the atmosphere. The samples were stored in a temperature- and humidity-controlled room away from direct sunlight until ready for analysis.

A separate sampling was conducted in August 2013 for the collection of samples for tritium-helium dating samples from seven selected production wells – three from Bukidnon Province and four from CDO. Samples for helium and other noble gases analyses were collected by flushing the water sample through annealed copper tubes, which are then pinched-off at both ends to prevent atmospheric contamination. The volume of the water samples inside the copper tubes is around 40 ml (Visser et al. 2014). Samples for tritium analysis were separately collected in double-capped 1-L PE bottles that are fully filled with no headspace and immediately sealed tight to prevent exchange with the atmosphere.

Stable Isotopes AnalysisLaser spectrometry of hydrogen and oxygen isotopes in water was performed at the International Atomic Energy Agency Isotope Hydrology Laboratory (IAEA-IHL), Vienna, Austria following the method described by

Wassenaar et al. (2014). Isotopic compositions are given as δ values, the relative deviations with respect to the standard value, the Vienna Standard Mean Ocean Water. Signifying R as the abundance ratio of the isotopic species, i.e. 2H/1H or 18O/16O, δ is defined by:

(1)

where δ is reported in ‰ (per mille, equivalent to 10–3). For deuterium-hydrogen (2H/1H), δD is used; for 18O/16O, the notation is δ18O. The reported maximum uncertainty of each measurement is ± 0.2‰ for δ18O, and ± 1.5‰ for δD.

The regression analysis of δD and δ18O in precipitation were determined by the method of PWRMA (Crawford et al. 2014), which was called as local meteoric water line (LMWL, δD = a*δ18O + b; Craig 1961). The parameters a and b were determined by the following equations:

(2)

(3)

where

(4)

and

(5)

The wi, yi, and xi refers to the fraction of monthly precipitation amount “pi ” to the total precipitation, δD and δ18O, respectively.

Tritium and Noble Gas AnalysesConcentrations of helium and other noble gases (e.g. Ne, Ar, Kr, and Xe) were determined at the IAEA-IHL, Vienna, Austria as described in Visser et al. (2014). Technical specifications of the instrument used are described in Matsumoto et al. (2017). Gases dissolved in the water samples inside the copper tube were extracted using a vacuum manifold. The extracted gases were inlet to a cryogenic and chemical getter cleanup system followed by analysis using a sector-field mass spectrometer for helium isotopes and a quadrupole mass spectrometer for 20Ne,



192

(9)

40Ar, and 84Kr isotopes (Stolp et al. 2010). Tritium samples were analyzed at the HYDRSYS LABOR Analytical Laboratory Ltd., Budapest, Hungary. Enrichment factors of 15–16 were obtained, with a limit of detection of 0.5 TU at a 95% confidence interval.

Tritium-Helium DatingThe 3H/3H age is defined as:

(6)

where λ is the 3H decay constant (= 0.05626 y–1; Lucas and Unterweger 2000), [3He]tritiogenic is the amount of 3He derived from tritium decay, and [3H] is the amount of tritium. The [3He]tritiogenic is estimated by separating it from the total [3He] dissolved in the groundwater, which has a number of sources noted as follows:

(7)

where [3Heeq] = 3He in solubility equilibrium with the atmosphere, 3Heexc = 3He due to excess air, and 3Heterr = terrigenic 3He (nucleogenic 3He together with mantle 3He). In Equation 7, only 3Hetot and 3Heeq are determined through measurements. 3Heexc and 3Heterr (if present in the sample) were determined by using measurements of 4He and Ne (3Heterr). The total 4He concentration measured in a groundwater sample can be written as:

(8)

where 4Hetot = total measured 4He concentration in the groundwater sample, 4Heeq = 4He concentration in solubility equilibrium with the atmosphere, 4Heexc = 4He concentration caused by excess air, and 4Heterr = terrigenic 4He concentration (radiogenic 4He plus mantle 4He). If no terrigenic helium is contained in the groundwater sample, 3Hetrit can be calculated by:

(10)

where is the recharge rate (mm yr–1), θ is the porosity of the aquifer, H is the thickness of the aquifer, z is the depth of screen below the water table (calculated by subtracting the static water level from depth to top of well opening), and τz is the groundwater age from tritium-helium dating.

RESULTS AND DISCUSSION

Isotopic Compositions (δD and δ18O) of PrecipitationThe 18O and deuterium composition of the precipitation samples collected from the three stations are summarized in Table 1. The δ18O range from –14.81 to –4.59 per mil while the δD values ranged from –109.2 to –25.97 per mil.

