Acta Amaz. - vol.16

THE INTERCEPTION PROCESS IN TROPICAL RAIN FORESTS: A LITERATURE REVIEW AND CRITIQUE (*) .

R. T. Clarke (**)

SÜMMARY

A noviw oi tht tLtQJioJuAn on m&aAuKem&ntò o{ naln mtzn. íntzucdptÁon ptiocçAòzA

by ^oheAtis i& madt. ln{)OtwiaXÁ.on on A^nÁca, C2.ntn.0l and South AmztvLca, and A&ia ÃJ> QÒJ&I.

A gene.Aal anatgòLt, lò made. and thz me.d to kun.tho.h. tko.. mimuAomuntò undoA ^toXd condÁ.

-txonò ÍÁ òtAQAòzd.

(*) Trabalho apresentado no "International Workshop on Precipitation and Water Recy - cling in Tropical Forests", 10-13 de novembro de 1981 - CENA.

(**) Institute of Hydrology Wallingford, Oxford, England.

ACTA AMAZON ICA,

)6/17 (n? único):

1986/87.

BACKGROVm

In temperate latitudes, interception as a component of the total evaporative loss

from forests has been intensívely studíed (see, for example, Helvey & Patrick, 1965;

HeJvey, 1 9 6 7 ; Lawson, 7967; BJake, 1975; Calder, 1 9 7 6 , 1 9 7 8 ) and its importance welj_

established. Field studies of the interception process have given results which, in

rnany cases, have been expressed as empirical regression equations of the form I = aP + b

G

where I is the depth of water intercepted and lost by evaporaiion, P. is the gross rain

b —

fali incident upon the forest canopy, and a, b are regression coefficrents; see, for

example, Penman (1963)- Many authors, however, Blake (1975) being one, expressed the

relation in the alternative form (l/P„) = b / P + a expressing the fact that the

percen-G percen-G

tage interception loss tends to be hyperbolically related to gross rainfall. In con

-trast to this empirical approach, Rutter et ai. ( 1 9 7 0 , Gash & Morton have constructed

physically-based computer models using as inputs rainfal? and the meteorológica!

varia-bles controlling evaporation to compute a running water balance of a forest canopy with

known structure, thereby producing estimates of interception losses. More recently,

Sellers & Lockwood (1981) have developed a multilayer crop model for the simulation of

interception loss and crop transpiration which, they consider, represents a considera

-b)e improvement in terms of physical realism over the Rutter model; Seilers and Lockwood

also suggest that the Rutter model substantial1y underestimates the interception from

a pine forest during and following storms of low intensity. Shuttleworth and Gash (1982, in press) have suggested that this difference between the Rutter and Sellers & Lock -wood models is likely to have arisen from d i f f e r e n c e s in the formulation of the aerody­

namic transfer m e c h a n i s m .

Models like those of Rutter and S e l l e r s & L o c k w o o d , have pratical disadvantages in that they require hourly meteorological d a t a , w h i c h are seldom a v a i l a b l e , and a complex computer prog ram ,which is time c o n s u m i n g to construct and o p e r a t e . A l t e r n a t i v e m e t h o d s o f assessing interception losses w h i c h a r e less demanding of computet ional effort

have been developed by Gash ( 1 9 7 9 ) ; Gash's tr,odel retains some of the s impl i ci ty of the regression approach described a b o v e , w h i l s t including much of the fundamental physical reasoning implicit in the Rutter m o d e l .

Over the past d e c a d e , process studies such as those of Stewart (1977) have resulted in greatly increased understanding of the d e p e n d e n c e of interception loss on climate and canopy structure (eg Gash e t a ) . , 1980) such that q u a n t i t a t i v e predictions of the effects of forests on w a t e r resources have been possible (Calder & N e w s o n , 1979) . Whilst studies o f the interception process in temperate forests are fairly plentiful and w e l l -d o c u m e n t e -d , information about the interception c h a r a c t e r i s t i c s of tropical forest is much more fragmentary despite the fact that such forests c o n s t i t u t e a major natural

resource of those countries in which they lie (Klinge et a l . , 1 9 8 1 ; L a i , 1 9 8 1 ) . The purpose of this paper is to d r a w together some of the fragments and to fit them together to see what p i c t u r e , if a n y , e m e r g e s ; it begins by a s s e m b l i n g the e v i d e n c e on a regional b a s i s . Units of quantities given in the paper are h e t e r o g e n e o u s but a re as given in the papers quoted.

AFRICA

Nye ( 1 9 6 1 ) reported the results of a study in rr.oist tropica) forests near K a d c , Ghana, The forest was m a t u r e secondary growth with a w i d e range of species;at its base w a s a fairly open shrub layer underlying a dense lower storey between 15 and 50 f e e t , with a scattered upper storey extending to 130 feet. Occasional m a s s i v e - c r o w n e d e m e £

gents extended even h i g h e r . Basal area at breast height w a s ]k$ f t?/ a c r e .

Annual rainfall in the region is 65 inches a y e a r , w i t h a long-rains season from March to J u l y , and a short-rains season from September to D e c e m b e r . In Nyr.'s study ,the

throughfall penetrating the canopy w a s measured in 6 (later 8) 8-inch c o p p e r g a u g e s ; three other gauges of the same type w e r e sited in c l e a r i n g s , to give an e s t i m a t e of gross r a i n f a l l . Initially, gauges w e r e read w e e k l y , but later daily.

In the year 1 95 8- 59 , the measured g r o s s preciρitation , , was 7 2 , 7 inches falling in I 6 7 d a y s , w i t h throughfa 1 1, T , o f 6 1. 5 inches. Initially, siemflow w a s a l s o measured for a two-month p e r i o d , but since less than o ne per cent of gross rainfall w a s found to flow dawn trunks its measurement w a s d i s c o n t i n u e d . Wye's estimate of interception loss, I,over the year was therefore 16¾; this figure is subject to some e r r o r . h o w e v e r , b e c a u s e the collector vessels used for measuring throughfall were small and it is known that throughfall is highly v a r i a b l e both spatially and in time. Wo information was given

about the size of t h e clearings in w h i c h w e r e sited the three gauges giving e s t i m a t e s of above-canopy p r e c i p i t a t i o n , nor of their v a r i a b i l i t y in c a t c h .

