SOME POST-HARVEST MANAGEMENT PRACTICES ON RATOON … Moberly Som… ·  · 2014-02-21SOME POST-HARVEST MANAGEMENT PRACTICES ON RATOON CANE IN THE RAIN-FED ... from the bare soil

SOME POST-HARVEST MANAGEMENT PRACTICES ON RATOON CANE IN THE RAIN-FED AREAS

OF THE NATAL SUGARBELT

P. K. Moberly and R . K. McIntyre S. A . Sugar Association Experiment Station, Mount Edgecombe 4300 Sout African

Republic

ABSTRACT

Because of the increasing incidence of burning in preference to trashing . in the rainfed areas of the Natal Coast Lowlands, alternative methods of conserving soil and moisture were investigated. Basin-listing proved to be moderately effective but was not considered as practicable an alternative to a trash blanket as is the mulch from the burnt tops which are scattered over the soil surface. In terms of cane yield, the response of ratoon crops starting in summer to burnt tops was about 60% as effective as a full trash blanket. In the colder winter months, scattering of the burnt tops was superior to a trash blanket in terms of cane yield and the response of ratoon crops to a soil mulch was considerably lower.

Yield responses from ratoon crops grown in the moderately deep and well-drained soils are greater with a trash blanket or when burnt tops are scattered, compared with the shallow, poorly-drained soils or those which were very sandy. A mulch of any kind on alkaline sandy soils promoted ratoon chlorosis. The treatment effects were measured in terms of cane yield only and the most important long-term effects on soil and water conservation were not quantitatively assessed. Many observations made under field conditions however, indicate that run-off substantially reduced by any soil mulch, even when it is present in small quantities.

INTRODUCTION

In the cane growing areas of the Natal Coast Lowlands there has been a trend towards burning cane before harvesting rather than cutting green cane to preserve the trash mulch. During the past decade the change, in terms of the percentage of the area where trashing is practised, has been roughly as follows:

Zululand from 30-25% North Coast from 80-60% South Coast from 50-20%

The reason for the preference for burning cane are numerous and generally well founded but are nevertheless cause for considerable concern from the agrono- mic viewpoint.

A labourer can cut and stack about 40% more burnt cane than green cane. Furthermore, grab-loaders cannot be used efficiently in a trash field; the amount of extraneous material delivered to the mill is much less when'the cane is burnt; better pay-loads in transport vehicles are obtained with burnt cane; greater morta- lity of the cane borer Eldana saccharina can be postulated with the burning; and

there is increasing confidence that in the absence of a trash blanket, weeds can be controlled by using selective herbicides.

The decision to burn should be taken after considering all of the advantages and disadvantages. The substantial average yield increase that may be expected from the use of a trash blanket on the Natal Coast has been demonstrated and assessed in many experiments conducted during the past 40 years (Pearson, Thompson, 6' 72 and Wilson 9 ) .

Pearson reported a mean yield increase of 19 tclha (tons cane per hectare) per crop from six ratoon crops on a well drained Rydalvale clay loam soil covered by a trash blanket. The mean response to a trash blanket in five experiments conducted on a range of soils was reported to be 9 t canelha or 13%, with large responses being obtained on soils derived from Dolerite and Dwyka Tillite, and small or no responses on soils derived from Schists and Table Mountain Sandstone. Thompson reported that the mean response due to a trash blanket on the Rydalvale soil at Mount Edgecombe from nine ratoon crops over three succesive crop cycles, was 9,2 t canelhalyear, which is equivalent to 13,8 t canelha for the average crop length of about 18 months. On a Waldene sandy loam the mean response was found to be 8 t canethalyear over four ratoon crops in a single crop cycle.

