PEDOLOGIE Bulletin de la Société BeIge de Pédologie, édité avec I 'aide financière de la Fondation Universitaire et des Ministères Belges de I 'Education et de la Culture Française et Néerlandaise Tijdschrift van de Belgische Bodemkundige Vereniging, uitgegeven met de financiële steun van de Universitaire Stichting en van de Belgische Ministeries van Opvoeding en Nederlandse, resp. Franse Cultuur
1986
XXXVI-l
Comité de Rédaction - Editorial Board - Redactiecomité P. Bullock (Rothamsted, U.K.), J. D'Hoore (Leuven, België) R. Dudal (Leuven, België), R. Frankart (Louvain-Ia-Neuve, Belgique), M. Girard (Grignon, France), G. Hanotiaux (Gembloux, Belgique), M.H.B. Hayes (Birmingham, U.K.), A. Herbillon (Vandoeuvre-Nancy, France), U. Schwertmann (Munchen, BR. Deutschland), C. Sys (Gent, België), M. Van Ruymbeke (Gent, België) Chief-editor : W. Verheye (Gent, België)
0/1986/0346/1
PEDOLOGIE is edited by the Belgian Society of Soil Science, Krijgslaan, 281, B-9000 Gent, Belgium. Subscript ion : 800 BF net per year (banking costs at subscribers expense) + mailing costs. Additional information to : The Chief-editor of Pedologie, Krijgslaan, 281, B-9000 Gent, Belgium.
2
BELGISCHE BODEMKUNDIGE VERENIGING
SOCIETE BELGE DE PEDOLOGIE
Samenstelling van de Raad van Beheer voor 1986 Composition du Conseil d'Administration pour 1986 Président Voorzitter
L. Baert
Vice-présidents Ondervoorzitters
R. Gombeer L. Mathieu
Secrétaire général Secretaris-generaal
w.
Secrétaire-trésorier Secret ar is-scha t bewaarder
R. Vermeire
Verheye
Membres Leden
L. J. G. K.
PRESIDENT D'HONNEUR
SECRETAIRES GENERAUX HONORAIRES ERE-SECRETARISSEN GENERAAL
ERE-VOORZITTER
Bock, J. Chapelle, J. Dufey, feyen, J.P. Heek, G. Hofman, Stoops, A. Van der Beken, Vlassak, Oudal.
R. Tavernier J. Ameryckx C. Sys
J. Baeyens
ANClENS PRESIDENTS OUD-VOORZITTERS V. f. L. G. A. G. L. A. G.
Van Straelen t Jurion t De Leenheer t Manil t Van den Hende Scheys Sine t Cottenie Hanotiaux
(1950-1953) (1954-1955) (1956-1957) (1958-1959) (1960-1961) (1962-1963) (1964-1965) (1966-1967) ( 1968-1969)
M. A. P. J. M. R. f. J.
De Boodt Herbillon Avril D'Hoore Van Ruyml;>eke frankart Delecour Lozet
(1970-1971) (1972-1973) (1974-1975) (1976-1977) (1978-1979) (1980-1981) ( 1982-1983) (1984-1985) 3
PEDOLOGIE, XXXVI-I, p. 5-15, 4 tables, Ghent, 1986.
ANAL YSIS OF THE SUITABILITY OF BELGIAN SOILS FOR GROWING FINE CARROTS J. VANDAMME R. BISTON
Abstract A suitability study on soils which grow fine carrots was assessed by intensive sampling of some 266 field plots, randomly clistributed over 8 soil classes. The study lasted 4 years, As product ion parameters, not only the gross yield on each plot was determined but also the marketabie yield (being the fraction of carrots with a diameter between 13 and 25 mm) and the canning yield (being the peeled marketabie yield). The statistical analysis of the data revealed a st rong annual effect on yield. Although the sandy silt loam soUs were most productive over the 4 year period it was assumed that the more sandy soils be the most suited for growing carrots, provided optimal water . conditions were present. Key-words Fine carrots, soU suitability.
1. INTRODUCTION
The paper analyses the suitability of Belgian soils for the production of fine carrots. The current research was similar in approach to previous soil suitability studies conducted by the Cent re for Horticultural Soils of the Faculty of Agricultural Sciences, Katholieke
J. Vandamme - Senior researcher at the Centre for Horticultural SoUs, Faculty of Agricultural Sciences, Katholieke Universiteit Leuven, Kardinaal Mercierlaan 92, 3030 Leuven (Heverlee), Belgium. R. Biston - Former research assistant at the National Institute for the Improvement and the Conservation of Vegetables (INACOL), Wezembeek-Oppem, Belgium. 5
Universiteit, Leuven, Belgium. Since more than 20 years, this cent re evaluates the suitability of Belgian soils for different horticultural crops, like strawberries, asparagus, Brussels leave, tomatoes, salsify, bush beans and lettuce. In Belgium, carrots are mainly grown for the canning industry and to some extent for table consumption. The annual production amounts on average nearly 100,000 tons of which only 15% are not canned. The area occupied for the product ion of fine carrots is roughly 2,000 ha, representing 5% of the total Belgian horticultural area reserved for vegetable growing. The annual gross income of this area represents more than 300 million Belgian francs (L.E.I., 1983). 2. MATERlALS AND METHODS The method adapted by the Centre of Horticultural Soils to evaluate the suitability of soils for growing vegetable or soft fruit crops is mainly based on a detailed field survey of a large number of field plots. For this purpose fine carrots of the cultivar "Amsterdamse bak" have been grown on a number of field plots, under normal farming conditions. During 4 successive growing seasons, some 266 plots, spread at random over 8 soil classes, have been monitored. In Belgium fine carrots are traditionally sown between April 25 and June 15. They are harvested in August and September, having a growing season of 3 to 4 months. The fertilizer application on the studied plots is on average 90 kg/ha nitrogen, 108 kg/ha phosphorus and 150 kg/ha potash. All the plots have been treated with birlane to control carrot fly attack and with herbicides on the base of linuron to keep the plots weed free. The plots clearly infected by nematodes have not been taken into consideration. The fine carrot producing area in Belgium represents 8 different soil classes. According to . the local Belgian soil classification procedure (Tavernier and Maréchal, 1958) those 8 soil classes can be subdivided into four textural (Verheye and Ameryckx, 1984) and two drainage classes. The drier soils comprise the weIl drained profiles with a winter water table below 60 cm. The more humid soils do have a shal10w water table in winter, while in summer the water table drops to about 120 to 150 cm bel ow surface. The texture of the soils analysed ranges from sandy silt loam to sand. The soils have not been classified according to profile differentiation, because in most situations the typical diagnostic features of profile development are faded away by repeated deep ploughing. The climate of the 4 seasons during which the field survey was conducted can be characterized as follows. In 1972 the growing season for fine carrots, from May throughout August, has been more 6
wet, overcast and cooler than normal. Only 625 hours of sunshine were recorded versus 817 hours for a normal year. The mean daHy temperature (14.9°C) was slightly below the normal mean temperature (15.2°C) and about 22 11m 2 more rainfall was recorded than in a normal year, during which 293 11m 2 rain is accumulated. In 1973 sunshine duration was quite normal (820 versus 817 hours), the ave rage temperature was slightly above normal (16.7 versus 15.2°C) and about 80 11m 2 less rainfall was recorded. The growing season in 1974 can be classified as quite normal regarding air temperature and rainfall. Thé season was however less sunny; about 60 hours less sunshine was recorded. The driest, warmest and sunniest season of the observation period was the 1975 season. The total rainfall recorded was 148 11m 2 versus 293 1Im 2, the average air temperature rose about 1.5°C above the norm al mean air temperature and sunshine was 43 hours more. The climate during the observation years (1972-1975) was thus not extreme. In 1972 the weather w"as on average slightly below normal, in 1975 above normal. The crop studied has not been exposed to extreme climatic conditions. Only in 1973 and 1975, depending from the water delivery capacity of the soit the fine carrot crop might have been exposed to water stress. This stress, if present, was more pronounced during the 1975 season than during the 1973 season. Each season, a number of farming and pedological observations were registered on the field plots selected over the different soH types. The survey always took place at harvest time. The size of the area sampled on each plot was about 2 m 2 large. The leaves were removed and the roots harvested manually. The harvested roots were cleaned, graded in 6 classes based on the diameter of the crown « 13 mm, 13-16 mm, 16-20 mm, 20-25 mm, 25-30 mm and > 30 mm), and the yield per class was determined. The carrots of the classes 13-16 mm, 16-20 mm, 20-25 mm and 25-30 mm were peeled and canned in order to define the canning yield of each plot. In addition to the larger sample a smaller sample of 30 plants (Ieaves and roots) and a soit sample were taken for examination and analysis in the laboratory. Besides some physical measurements on those plants like length, crown diameter and deformations, the leaves and the roots were screened on the N, P, K, Mg and Ca content. The canned roots we re only analysed on the nitrate content. On the soH sample taken of the plough layer following chemical parameters were determined pH-H 2 0, the concentration of P, K, Mg and Ca in mg/l00 g dry soil. 3. RESUL TS AND DISCUSSION The observed yield data, like net, marketable and canning yield 7
Table 1. The annual net product ion of fine carrots in kg per 100 m 2 and the num ber of field plots samples in function of the soil class. Year Parameter
Sandy silt loam moist dry
Light sandy loam dry moist
Sand -
Loamy sand dry
moist
dry
moist
1972 Number of plots Yield
2 848.5
4 794.3
7 817.1
7 759.7
9 815.4
2 4 771.0 1021.8
2 597.0
1973 Number of plots Yield
6 725.4
10 775.0
8 650.0
7 662.6
10 810.2
5 752.4
15 813.2
7 854.6
10 8 9 10 1974 Number of plots 1576.1 1164.5 1049.4 1161.5 Yield
11
764.9
12 765.6
12 602.6
8 761.6
15 678.5
18 750.5
13 681. 7
10 790.8
1975 Number of plots Yield
4 747.7
5 664.5
8 834.0
8 718.4
are markedly influenced by the climatic situation. As can be seen in table 1, giving the annual average net production per soil class, the average net yield varies considerably from year to year. It can be observed as weIl that yearly the maximum and minimum yields are not always obtained on the same soil class. Even a given soil class provides in one year the highest yield, and in another year the lowest one. The high variation is mainly due to miscellaneous influences, the effect of which partially is controlled by the number of repetitions per soil class. In order to normalize the measured yields those were corrected for climatic differences. Of the field plots within a soil class the product ion was averaged and compared to the average production harvested on those plots in 1975. The product ion of last year was taken as reference for establishing the climatic correct ion factor. In a next step the ratio of the average product ion for a given year to the average product ion of the year 1975 was multiplied by a weighing factor to account for the unequal number of field plots analyzed from year to year. The weighing factor was calculated according to the formula k 1k 2/(k 1+k 2), where k 1 and k 2 are the number of the field plots examined in the years of which the annual product ion is compared. As aresuit for each observation year, with exception for the reference year, a correct ion factor per soil class (8 in total) was obtained. Each year, these correct ion factors were further averaged over the soil classes, wi th the sum of the weighing factors as divisor. The same procedure was applied to the total yield collected over the four years. The average annual value of the
8
net yield, adapted as indicated above, is listed in table 2, together with its relative value with re gard to the average annual net yield obtained in 1975. Finally, the climatic correction factor by which the individual net yields were divided each year was calcuJated by putting the relative value of the total yield equal to 100 and by adjusting the relative value of the adapted, average annual net yield of each observation year to this number. In this way the correct ion factors were obtained in order to balance the net yield for the year effect, as listed in table 2. A similar procedure was applied to define the climatic correct ion factor for the marketabie and canning yield.
Table 2. Average annual net, marketabie and canning production as observed over the different soil classes analysed. The net product ion is expressed in kg per 100 m 2, while the marketabie and canning productions are expressed as a fraction of the net yield, the yield parameters being corrected for the year effect. Year
Number of Ada12ted net }::ield in plots kg/100m 2 % of the 1975 yield
Correct ion factor to balance the year effect
Marketabie yield in % of the net yield
Canning yield in % of the net yield
1972 1973 1974 1975
37 68 80 81
839.3 788.0 920.1 741.6
113 106 124 100
1.023 0.960 1.121 0.940
76.7 72.4 72.4 74.1
55.7 58.4 53.4 52.4
Total
266
820.7
111
1.000
73.4
54.7
In table 2 the average corrected marketabie and canning yields are expressed in percent of the average corrected net yield. The marketabie fraction ranges between 72.4 and 76.7%, while the canning yield varies between 52.4 and 58.4% of the net yield. About 1/4 of the harvested roots have a diameter smaller or larger than the acceptable diameter for canning purposes, and another 1/4 of the marketabie yield is lost in the peeling process, resulting finally in an average canning fraction of 55% of the net yield. The corrected yield parameters were further subjected to a classic analysis of variance in order to specify the effect of soil class, texture and drainage class, on the soil productivity with regard to fine carrots (Deckers et al., 1980). The results of this test are given in table 3. This table summarizes respectively for the net, 9
Table 3. Corrected mean value of the net, the marketabie and the canning yield of fine carrots in kg per 100 m 2 as a function of the soil class. In addition to the mean values the coefficient of variation and the results of the Duncan test are given. Variabie Variabie Number of plots analyzed
Light sandy loam Loamy sand dry dry moist moist dry
Sand
Total moist
22
27
32
32
45
37
44
27
1073.5 33 143* a**
860.0 29 115 b
840.0 29 112 b
838.0 28 112 b
766.9 29 102 b
774.0 25 103 b
750.9 37 100 b
801.3 35 107 b
820.7 32 109
Marketabie yield X Cv % D
749.3 29 135 a
597.7 33 107 b
601.8 33 108 b
605.2 32 109 b
568.0 27 102 b
602.2 26 108 b
556.6 39 100 b
621.0 35 112 b
602.8 33 108
Canning yield X Cv % D
565.3 29 133 a
438.4 33 103 b
444.6 32 105 b
450.9 33 106 b
425.1 28 100 b
443.7 28 104 b
424.7 39 100 b
451.2 38 106 b
448.7 33 106
Net yield X Cv % D
o
Soil class Sandy silt loam dry moist
266
* Relative yield in percent of the yield on the dry sandy soils. ** Results of the Duncan multiple range test. Averages followed by a different letter are significantly different at the 5% probability level.
I
the marketabIe and the canning yield per soil class, the average yield parameter obtained over the 4 years that the study lasted, the variation coefficient, the relative value of the average yield parameter expressed in percent of the average yield obtained on the dry sandy soils, and the results of the Duncan multiple range test. Finally, table 4 lists the linear correlation coefficients calculated between the different yield parameters, the soil texture and the drainage class. As can be se en from table 3 the dry sandy silt loam soils produce on average significant higher carrot yields than the other studied soil classes. On average those soils produced 33% more carrots. Analyzing more precisely the data in table 1 reveals that the apparent conclusion from table 3 is induced by the extremely high yield obtained on the dry sandy silt loam. soils in 1974. During the other years the product ion of fine carrots on those soils reached only between 45 and 55% of the ' excessive carrot yield recorded in 1974. With regard to the lighter soils the light sandy loams are more although not significantly productive than the sandy profiles. On the latter soils the highest production is obtained on the profiles with a more favourable moisture delivery capacity, i.e. on those soils surveyed as having a poorer drainage. The observed trends in net, marketabIe and canning yield are very similar due to the high correlation (tabie 4) that exists between those yield indicators. Table 4 shows also the relation that exists
Table 4. Correlation coefficients between some product ion parameters, soil texture and drainage class. Production parameters Net yield Marketabie yield Canning yield Number of carrots per m 2 Root length Diameter of the root crown
' Variabie net yield 1 0.943 0.901
Marketable yield
Canning yield
1 0.960
Soil texture
Drainage class
1
0.276 0.186 0.173
ns ns ns
0.293 0.138
0.323 +
0.328 ns
ns 0.163
ns 0.199
0.364
0.296
0.267
ns
ns
Total number of field plots is 266; rO.Ol 0.100 < r < 0.120, ns for r < 0.100
0.159; rO.05
0.121;
+
11
for
bet ween the number of carrots per unit area, the diameter of the root crown and the yield parameters. The differences found between the correlation coefficients relating root density and thickness of the root and net, marketabie and canning yield respectively are minor, indicating that both affect final yield in the same way. Both parameters, root density and root diameter, are not related to texture nor to the drainage class. The root length on the contrary is correlated with texture and drainage class. Carrot roots are on average longer on the finer textured and the more humid soils. Root length is also correlated with net yield, although less pronounced than root density and root diameter. The marketabie yield is only slightly related with root length, while canning yield and root length are not correlated at all. The net, marketabie and canned yield is significantly affected by soil texture. The effect of the drainage class on those 3 yield parameters is masked by the soil texture. 4. CONCLUSIONS The field survey has revealed that yields may vary up to 24% from one year to another as a consequence of climatic differences. The impact of soil class on yield was less clear and was not constant throughout the observation periode Accumulated over the 4 years, the dry sandy silt loam soils produced most carrots, up to 33% and more than the accumulated yield harvested on the dry sandy soils. However on a year basis the carrot yield produced on those soils can vary considerably. The problem on those soils is that with bad weather conditions the mechanical harvest operation is hindered, and that quite a volume of soil sticks to the harvested roots, which induces considerable breakage of the roots and harvest losses. The low yielding capacity of the dry sandy soils is most probably due to the insufficient water delivery capacity of those soils. On average, the humid sandy soils provided every year a stabie and acceptable fine carrot production. The medium textured soils and the moist sandy soils can be regarded as the most suitable for fine carrot growing on all the soil classes examined. Soil tillage and harvest operations are far more easier on these soils, the production is less subject to changes and each year an acceptable yield level is within the capability of good farmers. Those conclusions have been subscribed . by Buishand (1977) who found that in the Netherlands the most suited soils for fine carrot product ion are the sandy soils with ample water supply throughout the growing season, providing that excess rainfall is drained off. Also Vulsteke and Biston (1975) noticed a slight increase in quantity and quality of fine carrots on humid sandy soils in Belgium. Yield increase was not significant and some years marked a high variability in yield from one farm to another. 12
It is believed that the soil structure of the top layer of the seed-
bed at seeding, of the plough layer and of the underlying profile during the growing season are key factors in controlling yield level and yield quality. Subsoiling forexample improves root development and root ramification, allowing the cr op to explore a larger water and nutrient reservoir (Rutten, 1978). ACKNOWLEDGEMENTS The research was entirely financed by the Institute for Encouraging Scientific Research in Industry and Agriculture (IWONL), Brussels, Belgium. The authors wish to extend their sincere appreciation to professor Jan feyen of the Laboratory of Soil and Water Engineering, faculty of Agricultural Sciences, K. U. Leuven, Belgium for his help in reviewing the manuscript.