Hopkins ( i 9 6 0 ) g i v e s results from a study in the Mpanga Forest R e s e r v e . U g a n d a , at an altitude of 1160 m w h e r e mean annual rainfall is 1 1 3 0 m m . Besides giving a compre -hensive botanical description of t h e f o r e s t , Hopkins m e n t i o n s that it possessed both an understorey and an incomplete upper storey w i t h lianas a b o v e 30 m . A 1 2 0 ft tower extended into the c a n o p y , and three 7.6 cm glass-funnel gauges w e r e sited a b o v e the canopy, extending from three of the four corners o f t h e tower at k heights ( 3 0 f t , 6 0 f t , 90 and 1 2 0 f t ) ; a second set of four gauges - a l s o glass-funnel type - were m o u n t e d at ground level .

Hopkins did n o t m e a s u r e s t e m f l o w , a l t h o u g h he remarks that the botanical structure of the herb-layer understorey would result in channelling along plant s t e m s . H e recor-ded gauge catches during six intervals of t h e main w e t s e a s o n , concluding that about 35¾ of the rain falling on the forest is intercepted before it reaches the g r o u n d ; almost all of this interception took p l a c e between ground-level and 9- 2 m h e i g h t , occurring because of the herb layer.

O n e point about w h i c h Hopkins m a k e s no comment is that the uppermost ( 1 2 0 ft) gauges on the tower appeared to catch 1 1. 2 per cent m o r e than the " o f f i c i a l " g r o u n d l e -vel raingauge at M p a n g a , some 300 m t o the e a s t ; n o m e n t i o n is g i v e n , e i t h e r , of the variability amongst the three gauges sited a t each level o f the t o w e r .

In a later study on the o t h e r side of t h e c o n t i n e n t , Hopkins (1965) set o u t ten ground-level glass-funnel gauges o f 7.6 cm d i a m e t e r beneath a stand of the Olokomeji Forest R e s e r v e , N i g e r i a . M e a s u r e m e n t s w e r e recorded for the period O c t o b e r 1965 to February 1969 and c a t c h by these ten g a u g e s w a s compared w i t h catch by a gauge at t h e Olokomeji Forest S t a t i o n . N o actual m e a s u r e m e n t s a r e g i v e n , but H o p k i n s states that throughfall w a s recorded as 9 7 ¾ . A g a i n , n o m e a s u r e m e n t s o f s t e m f l o w w e r e made,a1 though presumably this component m u s t have been very s m a l l . Hopkins comments on t h e variabi -lity in catch o v e r short d i s t a n c e s by his ten ground-level g a u g e s , supporting this statement by some q u e s t i o n a b l e statistical s i g n i f i c a n c e (_t) tests in w h i c h each gauge is compared w i t h every o t h e r . In some i n s t a n c e s , a l s o , his g a u g e s w e r e found to b e full so that t h r o u g h f a l l , although h i g h , m a y have been u n d e r e s t i m a t e d . Taking account of this possible underestimation and the a b s e n c e of s t e m f l o w m e a s u r e m e n t , a figure of 97¾ for throughfall must be open to some d o u b t .

A l t h o u g h their results may have doubtful relevance t o w o r k on t h e A f r i c a n conti -n e -n t , Vaugha-n & W i e h e ( 1 9^ 7 ) studied the i-nterceptio-n c h a r a c t e r i s t i c s of the Upla-nd climax forest of M a c a b é , in the interior of M a u r i t i u s ; here t h e mean annual rainfall is about 3 1 7 5 m m , "frequently varying by i 750 m m " . V a u g h a n & W i e h e used o n e raingauge at an " e x p o s e d " site in a clearing o u t s i d e the f o r e s t ; about half a m i l e away,and within the f o r e s t , three 4-inch raingauges w e r e installed at random on the forest floor,about 10 yards s p a r t . For the y e a r 1 9 3 9 , catch by the exposed g a u g e w a s 3 0 9 ^ m m , w h i l s t c a t c h in the forest gauges w a s 2 3 7 5 , 1 7^ 5 and 2 3 3 6 m m . Direct rainfall on the forest f l o o r , the authors s t a t e , w a s found to be almost two-thirds of the total e x p e r i e n c e d by the

exposed s t a t i o n . They also m a k e the interesting comment that "relatively m o r e rair reaches the Forest floor during the warm months of heavy downpours than in the cooler months when ihe rain is less heavy and the wind velocity lighter".

With only three forest-floor gauges of small d i a m e t e r , the precision of Vauqhan :i W i e h e ' s estimate of throughfall must be low; n e i t h e r was stemflow m e a s u r e d .

A more recent study is that of Jackson (1971, 1975) working in the West Usambara mountains of northern T a n z a n i a , T h e f o r e s t , described as an Intermediate Evergreen Forest CoT.munity, lies between 1300 and 18 0 0 rn a l t i t u d e , and is of the most luxuriant type in East A f r i c a ; the 32- y e a r mean annual rainfall is 1338 m i . Jackson descr i ties the botanical composition of the f o r e s t , w h i c h is very rich in s p e c i e s . Throughfall was measured using 20 standard 5-inch gauges randomly located on an 80 ft χ 80 ft square grid; steinflow was measured by spiral gutters of fabric and b i t u m e n , attached to 20 trees. Gross rainfall w a s measured by means of 5 raingauges sited in a clearing some 70 rn s q u a r e , the e d g e of w h i c h w a s 45 m from the canopy n e t w o r k , J a c k s o n studied the

interception losses for different storm s i z e s , and for both short and longer p e r i o d s , Stemflow w a s s m a l l , of the o r d e r o f 1¾ o f gross r a i n f a l l , and catches by h i s t hroughfa 11 gauges very v a r i a b l e . Over a period April to September 1 9 6 9 , measured interception loss, for a total gross precipitation of 839 m m , was only 4 . 2 ¾ ; h o w e v e r , Jackson argues that interception losses for large storms ar e poorly d e f i n e d , a n d , using a statistical a n a l ^ s i s , d e r i v e s a " c o r r e c t e d " figure of about 16¾.