These substantial increases in cane yields when a trash blanket was used in rain-fed areas of the Natal Coast were shown by Thompson to be due primarily to the conservation of moisture through reductions in evaporation and run-off from the bare soil and the advantage was gained almost entirely during the period before a full canopy was formed. There was no benefit from a trash blanket when supplementary irrigation water was applied. Small improvements due to the trash mulch were measured in certain physical characteristics of the top soil. The trash mulch did not contribute to the nutrient content and instead there was some evidence that the response to trash was slightly greater in the presence rather than in the absense of fertilizer.

In 1947 an experiment was established on a fine sandy loam soil to compare the effects on ratoon cane yields of a trash blanket; burning and raking the tops off the plots; and burning with the tops left unraked. The results from three ratoon crops were reported by Pearson who showed that the mean yield response to a trash blanket was 14%, and 8% to the burnt tops left on the plots when compared with the burning and raking treatment. The benefits obtained from leaving the tops unraked did not have any practical significance at that time because the land had to be cleared so that weed cultivators could be used.

The trend towards increased burning of cane, the introduction of effective herbicides and the reported yield response from leaving burnt tops unraked (Pearson ') prompted further investigation into practices that might reduce the harmful results of burning.

MATERIALS AND METHODS

Soils

Experiments were conducted on a range of soils with markedly different cha- racteristics such as texture, depth and drainage. Brief descriptions of the relevant soil series (Macvivar 3, are:

Rydalvale: A black montmorillonitic clay over decomposing dolerite parent material; good depth and moderately well drained.

Msinga: A deep red porous sandy clay loam, with no visible structure but excellent internal drainage. Swartland: A dark grey porous sandy clay loam merging via illuvial tongues to weathered Middle Ecca sandstone; moderately deep and no drainage problems. Shortlands: A deep red, moderately porous clay with blocky structure; good internal drainage. Inanda: A deep, humic, dark brown, loamy clay topsoil, overlying a deep, structureless, well drained red subsoil. Clansthal: A very deep red porous loamy sand; excellent internal drainage. Fernwood: A very deep grey porous sand; excellent internal drainage, but commonly associated with serious nematode problems. Waldene: A shallow grey fine sandy loam on a porous layer of iron concretions overlying heavy clay; poorly drained and highly erodible. Glenrosa: A grey loamy sand merging via illuvial tongues to weathered granite; shallow to moderately deep; no drainage problems, highly erodible. Cartref: A grey sand overlying Table Mountain Sandstone, generally shallow with drainage problems at a sandlrock interface erodible.

Plot Size

The burning and trashing treatments were carried out on plots which were larger than usual and comprised eight cane rows, 10-12 m long and 1,4 m apart. The net harvested plots comprised the inner four rows with 1,O or 1,5 m discards at each end of the plots. Where tractor-drawn equipment was required. to rip and to form basins, a row length of 18 m was necessary.

In trials where the burnt tops were raked off and the effects compared with leaving the scattered tops, the normal plot size of six cane rows, 10 m long and about 1,4 m apart was adequate because the side-raking of tops was done by hand.

Desing

Latin Square or Random block designs were generally used and replicated six times.

Burning and Trashing

The plots were not large enough to allow the cane in the appropriate plots to be burnt whilst it was standing because the fire would easily jump the breaks into control plots scheduled for trashing. It was therefore necessary to cut the un- burnt cane in all plots first, to weigh the cane, to clear the breaks between all plots of trash, and then to burn the trash in the appropriate plots a few days later. This method of burning is usually less efficient than the burning of standing cane and a greater mass of singed tops remains.

Sometimes the singed tops had to be spread evenly over the soil, but generally little disturbance was necessary to obtain a reasonable distribution. The extent to which the singed tops covered the soil was assessed visually and1 varied between 20 and 70%, depending of factors such as season, the amount of dew on the foliage, the age of the cane, and the size of the crop.

346

I Weed Control

The herbicides used on various cccasions (in kg or L of product per ha) were: diuron + Sencor (2 + 2); diuron + 2,4-D or MCPA or Actril DS (2,5 + 2 or 4 or 1,25); Lasso + atrazine (5 + 2); and Velpar + diuron (0,5 + 2,O). Wherever herbicides were used, they were applied to all plots to avoid bias in the event of any phytotoxic effects on sugar cane. Hand hoeing was carried out where necessary.

