Ferdinnd Glsim , Mohmmd H. Goli ,*, Yong Sng Kim , Clncy Iyekr

a College of Natural and Applied Sciences, University of Guam, Mangilao, Guam, 96913, USA

b Water and Environmental Research Institute of the Western Pacific, University of Guam, Mangilao, Guam, 96913, USA

Keywords:Nitrate Composted organic wastes Inorganic fertilizer Calcareous soil

ABSTRACT The purpose of this study was to quantify the release of nitrate into vadose zone as well as the nitrogenholding capacity of compost applied on calcareous soils of northern Guam amended with both inorganic fertilizer and composted organic waste,to examine potentially adverse effect of these nutrient materials to groundwater quality. Three different nitrogen levels each of the composted organic waste and the inorganic fertilizer were applied to corn (maize) (farmland) study plots for three consecutive seasons,two dry and one rainy season.During each season,soil organic matter(SOM,w/w%)content and carbonnitrogen ratio (C/N, w/w%), were determined for analysis of the nitrogen-holding capacity of these calcareous soils.Nitrate levels in soil pore water were also determined for study of potential groundwater contamination. For three seasons, compost plots showed higher SOM% contents and lower C/N% than fertilizer plots. That is, compost-treated soils showed higher SOM content and adsorbed more nitrogen under normalized soil mass than did fertilizer-treated soils. Nitrate levels in pore water were generally higher on compost plots during early stages of corn but were generally higher on fertilizer plots during active leaf, tasseling, and maturity stages. During the rainy season, nitrate leachate seemed to increase,but this trend could not be confirmed because rainfall amounts were not measured. Overall, composted organic waste proved to be a good amendment for soil productivity and agricultural sustainability while reducing nitrate leachate from northern Guam farmland.

1. Introduction

The increasing use of inorganic fertilizers and disposal of wastes from animal farms are among the major contributors to elevated nitrate levels in groundwater resources(Shamrukh et al.,2001).In northern Guam, the majority of farmland has poor soils with low productivity and is not ideal for subsistence farming.The soils’poor chemical and physical properties, especially the low soil organic matter, may contribute to the excessive application of commercial inorganic fertilizers,which can negatively affect the integrity of the vadose zone and aquifer (Golabi et al., 2004). The Northern Guam Lens Aquifer (NGLA) (Jenson et al., 2006) consists of limestone bedrock overlying volcanic rocks.Estimated hydraulic conductivity in this region ranges from 1005 to 6004 m/day(Ayers&Clayshulte,1984). Guam is a tropical island with a mean temperature of 26°C ± 2°C (Karolle,1991) and annual rainfall of approximately 2540 mm, with distinct dry and rainy seasons (Lander, 1994).Currently, over 80% of fresh water is being pumped from the northern Guam aquifer by about 120 production wells(Bendixson,2013).Improper use of inorganic fertilizers may therefore adversely affect groundwater quality in the region.

Composted organic wastes have been shown to be usable as a major nutrient source that minimizes nitrogen leaching because of its high organic content,which increases abiotic sorption.Levanon et al. (1993) reported that higher organic-matter content in soils increased abiotic sorption as well as biotic degradation processes of synthetic chemicals,resulting in lower leaching of these chemicals.Another advantage is that composting recycles organic wastes rather relegating them to landfill or incineration, particularly advantageous to a small island such as Guam. Nevertheless, little is known about the environmental impact of using composted organic wastes above the NGLA.

The purpose of the study reported here is therefore to quantify the release of nitrate into the vadose zone when corn(maize) was cultivated in soils of northern Guam amended with both inorganic fertilizer and composted organic waste. It was assumed that application each of inorganic fertilizer and composted organic waste having same amounts of nitrogen equivalent to soils affects different nitrogen holding capacities and nitrate release from the soils to vadose zone.For analysis of nitrate levels from pore water,suction-cup lysimeters were installed at two different depths in a total of 28 study plots. The experimental design called for three treatment levels each of composted organic waste and inorganic fertilizer,each replicated four times.The remaining four study plots received no treatments and served as controls.The experiment was conducted over three corn cropping seasons. During corn cultivation, soil pore-water samples were collected weekly. Nitrate leachate data were grouped for analysis according to three corn growth stages. For examination of effects on nitrogen adsorption,soil organic matter and carbon-nitrogen content in soil were also measured.

