Soil & Tillage Research 56 (2000) 1535

The soil and climate characterisation of benchmark sites for lowland rice-based cropping systems research in the Philippines and Indonesia
B.M. Schafera, G. Kirchhof b,*
b

School of Agriculture and Horticulture, The University of Queensland, Gatton, Qld 4343, Australia School of Land and Food Sciences, The University of Queensland, St Lucia, Brisbane, Qld 4072, Australia

Abstract Soil morphological, physical, chemical and mineralogical properties are described at ve locations in major rice (Oryza sativa L.) growing areas of the Philippines (two sites) and in Indonesia (three sites) which were selected for lowland rice-based cropping systems research. The data were used to classify the soils into the local soil series, soil taxonomy and The Australian Soil Classication systems. These data were intended to facilitate transfer of knowledge of improved farming systems technology to other lowland rice growing areas in the regions. The soils were classied as Andsisols, Inceptisols and Vertisols, and were characterised by clay contents ranging from 370 to 870 g kg1 and cation exchange values ranging between 17 and 68 cmol (p) kg1 for whole soil. pH values were neutral to mildly alkaline. Land surface and root zone attributes were qualitatively evaluated for limitations to post-rice crop production by interpretation of modied surface and sub-soil properties associated with rice production. Leakiness of bunds was also examined and mainly attributed to biological activity and for the development of drainage channels. Climatic data are presented for each of the ve sites and the characteristics for potential rainfall incidence are given for the post-rice dry season crop period. The soil sites selected have a range of properties which are deemed to represent large areas of soils used for rice production in these two countries. # 2000 Elsevier Science B.V. All rights reserved.
Keywords: Rice soils; Pedology; Soil properties; Climate

1. Introduction Soils used for lowland rice-based cropping systems in Indonesia and the Philippines are characterised by high surface clay content and the fact that they are puddled to aid water retention during the inundation period of the rice crop cycle. Following a rice crop, the potential to use sub-surface soil water for upland crop
* Corresponding author. Present address: NSW Agriculture, PMB 944, Tamworth NSW 2340, Australia. Tel.: 61-2-67-63-1147; fax: 61-2-67-63-1222. E-mail address: gk@mpx.com.au (G. Kirchhof).

production entails amelioration of the adverse effects of puddling on surface soils. For successful crop establishment the rainfall incidence following rice is critical. Consequently soilclimate constraints are a major consideration at the early stage of the dry season crop cycle and seed bed preparation together with timing of the establishment phase is critical (Rahmianna et al., 1996). Two sites in the Philippines and three sites in Indonesia were selected as benchmark sites for an international collaborative project to investigate soil management strategies to increase yields of dry season crops following rice. Components of these

0167-1987/00/$ see front matter # 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 1 9 8 7 ( 0 0 ) 0 0 1 2 0 - 3

16

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

management strategies include various soil tillage practices, application of mulch and amendments and agronomic practices. Comparison across the ve sites requires detailed characterisation of soil properties using standard description terminology and analytical methods for classication, according to soil taxonomy (Soil Survey Staff, 1994). Characterisation is also required to facilitate transfer of research outcomes to other rice growing regions in terms of the combined effects of soilclimate attributes and tillage treatments on potential crop establishment, growth and yield of dry season crops. This paper provides detailed descriptions of the soils and climate at the ve experimental sites. This should avoid the use of incomplete soil and climate descriptions in the following papers from the project. In the Philippines the soils of the Bulacan Province have been described, mapped and classied in detail (Soil Survey Division, 1987). Soil description and analytical characteristics were provided for soil near the site at Manaoag in the Pangasinan Province. However the data were not correlated with soils in the region and terminology used was insufcient for classication. The soils of East Java have been variously described and classied at the regional level but limited data were available for the selected experimental sites. In South Sulawesi, the soils of the Maros Agricultural Research Station have been described (Ali and Sawijo, 1982) although the experimental site was outside of the area surveyed. The climate in the Philippines is governed by northeast and southwest monsoon air streams. The northwest monsoon originates in the cold Asiatic winter anticyclone and produces a distinct dry season from around October to May. The southwest monsoon originates as an Indian Ocean anticyclone during the southern hemisphere winter. It usually commences in early May, reaches its maximum inuence in August and abates in October. Monsoonal rains can occur as early as April and May and persist until November. The Philippines is located in a region which is recognised as having the greatest frequency of tropical cyclones (typhoons) in the world (Flores and Balagot, 1969). They produce rainfall between May and December with a mean monthly frequency of greater