( Fertilizer

Unless otherwise stated, the nutrients applied were those recommended by the Fertilizer Advisory Service of the Experiment Station of the S. A . Sugar Associa- tion and the same amounts were applied to all plots regardless of the treatments.

Rainfall

The average rainfall in the areas where the experiments were conducted is between 800 and 1 000 mmtyear and the distribution is generally poor. Since rainfall is such an important factor when assessing the merits of a soil mulch, the rainfall recorded for a particular crop is generally given here as the percentage of the long-term mean rainfall for the area (L. T . M.) .

I I Basing- Listing

In the trial were basin-listing was tested, the basins were made by a tractor- mounted implement built by the Engineering Departament of the S. A . S. A.Ex- periment Station, Mount Edgecombe (Anon I ) , and based on the design of a tie-ridger built by the National Institute of Agricultural Engineering, Silsoe, En- gland. The basins were formed in the inter-row. The average depth was 100- 180 mm and the average diameter about 60 mm (Fig. 1). When the soil was not friable enough it was necessary to rip the inter-row with spring tines befores attempting to form basins.

RESULTS AND DISCUSSION

Basin- Listing

First and second ratoon crops growing in a shallow Glenrosa soil at Shaka's Kraal on the North Coast were used to test the effects of basin-listing or a trash blanket compared with raking off the singed cane tops after burning. The trash and tops in the plots used to test basin-listing were also burnt and raked off. The basins were formed twice in each crop; soon after harvest and again before the leaf canopy was formed. The results are given in Table 1.

In the first ratoon crop there was no response to a trash blanket or basin-listing probably because the rainfall distribution was good and the yields were moderate.

1 i"a61e 1. Yield Results From Two Ratoon Crops Whlch Recelved Various Water Conservat~on Treatments

t Canelhalyear t Sucroselhalyear Treatments

1R 2R 1R 2R I

I Trash blanket 62 54 * 8 2 6,8 i Basin-listed 64 50 8,7 6,s Bare soil 62 42 8,O 5 ,4 I

Mean 63 49 8,3 6,3 , LSD (P = 0,05) 8 10 0,95 1,40

* Statistically significant at the 5% level.

Rainfall distribution was poor during the second ratoon crop and a substantial yield response of 12 t canelhatyear (P = 0,05) resulted from the trash blanket, and , 8 t canelhalyears (ns) from basin-listing. The basins held water effectively after rain. O n such relatively poor soils, which have low infiltration rates and water- holding capacities and where the land slopes appreciably (the experiment site had an average slope of 15%), basin-listing can apparently be beneficial but not as effective as a full trash blanket. 1

I

Figure I . Basin formed by basin listing implement

Cylindrical gypsum resistance blocks were installed at depths of 150, 450, 750 and 1 050 mm and the readins showed that there was an increased soil moisture content under the trash blanket, but there was very little change due to the basin-listing treatment. Using run-off lysimiters on a similar soil with an 11% slope at Shakaskraal, Thompson showed that basins were as effective as a trash blanket in reducing run-off but they are unlikely to have an effect on evaporation losses. Basin-listing is therefore a practice which should help to conserve soil and water in the absence of a trash blanket.

Unfortunately, once the basins have been formed, subsequent in-field tractor operations become difficult because of the bumps. Before applying herbicide or fertilizer mechanically, the basins would have to be broken and reformed again to avoid harm to the operator and damage to the tractor. The operation would be difficult to conduct on slopes greater than 20%.

A comparison on a Waldene Soil of the Effects on Ratoon Cane Yields of a Trash Blanket; Burning and Leaving the Burnt Tops Scattered; and Raking off the Tops

The experiment conducted by Pearson at Shakaskraal was continued for another six ratoon crops and the yields obtained from all seven ratoon crops are shown in Fig. 2.

