P l a n t a n d Soil X V I , n o . 1
NITRATE
February
DISTRIBUTION
1962
IN TROPICAL SOILS
IIk DOWNWARD MOVEMENT AND ACCUMULATION OF NITRATE IN THE SUBSOIL by R. W E T S E L A A R Division of Land Research and Regional Survey, Commouwealth Scientific Industrial Research Organization, Canberra, A.C.T. Australia
INTRODUCTION
Nitrate accumulation in tropical subsoils under continuous cultivation has been reported by m a n y investigators in a number of different countries 6 7 9 10 11 13 14 17. High nitrate contents have also been found in the subsoil of Tippera clay loam at Katherine, N.T., Australia, the environment of which has been described in Part I of this series 19. In this soil 500 pounds per acre of nitrate nitrogen accumulated in the top 10 ft of the soil, with a peak between 5 and 6 ft depth, after six years of bare fallow. In Part I of this series the accumulation of nitrate near the soil surface after a long dry period was described, and in Part II 20 it appeared that this was due to capillary movement upwards of nitrate which had been formed at greater depth. It was also found that the upward movement in Tippera clay loam was restricted to the top 11 ft of the soil. This implies that any nitrate which has been leached below this depth cannot be recovered in the topsoil by capillary movement. P h i l l i p s 15 has shown on the same soil type at Katherine that 95 per cent of the root mass of grain sorghum is restricted to the top 16 inches of the soil. Consequently substantial losses of nitrate can occur by leaching beyond the root zone. Since the nitrogen status of the soils ill the Katherine region is relatively poor (topsoil 0.04 to 0.07 per cent total nitrogen) and nitrogenous fertilizers are expensive because of transport costs, it was desirable to have a --
19--
20
R. WETSELAAR
better insight into the rate of removal of nitrate out of the topsoil and the depth at which subsoil accumulation occurred. It was decided to study the leaching phenomenon in situ and to compare the movement of applied nitrate with applied chloride on bare fallow soil in order to isolate biological side effects. Virgin soils virtually free of nitrate in the whole profile were selected. The following three soil types were studied - Tippera clay loam, Blain sand and Florina sand. Their primary characteristics have been described earlier 19 2o METHODS
(a) Experiment z In the 1957/58 wet season sodium nitrate and sodium chloride were applied to bare fallow Tippera clay loam in separate triplicate plots of 6 × 22 yards each, in a fully randomized design, on December 10, December 27, and February 21. Sodium nitrate was hand-broadcast at the rate of 2000 lb/acre in four split applications. Sodium chloride was applied in the same way and at an equivalent anion rate to facilitate direct comparison of anion concentrations in the soil samples. Control plots, receiving no salt, were included. A t the end of the wet season, in May 1958, samples were collected at 6-inch intervals to 5 ft depth on the different plots. By that time the plots to which the salts were applied on December 10 had received 23.7 inches of rain. Those which received salts on December 27 and February 21 had received 15.4 and 7.4 inches respectively since time of application. The plots were regularly harrowed for weed control and to aid permeability. Three holes per plot were augered,, samples at the same depth being bulked and subsampled. The rest of the bulk sample was replaced in the augerholes.
(b) Experiment 2 In order to collect further data to relate amount of rainfall with anion movement, the plots to which NaCI was applied on February 21, 1958, were again sampled in the same way, at 6-inch intervals, during the following wet season (1958159) on December 22, J a n u a r y 14, and February 3. Between the date of salt application and date of profile sampling 11.8, 18.5, and 21.3 inches of rain were received respectively.
(c) Experiment3 To compare anion movement in different soil types, sodium chloride at the rate of 2000 lb]acre was handbroadcast on triplicate plots each 4 × 12 i t on December 3, 1958, on Blain and Florina soil. Before the salt application,
NITRATE DISTRIBUTION IN TROPICAL SOILS. III
21
samples were collected to evaluate the actual cloride content of the soil to be used as control values. Samples were taken in 1-ft increments in the top 5 ft of the soil on J a n u a r y 13 and again at the end of the wet season after the last rains. F r o m the time of salt application to the J a n u a r y sampling the Blain site had received 8.5 inches of rain and the Florina site 7.7 inches, and between application and sampling at the end of the wet season the Blain site h ad received 23.4 inches and the Florina site 18.0 inches of rain. All plots were kept bare fallow by regular harrowing of the top 3 inches of the soil.
