Field studies on controlled drainage and recycling irrigation drainage for reduction of nutrient loading from arable land

War. Sci. Tech. Vol. 33, No. 4-5, pp. 333-339, 1996. e.> Pergamon Copyright CC> 1996 IA WQ. Published by Elsevier Science Ltd.

Pnnted m Great Bntam. All nghts reserved.

PH: S0273-1223(96)00249-1 0273-1223/96 $15'00 + 0'00

FIELD STUDIES ON CONTROLLED DRAINAGE AND RECYCLING IRRIGATION DRAINAGE FOR REDUCTION OF NUTRIENT LOADING FROM ARABLE LAND

Maija Paasonen-Kivekas,* Tuomo Karvonen,* Pertti Vakkilainen,* Noor Sepahi,* louko Kleemola** and Matti Teittinen**

* Helsinki University of Technology. Laboratory of Water Resources Engineering. Konemiehentie 2. FlN-02IS0 Espoo. Finland •• University of Helsinki. Department of Plant Production. P.O. Box 27. FlN-OOOJ4 University of Helsinki, Finland

ABSTRACT

Three on-farm trials were established to evaluate the suitability of water table management for Finnish growing conditions. The sites differed in soil texture, topography and cultivation. Drainage control was managed through specific wells in collector pipes. In recycling, a reservoir stored drainage water discharging from the fields. This water was used for subirrigation through conventional drainage or a dual level irrigation-drainage system. Hydro-meteorological variables were monitored continuously and real time transfer of the data was carried out via radlOlink and microcomputers. Quality of surface and subsurface waters was surveyed by manual sampling from the weirs, piezometers and reservoirs. Physical soil properties and mineral nitrogen (N) from several soil profiles were determined. Furthermore, above ground biomass, and N content and yield of crop were observed. In fine sandlloamy sand, subirrigation and controlled drainage raised groundwater table on average 80 cm compared to the reference areas. N concentration in the reservoirs declined significantly during May-August. Nitrogen in the piezometers and soil showed considerable spatIal and temporal variation within a single field. Evidence of the relationship between groundwater level and N concentration was observed. However, no unique correlation from the existing data could be detected. N yield of cereals was 10-50% higher in the controlled drainage and subirrigation areas compared to the reference areas. Most of this extra N was allocated to grains and removed from the fields reducing N load into the environment. Copyright © 1996 IA WQ. Published by Elsevier Science Ltd.

KEYWORDS

Agricultural pollution; controlled drainage; nitrogen; subirrigation,

INTRODUCTION

In Finland, agriculture induces the major part of the total nitrogen and phosphorus load to water courses (Rekolainen, 1993). The most serious effect due to the nutrient losses is extensive eutrofication of lakes,

333

334 M. PAASONEN-KIVEKAS et al

rivers and coastal waters. Whereas, nitrate contamination of aquifers is not a problem, in Finland, like in many other European countries. Degradation of groundwater quality due to cattle rising and crop production has been observed mainly in single wells (Huttunen, 1995). Several methods for abatement of nutrient leaching have been investigated during the last decade. One of the promising techniques might be water table management using controlled drainage and recycling of runoff waters. The methods promote sustainable crop production with decreased pollution and improved crop yield.

Water table management has been widely used in the Midwest and Southwest Regions of the United States. Originally controlled drainage was developed for subsurface irrigation but later the environmental benefits came up (Skaggs, 1987). A recycling subirrigation-drainage system has been investigated by Melvin and Kanwar (1995) from the point of view of the environment and crop production. According to the extensive field experiments nitrogen and phosphorus loading decreased about 45% through drainage control compared to conventional drainage (Evans et al.,. 1989, 1995). The reduction of nutrient load has been almost entirely due to the decline of total outflow (Evans et aI., 1995; Bendoricchio and Giardini, 1994). Some studies have also showed substantial reduction of nitrate concentrations at shallow water table depth (Madramootoo et a.,. 1993; Kalita et al .. 1993). The decline of total runoff is attributed to the storage capacity of the field and high evapotranspiration. The wet soil possibly enhances denitrification reducing nitrate to nitrogen gas. Furthermore, increased crop yield was achieved through favorable soil moisture conditions and more efficient use of soil nutrients. In Finland, a preliminary study was made on the effect of subirrigation on potato yield (Ahonen, 1991). In the subirrigated area the yield increased 10% and the amount of potato scab declined 50% compared to the area without subirrigation.