than 0.5 throughout the year, but generally less than 1.0 between January and May. The climate in Indonesia is dominated by monsoonal air streams which are at opposite times of the year to those in the Philippines. Rainfall in Java is largely affected by the position of the intertropical convergence zone which passes through twice annually. It is inuenced by the mountainous areas of Borneo and Sumatra (Sukanto, 1969). Although the wet and dry seasons are distinct, a moderate amount of rainfall occurs during the dry season which results in rainfall throughout the year. In southeast Sulawesi, a distinct dry season occurs but wet season rainfall is considerably higher than that of Java due to the inuence of the landmasses and mountains of Borneo. Compared to the Philippines, Indonesia has a very low incidence of tropical cyclones. 2. Methods Soil pits were hand dug to a minimum depth of 1.5 m to expose a vertical face of soil within a rice paddy and also to expose a vertical face of the associated bund. The proles were described using terminology proposed by McDonald et al. (1984) with minor modication by the use of consistence terms proposed by Soil Survey Staff (1951). This modication was made to facilitate communication with the professional workers in the two countries. The sites and prole exposures were photographed to provide a visual record. Soils were sampled for laboratory analysis by taking bulk samples from the designated horizons. These samples were analysed by the CSIRO Laboratories located in Canberra and Adelaide, Australia (Ringrose-Voase et al., 1996). Chemical methods follow the Australian Laboratory Handbook of Soil and Water Chemical Methods (Rayment and Higginson, 1992). Particle size analysis was determined by the sedigraph method (Hutka and Ashton, 1995) and mineralogy of the clay fraction was analysed semi-quantitatively by X-ray diffraction (Raven, 1995). The soils were classied according to soil taxonomy (Soil Survey Staff, 1994) and The Australian Soil Classication (Isbell, 1996). Soil survey reports and local information gained from professional workers associated with the program were used to identify soils

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

17

classied at the series level and to assist in the other classication systems. 3. Soil descriptions 3.1. Benchmark Site 1, San Ildefonso Location the Philippines, Luzon Island, Bulacan Province, San Ildefonso (see Fig. 1), Barangay Buenasita, Central Soil and Water Resources Research Station soil series: Mahipon series; soil taxonomy: Ustic Epiaquert and Australian Classification: Endocalcareous, Mottled, Epipedal and Aquic Vertosol gently sloping (38) to undulating relief; site component, slightly dissected lower piedmont footslope fringing a closed depression colluvium derived from sandstone, shale and limestone surface: well drained and internal: impeded and slowly permeable rice-based cropping

Classification

Topography

Parent material Drainage Land use

Prole morphology Ap1, (mixed, puddled) 013 cm, dark yellowish brown (10 YR 4/4), dark greyish brown (10 YR 4/2 moist), very dark grey (10 YR 4/1), dark brown (7.5 YR 4/4) (dominant) mixed, gravelly, fine sandy clay; moderate medium, 510 mm, angular blocky; rough ped fabric; dry extremely hard, moist firm, wet sticky and plastic. Common fine roots with rusty mottling on walls of very fine macropores (root channels). 2530%, 14 mm, sub-rounded, cemented ferromanganiferous nodules. Occasional 7 cm diameter sub-rounded pebbles of dolerite. Field pH 6.0. Clear discontinuous wavy with tongued pockets of gravelly fine sandy clay to: A12, 1327 cm, light brownish grey (10 YR 6/2 moist) with dark grey (10 YR 3/1), dark greyish brown (10 YR 4/2) many medium distinct mottles; gravelly fine sandy clay; moderate, 510 mm, angular blocky; rough ped fabric; dry extremely hard, moist firm, wet sticky

and plastic. Few very fine macropores. 510% soft to hard ferro-manganiferous nodules. Occasional subangular quartz and feldspar crystals. Field pH 6.0. Discontinuous wavy to: B21, (Mn) 2767 cm, light olive grey (5 Y 6/2 moist) with many, medium distinct brownish yellow (10 YR 6/6 moist) mottles; gravelly medium clay; strong, 20 50 mm, lenticular with intersecting slickensides parting to 1050 mm, angular blocky and 25 mm lenticular; smooth ped fabric; dry hard, moist firm, wet sticky and plastic. Occasional 25 mm diameter quartz gravels with translucent iron coatings. Common very fine pores and roots. Field pH 6.5. Gradual wavy to: B22, (Ca) 6797 cm, light olive grey (5 Y 6/2 moist) with many medium distinct brownish yellow (10 YR 6/6) mottles; medium clay; strong, 2050 mm lenticular with intersecting (30608) slickensides parting to 1050 mm angular blocky and 25 mm lenticular; smooth ped fabric; dry hard, moist firm, wet, very sticky and plastic. Field pH 8.0. 5%, sub-rounded, soft CaCO3 nodules. Occasional black manganese nodules. Gradual wavy to: B23, 97140 cm, light olive grey (5 Y 6/2 moist) medium clay; strong, 25 mm lenticular with slickensides on compound ped surfaces; smooth ped fabric; moist friable, wet sticky and very plastic. Occasional black subrounded manganese nodules. Field pH 8.0. Clear to: C, 140150 cm (continuing), weathered shale. 3.2. Benchmark Site 2, Manaoag Location the Philippines, Luzon Island, Pangasinan Province, Manaoag (see Fig. 2), Barangay Calmay and farmers field soil series: San Manuel silty clay loam; soil taxonomy: Typic Ustropept and Australian Classification: Haplic, Eutrophic, Grey and Dermosol backslope (<18), recent levee recent alluvium surface, subject to seasonal flooding; internal: moderately to well drained and moderately permeable cropping

Classification

Topography Parent material Drainage Land use

18

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

Fig. 1. Benchmark Site 1, San Ildefonso, Ustic Epiaquert.