REFERENCES Buishand, T. (1977) Teelt van bos- en waspeen. Proefstation voor Akkerbouw en de Groenteteelt in de volle grond te Alkmaar, the Netherlands. Deckers, J .A., Scheys, I. & fe yen, J. (1980) Studie van de bodemgeschiktheid voor kers. Agricultura, 28 (3) : 357-400. L.E.I. (1983) Landbouwstatistieken 1983. Landbouw Economisch Instituut (LEI), Ministerie van Landbouw, Brussel, België. Rutten, J. (I978) Fytotechnische aspecten van de teelt van wortelen, Daucus carota L. Ingenieursverhandeling, Fac. Landbouwwet. K. U. Leuven, België, 160 p. Tavernier, R. & Maréchal, R. (1958) Carte des associations de la Belgique. Pédologie, 8 : 134-182. Verheye, W. & Ameryckx, J. (1984) Mineral fractions and classification of soil texture. Pedologie 34 (2) : 215-225. Vulsteke, G. & Biston, R. (1975) Enquête over de invloed van het bodem type op de kwaliteit van wortelen. Bulletin Bimestriel, INACOL, 26 (7-8) : 224-230. 13
Analyse van de bodemgeschiktheid van enkele Belgische gronden voor de produktie van fijne wortelen Samenvatting Over een periode van 4 jaar is de geschiktheid van 8 Belgische bodemklassen onderzocht voor de teelt van fijne worteltjes in hoofdzaak bestemd voor de conservenindustrie. Daartoe werden in totaal 266 proefvlakken, bewerkt op de traditionele manier, geanalyseerd. De bodem typen waarop deze proefplekken lagen behoorden tot een vochtige of een droge drainageklasse, met een textuur variërend van zand tot zandleem. De bodemgeschiktheid werd bepaald aan de hand van volgende opbrengstparameters : lengte en diameter van de wortel, worteldichtheid, netto-, verkoopbare- en inmaakopbrengst. Na correctie van de opbrengstparameters voor het klimaatseffekt werden deze gecorreleerd met de textuur en de drainageklasse. De analyse van de veldenquêtes toonden aan dat de opbrengst sterk kan wisselen van jaar tot jaar, dat deze variatie sterk bodemgebonden is en dat de impact van de bodem op de opbrengst sterk wordt overschaduwd door het klimaat. Globaal genomen zijn de goed ontwaterde zandleemgronden het meest productief. Zij produceerden over de 4 jaar 30% meer dan de droge zandgronden. De productie op deze gronden is echter variabel. Tenslotte werden de lichte zandleemgronden en de zandgronden met optimale vochtvoorziening als de meest geschikte productiegronden voor de teelt van fijne wortelen weerhouden omwille van hun stabiele productiecapaciteit.
Analyse de l'aptitude de quelques sols belges pour la product ion de petites carottes Résumé Une étude de 4 années a porté sur I 'aptitude de quelques sols belges pour la culture de petites carottes. Elle était basée sur I'analyse de 266 parcelles, cultivées d'une maniêre traditionelle, réparties sur des sols secs et humides, et avec une texture sableuse à limonosableuse. Les paramêtres de product ion évalués se refêrent à la longueur, Ie diamêtre et la densité des carottes, ainsi qU'à la production nette, utile et prête à la conservation. Aprês correction pour l'effet climatologique Ie rendement a été correlé avec la texture et la classe de drainage. A partir des données expérimentales les conclusions suivantes peuvent être tirées : Ie rendement peut varier considérablement d'une année à I 'autre, les variations de product ion ne sont pas du même ordre pour toutes les classes de sol examinées, et finalement
14
l'impact du sol sur la production est camouflé par les variations climatologiques. En général, les sols sur limon sableux bien drainés donnent les meilleurs rendements, c.à.d. plus de 30% supérieurs à la product ion obtenue sur sols sableux secs; toutefois Ie rendement sur ces sols peut être fort variabie. Les sols sableux secs se caractérisent par une déficience en eau pendant la saison végétative. Les sols les plus aptes pour la culture des petites carottes sont les sols sur limon sableux léger et les sols sablonneux humides. Ils produisent chaque année un rendement acceptable.
15
PEDOLOGIE, XXXVI-I, p. 17-31, 6 tables, Ghent 1986
ASSESSMENT OF THE OPTIMUM SOIL NUTRIENT COMPOSITION FOR FINE CARROTS GROWING UNDER BELGIAN F ARMING CONDITIONS ]. VANDAMME R. BISTON
Abstract In addition to the physical crop and yield parameters monitored on a large number of field plots grown with fine carrots for evaluating the suitability of 8 Belgian soils for this crop, also the chemical composition of the top soil has been followed up during 4 consecutive growing seasons. The data monitored were subjected to factor analysis in order to define the optimum soil nutrient conditions for assuring good quality and quantity carrot yield. A nutrient advice has been established for sandy silt loam, light sandy loam and sandy soils. Special emphasis 'has been paid to the effect of slurry on yield and quality. The quality parameters examined referred to root deformation, colour and degree of black spot disease. The nutrient reserve of all the soils examined was sufficient to match cr op demand. It has been shown on the contrary that on some soils the nutrient content should be decreased for obtaining a better quality, a redder colour and less appearance of black spot disease. Key-words Carrots, soil nutrients, crop quality.
]. Vandamme - Senior research er at the Centre for Horticultural Soils, Faculty of Agricultural Sciences, Katholieke Universiteit Leuven, Kardinaal Mercierlaan 92 ; 3030 Leuven (Heverlee), Belgium. R. Biston - Former research assistant at the National Institute for the Improvement and the Conservation of Vegetables (INACOL), Wezembeek-Oppem, Belgium. 17
1. INTRODUCTION Soil suitability studies normally consist in evaluating the physical, agropedological suitability to produce crops given the climatic variability, and the chemical fertility from which the optimal soil nutrient composition in function of the erop can be derived. The ability of 8 Belgian soil classes for producing fine carrots, different in texture and drainage behaviour, has been examined in a previous paper (Vandamme and Biston, 1986). This study revealed that although the weIl drained sandy silt loam soils produce the highest accumulated carrot yield over a period of 4 years, the yield fluctuation from year to year limits the overall suitability of these soils. The more stabIe producing fine carrot soils were the lighter textured soils with adequate water supply. In the following paper emphasis is paid to the relation between yield level and yield quality, and the chemical composition of the top layer and the erop. For each of the soil classes examined the optimal nutrient composition for fine carrot growing is derived. 2. MATERIALS AND METHOOS Fine carrots of the cultivar 'Amsterdamse bak' have been grown for 4 years on a large number of field plots. In total 266 plots, distributed at random over 8 Belgian soil types were examined. At harvest time a sample of 30 plants, subdivided in leaves and roots, and a soil sample of the plough layer were collected for chemical analysis. The physical erop and yield Rarameters were monitored on a much larger sample area of 2 m 2 in size. The nutrient composition of erop and soil were determined at the laboratory of the Cent re for Horticultural Soils, Heverlee, Belgium. The extractant used for the chemical analysis of the soil was ammonium lactate and acetic acid, with a pH-level of 3.7. The cation composition was determined by atomie adsorption photometry, while the phosporous content was obtained by the calorimetrie method. The pH was measured by means of a calom'e l electrode. The erop analyses were performed by wet ashing (HCI0 4 and HN0 3). The nitrogen content was determined applying the Kjeldahl method, while phosphorous and the other cations were determined as specified above. The nitrate content of the canned roots, colour (degree of redness) and degree of black spot disease was performed by the laboratory facilities of INACOL. The physical erop and yield parameters monitored were net, marketable and canning yield, plant density (being the number of carrots per unit area), root length and diameter of the crown of the root. As explained in a previous paper (Vandamme and Biston, 1986) the yield data were corrected to balance the effect of climatic varia18
Table 1. Mean value (X) and variation coefficient (C) of the nutrient composltlOn of the plough layer, the carrot leaves and the carrot roots at harvest time, split up per soil texture. Nutrient composition
Number of plots analysed
X
Cv
X
Cv
X
Mg Cv X
Cv
49 64 153
Leaves in meq/ 100 g dry matter Sandy silt loam Light sandy loam Sand/loamy sand
49 64 153
15.0 65 91.8 41 16.7 29 15.1 50 102.0 30 18.2 41 14.9 62 106.1 43 20.6 30
Roots in meq/l00 g dry matter Sandy silt loam Light sandy loam Sand/loamy sand
49 64 153
26.0 49 25.9 52 24.9 42
6.17 10 5.97 9 5.69 9
26.2 42 24.3 36 20.8 39
-
21.4 41 22.8 34 21.4 34
73.0 36 77.8 37 81.1 30
7.5 69 6.7 40 6.9 63
8.5 28 8.8 28 9.7 23
X
N0 3(*)
N
Ca
Soil in mg/100 g Sandy silt loam Light sandy loam Sand/loamy sand
(*) Nitrates in ppm in the canned carrot roots. <0
K
P
pH-H 2O
Cv X
Cv X
Cv
143.6 48 95.1 59 75.5 42
110.9 35 126.9 22 86.5 35 135.5 21 82.6 31 143.9 16
16.5 35 17.5 31 14.5 27
81. 2 32 89.3 33 94.8 27
203 252 286
66 57 50
Table 2. Correlation coefficient between some production parameters and the nutrient status in the root at harvest and the nitrate content in the canned roots. Product ion parameters
Sandy silt loam canning yield plant density root diameter root length Light sand loam canning yield plant density root diameter root length Sand/loam sand canning yield plant density root diameter root length Sandy silt loam
Nutrients in the carrot roots at harvest Nitrate in mg per 100 g dry matter in the canned roots in P Mg K ppm N Ca ns - 0.295 ns +
+
ns ns ns
ns - 0.279 0.267 ns
ns ns 0.312 0.292
ns - 0.256
0.265
+
+
ns
ns
0.339 ns +
0.288
ns ns ns 0.440
ns 0.440 ns
ns
ns 0.304
ns +
0.357 ns
0.392 ns - 0.391 ns
+
0.352
ns ns ns ns
- 0.287 - 0.274 ns ns
ns ns
ns ns ns
ns
+
+
ns
n = 49; rO.Ol = 0.361; rO.05 = 0.279;
+
or - for
r>0.05 = 0.197 to 0.278; ns for r < 0.197 Light sandy loam : n = 64; rO.01 = 0.318; rO.05 = 0.244;
+
or - for
r>0.05 = 0.160 to 0.243; ns for r < 0.160 Sand/loamy sand: n = 153; rO.Ol = 0.208; rO.05 = 0.159; + or - for r>0.05 = 0.100 to 0.158; ns for r < 0.100 tions from one season to another. The quality of the fine carrot crop harvested was measured upon the root shape, colour and appearance of black spots. From a first analysis it became clear that the variation in yield, length and diameter of the root, fertility level of the soil and nutritional composition of leaves and roots on the sandy and loamy sand soils were minor to negligible. In addition, the data revealed 20
th at the drainage class did not affect the chemical fertility of the top layer nor the chemical composition of the crop. As a consequence, the data were ·presented and treated statistically as belonging to 3 distinctive textural classes : the sandy silt loam, the light sandy loam, and the sand/loamy sand soil class (tabie 1). The statistical analyses performed on the data set monitored were respectively the correlation matrix bet ween canning yield, plant density, root diameter and length, and the chemical composition of the plough layer, the fresh leaves and roots, and the nitrate content in the canned roots (tabie 2), and the factor analysis using the unrotated factor loadings for the principal components (Deckers et al., 1980). For each of the 3 textural classes up to 6 factors have been calculated, trying to explain maximum of the variance. Factor loadings of less than 0.300 were not considered as significant. The results of the factor analysis were transferred into quantitative data by multiplying the factor loading by the standard deviation. Sy doing so it was possible to define for each textural soil class specified the effect that would result by modifying the average fertility situation of the plough layer. In order to derive the optimum fertility level of the plough layer for maximizing product ion and yield quality the measured field Table 3. Estimated optimum nutrient composItIOn of the plough layer and nitrogen content in the leaves and the roots for maximum yield level and quality as a function of the soil texture class. Variabie Sandy silt loam
Soil texture class Light sandy loam Sand/loamy sand
Soil, in mg/l00g pH-H 2O P K Mg Ca
6- 6.5 25- 30 20- 25 6- 8 120-150
Total N in the leaves, in meq/ lOOg dry matter
100-120
100
-120
120
-140
Total N in the roots, in meq/ lOOg dry matter
75- 90
75
- 90
80
- 90
Nitrates in the canned roots, in ppm
120-180
120
-200
200
-250
5.8- 6 25 - 30 20 - 25 6 - 8 90 -120
5.7- 5.9 20 - 25 18 - 20 6' - 8 60 - 80
21
Table 4. Fraction of deformed roots (splitted, curled and green topped roots) in relation to canning yield, soil texture and soil fertility. Variabie Splitted roots
Type of deformation Curled roots Green topped roots
Percentage of deformed roots %
Cv Correlation coefficient canning yield soil texture pH-H 2O P (soil) K (soil) Mg (soil) Ca (soil) N (leaves) N (roots)
5.0 180
9.7 98
14.9 154
-
- 0.165 0.190 ns 0.126 ns ns ns ns
- 0.161 ns ns ns 0.165 ns - 0.186 0.275 0.160
n = 266; rO.01 = 0.159; rO.05 0.120; ns for r < 0.090
+
ns ns ns ns ns ns ns 0.121;
+
+
or - for r>0.05
0.090 to
data were split into 4 classes as a function of the canning yield. Following yield classes were considered : class I with a canning yield above 600 kg per 100 m 2, class II with a yield ranging from 450 to 599 kg per 100 m 2, class III combining those plots with a yield bet ween 300 and 449 kg per 100 m 2, and class IV grouping the plots with a yield smaller than 300 kg per 100 m 2. For each of the classes the average soil fertility level, nitrogen content in the leaves and the roots, and the nitrate content in the canned roots were calculated, on the basis of which the optimum nutrient levels per soil texture class were projected (tabie 3). Tables 4, 5 and 6 summarize the results of the statistical analyses performed between root deformation, colour and black spot appearance at one side, the soil fertility, the nitrogen content in leaves and the chemical composition of roots at the other side. The tables represent the linear correlation coefficients and the average content of the different chemical components in soil, leaves and roots measured.
22
Table 5. Colour gradation of fine carrots in relation to yield and soil fertility. Variabie
Correlayellowish tion coefficient red
Colour gradation class deep red
pure or- mixted ange red orange red
Number of plots
10
53
35
14
Canning yield in kg per 100m2
380.5
438.1
469.8
415.0
Soil fertility level in mg per 100 g pH-H 2 O p K
Mg Ca N content in the leaves in meq per 100 g dry matter Nutritional level in the roots in meq per 100 g dry matter N P K
Mg Ca
rO.01
6.44 21.0 21.0 3.7 104.4
5.72 21..1 17.2 4.3 72.7
5.85 19.3 18.1 4.4 78.3
5.99 25.8 28.3 5.9 77.0
112 - 0.091
0.098 - 0.036 - 0.018 - 0.236 0.087
119.7
142.0
139.5
145.7
- 0.116
62.0 18.0 59.0 7.5 15.5
81.5 18.0 74.5 8.4 12.4
85.4 24.8 78.3 8.1 12.9
98.3 26.4 76.6 8.6 14.2
- 0.321 - 0.263 - 0.151 0.000 - 0.002
0.243; rO.05
0.186
23
Table 6. Comparison of carrot yield, soil fertility, nitrogen content in the leaves and nutrient composition of the roots between affected and non-affected plots with black spot disease, as a function of the soil texture class. VariabIe
Sandy silt loam
Light sandy loam
Sand and loamy sand
affected non-affected affected non-affected affected non-affected Number of plots
5
35
10
38
17
80
Canning yield in kg per 100 m 2
465.7
521.7
387.2
453.2
418.8
428.3
6.24 3l.8 25.4 7.6 143.6
6.10 23.2 18.7 5.2 133.4
134.4
12l.6
146.0
128.8
137.1
144.9
102.8 35.8 90.0 9.1 20.6
73.2 23.7 75.7 8.1 16.2
104.5 29.4 98.4 9.4 16.9
86.1 23.4 78.9 8.3 17.7
94.2 2l.2 84.5 9.5 15.6
95.8 25.0 84.7 9.4 14.6
Soil fertility in mg per 100 g pH-H 2 O P K
Mg Ca N content in the leaves, in meq per 100 g dry matter Nutritional level in the roots, in meq per 100 g dry matter N p t-,j
.+::-.