The interception estimates o f Pereira ( 1 9 5 2 ) , following work by W i m b u s h ( 1 9 4 7 ) ,a r e amongst the most c o m p r e h e n s i v e , but are to some extent atypical In that they refer to tropical forests at an a l t i t u d e of about 8,700 ft, Pereira worked w i t h a 20- y e a r olrt plantation o f Montieroy C y p r e s s (C. m a c r o c a r p a ) and w i t h nearly natural b a m b o o forests

(Arundinaris a l p i n a ) . A total of 41 separate gauges was distributed in a l-acrc plot of each c o v e r , measuring "rainfall p e n e t r a t i o n , drip from the c a n o p y , and runoff down the s t e m s " ; throughfall and s t e m f l o w , although evidently both measured in Lottf1 ,were not discussed separately. For the six-year period 1 9^ 6 - 5 1 , and including the e x c e p t i o n a l l y dry year 1 9 4 9 , Pereira's results were as in T a b l e I,

Both in terms of length of record and intensity o f s a m p l i n g , t h e r e f o r e , Pereira's results d o m i n a t e ; h o w e v e r , they cannot be c o nsi d ere d typical of lowaltitude and natu -ral1y-occurri nq T A B L E 1 Y e a r ; 1946 1947 1948 1949 1950 1951 P„ ( i n c h e s ) : 41.01 60.60 45,43 2 5 , 5 2 4 1. 5 2 6 0. 1 8 ci Interception by c y p r e s s : 1 7. 2 1 9. 4 2 2 . 7 3 2 . 0 28.6 2 9. 0 Interception by b a m b o o : 13-4 1 4. 5 2 1. 5 2 5 . 8 2 1. 5 2 7 . 2

f o r e s t s , whilst m e a s u r e m e n t of stemflow for a stand of bamboo w o u l d have been fraught with difficulty,

CENTRAL A N D SOUTH A M E R I C A

No systematic attempt has been m a d e to explore the Spanish and Portuguese journals for material a p p r o p r i a t e to this p a p e r ; h o w e v e r , it is known that the UNDP/WMO Project B R A / 7 2/ 0 1 0 "Hydrology and Climatology of the Brazilian Amazon River B a s i n " , which uses personnel from the Brazilian agency SUDAM and in which many other agencies co-operate (Da Silva N e t o e t a ] . ,1 9 7 9 ) collects n o observation of forest intercept ion losses (Basso, personal c o m m u n i c a t i o n ) . This o m i s s i o n is r e m a r k a b l e , since "Sao objetivos adicionais do e s t u d o , as previsoes do tempo e de enchentes e m e l h o r a m e n t o no planejamento do uso da terra". Fragmentary information on interception by forests in Central and South A m e r i c a is then the following.

Hopkins (I960) quotes results from Freise ( 1 9 3 6 ) w h o worked in forests in Brazil which were termed " s u b t r o p i s c h e n " but w h i c h w e r e "probably not very different in

struc-ture from Tropical Rain F o r e s t s " ( R i c h a r d s , 1 9 5 2 ) . T w o l o c a l i t i e s , just over 500 km apart, were studied, and in each locality o b s e r v a t i o n posts w e r e set up in a forest c l e a r i n g , near the forest e d g e , and well w i t h i n the f o r e s t . At each forest post there were two rainfall g a u g e s , one 1 . 5 m above ground-level and the other attached to the top of the trunk of the highest tree from w h i c h the branches had been removed. T h e s e

rain-fall gauges w e r e recorded daily for 27 years in o n e locality and for 2 1 years in the other. Mean annual rainfall recorded at 1 . 5 rn above ground-level varied from 2 9 . 8 ¾ to 3^.8¾ of that recorded in the c r o w n s . When most of a month's rain fell a s thunder s h o w e r s , 2 6.7¾ of the crown rainfall reached the 1 . 5 m g a u g e , w h e r e a s when most of month's rain fell as long continued fine r a i n , 3 5 . 5 ¾ reached the ground.

No information has been found on the comparison of catches by gauges within the forest, with catches by gauges in the forest c l e a r i n g s . Freise's results also appear considerably a n o m a l o u s , h o w e v e r , by virtue of the very large interception losses that they imply, and by virtue o f the fact that interception losses appear to be less, not g r e a t e r , during period of long continued fine rain.

Heuveldorp (1979) has presented some results for the A m a z o n Ecosystem MAB Pilot Project at San Carlos de R i o N e g r o , n e a r the border between Venezuela and B r a z i1. A s p a r t of a larger study, funded by the German Research Foundation DFG and the A m e r i c a n N a t i o -nal Science F o u n d a t i o n , total rainfall is collected at an unstated number of canopy level g a u g e s ; throughfall is collected in four 6-metre long cut-open PVC t u b e s , w h i c h provide a total collecting area of 3 - 5 m2. Throughfall is also collected in two sets of 20 tins of unspecified d i m e n s i o n , one set of w h i c h is systematically distributed on a podsol s i t e , the other on laterite. The site ( 1 1 9 m a l t i t u d e ) is located near the c o n -fluence of Casiquiare and Rio Negro in T e r r i t o r i o Federal A m a z o n a s , and is typical of the black water areas draining towards the A m a z o n . T h e forest is described as "natural w e a k l y - s e a s o n a l , predominantly e v e r g r e e n , equatorial-tropical lowland f o r e s t " .

H e u v e l d o r p presents results in diagrammatic form for the period July 1 9 7 5 to September 1 9 7 7 ; as a guide to the rainfall r e g i m e , the annual total for 1976 (read from the g r a p h ) w a s about 3900 mm. Throughfall and drip from leaves in both stands averaged 87¾ of total incoming p r e c i p i t a t i o n : "a surprisingly large p r o p o r t i o n , especially as

stemflow, has not been i n d u e d . The amount of s t e m f l o w , h o w e v e r , is relatively s m a l l , hardly exceeding the error of o b s e r v a t i o n and about equal to the e v a p o r a t i o n of inter -ception w a t e r at the soil s u r f a c e " .