2. Materials and methods

2.1. Study site

Northern Guam Geology: Guam is an island 51 km long and 15 km wide located in the western Pacific Ocean, the largest and southernmost of the Mariana Islands(Fig.1).It consists of volcanic mountain uplands in the south and limestone uplands underlain by volcanic rock in the north. The limestone surface forms a plateau 60-180 m above sea level and occupying about half of the island.The basement volcanic rock beneath the NGLA is the late Eoceneearly Oligocene Alutom Formation (Tracey et al., 1964). Surface soil of northern Guam consists mainly of weathered limestone.Over 80% of exposed limestone on Guam originates from Mariana limestone. Barrigada limestone is mostly exposed on northern Guam(Schlanger,1971).Study plots were located at the University of Guam farmland research station in the village of Yigo(Fig.1).The soil underlying the study site is formed in sediment over both Mariana and Barrigada limestone (Young,1988).

Fig.1. Limestone distribution in northern Guam (Image from hydroguam.net).

2.2. Experimental design

Twenty-eight study plots, each 7.01 m by 6.09 m, were established(Fig.2).Plots were separated from one another by 5 m in one dimension and by 12 m in the other.Two main water lines fed eight water drip lines per study plot, which were set up approximately 91 cm apart. Timers were set to turn the water on twice a day for 2 h. Randomized experimental design of twenty-eight study plots applied to University of Guam farmland is shown in Fig.2.Corn was planted in the plots (Table 1). Two types of corn seeds were purchased from University of Hawaii.Hybrid sweet#8 was planted in 2014,and hybrid super-sweet#10 was used in 2015 and 2016.The result of corn yield in this experiment is being reported in a separate manuscript (Golabi et al., 2020). As corn ears neared the harvest stage, irrigation water was reduced to twice a day for 1 h.Adjustments were also made during storms, and dry or rainy seasons,to control erosion and guard against overwatering.Each study plot was equipped with two ceramic suction cups in vacuum lysimeters (pore size: 1.3 μm), installed at depths of 61 cm (upper sampler) and 122 cm (lower sampler) for collecting pore-water samples during corn cultivation (Fig. 3). Negative pressure of 30-50 centibars were ideal for this study.

Composted organic waste (C) or inorganic fertilizer (F) was applied to each plot, except for the four plots that served as controls. The application rates were set up as three different nitrogen treatment levels with four replications that granted names like Block I, II, III and IV (Fig. 2). To each fertilizer-treated plot(7.01 m×6.09 m),one of three amounts of gaviota brand fertilizer,all-purpose commercial inorganic regular fertilizer, 6.6 kg (F30),12.7 kg(F60),and 19.3 kg(F90),was applied.The fertilizer used in this study was balanced with 16-16-16 (N, P, K).16-16-16 means that the fertilizer contains 16 w/w% of nitrogen,16 w/w% of phosphorus and 16 w/w% of potassium. To each compost-treated plot(7.01 m×6.09 m)one of three amounts of compost,287.4 kg(C30),574.8 kg (C60), and 862.2 kg (C90), was applied (Table 1). The amounts of each compost of C30,C60 and C90 applied were chosen to be equivalent to nitrogen content of F30, F60 and F90, respectively.Before the application of organic compost to each plot,each of 30, 60, and 90 tons of organic compost per acre (= 4046.7 m2)were prepared at another location of University of Guam farmland(Fig.4).Thus,numbers in symbols of C30,C60 and C90 were simply adapted from number tons of each organic compost to easily compare with data collected from each plot(Table 1).The compost consisted of restaurant food, paper wastes, and woodchips discharged from Anderson Air Force military base in Guam and hog and chicken manures from local poultry and hog farms.