(d) Determination Soil nitrate was determined spectrophotometrically on 10-g subsamples, using phenoldisulphonic acid according to the method of S t e i n b e r g s (personal communication). Chloride content of samples collected in the 1957/58 season was determined by using 0.1 N AgNO3. After filtering the excess AgNO3 was back titrated with 0.1 N NH4CHS. The 1958/59 samples were determined as described in Part I I of this series 20 The mean m o v e m e n t in inches of the anion into the soil profile was determined according to B a t e s and T i s d a l e 2, i.e. the mean of the percentages of the total applied anion found in each sampling layer weighted by the distance (in inches) from the centre of each layer to the surface of the soil. The product of the top 6 inches was given the value of 0. The calculated mean movements of the anions were in good agreement with the calculation of the "mean chloride depth" as given by G r o e n e w e g e n 5. All calculations are based on the net anion content of oven-dry soil, t h a t is, after deduction of the control values as evaluated by the control plots. RESULTS
(a) Exlberiment z In Experiment 1 on Tippera soil, significant differences in anion content were recorded at all depths to 3½ ft between the control and chloride plots, and to 4 ft between the control and nitrate plots, The distribution of the net chloride and nitrate content over the 5-ft profile of these plots is given in Figures 1A and B respectively, together with the standard errors for each sampling depth. On the chloride plots significant decreases in net chloride content due to increasing amounts of rain were obtained in the top foot layers, whilst in the 2- to 3-ft layers the chloride content increased significantly. On the nitrate plots the net nitrate contents decreased significantly in the top ft of the soil and increased significantly in all layers in the subsoil between 1½ and 3½ ft. When comparing the net content of nitrate with chloride at each
22
g. WETSELAAR
depth the only significant differences between the two anions were in the 1½-2 It layer after 7.4 inches of rain, and in the three layers 0 to 1½ ft after 15.4 inches of rain. After 23.7 inches of rain the actual recovery of the applied anions was 100.0 per cent for the chloride and 125.8 per cent for the nitrate. It appeared t h a t the formation of nitrate in the top soil was stimulated b y sodium
Cl'(ppm) 0
!i
200
400
i
i
X
I
NO3 (ppm) 600
0
S.E.
_2o '
+Jo
I
I
•
200 •
~.
"Sa-
/
S.E. I
l/"
•
(
+60
"
/i;/
1"*'4
/;"~
•
i
1"-'-4
I
"a
60O
I
--60
,
c" 2.
400
i
V*4
]
14t
/
3.
./ /7 °
d
,-li/g
//
I
ld3
/ /
4-
rl I
M
W
S
A
B
Fig. 1. Applied anion distribution in top 5 feet of Tippers soil 1957/58 after 7.4 ( ), 15.4 ( - - - - - - ) , and 23.7 (. . . . ) inches rain. nitrate application, since the control values had already been deducted. This would explain the high nitrate values in the top 1-~ It of the soil after 15.4 inches of rain as compared with the chloride values. Apparently this did not result in a different distribution between nitrate and chloride after 23.7 inches of rain. Furthermore there was no significant difference between the mean m o v e m e n t of the two anions after different amounts of rain (Table 1). The difference in mean movement due to the a m o u n t of rain was significant (P = 0.05) (Table 1). (b) Experiment 2 In Experiment 2 on Tippera soil, the changes in chloride content at different depths during the 1958/59 wet season on the plots
23
N I T R A T E D I S T R I B U T I O N IN TROPICAL SOILS. I I I
TABLE 1 Mean movement (in inches) of applied nitrate and chloride after different amounts of rainfall during the 1957/58 wet season, in Tippera clay loam Anion 7.4
]
Rainfall, inches 23.7 laA [
NOaC1- . . . . . . . . . .
5.5 9.3
14.5 20.0
27.6 27.1
Mean S.B. ~= . . . . . . . .
7.4
17.2 1.30
27.3
J
S.E. ±
Mean 15.9 18.8
1.06
which had received NaC1 on February 21, 1958, were of a similar nature to those of Experiment 1. The mean movements after different amounts of rain are given in Table 2 and indicate significant differences (P = 0.05) due to different amounts of rain. TABLE 2 Mean movement of applied chloride after different amounts of rainfall during the 1958/89 wet season, in Tippera clay loam Rainfall (inches) . . . . . . . . . Mean movement (inches) . . . . . .