K
Mg Ca --------
5.99 28.5 21.0 5.6 89.2
5.94 22.9 20.6 4.6 87.9
5.97 19.4 19.4 5.3 83.5
5.73 20.1 19.1 4.9 64.6
I
3. RESUL TS AND DISCUSSION The chemical composition of the soils is, within the textural classes in which the soils have been classified, rather uniform. The soils do differ spatially considerably in Ca and Mg content. Only the variation of the Ca content in the plough layer is reflected in the Ca content of the leaves, surprisingly not in the Ca content of the roots. The lower Ca content in the carrot canopy on the light sandy loam, the sand/loamy sand soils is compensated by a higher K, Mg and N content, which is also reflected by a higher nitrate content in the canned roots. No relation was found between the fertility level in the plough layer and the chemical composition of the leaves and the canning yield, plant density, root length and diameter. A more distinctive relation was found bet ween the yield parameters and the chemical composition of the roots at harvest .time and the nitrate content of the canned roots as can be seen in table 2. On all the plots examined there seems to be a significant negative relation between the plant density and the nitrogen content of the roots. The phosphorous content of the roots harvested on the light sandy loam soils seem to be positively correlated with the root diameter and the root length. On the sandy silt loam soils the measured yield parameters seemed to be fairly weIl correlated with the potassium content of the roots. The effect of the Mg and Ca contents on the yield parameters examined seemed to be less evident. Particularly on the sandy silt loam and the light sandy loam soils canning yield, plant density and root diameter seem to be negatively correlated with the nitrate content in the canned roots. From the above it can be concluded that the N uptake by fine carrots increases by decreasing plant density. So far it is not clear why under these circumstances the plant density does not affect the uptake of the other cations. The most acceptable explanation might be found in the relatively high slurry dumping on some of the field plots examined. It is known that this practice results in a deterioration of the soil structure and a reduced emergency of the carrots. At the same time the high slurry application leads to a higher N availability in the soil profile throughout the growing season. Also Van Tilburg (1979a) found in his field test a negative relation bet ween carrot density and yield, and nitrogen sypply; the latter could either originate from slurry or mineral fertilizers. However, Wiebosch (1961) and Linser and Zeid (1973) have proven that a minimum N supply is needed to sustain the accumulation of dry matter, carotene and proteins. Overdoses of nitrogen result, according to the former authors, in a reduction in yield quality and in a too high nitrate content in the canned roots. The positive effect of the potassium content in the roots at harvest time found on some of the
25
yield parameters confirms the hypothesis formulated by von Habben ( 1973). With respect to the canning yield results shown in table 6 the sandy silt loam soils are most productive and produce up to 15 and 20% more than _the light sandy loams and the sand/loamy sand soils respectively. This difference in yield is mainly accounted due to a higher plant density as weIl as to a larger size of the roots. According to the results of the factor analysis yield increase on the sandy silt loam soils can be induced by increasing the nutrient status, especially of K and P in the plough layer and by reducing simultaneously the availability of Mg and N. Yield increase on the light sandy loam soils can be accomplished by lowering the level of all cations determined except for Ca, and by trying to reduce the nitrogen content in the crop. The same conclusions with regard to the fertility status, although more pronounced, could be drawn from the results of the factor analysis applied to the data assembied from the sand/loamy sand field plots. The result of the factor analysis translated into a farmer's useful advice has been summarized in table 3, for each of the textural soil classes examined. This table lists per textural class the optimum nutrient composition of the soil which should be approached to guarantee maximum yield level and quality. Confrontation of those levels to the actual nutrient situation indicates that the N level should be reduced. The pH-H 2 0 and the Ca content are also texturedependent and must be higher on finer soils. Similarly, the Pand K level should be increased on those soils. The same conclusions were derived for salsify (Appelmans and Vandamme, 1982), but not for tomatoes (Vandam me and Lamberts, 1978), beans and lettuce. Those crops require a higher fertilizer application on the more sandy soils. A high quality crop consists of equally shaped, non ramified roots. Under some conditions the roots might be splitted, curled or do have green tops. Curled and ramified roots hinder considerably mechanical harvest and peeling, resulting to an increase of yield loss. Table 4 lists the percentage of splitted, curled and green topped roots respectively counted on the field plots examined. The average percentage of occurrence is not dramatically increased but the spatial variability is quite high. The canning yield is most directly influenced by the number of green topped and splitted roots. The relation bet ween root deformation and soil texture is not very clear, except that the frequency of occurrence of green topped roots seemed to be affected by the texture. There is almost no significant relation between soil fertility and the presence of curled roots and green topped roots. Only the phosphorous content of the top layer seems to induce the occurrence of green topped roots. Root ramification is negatively related with the
26
soil Ca content and positively with the N content in leaves and roots. Both indicate that soil structure might be the direct cause of root ramification (Appelmans and De Leenheer, 1968; Dowker and Jackson, 1977). As specified earlier, structure deterioration is induced by the dumping practice of slurry. Root colour gradation, varying from deep red over orange red to yellowish red is, according to the data given in table 5, strongly affected by the soil fertility level and as a consequence by the chemical composition of the roots. The richer the pH-H 20 and the Ca content of the soil the more red the roots are coloured. Phosphorous and potassium are weak, although negatively related with the root colour. The relation between root colour and Mg content of the plough layer is significant, the carrots being more yellowish by a higher Mg content. Similarly, nitrogen is reducing the red colour of the roots. Also Linser and Zeid (1973) found a positive correlation between a high nitrogen 'content in leaves and roots, and the number of yellow carrots. The black spots on the roots have been considered as a symptom of cavity spot or a similar disease. The appearance of black spots was noted on 32 of the 185 plots examined. The disease was more widespread in cool and humid years than in dry and warm growing seasons. AIso, although not significantly, less field plots on sandy silt loam soils were affected. The drainage intensity of the soils examined did not affect the occurrence of black spots. This might probably be explained by the fact that the water table in the summer on the moist soils never exceeds the water table depth of 70 cm. Above this depth, it means th at the percentage of black spot infect ion strongly increases for water table depths shallower than 70 cm below surface (Lentjes and Brouwer, 1980). These authors, just like Hulshof and Maenhout (1981), claim that infection of black spot disease can be initiated by abrupt changes in soil moisture content. Comparison between the affected and non-affected plots did not indicate that the average yield on the affected plots was lower (2% on the sand/loamy sand soils, 17% on the light sandy loam soils and 12% on the sandy sil t loam soils), the nutrient status of the soil was higher, and with exception of the sand/loamy sand soils the leaves on the affected plots contained more nitrogen. The infected roots on the sandy silt loam and the light sandy loam soils contained more cations than the non-affected söUs, jn particular significantly more N and P. On the sand/loamy sand soils the affected and non-affected roots had similar chemical composition. The relation found between the degree of infection and the nutrient content of soil and root disagrees with the experience of Maynard et al. (I961, 1963), who quoted an increase in black spot 27
infect ion by lack of Ca. Perry and Harrison (1979), who checked the statement made by previous authors, found no relationship at all. On the contrary, they proved that the degree of infection was related with soil compaction and degradation of soil structure. In a similar way field experiments run by the canning-factory NOLIKO (Belgium) do indicate that black spot infection is strongly reduced by liming and deep ploughing (personal communication). With regard to the influence of N on black spot appearance proved in this study, Van Tilburg (197gb) was not able to trace any link bet ween both, neither for the mineral nor for the organic nitrogen in the soil.
4. CONCL USIONS The farming practice for fine carrots in Belgium is such th at all of the 8 studied soil classes do contain a sufficient storage of nutrients. There is no direct need to increase the fertilizer application. On the contrary the nutrient content of the plough layer should be lowered gradually, on the coarse textured soils more than on the fine, in order to improve fruit quality and to decrease the degree of infection of the roots by black spot disease. An increase of the Ca content might improve the root colour, but on the other hand according to the results of our experiments it might induce the appearance of black spot disease. The farmers growing fine carrots should pay particular attention to the nitrogen application. An overdosis of nitrogen, in the mineral or the organic form, reduces plant density and indirectly affects yield and enhances the ramification of the root. Furthermore, overapplication of nitrogen induces a paler root colour and promotes most likely the appearance of black spot disease. Therefore it is recommended to limit the nitrogen supply between 60 and 80 units per ha. ACKNOWLEDGEMENTS This research project was financed by the Institute for Encouraging Scientific Research in Industry and Agriculture (IWONL), Brussels, Belgium. The authors are largely indebted to the managers of the different canning factories for their assistance in getting the harvested carrot roots properly processed. Specially the coordinating task of ir. Seeger, direct or of INACOL at Wezembeek-Oppem, is very much appreciated. Finally, the authors wish to extend their sincere gratitu de to professor Jan Feyen of the Laboratory of Soil and Water Engineering, Faculty of Agricultural Sciences, K.U.Leuven, Belgium for his editing assistance.
28
REFERENCES Appelmans, F. & De Leenheer, L. (1968) Le pourcentage de betteraves fourchues comme indice de la compaction du sol et sa corrélation avec Ie rendement. Pédologie, 18 (3) : 333-350. Appelmans, F. & Vandamme, J. (1982) Optimale teeltcondities voor schorseneren op de verschillende bodemtexturen. Public. 13 van het Studiecentrum voor Tuinbouwgronden, Heverlee, Belgium, 79 p. Deckers, J.A., Scheys, I. & Feyen, J. (1980) Studie van de bodemgeschiktheid voor kers. Agricultura, 28 (3) : 357-400. Dowker, B.D. & Jackson, J.C. (1977) The effects of environments within a site on the performance of carrot cultivars. Journ. Hort. Sc., 52 : 299-307. Hulshof, J .A. & Maenhout, O.A. (1981) Pok i n waspeen. Mededeling van het Proefstation voor de Akkerbouw in volle grond, Lelystad-Alkmaar, Nederland, 36 p. Lentjes, P. & Brouwer, H. (1980) Pok bij peen. Ingenieursverhandeling, Vakgroep Tuinbouwplantenteelt, Landbouwhogeschool, Wageningen, 53 p. Linser, H. & Zeid, F.A. (1973) Der Einfluss der Stickstoffversorgung auf den Proteingehalt von Daucus carota. Zeitschrift für Pflanzenernährung und Bodenkunde, 136 (2) : 156-164. Maynard, D.N., Gersten, B., Vlach E.F. & Vernell, H.F. (1961) The effects of nutrient concentration and calcium levels on the occurrence of carrot cavity spot. Proc. Am. Soc. Hort. Sc., 78 : 339-342. Maynard, D.N., Gersten, B., Young, R.E. & Vernell, H.F. (1963) The influence of plant maturity and calcium level on the occurrence of carrot cavi ty spot. Proc. Am. Soc. Hort. Sc., 83 : 506-510. Perry, D.A. U Harrison, J. G. (1979) Cavity spot of carrots. J. Symptomatology and calcium involvement. Ann. Apl. Biol., 93 : 101-108.
29
Vandamme, J. & Lamberts, D. (1978) Bodemgeschiktheid en optimum teelt- en voedingscondities voor tomaten. Agricultura, 26 (3) : 234-352. Vandamme, J. & Biston, R. (1986) Analysis of the suitability of some Belgian soils for growing fine carrots. Pedologie, 36 (1) : 5~ 15. Van Tilburg, A. (1979a) De invloed van varkensdrijfmest op de opbrengst en de kwaliteit van waspeen. Mededeling van het Proefstation voor de Akkerbouw en de Groenteteelt in volle grond, Lelystad-Alkmaar, Nederland, 20 p. Van Tilburg, A. (1979b) De pokziekte in de waspeen. Mededeling van het Proefstation voor de Akkerbouw en de Groenteteelt in volle grond, Lelystad-Alkmaar, Nederland, 7 p. von Habben, J. (1973) Einfluss der Stickstoff und Kaliumdüngung auf Ertrag und Qualität der Möhre (Daucus carota L.). Landwirtsch. Forsch., 26 (2) : 156-172. Wiebosch, W.A. (1961) Literatuuroverzicht van peen of wortel. Mededeling van het Proefstation voor de Akkerbouw en de Groenteteelt in volle grond, Lelystad-Alkmaar, Nederland, 6 p.
Bepaling van de optimale voedingstoestand voor de kweek van fijne wortelen onder Belgische praktijkomstandigheden Samenvatting Aansluitend op de bepaling van fysische gewasgrootheden en opbrengstparameters op een groot aantal proefvlakken, teneinde de geschiktheid van 8 Belgische bodemklassen voor de produktie van inmaakwortelen te bepalen, werd eveneens de vruchtbaarheidstoestand van de bouwlaag gedurende deze 4 jaren durende studie opgevolgd. De opgemeten voedingsgegevens werden onderworpen aan faktoranalyses om de optimale bodemvoedingstoestand te definiëren, vereist om een kwantitatief en kwalitatief goede oogst te garanderen. Als praktisch resultaat werd voor de drie onderscheiden textuurklassen zandleem, licht zandleem en zand, de na te streven voedingstoestand van de bouwvoor gespecifieerd. In de studie werd eveneens aandacht besteed aan het effect van drijfmest toediening op de opbrengst en 30
de kwaliteit. Als kwaliteitsindicatoren werden volgende grootheden op het gewas bepaald : misvorming van de wortel, kleur en graad van pokaantasting. . De onderzochte bodems bleken over een voldoende reserve aan voedingsstoffen te beschikken. Er kon integendeel aangetoond worden dat sommige gronden chemisch te rijk zijn om goede kwaliteitswortelen voor inmaak te produceren. Gesteld werd dat de stikstofbemesting de dosis van 60 tot 80 eenheden per ha niet zou mogen overschrijden.
Définition des conditions nutritives optimales pour la product ion de petites carottes sous les conditions agricoles belges Résumé En complément à 1'étude de 8 sols belges pour la product ion de carottes, une recherche approfondie a été effectuée sur les conditions nutritives de la couche arabie, pendant les 4 années de ce projet. Les données d'analyse qui caractérisent Ie niveau nutritif des sols ont été soumises à une analyse factorielle. Les résultats ont permis de déduire pour les 3 classes texturales (les sols sur limon sableux, les sols sur limon sableux léger et les sols sablonneux) de déduire les conditions nutritives optimales pour fournir une bonne product ion de petites carottes. Dans Ie cadre de cette recherche I'effet de l' application de lézier sur la quantité et la qualité a été étudié. La qualité des petites carottes a été évaluée à part ir du pourcentage des carottes ramifiées, la couleur et Ie degré de taches noires sur les carottes. Le niveau nutritif moyen des sols étudiés était suffisamment élévé pour garant ir une bonne production. Au contraire, on a pu démontrer que la couche arabie de quelques sols examinés, ayant un état nutritif trop élevé, a un effet négatif sur la product ion, surtout sur la qualité des carottes. A partir de ces résultats on a conclu que la do se d'azote ne peut pas dépasser 60 à 80 unités par ha.
31
IPEDOLOGIE,
XXXVI-I, p. 33 - 43, 2 fig., 4 tab., Ghent, 1986.
EVALUATION OF PROFILE AVAILABLE WATER CAPACITY. 3. A MODEL FOR ESTIMATING PROFILE AVAILABLE WATER CAP ACITIES FOR WHEAT ON SOILS UNDER IRRIGATION, USING SIMPLE PHYSICAL AND CHEMICAL SOIL PROPERTIES L. BOEDT W. VERHEYE
Abstract Multiple regression analysis is used for predicting profile available water capacities (PAWC) for wheat using physical and chemical soil properties as input. The regression model is based on data collected during field experiments whereby PAWC for wheat was determined. It is expected that the model will be useful in land evaluation studies. Keywords PAWC, multiple regression analysis.
1. INTRODUCTION The capacity of soils to retain water and make it available to plants is an extremely important determinant in the evaluation of the suitability of a soil for plant growth and for irrigated agriculture. Direct experimentation, with a range of important crops of the studied area, to assess available water is obviously the best method, but this practice is of ten not possible or is too tedious for researchers involved in land evaluation and in planning. A close L. Boedt - Formerly Dept. Biology, Ecopedology, Universitlaire Instelling Antwerpen, Universiteitsplein, 1, B-2610 Wilrijk. W. Verheye - National fund for Scientific Investigations, Belgium, Labo Algemene Bodemkunde Rijksuniversiteit Gent, Krijgslaan 281, 9000 Gent and Dept. Biology, Ecopedology, Universitaire Instelling Antwerpen, Universiteitsplein, 1, B-2610 Wilrijk, Belgium. 33
approximation of this capacity of soils would in many situations be very helpful. The determination of this soil parameter in the classical concept of soil available water (being the difference between soil moisture content at -10 or -33 kPa and -1500 kPa) is quite laborious and often fails to give reliable results when compared with field observations of available water (Boedt and Verheye, 1985a and 1985b). This approach is however still succesful with many workers in the field of irrigated agriculture. Tables relating soil available water to soil texture (F.A.O., 1979) are used to make quick field assessments of profile available water. Gupta and Larson (1979) developed mode Is for estimating soil water retention over a wide range of soil matric potentials from particle size distribution, organic matter content and bulk density. Their regression mode Is predict soil available water reasonably well within the limited geographical area where the raw data to develop their models were collected, but of ten fail to give reliable results for soils containing only small amounts of silt. This makes the models less suitable for tropical areas. In an alternative approach Cassel et al. (1983) developed models for estimating in situ potential extractable water (PLEXW). Their models are based on numerous field observations and give results that are easily applicable in real life conditions. They originated, unfortunately from observations under rainfed farming conditions and, therefore, their application for irrigated agriculture is limited due to the use of a very harsh lower limit, at which plants virtually die. Ritchie (personal communication) suggested the use of a fraction of PLEXW as allowable depletion for irrigated crops. In a first instance this fraction is arbitrarily chosen and allowable depletion is later determined by "trial and error" methods. It is obvious that th is approach has its limitations for application in irrigation and scheduling. Hensley and De Jager (1982) predicted "profile available water capacities" (PAWC) for untested sites using rooting characteristics, bulk density and soil water contents at -10 and -1500 kPa. This model specified available water for irrigated crops quite accurately but the model has the disadvantage that the observations used to develop the model originated from a limited geographical area and that some soil parameters which are needed, are difficult to determine accurately (soil water content at -10 and -1500 kPa, bulk density). Buccheim and Ploss (1977) used the ratio of the extraction rate in the top zone to the extract ion rate in the total soil profile (rooting depth) to determine the allowable depletion for a given soil. They thereby recognised the importance of rooting depth in estimating available water in a soil profile. Laker (1982) proposed a model for predicting PAWC for irrigated maize on medium-textured to 34
clayey Ciskeian soils, using only rooting depth as independant variable. The model allowed for a quick assessment of available water in the soil profile; while maintaining a high degree of accuracy. The model originated because the data of Hensley and De Jager (1982) revealed that rooting depth was the absolutely dominant factor determining PAWC for these soils. 2. PROCEDURE During this study PAWC was determined at several sites on soils showing a wide variety in pedogenetic characteristics and rooting depths. As it is difficult to estimate the contribution of a water table to the moisture delivery capacity of the soil, it was decided that the selected soils should not have a water table within 7 m of the soil surface. The upper and lower limit of PAWC used during the determination are described by Boedt and Verheye (1985a). At each of the experiment al sites detailed soil profile descriptions were made and samples of the diagnostic horizons were taken for analysis. From the data collected in the field and determined in the laboratory a number of variables, expected to have an important impact on soil available water were selected : effective rooting depth, organic carbon, cation exchange capacity, soil structure, silt plus clay content, in situ measured field capacity. Pseudo-sand and pseudo-silt physically act as sand or silt when soil water retention characteristics are concerned. It is therefore im portant that the textural analysis of the soil should be done without destroying cementing agents such as iron. It is obvious that effective rooting depth has an important impact on profile available water. This impact has two aspects : (1) the total depth of the soil profile is of importance (the deeper the effective rooting depth, the more water is available, (2) the depth at which aspecific soil layer occurs has an influence on the amount of water that can be extracted from it. Pedogenetic horizons with identical characteristics, but occuring at different depths in the soil profile, will, because of the inherent rooting pattern of the crop, contribute differently to the profile available water. The closer the occurrence of the horizon to the surface the higher this contribution wilt beo The amount of organic matter present in the soil as such does not influence much the moisture holding capacity but the impact of humus on soil structure, porosity and water storage is very important (Buckman and Brady, 1969). Soil texture, especially the fine fraction (silt and clay), has an important impact on the soil moisture retention characteristics (figures 1 and 2). In figure 1 field capacity and permanent wilting point (the upper and
35
30----------------------------------------~
... ..... . .. ........ ...... . ... . .. ..... ..... . . ... ........ .
24
....
Cl: W I-
.. .........
