EFFECTS OF TREE ROOT COMPETITION ON AVAILABILITY OF SOIL AND PLANT NUTRIENTS, SOIL WATER AND LIGHT INTERCEPTION IN HEDGEROW INTERCROPS WITH DIFFERENT TREE SPECIES IN THE MID-COUNTRY INTERMEDIATE ZONE

This investigation determines the inter-species differences in root competition exerkd by six tree species (Cr~lliun.rlru cc~loih.yrsz~s, Desm,ocliurrt, rar~so~zii, Flen~,ir~girr :o~agrstu., Gliricirliu scpiur~r, Cussiu. spectabilis and 7lfh.ol7.i~ dir~ersifolio) on mung bean (Vigrr.u rudiafa) growing as hedgerow ~ntercrops in the mid-country intermediate zone of Sri Lanka. All intercrops had significantly lowe~.plailt nutrient (N, P, K) contents than sole mung heail indicating significant competition hy hedges for nutrients. This was conirmed by increased plant nutrient contents when a 1.5 m deep ti:ench was cut between hedges and crop rows. Competition was lowest by Gliric~rlicl and highest by Cussiu. This was confirmed hy soil nutrient analysis a t sowing and 11al.vesting of mung bean . All intercrops showed increases of exchangeable soil K and available P (except with Cliricidiu) during the cropping season. The fraction of incoming radiation interccpted was higher near hedges than in mid-plot indicating competition for light whicl~ was highest hy Ti.thon.iol (greater canopy spread) and lowest by Desnrodiunr and Flerrt.i~7.gin (lower canopy spread). Soil water content (SWC) was greater in sole mung bean than in intercrops indicating root competition for water. This was confirmed hy increased SWC wit11 trenching. Competition fol. water was mitigated to some extent by soil wate~, conservation oio the shading effect of canopics such as Tif.lron.ia. Based on the above results, it is concl~~ded that tree roots ofhedgerow intercrops exel-ted significant competition with the annual crop for ahsorption ofnutrients and water. In the present environment, out of the tree species tested, Glir.icirlicr. exerted the least competition for n ~ ~ t r i e n t s while Titl~,onic~. exerted the lcast competition for water This information could be used in the selection of tree species for hedgerow intercropping in thc mid-count~y intermediate zone of Sri Lanka.


INTRODUCTION
Competition for above-and below-ground resources is an unavoidable phenomenon in simultaneous agroforestry systems.'-Torlette t al.%nd De Costa and Chandrapala" have shown that interception of incident radiation by hedgerows reduced the light available for the annual crop in the alleys.Moreover, Singh e t aL7, Szott et at?, Corlett et al.%nd De Costa and Chandrapa1a"ave specifically shown that absorption of soil nutrients and water by tree roots reduces the availability of these growth resources to the associated annual crop.However, there is scope for minimizing competition and consequent crop yield reductions by careful choice of --* Col~esponding author tree species.In a previous paper, De Costa and Chandrapala4 reported significant variation between different hedgerow tree species in their root and shoot competition exerted on mung bean (Vigna radiata (L.) Wilczek) in hedgerow intercropping (HI) systems.The general objective of this investigation was to determine the inter-species variation in root competition for nutrients and water and shoot competition for light and thereby identify tree species suitable for HI in the mid-country intermediate zone of Sri Lanka.The specific objectives were : (a) To determine the effects of root competition (C,) from different tree specjes on plant a n d soil nutrient contents and soil water contents; and (b) To determine the inter-specjes variation in competition for light by examining radiation interception.