It is not clear w h e t h e r the PVC c o l l e c t o r s , or the t i n s , or b o t h , w e r e used In the measurement of t h r o u g h f a l l . T h e w i d t h of the PVC c o l l e c t o r (3.5/24) seems to have been about 14 c m , with the collector semicircular in cross s e c t i o n ; if this w a s the c a s e , splash could have resulted in an underestimation of throughfall,

Kline & Jordan ( 1 9 6 8 ) , as part of a study of tritIum movement in soi I of a t ropi cal r a i n f o r e s t , investigated a site in the El Verde forest on the Luquillo Mountains of Eastern Puerto R i c o . They g i v e no details of the nature of the forest c o m p o s i t i o n , but recorded throughfall at the forest floor by means of two raingauges. Canopy rainfall was measured by o n e standard tipping-bucket r a i n g a u g e . Over a 2 1 0- d a y p e r i o d , 1840mm of

rain w e r e recorded at canopy l e v e l , 370 mm at the forest floor; interception was there-fore 25'¾. No measurement of stemflow is m e n t i o n e d .

Also in the El Verde forest, Sol I ins S Drewry (1970) studied the water balance of so:ve individual tree c a n o p i e s . A catch funnel was mounted on a 92 ft t o w e r , well above the c a n o p y , to collect r a i n f a l l , w a t e r being routed to an 8-inch tipping.bucket at the base of the trees. Throughfall was also measured by two unmodified gauges (also presu-mably 8-inch) collecting the catch from 16 c o l l e c t o r s placed in a c i r c l e . Stemflow was

recorded by means of a 1 inch rubber hose attached around the tree.

i\ twelve-month study of one tree of the species Manilkara b i d e n t a t a (Ausubo) showed that of the incoming r a i n f a l l , about 5 0 Í penetrated the canopy as throughfall D a c r y o d e s excelsa (Tabonuco) and C r o t o n p o e c i l a n t h u s ( S a b i n o n ) , studied during the relatively wet summer m o n t h s , showed much less tendency to intercept rainfall (70¾ throughfat1,t¾ stem flow) d e p e n d i n g , h o w e v e r , on the intensity and duration o f the rain.For the M a n i l k a r a , there was also m u c h variation in monthly interception t o t a l s , w h i c h range from 30¾ to 30 of gross p r e c i p i t a t i o n .

Soil ins U Drewry also quote results obtained by Clegg (1963) w h o , in a 130-inch annual rainfall region of m o n t a n e forest (El Y u n q u e ) recorded an interception percentage of 5^¾ (varying from 42 to 77¾)·

Clements f. Colon ( 1 9 7 5 ) reported further results from the El Verde forest. In a 20 χ 20 m p l o t , six troughs (8 inches w i d e , 24 feet long, 10¾. slope) of ga 1 vanized meta 1 collected t h r o u g h f a l l ; gross rainfall was measured by a I m?- stainless steel collector

rr.ounted on a 72 ft tower. All trees w i t h dbh of 3 Inches or more (45 trees a l t o g e t h e r ) w e r e fitted w i t h stemflow c o l l a r s .

Clements c Colon calculated the regression of throughfall and stemflow o n gross rainfall for 1 3 7 storm e v e n t s over a 12-month period. Throughfall varied from 6 7 % f o r u storm of 0.1 inch to 92.5¾ for a I inch storm; stemflow as a percentage o f r a i n f a l l , varied from 1.4¾ for a 0.5 inch rainfall to 5.1¾ for a 1 inch storm. T h e g o o d n e s s of

fit of the linear regressions quoted by Clements & Colon cannot be assessed,and their diagrams show only the lines without any accompanying scatter of 137 e v e n t s , Depth of canopy storage was estimated a s 1.2 n«n (0,03 i n c h e s ) .

On the Central A m e r i c a n m a i n l a n d , Read (1977) carried out a m i c r o c l i m a t i c study near the site of the dam on the B a y a n o R i b e r , 70 km east of the Pacific entrance of the Panama C a n a l . T h e forest has been uncut up to 1 9 7 3 , but in the dry season January-April 1974 there w a s e x t e n s i v e clear c u t t i n g , followed in 197** and 1 9 7 5 by secondary growth of large-leafed and soft wood t r e e s . Gross rainfall w a s measured by o n e standard gauge above the c a n o p y , and throughfall w a s measured some 500 m away by means cf a heavy sheet metal trough, 1 0 m long and 10 cm w i d e ; s t e m f l o w , it a p p e a r s , w a s not m e a s u r e d . For the year May 1973-April 1 9 7 4 , rainfall at the top of the canopy w a s 2031 m m , w i t h total throughfall 1 1 8 7 mm (interception w a s 844 m m , 4 2 % ) . For the following y e a r , gross rain fall w a s 2 0 1 6 mm w i t h throughfall 1047 mm (interception 969 m m , 4 8 ¾ ) .

ASIA

In West P a k i s t a n , Raeder-Roitzsch & Masrur (1969) described a study in a 0.4 acre stand of medium-aged ( 4 0-50 y e a r s ) C h i r pine (Pfnus r o x b u r g h ï i ) . The stand w a s at an elevation of 5500 ft on a northwest facing 2 5 s l o p e ; average height of the trees w a s 60 f t , and m e a n dbh 15 inches. Mean rainfall at the site w a s about 50 inches.

Five raingauges (5-inch diameter standard) w e r e rotated at random amongst 25poi ηts lying on an 8 ft grid underneath the c a n o p y . Gross rainfall w a s measured by two 5-inch gauges in a large clearing near the forest s i t e ; stemflow w a s measured on three rando­ mly-selected trees by m e a n s of collar a t t a c h m e n t s w h i c h fed the stemflow into the d r u m s .

For the year December 1967 to N o v e m b e r 1 9 6 8 , net rainfall under the trees w a s 92¾; in absolute t e r m s , net rainfall w a s 53 inches out of a gross rainfall o f 5 7. 5 5 inches. After subtracting the very small amount of s t e m f l o w , interception loss was 8¾ of gross rainfall. The low figure w a s ascribed by Raeder-Roitzsch & Masrur to the o p e n n e s s of the Chir pine stand (canopy closure 0 . 4 ) .