Experiments were conducted from April through July in 2014(Year 1,wet season),from November through March in 2015(Year 2,dry season),and from September through December in 2016(Year 3,wet season).Monthly precipitation data from the National Weather Service is shown in Table 2. The rain gauge was located at Guam International Airport, about 10 km from the University of Guam’s farmland research station, so rainfall data might not be correct for the farmland under study. Average monthly precipitation during cultivation was 103 mm for Year 1 and 115 mm for Year 2.In Year 1,13 days before harvest, before the final the fnial pore-water sampling,a heavy tropical storm hit Guam.Rainfall was exceptional for a dry season, so these precipitation data were not included in the average value for the Year 1 season. Average precipitation of Year 3 was 302 mm,about 3 times that of Year 1 and the Year 2.During the experiments,composted organic waste was applied one week before planting, whereas inorganic fertilizer was applied two weeks prior and six weeks after when the corn start to tassel.Composted organic waste was applied only during Years 1,and Year 3,whereas inorganic fertilizer was applied during all three years.

Fig. 2. Randomized experimental design of twenty-eight study plots applied to University of Guam farmland.

Table 1 Application rates of composted organic waste and inorganic fertilizer.

Fig. 3. Vacuum lysimeters, (a) installation, (b) photo installed in study plots, and (c) photo of pore-water sampling.

2.3. Chemical analysis

Soil organic matter(SOM) w/w% content was determined from soil samples by the application of Walkley-Black method (Sparks et al.,1996). It was measured once for each plot before harvest.

Carbon-to-nitrogen (C/N, w/w%) ratio (C% over N% in top soil)was also measured, by a carbon and nitrogen analyzer instrument(FlashEA 1112 SERIES, Theremo Electronic Corporation). For determination of C/N, top-soil samples were collected from each plot before planting and again after harvest. Before analysis, soil samples were air-dried and sieved with a 2-mm mesh screen. The samples were then milled in a coffee grinder and sieved again with a 0.023 mm mesh screen. Soil pH was measured with an Oakton glass electrode pH meter.For measurement,a ratio of one part soil to one part water was used, but occasional two parts water were used as a result of soil texture(Sparks et al.,1996).In all seasons,pH ranged between 6.7 and 7.4. According to Motavalli and Marler(1998) the optimum pH range for planting corn is 5.5-7.5, so conditions of the study plots were appropriate for corn cultivation.

Nitrate (NO3--N) levels were analyzed from pore-water samples taken by the same lysimeters described above. After collection, the leachate samples were stored in 50-mL capped containers,placed in a cooled insulated box, and transported to the soil lab for further analysis. Pore water sampling was conducted once a week throughout the growing season. Nitrate was analyzed by the flow injection analyzer(QuickChem 8500,Lachat Instruments).Detection limit of the analyzer ranged between 0.11 to 10.00 mg/L of nitrate.Thus,samples were diluted with DI-water prior to analysis if values of the first analysis were out of reliable detection range.If values of nitrate showed below 0.11 mg/L,those values were estimated with a linear interpolation method.Thus,some of average nitrate values in the result section might not be perfectly accurate if the average values included values less than 0.11 mg/L. All raw nitrate data was shown in appendix(Table A4,A5,A6,A7,A8 and A9).

Corn growth stages are usually classified as emergence,nth leaf(1-10), tasseling, silking, and maturity (Çakir, 2004). In our experiment, these stages were grouped into three-vegetation stage I,vegetation stage II,and production stage-to facilitate study of the role of nitrate leachate at different nutrient consumption stages.Vegetation stage I included emergence and production of up to 3 leaves,on assumption that nitrate consumption during these is low. Vegetation stage II included production of 4-10 leaves and tasseling (the stage of greatest growth), and the production stage included silking and maturity (the period of crop production).