METHODS AND MATERIALS
The experiment was carried out under rainfed conditions from June to October, 1998 at Pallekelle in the mid-country (367 m above mean sea level) intermediate (rainfall 1400 mm yr-') zone (IM,) of Sri Lanlra (7-8"N & 80-81°E) on a moderately well-drained soil belonging to Rhodudults.Adetailed description of the experimental procedure is given in De Costa and C h a n d r a ~a l a .~ Only the essential details are given here.
Measuremen.ts: N , P and K contents in mung bean were measured 50d after sowing (i.e. a t 50% flowering stage).Plant samples were taken a t 30, 180 a n d 360 cm from hedges and their nutrient contents were analyzed separately.Total N was measured on samples digested with Conc.H,SO, acid using the micro-Kjeldahl method.Tota1 P was measured colorimetrically on ashed samples which had been digested with nitric acid.The dry ash methodlo was used to measure plant K content.
Major soil nutrient contents (i.e. total N, available P and exchangeable K) and soil pH were measured a t the beginning and end of the cropping season.llSoil samples were taken from both top soil (0-10 cm depth) and sub-soil (20-40 cm) at 15, 165 and 345 cm from hedges.Available soil N was extracted by KC1 and was determined by the micro-Kjeldahl method.Available P was extracted by NaHCO, and was determined colorimetrically.Available K was extracted by ammonium acetate and was determined by the flame photometer.Soil water content in top (0-10 cm) and sub (20-40 cm) soil was measured gravimetrically at 2 week intervals from sowing up to 10 weelrs after sowing of mung bean.Samples were talcen at 15 and 200 cm distances from hedges.The fraction of incomlng radiation intercepted (0 by the combined canopies of hedges and mung bean was measured in trenched treatments using a pair of tube solarimeters a t 2 week intervals from sowing up to 10 weeks after sowing.Measurements off were made a t 15 and 200 cm from hedges.During each measurement one solarimeter was kept in the open to measure incomjng radiation while the other was kept a t ground level inside t h e plot a t specified distances from the hedge.Voltage output from solarimeters were recorded and integrated by microvolt integrators.All measurements were made around mid-day (i.e. from 1000 1 1 to 1400 h) and each measurement was done for a period of 15 min.

RESULTS
Plant nutrient contents: Plant nutrient contents did not show significant (p<0.05)differences or consistent variation with distance from hedges (data not shown).Therefore, the respective nutrient contents were averaged across different distances.
Trenching significantly (p<0.001)increased N content in mung bean grown with all tree species (Table 1).Mung bean under Gliricidia and Cassia had the highest and lowest N contents respectively in both TR and NTR.Only the mung bean with Cassia.andFlem.in.gia had significantly (p<0.05)lower N than the sole crop control under both trenching treatments.The rest had plant N contents whjch were not significantly (p<0.05)different from that of the control (Table 1)'.P contents of mung bean in all HI were significantly (p<0.05)lower than that of the control in both trenching treatments.However, differences between tree species were not significant at p=0.05.Although, mung bean under the TR had higher P than the corresponding NTR except in the crop with Desnzodiunz (Table I), the difference was not significant (p=0.05).Different hedge species and trenching had highly significant (p<0.0001)effects on plant K contents.111 both TR and NTR, mLmg bean with all tree species had significantly lower K than the sole crop.Among HI, mung bean with Gliricidia and Cassia had the highest and the lowest K (Table 1).
TR significantly increased the K content of mung bean with all hedge species as compared t o NTR.Initial soil nutrients: Initial soil N, P and K showed highly significant (p<0.0001)variation between different hedgerow species (Table 2).Experimental plots which were under the sole crops in the previous season (and designated to have sole crops in the present season also) had significantly greater initial soil N, P and K than all HI plots except those with Gliricidia.Plots with Gliricidia showed significantly greater initial soil nutrients than even the sole crop control.On the other hand, initial soil nutrient contents did not differ significantly (p=0.05) between plots designated to have TR and NTR treatments.Initial soil nutrients did not differ

Table 1: Major nutrient contents of mung bean grown with different hedgerow tree species with (TR) and without (NTR) trenching
Phosphorus Potassium n e e (mg/100 g dry wt.) (mgIl00 g dry wt.) specles