: ::... :::::: ... .....::..: .. ......... .. ..... ... ..... .
::. : 3 .. ..
--'
o
Cf)
18
u.
c:x::
2 0 E
~
u E E
12
........
Sond
Sondy
loom
Loom
SJIt
Clay
Loom
Loom
Clay
fig. 1. The general relationship between soil water characteristics and soil texture (from Buckman and Brady, 1969).
. . .
.
,
20
.
,
.-Z &AI
u
a:: 15
~
-
~
-
r
&AI
:i :>
.J
10
0
> I
• )(
-
5
&AI j
CL
0
I
1
1
.L
~
I
la
la
I
lil
.
I
I
. .
I
si aicl cl Icl IC lie c
SOl L TEXTURE
fig. 2. Change in PLEXW value with soil texture (from Ratliff et al, 1983).
36
lower limit of available water in the classical concept) are shown as functions of the soil texture. The same may be observed in figure 2 where thechange in PLEXW-value with soil texture is shown (Ratliff et al., 1983). This relation between soil texture and soil available water is especially prominent in light textured soils. In sandy textural ranges available water increases with increasing clay content. This increase continues until a certain textural composition is reached, above which the available moisture remains more or less constant. This is clearly illustrated by figures 1 and 2 : the available water between field capacity and permanent wilting point and the PLEXW - values found on soils with textures heavier than sandy loam remain almost constant. The cation exchange capacity is a reflection of the combined influence of the amount of organic matter and clay present in a soil and of its clay mineralogy and was therefore incorporated in the model. Soil structure has an important impact on root development and porosity and thereby influences considerably the availability of soil moisture. In strongly structured horizons certain limitations are put on the ramification of roots. Roots grow and develop around the structural units following the path of the least resistance. Strongly developed clay cutans prohibit the penetration of roots into the structural units. As a consequence Ie ss water is extracted from the horizon in which these strong structural units occur. This is illustrated by the extract ion pattern from a soil showing a strongly developed natric B-horizon (tabie 1). Table 1. Quantities of water extracted at first stress (indicated by pre-dawn leaf water potential readings) by wheat on a BEAU-soil having a strongly structured B-horizon (from Boedt and Verheye, Soil depth 0- 100 100- 200 200- 300 300- 400 400- 500 500- 600 600- 700 700- 800 800- 900 900-1000
Water extracted (mm/100 mm of soil) 16.6 13.6 12.1 10.6 5.6** 6.3 6.3 7.8 6.6 3.0
** start of the strongly structured B-horizon.
37
Table 2. Soil data used to develop PAWC models for wheat. D.I. (cm)
A.W.I. (mm/l0cm)
O.C.
C.E.C. (meq/l00g)
S.I.
F.C.
SICL
(%)
(%)
(%)
(%)
10 31 54 14 39 59 12 37 67 142 38 88 150 8 61 148 10 35 70 110 5 18 46 79 106 11 28 45 74 10 45 85 11 28 48 73 91
13.0 12.9 12.9 12.9 14.8 15.5 8.7 8.3 8.2 7.7 8.1 7.7 6.4 9.8 9.1 9.2 23.3 11.6 9.2 7.3 19.1 16.3 12.7 7.7 3.8 22.0 18.0 13.4 11.4 24.3 15.7 8.1 14.8 11.7 7.3 7.3 3.3
0.26 0.16 0.13 0.24 0.17 0.10 0.23 0.18 0.11 0.10 0.23 0.11 0.11 0.21 0.12 0.07 1.03 0.71 0.31 0.26 0.56 0.51 0.43 0.23 0.13 0.94 0.80 0.47 0.23 0.72 0.64 0.40 0.54 0.51 0.33 0.11 0.10
6.4 8.8 11.0 12.6 19.2 16.4 7.0 4.8 7.0 12.0 6.0 5.6 8.0 5.2 7.2 6.0 22.8 12.2 22.4 29.0 16.0 15.6 18.0 15.0 17.0 14.8 17.0 18.0 21.0 20.4 22.0 24.0 12.4 14.4 32.8 24.4 28.4
40 24 38 41 52 69 28 45 41 41 38 52 42 36 35 38 27 30 41 48 29 30 30 34 40 36 28 54 43 27 37 33 34 35 65 66 68
18.3 19 .. 7 21.0 23.2 31.5 36.8 13.0 14.8 17.3 22.2 13.2 16.6 20.5 13.0 13.4 20.5 33.0 33.4 35.0 43.0 26.8 26.8 26.8 26.8 26.8 32.0 31.2 29.6 31.0 31.2 29.4 29.6 24.8 25.0 33.0 37.0 32.5
8.1 14.1 13.8 16.8 29.2 29.9 8.6 10.0 11.2 14.4 7.2 11.2 13.6 8.9 11.0 13.4 80.3 84.3 86.0 85.9 56.7 52.9 57,9 55.8 52.9 74.8 74.0 76.2 81.8 71.8 74.1 73.0 51.8 48.4 79.7 70.8 62.2
A.W.I.
=
D.I. O.C. C.E.C. 5.1. F.C.
=
SICL
=
= = = =
available water index : available water (mm) per 10 cm of a pedogenetic horizon. depth index (cm) organic carbon content (%) cation exchange capacity (meq./ lOOg. soil) structure index (%) soil water content at depth index at field capacity (measured in situ) (%) silt + clay content (%)
In weakly structured horizons there is hardIy any resistance to root ramification. This results in an evenly distributed root system allowing a more effieient water uptake. In an attempt to quantify this qualitative soil characteristie the ratio of (Na + Mg) to total exchangeable cations expressed as a percentage, was calculated. This was done because it was found during the field observation of the soil structure th at weIl structured soil horizons always contained relatively high amounts of Na and Mg. This ratio was called structure index (S.I.) and it gave a fairly good reflection of the degree of structure encountered. Field capacity was selected to be used in the model because it reflects the totality of factors affecting water holding capacities of soils. 3. RESULTS The available water data, expressed as available water index (A.W. I. = amount of available water in a 10 cm thick layer of a specifie pedogenetie horizon), are given in table 2. The data are grouped per pedogenetic horizon. This grouping per pedogenetie horizon was done because of the importance that certain pedogenetie characteristics have on soil available water (Boedt and Verheye, 1985a). Depth index refers to the distance (expressed in cm) between the surface and the top respectively lower boundary of the pedogenetie horizon divided by two; it reflects the depth at whieh the horizon is occuring. These different variables are used together in a simple multiple linear regression for predieting PAWC. Table 3 shows the correlation coefficients found between the selected variables and available water index for the horizons identified in table 2. It is clear from table 3 that organic carbon content and depth index have the most dominant impact on the available water index. The influence of the dep th index was expected from the observed extract ion patterns (Boedt and Verheye, 1985b). The high correlation coefficient found for organic carbon content can be explained by (1) the general decrease in organic carbon with depth, whieh is more or less parallel to the soil water extraction pattern and by (2) the intrinsic water holding capacities of organic matter. The apparent low correlation coefficients found for the structure index, water content at field capacity and (silt + clay) content must be interpreted with caution. Indeed, the relation between those parameters at the one hand and available water at the other hand is overshadowed by the influence that the depth, at which a specific horizon occurs, has on the amount of water that is extracted from it. Clay mineralogy does not seem to have an important impact on soil availbIe water and this resuits in a low correlation coeffieient between C.E.C. and A. W.I. This parameter was therefore no longer used in the model.
39
Table 3. Corrleation coefficients between the different soil variables and available water index (A.W.I.) for wheat.
A.W.I.
D.I.
o.c.
C.E.C.
S.I.
f.C.
SICL
-0.65
0.78
0.08
-0.42
0.25
0.28
A simple multiple regression analysis with the available water index as the dependent variabie and the five remaining parameters as independent variables was th en executed. The obtained equation, which predicts the available water in a 10 cm thick soil layer of a specific pedogenetic horizon, was : y = 4.6 - 0.03xl + 15.84x2 - 0.07x3 + 0. 51x 4 - 0. 15x 5 (1) with y zon.
=
available water per 10 cm of a certain pedogenetic hori-
Xl depth index (cm) x2 organic matter content (%) x3 structure index (%) x4 water content at field capacity (%) x5 silt + clay content (%). r2 0.81 f(5,36) = 26.63 (significant at 0.01 level). The profile available water capacity (PAWC) for the total soil profile is given by i=n PAWC =L: di (4.6 - 0.03xl + 15.84x2 - 0.07x3 + 0.51x4 -0.15x5) (2) i=1 10 with n = number of the ith pedogenetic horizon. d = thickness of. the ith pedogenetic horizon in cm. xl to x 5 as for equation (1). This regression could be tested for wheat on two soils from the KAL- and BEAU series (Boedt and Verheye, 1985a). The KAL- and BEAU soil series define respectively deep sandy aeolian soils of the Kalahari formation and ' moderately deep to shallow weathered soils developed over a mudstone, sandstone and dolerite complex. The raw data of these two soils are listed in table 4. They were not used in the multiple regressi0n analysis that resulted in equation 1. Hensley and De Jager (1982) determined for both soils P AWC for wheat in situ. for the KAL-soil they found a PAWC value for mature wheat of 132 mm, while the model predicted it to be 136.7 mmo for the BEAU - soil they observed in successive years 125 mm and 137 mm as PAWC values for wheat, while the proposed equation predicted 124.5 mmo 40
Table 4. Basic soil properties for a KAL - resp. BEA U soil (derived from Hensley and De Jager, 1982). KAL - soil Horizon Ap B21 B22 B23
D.I. 11 40 89 153
O.C. 0.23 0.18 0.11 0.10
S.1. 30 31 29 51
f.C. 13 14 16 17
SICL 11.1 12.1 12.5 14.0
O.C. 0.49 0.33 0.18 0.10 0.10
S.1. 22 23 33 36 30
f.C. 27.5 26.0 27.0 27.0 27.5
SICL 30.9 32.5 34.2 57.3 58.0
BEAU - soil Horizon Ap B21 B22 B31 Cl
D.1. 13 33 50 70 90
4. DISCUSSION Although the proposed model is an improvement on the of ten inaccurate determination of available water in the classical concept, the model still requires the input of a range of routine laboratory and field data. In many cases these data are not available for fieldworkers in land evaluation. Therefore a simplified model, with two independent variables, that can be determined easily in the field by experienced fieldworkers (depth index and silt + clay content), is additionaly proposed. The multiple regression analysis yielded the following equation : (3) y = 14.2 - 0.08xl + 0.04x2 with xl depth index (cm) x2 = silt + clay content (%) r 2 = 0.47 f(2,36) = 15.38 (significant at 0.01 level). To obtain the profile available water capacity for the total soil profile equation (4) is used : i =n (4) PAWC = ~ di (14.2 - 0.08xl + 0.04x2) 1=1 10 This equation is obviously less accurate than formula (2) but it has the advantage that it can be used without preliminary laborato41
ry determinations : an experienced soil surveyor can accurately estimate soil textuTe from a field observation, effective rooting depth can be deducted from the soil profile characteristics. for the KAL - and BEAU soils that were used to test the accuracy of equation (2) formula (4) yields the following estimates of PAWC : 134.9 and 119 mmo These values compare reasonably weIl with the field determinations of PAWC for wheat for these soils. 5. CONCLUSIONS Significant multiple regression equations could be developed for predicting PAWC for wheat. The regression model gives the possibility to estimate PAWC from routine laboratory data and/or from some simple in situ soil observations. The model, as it stands for the moment, is applicable to soils belonging to a considerable range of textural classes. However, because it has been developed on data collected on soils within a limited geographical area it would be useful to collect additional data to improve the widespread applicability of the model. 6. ACKNOWLEDGEMENTS The obtention of a "NfWO-Krediet voor Navorsers" grant permitting part of the data collection used in this paper is highly acknowledged by the second au thor.
REFERENCES Boedt, L. & Verheye, W. (1985a) Evaluation of profile available water capacity. 1. The conceptual approach. Pedologie, 35( 1) : 55-65. Boedt, L. & Verheye, W. (1985b) Evaluation of profile available water capacity. 2. Application to irrigation of soils with different properties. Pedologie, 35( 1) : 67-89. Buchheim, J.f. & Ploss, L.f. (1977) Computerized irrigation scheduling using neutron probes. Publ. ASAE, St. Joseph, Michigan. Buckman, H. O. & Brady, M. C. (1969) The nature and properties of soils. Collier-Macmillan, Londón.
42
J
Cassel, D.K., Ratliff, Models for estimating physical and chemï.cal Soil Sci. Soc. Am. J.,
L.F. & Ritchie, J. T. (1983) in situ potential extractable water using soil properties. 47 : 764-769.
F.A.O. (1979) Soil survey investigations for irrigation. F.A. O. Soil Bulletin 42, Rome. S.C. Gupta & W.E. Larson (1979) Estimating soil water retention characteristics form particle size distribution, organic matter percent and bulk density. Water Res. Research, 15(6) : 1633-1635. Hensley M. & De Jager, J.M. (1982) The determination of the profile available water capacities of soils. University of Fort Hare, Alice, 282 p. Laker, M.C. (1982) A provisional simple model for estimating PAWC for maize on certain Ciskeian soils. In : Hensley, M. and De Jager, J .M. (eds.) The determination of the profile available water capacities of soils. University of Fort Hare, Alice, 282 p. Ratliff, L.F., Ritchie,J.T. & Cassel, D.K. (1983) . Field measured limits of soil water availability as related to laboratory-measured properties. Soil Sci. Soc. Am. J., 47 : 770-775.
Evaluatie van de beschikbare watercapaciteit in het bodemprofiel. 3. De ontwikkeling van een model voor het schatten van de beschikbare waterkapaciteit in het profiel aan de hand van enkele eenvoudige fysische en chemische bodemkarakteristieken. Samenvatting Door middel van een meervoudige lineaire regressie analyse wordt een model ontwikkeld dat de beschikbare waterkapaciteit (PAWC) voor tarwe schat. Het model is gebaseerd op gegevens verzameld tijdens in situ bepalingen van PAWC voor tarwe. Het model laat een vlugge appreciatie toe van het beschikbaar water in een bodemprofiel en zou van waarde kunnen zijn in de evaluatie van gronden voor irrigatiedoeleinden.
43
Evaluation de la capacitê de la teneur en eau disponible dans Ie profil du sol. 3. Dêveloppement d'un modêle pour estimer la capacitê de la teneur en eau disponible pour du blê irrigê à l'aide des caractêristiques physiques et chimiques du sol Résumé Au moyen d'une analyse de rêgression multiple un modêle est dêvelopê pour prêdire la capacitê de rêtention en eau disponible pour Ie blê. Le modêle, basê à I'origine sur des observations de terrain, permet une apprêciation rapide de la teneur en eau disponible et peut être d'une valeur considêrable pour I 'êvaluation des sols pour des cultures irrigêes.
44
PEDOLOGIE, XXXVI-I, p. 45-57, 4 tab., 2 fig., Ghent, 1986
IDENTIFICATION LEGEND FOR SOIL SERIES DEVELOPED IN THE UPPER PLIOCENE DEPOSITS OF BANGLADESH M. HASSAN R. TA VERNIER
Abstract Soil series identi fied and described by the Soil Resource Development Institute (SRDI), Gvt. of Bangladesh, for soils of the Upper Pliocene Deposits, have been recharacterized accurately in the light of Soil Taxonomy. Spatial relationship between the soil series in a catena has been shown on a schema tic diagram and a field legend containing the genetic differentiae have been proposed for the soil series. This legenè, will serve as a field guide for the identification of soil series developed in this landform and subsequently will be helpful for their classification. Keywords Estuarine, extrapolation, ferrolysis, illuviation, legend.
1. INTRODUCTION
The Upper Pliocene sediments, locally known as Madhupur Clay, occupy the central part of Bangladeh lying bet ween 24°00' and 24°45'N latitudes and 89°45' and 90° I5'E longitudes. Th'is landscape looks like a dissected terrace consisting of a series of uplifted fault blocks bounded to the west by prominent escarpments and is gently tilted towards the east. The northern part consists of broad, nearly level upland with narrow drainage valleys. It consists of an unconsolidated homogeneous clayey sediment. The thickness of the clay layer va ri es from 8 m ne ar Dhaka city and is gradually becoming thinner in the west (Wadi a, 1957; FAO, 1971). M. Hassan - Or. Sc., Head Soil Science Division, Forest Research Institute, Chittagong, Bangladesh. R. Tavernier - Prof. dr. em., State University Ghent, Belgium.
45
Bangladesh has a humid megathermal cimate (A'), summer days are longer and hotter. Day length and temperature jointly contribute to the variations in seasonal temperature efficiency (Thornthwaite, 1948). The flux of water being negative under such conditions (Tavernier and Eswaran, 1977), a net loss of bases takes place. The hot humid climate of Bangladesh is favourable for astrong chemical breakdown of the silicate minerals and subsequent leaching of bases and silica with the percolating water. Though kaolinite should be the stabie mineral under such conditions, the clay minerals present in soils of the Madhupur clays are mixed in composition showing no evolutionary trend. The Soil Resources Development Institute (SRDI) of the Government of Bangladesh had identified and described twenty one soil series in this landscape during the reconnaissance soil survey (Anon., 1963, 1965). This characterization is not complete for the purpose of a genetic soil classification, because one soil series according to these characters may occupy different pI aces in the higher categories. In order to remove this anomaly, the objectives of this paper are to correlate the soil series developed on the Madhupur clays, to characterize them accurately for the purpose of a genetic soil classi fication and finally to prepare an identification legend for the soi I surveyors. 2. PROFILE DEVELOPMENT Appreciable loss of clay from the topsoil is observed in soils of this landform. This is mainly due to run-off and partly due to clay illuviation. The granulometric data do not show a significant clay bulge in the lower layers commensurable to the amount of clay lost from the topsoil (fig. 1). But the conspicuous clay cutans in the subsoil, observed in the field and in thin sections in some deeply weathered soils, indicate clay illuviation (Brammer, 1971; Hassan, 1982). Illuviation cutans present in the B horizon are adequate (1 % by volume) for designating them as argillic. Soils occurring on shallowly weathered terraces have a cam bic horizon. The drainage sequence shows a good correlation with the grade, class and type of the subsoil structure. Poorly to imperfectly drained pedons have a medium and coarse angular blocky structure while the weIl drained ones have a medium and fine subangular blocky structure. The subsoil colour also shows a good correlation with the drainage sequence and relief. Poorly drained soils occurring in concave and level topography are grey, while the freely drained soils occurring on slightly convex topography are strong brown. Variation in the content of free iron oxide with the drainage sequence may be an indication of the neosynthesis of iron minerals (Hassan, 1982) in 46
0
60 %of cia
40
10
KASHIMPUR
SERIES
30 NOADDA SERI ES
-50 E ~
GERUA SERIES
.c.