11.8
18.5
~
~
21.3 ~
S.E. ~
-
30. "n, I: v
'~ 2.0. ~9
K >
o E
Io
,.-..
1:3 E
0
f/ ,b
2'o rainfall
3'o (in.)
Fig. 2. R e l a t i o n b e t w e e n m e a n d o w n w a r d m o v e m e n t of a p p l i e d a n i o n s a n d r a i n f a l l o n T i p p e r a soil.
In Figure 2 the mean movement of the anions in Experiments 1 and 2 on Tippera soil is plotted against rainfall (chloride and nitrate in 1957/58 and chloride in 1958/59). The relationship between mean movement and rainfall is expressed by the equation y = r e x , in which m represents the mean movement per inch of rainfall. The adjusted line, as given in Figure 2, is represented by
R. WETSELAAR
24
y = 1.075x, with a s t a n d a r d error of :b 0.0473 for r~. The correlation between rainfall and mean m o v e m e n t was r - - - - + 0 . 9 4 6 (P = 0.01). The percentage applied chloride either retained in or lost from different depths of soil during the 1957/59 period (Experiments 1 and 2) is given in Figure 3.
"(3IOO~
t
0ol
1
[o
\
\
\
o
~
.oo
•
o
,6
o
~,~ o
~
IOO
2'0
rainfall
(in.)
Fig. 3. Relation between rainfall and percentage of applied chloride retained in or 10st from different thicknesses of topsoil. 0-6 (®), 0-12 (o), 0-18 (E3), and 0-24 (I) inch layer. (c) Experiment 3 For Blain and Florina soil (Experiment 3) the chloride distribution of the applied anion after different a m o u n t s of rainfall is given in Figures 4 and 5 respectively. Significant differences in net chloride content due to different amounts of rain were obtained in the 2-3 ft layer on Florina and in the 1-2 ft and 3-4 ft layers on Blain. The mean movements after the two a m o u n t s of rainfall are TABLE Mean
movement
3
of applied chloride after different amounts of rainfall during the 1958/59 w e t s e a s o n , i n F l o r i n a a n d B l a i n soil
Soil
Florina
[
R a i n f a l l (inches) . . . . . . .
7.7
18.0
S.E.
M e a n m o v e m e n t (inches)
8.0
17.0
4-1.05
. . .
]
Blain 8.5
23.4
S.E.
21.6
39.8
4-1.48
NITRATE DISTRIBUTION IN TROPICAL SOILS. III
25
given in Table 3 for both soil types. The rainfall/mean movement relationships were y = 0.99x for Florina and y = 2.12x for Blain. On Florin• 96 per cent of the applied chloride was recovered after CI- (ppm) aoo 300
,00 ,
,
S.E.
400
I
t
I
I
•
--70
"%
v
~"
+7O
I
•
l
2
Ck ~9 "O 7~ 3.
o
!
/
/
/
//
5
Fig. 4. Applied chloride distribution in top 5 feet of Florina soil, 1958/59, atter 7.7 (--) and 18.0 ( - - - ) inches rain. CI- (ppm) tO0
200
,
P_ 5 2 1:3.
i
300 i
400
S.E.
' Jo
'+&
,/ I
•
I
~3-
t" //*I
/
o/
Fig. 5. Applied chloride distribution in top 5 feet of Blain soil, 1958/59, after 8.5 (--) and 23.4 ( - - - ) inches rain.