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

19

Fig. 2. Benchmark Site 2, Manaoag, Typic Ustropept.

20

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

Prole morphology Ap1, 011 cm, very dark greyish brown (10 YR 3/ 2 moist) silty clay; coarse polyhedral with conchoidal faces (primary), breaking to compound prismatic (secondary) breaking to strong, 25 cm polyhedral (tertiary); rough ped fabric; dry extremely hard, moist very firm, wet sticky and plastic. Common fine roots. Field pH 6.5. Abrupt smooth to: Ap2, 1138 cm, dark greyish brown (10 YR 4/2 moist) silty clay; coarse polyhedral with conchoidal faces breaking to compound prismatic, further breaking to strong, 25 cm polyhedral; rough ped fabric; dry extremely hard, moist very firm, wet stick and plastic. Common fine roots. Field pH 6.5. Arbitrary clear smooth to: A13, 3869 cm, dark greyish brown (10 YR 4/2) silty medium clay; compound prismatic breaking to moderate fine 2 5 mm polyhedral with well developed organans, rough ped fabric; dry hard, moist friable, wet sticky and plastic. Occasional roots. Field pH 7.0. Arbitrary clear smooth to: A14, 69104 cm, dark greyish brown (10 YR 4/2 moist) medium clay, compound prismatic 25 cm breaking to medium 37 mm polyhedral with well developed organans; rough ped fabric; dry extremely hard, moist friable, wet sticky and plastic. Occasional roots. Field pH 7.5. Arbitrary clear to: AC1, 104117 cm, very dark greyish brown (10 YR 3/2) common medium distinct dark yellowish brown (10 YR 4/6) mottles; silty loam; compound prismatic breaking to medium, 37 mm polyhedral; rough ped fabric; dry hard, moist friable, wet sticky and plastic. Occasional roots. Field pH 7.5. Arbitrary to: AC2, 117160 cm (continuing), dark yellowish brown (10 YR 4/6) many medium distinct, very dark greyish brown, mottles (organic); silty clay loam; compound prismatic breaking to medium 37 mm, polyhedral; rough ped fabric; dry extremely hard, moist firm, wet sticky and plastic. Field pH 7.5. 3.3. Benchmark Site 3, Ngale Location Indonesia, East Java, Ngawi, Ngale Experiment Station of the Classification

Topography Parent material Drainage Land use

Research Institute for Legume and Tuber Crops (RILET) (see Fig. 3) and Malang (formerly MARIF) soil series: unnamed; soil taxonomy: Chromic Epiaquert and Australian Classification: Haplic, Epipedal and Aquic Vertosol almost flat (<18) floodplain recent alluvium derived from limestone and intermediate to basic volcanic ash surface: subject to seasonal flooding and internal: poorly drained, slowly permeable lowland rice

Prole morphology Ap1, (puddled) 020 cm, dark grey (2.5 Y 4/0 moist) many fine distinct rusty mottles; medium clay; massive breaking to coarse angular blocky further breaking to fine, 23 mm angular blocky; rough ped fabric; wet very sticky, very plastic. Many fine roots. Field pH 8.5. Abrupt smooth to: Ap2, (puddled) 2033 cm, dark grey (2.5 Y 4/0) many fine rusty mottles; medium clay; strong medium (57 mm) angular blocky; wet very sticky, very plastic. Common fine roots. Field pH 8.5. Clear smooth to: B21, 3372 cm, dark grey (2.5 Y 4/0 moist) common medium distinct rusty mottles in top 1015 cm; heavy clay (superplastic) strong coarse (714 cm) lenticular with strong intersecting slickenslides breaking to strong fine 25 mm lenticular, smooth ped fabric (few macropores); wet very sticky, very plastic. Common fine roots. Field pH 8.5. Gradual smooth to: B22, 72160 cm, dark grey (2.5 Y 4/0 moist) heavy clay (superplastic) strong coarse (1015 cm) lenticular breaking to fine 3 5 mm lenticular, well developed intersecting slickenslides; smooth ped fabric with few macropores; wet very sticky, very plastic. Common fine roots. Field pH 8.5. Gradual smooth to: B23, 160170 cm, dark grey (2.5 Y 4/0 moist) medium clay as above. Occasional sub-rounded CaCO3 nodules. Field pH 8.5.

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

21

Fig. 3. Benchmark Site 3, Ngale, Chromic Epiaquert.