Trash blanket . . .

Burnt, tops raked off 7

Irnt, tops left scattered -

1 2 3 4 5 6 7

Number of ratoon crop

Figure 2. Yields in t canelhalyear from seven ratoon crops which received trash treatments.

If the plots where the trash was burnt and the tops raked off are used as a control, there was a response to the burnt tops left scattered in all of the seven crops, and to the trash blanket in six out of seven crops. The mean responses in t canelhalyear to a trash blanket in six out of seven crops. The mean responses in to are halyear to a trash blanket and burnt tops left scattered for the seven ratoon crops were 5,O and 4,7 respectively, which represent 9% and 8,5% increases over

burning and raking. Thompson compared the effects on cane yield of nil, 16, 36, and 56 t of trash per ha, and found that the lowest amount of trash was best, yields decreasing with increasing quantities of trash. It was also shown by Thompson that a trash blanket intercepts as much as 50% of rainfall totalling 2,5 mm but only 9% of a 25 mm rainfall. It is not surprising therefore that small quantities of crop residue were beneficial to the following crop.

A Comparison of the Effects on Ratoon Cane Yields of Burnt Tops Raked off or Left Scattered

In order to compare further the effects on ratoon cane of ranking burnt tops into windrows compared with leaving the tops scattered, the two treatments were introduced into a number of experiments which had already been established to investigate subjects such as nematicide efficacy, herbicide phytotoxicity, variety performance and nutrient responses. By applying each treatment to half the number of replications, comparisons were made on 14 different crops. The res- ponses in terms of tons cane per hectare per annum to the burnt tops left scattered compared with burnt tops raked into windrows are shown in Table 2 for ratoon cane grown on clay, loam and loamy clay soils.

Table 2. The Response of Ratoon Cane, Grown on Loamy Clays and Clay Loam Soils, to Burnt Tops Left Scattered Compared with burnt Tops Raked Into Windrows (Control Plost)

Rainfall as % of Response to Yield of

L. T. M. Scattered Tops Soil Month Control Plots

Crop (site) Cut First Whole

tclhalyear tclhalyear % 4 mths Crop

Rydalvale 2R November 147 109

(Mt. Edgecombe) Rydalvale

3R January (Mt. Edgecombe) Shortlands

4R November 146 106 73 Nil Nil (Mtunzini) Inanda

1R September 95 89 (U. Tongaat) Inanda

2R February 105 109 (U. Tongaat) Inanda 3R August 8 1 99 65 (U. Tongaat) Mean 75

* Statistically significant at the 5% level.

The average response is slightly ower than that on the Waldene soil at Shakas- kraal for seven ratoon crops. The largest responses to burnt, tops left scattered (+ 7 t canelhalyear, P = 0,05) were obtained from crops cut in late summer, whilst a small negative effect was measured when the tops were left scattered in early summer at the high altitude of Upper Tongaat. Rainfall on the crop, expressed as a percentage of the L. T. M. was not always well correlated with the magnitude of the treatment response.

The responses obtained to burnt tops left scattered on sandy soils are given in Table 3 .

Table 3. The Response of Ratoon Cane Grown in Sandy Soils to Burnt Tops left Scattered Compa- red With Burnt Tops Raked Into Windrows (Control Plots)

Rainfall as % of Yield of

Response to L. T. M. Scattered Tops

Soil Control Plots Crop Crop Start

(site) First Whole

tclhalyear tclhalyear % 4 mths Crop

Fernwood 2R November 191 131 62 - 1 - 2

(Empangeni) Fernwood

2R November 129 105 105 + 3 +3 (Mt. Edgecombe) Fernwood 3R November 100 101 101 - 2 - 2 (Mt. Edgecombe) Fernwood

4R November 49 60 38 NII (Mt. Edgecombe)

Nil

Clansthel 3R September 142 118 114 - 2 -2 (Umhlanga) Clansthal

2R April 88 119 86 -20 * - 23 (Umhlanga) Clansthal

3R August 110 93 94 +3 (Umhlanga)

+3 .