Cu.11 iandra
Note: Each value is t h e mean of measurements a t three distances from hedgerows (30, 180, and 360 cm) in three replicate plots of each species x trenching treatment cornhination.Note: Each value is the mean over two soil depths and three distances horn the hedgerows.See text for hrtller explantation.
significantly (p=0.05) between top and sub-soil layers or between hfferent distances from hedges (data not shown).
Soil nutrient contents at harvest of mung bean: Soil nutrient contents at harvesting of mung bean varied significantly with different tree species, trenching I treatments, soil depth and distance from hedges.In order to retain clarity of presentation, effects of tree species and trenching are given in Table 3 and those of soil depth and distance from hedges a r e shown in Table 4 .This separate presentation was possible because there was no significant (p<0.05)interaction between the two above sets of factors.Note: Each value is the mean ovel.two soil depths and three distances fiom the hedgerows.See text for further explanation.
In both TR and NTR treatments, soil N at harvest was significantly (p<0.01)lower than sole crop in all HI except those with Gliricidiu (Table 3).While trenching increased final soil N in all HI as compared t o the respective NTR treatments, the increases were most prominent in HI with Desmodium, Tithonia and Calliandra.The overall effect of trenching on final soil N was highly significant (p<0.0001).Similar to the pattern shown by soil N, the final available soil P was significantly (p<0.05)lower than in the sole crop in all HI except those with Gliricidia in both trenching treatments (Table 3).Trenching significantly (p<0.0001)increased the final soil P in all HI with the greatest increases in HI with Gliricidia and Calliandra.Final soil exchangeable K showed a different pattern to those of N and P with all HI having significantly (p<0.0001)greater final soil K than the sole crop (Table 3).both trenching treatments, HI with Gliricidia had significantly greater final soil than the rest which did not differ significantly (p<0.05)among themselves.Trenchi significantly (p<0.001)increased the final soil K in all HI.
Soil N a t harvest of mung bean increased significantly (p<0.0001)wi increasing distance from hedges in both top soil (0-10 cm depth) and sub-so (20-40 cm depth) layers (Table 4).However, the respective final.soil N did n differ significantly (p<0.05) between the two soil layers.Similar to N, both fin soil P and Kincreased significantly (p<0.001) with increasing distance from hedg in both soil layers.In addition, the respective P and K were significantly greater top soil as compared to sub-soil.Changes in soil nutrients during the cropping season: Table 5 shows t E changes i11 major soil nutrients during the cropping season.In TR treatments, I with Desmodium and Flemingia showed increases in total soil N whereas the re of HI and the sole crop showed decreases in soil N. Soil N depletion under TR w highest in HI with Gliricidia and least in HI with Cassia and Callian.dra.Und NTR, all HI except that with Flenzin.giashowed depletion of soil N which was great than the respective depletion under TR.Here also, the highest N depletion w shown by HI with Gliricidia and the least was by HI with Cassia.In contrast to : soil K showed an increase in all HI and the sole crop under both TR and N'l treatments (Table 5).Moreover, all HI had greater K enrichment than so crop under both trenching treatments.On the other hand, trenching slight increased the respective soil K enrichment in all HI.Soil P also showed ; enrichment during the season in all HI except those with Gliricidia.In contrast, there was a soil P depletion in sole crops.Trenching increased P enrichment in all HI where there was an enrichment and decreased P depletion in the rest.Highest P enrichment was shown in HI with Cassia (in TR) and Gliricidia (in NTR).

Radiation interception:
The fraction of incoming radiation intercepted (R), at both 15 (R,,) and 200 (R,,,) cm distances from hedges, showed significant (p<0.0001)variation between different tree species on all dates of measurement (Fig. 1.a & b) with the exception of R,,, a t 6 weeks after sowing (WAS).The inter-species variation was clearer a t 15 cm from hedges than at 200 cm.Except for HI with Flentingin and Desmodium, all, other HI had significantly (p<0.05)greater R,, throughout the duration of measurement.HI with Tithonia had significantly (p.cO.05) greater R,, than the rest from 6 WAS 0nward.s.HI with Cassia., Gliricidia and Callia,n,dra also had significantly (p<0.05)greater R,, than the sole crop throughout the measurement period.In general, R,,, was lower than R,, in all HI throughout the season.However, inter-species variation of R,,, (Fig. 1.b) was less consistent as compared to R,, (Fig. 1.a).During the first 4 weeks, sole crops had significantly (p<0.05)greater or equal R,,, than all HI.At 6 WAS, R,, of all HI did not differ significantly (p<0.05) from the sole crop.From then onwards, HI with Tithonia had significantly (p<0.05)greater R,, than all other crops includjng the control.Among the rest of HI, those with Cassia, Gliricidia and Calliandra had significantly (p<0.05)greater R,,, than those with Flenzin.gia and Desmodir~m..   Soil water content (SWC): Seasonal variation of SWC in top and sub-soil layers with (TR) and without (NTR) trenching is shown in Fig. 2.There was a common overall pattern with SWC increasing up to 4 WAS and then decreasing until.harvest of mung bean.The substantial rainfall that occurred during the initial 4 week period after sowing (as shown in Table 1 of the companion paper*) was responsible for the increase in SWC up to 4 WAS.Anotable feature in the seasonal variation of SWC in both HI and sole crop plots was t h a t the decrease of SWC jn top soil was substantially reduced during the final 2 week period from 8 to 10 WAS (Fig. 2.a & b).This was probably because of the decrease in water use by mung bean during this final phase.In contrast, soil water extraction from the sub-soil layer during the final fortnight continued at almost the same rate as before.Except a t 2 WAS, there were highly significant (p<0.0001)differences between hedgerow species in SWC in both soil layers as well as in both TR and NTR (Fig. 2).In general, on most days of measurement in all four treatment combinations (2 soil layers x 2 trenching treatments), the sole crop had significantly greater SWC than all HI except Tithonia.On some days (eg. at 6 WAS in top soil), HI with Tithonia had significantly greater SWC than even the sole crop.In the top soil, HI with Callia,ndra., Gliricidia andFlemingia tended to have lower SWC under both TR and NTRwhereas HI with Cassia and Desmodiunt tended to have higher SWC.HI with Tithonia had significantly (p<0.05)greater sub-soil SWC as compared t o the rest of HI.On the other hand, HI with Flemir7.giahad-significantly lower sub-soil SWC on most days under both TR and NTR.Although not apparent in Fig. 2, trenching increased SWC in both top and sub-soil layers.
Table 6 shows the effect of trenching and distance from hedgerow (D) on top and sub-soil SWC.Trenching significantly (p<0.0001)increased SWC in both top soil (except a t 8 WAS) and sub-soil (except at 6 WAS).Except at 6 WAS, D had a significant (p<O.05)effect on top soil SWC in both TR and NTR.At 2,8 and 10 WAS, top soil SWC was significantly higher closer to the hedgerow (at 15 cm) whereas the opposite was observed a t 4 WAS (Table 6).In contrast to top soil, D did not have a significant (p<0.05)effect on sub-soil SWC on 2,4 and 6 WAS.However, at 8 and 10 WAS, sub-soil SWC was significantly greater at 15 cm than at 200 cm in both TR and NTR.

