Effects of Light Intensity on Growth of Some Tropical Forage Species. I. Interaction of Light Intensity and Nitrogen Fertilization on Six Forage Grasses1

In order to determine the response of some tropical grass species to low light situations such as under plantation crops, seasonal cloud cover, etc., six tropical forage grasses were evaluated over a 20 month period on an Oxic Haplustoll in Hawaii (100 m above sea level) under four light regimes (100, 70, 45, and 27/, daylight using polypropylene netting) in the field. The forage grasses evaluated were: Brachiaria brizantha, B. miliifirmis, Digitaria decumbens, Panicum maximum, Pennisetum clandestinum, and P. purpureum. Dry matter yields of N-fertilized (865 kg N ha' yr) grasses were highest at 100 and 70% daylight (16 to 40 metric tons of dry matter (DM) ha' yr), with P. maximum and P. purpureum having highest yields. Under 277, daylight, yields were 8 to 15 tons, with P. maximum, B. brizantha, and B. miliiformis having highest yields. When no N was added maximum yields in tons ha' yr' were B. mWiformis, 9.2 at 27% daylight; D. decumbens, P. maximum, and B. brizantha, 18.5 to 15.0 at 45% daylight; and P. clandestinum, 9.2 at 70% daylight. P. purpureum without N yielded 30 tons at full daylight, apparently because its very extensive root system invaded adjacent N-fertilized plots. Percentage of DM in the forage decreased with shading and N fertilization. Percent N increased with decreasing light intensity (from 1.0 to 1.6% in the minus-N and from 1.2 to 1.9% in the plus-N treatments). Slight acetylene reduction activity was found in soil cores beneath all species, except B. brizantha, indicating that the plants were almost entirely dependent on soil and fertilizer N for growth. Sward height increased significantly with decreasing light intensity and N fertilization. Concentrations of P, K, Ca, Mg, S, Cu, and Zu tended to be higher in shaded forage, higher in N-fertilized forage (except for P and Zn), and generally higher during the cool season. Thus under N-deficient conditions, most yield and forage quality parameters were enhanced under moderate shade. Conversely, the tropical grasses studied generally responded to N fertilization only under conditions of moderate to high solar radiation. Root weight data and observations on the rate of recovery after clipping indicate the shaded pastures would require careful management to avoid excessive depletion of root reserves, either by lenient grazing to maintain high leaf areas or by allowing an extended recovery period. Additional index words: Shading, Acetylene reduction, Nfixation, N percentage, Root weight, Dry matter content, Height, LAI, Brachiaria brizantha, Brachiaria mitiiformis, Digitaria decumbens, Panicum maximton, Penxisetum cfandestiuuan, Penniwtum purpuretan. THE climate of tropical areas is characterized by intense year-round solar radiation above the atmosphere. However, during the wet season, solar radiation re aching the earth's surface may be quite low due to dense cloud cover. In Hawaii, solar radiation in different areas varies from 429 to 704 cal cm2 day in June and from 244 to 426 cal cm day' in December (Yoshihara and Ekern, 1977). In many areas where tree crops are grown (e.g., coconut, oil palm, timber, firewood, or rubber trees), forage crops and pastures may be grown in the partial shade of these trees, especially during the first few years. Thus it is often necessary to grow forage grasses under low-light conditions in the tropics. Decreased dry matter (DM) production under reduced light intensity has been found with many grass and legume species (Blackman and Templeman, 1938; Burton et al., 1959; Deinum, 1966; Ludlow et al., 1974), but the magnitude of the effect on growth depends on the stage of growth and on concurrent temperature, moisture, and nutrient (especially N) stresses. Positive interactions between light intensity and N application have been reported for several temperate grasses (Blackman and Templeman, 1938; Burton et al., 1959; Deinum, 1966). For example, Blackman and Templeman (1938) reported that shading to 61 and 44% of daylight did not affect the yield of Agrostis tenuis Sibth and Festuca rubra L. at low N fertilization. When N fertilizer was added, the yield increased dramatically in normal daylight, but not in shade. Deinum (1966) reported that DM and N yields were higher at reduced light intensities (100 cal cm day') when N was deficient. However, when yields of stubble and roots were included, total yields were equal at all light intensities, since the plants adapted to the lower light intensity by translocating a lower proportion of the assimilates to the roots. Increased shoot: root ratios under low light intensity have also been reported for both tropical and temperate grasses and legumes by others (Burton et al., 1959). Shading of grasses generally causes elongated growth, reduced specific leaf weight (SLW), and increased leaf area ratio (LAR), (Earley et al., 1966; Paulsen and Smith, 1969; Ludlow et al., 1974; Pyon, 1975). Plants grown in the shade are also more succulent, resulting in a lower DM percentage (Gordon et al., 1962). Nitrogenase activity (estimated by acetylene-reduction assay) has been measured in some rhizosphere associations in recent years (Schank and Day, 1977). However it is not known to what extent this N-fixation activity may be affected by light intensity. The N and mineral contents (dry weight basis) of grasses usually increase with shading in contrast to the legumes which are not greatly affected in this respect (Bathurst and Mitchell, 1958; Burton et al., 1959; Gordon et al., 1962). Cunningham and Nielsen (1965) reported that total cations of several crop species 'Journal Series No. 2486 of the Hawaii Agric. Exp. Stn. This work is part o£ a dissertation presented by the senior author in partial fulfillment of the requirements for the Ph.D. degree in Agronomy at the Univ. of Hawaii and was supported jointly by the Hawaii Agric. Exp. Stn. and the USAID Contract ta-C-1207 (NifTAL Project, P. O. Box O, Paia, HI 96779). Received 12 Feb. 1980. 