Cartref 3R November 110 90 63 + 5 +7

(Paddock) Mean 83 - 2 - 2

* Statistically significant at the 1% level.

Generally the effect of burnt tops left scattered was negligible on sandy soils but there was an indication (ns) of a positive response in a ratoon crop cut in November on a Cartref soil, and a severe yield depression from a crop treated in autumn on the Clansthal loamy and. The latter effect was probably due to the iron chlorosis that trash or burn&ops caused in ratoon cane grown on an alkaline sandy soil.

A Comparison on Various Soils of the Effects on Ratoon Cane of a Trash Blanket, Burnt Tops Left Scattered and Burnt Tops Raked Into Windrows

In order to compare the response of ratoon cane to a trash blanket with that from burnt tops left scattered, another series of experiments on six different soils was established. The three treatments were, a trash blanket, burnt tops left scattered and burnt tops raked off. The results from nine different crops have been evaluated but those of a ratoon crop cut in May on the Clansthal loamy sand, where yields were markedly suppressed by both the mulch treatments because of iron chlorosis, have been excluded. On average, the response to a trash blanket was only 6 tcanelhalyear (9%) and 5 tcanelhalyear (7%) from a trash blanket and

35 1

burnt tops left scattered respectively. By classifying the soils at the experiment sites into ((good>) (moderately deep and well drained, Table 4) and <<poor>> (sha- llow, poor to moderately drained, or associated with iron chlorosis, Table 5 ) and by sub-dividing the crops into either a summer or winter start, the trends of the responses can be studied.

Table 4. The Response of Ratoon Cane, Grown in aGoodx Soils, to a Trash Blanket or Burnt Tops Left Scattered Compared With Burnt Tops Raked Into Windrows (Control Plots)

Response to Rainfall as

% Ground Yields of % of L.T.M. Trash Burnt Tops

Soil Cover by Control Crop Blanket

(Site) Scattered

Burnt Plots First Full

Tops t canelhalyear 4 mths Crop t cane1 , t cane1 ol

10 . halyear halyear "

Crops starting in summer

Rydalvale (Mt.Edgecombe) 1R 65 66 70 46 +14* +30 +8 +17 Msing a (Tongaat) 2R 109 85 60 114 +6* +5 +6* +5 Msinga (Tongaat) 3R 35 84 40 57 +9* +16 +3 +5

Mean 72 + lo +14 +6 +8

Crops starting in winter

Rydalvale (Mt. Edgecornbe) 2R 106 118 40 88 +7 +8 +8 +9 Swartland (Renishaw) 2R 61 97 20 75 -5 -7 -2 -3

Mean 82 +1 +1 + 3 +4

* Response statistically significant at 5% level.

For ratoon crops starting in summer there is evidence that the response to a trash blanket and tops scattered is greater on the better soils. This confirms the results of the two long-term experiments conducted on the Rydalvale and the Waldene soils at Mount Edgecombe and Shakaskraal respectively where the mean response to a trash blanket from 16 ratoon crops on the Rydalvale soil was 12 tclhalan and 6 tclhalan from 11 ratoon crops on the Waldene soil ( ~ o b e r l ~ ~ ) .

For ratoon crops starting in summer, tops left scattered had about 60% of the effect of a full trash blanket.

