PertanikaJ. Trap. Agric. Sci. 17(1): 21-26 (1994) ISSN: 0126-6128© Universiti Pertanian Malaysia Press

Response of Selected Tropical Grasses to Irrigation with Brackish Water

U.R. SANGAKKARADepartment ofCrop Science

Faculty ofAgricultureUniversity of Peradeniya,

Sri Lanka

Keywords: tropical grasses, irrigation, brackish water

ABSTRAK

Program pembangunan pertanian di kawasan-kawasan berpasir zon kering Sri Lanka telah mencadangkanpenubuhan foder yang sesuai untuk ternakan di bawah program pengairan menggunakan air masin bawahtanah. Satu kajian telah menilai tindak balas tiga jenis rumput tropikal yang popular dan satu spesis asli yangditanam di tanah berpasir terhadap pengairan menggunakan air masin yang mempunyai konduktiviti 0.5-6.0ms. cm. Kajian telah dijalankan selama 6-7 bulan dengan pengairan yang disesuaikan dengan musim kemarau.Pertumbuhan ketiga-tiga spesis dipengaruhi oleh peningkatan konduktiviti air. Brachia mutica adalah spesisyang paling terjejas. Penghasilan Papalum dilatatum pula tidak begitu terjejas dengan peningkatan kemasinanair. Panicum maximum memberi penghasilan tertinggi dan membekalkan kuantiti foder yang berpatutan didalam semua keadaan. Walau bagaimanapun, spesis yang diperkenalkan memberi penghasilan yang lebihtinggi daripada spesis asli kecuali Brachiara pada tahap konduktiviti yang lebih tinggi. Data menunjukkankesan kemasinan ke atas kadar penghasilan spesis pilihan dan dibentangkan di sini hasil kajian mengenaipenanaman rumput menggunakan spesis-spesis ini di bawah sistem pengairan di kawasan berpasir pada musimkemarau.

ABSTRACT

Agricultural development programmes in the sandy regions of the dry zone of Sri Lanka envisage the establish­ment of suitable fodder for livestock under irrigation programmes using brackish underground water. A studyevaluated the response of three popular tropical grasses and the natural species to irrigation with brackish waterhaving conductivities ranging from 0.5-6.0 ms. cm, when established in a sandy soil. The study was carried outfor 6-7 months, with regular irrigation to correspond to the dry season. Growth of all three species was affected byincreasing conductivity of water. Brachiaria mutica was affected to the greatest extent. Yields of Paspalumdilatatum were reduced to a lesser extent with increasing brackishness of water. Panicum maximum producedthe highest yield and provided a significant quantity offodder in all treatments. However, the introduced speciesoutyielded the natural species with the exception of Brachiaria at the higher levels of conductivity. The dataillustrated the effect of brackishness on the yielding ability of the selected species. The practical significance of thestudy in terms ofpossible uses of these species in growing grasses under irrigation in the sandy tracts for herbageproduction in the dry season is presented.

INTRODUCTION

The production of most plant communities isdependent on the availability of adequate soilmoisture. This phenomenon is also true ofgrasslands of the tropics, which are subject toperiods of severe moisture stress. Thus, totalannual rainfall (Ie Honerou and Hoste 1977),annual and particularly monthly rainfall pattern(Duncan and Woodmansee 1975) or progressiveseasonal total rainfall combined with

temperature (Naveh 1982) can be correlatedwith herbage yields in most tropical rangelands.

Most rangelands in the tropical dry regionsare considered resilient due to their capacity topersist through fluctuating yearly weather pat­terns (Naveh 1982). However, their persistenceis associated with the presence of volunteer earlymaturing species. These species which are tus­socky in nature tend to reduce their productivitywith time (Biddescombe 1987). The lack of ad-

U.R. SANGAKKARA

equate fertility, the presence of alkali soils andthe absence of proper grazing management alsoaffect their productivity.

The rangelands ofSri Lanka account for some700,000 ha of land, spread primarily over the dryzone (Rajaguru 1986). They consist of grassesand tussocky perennial shrubs which providelow quality herbage. Grasses such as Imperata,Cynodon and Panicum dominate these rangelandswhich are found over a wide range of soils. Theproductive capacity of these rangelandsespecially in the dry zone, is affected by theunevenly distributed rainfall due to thepresence of a dry period for 5-6 months of eachyear. Thus seasonal productivity is a commonfeature in these rangelands.

The rangelands of the north western coastalregions ofSri Lanka are characterised by the pres­ence of alkali and sandy soils. The species grow­ing on these rangelands also tend to be coarseand tussocky, and their productivity is dependenton rainfall and soil conditions. Thus during dryperiods the species tend to wither, leaving deador unpalatable material for animals. Another char­acteristic feature of this region is the presence ofa high water table (304 m below soil surface).The quality of the water varies with location andseason and the conductivity ofwater changes rap­idly (0.4-8 #ms/cm). This ground water, whichis easily accessible, could successfully be usedfor production of fodder crops to provideadequate food for livestock during the dry period.