Mostly, the isotopic signatures of collected rainwater samples during post-monsoon (December–April) season are relatively more enriched in heavier isotopes (i.e. δD and δ18O are more positive) than the isotopic signatures of collected rainwater samples during monsoon (May–November) season. The behavior may be attributed to the shift in the prevailing winds systems in the Philippines between the Southwest monsoon, which usually prevails from July to September, and the Northeast monsoon, which prevails from November to February (Estoque et al. 2000). Gerardo-Abaya (2005) reported similar behavior in isotope data in rainwater samples collected from different parts of the Philippines. The phenomenon also reflects the effect of rain amount (i.e. “amount effect”) to the isotopic composition of the rainwater. A higher volume of rainwater was collected during the monsoon season than during the post-monsoon season in this study.

where Rtot = measured 3He/4He ratio of the sample, Ratm = 3He/4He ratio of air (1.384*10-6), α = solubility isotope effect (ca. 0.983)), and S = salinity in ‰. The conversion factor for 3He from cm3STPg–1 to TU is 4.021*1014/[(1–S)/1000)] TU/cm3STPg–1.

Based on the calculated groundwater ages, the recharge rate was estimated using the formula (Cartwright et al. 2017):

The LMWLs calculated based on PWRMA method were δD = (8.26 ± 0.21) δ18O + (11.56 ± 1.88), δD = (8.18 ± 0.26) δ18O + (11.10 ± 2.28), and δD = (7.88 ± 0.0.30) δ18O + (7.74 ± 2.71) for stations S3, S2, and S1, respectively. While S3 and S2 show almost similar meteoric water lines, S1’s meteoric water line significantly differs from the other two, possibly due to its different topographic features of the neighboring catchment and its higher elevation.



193

The isotopic signatures of the rain samples were also used to calculate the isotopic lapse rate based on the altitude effect. Months during which a high amount of rainwater was collected, specifically July– September, were selected for this purpose (Figure 2). A lapse rate of δ18O = –0.0034 h – 7.14 was calculated, where h is the altitude or recharge. This indicates a decrease of –0.34 ‰ in δ18O per 100-m altitude. The value is more than the global average isotopic lapse rate of –0.28 ‰ δ18O/100m (Poage and Chamberlain 2001). Reported lapse rates for δ18O and δ2H worldwide lie within the ranges of −0.1 to −0.6 ‰ × 100 m−1 and −0.5 to −4 ‰ × 100 m−1, respectively (Windhorst et al. 2013). The large lapse rate may be related to the orographic steepness of the study area.

Bukidnon Province showed high variation, with samples from higher altitude having lower isotopic values than samples having a lower altitude. In contrast, groundwater samples from CDO showed minimal variation except for groundwater samples from the coastal well field, which have the highest isotopic values. Some samples from the same well field have isotopic values that deviated from the LMWL, possibly indicating mixing with intruding seawater as these samples also exhibit higher electric conductivity (> 2000 μS/cm) than samples from the nearest well field (i.e. Central CDO).

Shallow groundwater samples have isotopic values ranging from –9.03 to –7.45 ‰ for δ18O and –59.50 to –49.63 ‰ for δD. The samples show distinct isotopic values, with samples at lower altitudes having higher isotopic values than the samples in higher altitudes. This strongly indicates recharge from local precipitation.

Although in general, groundwater samples from CDO have higher isotopic values than groundwater samples from Bukidnon, it is observed that isotopic values of some

Table 1. Stable isotopic composition of monthly precipitation samples collected from the three rain sampling stations.

Station name Location Altitude

(masl)a n Total rainfall amount (mm)

Isotope Range Weighted meansDeuterium

excessδ18O, ‰ δ2H, ‰ δ18O, ‰

δ2H, ‰Min. Max. Min. Max.

S1 8.16096, 125.11923 658 13 1700.60 –11.67 –2.28 –86.56 –21.12 –8.63 –61.10 7.96

S2 8.35625, 124.863417 570 13 1144.78 –11.9 –4.59 –85.71 –25.97 –8.48 –58.20 9.61

S3 8.50885, 124.7711 18 16 1315.26 –14.81 –5.13 –109.2 –28.2 –8.46 –58.36 9.36

Figure 2. Estimation of the altitude of recharge of groundwater using the isotopic lapse rate calculated based on δ18O of rainwater samples collected for the months of June – September.