ä. 70 ~
"ö
90
110
CHHIATA SERIES
I
I
Fig. 1. Vertical distribution of clay in four soil profiles.
some of these profiles. Data on physical properties and (dithionite) extractable iron contents of several selected soil series representing the drainage sequence are given in table 1. The clay fraction of weIl drained soils of deeply weathered terraces contain 40 to 50 per cent kaolinite (Habibullah, 1971; Hassan, 1982). Vertical homogeneity in composition and content is an indication of inherited origin of the clay minerais. Mineralogy of the clay fraction, CEC and surf ace a.rea representing the drainage sequence are summarized in table 2. From these data two sets of clay mineralogical characters are observable. Soils developed on shallowly weathered (*) terraces contain illite, kaolinite, smectite, vermiculite and intergrade minerals while soils of deeply waethered terraces contain kaolinite and illite. Goethite and haematite may occur in some of the weIl drained profiles.
(*) ShaIlowly weathered terraces include a deep, compact clay deposit which is altered by a secondary weathering sequence down to 10-100 cm depth. In deeply weathered terraces, the unaltered clay layer is not observable wi thin 3 m depth.
47
Table 1. Some data on physical properties and extractable iron contents of several selected soil series (subsoil). Soil series
Drain- Colour age
Tex- Structure ture
Khilgaon
Poor
Grey
Sic
Massive/Pr None coarse
0.2
Chhiata
Poor/ Imp
Olive grey
Sic
Abk coarse Cambic* /medium
2.5
Gerua
WeIl
Brown
Sic
Sbk fine
Cam bic
3.5
Chandra
Imp/ Poor
Greyish brown
Cl
Abk medium
Argillic
3.1
Tejkunipara
WeIl
Yellow brown
Cl
Sbk fine
Argillic
4.5
Weak red Cl to red
Sbk fine
Argillic
8.5
Kashimpur WeIl
Diagnostic subsurface horizon
Dithionite extracted Fe203 % clay
Sic = Silty clay; C = Clay; Cl = Clay loam; Imp = Imperfect; Abk = Angular blocky; Sbk = Subangular blocky; Pr = Prismatic . . * Clay differentiation is mainly due to run-off; effect of ferrolysis is limited within the topsoil (0-10 cm) and along the ped-faces. Table 2. Clay mineralogical data (subsoil). Drainage
Khilgaon
Poor
33.3
278
Illite, kaolinite, smectite, vermiculite and ihtergrade minerals
Chhiata
Poor to imperfect
32.2
266
- ditto -
Tejkunipara Mod. weIl
18.0
155
Kaolinite, illite
Kashimpur
17.6
149
Kaolini te, illi te, goethi te (trace) and haematite
Well
CEC meq./l00g clay
Surface area m 2 /g
Soil series
Clay minerals
48
3. SPATIAL RELA TIONSHIP Spatial relationship between the soil present in a catena is shown in figure 2. This relationship has been established based on the fieldwork from airphoto-interpretation and from the extrapolation of soil properties attributed by the activities of soil forming factors (J enny, 1971). Topography and parent materials which regulate the drainage are probably the main factors responsible for their contribution in differentiating the soil series within the catenas. The effect of time is rather less pronounced. Irregular no hor izon prismatic clay. grey mottled
Level
Undulated Cam bic horizon I angular blocky isub-angu,ar blocky : cloy, greyish I cloy. brown mottled I mottled I
I
Level Argillic horizon : sub-angular blo~ky Icloy loam , strong brol.Jn : mottled : non mottled
Ocm
100 cm "
"
Khilgaon Alluvial valley
I
.....
Ch~ndra
..'
,Chhio.ta ,. ' I Gerua
I
Kas~im~ur
I
, ShallolNly INeathere~ I (S'w'l terrace
Deeply loIeather4 (DVj ttrrace
Fig. 2. Spatial dis tri but ion of soil series in rel at ion to relief and parent material.
4. SOIL SERIES The series is the lowest category in Soil Taxonomy (Soil Survey Staff, 1975). The differentiae in the range permitted for a soil series are narrower than the range permitted in defining the soil family and the units of higher categories. The differentiae for soil series belonging to a family shall meet the following requirements: (i) properties observable or inferable with reasonable accuracy and (ii) the range of propertïes are significantly greater than the normal errors of measurement, observation or estimate by qualified field personnel. Differentiating criteria used 49
Table 3. Correlation table for soil series (characteristics given refer to subsoils). Upland soils Deeply weathered
VaIley soils
Shallowly weathered
Terrace slopes
Cambic B Argillic
Cambic B
Undulated
Level
WeIl.mod. weIl drained :
Poorly to imperfectly drained :
Eroded zone
Deposition zone
i) WeIl drained
Kashimpur series; redder than 5 YR Tejgaon series; yellowish brown to strong brown
Belabo series strong brown to pale brown; non-mottled
Gerua series; brown, intermittent B2 t
Salna series; Noadda series brown, no mottled subsoil intermittent B2 t, mottled ii) Poorly to imperfectly drained subsoil Tejkunipara series; mottled subsoil
Dem ra series; olive, mottled subsoil Chhiata series; grey, mottled subsoil
WeIl drained :
Poorly Poorly drained : drained : Sayek and Pyati Khilgaon series; may be Kalma series; grey --merged with series; to dark grey Tejkuni para grey, and Noadda mottled Karail series; series subsoil burried respectively histic topsoil within 100 cm
Chandra series; greyish, mottled subsoil Ul
o
-- -- -
---
-
-
--
-
-
in the Soil Survey Manual (Soil Survey Sta ff, 1960) have been followed for preparing this legend. Soil series, their occurrence and important diagnostic characteristics have been summarized in table 3. 5. LEGEND TO SOIL SERIES IDENTIFICATION A.
Soils occurring on level, deeply weathered terrace materiais.
A 1.
Have strongly structured clayey subsoil with high base saturation and have a CEC of less than 24 meq/l00 g clay by NH 4 0AC. Used extensively for cultivation and for Sal (Shorea robusta) forests.
A 1a. Have an argillic horizon; base saturation (by NH 4 0AC) is more than 50 per cent in the argillic horizon; mixed kaolinitic; level. 1. Have weIl drained, strongly s~ructured, brownish or reddish (7.5 YR 5/6 to 6/6) clayey subsoil. Clay content does not increase up to 20 per cent within a vertical distance of 7.5 cm or 15 per cent wi thin a vertical distance of 2.5 cm at the upper boundary of the argillic horizons. Tejgoan series. 2. Have characteristics similar to 1 but with an argillic that has, throughout its thickness, a hue redder than (2.5 YR 5/4-10 R 5/4-5/6, dry); value (moist) of less and value (dry) not more than 1 unit higher than the value. Kashimpur
horizon 5 YR than 4 moist series.
3. Have motties with chroma of 2 or less in some part of the upper 25 cm of the argillic · horizon and are saturated in the mottled layer at some depth during most years. Tejkunipara series. 4. Have an aquic soil moisture regime; have characteristics associated with wetness, namely motties, or iron manganese concretions of > 2 mm in diameter or chroma of 2 or less immediately below the Ap or AI. Chandra series. Have moderately structured clayey subsoil; high base saturation. Used extensively for cultivation an'd for Sal (Shorea robusta) forests. A 2a. Have a cam bic horizon; mixed kaolinitic, level. 1. Have a weIl drained, moderately structured yellowish brown (Io YR 5/4-5/8, dry) to strong brown (7.5 YR 5/6-5/8, dry) clayey subsoil; base saturation (by NH 4 0AC) in the cambic 51
horizon is more than 50 per cent; no motties within 60 cm from the surface which have a chroma of 2 or less. Belabo series. 2. Have characteristics similar to 1 but have characteristics associated with wetness immediately below the Ap or A 1 or have motties of chroma 2 or less within 60 cm from the surface. Noadda series. B.
Soils occurring on level to undulating, shallowly weathered terrace materiais.
BI.
Have brownish (7.5 YR to 2.5 Y, dry), freely drained, clayey subsoil occurring on undulated topography. Used mainly for cultivation and for forests.
B 1a . Have a cam bic B, base saturation of 60 per cent or more (by NH 4 0AC) between depths of 25 and 75 cm; mixed illitic. 1. Strong brown (7.5 YR 5/6-5/8, dry) to yellowish brown (10 YR 5/4, dry), have no motties th at have chroma of 2 or less within 60 cm from the soil surface; may have intermittent clay coatings in the cam bic horizon « . 1 % in thin section}. Gerua series.
2. Have characteristics similar to 1 but may have motties of chroma 2 or less wi thin 60 cm from the soil surface. Usually the subsoil colour is paler (10 YR 5/6, dry to 2.5 Y 5/4, dry) than Gerua series. Salna series. B2 . . Olive grey (5 Y 5/2, dry) to olive (5 Y 5/4, dry), imperfectly drained, clayey subsoil occurring on level terrace. Used extensively for paddy cultivation. B2a . Have a cam bic hor i zon; groundwater stands at or near the surface for some part of the year; have an aquic soil moisture regime; mixed illitic. 1. Have a horizon within the upper 75 cm that has motties with chroma higher than 2; C horizon occurs at a depth of 25 cm or less. Demra series.
2. Have chroma of 2 or less in the mottled layer within a depth of 50 cm from the soil surface; have a fj-iable (moist) subsbil; have evidence of ferrolysis* (Brinkman, 1970) in topsoil and subsoil.
Chhiata series.
* Silt skins along the ped fa ces and pores. Ferrolysis involves iron in an orderly sequence of oxidation-reduction CYcles, leaching of cations displaced by iron during reduced phase and finally breaking down of the clay structure in the beginning of the oxidation phase of each cycle. 52
c.
Soils occurring in valleys.
Cl.
Have grey (5 Y 5/1, moist) to olive (5 Y 5/4, dry), imperfectly to poorly draihed subsoils occurring in the run-off zone of valleys. No di agnost ic subsoil horizon present. Used mainly for paddy cultivation.
CIa. Have motties of chroma 2 or less within 60 cm from the surface; groundwater stands at or near the surface for part of the year, mixed illitic. 1. Have grey, mottled, clayey subsoil; an aquic soil moisture regime; may have evidence of ferrolysis in topsoil. Kalma series. C 2.
Have grey (5 Y 5/1, dry) to dark grey (5 Y 4/1, moist) poorly drained clay subsoil occurring in the alluvial zone of valleys; no diagnostic subs6il horizon; groundwater stands at or near the surface for the greater part of the year; mottled grey or nonmottled blueish subsoil. Used for paddy cultivation, part remains perennially flooded.
C 2a . Have aquic soil moisture regimes; organic carbon distribution is irregular within a depth of 1.25 m from the soil surface; mixed smecti tic. 1. Have a buried histic layer or a topsoil that has its upper boundary within 1 m from the soil surface. Karail series. 2. Have characteristics similar to 1 but lacks a buried histic layer or a topsoil within 1 m from the surface. Khilgaon series. Remark Leached soils having a base saturation of less than 50 per cent (by NH 40AC) and the intergrades between Alfisol and Oxisol which may be encountered during detailed investigations shall have to be characterized and accomodated at appropriate places. 6. SOIL SERIES CORRELATI ON Recogni zable di fference in diagnostic properties signi ficant for genetic differentiation and for land-use potentials have been considered for correlating the soil series developed on Madhupur clays. Table 3 shows the relationship of soil properties with landform, relief and drainage conditions. 7. DISCUSSION Soil series for this legend have been split based on the presence 53
of diagnostic horizons, soil moisture regimes, subsoil colour, vertical distribution of clay and base saturation per cent in the subsoil layers. Deeply weathered upland soils have three drainage categories affecting the land-uses. Some soils developed on deeply weathered terrace materials have an argillic horizon, while the others have a cam bic horizon. Aquic characters diagnostic for the Chandra series have probably originated either due to continuous paddy cultivation, by ponding of rain water or locally, due to raised groundwater tables in the monsoon season. WeIl drained soils developed on shallowly weathered terrace materials occur on undulating topography while poorly to imperfectly drained soils occur on level topography. Landuses vary mainly due to the variation in relief which controls the internal and external hydrology of the soils of this region. Differentiae between Tejgaon and Kashimpur series are based strictlyon the subsoil colour, between Tejgaon and Tejkunipara series on subsoil motties, between Tejkunipara and Chandra series on drainage conditions. Noadda and Belabo series have been split depending on the dominance of motties in the subsoil. Noadda series has a pal er subsoil colour than the Gerua series. Their main difference from Tejgaon and Tejkunipara series is the lacking of an argillic horizon. Gerua series has been split from the Sal(1)a series by the presence of a intermittent (Soil Taxonomy, 1975) B2 t in the former. Originally, Salna series was characterized as having a pale yellow subsoil, while Gerua series was having a brown subsoil. Demra and Bhatpara series have been merged together and renamed as Demra. Demra and Chhiata series represent a similar drainage condition, but Demra series is olive either due to alesser period of saturation than the Chhiata series or by inheritance. Moreover, in Chhiata series the effects of ferrolysis are more pronounced. Pyati and Sayek series are probably the eroded phases of the Tejkunipara of Noadda series. These soils are, therefore, merged with Tejkunipara and Noadda series respectively. These soils occupy only a small percentage, therefore, their merging with Tejkunipara and Noadda series will not bring important changes in the soil map. Kal ma series occurs on valley slopes subjected to erosion. Range of characters has been narrowed down in this legend to accomodate only the grey valley soils. Kalma differs from Chhiata series by its occurrence in valleys and from Khilgaon series by occurring in a zone of erosion. Differentiae range of Khilgaon series is broad enough to accommodate grey to dark grey and poorly to very poorly drained valley soils having no diagnostic horizon. This drainage range has identical land-uses, development potential and genet ic aspects. Karail series differs from Khilgaon series by the presence of a buried histic layer or a topsoil wi thin 100 cm from the surface and has different land-uses. 54
Sehematie classifieation (Soil Taxonomy, 1975) of the soil series based on the proposed eharaeteristies is given in table 4.
Table 4. Sehematie elassifieation of the soil series. (Differentiating subgroup eharaeteristies are mentioned bet ween the parenthesis). Soil series
Soil Taxonomy Great group
family
Hapludalf Haplustalfs Paleudal fsl Rhodustal fs Hapludalfs (aquie) Haplaqual fs Eutroehrepts Eutroehrepts (aquie)
Clayey, mixed, kaolonitie, hyperthermie level - ditto -
Deeply weathered terraees Tejgaon Kashimpur Tejkunipara Chandra Belabo Noadda
-
ditto ditto ditto ditto
-
Shallowly weathered terraees - undulated Gerua
Eutroehrepts
Salna
Eutroehrepts (aquie)
Clayey, mixed, illitic, undulated - ditto -
- level Demra Chhiata
Haplaquepts (aerie) Haplaquepts (typic)
- di tto, level - ditto -
Sloping valleys Kalma
Haplaquentsl Haplaquepts
- ditto - sloping.
Karail
Haplaquents (thapto-histie) Haplaquents (typie)
- ditto - smeetitie
Khilgaon
- ditto -
55
ACKNOWLEDGEMENT The authors are indebted to Mr. H. Brammer, FAO Land-use advis or, to the Ministry of Agriculture, Dhaka, for his criticism and suggestions and to Professor Sultan Hassain, who took pains in reading the paper.
REFERENCES Anon. (1963) Reconnaissance Soil Survey Report of Dhaka district (Official report). Soil Res. Dev. Instit., Dacca. Anon. (1965) Reconnaissance Soil Survey Report of Tangail sub-division of Mymemsingh district (official report). Soil Res. Dev. Instit., Dacca. Brammer, H. (1971) Coatings in seasonally flooded soils. Geoderm a, 6 : 5-16. Brinkman, R. (1970) Ferrolysis, a hydromorphic soil forming process. Geoderma, 3 : 199-206. FAO (1971)
Soil Survey Project, Bangladesh. Soil Resources. AGL : SF/Pak 6. Technical Report 3. Habibullah, A.K.M. (I 971) Clay mineralogy of the seasonally flooded soils of East Pakistan. J. Soil Sci., 22 : 179-190. Hassan, M.M. (1982) Characteri zation and pedogenetic study of some forest soils of Bangladesh. Doctorate Thesis, RUG, Belgium.
Jenny, H. (1971) Factors of Soil Formation. McGraw-Hill, New York. Soil Survey Staff (1960) Soil Survey Manua!. S.C.S.-USDA Handbook No. 18. Soil Survey Staff (1975) Soil Taxonomy. A basic system of soil classification for making and interpretation of soil surveys. S.C.S.-USDA Handbook No. 436. 56
Tavernier, R. & Eswaran, H. (1977) Soil properties changes as a function of soil genesis in humid topics. Proc. Clamatrops, Kuala Lumpur. Paper No. 10. Thornthwaite, C.W. (1948) An approach towards a rational classification of climate. Geograph. Review, 33 : 55-97. Wadi a, D.N. (1957) Geology of India (revised ed.). McMillan Co., Ltd., London.
Identificatie-legende voor de bodem series ontwikkeld in de BovenPleistocene afzettingen van Bangladesh Samenvatting Bodemseries van gronden ontwikkeld op de Boven-Pleistocene afzettingen van Bangladesh, die eerder werden ge'fdenti ficeerd en beschreven door het lokale· Soil Resource Development Institute (SRDI), werden opnieuw gedefinieerd in het licht van Soil Taxonomy. De geografische spreiding van deze bodems volgens een klassieke catena werd aangetoond aan de hand van een schematisch diagram; een terreinlegende omvattende de genetische verschillen van deze series werd voorgesteld. Deze legende kan gebruikt worden als terreingids voor de identificatie van de bodemseries ontwikkeld binnen dit landschapspatroon en zal derhalve van nut zijn voor klassifikatiedoeleinden.
Légende d'identification pour les séries de sols développés dans les dépots du Pliocêne Supérieur au Bangladesh Résumé Les séries de sols développés sur les dépots du Pliocêne Supérieur du Bangladesh et préalablement identi fiés et décrits par I'Institut de Pédologie local (SRDI), ont été réétudiés en vue d'une caractérisation dans Taxonomie des Sols. La distribution géographique des sols selon une toposéquence classique est illustrée par un diagramme schématique; une légende de t~rrain, comprenant les différences génétiques des séries est présentée. Cette légende peut être employée comme guide de terrain pour I'identification des séries de sols développés dans ce paysage et sera utile pour leur classification.
57
I
J
I PEDOLOGIE,
XXXVI-I, p.
59 ~ 73, 3 fig., 3 tab., Ghent, 1986.