Agricultural development programmesenvisage the improvement of grazing lands of thedry regions which have an immense potentialfor producing adequate feed for livestock(Rajaguru 1986). In the north western region,the development programmes envisage the useof ground water. However, salinity of the wateraffects the growth of most grass species(Whiteman 1980), and all tropical species donot respond similarly to brackish water (Bogdan1977). Studies and surveys (e.g. Russel 1976)illustrate different degrees of salt tolerance oftropical herbage species.

Studies on the comparative performance oftropical herbage species under irrigation, espe­cially with brackish water, are not widely reportedin Sri Lanka. However, the use ofavailable groundwater is an important phenomenon in improvingrangelands. A case study was carried out on asandy tract in the north western region of Sri

Lanka to evaluate the performance of threeselected grasses, and to compare their perfor­mance with that of the natural species when pro­vided with irrigation water having different con­ductivities. The principal objective of theprogramme was to study their performance inthe dry season as productivity is generally ad­equate in the wet season for animal require­ments.

MATERIALS AND METHODS

The study was carried out at a site on a sandytract, which is typical of the soils of the region,over a period of eight months that correspondedto the annual dry season. The selected sitecontained the indigenous species prior to thedevelopment of the trial. The water table of thelocation was at a depth of 3.5-4. m from the soilsurface.

The species selected were Brachiari mutica,Paspalum dilatatum and Panicum maximum whichare grown widely in different environments of thetropics (Bogdan 1977). The experimental de­sign used was a randomized block design withthree replicates.

The grasses were established from uniformvegetative propagules on 2 x 3 plots at the onsetof rains and maintained for one year. The spac­ing adopted was 30 x 40 cm. In addition, similarsized plots of the indigenous species were alsomaintained for comparison. A uniform fertilizerrate equivalent to 800 Kg/ha ofa 15: 10: 15 N:P:Kmixture was applied once in six months to allplots prior to the beginning of the trial.

The herbage of all plots were cut to a uni­form height of 4 cm at the end of the rainy sea­son (early January - 15 months after planting),before the experiment was initiated. During theexperimental period, which corresponded to thedry season, the separated plots of each specieswere irrigated manually with ground water hav­ing four different levels ofbrackishness (measuredby conductivity, using a standard conductivitymeter once a month). The water, obtained fromadjacent sources (wells) had a mean conductivityof 0.5 ms/cm ± 0.06; 2.1 ms/cm ± 0.13; 4.2 ms/cm ± 0.42 and 5.9 ms/cm ± 0.21 over the experi­mental period. These treatments were termed C1,C2, C3 and C4 respectively. The rate of applica­tion of water was 30 I per plot at 8-10 d intervals,to correspond to flood irrigation as practised byfarmers of the region. The herbage of each plot

22 PERTANlKAJ. TRap. AGRIC. SCI. VOL. 17 NO.1, 1994

RESPONSE OF SELECTED TROPICAL GRASSES TO IRRIGATION WITH BRACKISH WATER

was harvested at two monthly intervals over theexperimental period. Thus, four harvests weretaken from February to August by cutting to aheight of 5 cm above ground level. The herbageharvested was dried at 80°C to a constant weightand dry matter contents were determined.

RESULTS AND DISCUSSION

The climatic parameters during the experimen­tal period and important soil characteristics ofthe site are presented in Table 1. The rainfall overthe period, the temperature and estimated evapo­transpiration rates are similar to the long term(10 year) mean values of the region. This indi­cates that the climate of the season of study con­formed to that of the general climatic parametersof the region.

The total rainfall over the experimental pe­riod was 368 mm, which was approximately 25%of the annual rainfall. This is a characteristic fea­ture of the dry season on the dry zone ofSri Lanka

wet season high rainfall could leach the soil, thuspreventing the build up of excessive levels ofsalinity with time.

The soil characteristics illustrate the sandynature of the site, and the relative highconductivity. The pH suggested alkali conditionsand the percentage organic carbon was very lowdue to the rapid degradation of any addedmaterial. The water holding capacity of the soilwas also low. (Table 1). These parameters,coupled with the climatic conditions of the dryseason, make crop production difficult in thisregion under rainfed conditions.

Dry Matter Productivity

The productivity of natural species and ofBrachiaria when grown with irrigation usingground water of different brackishness over thedry season is presented in Table 2. The dry matterproductivity of both species was less at the secondharvest due to the low rainfall. Thereafter, yields

TABLE 1Mean climatic and selected soil parameters of the experimental site.