Elsewhere in the Indian s u b c o n t i n e n t , Ray ( 1 9 7 0 ) worked in two pure s t a n d s : one of Shorea robusta (plot area 83 m2 containing 14 trees at a density of I 6 7 8 trees per hec-tare; average tree height in December 1968 w a s 8 . 7 m , and a v e r a g e g i r t h a t breast height 34 cms) and the other of A l s t o n i a scholaris (plot area 400 m2 containing 67 trees at a density of 1675 trees per h e c t a r e ; average tree height w a s 8.3 m , and average girth at breast 48 c m ) . T h e two plots w e r e about 500 m a p a r t .

Daily rainfall w a s measured by a standard g a u g e in the o p e n , being sited 1 1 8 m and 59Ο m from the p l o t s . Throughfall w a s measured in small troughs (35 cm χ 2 4 cm χ Π cm deep) placed randomly on the floors of the p l o t s ; 36 such troughs w e r e placed in the

larger p l o t , 12 in the smaller. Stemflow w a s a l s o measured by conical tiη co 1 1 ars fixed to tree s t e m s , but the number of trees measured is not stated.

Measurements w e r e taken in the two rainy seasons of 1 9 6 7 and I 9 6 8 . Total inter -ception losses are not g i v e n , but losses are given for rainfall " g r o u p s " of increasing m a g n i t u d e ; for storms between 1 and 10 m m , interception losses for A . scholaris and S . robusta w e r e 30.9¾ and 33-7¾ respectively.

Low ( 1 9 7 2 ) studied interception losses in the humid forested area at L a w i u , in the Sungei Lui C a t c h m e n t , West M a l a y s i a .

Ho account of the forest composition is g i v e n ; o ne raingauge recorder w a s placed under a 25 ft specimen of Shorea c u r t i s i i , with its funnel above the c a n o p y , whilst 3 non - record i np gauges (dimension u n s p e c i f i e d ) w e r e placed randomly under the canopy to measure throughfall. Stemflow w a s apparently not m e a s u r e d . During the period 17 April

1 9 6 9 to II September 1969, for which unbroken records w e r e a v a i l a b l e , gross rainfa 11 was 3 8 . 8 9 inches, w i t h an interception loss of 1 4. 2 7 inches, representing 36¾ of rainfall.

Low's results must be open to some d o u b t . Of the three ground-level g a u g e s , o n e was stolen shortly after the work b e g a n , and a n o t h e r w a s removed. For the period defined a b o v e , t h e r e f o r e , the instrumentation w a s limited to one gauge recording g ross r a i n f a l l , and one recording throughfall. Th e q u e s t i o n a b l e nature of the results is a l s o sugges ted by the fact that between 2 1 and 28 August 1 9 6 9 , zero interception was recorded although 3 inches of rain w e r e m e a s u r e d .

Elsewhere in M a l a y s i a , Kenworthy ( I 9 6 9 ) reported results from a 3 1 . 5 ha site of Hill O i p t e r o c a r p Forest in the Ulu Gombak Forest R e s e r v e , S e l a n g o r . T h e area lies bet-ween 8OO and I8OO f t , the lower slopes having been selectively felled some 1 2- 1 5 years previously, w h i l e the upper reaches remained under primary forest. Rainfall w a s n e a s u -red by an unspecified number of r a i n g a u g s s ; no information is g i v e n regarding their position. Throughfall was apparently measured by an unspecified number of 1 m e t r e square trays. Stemflow vias not m e a s u r e d . For the period August 1968 to January 1969 , ra i η fa 1 1 is quoted as 2500 m m , and interception as 450 mm ( 1 8 ¾ ) ; this e s t i m a t e of interception loss must be regarded with some r e s e r v e , h o w e v e r , both because of the a b se n c e of any justification for neglecting stemflow and because of the absence of information about the instrumentation used.

Tangtham ( 1 9 8 1 ) states what will already be clear to readers of this p a p e r , "The proportion of rainfall intercepted by tropical forests is very v a r i a b l e . In tropical Southeast A s i a , interception measured in T h a i l a n d indicated high values of 6 3 per cent gross annual rainfall for the natural teak f o r e s t , 61 per cent for dry d i p t e r o c a r p ; 9 per cent for hi11-evergreen f o r e s t ; and only 4 per cent for the dry evergreen forest

(Chunkao et a l . , 1 9 7 1 ) . T h i s kind of investigation in the lowland r a i n f o r e s t o f Ma 1 aysia gave a value of 22 per cent ( M a n o k a r a n , 1 9 7 9 ) . It was also stated by Raros ( 1 9 7 9 ) that intercepted w a t e r may exceed 60 per cent of gross rainfall for the humid tropical forest in the P h i l i p p i n e s " ,

Finally, results are becoming a v a i l a b l e from a c o l l a b o r a t i v e study between the United Kingdom Institute o f Hydrology and the Indonesian Institute Pertanian Bogor IPB

(Calder et a l. , 1 9 8 1 ) . A site has been instrumented in the Janlappa Nature Reserve,about 55 km northwest o f Bogor at an a l t i t u d e of 6080 m above sea level. T h e r e s e r v e , c o n -taining one of the few remaining areas o f lowland tropical rainforest in West J a v a , is 32 ha in e x t e n t , and is divided into 21 blocks by inspection paths o f 3 m w i d t h .

Gross r a i n f a l l , net r a i n f a l l , and other meteorological v a r i a b l e s a r e recorded at 5-minute intervals by two a u t o m a t i c weather s t a t i o n s ; these h a v e been modified somewhat frcrn a standard Institute of Hydrology d e v i c e . One weather stations is mounted in a

clearing, over scrub v e g e t a t i o n , whilst the other is mounted above the forest canopy o n a 1** m bamboo tower. Beneath the canopy a r e two large ( 3 7 . 7 m2) plastic net-rainfall

gauges, which record the total of throughfall and stemflow.

Calder e t al. show some o f their extensive data o n 5 minute totals of gross and net rainfall. For the month of September 1 9 8 0 , the measured interception ratio w a s 12.6¾; however, September is on a v e r a g e the driest month at J a n l a p p a , and further data are being collected to determine an annual interception ratio.