EEN DIGITAAL TERREINMODEL ALS HULPMIDDEL BIJ HET SPECIFI~REN VAN HET DRAINAGEGEDRAG VAN DE BODEM G. WYSEURE
.S amenvatting Het effekt van de topografie op het drainagegedrag van bodems werd onderzocht. De helling speelt hierbij een rol als maatstaf voor de intensiteit van de laterale flux, terwijl de divergentie een goed criterium blijkt te zijn om verschillen in bodem vochtigheid aan te duiden. Beide parameters werden vergeleken met de draineringsklasse, zoals gedefinieerd voor de Belgische bodemkaarten, en toegepast op het bekken van de Mark. Helling en divergentie van het terrein werden berekend voor elk knoop-punt van een gelijkzijdig netwerk dat over het stroombekken werd getekend en vervolgens vergeleken met de draineringsklasse van dit punt. Deze vergelijking toont aan dat de divergentie beschouwd kan worden als een nuttig criterium bij de specificatie van het draineringsgedrag van de bodems. .S leu te Iwoorden
Topografie, divergentie, bodemdrainering.
1. INLEIDING
Bodemkaarten zoals de Belgische bodemkaart bevatten informatie voor de hydrologische interpretatie van rivierbekkens. Het drainagegedrag is het voornaamste criterium om de bodems in te delen volgens hun deelname in het afstromingsproces. Het rivierbekken van de Mark, bijrivier van de Dender, werd als case-studie in dit artikel behandeld. De bodem- en de topografische G. Wyseure - Assistent aan het Laboratorium voor Bodem- en Waterbeheersing, Afdeling Landbeheer , Faculteit der Landbouwwetenschappen, K.U. Leuven, Kardinaal Mercierlaan 92, 3030 Leuven, België. 59
kaarten van het stroom bekken werden getransformeerd tot digitale bestanden. Daartoe werd een rooster met maaswijdte van 100 x 100 m over de kaarten gepositioneerd en werden de eigenschappen van de knooppunten, de terreinhoogte en het bodem type, onder digitale vorm opgeslagen. Een dergelijk bestand wordt meestal aangeduid als een digitaal terrein en een digitale bodemkaart. Voorkeur werd gegeven aan een roostermethode, omdat deze een eenvoudige overlay van verschillende kaarten, in concreto topografische en bodemkaart, mogelijk maakt. Voor inbreng en weergave van kaarten wordt meestal een vectormethode verkozen. Hierbij worden de coördinaten van de grenslijnen of contourlijnen gedigitaliseerd. De vector en rooster informatie zijn d.m.v. computer programma's omzetbaar. Daar het bekken van de Mark met een totale oppervlakte van 190 km 2, 30% bodems met drainageklas d (matig gleyige gronden) en 7% met drainageklas complex D (zwak tot matig gleyige gronden) bevat, is een verdere specificatie van het· drainagegedragop basis · van topografische criteria, zoals af te leiden uit een digitaal terrein, nuttig. 2. BODEMFYSISCHE INTERPRETATIE VAN DE TOPOGRAFIE Hoewel tot eenzelfde neerslagregime behorend vertonen aan elkaar grenzende bodems, dikwijls belangrijke verschillen in hun waterhuishouding. Twee processen zijn hiervoor verantwoordelijk : de drainage van een grondwatertafel, permanent of stuwend en de laterale stroming bij hellingsgronden met een diep freatisch oppervlak. De drainaget0estand in vlakke gebieden met permanente grondwatertafel wordt vooral bepaald door het ontwateringsniveau, de reactiefactor van het gebied en de bergingscoëfficiënten van de opeenvolgende bodem lagen. Meestal is de bodemkaart van vlakke gebieden qua drainage-intensiteit vrij goed gedetailleerd. Daar niveauverschillen van de grootte-orde van 0,1 m in vlakke gebieden voorname verschillen in waterhuishouding teweegbrengen is een topografische kaart hier onvoldoende nauwkeurig. Daartegenover staat dat de vochttrap van hellingsgronden minder gedetailleerd is weergegeven op de bodemkaart. De relatieve hoogteverschillen op de topografische kaart van deze zone zijn echter meer uitgesproken. Voor deze zones is het a priori niet onmogelijk om met behulp van de topografische kaart het drainagegedrag beter te omschrijven. Veldobservaties hebben onder andere aangetoond dat vochtige zones zich uitbreiden of inkrimpen naargelang de neerslaghoeveelheid (Anderson en Burt, 1978). De variatie in oppervlakte gebeurt niet willekeurig maar volgens bepaalde patronen die o.a. samenhangen met de topografie. Een aanvullende karakterisatie van het drainagegedrag laat toe om continue overgangen op te stellen 60
z (m)
~---------------,
I x.
+~ (m)
Fig. 1. Grafische voorstelling van de terreindiscretisatie in de x-richting. ten einde discrete classificaties van de bodemkaart in elkaar te laten overgaan. Binnen de horizontale dimensie kan het continu verloop van de hoogte van het terreinoppervlak gediscretiseerd worden door op gelijke eindige afstanden, l:1 x, de hoogte van het maaiveld zi te registreren (figuur 1). Gebruik makend van deze discretisatie kan een controlevolume gedefinieerd worden met als bovengrens het maaiveld, als ondergrens de onderkant van de B-horizont of eventueel de bovenkant van het tertiair substraat en als laterale grenzen vertikale halverwege twee opeenvolgende knooppunten (figuur 2). De balans van de fluxen doorheen de grenzen van dit controlevolume berekent de vochtverandering binnen dit volume per eenheid van horizontale afstand en 'per tijdstàp : (l:1S}/(llx.l:1 T) = I - D + FI - FO waarin l:1S de verandering in vocht inhoud in m 3 /m in het controlevolume voorstelt, l:1x de horizontale afstand tussen 2 knooppunten in 61
z
BODEMOPPERVLAK
(m)
A - HORIZON
---------~, ,
"
K;I
"
,"
---
I
:
!, I I
z.
1
I I
I I
B-HORIZO~
D
tJ.x
tJ.x
x.
1
(m)
x
Fig. 2. Definitie van het controlevolume met aanduiding van de grensv lakken. m, 6 T de lengte van de tijdstap in sec, I de netto inkomende flux doorheen het bodemoppervlak in m/sec, D de uittredende flux aan de ondergrens in m/sec, FI de laterale influx in m/sec en FO de laterale out flux in m/sec. De laterale fluxen worden positief gerekend wanneer ze tegengesteld gericht zijn aan de x-as. Het is praktisch niet haalbaar om al de termen van vergelijking (1) voor elk knooppunt van een regionaal gebied te kwantificeren. Daarom wordt in deze studie enkel de invloed van de topografie op de termen van vergelijking (1) geanalyseerd. De ruimtelijke herverdeling van vocht in het landschap wordt vooral toegeschreven aan oppervlakkige afstroming en laterale fluxen doorheen het profiel. Oppervlakkige afstroming ontstaat doordat plaatselijk de neerslagintensiteit de infiltratiecapaciteit van de bodem overtreft en vervolgens het neerslagexces afstroomt. Het is belangrijk hierin twee fasen te onderscheiden (i) vooreerst het ontstaan van het neerslagexces en (ii) het eigenlijk afstromen. Het is 62
duidelijk dat afstromen slechts kan indien voldoende helling aanwezig is. Vandaar dat oppervlakkige afvoer dikwijls vrij eenzijdig geassocieerd wordt met hellingen. Het verband tussen het ontstaan van neerslagexces en helling is echter niet zo duidelijk. Oppervlakkige afstroming geschiedt daarenboven zelden als een dunne film die afstroomt over het terrein. Meestal is de afstroming gelokaliseerd in kleine geultjes en depressies van het terrein, zodat er geen continue herverdeling doorheen het landschap plaatsvindt. De vochtaanrijking blijft meestal beperkt tot de stroombanen, zoals greppels, geulen en grach ten. De netto inkomende flux I en de uittredende flux 0 uit (1) zijn slechts bepaalbaar door uitgebreide meting en als dusdanig (voorlopig) niet af te leiden uit de kaart informatie. De laterale fluxen in de bodem resulteren daarentegen wel in een continue herverdeling van vocht in het landschap. De ogenblikkelijke laterale water flux doorheen het vertikaal grensvlak van het in figuur 2 gedefinieerd controle volume kan als volgt worden berekend : F =
f
Zz
Zo
dil' K (8) ----- cos a dz
(2)
dX
waarin F de flux is doorheen het vertikaal grensvlak in m/sec, en z, Zo en Zz respectievelijk de hoogte, de hoogte van dè onderkant van het controlevolume en de hoogte van het maaiveld langs de laterale grens voorstellen; K(8) de hydraulische conductiviteit is in m/sec, dll'j d x de gradiënt is van de hydraulische waterpotentiaal (dimensieloos) en cos a de cosinus is van de hoek van de normale op het maaiveld met de vertikale. Gespecifieerd naar de in- en uitgaande flux wordt vergelijking (2) respektievelijk Zz FI = z K1(8) ~~-cos a)1 dz o dX
J
~cos a)
OX
0
dz
Daar binnen een bodemtype de vertikale profielopbouw vrij uniform is en het vochtgehalte 8 tussen twee opeenvolgende knooppunten weinig van elkaar verschilt geldt dat : K(8)
KI (8) = KO (8)
zodat
[~ dX
cos a)I -
~ dX
cos a)O ) dz
Op regionale schaal is de K(8 )-relatie fysisch niet bepaalbaar
63
voor een· groot aantal knooppunten. De geometrie van het stromingsdomein wordt bepaald door de discretisatie en door de profielopbouw. De hydraulische potentiaal, welke opgebouwd is uit twee componenten, nl. de matrix-potentiaal en de zwaartekrachtspotentiaal, wordt bepaald door het vochtgehalte en het terreinverloop. De gradiënt van de matrix-potentiaal in de horizontale richting is heel wat kleiner dan de gradiënt van de zwaartekrachtspotentiaal zodat we ons kunnen beperken tot de invloed van de zwaartekracht. Daar de waarde van cos Ct nagenoeg 1 is kan in het kader van de doorgevoerde vereenvoudiging uitdrukking (2) herleid worden tot : fI
= f (( z i + 1
- z i) / ~ x )
(3) waarin fI en FO respectievelijk functie gesteld worden van de gradiënt van het bodemoppervlak. Invoering van vergelijking (3) in vergelijking (1) laat verder toe van de vochtverandering in een controlevolume per eenheid van tijd te benaderen door middel van de uitdrukking:
~S/~ T = f 2 (( zi+ 1 + zi-l - 2Zi)/(~x)2)
(4)
De term tussen rechthoekige haakjes in vergelijking (4) is een maat voor de terreinconcaviteit of divergentie. De vergelijkingen (3) en (4) tonen aan dat de gravitatie- of hellingsgradiënt kunnen beschouwd worden als een maat voor de grootte van de laterale fluxen en de vocht veranderingen. Een concaviteit of positieve divergentie doet het vochtgehalte relatief sneller toenemen dan de andere minder concave terreingedeelten. Een convexiteit of negatieve divergentie leidt tot relatief snellere ontwatering. Voor een twee-dimensionaal rooster met in de knooppunten de hoogte van het bodemoppervlak kunnen de gradiënt en de divergentie van het landschap als volgt worden benaderd : Si,j
= I grad (z(x,y) ) I = { ( (zi+ l,j - Zi_l,j)/(2&x)) 2 + ((Zi,j+l - Zi,j_l)/(2~y))
divi,j
div (grad(z(x,y))
2
I
}"2
(5)
= ((zi+l,j + Zi-l,j - 2Zi,j)/(~x) 2) +
.. 1 + Z·I,J. 1 - 2z·I,J.)/(~y)2) ( (z I,J+
(6)
waarin Si,j en div i,j respectievelijk de gradiënt van de terreinhoogte en de terreindivergentie voorstellen van het knooppunt i,j. Aangezien de gradiënt een vector is, is de grootte gelijk aan de resultante van de x en y componente. De divergentie is de Laplaciaan van de ter-
64
reinhoogte en is een scalaire waarde. Louter mathematisch kunnen de begrippen helling en divergentie ook gedefinieerd worden als de eerste en de tweede afgeleide (Laplaciaan) van de terreinfunctie z = z(x,y). De begrippen zijn dan uitsluitend op de geometrie van het terreinoppervlak gebaseerd (Papo en Gelbman, 1984). De beide begrippen kunnen echter ook een bodem fysische interpretatie toegekend worden. Alternatief kan men, uitgaand van de algemene continuiteitsvergelijking voor bodemvocht, tot hetzelfde numerisch analoog komen zoals weergegeven in vergelijking (6) (Wyseure et al., 1985). 3. TOEPASSINGEN OP HET RIVIERBEKKEN VAN DE MARK Een digitaal terrein werd opgesteld van het bekken van de Mark in het kader van een hydrologische studie (De Meyer, 1982). Naast de topografische informatie werd van elk knooppunt van het rooster ook het bodem type en het bodemgeqruik gedigitaliseerd en in het gegevenbestand opgenomen. Het grid toegepast op het 190 km 2 groot stroomgebied had een maaswijdte van 100 x 100 m. Het bestand bevatte 19051 velden. Een veld van het bestand is volledig opgevuld met de hierboven opgesomde terreininformatie van een knooppunt. De terreinhoogte werd opgenomen in m, het bodem gebruik werd gekwantificeerd volgens een 7-delige kode en het bodem type volgens de nomenclatuur van het Centrum voor Bodemkartering (Gent, België). Een tweede bestand bevat de positionering van de knooppunten. Figuur 3 geeft een voorbeeld van de ruim telijke spreiding van de drainageklassen van het Mark bekken. Het Mark bekken vormt een overgangsgebied tussen de zandleemstreek met in het oosten uitlopers van het Zuid Vlaamse heuvelland en de leemstreek van Midden België. De leemgronden, die het meest voorkomen (ongeveer 71 % van het areaal wordt ingenomen door A texturen), situeren zich vooral op de plateaus welke diep doorsneden zijn. De zandleemgronden welke ongeveer 15% innemen van het stroomgebied situeren zich tussen de plateau en vallei gronden. Deze worden gekenmerkt door een golvend tot geaccidenteerd reliëf met meestal een tertiair substraat op geringe diepte. Het substraat is doorgaans kleiIg en dagzoomt plaatselijk. Het laagste gedeelte van het bekken, de alluviale gronden zijn hoofdzakelijk lemig tot kleü'g van textuur en hydromorf. Het stroomgebied strekt zich uit over volgende bodemkaarten: Geraardsbergen 100 W (Louis, 1975) Denderwindeke 100E (Louis, 1975), Sint-Kwintens-Lennik 101E (Louis, 1957) Lessines 113E (Honnay, 1975), Bever 114W (Louis, 1970), Enghien 114E (Louis, 1956), Rebecq-Rognon 115W (Louis, 1959), Lens 127W (Louis, 1957) en Soignies 127E (Louis, 1957). De letter-typering van de Belgische bodemkaart bleek, mits minimale aanpassingen, geschikt voor digitalisering. Elk bodem type wordt daarbij gekarakteriseerd door 5 tekens. Het eerste teken is optioneel 65
Fig. 3. Drainagekaart van het Mark bekken. Een maaseenheid meet 100 x 100 m. Legende van de kaart : . is de drainageklas e, f en g, § is de drainageklas h en i, ~ is de drainageklas d en D, B is de drainageklas b, c en (A) en 0 zijn de niet gekarteerde gronden. en wordt slechts aangewend wanneer op een diepte geringer dan 125 cm beneden het maaiveld een textureel van de bovengrond afwijkende laag voorkomt. Het tweede teken of karakter wordt steeds 66
Tabel 1. Lettercode van de natuurlijke draineringsklassen voor de zandleem- en de leem streek (volgens het Centrum voor bodemkartering Gent, België). Lettercode
Definitie
• b •
niet gleyig zwak gleyig matig gleyig zwak tot matig gleyig sterk gleyig met reductiehorizont zeer sterk gleyig met reductiehorizont gereduceerd sterk gleyig. zeer sterk gleyig
• c •
• d • · D.
· e . • f •
• g • • h • · i .
ingevuld en staat voor de textuurklasse van de bovengrond. De natuurlijke drainageklasse, zoals waargenomen tijdens de kartering, wordt door een derde, verplicht karakter gedefinieerd. Het vierde karakter duidt op de profielontwikkelingsgroep, terwijl het vijfde optioneel karakter slaat op een variante in profielontwikkeling of in de aard van de ondergrond. De lettersymboliek met legende van de natuurlijke draineringsklassen is weergegeven in tabel 1. Door middel van de vergelijkingen (5) en (6) werden voor elk knooppunt van het rooster de topografische parameters helling en divergentie berekend. Daaropvolgend werden de per knooppunt afgeleide topografische parameters geconfronteerd met het in elk knooppunt voorkomend bodem type. Voor een snelle en efficiënte verwerking van de dataset, het afleiden van de topografische criteria, de statische behandeling van de grootheden en het in kaart kunnen weergeven van bepaalde gebiedskarakteristieken werden alle basisgegevens zoals hoogte, bodem type en bodem gebruik in een twee dimensionele matrix geplaatst. De topografische criteria werden berekend met behulp van de terreinhoogte van de vier omliggende punten en van het punt zelf. Daar de punten van de waterscheidingslijn van het bekken niet omgeven zijn door 4 punten werden deze geweerd uit de berekeningen. Uit de eigenschappen van de waterscheidingslijn, die op topografische basis een ondoorlaatbare randvoorwaarde voorstelt, volgt onmiddellijk dat de helling in elk punt van de waterscheidingslijn nul is en de divergentie negatief. De statistische verwerking van de dataset geschiedde door middel van BMDP programma's (1981), het kartografisch weergeven door middel
67
Tabel 2a. Frequentie overzicht van de natuurlijke drainageklassen van het stroom bekken van de . Mark. Vochttrap (1) • • • • •
b • c • d • O. h •
· i
.
· e • • f • • g •
(%) (2) 20,6 31,7 29,9 6,6 1,9 0,3 5,4 3,3 0,3
(1) Voor de verklaring van de vochttrap zie
tabel 1 (2) Procentueel aandeel van de gekarteerde gronden.