22

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

3.4. Benchmark Site 4, Jambegede Location Indonesia, East Java, Jambegede (see Fig. 4), 20 km south of Malang at the Research Institute for Legume and Tuber Crops (RILET) and Malang (formerly MARIF) soil series: unnamed; soil taxonomy: Anthraquic Hapludand and Australian Classification: Mottled, Sodic, Eutrophic and Black, Dermosol levee backslope (1.58) weathered andesitic tuff and ash surface: seasonal flooding and internal: moderately well drained, moderately permeable rice-based cropping systems

Prole morphology Ap, (puddled) 018 cm, very dark grey (10 YR 3/1 moist) admixed with partly humified crop residue: dark yellowish brown (10 YR 4/6) mottling at admixture face; light clay; blocks 1015 cm cracking coincidentally with plant row distribution. Blocks break into 57 mm sub-angular with rusty brown cutans. Some evidence of platiness 25 cm width at base of puddling depth. Earthy fabric; moist friable, wet sticky and plastic. Common roots. Field pH 77.5. Clear smooth to: A12, 1845 cm, very dark grey (10 YR 3/1 moist) common medium faint yellowish brown (10 YR 5/8) mottles; light to medium clay; strong coarse, 2040 mm sub-angular blocky with incipient conchoidal faces on upper surfaces; rough ped fabric and macropores; moist friable to firm. Few reddish brown (2.5 YR 4/6) earthy scoria to 4 cm diameter. Field pH 88.5.

Classification

Topography Parent material Drainage Land use

Fig. 4. Benchmark Site 4, Jambegede, Anthraquic Hapludand.

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

23

Diffuse smooth to: A13, 4562 cm, very dark grey (10 YR 3/2 moist) common coarse distinct yellowish red mottle, medium clay; coarse angular blocky to strong polyhedral (57 mm), rough ped fabric and common macropores; moist friable to firm. Field pH 77.5. Few roots. Occasional (<5%) lumps 25 cm diameter soft reddish brown scoria. Gradual smooth to: AB, 6280 cm, dark brown (7.5 YR 3/2 moist) with many coarse distinct grey (5 Y 4/1) mottles on ped surfaces and dark brown (7.5 YR 3/4) coatings in root channels; lightmedium clay; strong, fine 25 mm angular blocky; rough to smooth ped fabric; moist firm. Field pH 77.5. Occasional roots. Gradual smooth to: B2, 80160 cm, dark brown (7.5 YR 3/2 moist) with common medium distinct mottles; light clay; strong, angular blocky (2 5 mm), smooth ped fabric; wet sticky and plastic. Field pH 7.5. Occasional roots. 3.5. Benchmark Site 5, Maros Location Indonesia, South Sulawesi, Maros Baru, Buloe and farmers field 10 km from the Research Institute for Maize and non-rice grain crops at Maros (see Fig. 5) (formerly MORIF) soil series: Bentomanaik (tentative); soil taxonomy: Aeric Tropaquept and Australian Classification: Ferric, Dermasolic, Redoxic and Hydrosol alluvial flood plain (<18) mixed alluvium derived from weathered basalt and karst seasonal inundation, water table to 5 cm surface. Rapid permeability

Classification

Topography Parent material Drainage

oxidation on exposure to many coarse distinct yellowish red mottles. Field pH 6.5. Abrupt smooth to: Ap3, (compacted) 1422 cm, greyish brown (10 YR 5/2 moist) clay loam; coarse sub-angular blocky breaking to medium (5 mm) angular blocky with some development of conchoidal faces on compound peds; rough to smooth ped fabric; wet sticky and plastic. Field pH 6.5. Abrupt smooth to: A14, 2241 cm, greyish brown (10 YR 5/2 moist) common distinct reddish brown mottle; clay loam medium sub-angular blocky breaking to fine (25 mm) sub-angular and angular blocky; rough ped fabric; wet sticky and plastic. Field pH 6.5. Clear smooth to: B2 FeMn, (plinthic) 4171 cm, greyish brown (2.5 Y 5/2 moist) many coarse distinct dark brown (7.5 4/4) mottles; light clay and silty clay lining the walls of 37 cm diameter root channels; medium (57 mm) moderate sub-angular blocky; wet sticky and plastic. Field pH 7.0. 20 30% soft ferro-manganiferous nodules. Gradual smooth to: B2 Fe, 71111 cm, greyish brown (2.5 Y 5/2 moist) common medium distinct yellowish brown (10 YR 5/6) and few medium distinct yellowish brown (10 YR 4/6) mottles and black segregations (10 YR 3/1); light-medium clay; smooth ped fabric; wet sticky and plastic. Field pH 7.0. Gradual smooth to: B22 Fe, 111150 cm, light brownish grey (2.5 Y 6/2) with common coarse distinct yellowish brown (10 YR 6/2) mottles and black (10 YR 2/1) manganese segregations; light clay; medium 57 mm, moderate sub-angular blocky; smooth ped fabric with large macropores occupying, 1040% of the surface area; wet sticky and plastic. Field pH 7.0. 4. Accessory eld soil prole and bund features Sealing of bunds during the rice cropping phase is regarded as a problem for water retention. Bunds were examined at some of the sites to provide a preliminary insight into the problem. Drainage channels created by faunal and oral activity during the cropping cycle appeared to be the main contributor to the problem. Concentrations of ferro-manganiferous nodules in surface horizons of the Ustic Epiaquert (Site 1) were

Prole morphology Ap1, (puddled) 012 cm, dark grey (5 Y 4/1 moist) (gleyed) with rusty oxidation around root mass; clay loam; puddled; rough ped fabric; wet sticky, plastic. Field pH 6.0. Abrupt smooth to: Ap2, (gleyed) 1214 cm, dark grey (5 Y 4/1 moist) organic clay loam; platey, rough ped fabric; wet sticky and plastic. Rapid

24

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

Fig. 5. Benchmark Site 5, Maros, Aeric Tropaquept.