Helsinki University of Technology, University of Helsinki and the Drainage Research Foundation initiated a research project on controlled drainage and subirrigation in Finnish growing conditions (Laikari and Karvonen, 1992). The experiments started in 1992-1993 continuing until 1996. The objective of the project is to find out the soil types and technical structures suitable for water table management. Furthermore, the influence of the control practices on nutrient load and on crop yield is investigated. Field scale measurements were established to study the techniques under normal cultivation. The research is conducted in close co-operation with the local farmers. Problems are expected due to the short growing season and characteristics of the experimental sites. Moreover, winter time may cause severe difficulties for control structures, instrumentation and sampling. In this study the experimental fields and results from 1993-1994 are presented. Mitigation of agricultural loading by water table management is discussed.

EXPERIMENTAL FIELDS AND SCHEDULE

The experiments were carried out at three locations with different soil characteristics. For controlled drainage, one on-farm trial was established in southern Finland (Kirkkonumrni) and one in western Finland (Lapua). A recycling system of drainage water was constructed in southern Finland (Kirkkonummi) and in northern Finland (Tyrnava). In recycling, a reservoir stores water discharge from the fields during wet periods and excess irrigation. At the Kirkkonummi site, recycling is carried out through conventional drainage pipes. At the Tyrnavii site, there are both conventional and dual level pipes for subirrigation•drainage.

Both in controlled drainage and subirrigation, water table is managed using specific wells in collector pipes.

The field measurements have been carried out by automatic monitoring and manual sampling. Meteorological variables, groundwater level and water discharge have been observed continuously and real time transfer of the data has been conducted via radiolink and microcomputers (Koivusalo and Hyvonen, 1994). Water samples have been taken from surface and subsurface runoff (V-type weirs), groundwater (piezometers) and the reservoirs. Physical properties of the soils have been determined in the laboratory. Spatial and temporal distribution of mineral nitrogen in the soil have also been measured. Above ground biomass and nitrogen content and yield of crop have been determined both in the areas under water table management and areas under conventional drainage (Teittinen et al .. 1994; Teittinen, 1995).

Controlled drainage and recycling irrigation drainage 335

The Kirkkonummi site consists of 5 ha clay fields with a steep surface slope. Controlled drainage is 'carried out in one field through several PVC wells. In the other field, a reservoir and new plastic pipes and control wells were constructed for recycling in 1993-1994. The crop was spring wheat, in 1993, and barley, in 1994-1995. Rate of nitrogen fertilizer was 120 and 100 kg Nlba, respectively. Drainage control has been managed throughout the year. Control of water table has been lowered for some weeks to facilitate crop establishment and harvest, in May and September. Three monitoring stations exist in different parts of the site. Groundwater table and quality have been observed from 10-24 piezometers at a depth of 1-2.8 m. Some of the control wells have a V -type weir for drainage flow measurements along a collector pipe. Samples on surface and subsurface runoff have been collected since August 1992 and groundwater samples since April 1993. Ammonium and nitrate nitrogen and moisture in the soil profiles have been observed in intervals of 15-90 days since May 1993 (Paasonen-Kivekiis, 1994).

The Lapua site is entirely 3 hectares including two separate fields. The site is located in the zone of acid sulphate soil. This soil causes severe problems both for cultivation and water pollution in the west coast of Finland. Fine sand exists in the soil surface until a depth of 50 cm and loamy sand deeper in the profile. One of the fields has controlled drainage and the other conventional drainage. The water table is controlled through one well in the drainage outlet. The fields are used for intensive potato cultivation. Groundwater level and quality have been observed from 8-12 piezometers since June 1993. Mineral nitrogen and pH from the 0-1501240 cm soil profiles have been determined several times during frost free period. Crop variables have been observed in 1995.