26
R. W E T S E L A A R
18.0 inches of rain, while the recovery on Blain was only 60 per cent after 23.4 inches of rain. This suggests that substantial amounts of the applied chloride on Blain moved into the subsoil below the maximum sampling depth of 5 ft. This could not be taken into account in the calculation of mean movement. If it could have been, the mean movement in this soil type after 23.4 inches of rainfall would have been greater. DISCUSSION
(a) Strati]ication o/ anions over soil depth In Experiment I significant differences between nitrate and chloride content were recorded in only 4 out of 30 soil-depth/ rainfall-amount combinations. This indicates strongly that nitrate and chloride ions move downwards in the same way and at the same rate. It is clear from Figure 1 that this movement is not due to a complete displacement of the soil solution by the leaching rain, but to gradual dilution out of the topsoil, giving rise to a near-normal distribution of the anion over the soil profile after 23.7 inches of rainfall on Tippera clay loam (Figure 1). This would be in harmony with D a y ' s 4 concept that the velocity of the infiltrated water varies from point to point in a porous system, such as a soil profile. This gives rise to a random downward movement of droplets originating from the same infiltrated drop of water. D a y refers to the mechanism of movement as "hydrodynamic dispersion". He found that when pure water in a column of saturated sand was displaced with salt water, the salt was distributed symmetrically above and below the centre of concentration in accordance with the normal Gaussian distribution. However, a marked downward "tailing" in the chloride and nitrate stratifications can be seen in Figure 1 after 7.4 inches of rainfall, and to a lesser extent after 15.4 inches, on Tippera clay loam. This indicates that, apart from the hydrodynamic dispersion, other structural factors of the soil influence the initial distribution of the anions after rainwater infiltrates the soil surface. The distribution could arise if certain droplets were allowed to move downward with less obstruction than would be expected from the structure of the soil. Since Tippera clay loam does not form any cracks on drying, the dominant clay mineral being kaolinite, it is possible
N I T R A T E D I S T R I B U T I O N IN T R O P I C A L SOILS. I I I
27
that the droplets concerned followed channels of partly decomposed roots. These channels have been shown to be present in this soil type. The departure from normality of anion concentration could not have been due to an increase of the sand fraction with depth, since the clay content of Tippera clay loam increases gradually from 20 per cent in the top 6 in. to about 50 per cent at 3-6 ft depth. A similar downward "tailing" occurred in Florina and Blain soil (Figures 4 and 5) though to a lesser degree. (b) Ef]ect o] rain]all on mean movement o/anions The high correlation (r = +0.946) between rainfall amount and mean movement of anions in Tippera clay loam (Figure 2) suggests that rainfall is the most important factor affecting the movement of the anion within one soil type. Though this has long been known this relationship has usually been studied in the laboratory 3 16 where the soil column was kept wet b y the percolating water. In the experiments reported here the mean movement per inch of rainfall (r~ values) for the different soil types is dependent upon all climatic factors governing the rate of drying and wetting of the soil in situ. These factors contribute to the standard error of ~. During intermittent dry spells in the wet season some capillary movement upwards of the anions could have taken place in the top foot of the soil ( W e t s e l a a r 20), but apparently this was not of any significant magnitude to reveal a measurable compensation of the downward movement. (c) E]]ect o] soil type on mean movement o] anions The r~-values for the chloride ion differed between the three soil types studied. They were 1.075, 0.99, and 2.12 inches for Tippera, Florina, and Blain respectively. Both Blain and Florina are sandy soils with a clay content of approximately 5 per cent in the topsoil. Florina sand, however, has a very high silt content in the top compared with Blain sand and considerable run-off takes place thus reducing the amount of leaching rainwater. No further effort has been made to relate r~-values with soil characteristics such as texture or porosity, but the difference between Tippera
2~
R. W E T S E L A A R
and Blain confirms that ~ is higher on more permeable soils, as has been found b y other investigators 2 s (d) Practical implications The main cash crops for the Katherine region are peanuts, cotton, and sorghum, and the main non-leguminous fodder crops bulrush millet, sudan grass, and fodder sorghum. All these crops respond to applications of nitrogenous fertilizer 3 1% Hence all nitrogen that is made available in the soil to the plant should be used to reduce expenditure on fertilizer nitrogen. In a crop sequence not including leguminous crops 85 pounds per acre of nitrate nitrogen is formed each year in the topsoil ( W e t s e l a a r and A r n d t * ) . W e t s e l a a r and N o r m a n 21 have shown that, when the nitrogen supply of an annual fodder crop depends on the nitrate formed in the topsoil during one wet season, not more than 50 per cent can be incorporated in the above-ground plant material. Furthermore, W e t s e l a a r and A r n d t * estimated that at least 50 per cent of this amount of nitrate nitrogen is formed in the pre-planting stage in early wet season (October to mid-December), owing to the long period of drying during the preceding dry season followed b y re-wetting ( B i r c h 1). When crops are sown in mid-December an average of 7½ inches of rain has already been received Is. From Figure 3 it can be seen that, b y that time, about 60 per cent of the nitrate nitrogen already formed has been leached out of the top 6 in of the soil. A period of approximately 4 weeks passes before substantial plant growth is made, during which most roots do not penetrate beyond 1 ft depth, but their mass is probably negligible compared with the total soil mass in that zone, particularly with the standard 3-ft crop row spacing. Thus, with a total rainfall of 16.72 inches between the start of the rains and major crop root development, about 75 per cent, or 30 pounds per acre of nitrate nitrogen, can be lost out of the top foot of soil (Figure 3). When further development of the plant and its root system has taken place more rain will have been received, and the actual amount of nitrate nitrogen * The maintenance of soil fertility in dryland agriculture in the Katherine-Darwin region, Northern Territory, Australia. P a n Indian Ocean Sci. Assoc. 4th Sci. Congress. Karachi, 1960, Section D.~'(Not y e t published).