For ratoons starting in winter the results indicate a comparatively small respon- se in two crops and a yield depression in two crops. The scattered tops tended to be more beneficial than the trash blanket in winter cut crops, esumably because

352

1 Table 5 . The Response of Ratoon Cane Grown In Medlum to Poor Solls, to a Trash Blanket o r Burnt Tops Left Scattered Compared With Burnt Tops Raked Into Windrows (ContrOl Plots)

Response to Rainfall as

% Ground Yields of % of L.T.M. Trash Burnt Tops

Soil Cover by Control Crop Blanket Scattered

(Site) Burnt Plots First Tops t cane/ha/year

4 mths Crop t cane1 , t cane1 o,

halyear " halyear "

Crops starting 1 j n summer

Waldene (Compensat~on) 1R 49 8 1 25 40 +4* + I 0 + 3 +8 Glenrosa (Renlshaw) 1R 57 54 20 44 + lo* +23 + 6 + I 4 Clansthal (Lmhlanga) >R 76 71 40 70 +5* I * +2 + 3

Mean 5 1 + 6 + I 1 +4 + X

Crops starting in winter

Glenrosa (Renlshaw) 2R 87 96 30 68 +5* +7 + 1 1 * * +t> Clansthal (Lmhlanga) 4R 30 108 45 67 - I X * * -27 -12** -18

' Response s ta t~s t~ca l ly $lgn~flcant at 5% level '" Response s t a t ~ s t ~ ~ a l l y \~gnll lcant at 1 % level

the scattered tops lowered soil 'temperatures less than the trash blanket. 'The reduced yield caused by both mulch treatments in the experiment at Kenishaw was thought to be associated with the crop starting at the beginning of winter and the site having a south-east aspect. The effects of' the various treatments on stalk population are shown in Figure 3.

Where a posi.tive response to a trash blanket is obtained, it is normal to find a lower stalk population of' taller and heavier stalks ('I'hompson (') comp:~red with the cane where the previous crop was burnt and the tops raked off. Where burnt tops were left scattered, the ef'f'ect on stalk population tended to be intcrmcdiate between that of' a trash blanket and burning with the tops raked off'. 'I'his is illustrated in Figure 4 which refers to the second ratoon crop on the Hydi~lvi~lc clay soil at Mount Edgccombc.

Trash blanket

a.**--m9 a * ............. Burnt, tops raked off

4 - \** , , , , , Burnt, tops left scattered Y-

I ~~/!L-- 40 I I 1 1 1 I

A S O N D J F M A M J J A S

Rainfall, 512 1/80 to 91518 1

O ~ c t u a l rainfall

Month

Figure 3. The effects on stalk counts of varlous trash treatments on a ratoon crop startlng In wlnter

/ at Renishaw.

Weed Control

In many experiments where burnt tops were left scattered, a1range of common- ly used herbicide mixtures, as described under <<Materials and Methods)) were applied, and many others have been used under field conditions. The large

' quantities of crop residues generally did not reduce the efficacy of herbicides. 'Only where herbicides were applied to fields where there was a large amount of

354

0.2

0 , , , , , , , , , , , , 80

N D J F M A M J J A S N D J F M A M A S

Month Month

Figure 4. The effect on stalk height and counts of various treatments on a ratoon crop starting in summer I

at Mount Edgecombe. I I

fresh powdery ash was the effectiveness of the herbicides reduced. Soil disturban- ce caused by side raking burnt tops into windrows invariably stimulated germina- tion of weeds while leaving the burnt tops unraked tended to reduce germination. It is therefore advisable to wait for rain to settle the ash and crop residues before applying herbicides onto the scattered tops.

Nutrition

The nutrient content of trash and tops varies with variety and many agronomic practices. Thompson ' reported that variety NCo 376, yielding 90 t canelha, contained 75 kg nitrogen, 8 kg phosphorus and 85 kg potassium in the tops and trash. There is no evidence to suggest that these nutrients become available in sufficient quantities to contribute significantly to the following crop.

Soil analysis indicated a slightly higher exchangeable potassium in plots where the cane was burnt compared with where it was cut green and trashed. After 23 years of cropping and no fertilizer application on a Rydalvale series soil the exchangeable K was 126 ppm where the cane was trashed and 140 ppm where it was burnt. The difference between burning and trashing on soil K was even smaller after fifteen years of cropping on a Waldene series soil but the trend was the same. In the most recent ratoon crop of the long-term burning and trashing experiment on the Rydalvale soil at Mount Edgecombe, the third leaf K Oh dry matter of cane in the plots that had not been fertilized since 1939 was 0,85 in the presence of a trash blanket; 0,88 where burnt tops were left scattered; and 0,67 where the burnt tops were raked off. The contribution made by the nutrients in the crop residue is small and not of any practical significance.