Besides nitrate, the pore-water samples were also analyzed for phosphate,in order to evaluate the phosphorus leachate trends,but phosphate was mostly undetectable in the samples,so its impact is not discussed here.

Fig. 4. Preparation of composted organic waste (a) and its application to study plots (b).

Table 2 Monthly precipitation data (from the National Weather Service) during the three experimental seasons.

3. Results and discussions

3.1. Soil organic matter

Fig. 5 shows averaged SOM w/w% content in soils collected before harvest from all study plots for three consecutive seasons.Recall that, this experiment was conducted with four replications.Thus, four of samples (n = 4) were averaged for each treatment,except samples of Year 2 C60 and C90 (n = 3) that one of four samples were lost due to contamination during delivery. Standard deviation(SD)of SOM data ranged between 0.47 and 2.32,and over 75% of average values had SD less than 1.00 (Appendix Table A1).For the three seasons, compost-treated plots had 1.8-4.3% higher SOM content than fertilizer-treated plots. SOM content increased about 0.2-2.1%in the order of C30,C60,and C90 except Year 1 C90.Before the Year 1 harvest,a heavy tropical storm hit Guam(Table 2)and may have affected SOM data for that season.

Although compost was not applied in Year 2 croping season,the trend in SOM content was nearly same as that of Year 1.This result might have arisen in at least ways: 1) Compost might have promoted microbial activities that created another source of organic matter in the soil (Zibilske, 1987). 2) In addition, corn roots and dead leaves produced during the previous season might have contributed to SOM as well.

SOM was higher in all treatments of Year 3 than in either Year 1 or Year 2. Re-application of compost in Year 3 might also have increased SOM. Fertilizer plots showed 0.5-1.3% higher SOM content than the control plots. Although application of inorganic fertilizer is not recommended for soil sustainability, its application might have helped to increase SOM slightly in the calcareous soil of northern Guam.

3.2. Carbon-nitrogen ratio in soil

C/N w/w% measured from top-soil samples taken before planting and after harvest are shown in Fig. 6. High C/N% means that percentage carbon mass is higher than percentage of nitrogen mass in the soils under study. Same as SOM samples, four of samples were averaged for each treatment except several samples that one of samples were lost (Fig. 6). SD values were also reported in Appendix (Table A2.). For all soil samples, C/N% from the compost plots was lower than that from fertilizer plots and control plots.As mentioned above, SOM w/w% in compost-plot soils was higher than that in fertilizer-treated soils (Fig. 5). Evanylo et al. (2008)reported that compost increased soil carbon. Soil organic matter is also known to increase nutrient content and water-holding capacity of soils while lowering their bulk density(Laird et al.,2001).Total nitrogen content of compost-treated soils,normalized by soil mass, would therefore be expected to be higher than that in fertilizer-treated and control soils. C/N% values for the three treatments in each season did not differ significantly.

Control plots showed C/N of about 40% or slightly less. At each sample analysis for each season, C/N% values at fertilizer-treated soils for both before planting and after harvest were higher than controls expect one case, Year 2 after harvest. Inorganic fertilizer therefore slightly increased SOM content, as mentioned above(Fig. 5), but the effect seems very low or negligible.

Fig. 5. Soil organic matter (SOM w/w%) in soils of study plots.

Fig. 6. Carbon-to-nitrogen ratios (C/N w/w%) in soils of study plots.