The area of the Tyrniivii site is 8 hectares. The relatively plain fields are fine sand until a depth of 1.2 m and finer textured soil deeper in the profile. In the dual level subirrigation-drainage system, shallow irrigation pipes are installed perpendicular to drainage pipes. The irrigation pipes are at a depth of 0.6-0.8 m with 4 m spacing. The respective dimensions for drainage pipes are 1.1 m and 30 m. The pipes in the other subirrigation-drainage system have a depth of 1.1 m and spacing of 14 m. The reservoir is located in irreclaimable land next to the fields, approximately 5 m below the highest point in the field. There are one well for distribution of irrigation water and several head control wells in the collector pipes. During dry seasons it is also possible to lead external irrigation water from the adjacent river. Besides the subirrigated areas, there is a reference area with no irrigation and drainage. In 1994, oat was grown in the studied subareas with a fertilizer rate of 91 kg Nlba. In the summer 1995, the crop was potato with manure application about 180 kg Nlba. The field observations include a weather station, 12 piezometers at a depth of 1.3-2.0 m and V-type weirs for subsurface drainage flow. Both the hydrometeorological and water quality measurements started in the summer 1994. Soil sampling for mineral N determination has been conducted from 0-150 cm soil profiles since June 1994.

The piezometers to collect water samplers and to observe piezometric heads were made of 0 30 mm PVC pipe. The water table depth is monitored every 15 minute by sensors in 3-4 tubes at all the sites. The water samples are taken from the piezometers with a manual pump in intervals of 1-4 weeks. The piezometers are pumped out one day before sampling, and the samples are collected on the following day. Water table depth is measured both in the state of exhausting and filling. The samples are analyzed for soluble nitrogen (NOr N, N0z-N+N03-N, NH4-N) and total nitrogen. Furthermore, phosphate (P04-P) and total phosphorus and pH are determined. Ammonium and nitrate nitrogen in the soil samples are analyzed using 2 M KCl•extraction.

RESULTS AND DISCUSSION

Water table management had a considerable effect on groundwater level and crop yield. Evidence of a relationship between groundwater table and nitrogen concentration was observed, but no unique correlation could be detected. In the summer 1994, subirrigation raised the groundwater level in the fine sand field (the Tyrniivii site) 60-90 cm higher than in the area with no irrigation and drainage. Towards autumn the difference declined, but the influence of drainage control could still be detected (Fig. I). At the Lapua site, the gr?undwater table in the controlled field was on average 80 cm higher compared to the uncontrolled field, In May-July 1994. However, later in the summer the groundwater level declined significantly in the

336 M. PAASONEN-KIVEKAs et al.

controlled area, too, due to the exceptionally dry weather. In the fine sand/loamy sand, the groundwater table responded almost immediately to control measures due to the favorable soil hydraulic conductivity.

At the Kirkkonummi site, groundwater level rose only in the immediate surrounding of the control wells. Subirrigation seemed to increase soil moisture in the vicinity of the drains. However, its influence on groundwater level remained indistinct during the first experimental year, 1995. This is attributed to the steep slope, low hydraulic conductivity and the earlier tile pipes in the area. Besides piezometers soil moisture measurements are needed, particularly in clay soils, to evaluate the effect of control practices.

... ! ~ "

July-August 1994 Septerrber 1994-January 1995

0 0

-40 -40 ... -80

! ~

-80

" -120 -120

-160 -160

PJ.OIIIIter PJ.ometer

Figure 1. The average water table depth (GWL) below soil surface at the Tymava site, in July-August 1994 and September 1994-January 1995. Piezometers of the area A = sub irrigation with conventional

drainage, B = subirrigation with dual level pipes and C = no drainage and irrigation.

Groundwater quality showed over all the site conditions high variation both in space and time. N03-N formed major part of total nitrogen in subsurface waters, in the Kirkkonumrni and Tymiivii site, as expected. At the Lapua site, low pH and high concentrations were characteristic for the area under controlled drainage. Whereas, N03-N was the dominant fraction in the reference area. In Kirkkonummi, the average total N concentration during June-October 1993 varied from 1.3 to IS mg Nil declining downhill. The pH-value, on average 6.S±O.4, and total phosphorus, O.7S±O.96 mg/l, showed no systematic distribution along the slope. In Tymiivii, N concentration in the subirrigated areas was on average lower than in the reference area both during the growing season and autumn 1994 (Fig. 2). However, the adjacent piezometers reflected high variation both in nutrient concentrations and pH-values. At the Lapua site, NH4-N formed even 95 % of the total N in the groundwater samples next to the controlled drainage outlet. A similar type of distribution was observed in the soil profiles where high NH4-N concentrations existed below groundwater table. pH of the water samples from the controlled area was generally below 3.5, and from the uncontrolled area almost 5.0.