Isotopic Composition of Groundwater SamplesIsotopic values of deep groundwater samples from Bukidnon Province varied from –9.19 to –7.87 ‰ for δ18O and –63.63 to –53.24 ‰ for δD. On the other hand, isotopic values of deep groundwater samples from CDO ranged from –8.4 to –6.58 ‰ for δ18O and –54.66 to –46.7 ‰ for δD (Figure 3). Groundwater samples from

Figure 3. The relationship between δD and δ18O of groundwater samples plotted on the LMWLs.



194

groundwater samples from Bugo and Gusa-Cugman-Agusan well fields overlap with the isotopic values of groundwater samples from the downstream section of Bukidnon Province. These groundwater samples are in the towns of Libona and Manolo Fortich, which are adjacent to the southern border of CDO. Based on models from previous groundwater assessment studies of the NWRB, the groundwater in the eastern section of CDO is recharged within these parts through the BF (NWRB 2016).

Isotopic signatures of some shallow well samples also overlap with isotopic signatures of deep groundwater samples, suggesting an interconnection between the shallow and deeper aquifer. Based on the general lithological profiles in the study area, it seems that a well-defined multilayer does not exist. It may be possible that the aquifer in the area is a large aquifer interspersed with clay layers in some area but this needs to be further verified.

The altitudes of the recharge of the groundwater were estimated using the previously calculated isotopic lapse rate (Figure 2). Average δ18O of the collected groundwater samples were substituted to the isotopic lapse rate to estimate the altitude of the recharge zones (Paternoster et al. 2008). Based on the available data, shallow wells near the coast are recharged from 35–80 masl, while inland shallow wells (more than 3 km from the coast) range from 100–200 masl. The groundwater in the production wells (i.e. deep groundwater) is estimated to be recharged at 200–400 masl. This is the same altitude as the CT, IL, and BF previously identified by groundwater flow models of NWRB as the recharge zones in CDO (Figure 4).

Figure 4. Front-view 3D images of the study area looking from Macajalar Bay. The arrows point to the inferred recharge zones based on the estimated altitudes of recharge.

Tabl

e 2.

Sum

mar

y of

mea

sure

d no

ble

gas a

nd tr

itium

con

cent

ratio

ns, a

nd c

alcu

late

d 3 H

/3 He

ages

. Num

bers

in re

d ar

e th

e lo

wer

lim

it of

cal

cula

ted

ages

due

to lo

w c

once

ntra

tion

of tr

itium

(i.e

. be

low

LLD

; LLD

= 0

.5 T

U).

The

Chi^

2 m

easu

res t

he fi

tnes

s of t

he m

odel

to th

e m

easu

red

valu

es –

the

low

er th

e nu

mbe

r, th

e be

tter i

s the

fit.

Sam

plin

g si

teH

e (c

m3 S

TP/

g)E

rror

Ne

(cm

3 ST

P/g)

Err

orA

r (c

m3 S

TP/

g)E

rror

Kr

(cm

3 ST

P/g)

Err

orX

e (c

m3 S

TP/

gE

rror

3 He/

4 He

Err

or

Kih

are

Wel

l, M

anol

o Fo

rtich

, B

ukid

non

4.75

E–08

6.65

E–10

1.90

E–07

2.00

E–09

3.11

E–04

5.67

E–06

6.63

E–08

1.96

E–09

1.00

E–08

1.66

E–09

1.49

E–06

2.70

E–08

BSW

M W

ell,

Mal

ayba

lay

City

, B

ukid

non

5.32

E–08

7.37

E–10

2.12

E–07

2.27

E–09

3.06

E–04

5.51

E–06

6.27

E––0

81.

78E–

098.

32E–

093.

76E–

101.

38E–

061.

84E–

08

Kib

anga

n Pu

mp,

Im

pasu

g-on

g,

Buk

idno

n1.

83E–

072.

53E–

098.

42E–

079.

39E–

096.

86E–

041.

21E–

051.

13E–

073.

34E–

091.

41E–

087.

01E–

101.

40E–

061.

56E–

08

CO

WD

PS

#14,

C

DO

4.96

E–08

6.76

E–10

1.93

E–07

2.07

E–09

2.96

E–04

5.30

E–06

6.12

E–08

1.93

E–09

8.38

E–09

6.18

E–10

1.68

E–06

3.96

E–08

CO

WD

PS

#23,

C

DO

5.71

E–08

8.31

E–10

1.94

E–07

2.16

E–09

2.97

E–04

5.24

E–06

6.57

E–08

1.96

E–09

9.50

E–09

4.64

E–10

2.35

E–06

2.67

E–08

CO

WD

PS#

21,

CD

O4.