DISCUSSION

The results d o not appear to be indicat i ve of any cons istent p a t t e r n , nor is it clear whether the differences can be attributed to measurement e r r o r s or w h e t h e r they are real differences. It is a l s o not possible to determine w h e t h e r , if they are real differences,

they are caused by differences in climate or forest s t r u c t u r e . It might be helpful to consider the likely causes of d i f f e r e n c e s in interception l o s s ; using the framework of

Gash ( 1 9 7 9 ) simplified by Shuttleworth ( 1 9 7 9 , Equation 7 3 ) w i t h the further s i m p l i f i c a

-tion that small storms w h i c h d o not saturate the canopy a r e neglected,the intercep-tion ioss can be expressed a s : I = ·- Ρ + nS w h e r e Ê and R a r e m e a n evaporation and rainfall

R

rates during r a i n f a l l , η is the number of rainfall events and S is the canopy capacity. In temperate climates typified by long periods o f low r a i n f a l l , the two terms o n the right hand side of this equation contribute about equally to the total loss.However, in a climate dominated by local convective r a i n f a l l , with s h o r t - d u r a t i o n , high intensity storms, R will be p e r h a p s an o r d e r of m a g n i t u d e greater than the typical l-2mm/hour found in temperate c l i m a t e s , and this might be expected to increase the relative size of the second term. Little is known about the size of Ε under tropica) c o n d i t i o n s ; low wind speeds would tend to decrease it (see the Mauritius w o r k a b o v e ) , higher radiation and vapour pressure deficits w o u l d tend to increase it. C o n v e r s e l y , in a m o n s o o n - t y p e cli mate with relativly few rainfall events but of long d u r a t i o n , the first term in the equation m i g h t be expected to d o m i n a t e . T h e above argument suggests that future studies of tropical interception should include not only m e a s u r e m e n t s of rainfall rate and an attempt at measurement of canopy c a p a c i t y , but a l s o meteorological m e a s u r e m e n t s to enable the estimation of evaporation rate during rainfall.

It is clear that the l o s s , by e v a p o r a t i o n , of rainwater intercepted by a tropical forest canopy can in some cases represent a considerable proportion of gross rainfall. It is important, a l s o , to remember that any quoted interception ratio represents o n e aspect only of w a t e r foss from the forest; if, during a n y given time interval, there had been no r a i n , there would have been n o interception l o s s , but w a t e r w o u l d still have been lost by transpiration. It is therefore necessary to consider the total w a t e r loss through both interception and t r a n s p i r a t i o n , if forest water use is to be assessed adequate 1y.

Where the interactions betwecn vjet l e a v e s and the a t mo s p he re (of wfiich intcrception loss is a phenomenon) and between d r y ]?aves and the a l mospíiere (of w h í t h t r a n s p i ra t i on loss is a phenomenon) a r e to be considered together, thcn the müsC s-jt isfactory way i n which total water loss can be assessed is by descriplion of the processes in p h y s i c a l terms. Their complexity has been wel ) described by Jatkson (1975 , o p . c i t .) ,who ;i I s o se ts out the d i f f i eu 1 t i es of obtainino the n;ea surement s necessary for implementa tioti of ( f o r example) the fiutter tradel , which requires h o u r l y greíá r a i n f a l l , v,'ind run and hourly mean t e m p e r a t u r e , humidity and net radiation above the c a n o p y . "TJie çotpptqfiefi s i ve a p -proach of Rui ter ... rvay lie liinitarf in ternns of appllcation e l & e w h e r e , becausii such complete i ns t rumenta t i on is imposs. i b I e" C j á c V s o n , 1 9 7 5 ) .

Whilst the difficulties of asaessing the tot.,1 wator loss fro.n tropical forests should not be underest i rra ted, they should not be exaggcrated e i t h e r . DurJng tíie six years since Jackson wrote the a b o v e q u o t a t i o n , great atlvances have beçn vr^ade i n the reliability and portability of the necâssiry hydro 1 og i ca I i nst runsen ta t i on . Solid statc menor y for i ns t rumen ts wi th low power requ i renien t s , and w i t b batteries rechargt-ab 1 e by solar p o w e r , is now a r e a l i t y , and it wi I I be unfortunate i f v/ork such as t h a t o f C a l d e r (198 I ,op.cí t.} is not repeated elsewhere within tropical forests, represent inq a s they do an imriensely important natural resource.

ACKSOWLEDCEMENTS

lan Calder and Ivan Wríght provided rrost o f the r e f e r e n c i a quotêd in this p a p e r ; John Gash read the draf t , m a d e substantial c o r r e c t i o n s , and provided Fu r ttter ma türi a ' . My debt to ali three is gratefully acknowledqed.

RESUMO

É f,aíta íur.a t c u ó s a o da ZÃdaiatuic iüiikf med-idai M p à O c e J-loí dí. i\it<LJicc.\Ka.o cia

água da chuva p c i í f - o i e v í a . s . Sao apfizi enfadai -tíiJijAmHçceA da Á^iica, da Amçftíeç Cen

-tiai e do Sul, e fia fota. f ^cita uma anítiAÇ. geAsf e i í t f }\ic.ítAada6 ai necp^ildadeò de.

References

Blake, J. G. 1 9 7 5 . The interception p r o c e s s . I n : C h a p m a n , T. G. & D u n i n , F. X . e d . -Prediction in Catchment H y d r o l o g y . C a n b e r r a , p . 5 9 - 8 1 .

Calder, 1. R. - 1 9 7 6 . The m e a s u r e m e n t of w a t e r losses from a forested area using a " n a t u r a l " l y s i m e t e r . J. H y d r o l . , 3 0 : 3 1 1 - 3 2 5 .

1 9 7 8 . T r a n s p i r a t i o n o b s e r v a t i o n s frc:r. <* spruce forest and comparisons w i t h predictions from an evaporation m o d e l . J. H y d r o l . , 3 8 : 3 3 - 4 7 .

Calder, 1. R. & N e w s o n , M . fJ. - 1 9 7 9 . Land use and upland w a t e r r e s o u r c e s in B r i t a i n : a strategic look. W a t e r R e s o u r . B u l l . , 15 ( 6 ) : 1 6 2 8 - 1 6 8 9 .