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

25

probably due to wetting and drying cycles of the surface soil which inundates during wet seasons. Current cropping practices have probably accelerated the process of nodule formation. Layering evident in the Ap horizon was deemed to be due to puddling. Tongues of Ap horizon soil material in A12 horizon were interpreted as inll wash of surface material into vertical cracks following saturation and dispersion of soil material with low liquid limits. Churning of material in the upper part of the prole was evidenced by the presence of up to 50 cm diameter pockets of contrasting sandy textured soil material which indicated that pedoturbation had occurred during pediment development. The boundary of the pocket was coincidental with major slickenside surfaces which indicated considerable internal mixing of surface and sub-soil materials. Vertical cracks up to 1 cm wide and at mean intervals of 50 cm were evident to 150 cm depth. Soil properties of the bund were found to be similar to the Ap horizon. The position of the bunds is not permanently located and they are reformed on a seasonal basis. This is probably due to the high shrinkswell property of the clay which would cause bund failure with decreasing water content and concomitant soil shrinkage. The Typic Ustropept soil located at Manaoag (Site 2) was observed to have well developed macropores throughout the prole which appeared to be old root channels. Few, large, vertical cracks 1 cm wide to 1 m depth and 1 m spacing were observed and may be due to wetting and drying following compaction. Compaction is evident in Ap horizons by the presence of domed conchoidal faces on upper surfaces of compound peds. Well developed organans on surfaces of peds indicated organic matter distribution to 1 m depth in the prole. At Ngale (Site 3) the bund associated with the Chromic Epiaquert prole was similar to soil material described for Ap1 except that the fabric was classed as earthy due to the presence of well formed macropores with rust coloured coatings which appeared to be old root channels. In the B22 horizon below the associated bund, 11.5 cm diameter, preferred drainage channels were observed inlled with dark grey (5 Y 4/1 moist) clay which indicated reduced conditions. These channels appeared to have been formed by macrofauna (crabs). A water table at 160 cm coincided with

accumulation of calcium carbonate. Dominant grey colours suggested that anaerobic conditions persist at least seasonally. Roots were evident to 1.5 m and were identied as rice plant roots. Moisture content throughout did not appear to be uniform and where structure was more strongly developed the soil appeared to be better drained. Patches of rusty brown mottles in the upper B21 were associated with concentrations of roots in more strongly structured soil and associated with soil cloddiness. The Anthraquic Hapludand described at Jambegede (Site 4) contained common worm castes 3 mm in diameter which occupy up to 20% by volume at depth. Surfaces of vughs (25 mm diameter) are convoluted suggesting preferred drainage channels are developed in some root channels. pH increase in the A12 horizon together with conchoidal face development indicated restricted drainage in the puddled zone. The bund associated with the prole was described as a very dark greyish brown (10 YR 3/2) clay loam. Rat burrows were evident and inlled with plant and soil material. Below the bund, soil material was found to be compacted and similar to the Ap horizon. Macroporosity was well developed throughout the prole due to root and earthworm channels. At Maros (Site 5) the mammilated walls of vughs ranging in size from less than 1 to 7 mm diameter were coated with grey oriented silt. Macroporosity was well developed throughout B2 horizons. A brownish black manganese patina on compound ped surfaces in the Ap1 (puddled) indicated poor drainage. In the Ap3 (compacted) cutans on compound peds were described as having a smooth ped fabric indicating a drainage restriction. Rough ped fabric is evident on primary units. Rapid oxidation on exposure also indicated strong anaerobic conditions. 5. Soil prole analytical properties The soils used for experimentation have more than 400 g kg1 clay (<2 mm) throughout the prole which conrms the initial site selection of clay soils (Table 1). The Ustic Epiaquert (Site 1) and Anthraquic Hapludand (Site 4) proles have signicant sand in the surface horizons which is probably due to colluvial transport of surface soil material. The soil at San Ildefonso has the greatest change in particle size

26

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

Table 1 Particle size distributions using USDA clay, silt and sand fractions Depth (cm) g kg1 in texture class (mm) Clay (<2) Silt Fine (220) San Ildefonso 011 1130 30, Ap Ta 3070 70110 Manaoag 011 1138 3869 69104 104117 117160 Jambegede 018 2040 4060 6080 80160 Ngale 014 1434 3454 5478 78110 Maros 012 1214 1422 2241 4171 71110 110150
a