NITRATE DISTRIBUTION IN TROPICAL SOILS. III
29
that can be taken up by the plant will depend on - (i) the depth of the fully developed root system, which is mainly governed b y the species 21, and - (ii) the total amount of rainfall received till the end of the growing season. Quite clearly substantial losses of nitrate nitrogen can occur, particularly in the case of shallow-rooted crops such as sorghum 15 21. Under such conditions it can be expected that the crop will not be able to take up all the nitrate nitrogen that is made available during the wet season. W e t s e l a a r and N o r m a n 21 compared the ability of different fodder crops in recovering nitrate nitrogen that had accumulated at about 5 ft soil depth in previous seasons. Bulrush millet appeared outstanding in this respect. Other means to prevent or overcome leaching losses have been discussed by W e t s e l a a r and A r n d t * They included early planting, close spacing of crops, and temporary immobilization of the nitrate nitrogen in the topsoil by crop residues at the beginning of the wet season. The main objective in arable cropping in the region will always be to utilize, or to conserve for future utilization, available soil nitrogen. This m a y be accomplished with suitable agronomic practices and the use of deep-rooting crops. The long-term maintenance of soil nitrogen status can be achieved by incorporating leguminous crops in the rotation. SUMMARY T h e d o w n w a r d m o v e m e n t of t h e n i t r a t e ion i n t h e t o p 5 f t of T i p p e r a clay l o a m w a s followed b y a p p l y i n g s o d i u m n i t r a t e a t t h e r a t e of 2000 p o u n d s p e r acre t o b a r e fallow soil a f t e r d i f f e r e n t r a i n i n t e r v a l s d u r i n g t h e 1957/58 w e t season. A t t h e e n d of t h e s e a s o n soil s a m p l e s were collected a t 6-inch i n t e r v a l s a n d c o m p a r e d w i t h s a m p l e s f r o m a d j a c e n t p l o t s t h a t r e c e i v e d a n e q u i v M e n t a n i o n q u a n t i t y of s o d i u m chloride a f t e r t h e s a m e r a i n i n t e r v a l s . A f t e r 23.7 i n c h e s of r a i n t h e d i s t r i b u t i o n of n i t r a t e a n d chloride a n i o n s t h e soil profile was n e a r l y i d e n t i c a l a n d i t w a s c o n c l u d e d t h a t t h e a n i o n s are e q u a l l y m o b i l e in t h i s soil. T h e m e a n m o v e m e n t of t h e a n i o n was 1.075 inches for e a c h i n c h of rainfall. A h i g h p o s i t i v e c o r r e l a t i o n of 0.946 was o b t a i n e d b e t w e e n m e a n m o v e m e n t a n d rainfall. T h e d o w n w a r d m o v e m e n t of b o t h a n i o n s o u t of t h e topsoil a p p e a r e d t o b e e n h a n c e d b y c h a n n e l s left b y p a r t l y d e c o m p o s e d roots. * See footnote p. 28.
30
R. W E T S E L A A R
A p p l i c a t i o n of s o d i u m c h l o r i d e o n s a n d i e r soils r e v e a l e d a m u c h h i g h e r m e a n m o v e m e n t of a n i o n s o n B l a i n s a n d t h a n o n T i p p e r s c l a y l o a m a f t e r equal amounts of rain, but on Florins sand with a high silt content in the t o p s o i l t h e m e a n a n i o n m o v e m e n t a p p r o a c h e d t h a t of t h e c l a y soil. T h e d i f f e r e n c e is e x p l a i n e d i n t e r m s of l o w i n f i l t r a t i o n r a t e i n t o t h e F l o r i n s soil. T h e p r a c t i c a l i m p l i c a t i o n s of l e a c h i n g o f n i t r a t e i n T i p p e r a soil a r e b r i e f l y discussed.