DISCUSSION
Plant and soil nutrient contents: Comparison of p.lant nutrient contents between sole crops and HI indicated whether the presence of hedges increased or decreased the availability of respective nutrients to mung bean.This experiment clearly showed that hedges decreased the availability of all three major nutrients (N, P & K) to mung bean.The fact that this was due to competitive absorption of nutrients by hedgerow roots was confirmed by the trenching treatment (TR) which showed increased nutrient availability as compared to the non-trenching treatment (NTR).Although competition for N and K could be removed quickly by trenching (as shown by significant increases in TR), results of this work indicate that competition for P may not be readily removed by trenching (as shown by small, statistically non-significant increases in TR).Gunasena e t aL.12 also observed that HI decreased the availability of available P in both the top (0-10 cm) and sub-soil (10-40 cm) layers.However, these observations contradicted the findings of Haggar e t a1.l" Hands e t al.14 and Hulugalle and Ndi15 t h a t HI increased plant available P. Differences between hedge species clearly indicated that competition for nutrients (especially for N and K) was least by Gliricidia and highest by Cassia.In fact, the variation in plant nutrients observed in our experiment is due to the net result of two opposing effects, i.e. the competitive removal of nutrients by hedgerow roots and addition of nutrients through hedge prunings.Therefore, the high level of competition by Cassia could be either due to its higher root competition or lower nutrient additions through prunings.It is likely that both the above factors were responsible.Cassia had the highest above-ground biomass production among the tree species4 and therefore could be expected to exert greater competition for nutrients and also have'an extensive root system.lGRao et al.17 observed that the root system of Gliricidia in hedgerow intercrops was less extensive than that of Cassia, thus confirming our observation of greater root competition by Cassia.Moreover, as biological N, fixation does not occur in Cassia, it is likely to have less N in its prunings.
The above conclusions based on observed variation of pl ant nutrient contents were confirmed by soil nutrient contents.Greater removal of soil N, P and K by HI was confirmed by the significantly lower initial values of N and K m HI as compared to sole crops.Moreover, the lower competition from Gliricidia was confirmed by the significantly greater initial soil N, P and K contents in plots w t h Gliricidia as compared to other tree species.The lower final soil N and P contents in HI as compared to sole crops and increased final soil N, P and K contents with trenching further confirmed the root competition for nutrients in HI.The highest values of final soil N, P and Kin HI with Gliricidia further confirmed its lower level of competition for nutrients.
On the other hand, a notable finding of our experiment is the ability of all hedgerow species to enrich soil K during the course of the cropping season.A similar observation was made by Salazar c t a1.18 and in the previous season of this worls.which was not reported in De Costa and ChandrapalaG because of the need to confirm it.This increase in soil K could have been caused by the release of K due to fine root turnover during the cropping season and especially at its beginning when hedges were pruned.Moreover, some of the Kfixed within clay minerals could have been released to the soil solution during the cropping season1" and this process could have been facilitated by hedge root systems.However, significantly greater plant K contents in sole crops than i n HI, show t h a t despite the increased availability of K in the soil, mung bean in HI faced competition from tree roots during absorption of K.This was confirmed by the significantly greater K, in both plants and soil at harvest, in trenched plots.Here again, Gliricidia showed the lowest competition.The observation of Schroth and Zech2" and Rao et al." that Gliricidia has a lower capacity for root production as compared to most hedgerow species could explain its low competition in the present experiment.Further evidence to demonstrate the significant root competition from hedgerows for all major nutrients is provided by the observations of increased availability of N, P and K with increasing distance from hedgerows.
Another notable observation was the absence of significant differences in plant nutrient contents a t different distances from hedgerows despite signifjcant variation in soil nutrient contents with distance from hedgerows (Table 4).This was probably because the hedgerow roots had spread the entire width of the alley between the two hedges and were competitively absorbing more nutrients from the soil.Recently, many workers" 21s2%ave observed that after a few years from establishment, the hedgerow roots traverse the whole width of the alley.If this occurs and if the hedgerow roots can competitively absorb more nutrients, it is not surprising t h a t the plant nutrient contents within the alley do not vary significantly with distance from the hedgerow.