84E–

087.

17E–

101.

92E–

072.

13E–

092.

85E–

045.

00E–

066.

16E–

081.

84E–

098.

84E–

094.

31E–

101.

40E–

061.

63E–

08

CO

WD

PS#

10,

CD

O6.

99E–

081.

00E–

092.

79E–

073.

06E–

093.

80E–

046.

69E–

067.

84E–

082.

36E–

091.

02E–

085.

05E–

101.

43E–

061.

56E–

08

Groundwater AgesGroundwater ages determined by the 3H/3He dating technique were calculated with the assumption that there is no contribution of terrigenic helium from the mantle (Table 2). This is valid to some degree since Mt. Kitanglad is already extinct and the study area is mostly underlain by QAI. Low levels of tritium (< 0.5 TU; i.e. the LLD of



195

the analytical laboratory) prevented reliable age dating for four groundwater samples. The 3H/3He ages of the groundwaters ranged from 8.5 to > 71.6 yr. There is no discernible trend in 3H/3He age with depth, which may be related to the large distances between the wells sampled in this investigation (Figure 5).

The consistency of the estimated 3H/3He ages was checked against the rain tritium history. This is done by summing the observed 3H and tritiogenic 3He for each sample, in tritium units, and plotting these against the recharge year (observation year minus the 3H/3He age). There are no existing continuous rain tritium records for the Philippines except in 1961, 1964–1966, and 2000 onwards. The nearest Global Network of Precipitation stations with rain tritium records since the 1960s are Hong Kong, China; Jakarta, Indonesia; and Tokyo, Japan (IAEA and WMO 2018). The recently reconstructed rain tritium history of Tokyo, Japan from 1950–2018 was used by scaling it to 0.18 for the anthropogenic period (1963–1979) and 0.53 for the remaining years (Gusyev et al. 2016, 2019; Chatterjee et al. 2019). The comparison shown in Figure 6 indicates a reasonable agreement between estimated tritium in the recharge rain and actual rain measurements except for some samples, which have low tritium concentrations (< 0.5 TU). One of these samples has a recharge year prior to the 1950s. Groundwater with recharge year prior to 1950 is not representative of actual recharge year but is an artifact of mixing. The other samples are affected by mixing with younger groundwater (Price et al. 2003). This is notably exhibited by the well with depths of below 40 m but with groundwater age of at least 17.4 yr. This supports the earlier assumption of the absence of a well-defined multilayer aquifer or the presence of a large interconnected aquifer where young and old components mix.

Recharge RateThe recharge rate was estimated from the groundwater sample with measured tritium (e.g. COWD Deepwell No. 14), which is located in the Macasandig Aquifer. Water in this well has an apparent age of 38.6 ± 5.8 yr. The average porosity of 0.30 ± 0.15 was assumed for the aquifer due to its high contents of mixed sand and gravel. The well has its topmost screen at 62.03 meters below ground level (mbgl) and has a static water level of 17.52 mbgl. Although the NWRB estimated the shallow aquifer in the study area to be 10–40 m thick, there is no estimation for the thickness of the deeper aquifer. However, at certain places, the aquifer has been drilled up to 255 m. Assuming this number to be the maximum thickness of the aquifer, a recharge rate of 380 ± 52 mm/yr was calculated, which corresponds to at least 19% of the annual average rainfall in the city. This is more than twice the recharge rate estimated by NWRB using their

Figure 5. The 3H/3He age (years) of groundwater sites with sampling depth.

Figure 6. The sum of 3H and tritiogenic 3He in groundwater samples as a function of the recharge year. The blue line represents the tritium measurements measured in Tokyo, Japan scaled to Philippine Latitude at 0.18 for the anthropogenic years (1963–1979) and 0.53 for the remaining years. The yellow, violet, and green lines represent the tritium concentrations measured in Manila in 1961, and 1964–1966; Diliman, Quezon City in 2000–2009; and Philippine Nuclear Research Institute, Quezon City in 2017 onwards. The recharge year of groundwater sites prior to 1950 is not representative of the actual recharge year but is an artifact of mixing. Some groundwater samples, specifically those with < LLD tritium concentrations, do not match tritium in rain, possibly due to dispersion and mixing processes.



196

precipitation-based method, which assumes that only 10% of the annual average rainfall (i.e. 156 mm/yr) is recharged into the aquifer (NWRB 2016).

The high recharge rate may be due to the presence of fractured zones, particularly in the IL, which is identified by the NWRB as a recharge area of groundwater in the study area. Average recharge estimates from hydrogeological studies of tropical limestone aquifers typically lie within the range of 20–30% of annual rainfall (Vacher and Quinn 1997).