Sand Coarse (2050) 80 56 62 24 05 57 52 49 68 88 113 94 76 79 75 75 27 39 14 12 13 67 61 56 70 72 74 69 Total (250) 283 178 226 173 208 421 430 489 430 411 431 362 309 317 289 302 230 152 107 93 115 511 500 435 454 369 351 356 Very fine (50100) 74 45 64 46 09 23 23 21 42 58 65 79 60 41 38 45 17 13 08 13 15 22 23 24 34 50 52 74 Fine Medium Coarse (100250) (250500) (5001000) 113 78 110 59 04 19 11 12 27 28 33 86 63 41 38 35 11 10 05 06 06 08 07 13 14 41 50 64 57 107 83 65 09 06 01 01 01 02 01 18 11 06 05 05 02 01 02 02 02 01 01 01 02 05 04 03 29 56 56 52 05 06 00 00 01 01 00 03 01 01 00 00 01 01 00 01 01 00 02 01 01 04 04 02 Very coarse Total (10002000) (502000) 37 48 92 66 01 02 00 00 02 00 00 00 00 00 00 00 00 00 00 00 00 00 00 01 01 04 83 03 310 333 405 288 27 56 36 35 73 90 99 185 135 89 81 85 31 24 15 22 24 30 33 40 52 104 192 146

407 489 369 539 765 523 534 476 497 499 470 453 556 594 630 613 739 824 878 885 861 459 467 525 494 527 457 498

203 122 164 149 203 364 378 440 362 323 318 268 233 238 214 227 203 113 93 81 102 444 439 379 384 297 277 287

Tongues of Ap material at 30 cm depth.

distribution over depth with clay content increasing from 400 g kg1 at the surface to 750 g kg1 at depth. The clay content of the soil at Jambegede increases with depth from 450 g kg1 near the surface to 600 g kg1 at depth, with a corresponding decrease in sand content. The Chromic Epiaquert (Site 3) is a heavy clay soil with clay content increasing from 740 g kg1 in the

surface to 880 g kg1 at depth. The Typic Ustropept (Site 2) and Aeric Tropaquept (Site 5) have similar particle size distributions with 450500 g kg1 clay throughout. Sand content increases with depth from 30 g kg1 near the surface to 100 g kg1 at Site 2, and 200 g kg1 at Site 5. This characteristic of these two proles suggests that the soil material is of alluvial origin.

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535 Table 2 Chemical properties Depth (cm) Organic carbon (g kg1) pH EC (1:5 extract dS m1) Exchangeable cations Ca (cmol (p) kg1) 11.5 12.7 9.1 15.2 27.3 36.0 41.3 42.1 33.5 31.5 26.7 9.6 9.1 8.1 7.4 7.1 52.4 51.5 51.3 57.4 52.4 5.7 6.0 9.7 10.3 11.9 11.5 12.1

27

CEC Na (cmol (p) kg1) 0.58 0.87 0.76 1.05 1.29 0.42 0.40 0.34 0.31 0.25 0.35 0.27 0.52 0.80 0.76 0.77 0.32 0.35 0.39 0.44 0.42 0.20 0.27 0.25 0.33 0.49 0.47 0.43

Mg (cmol (p) kg1) 7.7 9.5 7.3 12.6 22.3 10.6 9.7 7.5 4.4 4.5 5.0 6.7 6.9 5.7 5.0 5.0 14.5 14.6 13.8 16.1 15.3 4.8 5.1 7.9 8.4 9.3 9.2 10.1

K (cmol (p) kg1) 0.22 0.11 0.00 0.00 0.00 1.02 1.16 0.70 1.85 1.26 0.38 1.74 1.48 2.71 2.94 4.00 0.89 0.55 0.31 0.19 0.31 1.56 0.86 0.79 0.54 1.01 1.20 1.67 21.4 23.4 17.4 28.2 49.1 43.6 52.8 49.2 40.7 42.3 33.3 18.6 17.2 17.6 16.6 16.5 69.5 68.7 62.6 74.2 68.0 15.6 14.9 17.6 19.9 22.2 22.9 24.9

San Ildefonso 011 1130 30, Ap Ta 3070 70110 Manaoag 011 1138 3869 69104 104117 117160 Jambegede 018 2040 4060 6080 80160 Ngale 014 1434 3454 5478 78110 Maros 012 1214 1422 2241 4171 71110 110150
a

11.65 4.14 4.29 2.56 1.28 11.30 9.83 8.29 6.38 4.77 4.84 11.67 7.84 4.67 4.23 4.04 14.32 11.47 7.48 6.46 5.64 16.32 15.06 6.88 6.15 2.67 2.30 1.97

6.54 7.35 7.42 7.99 7.80 7.71 7.61 7.55 7.64 7.76 7.77 7.60 7.23 7.40 7.56 6.98 7.05 7.20 7.21 7.01 7.64 4.88 5.06 6.25 6.63 6.75 6.90 6.97

0.090 0.062 0.059 0.069 0.140 0.112 0.104 0.120 0.101 0.094 0.092 0.097 0.178 0.079 0.073 0.070 0.103 0.111 0.108 0.064 0.060 0.106 0.068 0.059 0.054 0.066 0.062 0.057

Tongues of Ap material at 30 cm depth.