Tabel 2b. Frequentie overzicht van de terreinhoogte van het stroombekken van de Mark. Hoogte interval (m) 10 20 30 40 50 60 70 80 90
- 20 - 30 - 40 - 50 - 60 - 70 - 80 - 90 -100
(%) (1) 1, 1 7,3 14,0 22,2 23,8 19,2 8,9 2,7 0,9
(1) Procentueel aandeel berekend op alle punten van het stroombekken. van arcerings- (Boon, 1982) en contour plot subroutines (Van Houte, 1977). De frequentieoverzichten van de drainageklas, de terreinhoogte, -helling en -divergentie zijn samengevat in tabeLlen 2(a), 2(b), 2(c) en 2(d). De frequenties zijn bepaald door telling van de eigenschappen van de punten. Op de waterscheidingslijn vielen ongeveer 4% 68
Tabel 2c. Frequentie overzicht van de terreinhelling van het stroombekken van de Mark. He llingsin terva I (%) 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0
-
0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5
(%) (1)
8,6 10,2 12,7 14,6 16,6 11,2 10,0 9,7 6,4
(1) Procentueel aandeel berekend op alle pun-
ten van het stroombekken.
Tabel 2d. Frequentie overzicht van de divergentie van het stroombekken van de Mark. Divergentie interval (x 10- 4 )
(%) (1)
-25,0 -15,0 - 5,0 + 5,0 +15,0 +25,0 +25,0
0,4 1,4 11,2 71,4 14,5 0,9 0,2
-25,0 -15,0 - 5,0 + 5,0 +15,0
-
(1) Procentueel aandeel berekend op alle punten van het stroombekken. van het totaal aantal (19051) snijpunten. Via de variantieanalyse (ANOVA) werd getracht na te gaan of de gemiddelde van de terreinhoogte, helling en divergentie significant verschillend waren voor de verschillende natuurlijke drainageklassen. De respectievelijke Ftesten waren telkens significant. Tabel 3 geeft per drainageklas de gemiddelde waarde (x) en de standaardafwijking (sx) van terreinhoogte, -helling en -divergentie.
69
Tabel 3. Het gemiddelde en de standaardfout op het gemiddelde van de terreinhoogte, helling en divergentie per drainageklas.
• • • • •
b • c • d • D. e •
• f •
• g • • h • • i •
Divergentie (10- 4 )
Hoogte (m)
Helling (%)
x
Sx
x
Sx
x
Sx
57,8 55,3 50,7 53,1 36,9 37,6 41,7 37,4 66,5
(0,26) (0,18) (0,22) (0,46) (0,44) (0,61) (2,40) (0,84) (1,40)
3,9 3,4 2,9 4,7 2,3 2,0 3,0 2,8 2,5
(0,046) (0,025) (0,025) (0,076) (0,069) (0,078) (0,320) (0,130) (0,240)
-2,2 -0,036 0,44 -3,06 5,5 5,4 6,3 4,1 5,8
(0,11) (0,070) (0,066) (0,21) (0,21) (0,28) (1,2) (0,33) (0,97)
0,12
3,3
0,017
0,116
0,046
Gemiddelde 52,3
4. BESPREKING VAN DE RESULTATEN De hoogte blijkt een te weinig gevoelig, topografisch criterium te zijn om de bodems van het bestaande stroombekken in te delen in droog of nat. De verschillen in gemiddelde hoogteligging van de gronden binnen de droge drainageklassen is gering. Bij de natte bodems liggen de zeer natte gronden gemiddeld hoger dan de natte. Vooral de gronden met een stuwwatertafel, behorende tot de drainageklas i, liggen gemiddeld hoog. Hoger gelegen slecht gedraineerde gronden zijn meer onderworpen aan zijdelingse watertoevoer van hellingen dan de vlakkere, lager gelegen bodems. Volgens de gegevens in tabel 3 bestaat er geen rechtstreeks verband tussen de drainageklas en de terreinhoogte. Met andere woorden de hoogte kan niet als een bruikbaar criterium gehanteerd worden om de natuurlijke drainageklas meer expliciet te karakteriseren. De gemiddelde helling van de gronden, berekend voor de verschillende drainageklassen van het Mark bekken, varieert van 2 tot 4,7%. Het gehele gebied kan dus als zwak hellend gekarakteriseerd worden. Binnen de drainageklassen b, c, d, e en f hebben de goed gedraineerde gronden een grotere helling dan de matige en de onvoldoend gedraineerde. Het verschil in helling tussen de natte gronden (de drainageklassen g, h en i) is minder duidelijk bij een gebrek aan voldoende waarnemingspunten. Opvallend is dat de gronden behorend tot de complexe drainageklas D gemiddeld de grootste helling bezitten. De drogere gronden, de gronden met drainageklas b, c en d onderscheiden zich duidelijk van de overige in divergentie. De diver70
J
gentie is laag positief of negatief, hetzij licht concaaf tot vlak, tot uitgesproken convex. De gronden behorend tot de complexe drainage klas 0, zijn het sterkst convex. De gronden behorend tot de natte drainageklassen zijn uitgesproken concaaf. De verschillen in concaviteit van de gronden, behorend tot de natte klassen, zijn eerder gering. Het verschil tussen de natte gronden (e, f en g) wordt veeleer bepaald door de diepte van het grondwater en zijn schommelingen dan door laterale stroming. De terreindivergentie van de gronden met stuwwatertafel, (h en i) is ook sterk positief hetgeen er op wijst dat naast de aanwezigheid van storende lagen of horizonten er ook een belangrijke topografische invloed is die bijdraagt tot de natheid van deze gronden. De gronden behorend tot de com plexe drainageklas 0, nemen een zeer bijzondere plaats in. De divergentie van deze gronden is gemiddeld het meest negatief en de terre.i nhelling het grootst. Volgens deze eigenschappen zouden de gronden behorend tot de complexe drainageklas 0 de droogste moeten zijn. De interne drainage van deze gronden is over het algemeen zwak of onvoldoende ingevolge de aanwezigheid van een klei-aanrijkingshorizont, een klei- of een klei-zandsubstraat op geringe diepte. De combinatie van een langzame interne en van een excessieve externe drainage maken dat het drainagegedrag van deze gronden moeilijk definieerbaar is. Daarenboven bemoeilijkt het substraat soms de interpretatie van de intensiteit en de diepte van voorkomen van de roestverschijnselen. Hoewel voor de berekening van de divergenties hogere eisen gesteld worden aan nauwkeurigheid van de hoogtebepaling en aan de fijnheid van het rooster dan voor het berekenen van de hellingen laat de divergentie toe dat aan sommige ontwateringsklassen, zoals bv. de overgangsklasse d en de klasse complex D, een duidelijke plaats wordt toebedeeld. 5. BESLUIT
De helling en de divergentie van elk punt in het landschap kunnen door middel van een numerisch eindig verschillen analoog berekend worden vertrekkend van de terreinhoogte in de snijpunten van een equidistant rooster. Binnen het bekken van de Mark is de divergentie het topografisch criterium dat het best in staat bleek om de droge drainageklassen, met inbegrip van de overgangsklas d, te karakteriseren. Meer nog dan de helling blijkt de divergentie een topografisch hulpmiddel te zijn om het drainagegedrag, zoals weergegeven op de bodemkaarten, beter te specifiëren.
71
LITERATUUR Anderson, M.G. & Burt, T.P. (1978) Towards more detailed field monitoring of variabie source are as. Water Resources Res., 6 : 1123-1131. Boon, W. (1982) Arcer ingssubrou tines. Labo voor Landtechniek. Faculteit der Landbouwwetenschappen, K.U. Leuven (niet gepubliceerd). De Meyer, J. (1982) Hydrologische analyse van de Marke, bijrivier van de Dender. Ingenieursverhandeling van de Faculteit der Landbouwwetenschappen, K. U. Leuven, 83 p. Honnay, J.P (1975) Kaartblad van Lessines 113E. Centrum voor Bodemkartering, Gent, België. Louis, A. (1956) Kaartb lad van Enghien Il4E. Centrum voor Bodemkartering, Gent, België. Louis, A. (1957) Het kaartblad van Sint-Kwintens-Lennik 101E, Lens l27W en Soignies l27E. Centrum voor Bodemkartering, Gent, België. Louis, A. (1959) Kaartblad van Rebecq-Rognon 115W. Centrum voor Bodemkartering, Gent, België. Louis, A. (1965) Kaartblad van Denderwindeke 100E. Centrum voor Bodemkartering, Gent, België. Louis, A. (1970) Kaartbléld van Bever 114W. Centrum voor Bodemkartering, Gent, België. Louis, A. (1975) Kaartblad van Geraardsbergen 100W. Centrum voor Bodemkartering, Gent, België. Papo, H. & Gelbman, E. (1984) Digital Terrain Models for Slopes and Curvatures. Photogrammetric Engineering and Remote Sensing, 50, (6) : 695-701. Van Houtte, J. (1977) Fortran IV programma voor contour mapping van een equidistant grid. Departement Metaalkunde, Faculteit der Toegepaste Wetenschappen, K. U. Leuven (niet gepubliceerd). 72
Wyseure, G., Feyen, J. & Boon W. (1985) Topographic analysis as an aid in the specification of the drainage behaviour of soils. In preparation.
A digital terrain model as a technical instrument for the specification of the drainage behaviour of soils. Summary The effect of topography on the drainage behaviour of soils is analyzed. The slope is shown as a measure for the intensity of the lateral flux, while the divergency appears to be a good criterion for specifying the changes in soil moisture. Both parameters defined were compared to the drainage classification, as specified in the Belgian soil map, with application to the Mark catch ment area. The slope and the divergency of the land were calculated for each modal point of an equidistant grid that was super-imposed on the catchment and compared with the drainage classification of those points as defined on the Belgian sbil map. The comparison reveals that the divergency can be considered as a useful criterion in the specification of the drainage behaviour of soils.
L 'interprêtation du dynamisme de drainage des sols à l'aide d'un modèle digital de terrain. Rêsumê L 'effet de la topographie sur Ie comportement du drainage des sols est analysê. La pente se montre plutot comme une indication de 1'intensitê des flux latêraux dans Ie sol, tandis que la divergence est un bon critère pour Ie dynamisme des variations de la teneur en eau. Ces deux paramètres, définis du point de vue de la physique de sol, ont êté comparês à la classification du drainage naturel des sols, comme dêfini sur la carte pédologique beIge. La comparaison indique que la divergence peut être considérêe comme un critère utile pour 1'interprêtation du drainage naturel des sols.
73
IPEDOLOGIE, XXXVI-I, p. 75 - 84, 5 fig., Ghent, 1986.
VARIABILITIES IN THE LENGTH OF THE GROWING PERIOD APPLIED TO PUGLIA, ITALY M.M. BELLOCCHIO M. FALCIAI P. TRUCCHI
Abstract The concept of the growing period, as defined by F AO t has been applied for Puglia, southern Italy. Variabilities in the leng th of the growing period were first studied in relation to different methods in computing the potential evapotranspiration and on the basis of different levels of available water and effective rainfall/gross rainfall ratios. It was hereby found that, owing to a lack of complete climatic parameters, the FAO method is not always applicable. It was furthermore observed that other formulas lead to different and contrasting results. l The impact of available water and effective/gross rainfall is different on the length of the climatic characteristics of the different ar_eas. It has particularly been noticed that the growing period strictly depends on summer rainfall. A frequency analysis of the growing period va lues, calculated for the 1951-70 period, has confirmed that those values are not very reliable when average temperature and rainfall data are used. A month by month computation of the growing period fits better in mediterranean areas which are characterized by considerable summer rainfall variations. Key words Growing period, agro-climatic land evaluation.
M.M. Bellocchio, M. Falciai and P. Trucchi - Istituto di Idronomia, Università di Firenze, Via S. Bonaventura, 13, Firenze, Italia. 75
1. INTRODUCTION The concept of the growing period (GP) method has been examined and applied in Puglia, southern Italy, in the framewoik of a research program me, the aim of which was to define a parametrical methodology for an agroclimatic land evaluation and classification in mediterranean areas. This study was undertaken as a joint programme between the Junta de Andalucia, Agencia de Medio Ambiente, Spain and the Institute of Geopedology, University of florence, in conjunct ion with the Research Institute of Bari, Italy. The method, proposed by the fAO and applied to the African continent, defines the growing period (GP) as the period of the year when neither the air temperature, nor the soil water availability limits cr op production. More precisely, the GP is the period, in days, during the year when R (Rainfall) equals or exceeds half the PET (Potential Evapotranspiration), plus a number of days required to evapotranspire the water reserve stored in the soil in the previous months. A further factor we have taken into account is the average temperature which, for each erop, must exceed a minimum value in order to achieve a normal photosynthetic activity. In applying this to the Puglia region we have considered the AET /PET ~ 0,5 relationship. In this context the potential evapotranspiration or PET has been calculated following the Thornthwaite formula, owing to the lack of the climatic data other than temperature. Actual evapotranspiration or AET has been calculated as the sum of the rainfall (R) + a fraction of the available water (AW) in the soil not exceeding 75% of the AW in the previous month. Moreover, we have chosen a minimum temperature value of 15°C and 200 mm of water storage. A map, indicating the length of the growing period has been drawn using the data from 57 meteorologicel stations (fig. 1) in the period 1951-1970. Before computing the growing period in the whole region variations of the length of GP, in relation to the standards vartation, were studied in Bosco Umbra and Taranto; these two stations reflect the extremes in altitude, latitude, summer rainfall and temperature of this region. The standards referred to concern different methodologies in the PET computing, different levels of AW (Available Water), a varying ER/GR (Effective Rainfall/Gross Rainfall) ratio and a frequency analysis of the GP. 2. INfLUENCE Of PET ESTIMATION METHOD
Besides the Thornthwaite formula, three of the four methods applied by the fAO (1977) for estimating the PET have been considered : a) Blaney and Criddle, b) Radiation and c) Penman.
76
:j
Fig.l. Length of growing period (expressed in days) in the area concerned.
SP (day.)
120
1-:.:.:,;,1Boseo
100 .
D TARANTO •
.. .. UMBRA
80.
&0
r;:;'1-.
40
20
Blaney Criddle
Radiation
Penman
Thornthwalte
Fig. 2. The influence of PET estimation method, on the growing period. In studying the variation of the GP in relation to different methods of estimation, average temperatures and rainfall (ER/GR = 100%) have been used, and a minimum temperature value of 15°C and a 200 mm AW have been considered. The results are illustrated in figure 2. There is a GP constant leng th when the PET is calculated by the three FAO methods (34 days for Bosco Umbra and 43 days for Taranto). Longer and reversed GP va lues are however observed when PET is calculated by the Thornthwaite formula (124 days for Bosco Umbra and 73 days for Taranto). 3. INFLUENCE OF AW The dependence of the growing period from available water variability has been investigated for AW values of 100, 150, 200, 250, 300 mm. PET has been calculated by the Thornthwaite formula; an AET / PET ~ 0,5 an ER/GR = 100% and a temperature of 15°C have been fixed. The results at the two stations are very different. In Bosco Umbra the GP is independent of the AW levels (124 days); in Taranto th GP increases from 43 to 104 days when AW passes from 100 to 300 mm (fig. 3).
78
GP
BOSCO UMBRA
(days)
120
---------------
100 TARANTO
80 60 40 20
100
150
200
250
300 AW (mm)
fig. 3. The influence of available water capacity of the soil (AW) on the growing periode
These different patterns can be explained by different values of rainfall deficit (RD) in the summer months. In Bosco Umbra the rainfall deficit is low and the soit moisture reserve is scarcely utilized; hence, the GP length is not influenced; in Taranto however, there is a considerable deficit and a lot of the reserve is utilized, insofar that the GP length is greatly influenced. 4. INfLUENCE Of ER/GR The ratio ER/GR is difficult to determine. This value depends on the intensity of the rainfall, in relation to the infiltration velocity; it also depends on the amount of rainfall, in relation to the soil storage capacity, and on the PET value. The influence of different ER/GR values (80, 60, 40, 20%) on the GP has been investigated using the same va lues of PET, AW and temperature. In relation to the ER/GR decrease in both stations, the GP decrease is considerable, but particularly so in Bosco Umbra where the summer rainfall is more pronounced (fig. 4).
79
GP
(days)
120
I
r---
Basca UMBRA " I I
100
I
I
80
·1
TARANTa
60 40
20
20
40
60
80
100 ER/ ("') SR
Fig. 4. The influence of ER/GR.
5. FREQUENCY ANAL YSIS The standards for the computation of the GP refer to average va lues of temperature, rainfall and PET. As a matter of fact, when the GP calculation is based on average data, it results, as De Pauw (1983) noticed, in a considerable over-evaluation in environments characterized by considerable rainfall variations. This fact has been found out calculating a month-by-month-GP for the twenty years between 1951 and 1970. PET has been obtained using the Thornthwaite formula and taking AW = 200 mm and ER/ GR = 100%. In this respect it could be noticed from figure 5 that the ave rage GP is longer than the mean of the GP va lues. The difference between those two GP va lues is 17 days for Bosco Umbra and 5 days for Taranto. Further considerations can be made : - while in Bosco Umbra the range of GP variability is 112 days because of the remarkable variations in summer rainfall , this range is smaller in Taranto where summer rainfall does not vary so much from one year to another; - the only mean of thè GP va lues is neither reliable nor useful by itself in order to forecast. For example, if we want a guaranteed 80
GP
(days)
130
/"
/~~ .,. ... --"'" .
,,"
~
Boseo UMBRA
. 120
,,'"
100
,,'
I.
80 I
I
I
I
I
J
. TARANTO
J ./
60 40 20
c E
~
GI ~
.&:;
0 Ol
c IV
GI
E 111
Q.
c
til
=ë GI E
. <-'
a.:
...GIU
~
-...
<-'
DI
~
GI
>
•
~
,.: 111
E
s! 0
Fig. 5. Frequency analysis results. GP for 9 years out of 10, we should pass from 106 days to 52 days in Bosco Umbra and from 68 days to 55 in Taranto. 6. CONCLUSIONS The study confirms, as other authors have already shown, the difficulties of the GP calculation in mediterranean environments characterized by considerable summer rainfall variations. Besides, in the study area, as weIl as in many other are as, difficulties and problems connected with the ETp estimate method still persist; in fact, not always other climatic data besides temperature are available. Furthermore is emphasized the importance of a GP frequency analysis.
81
REfERENCES De La Rosa, D. & Magaldi, D. (1982) De un sistema de evaluacion de tierras para regi.o nes mediterranes. Soc. Esp. Cien. Suelo, Madrid. De Pauw, E. (1983) The concept of dependable GP and its modelling as a tooI of land evaluation and agricultural planning in the wet and dry tropics. Pedologie, 32 (3) : 329-48. F.A.O. (1976) A framework for land evaluation. FAO Soil Bull., 32, Roma. , F .A.O. (1977) Crop water requirements. FAO Irrigation . and Drainage Paper, 24, Roma •. F.A.O. (1978) Report on the Agro-ecological zone projects. Methodology and results for Africa. World Soil Resources, Report, 48, Roma. Higgin~" G.M. & Kassam, A.H. (1981) The F.A.O. agro-ecological zone approach to determination of land potential. Pedologie, 31 (2) : 147-68.