Calder, Τ. R.; M u r d i y a r s o , D. W r i g h t , I. R. - 1 9 8 1 . IH-IPB Study of Evaporation from Lowland Tropical R a i n F o r e s t , J a v a . I n s t i t u t e of H y d r o l o g y . P r o g r e s s R e p o r t ( 1 ) .

Chunkáo, K.; T a n g t h a m , N . ; U n g k u l p a k d i k u l , S. - 1 9 7 1 . M e a s u r e m e n t s rainfall in early wet season under hilt and d r y - e v e r g r e e n , n a t u r a l teak, and d r y - d i p t e r o c a r p forests of T h a i l a n d . R O G - M A W a t e r s h e d R e s e a r c h Bull ( 1 0 ) .

Clegg, A. G. - 1 9 6 3 . Rainfall interception in a tropical f o r e s t . C a r i b b e a n F o r e s t e r , 2 4 ( 2 ) : 4 5 - 5 9 .

Clements, R. G. & C o l o n , J. A. - 1 9 7 5 . The rainfall interception p r o c e s s and m i n e r a l cycling in a m o n t a n e rain forest in eastern Puerto R i c o . I n : H o w e l l , F. G. G e n t r y , J. B.; S m i t h , Μ. H. e d s . - M i n e r a l C y c l i n g in S o u t h e a s t e r n E c o s y s t e m s . US Energy Research and development A d m i n i s t r a t i o n .

Da Silva N e t o , J. B. ; T e r e z o , Ε . ; B a s s o , E . ; O s ó r i o , C ; C a r v a l h o , R. ; S o l o m o n , S.-1979. Estudo da h i d r o l o g i a e climatologia da b a c i a A m a z ó n í c a B r a s i l e i r a . I n : III S i m p ó s i o B r a s i l e i r o d e H i d r o l o g i a , 1 : 1 8 0 - 1 9 2 .

F r e i s e , F. - 1 9 3 6 . D a s B i n n e k l i m a van UrwaUlern in S u b r t r o p i s c h e n B r a s i l i e n . P e t r m a n n s M i t t . , 8 2 : 3 0 1 - 7 , 3 4 6 - 3 4 8 .

G a s h , J. H. C. - 1 9 7 9 . An analytical m o d e l of rainfall interception by f o r e s t s . Q u a r t . J. R o y . M e t . S o c . , 1 0 5 : 4 3 - 5 5 .

G a s h , J. H. C. & M o r t o n , A . J. - 1 9 7 8 . An a p p l i c a t i o n of the Rutter model to the e s t i ­ m a t i o n of the interception loss from Thetford F o r e s t . J. H y d r o l . , 3 8 : 4 9 - 5 8 .

G a s h , J. H. C ; W r i g h t , I. R.; T J o y d , C. R. - 1 9 8 0 . C o m p a r a t i v e e s t i m a t e s of interception loss from three coniferous forests in Great B r i t a i n . J. H y d r o l . , 4 8 : 8 9 - 1 0 5 .

H e l v e y , J. D. - 1 9 6 7 . I n t e r c e p t i o n b y Eastern W h i t e P i n e . W a t e r R e s o u r R e s . , 3 ( 3 ) : 7 2 3 - 7 2 9 .

H e l v e y , J. D, & P a t r i c , J. H . - 1 9 6 5 . Canopyand litter interception of rainfall by h a r d w o o d s of eastern United S t a t e s . W a t e r R e s o u r . R e s . , 1: 1 9 3 - 2 0 6 .

" e u v e l d o r p , J. - 1 9 7 9 . The international Amazon M A B Rainforest Ecosystem Pilot; Project at San C a r l o s de Rio N e g r o : m i c r o m e t e o r o l o g i c a l studies.In: T r a n s a c t i o n s of the Second International M A B - I U R O W o r k s h o p o n Tropical Rainforest E c o s y s t e m s R e s e a r c h . E d . S. A d i s o e m a r t o & Ε . F. B r u n i g . H a m b u r g - R e i n b e k Special R e p o r t , 2.

H o p k i n s , B. - I960. Rainfall interception by a tropical forest in U g a n d a . East A f r i c a n A g r i c . J o u r . , 2 5 : 2 5 5 - 2 5 8 .

· 1 9 6 5 . Vegetation of the Olokomeji F o r e s t R e s e r v e , N i g e r i a . I I I . The microeli -mate w i t h special reference to their seasonal c h a n g e s . J. E c o l . , 53 (1) 1 2 5 - 1 3 8 .

J a c k s o n , I. J. - 1 9 7 1 . P r o b l e m s of throughfall and interception assessment under tropical f o r e s t . J. H y d r o l . , 1 2 : 2 3 4 - 2 5 4 .

- 1 9 7 5 . R e l a t i o n s h i p s b e t w e e n rainfall p a r a m e t e r s and interception b y tropical f o r e s t s . J. H y d r o l . , 2 4 : 2 1 5 - 1 2 3 1 .

K e n w o r t h y , J. B. - 1 9 6 9 . W a t e r b a l a n c e in the tropical r a i n f o r e s t : a preliminary study in the U l u G o m b a k Forest R e s e r v e . M a l a y N a t . J., 2 2 : 1 2 9 - 1 3 5 .

K l i n g e , H . ; F u r c h , K . ; I r m l e r , U . ; J u n k , W. S. - 1 9 8 1 . Fundamental ecological Parameters in A m a z o n i a , in R e l a t i o n to the P o t e n t i a l Development of the R e g i o n . I n : L a i , R. & R u s s e l l , E . W . e d s . - T r o p i c a l A g r i c u l t u r a l H y d r o l o g y , p . 1 9 - 3 6 .

K l i n e , J. R. & J o r d a n , C. F . - 1 9 6 8 . T r i t i u m m o v e m e n t in Soil of a T r o p i c a l R a i n F o r ­ e s t . S c i e n c e , 1 6 0 : 5 5 0 - 5 5 1 .