A. Climatic Factors

Rainfall (mm/month)

Mean monthly rainfall (mm/month)

Jan119

140

Feb54

41

Mar11

18

Apr12

16

May25

31

Jun49

52

Jul65

74

Aug21

15

Sep12

08

Mean monthly temp. Co.

Mean monthly temp. for 10 years

Estimated ET (mm/day)Mean of5 years (mm/dy)

29.1 29.5 29.1 29.9 30.4 30.8 30.1 29.9 30.4

28.4 29.2 29.3 29.6 30.1 30.5 29.9 30.3 30.8

4.7 6.1 6.2 6.0 5.4 5.0 6.1 6.2 5.84.8 6.1 6.5 5.9 5.9 5.2 6.5 6.3 5.9

B. Soil Factors

Texture

Water holding capacity

% Organic matter

Sandy (79.5% ± 1.59 sand)

14.5% ± 2.42 v/v

0.094% ± 0.013

pH - 7.98 ± 0.44 (1:25 H20)

Conductivity - 2.98 ms/cm ± 0.23

(Domros 1974). The high temperature resultedin high evapotranspiration rates. Thus, the quan­tity ofwater lost to the atmosphere exceeded rain­fall, resulting in the upward movement of brack­ish water from the ground water table. This, assuggested by Hanson et al (1979) increased thesalinity level of the soil especially in the dry sea­son, which is characteristic of this region. In the

increased due to some rainfall during themonsoon, which however, was less than theestimated evapotranspiration rates. Thephenomenon affected both species even with thesupply of irrigation in all treatments, irrespectiveof the increasing brackishness. A comparison ofthe productivity of the two species illustrated thehigh yielding ability of Brachiaria when irrigated

PERTANIKAJ. TROP. AGRIe. SCI. VOL. 17 NO.1, 1994 23

U.R. SANGAKKARA

with good quality water. This indicated thepotential of introducing high yielding grassesto provide adequate fodder for livestock by theprovision of good quality irrigation water whichis available in some parts of this region. However,with the increasing brackishness of water, theproductivity of Brachiariawas less but greater thanthat of the natural species found in the region.This suggests that if the available ground water

had a higher conductivity, the use of indigenousgrasses would be a more viable source of fodderthan species such as Brachiaria which areconsidered susceptible to salinity (Bogdan1977). Table 3 presents the productivity ofPaspalum and Panicum over the experimentalperiod. At all harvests the herbage yields of bothspecies exceeded those of the natural speciesand Brachiaria. This illustrates the greater

TABLE 2Response of natural species and Brachiaria to irrigation with brackish water

Treatments

Species

atural Species

S.E. (Mean)

Irrigation

Cl

C2

C3

C4

HI

51

44

33

28

2.47

Herbage Yields (g/dry matter/m2)

H2 H3

18 24

13 20

11 16

9 12

1.86 4.23

H4

39

31

26

19

5.12

Brachiaria

S.E. (Mean)

ClC2C3C4

85 31 6961 20 4524 14 3010 7 10

7.14 2.77 8.04

81592115

9.12

C1 = brackishness of 0.5 ms/cm ± 0.06C2 = brackishness of2.1 ms/cm ± 0.13C3 = brackishness of 4.2 ms/cm ± 0.42C4 = brackishness of 5.9 ms/cm ± 0.21

TABLE 3Response of Paspalum and Panicum to irrigation with brackish water

Treatments

Species

Paspalum

Herbage Yields (g/dry matter/m)

Irrigation HI H2 H3

Cl 128 49 80

C2 104 41 70

C3 91 33 61

C4 80 25 53

H4

99

75

66

45

S.E. (Mean)

Panicum

S.E. (Mean)

C1C2C3C4

3.05 6.11 5.08

178 70 95154 58 76136 51 61121 46 54

4.17 8.82 9.17

4.12

135118101

89

5.98

Cl = brackishness of 0.5 ms/cm ± 0.06C2 = brackishness of2.1 ms/cm ± 0.13C3 = brackishness of 4.2 ms/cm ± 0.42C4 = brackishness of 5.9 ms/cm ± 0.21

24 PERTANIKAJ. TROP. AGRIC. SCI. VOL. 17 NO.1, 1994

RESPONSE OF SELECTED TROPICAL GRASSES TO IRRIGATION WITH BRACKISH WATER

500

Fig. 1: Effect of brackishness of irrigation water on cum­ulative herbage yields

adaptability of these two species to differentqualities of irrigation water. A comparison of theproductivity of the two species (Table 3)indicates the higher yields of Panicum at allharvests in all treatments. This clearly illustratesthe suitability of Panicum as an adaptable speciesfor the dry regions; and its capacity to producerelatively higher yields especially in the dryseason when provided with irrigation usingbrackish water. The data also confirmed earlierreports (e.g. Russel 1976, Whiteman 1980) ofthe tolerance of Panicum to alkaline conditions.The cumulative yields obtained over the seasonfrom the four types or herbage when irrigatedwith water of different levels of brackishness arepresented in Fig. 1. The data illustrate the pooryielding ability of the natural species irrespectiveof the different types of irrigation water. Thissuggests that these rangelands require improvedspecies to increase their productivity even in thepresence of irrigation water as the natural speciesdo not seem to have the inherent capacity torespond to irrigation even with good quality water.Fig. 1 also illustrates the high yielding ability ofPanicum in dry season environmental conditionssupplemented with irrigation using water ofdifferent levels of brackishness. The yields ofPaspalum which are lower than those of Panicumshowed greater adaptability than Brachiana.