3.3. Nitrate concentration trends

Fig. 7 shows changes in average nitrate concentrations in samples taken by both upper and lower vacuum lysimeters(Fig.3)from all experimental plots at each growth stage during three seasons.Total periods of cultivation for Year 1, 2 and 3, were 99 days,139 days, and 91 days, respectively. Soil porewater were sampled on a weekly basis at all plots.Total number of sampling for Year 1,2 and 3 were 10-,12- and 10-times during periods of cultivation. Recall that replication was classified as Block I, II, III and IV for each treatment (four replication). In this data analysis, analyzed nitrate values were averaged at each corn growth stage that are; vegetation stage I,vegetation stage II,and production stage.Thus,number of samples that were averaged ranged from 6(n=6)to 20(n=20).Some of samples that were lost due to failure of sampling control prior to analysis did not include in the averaged nitrate values.About 55% of weekly samples had SD values less than 5.00 that showed relatively good correlation among replication samples.However,some weekly replication samples showed SD values up to 100.00 that indicated data correlation is very low. These correlations were more observed at fertilizer treated plots,especially from soil porewater samples collected within 1-2 days after rain events.Thus, we assumed that fertilizer treated plots tend to generate more localized water channel by rainwater that resulted in localized nitrate concentrations in soil porewater. However, these high SD values could not be quantitatively analyzed in this study. Raw data of nitrate concentrations and SD values can be found in in appendix (Table A4, A5,A6, A.7, A8 and A9).

At vegetation stage I, when nitrate consumption was relatively low, nitrate leachates at both upper and lower samplers on all compost plots were nearly same or higher than those at all fertilizer plots. Compost-treated soil contained higher organic matter and nitrogen mass than fertilizer-treated soil (Figs. 5 and 6), so compost-treated soil rather increases nitrate leachate by soil-water partitioning.This trend was evident under all three treatments and in both dry and rainy seasons.

Fig. 7. Nitrate levels in pore water collected from vacuum lysimeters mounted at two depths(61 cm, Upper, and 122 cm, Lower) for all study plots at three growth stages of corn(maize). Vegetation Stage I, emergence and production of up to 3 leaves; Vegetation Stage II, production of 4-10 leaves and tasseling; Production Stage, silking and maturity.

At vegetation stage II and the production stage,for Years 1 and 2,nitrate levels measured at both lysimeters were higher on almost all fertilizer plots than on compost plots. As uptake rates of nutrients by corn increased,nitrogen mass in both compost-treated and fertilizer-treated soils decreased. On the other hand, organic soil attracted nonadsorbed nitrogen more strongly according to soilnitrogen equilibrium, whereas fertilizer-treated soil, which contained low organic matter, kept releasing more nitrogen to pore water. At vegetation stage II and the production stage, the higher nitrate levels in pore water on fertilizer-treated plots might have been caused by low adsorption capability of fertilizer-treated soils.

In Year 1,nitrate levels were higher at F60 than at F90.In year 2,nitrate levels were slightly higher at F90, but the levels did not make much difference.Stoichiometrically,F90 contained 1.5 times as much nitrogen as F60,but in our experiment,no distinguishable difference between the F60 and the F90 plots was observed. If the amounts of inorganic fertilizer applied to F90 plots were excessive,northern Guam calcareous soil might resist excessive nutrients and not release nitrogen to pore water stoichiometrically and proportionally.

Recall that composted organic waste was not applied during Year 2.In our experiment,nitrate levels in pore water at both upper and lower samplers ranged from 0.7 to 21.5 mg/L for Year 1 and from 0.2 to 24.7 mg/L for Year 2.Although nitrogen was consumed by corn during the Year 1, nitrate leachates were not significantly different during Year 2.Several reasons are possible.Compost plots might contain enough nutrients to supply corn for several years.In addition,soil organic matter can increase nutrient-holding capacity and improve soil quality and therefore long-term sustainability(Laird et al., 2001).

In Year 3, corn was planted during the rainy season, and composted organic waste was reapplied.The pattern of nitrate levels in compost and fertilizer plots were similar to those of Years 1 and 2 except during vegetation stage I. SOM% contents of compost plots were higher in Year 3 because of the reapplication(Fig.5),and the higher SOM might have increased mass of nitrogen in soil and therefore released higher amounts of nitrogen into the pore water during vegetation stage I. In Year 3, control plots showed higher leaching of nitrate than in Years 1 and 2. During this rainy season,average precipitation was about three times higher(Table 2)than in Years 1 and 2. We therefore assumed that frequent and heavy rainfall might have elevated levels of nitrate leachate in pore water regardless of depths of sampling, but this point could not be confirmed in our experiment because no rain gauge was available on this farmland.