L a i , R. - 1 9 8 1 . D e f o r e s t a t i o n of T r o p i c a l Rainforest and H i d r o l o g i c a l P r o b l e m s . I N : L a i , R. & R u s s e l , E . W . - T r o p i c a l A g r i c u l t u r a l H y d r o l o g y . J o h n W i l e y , p . 1 3 1 - 1 4 0 .

L a w s o n , E . R. - 1 9 6 7 . T h r o u g h f a l l and S t e m f l o w in a Pine-Hardwood Stand in the Quachita M o u n t a i n s of A r k a n s a s . W a t e r R e s o u r . R e s . , 3 ( 3 ) : 7 3 1 - 7 3 5 .

L o w , K . S. - 1 9 7 2 . Interception losses in the Humid F o r e s t e d Areas (with special refer­ ence to SungaiLui C a t c h m e n t , West M a l a y s i a ) . M a l a y . N a t . J., 2 5 : 1 0 4 - 1 1 1 .

M a n o k a r a n , N . - 1 9 7 9 . S t e m f l o w , throughfall and rainfall interception in a lowland tropical rainforest i n P e n i n s u l a r M a l a y s i a . T h e M a l a y s i a n F o r e s t e r , 4 2 ( 3 ) : 1 7 4 - 2 0 1 .

N y e , P. H . - 1 9 6 1 . O r g a n i c matter and nutrient c y c l e s under m o i s t tropical f o r e s t . Plant and S o i l , 13 ( 4 ) : 3 3 3 - 3 4 5 .

P e n m a n , H . L . - 1 9 6 2 . V e g e t a t i o n and H y d r o l o g y . Coenjonwealth B u r e a u of S o i l s . T e c h . C o o n n m i c a t i o n ( 5 3 ) .

P e r e i r a , H . C . - 1 9 5 2 . Interception of rainfall by C y p r e s s p l a n t a t i o n s . E.African A g d c . J o u r . , 1 8 : 7 3 - 7 6 .

R a e d e r - R o i t z s c h , J. E . & M a s r u r , A. - 1 9 6 9 . O b s e r v a t i o n s on rainfall interception in Chir Pine (Pinus r o x b u r g h i i ) w o o d l a n d . P a k i s t a n J o u r . F o r e s t r y , 19 ( 1 ) : 9 9 - 1 1 1 .

R a r o s , R. S. - 1 9 7 9 . C r i t i c a l ecological c o n s i d e r a t i o n in r i v e r basin m a n a g e m e n t in Southeast A s i a . P a p e r prepared for Regional P o s t - G r a d u a t e T r a i n i n g W o r k s h o p on Integrated River Basin M a n a g e m e n t , 19 N o v . - 7 D e c , P h i l i p p i n e s .

R a y , M . P. - 1 9 7 0 . Preliminary o b s e r v a t i o n s of s t e m f l o w etc in A l s t o n i a s c h o l a r i s and Shorea robusta plantations at A r a b a r i . W e s t Indian F o r e s t e r , 96 ( 7 ) : 4 8 2 - 4 9 3 .

R e a d , R. G . - 1 9 7 7 . M i c r o c l i m a t e as Background Environment for E c o l o g i c a l S t u d i e s of I n s e c t s in a Tropical F o r e s t . J o u r . A p p l i e d M e t e o r o l o g y , 1 6 : 1 2 8 2 - 1 2 9 1 .

R i c h a r d s , P. W . - 1 9 5 2 . T h e T r o p i c a l R a i n F o r e s t . C a m b r i d g e .

R u t t e r , A. J.; K e r s h a w , Κ . Α . ; P. C ; M o r t o n , A. J. - 1 9 7 1 . A p r e d i c t i v e model of rain fall interception in f o r e s t s . I. D e r i v a t i o n of the m o d e l from o b s e r v a t i o n s in a p l a n t a t i o n of C o r s i c a n P i n e . A g r i . M e t . , 9: 3 6 7 - 3 8 4 .

S e l l e r s , P. J. & L o c k w o o d , J. G . 1 9 8 1 . A computer simulation of the effects of dif -fering crop types o n the w a t e r b a l a n c e of small c a t c h m e n t s over l o n g t i m e p e r i o d s . Q u a r t . J. R o y M e t . S o c , 1 0 7 : 3 9 5 - 4 1 4 .

Shuttleworth, W . J. & G a s h , J. Η. C. - 1 9 8 2 . Comment on p a p e r by Sellers and L o c k w o o d . Q u a r t . J. R o y M e t . S o c . , (in p r e s s ) .

Sollins, P. & D r e w r y , G. - 1 9 7 0 . A T r o p i c a l r a i n f o r e s t ; a Study of I r r a d i a t i o n and Ecology at E l V e r d e , P u e r t o R i c o . Ed. Η . Τ. Odum & R. F. P i g e o n . U S Atomic Energy Comission. D i v . of T e c h . I n f . C h a p . H. 1 0 .

Stewart, J. B . - 1 9 7 7 . E v a p o r a t i o n from the w e t canopy of a pine f o r e s t . W a t e r Resour. R e s . , 1 3 : 9 1 5 - 9 2 1 .

Tangthain, K. - 1 9 8 1 . Ecological a s p e c t s of w a t e r s h e d resources m a n a g e m e n t in humid trop ic.al e c o s y s t e m s . In: H A B / N E B / U S A I D R e g i o n a l T r a i n i n g C o u r s e on W a t e r s h e d R e s o u r c e s Management and E n v i r o n m e n t a l M o n i t o r i n g in H u m i d T r o p i c a l E c o s y s t e m s . C h i e n g m a i Thai

land, 25 M a y - 8 June 1 9 8 1 .

Vaughan, R. & W i e h e , P. 0. - 1 9 4 7 . S t u d i e s in the v e g e t a t i o n of M a u r i t i u s ( I V ) : Some notes on the interval climate of the upland climax f o r e s t . J. E c o l . , 3 5 : 1 2 6 - 1 2 5 .

Wimbush, S. H. - 1 9 A 7 . R a i n f a l l interception b y C y p r e s s and B a m b o o . E . A f r i c a n A g r i c . J o u r . , 1 3 : 1 2 3 - 1 2 5 .