Chemical and mineralogical analyses (Tables 2 and 3) show that the soils are neutral pH with the proles at Site 1 and Site 5 being mildly acid in their surface horizons. Cation exchange capacities vary from 20 cmol (p) kg1 in the soil at Site 5 to nearly 70 cmol() kg1 in the soil at Site 3. The exchange complexes are dominated by calcium and magnesium and show no imbalance of cations which would limit plant growth. These values are supported by the clay species identication which show a predominance of kaolinite and smectite.

The soils at Sites 2 and 5 also contain signicant proportions of vermiculite and illite, respectively. The clay species in the prole at Site 3 was predominantly smectite which is reected in the high cation exchange capacity values throughout. In contrast, the clay fraction of the prole at Site 2 contained 63% smectite in the surface which decreased with depth and is replaced by kaolinite. This suggests that the prole is layered with more recent alluvial depositions having higher smectite clay contents.

28 Table 3 Mineralogy of the clay fraction Site depth (cm) San Ildefonso 011 1130 3070 70110 Manaoag 011 3869 104117 117160 Jambegede 018 4060 80160 Ngale 014 3454 78110 Maros 012 2241 71110 110150

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

Mineralogical composition (% of <2 mm fraction) Illite 163 143 184 153 Kaolinite Smectite 355 294 347 296 194 296 367 388 6513 8617 8717 265 92 <5 5511 479 418 469 6319 7021 6115 6015 6316 5112 4810 469 338 133 123 7318 9024 >95 205 267 308 307 Vermiculite <5 103 165 186 144 144 <5 <5 Quartz 11 11 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 Goethite Anatase <1 <1 Trace Trace Trace Feldspar <1 <1 <1 <1 <1 <1 <1 <1 Cristobalite Trace Trace Trace

Swelling clay (% of whole soil)a

268 3410 338 4611 338 246 245 224 154 82 72 5413 7921 84 92 133 144 153

51 81 81 81

<1 <1 <1 <1

a Proportion of swelling clay on a whole soil basis is the proportion of smectite in the clay fraction multiplied by the proportion of clay (Ringrose-Voase et al., 1996).

The smectite content of the soil prole at Site 1 is similar to that at Site 2 near the surface but remains constant with depth. However the increase in percentage of clay with depth results in the swelling clay component increasing from 26 to 46%. In contrast, the clay fraction in the surface soil at Site 4 contained 33% smectite which decreased to 12% at depth.The clay fraction in the soil at Site 5 contained 20% smectite near the surface which increased to 30% at depth. The relatively constant clay content in the soil results in the smectite content of the clay fraction only changing marginally from 9 to 15%. 6. Potential soil limitations for plant growth Soil properties limiting plant growth were interpreted from the qualitative descriptions obtained for

each site and prole (Table 4). Diagnostic pedological features were used to assess the potential of the soils for dry season crops which require soil depths greater than that provided by the puddled surface soil layers. The interpretations were based on the general edaphic conditions and do not account for the specic crop requirements. The term hard-setting is used to dene the soil surface condition created by puddling and drying.

7. Climate The characteristics of the climate at each site are presented in Figs. 610. Limited data prevented detailed analysis in relation to crop risk analysis based on soilclimatic probability relationships and

Table 4 Qualitative descriptions of each site and soil prole Soils Land surface limitations Seasonal flooding Philippines Ustic Epiaquert (San Ildefonso) Typic Ustropept (Manaoag) Indonesia Chromic Epiaquert (Ngale) Anthraquic Hapludand (Jambegede) Aeric Tropaquept (Maros) None Slope Surface condition Hard-setting Hard-setting Limitations in root zone Internal drainage Poor, strongly mottled; Fe, Mn concentrations Moderately well drained, weak mottling Poorly drained (gleyed) Moderately well drained, weak mottling Rapidly drained; Fe, Mn concentrations Permeability Effective depth (cm) Slow to very slow Moderate 100150 100150 Depth (cm) DSWTa >50 >100 AWC estimate Drainage class

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

128 <18

Moderate Moderate

Poorly drained Well drained

Seasonal None Seasonal

<18 1.58 <18

Self-mulching Hard-setting Hard-setting

Slow to very slow Moderate Moderate

100150 100150 100150

<50 >100 >20

High Moderate Low-moderate

Poorly drained Well drained Poor in wet season, well drained in dry season

DSWT: dry season water table.

29

30

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

Fig. 6. Medium term average daily rainfall and probability for rainfall (1988 to 1995) at San Ildefonso.

only average values could be obtained to indicate potential rainfall during the post-rice period. Future work to establish probabilities for dry season crop establishment success rate would be a benet to extrapolate research ndings of the project to other regions. The probabilities of daily rainfall exceeding different values are calculated from the available long term average daily rainfall events.