Radiation interception:
Results on inter-species varlation of t h e fraction of radiation intercepted (R) showed that hedges exerted significant competition for light as shown by the greater R at 15 cm from the hedge (where both the hedge and crop would ~ntercept radiation) than a t 200 cm (where only the crop would ~ntercept radiation).As all hedgerow intercrops grew taller than mung bean, it is highly likely that a greater proportion of the increased R,, cm was being intercepted by hedge rather than crop.This was confirmed by the highest R,, in HI w ~t h Titho7zzu (which had the 'most,-spreading' canopy) and t h e low R,; in HI with Flcrningicr and Desmocllnr~~ (which had the 'least-spreading' canopies).Corlett et nL.5 also observed greater shading of annual crops when canopies of hedges were allowed to spread by less frequent pruning Soil water contents (SWC): The generally higher SWC in sole crops than in HI indicated that there was significant competitiori for soil.water from hedge roots.This was confirmed by the increase of SWC with trenching.Similar positi.veeffects were observed with root barriers by Ong et al.": and Rao et nl."Hauser and G i c l ~u r u ~~ also observed significant competition for water by hedges of Cassia and Acioa..However, hedges had a positive effect in conserving soil water, especially in top soil near hedges by their shading effect which probably minimized soil evaporation and crop transpirat,ion.This is supported by the significantly greater SWC observed in 1-11 wit11 2Ithon.i~which probably exerted the highest shad.ingeffect due to its spreading canopy.Chirwa et al."j also observed a positive effect of HI which had a 1.ower soil water suction than sole crops.This worlr clearly showed t h a t tree roots of hedgerow iritercrops exerted signify cant co~npetjtion with the annual crop for absorption of rlut,rients and water.However, tree species differed in the degree of competition exerted.In the present environment, out of'the tree species tested, Gliricidin exerted the least conlpetjtion for nut,ri ents while Tit11.oniaexert,ed the least competition for water.

Note:
Each value is the mean of6 tree species (plus control) and 3 replicates.Sig~ficance (or otllerwise) of' mean compai~isons hetween distances within each trenching treatment is shown.by' (or ns).Significance (or otherwise) of mean comparisons hetween trenching treatments within each distance is shown by -I (or 11,s).

Table 2 : Soil nutrient contents at sowing of mung bean grown with different hedgerow tree species with (TR) and without (NTR) trenching Total N (76)
(X lo.2)

Table 4 : Soil nutrient contents at final harvest of mung bean at differen distances from hedgerows in top soil (0-10 cm depth) and sub-soil (20-40 cm layers
Note: Each value is the mean over six tree species and two trenching treatments.See text fir fusth explanation.

Table 5 : Changes in soil nutrient contents during the cropping season of mgng bean grown with different hedgerow tree species with (TR) and without (NTR) trenching
Nutc: Each value is t,he difference Ijetween the ~espective values at harvest; (Tahle 3) and s o ~i r ~g (Tahle 2) of mung bean.A negal;ive value indicates nutrient depletion and a 1)ositive value iildicates , nutrient enrichment

Table 6 : Seasonal variation of soil water contents (% dry weight basis) at different distances from hedgerows in top soil ( 0-10 cm depth) and sub-soil (20-40 cm) layers for treatments with (TR) and without
(NTR) trenching.