Recall that, two vacuum lysimeters were installed at depths of 61 cm (upper sampler) and 122 cm (lower sampler) for collecting pore-water samples. At vegetation stage I, nitrate levels at lower pore-water samples were slightly lower than those at almost all upper samples. Main driving force of nitrate transport to vadose zone is concentration gradient between soil and water. Because nitrate consumption at top-soil layer (0-50 cm depth) was relatively low at the vegetation stage I,nitrate should be attracted and adsorbed at each soil layers as pore-water percolates downward.Thus,pore-water migration results in raised nitrate levels in lower soil layer,but because of that,nitrate levels of pore-water the lower soil layer decrease. On the contrary, at vegetation stage II and the production stage, most of nitrate levels collected at upper porewater samples were lower than those at lower samples. In general,average length of corn root is about 30 cm(Ball-Coelho et al.,1998), so corn absorbs nutrition and water mostly from top-soil layer. At vegetation stage I, nitrogen and its derivatives were adsorbed in lower soil layer.Thus,adsorbed nitrogen sources might result in relatively higher nitrate levels at the lower soil layer.However, a few of nitrate level data (Years 1 and 3 production stages) did not follow this pattern. As described before, heavy rainfall (Year 1 heavy storm during production stage, and Year 2 rainy season) might disturb ordinary nitrate migration pattern in this study.

4. Conclusion

Composted organic waste increased SOM as well as nitrogen content in northern Guam’s calcareous soils and increased nutrient-holding capacity and induced lower bulk density, resulting in improved soil quality and long-term sustainability.Inorganic fertilizer had little influence on holding nitrogen content in the calcareous soil. During the early stages of corn growth, nutrient consumption was low, and nitrate leachate into soil pore water might be higher in compost treated than in fertilizer-treated soils.In addition,high rainfall might have increased nitrate levels in soil pore water.At most growth stages,however,nitrate leachates were lower in compost-treated than in fertilizer-treated soils. Overall,application of composted organic waste is beneficial to northern Guam’s soil not only in minimizing nitrate leaching into groundwater but due to its residual effect,it continued supplying nutrients to the corn plants during the second year of production,even when no land application of compost was carried on during this period.The comparable crop yield from the compost treated plots with the yield from fertilizer applied plots during the second year(data not shown)was clearly an indication of this residual effect.

Furthermore, compost treated plots had 1.8-4.3% higher organic matter than fertilizer-treated plots during the entire study period. This indicated that, the organic matter content of the soil was improved continuously due to the land application of compost(Fig. 5) on plots under study.

Acknowledgement

This manuscript was the result of a project works which was supported by the‘Tropical and Subtropical Agriculture Research(TSTAR)’ and funded by the USDA-NIFA.

Appendix

Table A1 Raw data of Soil organic matter (SOM, w/w%) measured from soils of twenty-eight study plots.

Table A2 Raw data of carbon-to-nitrogen ratios in soils of twenty-eight study plots(before planting).

Table A3 Raw data of carbon-to-nitrogen ratios in soils of twenty-eight study plots(after planting).

Table A4 Raw data of nitrate levels in porewater of twenty-eight study plots at upper depth(Year 1).

Table A4 (continued)

Table A5 Raw data of nitrate levels in porewater of twenty-eight study plots at lower depth(Year 1).

Table A6 Raw data of nitrate levels in porewater of twenty-eight study plots at upper depth(Year 2).

Table A6 (continued)

Table A7 Raw data of nitrate levels in porewater of twenty-eight study plots at lower depth(Year 2).

Table A7 (continued)

Table A8 Raw data of nitrate levels in porewater of twenty-eight study plots at upper depth(Year 3).

Table A8 (continued)

Table A9 Raw data of nitrate levels in porewater of twenty-eight study plots at lower depth(Year 3).

Table A9 (continued)