The on site characteristics, like topography, small scale variation of soil properties and fertilizer application, are expected to cause the irregular distribution of nutrients. Therefore, a possible effect of water table management on groundwater quality is hard to distinguish from the total variation. In the previous studies by Madramootoo et al. (1993) and Kalita et al. (1993), reduction of N03-N concentration was observed in small plots and lysimeters with a stable water level and several replicates. At the Tymiivii site, the shallow groundwater table might have enhanced denitrification. Moreover, dilution due to subirrigation could also reduce the concentrations. Declining N concentrations according to the field slope have been observed in other studies, too (ref. Hallberg, 1989). Probable reasons are dispersion and dilution and high denitrification rates in the lower part of the field. The exceptional nitrogen concentrations and pH-values, at the Lapua site,

Controlled drainage and recycling irrigation drainage 337

are merely due to the acidity status of the soil. Acidity seems to vary on a very small scale even within a field. The high ammonium concentrations are most likely caused by nitrate reduction to ammonium (Patrick, 1982). Another reason can be retardation of nitrification due to the low pH and anaerobic conditions. Conventional drainage has been reliazed to contribute on formation of acidity (Palko, 1994). In dry season, water table management can reduce acidity preventing oxidation of sulphides deeper in the soil.

i; E -z

t ....

July-August 1994 September-December 1994

10 10

8 8

6 I 6 z

4 ! 4

2 2

0 0 A1 A2 B1 B2 B3 B4 C1 C2 A1 A2 B1 B2 B3 B4 C1 C2

PIezometer PlllZonwter

Figure 2. The average total nitrogen concentration of groundwater at the Tymiivii site, in July-August 1994 and September-December 1994. Piezometers of the area A = sub irrigation with conventional

drainage, 8 = subirrigation with dual-pipe system and C = no drainage and irrigation.

20~------------------------------------------~

15

~ .5. • Drain 1 z 10 • Drain 2 i o Reserwir ~

5

o April May June July Aug Sep Q:t

Figure 3. Nitrogen concentration in the reservoir and subsurface drains at the Kirkkonummi site. in April•October 1995 (monthly averages).

Nitrogen in drainage water had a seasonal pattern. as expected. At the Kirkkonummi site. heavy rainfall events occuring after fertilization induced nitrogen leaching from the soil surface into deeper layers. Accordingly. peak concentrations in the subsurface drains reached even 60 mg Nil and in the main ditch 20 mg Nil. The N levels in the reservoir stayed significantly lower than in the drainage waters (Fig. 2). However. the concentration difference diminished gradually along with the decreasing temperature. Similar phenomenon was also observed in the reservoir at the Tymiivii site. The findings support the results presented by Melvin and Kanwar (1995). Immobilization and denitrification and dilution are likely the main

M, PAASONEN"KIYEKAS

reasons for soil

biomass accumulation of the cereals increased both in the fields under controlled and to the reference areas, Moreover. N observed, In Kirkkonummi at harvest time 1993, N of area than in the area outside control. However, a very

summer, At the site, the N system was 51 % than in the reference area, The value for the system with conventional drains was 29%, Most of the extra N was allocated to and therefore removed from the field, High variation between caused in soil

CONCLUSIONS

on water table management was carried out in three on-farm trials, in Finland, Controlled ",,,'rlfTQrtl\" had a remarkable effect on level and in the fine sand

fields, However, of the soil remained indistinct of the on site characteristics, The N in the controlled fields reduced the

N load into the environment. Furthcrmore, concentration of water diminished in the reservoir the summer. In the acid soils, water table management is pv>"'''''PrI to reduce acid load into surface waters,

In the some arose related to the field scale measurements. lnflnence of water table control on outflow rates could not be detected from the data, It was cvident that effect of controlled on weather conditions and management This has also been observed in earlier studies (Evans et ai" 1995), Small scale expelrirrlents with to more reliable information on water than on-farm trials with

",,,,,prn,''''Bmr,r of water table management should be done from the of view of the crop and economy, fn controlled much lower investment are needed than in the

Sulllfl'lg,ltle,n, On the other hand, lower increase in the crop is also Economical benefits of a system on the type and of cultivated Location and

the necessary reservoir and of water table control needs attention in the

ACKNOWLEDGEMENTS

The authors thank Ms Aino Peltola, Mr Anni Taskinen and Me Matti Keto for Ollt the chemical and The authors also express their to Mr Rallno Peltomaa for technical help and