'Former research assistant and agronomist respectively, Dep. of Agronomy and Soil Science, University of Hawaii College of Agriculture and Human Resources, Honolulu, Hi 96822. Present address of senior author: Mailing Agricultural College, 8340 Malling, Denmark. 2 AGRONOMY JOURNAL, VOL. 73, MAY-JUNE 1981 were positively correlated with N content. Myhr and Saebo (1969) and Mayland and Grunes (1974) reported similar results. Little work has been done on the effect of shading on tropical forage grasses and legumes. However, such studies could be very useful since there are extensive areas in the tropics where grazing could be profitably carried out under coconut palms (Hugh, 1972; Javier, 1974) and other plantation crops. Also total solar radiation is greatly reduced in many tropical areas due to seasonal cloud cover. The experiments reported here were carried out to assess the effect of shading on the performance of six tropical grasses fertilized with N (356 kg N ha yr) or given no N fertilizer. MATERIALS AND METHODS Six grasses were grown on an Oxic Haplustoll near Paia, Hawaii (20°55'N and 156°22'W, 100 m above sea level) during the period July 1975 to February 1977. Grasses which were replicated three times included: corigrass (Brachiaria miliiform1s (Presl.) A. Chase), Mealani hybrid digitgrass (Digitaria decumbens Stent.), and guineagrass (Panicum maximum Jacq.). Grasses which were not replicated included: signalgrass (Brachiaria brizantha Hochst. ex. A. Rich.), kikuyugrass (Pennisetum clandestinum Hochst. ex. Chiov.) and napiergrass 3418 (Pennisetum purpureum Schumach.). Each species was grown under two N levels: without N ("minus-N"), and with 365 kg N hayr ("plus-N") applied as urea broadcast after each cutting (once every 8 weeks) to provide 56 kg N ha. The plants were grown under four different light regimes: 100, 70, 45, and 27% of full daylight, by shading with three densities of polypropylene screening. The screening was stretched over the plots 1.9 m above the ground, to allow air circulation and easy passage underneath. The sides facing east and west slanted down at a 45° angle to 1 m above the ground, to shade the plants from morning and evening sun. Each shaded block measured 12.2 X 13.4 m. Before planting, phosphate and potash were applied at a rate of 150 kg ha of P as treble superphosphate and 180 kg ha of K as muriate of potash. Lime was applied at a rate of 3,400 kg ha to increase the pH to about 6.3. An additional 80 kg ha P was applied one year later and an additional 90 kg ha K was applied at 6 months and 12 months after establishment. Watering was by subsurface drip irrigation using polyethylene tubes (13 mm inside diameter) buried 15 cm deep and spaced at 90 cm intervals. Emitter spacing was varied so that treatments receiving 100, 70, 45, and 27% daylight received water in proportions of 100, 89, 72, and 50 in order to compensate for differences in evapotranspiration. Three replications of a split-split plot arrangement were used with light regimes as whole plots, species as sub-plots, and N levels as subsub-plots. Each sub-sub-plot measured 2.0 X 2.6 m. The total area harvested for yield determination per plot was 2.7 m'. Gumeagrass was established from seeds and the other grasses were established by vegetative propagation. The experiment was established on 27 July 1975, shaded 2 months later, and the speties harvested every 8 weeks during the period November 1975 to February 1977. All species were cut at 4 to 7 cm above the ground using a small sickle-bar mower, except for guineagrass (cut at 15 cm) and napiergrass (30 cm). Measurements taken included: 1) green weight; 2) dry weight measured by drying the samples at 55 C to constant weight; 3) N content of the dried and ground samples by Kjeldahl digestion (0.5 g sample) followed by determination of NH4 using an automatic analyzer; 4) mineral content (P, K, Ca, Mg, Si, S, Cu, Zn) for the 9 March, and 25 August harvests by X-ray fluorescence quantometer; 5) sward height measured after 8 weeks regrowth on 5 Mar. and 24 Aug. 1976; 6) area:dry weight ratio of the foliage (leaves plus stem) and leafplus-stem area index at the 25 Aug. 1976 and 9 Feb. 1977 harvests based on the projected area of three weighed subsamples of chopped fresh forage per lot using an optical planimeter (Davis et al., 1966, modified); 7) nitrogenase activity (measured as acetylene reduction) of undisturbed soil cores (17 cm diameter X 20 cm depth) incubated for 24 hours with 5% acetylene (95% air) and subsequently measured for ethylene production by gas chromatography; 8) root dry weights based on a sampled soil volume of 30 x 30 cm in each plot at completion of the experiment. Results were analyzed statistically using split-split plot analysis of variance and the Bayes L.S.D. The results from the nonreplicated species were adjusted for replication effects using the procedures for augmented designs described by Federer and Rag havarao (1975). RESULTS AND DISCUSSION