400

300

Yield(gDM/m2

)

200

100

. Panicum Y = - 0.2914x + 2.96 ( r2 = 0.74)

PaspalumY =-0.~2.41 (r2 = 0.81)

.~ '~

Brachiaria Y =-0.5~~+ 3.95 (r2 =0.69)

.~ "'~.

.'"Natural Species Y = - 0.1142x + 1.89 ( r2 = 0.76)

.-----. ~~:,.

0.5 2.1 4.2 5.9

Level of brackishness (ms/cm)

Regression equations (Fig. 1) for the cumu­lative yield data of all species show the declineof yields of species with increasing brackishnessof irrigation water. Amongst the species tested,the natural species are least susceptible to in­creasing brackishness of irrigation water.Brachiana is affected to the greatest extent. Theyield decrease in Brachiana was approximately30% between the C1 and C2 treatments and by45% between C2, C3 and C4 treatments. Thisalso illustrates the unsuitability of Brachiana forthe rangelands of these regions, especially if ir­rigation is provided with water having a high levelof brackishness. The yield decline of PanicumPaspalum and the natural species is around 15 ­20% between the irrigation treatments. However,although the reduced yield with increasingbrackishness is greater in Panicum, its highestyields in all treatments highlight the adaptabilityof this species to the environmental conditionswith irrigation from different qualities of water.

CONCLUSION

In terms of practical benefits, the pasture im­provement programmes of this region are ofvitalimportance to the livestock industry (Rajaguru1986). The land available for pastures is generallynot suited for cropping due to the alkalinenature of the sandy soil, and the brackishness ofground water. Improvement of these rangelandsrequires adaptable species, especially to providefodder for the animals during the dry season. Theresults suggest the possibilities of using speciessuch as Panicum and Paspalum for this purpose.These species which are tolerant to brackishnesscould be useful in increasing the productivity ofthese rangelands in combination with the readilyavailable source of ground water, especiallyduring the dry period when other types ofvegetation succumb to the harsh environmentalconditions.

ACKNOWLEDGEMENTS

Gratitude is expressed to Mr H.H. Ratnayake forresearch assistance and the MARGA Institute forfunds.

REFERENCES

BIDDESCOMBE, E.F. 1987. The productivity ofmediterranean and semi arid grasslands. InEcosystems of the World 17 B-Managed grasslands,ed. RW. Snaydon, p 19-28. Holland: Elseiver.

PERTANIKAJ. TROP. AGRIC. SCI. VOL. 17 NO.1, 1994 25

U.R. SANGAKKARA

BOGDAN, AY. 1977. Tropical Pasture and Fodder Plants.475 p. London: Longman.

DOMRos, M. 1974. Agroclimate of Ceylon. 267 p. WestGermany: Steiner Verlag.

DUNCAN, D.A and R.G. Woodmansee. 1975. Forecast­ing forage yields from precipitation inCalifornia's annual rangelands. Journal ofRangeA1anagement28:327-329.

HANSON,]., O.W. ISRAELSON and G.£. STRINGHAM. 1979.Irrigation principles and practices. 417 p. London:John Wiley & Sons.

LE HONEROU, H.N. and C.H. HOSTE. 1977. Range­land production - annual rainfall relations inthe mediterranean basin and in the MricanSahelo - Sudanian Zone. Journal ofRange A1an­agement30: 181-189.

NAVEH, Z. 1982. The dependence of the productivityofa semiarid mediterranean hill pasture ecosys­tem on climatic fluctuations. Agricultural Envi­ronment7: 47-6l.

RAJAGURU, AS.B. 1986. Livestock development policyplan for Sri Lanka. Report ofthe Society ofResearchJor Native Livestock 11: 37-64.

RUSSEL, J. 1976. Comparative salt tolerance of sometropical and temperate legumes and tropicalgrasses. Aust. of Experimental Agric. and AnimalHusbandry 16: 103-109.

WHITEMAN, P.c. 1980. Tropical Pasture Science 390 p.London: Oxford University Press.

(Received 5 A1ay 1993)

26 PERTANIKAJ. TROP. AGRIC. SCI. VOL. 17 NO.1, 1994