CONCLUSIONS

If burning of cane before harvest should become necessary for economic rea- sons in the rain fed areas of the Natal sugarbetl, basin-listing would help to conserve soil and moisture but it is not considered to be a practical solution, particularly on steep slopes.

Leaving the burnt tops scattered was about 60% as effective as a trash blanket in terms of increasing yield in crops starting in summer. Omitting the practice of side-raking will reduce costs, decrease soil erosion and necessitate the use of herbicides, which were found to be equally effective when burnt tops were left scattered or raked into windrows.

For ratoon crops cut in winter the tops left scattered were more effective than a trash blanket in terms of cane yield, but both mulch treatments were less effective than they were in summer.

The responses in summer to a trash blanket and burnt tops left scattered was greater on "good" soils (10 and 6 t canelhalyear respectively) than on "poor" soils (6 and 4 t canelhalyear respectively). There was only'a small response to a trash blanket or burnt tops on the sands and loamy sands.

Both mulch treatments severely aggravated the problem of ratoon chlorosis when applied to alkaline sandy soils.

Responses have been measured only in terms of cane yield but Thompson showed that in twelve crops there was not a significant difference between the sucrose contents of cane that had been burnt or trashed at the time of the previous harvest. The greatest long-term benefit from using a soil mulch is the one of soil conservation particularly in the relatively steep areas of the Natal sugarbelt. It has unfortunately not been possible to measure the soil loss under the different treatments but it has been observed that a comparatively small amount of crop residue affords the soil a great deal of protection from raindrop bombardment and subsequent erosion.

REFERENCES

1 Anon (1970) SASA Expt Stn Ann R e p , 1969-1970, p 15 2 Macvlvar, C N (1937) Soils of the Sugar Industry S Afr Sug Ass Expt Statlon Bulletln

No 19 3 Macvlvar, C N , D e Vllllers, J M, Loxton. R F and Vester, E. (1977) So11 Class~f lca t~on A

Blnomlal System for South Afrlca Dept of Agrlcultural Technical Servlces p u b l ~ c a t ~ o n 4 Moberly, P K (1977) Trashlng Versus Burnlng of Sugar Cane Natal Reglonal Newsletter

Dept of Agrlcultural Techn~ca l Servlces publlcatlon 5 Pearson, C H 0 (1959) The Conception of Farmlng on Subt rop~ca l Soils Wlth the Use of the

Crop by-product, Trash, as a Means to Improve Ylelds Proc of the X Congress of the ISSCT Hawall, pp 578-588

6 Thompson, G D (1965) The Effects of Trash Conservation on So11 M o ~ s t u r e and the Sugar Cane Crop In Natal Proc S Afr. Sug Technol Ass . 39:143-157

7 Thompson, G D (1966) The Production of Trash and ~ t s Effects as a Mulch on the So11 and on Sugar Cane Nutrltlon Proc S Afr Sug Technol. Ass . 40:143-147