The average annual rainfall at San Ildefonso Vista (Site 1) is 1986 mm. Long term daily average rainfall and probability is given in Fig. 6. The wet season lasts from around mid-May through November with highly variable daily rainfall events. Post-rice crops are usually sown between early December and late February which is at the end of the wet season. In contrast to the abrupt change from the

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

31

Fig. 7. Recent average daily rainfall (1992 to 1994) at Manaoag, Pangasinan and probability for rainfall combined from 3 sites near Manaoag over 9 years.

dry to wet season, the change from wet to dry is relatively gradual which is important for dry season crop establishment. Although the probability for rainfall remains above 10% throughout the dry

season, the chances of crop damage from high intensity storms are minimal. Occasional typhoon activity during this otherwise dry period may occur in some years.

32

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

Fig. 8. Long term average daily rainfall and probability for rainfall (1981 to 1995) at Ngale.

Average annual rainfall recorded at Manaoag (Site 2) is 1486 mm. The dry season extends from about mid-October through May (Fig. 7). Compared to the San Ildefonso site, the transition from wet to dry season is more abrupt which can be explained by the lower probability of rainfall throughout the year. The wet season tends to commence with relatively large erratic rainfall events commencing in March

which become more regular in occurrence by May with the incidence of moderate rain (1050 mm per day). These rainfall events make the Manaoag area more suitable for post-rice cropping. Average annual rainfall at Ngale (Site 3) is 2179 mm. The wet season occurs during the period between October and April and is followed by a moderately wet dry season (Fig. 8). The transition

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

33

Fig. 9. Long term average daily rainfall and probability for rainfall (1983 to 1995) at Jambegede.

from wet to dry season is gradual with a high probability of effective rainfall for post-rice plant establishment and growth. Compared to the sites in the Philippines, post-rice crops are established earlier and at the end of the wet season. This is made possible due to the lower frequency of tropical cyclones and more reliable rainfall events. Average annual rainfall at Jambegede (Site 4) is 2202 mm. The characteristics

of rainfall distribution (Fig. 9) are regarded as being comparable to those at Ngale. Average annual rainfall at Maros (Site 5) is 3085 mm. Compared to East Java, the wet season is more intense with probabilities for daily rainfall events approaching 90% (Fig. 10). The transition from wet to dry season is abrupt although rainfall peaks in January and declines to a minimum in July. The

34

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535

Fig. 10. Long term average daily rainfall and probability for rainfall (1976 to 1993) at Balitjas Maros.

transition from wet to dry season occurs during the period from April to May and is comparable to the sites in East Java. Acknowledgements This work was funded by the Australian Centre for International Agricultural Research.

References
Ali, M.A., Sawijo, 1982. Detailed Soil Survey of the Maros Agricultural Research Station South Sulawesi. Centre for Soil Research, Ujung Pandangan, Ministry of Agriculture, Government of Indonesia UNDP and FAO. Flores, J.J., Balagot, V.F., 1969. Climate of the Philippines. In: Arakawa, H. (Ed.), Climates of Northern and Eastern Asia. Elsevier, Amsterdam, pp. 59204. Hutka, J., Ashton, U.C., 1995. Sedigraph particle size analysis of

B.M. Schafer, G. Kirchhof / Soil & Tillage Research 56 (2000) 1535 soil samples from the Griggward study area of the Wagga Wagga project. CSIRO (Australia) Division of Soils Technical Report No. 3/1995. Isbell, R.F., 1996. The Australian Soil Classication. CSIRO Publications, Melbourne. McDonald, R.C., Isbell, R.F., Speight, J.G., Walker, J., Hopkins, M.S., 1984. Australian Soil and Land Survey Handbook. Inkata Press, Melbourne. Rahmianna, A.A., So, H.B., Kirchhof, G., Sumarno, Adisarwanto, T., 1996. Crop establishment of legumes in lowland rice-based cropping systems. In: Management of Clay Soils for Rainfed Lowland Rice-based Cropping Systems. ACIAR Proceedings No. 70. Raven, M.D., 1995. Mineralogy of clay fractions from Indonesia and the Philippines. CSIRO (Australia) Division of Soils Mineralogical and Geochemical Services Report No. 187. Rayment, G.E., Higginson, F.R., 1992. Australian Laboratory Handbook of Soil and Water Chemical Methods. Australian

35

Soil and Land Survey Handbooks, Vol. 3. Inkata Press, Melbourne, 330 pp. Ringrose-Voase, A.J., Hutka, J., Clarke, M., Raven, M.D., Schafer, B.M., 1996. Analyses of soil proles at the experimental sites in Indonesia and the Philippines. In: Management of Clay Soils for Lowland, Rice-based Cropping Systems, ACIAR Project 8938. CSIRO (Australia) Division of Soils Technical Report No. 2/1996. Soil Survey Division, 1987. Soil Survey and Classication of the Bulacan Province Bureau of Soils and Water Management. Department of Agriculture and Food, Manila, p. 384. Soil Survey Staff, 1994. Keys to Soil Taxonomy. USDA Agricultural Handbook, No. 436. Government Printer, Washington, DC. Soil Survey Staff, 1951. Soil Survey Manual. USDA Agricultural Handbook, No. 18. Government Printer, Washington, DC. Sukanto, M., 1969. Climate of Indonesia. In: Arakawa, H. (Ed.), Climates of Northern and Eastern Asia. Elsevier, Amsterdam, pp. 215229.