ACKNOWLEDGEMENTS T h e c o n t i n u o u s a s s i s t a n c e in t h e field a n d l a b o r a t o r y of Mr. R . J. W r e n is g r a t e f u l l y a c k n o w l e d g e d . T h a n k s a r e d u e t o all s t a f f o f t h e K a t h e r i n e R e s e a r c h S t a t i o n f o r a s s i s t a n c e in t h e field. Received March 6, 1961
REFERENCES 1 B i r c h , H. F., The effect of soil drying on h u m u s decomposition and nitrogen availability. Plant and Soil 10, 9-31 (t958). 2 B a t e s , T. E. and T i s d a l e , S. L., The m o v e m e n t of nitrate nitrogen through columns of coarse-textured soil materials. Soil Sci. Soc. A m . Proc. 21, 528-8 (1957). 3 C.S.I.R.O. Aust. Division of Land Research and Regional Survey. Katherine Research Station Progress Report, 1946-56 (1959). 4 D a y , P. R., Dispersion of a moving salt-water b o u n d a r y advancing through saturated sand. Trans. Am. Geophys. Union 37, 595-601 (1956). 5 G r o e n e w e g e n , H., Relation between chloride accumulation and soil permeability in the Mirrool irrigation area, New South Wales. Soil Sci. BT, 283-8 (1959). 6 H a r d y , F., Seasonal fluctuations of soil moisture and nitrate in a h u m i d tropical climate (Trinidad, B.W.I.). Trop. Agr. (Trinidad) 23, 40 (1946). 7 J e w i t t , T. N., Field nitrates in Gezira soil. U. J. Agr. Sei. 47, 461-7 (1956). 8 L a r s e n , J. E. and K o h n k e , H., Relative merits of fall- and spring-applied nitrogen fertilizer. Soil Sci. Soc. Am. Proe. 11, 378-83 (1947). 9 L e u t e n e g g e r , F., Changes in the a m m o n i a a n d nitrate contents of tropical red loam as influenced b y manuring and mulching during a period of one year. E. African Agr. J. 22, 81-7 (1956). 10 M a r t i n , A. E. and Cox, J. E., Nitrogen studies on black soils from the Darling Downs, Queensland. I. Seasonal variations in moisture and mineral fractions. Australian J. Agr. Research 7, 169-83 (1956). 11 Mills, W. R., Nitrate accumulation in Uganda soils. E. African Agr. J. 19, 53-4 (1953). 12 N o r m a n , M. J. T. and W e t s e l a a r , R., Performance of annual fodder crops at Katherine, N.T.C.S.I.R.O. Aust. Div. Land Res. Reg. Surv. Tech. Pap. No. 0 (1960). 13 P a t h a k , A. N., Effect of seasonal variations on nitrate content of water soluble salts. Madras Agr. J. 40, 344-51 (1953). 14 P e r e i r a , H. C., Field measurements of water use for irrigation control in K e n y a coffee. J. Agr. Soi. 40, 459-66 (1957).
N I T R A T E D I S T R I B U T I O N IN TROPICAL SOILS. I I I
31
15 P h i l l i p s , L. J., The influence of land preparation on the yield of peanuts, sorghum, and cotton at Katherine, N.T.C.S.I.R.O. Australia. Div. Land Res. Reg. Surv. Tech. Pap. No. I (1959). 16 R o u s e l l e , V., Le movement des nitrates dans le sol et les consequences relatives a t'emploi du nitrate du soude. Ann. Sci. Agron. 1, 97-1t5 (1913). 17 S e h o f i e l d , J. L., A comparison of soil nitrate nitrogen values under bare fallow and after ploughing in various perennial tropical legumes and eowpeas. Queensland J. Agr. sei. 2, 170-89 (1954). 18 S l a t y e r , R. O., Agricultural climatology of the Katherine area, N.T.C.S.I.R.O. Australia. Div. Land Res. Reg. Surv. Teeh. Pap. No. 13 (t960). 19 W e t s e l a a r , R., Nitrate distribution in tropical soils. I. Possible causes of nitrate accumulation near the surface after a long dry period. Plant and Soil 15, 110-120 (1961). 20 W e t s e l a a r , R., Nitrate distribution in tropical soils. II. Extent of capillary movement during a long dry period. Plant and Soil 15, 121-133 (1961). 21 W e t s e l a a r , R. and N o r m a n , M. J. T., Recovery of available soil nitrogen by annual fodder crops at Katherine, Northern Territory. Australian J. Agr. Research 11, 693-704 (1960).