8 Thompson, G D (1966) Trash Mulching-~ts Effects on Soils and Crops I A South Afrlca World. Crops June, 1966, pp 62-65

9 Wllson, J (1960) Trashlng. S Afr Sug Ass. Expt S t a t ~ o n Bulletln No 12

QUELQUES PRATIQUES CULTURALES DE POST- RECOLTE SUR LES REPOUSSES DANS DES REGIONS

A IRRIGATION PLUVIALE DE LA CEINTURE SUCRIERE DE NATAL

P. K. Moberly et R. K. McIntyre Station expkrimentale de 17Association du sucre d'Afrique du Sud, Mount Edgecombe

4300, Afrique du Sud

Vu la predominance croissante du briilage sur l'kpaillage dans des re- gions 2I irrigation pluviale des terres bases de la c6te de Natal, les auteurs ont examine d'autres mkthodes pour la conservation du sol et de l'humidi- tC. La formation de creux s'est avCrCe modkrkment efficace, mais elle ne peut etre considerke comme une solution de remplacement B la couche de paille aussi efficace qu'une couche de bouts brfiles disperses sur le sol. Le rendement en cannes des repousses couvertes d'une couche de bouts briiles a kt6 environ 60% aussi efficace que dans le cas cas des repousses couver- tes d'une couche serree de paille. Durant les mois d'hiver, la mCthode des bouts briilks s'est rCv6lee supkrieure 21 celle de la couche de paille en termes de rendement en cannes, et la reaction des repousses B une couche protectrice a kt6 notablement infkrieure.

Le rendement des repousses cultivkes dans des sols modkrement pro- fonds et bien drain& est meilleur avec un couche de paille ou lorsque les bouts brfilks son disperses, qu'avec des sols peu profonds i faible drainage ou fortement sablonneux. La couche protectrice, quelle qu'elle soit, des sols sablonneux alcalins a provoquk la chlorose des repousses. Les auteurs ont mesure l'effet du traitement en termes de rendement en cannes seule- ment, mais ils n'ont pas determine quantitativement les effets les plus importants B long terme sur la conservation du sol et de l'eau. Pourtant, maintes observations r6alisCes sur le terrain indiquent que l'kcoulement est reduit substantiellement par n'importe quelle couche protectrice, mCme si celle-ci est peu Cpaisse.

ALGUNOS METODOS DE A T E N C I ~ N POSTERIORES A LA COSECHA RESPECT0 A LA CANA DE RETONO

EN LAS AREAS DE ABASTECIMIENTO PLUVIAL DEL C I N T U R ~ N AZUCARERO DE NATAL

P. K. Moberly y R. K. McIntyre Estacidn Experimental de la Asociacidn Azucarera de Africa del Sur,

Mount Edgecombe 4300, Africa del Sur

RESUMEN

Debido a la creciente incidencia de la quema con preferencia a1 despaje en las Areas de abastecimiento pluvial de las Tierras Bajas Costeras de Natal, se investigaron mCtodos alternos para la conservation del suelo y la humedad. La catalogaci6n por cuencas demostr6 ser moderadamente efectiva, per0 no se consider6 tan practicable para la alternativa de un colch6n de paja como la cobertura propiciada por 10s cogollos quemados que estiin diseminados sobre la superficie del suelo. En tCrminos del rendi- miento de la caiia, la respuesta de las cosechas de retoiios comenzadas en el verano, respecto a 10s cogollos quemados, fue alrededor del 60% tan efectiva como un colch6n completo de paja. En 10s meses mas frios del invierno, la diseminaci6n de 10s cogollos quemados result6 superior a1 colchon de pajas en terminos del rendimiento de la caiia y la respuesta de las cosechas de retoiios a una cobertura protectora del suelo fue considera- blemente inferior.

Las respuestas del rendimiento respecto a cosechas de retofios cultiva- dos en 10s suelos moderamente profundos y bien drenados son mayores con un colch6n de pajas o cuando 10s cogollos quemados estAn diseminados, en comparaci6n con 10s suelos poco profundos y deficientemente drenados, o aquellos que eran sumamente arenosos. Una capa protectora de cualquier clase sobre suelos arenosos alcalinos foment6 la clorosis en 10s retoiios. Los efectos de 10s tratamientos se valoraron en tCrminos del rendimiento de la caiia dnicamente y 10s efectos miis importantes a largo plazo sobre la conservaci6n del suelo y el agua no se evaluaron cuantitativamente. Sin embargo, muchas observaciones efectuadas bajo las condiciones del campo indican que el escurrimiento se reduce sustancialmente mediante cualquier capa protectora del suelo, incluso cuando Csta se encuentre en pequeiias cantidades.