Mr Kalevi Pelanteri and Mr Veikko for field work, The assistance of Mr Pertti and Mr Harri Koivllsalo in maintenance of the stations is The study has received financial support from the of Finland, Maa- ja Vesitekniikan Tuki Association, Kem.ira Foundation, of and and the Research Foundation,

REFERENCES

Ahonen, J, (1991), 1"0!7jalVeslAas:[1!1I7n

"Itllm,,,"!,,,,, in Finland),

Bendoricchio, G. Giardini, L (1994), A controlled drainage aeI110fl:,tra'ti()jl

Padova, pp, 768-769. in lialy, ESA

L,omroucu uramagc: ana rc\:ycung lrngauon unnnugc

Evans, R, 0., Gilliam, J, W. and Skaggs, R W. (1989). Effects afagricultural water table managemelll Oil drainage willer quuUt)'. Water Resources Research Institute of the University of North Carolina,Report No. Raleigh.

Evans, R. O. , Skaggs, R W. and Gilliam, J. W. (1995). Controlled conventional on water quality. Journal and Drainage, 121

Hallberg, G. R. Nitrate in ground water in the United States, In: Protection, Follett, R F. (ed.), Developments in Agricultural and Managed·Fores(,

Huttunen, L. (1995). Nitraat!! kohoamassa pohjavedenotlamoilla in nitrate concentrations at intakes in Finland), Vesitalous 36 (3),

Kalila, P. K., McKool, D, and Kanwar, R, S. (1993), Groundwater Management for Environmental Benefits. Proceedings of International Conference on Environmentally Sound Water Resources Utili:ation, Bangkok. Thailand, 8·11 November 1993,1·174-181.

KoivusaJo, H, and Hyvoncn, p, (1994), Slilirosalaojitus· ja padotuskastelu!utkimusrcn kockenuicn minausjiil:jesrelyr ja liedonkcruu· ja siirto·ohjeimis\o (Monit()ring and data transfer in the experimentalsiles of controlled drainage and suhirrigalion research) Vesitalouden laboratorion monislesalja 1994:6. Helsinki University Lab. of WaleI' Resources Engineering, Espo(), Finland,

Laikari, E, and Karvonen, T. (cds,) (1992), Saatosalojilus Koekenttien perustaminen (Controlled drainage Implementation of the experimental sites). Salaojituksen tutkimusyhdistyksen tleliote, 16, Helsinki, Finland.

Madramoo\oo, C. A" Dodds, G, 1'. and Papadopoulos, A. (1993), and environmental benefits of water·table management. J [rriga/ioll alld Drainage Engnrg, 119 (6),

Melvin, S, W. and Kanwar, R, S, (1995). Environmental and economic impacts a recycling suhirrigation-drainage system, In: Subirrigaliall and Control/ed Drainage, Belcher, H. W, and I)'ltd, F. M, Lewis Puhlishers,

Paasonen-Kivekiis, M. (1994). Distribution of soil nitrogen in an agricultural field, In: Spatial and temporal variability and interdependencies among Kettunen c! al. NHP.Report 36, Finland, 142-148,

Palko, J. (1994), Acid sulphate soils thcir agricultural and environmental prohlems in Acta Universitatis Ouiuensis, Series C, TECHNICA 75, Oulu, Finland,

Patrick, W. H. (1982), Nitrogen TransfonnatiollS in Suhmerged Soils, In: Nitrogen in Agricultural Soils, Stevenson, F. 1. (ed.), Agronomy No, 22, S5SA, Madison, Wisconsin, 449-466.

Rekolainen, S, (1993), Assessment and mitigation of agricultural water pollution. Publicatiolls oJ the Water and Environment Research Institute, 12, National Board of Waters and the Environment, Finland, Helsinki.

Skaggs, R. W, (1987). and management of drainage systems, In: nt Fifth National Drainage Symposium, ASAE,5t 1-12,

Teittinen, M., Karvonen, and Paasonen·Kivek!ls, M, (1994), Simulation of effects of controlled drainage on nitrogen leaching and growth of spring wheal. In: Modeling the Jate of agrochemicais and fertilizers in the environment, Giupponi, C. and Marani, A. (cds.), In!. Workshop, Venice(Jtaly), 3-5 March 1994.

Tcittincn, M, (1995), Influence of controlled drainage on soil water and nutrient balances of crops, In: Nordiskt lalllbruk i del "ya Europa, NJF XX:c kongress, Reykjavik 26-29.juni 1995. Nordiskjord bruksforskning, 77 (2),