Magaldi, D. & Ronchetti, G. (1979) Basi pedologiche della fertilità. L 'Italia agricola, anno 116, n 0 2. Magaldi, D. & Ronchetti, G. (1984) Report on the developing project for land evaluation in Italy on 1: 1.000.000 scale. 'In : J.C.F .M. Haans et al (eds) : Progress in land evaluation. A.A. Balkema, Rotterdam, pp. 57-66. Pinna (1977) L 'evapotranspirazione e il bilancio idrico secondo Thornthwaite. Climatologia, UT ET , Torino, pp. 341-344 and 433-440.
82
La variabilité dans la période de croissance. Application à la région de Puglia, Italie. Résumé Le principe de la période de croissance (GP), tel que défini par la FAO, a été appliqué à la région de Puglia. La variabilité de Ola période de croissance a d'abord été étudiée en fonction de la méthodologie adoptée pour Ie calcul de l'évapotranspiration potentielle (ETP), du niveau d'eau disponible (AW) et du rapport entTe la pluie efficace et totale (ER/GR). Il a été noté que, par manque de données climatiques completes, la méthode de la FAO n'est pas toujours applicabie. Il a également été observé que les differentes formules conduisent à des résultats différents ou même contrastés. Les valeurs de WA et du rapport oE R/GR influencent Ie GP suivant les caractéristiques climatiques du milieu; en particulier on a remarqué que la longeur du GP est nettement dépendante des pluies d'été. Une analyse de fréquence des GP, calculée sur vingt années, a confirmé que ces resultats sont peu dignes de foi, lorsqu'on considère les données moyennes de temperature et de plu ie; par contre, un calcul mensuel de la période de croissance est beaucoup plus pertinent dans les milieux méditerranéens, caractérisés par une forte variabilité des pluies d'été.
Riassunto Il metodo del periodo di crescita (GP), proposto dalIa FAO, è stato applicato alla regione pugliese. Prima si è tuttavia indagnato sulla variabilità del GP in funzione della metologia adottata per il calcolo delI'Evapotranspirazione potenziale (ETP), del livello di acqua disponibile (AW) ed infine del rapporto tra pioggia efficace e pioggia totale (ER/GR). Si è contatato come, non essendo sempre applicabili i metodi per la stima dell 'ETP proposti dalIa FAO a causa della mancanza di raccolte complete dei diversi parametri climatici, altre formule portino a risultati assai diversi e contrastanti. I valori della AW edel rapporto ER/GR influenzano diversamente il GP a seconda delle caratteristiche climatiche della zona; in particolare si è notato come la lunghezza del GP si strettamente dalle piogge estive. Una analisi di frequenza dei GP calcolati su 20 anni di osservazioni, ha conferrnato come questi siano poco attendibili quando si
83
parta da dati medi di temperatura e pioggia e come invece un calcola a calendaria sia assai piu rispondente in ambienti mediterranei caratterizzati da una forte variabilità delle piogge estive.
84
New publications
Nouvelles publications
Nieuwe uitgaven
Publications Committee of XI ICSMFE Proceedings of the Eleventh InternationaL Conference on Soil Mechanics and Foundation Engineering, San Francisco, 12-16 August 1985, 5 vol. A.A. Balkema, 1986, 2469 p., ISBN 906191 560 0 Price (cloth) : 76 US dollars or 62 UK Pounds per volume. As the First International Conference on Soil Mechanics and Foundation Engineering was held in 1936, these proceedings of the Eleventh International Conference are devoted to a half century of international exchange of knowiedge. The proceedings are organised around a framework of nine themes. For every theme there is a theme lecture, discussing the current state of knowledge and suggesting areas of future research. Technical papers provide supplementary information on every theme. The themes are : 1. Soil mechanics - Property characterization and analysis procedures. 2. New developments in field and laboratory testings of soils. 3. Geotechnical aspects of environmental control. 4. Piles and other deep foundations. 5. Geotechnical engineered construct ion. 6. Evaluating seismic risk in engineering practice. 7. Stability of natural deposits during earthquakes. 8. Comparison of prediction and performance of earth structures. 9. Geological aspects of geotechnical engineering. The proceedings consist of four volumes. Volume 1 contains eight of the theme lectures. The lecture of theme 8 will be included in a post-conference volume 5. The theme lectures have on the average about 50 pages of densily spaced text and figures and present very valuable states of the art. Volumes 2, 3 and 4 contain 443 technica I papers by authors from 47 countries. The papers are ordered by theme number and pertinent subdivisions. The technical papers have a limited number of pages and consequently give only a brief but mostly relevant discussion of theoretical and practical aspects of the conference themes. In conclusion, these proceedings constitute a very good review of the progress over the past 50 years, in soil mechanics and foundation engineering, all over the wor ld. However, these subjects might be less interesting for the soit scientist and also the rather high price is a negative factor. Nevertheless, when considered per page, the price is in the normal range of scientific literature, making a purchase of these proceedings worth while. F. DE SMEDT 85
J. Murphy et L.H. Sprey. Evaluation permanente du dêveloppement agricoie. Publ. 34, Intern. Instit. Land Reclam. and Improvement, P.O. Box 45, 6700 AA Wageningen, Hollande, 1984, 272 p., ISBN 90-70260-891. Commandes et informations supplêmentaires : ILRI, B.P. 45, Wageningen, Pays-Bas. L'êvaluation permanente est dêcrite par les auteurs du livre comme un systême de collecte et d'analyse de donnêes, suivi d'une rêtroaction (feedback) de l'information agricole pendant la vie d'un projet. Une telle procêdure d'êvaluation peut permettre aux directeurs de projet d' ajuster leurs activitês aux besoins contraintes du cultivateur et peut fournir en même temps aux planificateurs une information continuellement actualisêe sur les rêsultats obtenus. Dans Ie prêsent ouvrage les auteurs dêcrivent unI'programme d'êvaluation permanente appliquêe à l'agriculture de subsistance de la rêgion sêmi-aride en Afrique de l'Ouest, ou notamment les exploitations familiales pratiquent des cultures pluviales destinêes à l'autoconsommation. Quoique les nombreux exemples sont presque toujours originaires de cette zone Ie livre contient suffisamment de matiêre pour permettre des extrapolations à d'autres rêgions climatiques ou à d'autres types de projets de mise en valeur. Le livre se prêsente en\i deux parties. La premiêre partie, Principes Gênêrales, explique êtape par êtape comment organiser un service d'êvaluation permanente; elle inclut des chapitres tels que la dêlimitation du travail d'un service d'êvaluation, l'allocation des ressources d'un tel organisme, les rêgles gênêrales pour les interviews, la prêparation d'un program me d'enquêtes, Ie traitement des donnêes, etc••• La seconde partie, Mêthodologies, traite plus en dêtail Ie déroulement des activitês dêcrites dans Ie premier volet; y sont êgalement discutês !: la nature et la formation du personnel, Ie choix et Ie röle de l'êchantillon, les pratiques culturales et l'estimation de la product ion agricole et de 1'êlevage, ••• Ce livre s'adresse en premier lieu à ceux qui seront appelês à programmer et à analyser la collecte continue des donnêes et à surveil Ier la mise en oeuvre de ce program me, soit pour un pro jet soit pour un organisme êtatique. Pour l'agronome, Ie pêdologue ou tout autre expert faisant partie du groupe interdisciplinaire pour Ie dêveloppement rural l'ouvrage donne êgalement un aperçu global qui permet de mieux comprendre la mêthodologie de travail employêe dans certaines recherches sociologiques et/ou agro-êconomiques. Ce volume constitue la traduction française de la publication ILRI n° 32, êditêe en langue anglaise sous Ie titre "Monitoring and evaluation of agricultural change". eeci explique l'apparition dans 86
Ie texte de certains anglicanismes et d 'une terminologie souvent peu courante. W. VERHEYE
J. Murphy et L.H. Sprey Intoduction aux enquêtes agricoles en Afrique. Publ. 35, Intern. Instit. Land Reclam. and Improvement, P.O. Box 45, 6700 AA Wageningen, Hollande, 1984, 134 p., ISBN 90-70260-956. Commandes et informations supplêmentaires : ILRI, B.P. 45, Wageningen, Pays-Bas. Ce livre constitue un manuel d'accompagnement de la publication prêcêdente; il a particuliêrement êtê êcrit pour les êtudiants qui suivront une formation aux enquêtes °agricoles. Son but est de montrer aux enquêteurs pourquoi on leur demande de collecter des donnêes, comment elles sont utilisêes et comment la qualitê de leur travail influencera les rêsultats d'une enquête agricoie. L'ouvrage prêsente des informations gênêrales sur les pratiques culturales, les mathêmatiques, la gestion agricole et les mêthodes d'exploitation - matiêres qu'un enquêteur doit comprendre pour pouvoir s'acquitter correctement de sa tache. Le livre est divisê en leçons, d'un niveau três simple au dêbut et avec une difficultê qui s'accrott progressivement. A la fin de chaque chapitre se trouve une sêrie d'exercices ou de questions, dont les solutions sont donnêes ultêrieurement dans Ie texte. Ce livre a êtê publiê à 1'intention des enquêteurs, dont Ie niveau d'êducation est de quelques annêes d'êtudes secondaires. Ils y trouveront plus particuliêrement un exemple du type de travail d'enquêteur effectuê auprês des petits cultivateurs de l'Afrique Occidentale sêIJ1i-aride. Les surveillants de ces mêmes enquêteurs, qui en principe ont un niveau d'instruction plus êlevê et une certaine expêrience de la collecte des donnêes, tireront certainement aussi profit de la lecture de ce manuele W. VERHEYE
F. Andersson & B. Olsson (eds.) Lake Gardsjön. An acid forest lake and its catchment. Ecological Bulletins (Stockholm) 37. Publishing House of the Swedish Research Councils, P.O. Box 6710, S-11385 Stockholm, Sweden, 1985, 336 p., ISBN 91-86344-25-0. Price : 44 US dollars. 87
This volume gives a comprehensive documentation of an ecosystem approach in investigating the effects of acid deposition on soils and water. It contains an analysis of all ecological conditions in an acidified lake prior to liming. As ecosystem Lake Gardsjön and its surroundings was chosen because of its location in a region wi.tlh high deposition of air pollutants. Also recent information on the chemistry and biology of this lake was available. This bulletin is divided up into five parts, each of them containing a number of original publications which can be read as independent contributions. The first part, with 7 papers, handles the development of the Lake Gardsjön area. The aim of these papers is to give information on previous conditions in terms of land history, use and management up to the present-day situation. The second part, with 9 papers, deals with water and element cycling. Emphasis is put on pathway and residence time of water passing through the soU, on the type of buffering systems operation in the soil and especially on the role of aluminium as a soil buffering agent. The leaching from the soU into the lake of other elements is also studied. The third part, with 8 papers, handles biotic structures and relations. In these papers, the authors prove that it is to some extend a m~sinterpretation that a strongly acidified lake is a de ad lake. Changes in the foodwebs resulted in another relative distribution of primary and secundary production of a lake, contributing to the changes in lake chemistry. The fourth part, with 8 paper-s, covers sediment proper ties and processes. Changes oGcurring in the seddments are discussed and quantified. In the fifth part, with 3 papers, a s ynthesis is given (1) on the processes contributing to soil and water acidification, (2) on the floraand fauna of the acidified lake and (3) on attempts to quantify the sediment processes. This volume is of interest to soU and water biologists, environmentalists, ecologists, chemists as weIl as to experts in modeIling transport processes.
o.
VAN CLEEMPUT
D. Dent Acid sulphate soils : A baseline for research and devolopment. Publication No. 39 of the International Institute for Land Reclama88
tion and Improvement, 204 pp., 44 fig., 31 plates, 18 color plates; ILRI, P.O. Box 45, 6700 AA Wageningen, The Netherlands, ISBN 90 70260 98 O. Price : 38 US dollars. This most interesting and complete review on Acid Sulphate Soils subdivided into seven parts : Fundamental properties of acid sulphate soils. Management. Chemical and physical processes in acid sulphate soils. Field relationships, soil horizons and soil profiles. Soil classification. Soil patterns. Soil survey and land evaluation. The fundamental properties of acid sulphate soils are taken in review. Af ter situation of the proble'm and guidelines for field identification, the natural environment is described. The text is illustrated with a series of interesting plates and figures. Agronomic and environmental problems are commented and considerations with regard to social and economic implications have been formulated. In a series of alternative management strategies, flooded rice cultivation is compared with rainfed rice cultivation. Controlled high watertable management for perennial crops considers cultivation of oil palm and establishment of grasland. Reclamation techniques through drainage and salt water leaching are discussed. Fertilizer management is suggested for a series of utilization types. In the chapter on chemical and physical conditions, at tent ion is drawn on the potential acidity. As weil the processes related to formation of pyrite and oxidation are studied. Modelling the rate of acid production suggests, besides a static and a ,dynamic model, a method of calculation of the ra te of oxidation. The whole is illustrated. by a case study on the Gambi Barrage Scheme. The part on field relationships, soil horizons and soil profiles considers the sequential development of horizons in the tidal zone and the development of horizons foilowing drainage. The text is illustrated with color plates of profiles and thin sections and some profile descriptions complemented with analytical data. For the classification of acid sulphate soil the work recommends to use soil properties important for management and land-use planning. These characteristics are listed and discussed. Soil patterns of acid sulphate environments are described for the tidal zone, inter-tidal zone and reclaimed landscapes. These regional soil patterns are illustrated with extracts of soil maps. The last chapter describes soil survey and land evaluation of acid sulphate soils. The main objectives of a soil survey are considered is 1. 2. 3. 4. 5. 6. 7.
89
as : land-use planning, project design and implementation. The soil survey design is discussed and attent ion is drawn on the use of air photographs and Landsat satellite imagery. for the identification procedure one compares acid sulphate soils and potential acid sulphate soils. A model for full profile characterization is suggested and comments are formulated for the different characteristics. for land evaluation the successive stages in the interpretation adapted to this specific environment are listed; some land utilization types have been defined as weIl. A list of land qualities affected by characteristics of acid sulphate soils is suggested. The book of David Dent represents a most complete work on acid sulphate soils; it is a valuable contribution to the knowiedge, management and evaluation of this specific environment. It can be recommended to soil scientists, land-use planners and agronomists.
c.
SYS
G. Van de Goor and G. Zijlstra Irrigation requirements for double cropping of lowland rice in Malaya. Intern. Instit. Land Reclam. and Improvement (ILRI), Publ. 14, P.O. Box 45, 6700 Wageningen, The Netherlands, 1983, 56 p., 17 fig., 17 tables, ISBN 90 70260 840. Orders to : ILRI, P.O. Box 45, Wageningen, The Netherlands. This publication summarizes the results of a study of irrigation water requirements for lowland wet rice cultivation in Malaysia, both in the "wet" main-season and in the "dry" off-season. As many rice growing areas in Asia are more or less similar to the Malaysian situation, the conclusions of this work are believed to be of interest to many other countries as weIl. Af ter a description of the environmental conditions in the test are as (chapters 2 and 3) and a discussion of the factors determining the water requirements in rice fields, i.e. evapotranspiration, percolation, water height in the field and soil saturation (ch. 4), formulas are developed for irrigation requirements (ch. 5). With the aid of these figures a suitable irrigation capacity of new projects may be selected, taking into account the length of the presaturation period, the dependable rainfall and the effectiveness of water control. The closing chapters of this book discuss some measures to obtain efficient distribution of water and good water control in the off-season and give a review of the prospeets of double cropping as a permanent system with regard to yield level and soil conditions. This work constitutes a practical handbook for soil scientists and 90
rural planners having to deal with irrigated rice projects. W. VERHEYE
J.M. Van Staveren and D. Van Dusseldorp (eds) framework for regional planning in developing countries. Intern. Instit. Land Reclam. and Improvement (ILRI), Pub I. 26, P.O. Box 45, 6700 Wageningen, The Netherlands, 345 p. 1983, ISBN 90 70260 832. Orders to : ILRI, P.O. Box 45, Wageningen, The Netherlands. This book describes the guidelines for regional planning in predominantly rural areas and concentrates on the importance of concerted, multidisciplinary efforts in the rural planning procedures. The weIl documented study includes 6 major chapters and a series of technical annexes which discuss into more detail the approach and methodology in the different subdisciplines. Af ter an introduction and positioning of regional planning in the process of planned development (chapters 1 and 2), the interdisciplinary procedure of regional planning is discussed and the different stages in such procedure are com mented (ch. 3 and 4). Chf3.pter 5 advises on reporting methodologies and on the structure of such reports in order to en su re a well-ordered presentation of the of ten massive and complex information. Chapter6 is the core of the framework and focusses on "the planning of the planning process", i.e. on the coordination and integration of the various streams of essential information. This framework certainly does not profess to be a handbook for planning for each separate discipline nor does it claim to introduce any spectacular new elements or concepts. It merely reflects the systematized experience of contributors from different disciplines who have been involved in planning in various developing countries. It focusses on the relations between the activities that are fundamental for interdisciplinary cooperation. This very interesting volume can be of use for various groups of people : students, preparing themselves for positions that may bring them in contact with regional planning, will find here an excellent background information; experts preparing themselves for participation in the studies for a regional plan and who have not worked in such exercise before, and administrators involved in the preparation, supervision or implementation of regional plans will find useful information on the concepts and tasks of the different subdisciplines and on the way this infonnation has to be integrated by the team leaders and/or decision makers. W. VERHEYE 91
Pedologie, 1986 (XXXVI)-l,
92 p., Ghent, 1986.
SUMMARY
SOMMAIRE
INHOUD
J. Vandamme & R. Biston Analysis of the suitability of Belgian soils for growing fine carrots. J. Vandamme & R. Biston Assessment of the optimum soil nutrient composition for fine carrot growing under Belgian farming conditions. L. Boedt & W. Verheye Evaluation of profile available water capacity. 3. A model for estimating profile available water capacities for wheat on soils under irrigation, using simple physical and chemical soil properties. M. Hassan & R. Tavernier Identification legend for soil series developed in the Upper Pliocene deposits of Bangladesh. G. Wyseure Een . digitaal terreinmodel als hulpmiddel bij het specifiëren van het drainage gedrag van de bodem A digital terrain model as a technica I instrument for the specification of the drainage behaviour of soils.
5
17
33
45
59
M.M. Be llochio, M. Falciai & P. Trucchi Variabilities in the length of the growing period applied to Puglia, Italy.
75
New publications - Nouvelles publications - Nieuwe uitgaven
85
Druk Vita, 9750 Zingem 92