CONCLUSIONGroundwater in CDO is mainly of meteoric origin. Shallow groundwater is recharged by local precipitation, as manifested in its strong isotopic gradient relative to the distance from the coast and altitude. The deep groundwater is recharged on the downhill side of Bukidnon near its boundary with CDO. Isotopic signatures of the shallow and deep groundwater show an interaction between the shallow and deep aquifer, possibly due to the absence of a well-defined multilayer aquifer based on available lithological profiles. It is possible that the aquifer in the study area is a large aquifer that is interspersed only with clay layers in some areas.

Based on available data and the calculated isotopic lapse rate, shallow wells near the coast appear to be recharged from 35–80 masl. Deep groundwater is recharged from 200–300 masl. This is the same altitude as the IL, CT, and BF, which were identified by the NWRB as the possible recharge zones of groundwater in CDO.

A recharge rate of 380 mm/yr was calculated for CDO, which corresponds to more than 20% of the annual average rainfall in the city. This is twice the recharge rate calculated by the NWRB using their precipitation-based method, which assumes that 10% of the annual average rainfall is recharged to the aquifers. The high recharge rate is attributed to the presence of fractured zones in the recharge area, which acts as preferential flow paths.

ACKNOWLEDGMENTSFunding for this project was provided through the IAEA Water Availability Enhancement Project, with additional assistance from the Department of Science and Technology’s Accelerated Science and Technology Human Resource Development Program. We thank the NWRB, the Mines and Geosciences Bureau Region X, and the CDO Water District for the assistance provided during the sampling.

REFERENCESAL CHARIDEH AR, ABOU ZAKHEM B. 2010. Distri-

bution of tritium and stable isotopes in precipitation in Syria. Hydrolog Sci J 55(5): 832–843.

ALVIS-ISIDRO RR, SOLAÑA RR, D’AMORE F, NUTI S, GONFIANTINI R. 1993. Hydrology of the Greater Tongonan Geothermal System, Philippines, as deduced from geochemical and isotopic data. Geothermics 22(5/6): 435–449.

BOUCHAOU L, MICHELOT JL, QURTOBI M, ZINE N, GAYE CB, AGGARWAL PK, MARAH H, ZER-OUALI A, TALEB H, VENGOSH A. 2009. Origin and residence time of groundwater in the Tadla basin (Morocco) using multiple isotopic and geochemical tools. J Hydrol 379: 323–338. DOI: 10.1016/j.jhy-drol.2009.10.019.

CARTWRIGHT I, MORGENSTERN U. 2012. Constrain-ing groundwater recharge and the rate of geochemical processes using tritium and major ion geochemistry: Ovens catchment, southeast Australia. J Hydrol 475: 137–149.

CARTWRIGHT I, MORGENSTERN U. 2016. Using tritium to document the mean transit time and sources of water contributing to a chain-of-ponds river system: implications for resource protection. Ap Geochem 75: 9–19.

CARTWRIGHT I, CENDON D, CURRELL M, MER-EDITH K. 2017. A review of radioactive isotopes and other residence time tracers in understanding groundwater recharge: possibilities, challenges, and limitations. J Hydrol 555: 797–811.

CHATTERJEE S, GUSYEV MA, SINHA UK, MO-HOKAR HV, DASH A. 2019. Understanding water circulation with tritium tracer in the Tural-Rajwadi geothermal area, India. Ap Geochem 109: 104373. DOI: 10.1016/j.apgeochem.2019.104373.

CLARK I, FRITZ P. 1997. Environmental Isotopes in Hy-drogeology. Boca Raton, FL: Lewis Publishers. 328p.

CRAIG H. 1961. Isotopic variations in meteoric waters. Science (133): 1702–1703.

CRAWFORD J, HUGHES CE, LYKOUDIS SP. 2014. Alternative least squares methods for determining the meteoric water line, demonstrated using GNIP data. J Hydrol 519: 2331–2340. DOI: 10.1016/j.jhy-drol.2014.10.033.

DAVISSON ML, SMITH DK, KENNEALLY J, ROSE TP. 1999. Isotope hydrology of southern Nevada groundwater: stable isotopes and radiocarbon. Water Resour Res 35: 279–94.



197

Documents

Genetic Characterization of Philippine Saba Germplasm ...philjournalsci.dost.gov.ph/.../vol149no1/...saba_.pdf · collection of Saba accessions maintained at the National Plant Genetic