Growth of legume cover crops
The variety of cassava did not significantly affect the growth of LCC. A significant difference in the growth of the LCC was caused by the different species of LCC. (Fig 1) indicates the differences growth of LCC under cassava. Area coverage by
P. javanica was highest and linearly increased along with cassava plant age. Area coverage by
C. pubescens was the lowest. A previous study also indicated that LCC genotypes had different responses to shading
(Mauro et al., 2013). The rate of area coverage by
C. pubescens and
C. mucunoides was steady-state at period 4-6 MAP. The area coverage by the mixed LCC that was higher than
C. pubescens and
C. mucunoides was contributed from
P. javanica. The maximum area coverage by
C. mucunoides was 77.3% (8 MAP);
C. pubescens was 39.3%,
P. javanica was 98.1% and the mixed LCC was 84.8% that occurs at 10 MAP.
P. javanica Benth reaches 60-70% cover after about 4 months and 90-100% after 8 months (Halim, 2016).
P. javanica produced the highest biomass from shoot pruning as a source of soil organic carbon in situ. The total fresh weight biomass of the shoot pruning (Table 1) of
P. javanica was the highest (11,01kg 20 m
-2 or 5.505 ton ha
-1), with dry matter yield of 1.140 ton ha
-1.
Samedani et al., (2013) reported that
P. javanica in open space produced green fodder up to 30-50 ton ha
-1 (dry matter yield of 4-10 ton ha
-1). Average light transmitted along the cassava plant age was 47.18% (Fig 2) or shading by 52.82% reduced the biomass dry weight from shoot pruning of
P. javanica at about 28.5% compared to the minimum dry matter yield (4 ton ha
-1) resulted in
Samadeni et al., (2013). Mauro et al., (2013) mentioned that shading by 50% reduced the maximum growth rate up to 21%.
C. pubescens is one of the shade-tolerant legumes which can persist under 80% shade (Teitzel and Peng 2016). The area coverage of this LCC showed a little increase at the period of 6-10 MAP (Fig 1), during which light transmission range from 10 to 40% (Fig 2) or under 60 to 90% shading condition.
Peng and Aminah (2016) reported that under low light intensities (< 20% of transmitted light)
C. mucunoides leaves are reduced in size by 70% compared with leaves in full sunlight.
C. mucunoides grows rapidly and can cover the soil in 3-6 months after sowing and even sooner on newly cleared, fertile land. Under cassava the land area coverage by
C. mucunoides was markedly increasing at 5-8 MAP (40-65%) when the light transmitted increase by 10-40% (Fig 2). After the period area coverage decreased due to the senescence of the leaves.
Shading of cassava
Cassava varieties had a different canopy growth. There was a quadratic response to the number of leaves to plant age. The number of attached leaves of UJ-5 was more than Mangu, especially at the end of growth (6-8 MAP). However, the number of fallen leaves of UJ-5 was less than Mangu at 6 MAP (Table 2). The quadratic response equation is shown in (Table 3).
The maximum number of attached leaves of UJ-5 was higher by 29.6% than Mangu. The different number of attached leaves was caused by different branching architecture. Mangu variety has a single branch and UJ-5 has a dichotomous branch
(Fukuda et al., 2010).
The leaf area was determined by the number of attached leaves (Table 2). As well as the number of attached leaves, LAI of UJ-5 was higher than Mangu. The LAI was dynamic along with the plant age. Fig 2 shows a quadratic respond of LAI (Y) to the plant age (X), with the equation Y = -0,2424X2 + 2,6032X - 3,3006 (R² = 0,7351) for Mangu and Y = -0,1445X
2 + 2,0719X - 2,8114 (R² = 0,8265) for UJ-5. The maximum LAI of Mangu and UJ-5 were 5.14 at 5 MAP and 5.70 at 6 MAP, respectively.
Light transmitted from cassava canopy to above the LCC at period 4 -6 MAP ranged from 10 to 30%. The low light transmitted was caused by the steady area coverage of
C. mucunoides and
C. pubescens (Fig 1).
Soil properties
Table 1 shows that the average percentage dry weight of
P. javanica and the mixed LCC was lower (20.6%) than
C. mucunoides and
C. pubescens (22.5%). This indicated that the shoot of
P. javanica had more water content. Due to the high area coverage (Fig 1) and shoot water content,
P. javanica could maintain the soil moisture content higher than other LCC (Table 4). The field capacity and wilting point moisture content levels of the soil were 39.2% and 22.2%, respectively. The soil covered by the
P. javanica had a moisture content of 90% field capacity; higher than those covered by other LCC.
The soil organic carbon before planting was 1.96%, below the threshold of generally 2.0% for sustaining good soil quality
(Musinguzi et al., 2013). After planting cassava with and without legume at 10 MAP, the soil organic carbon increased to 2.25-2.38% (Table 4). Soil bulk density ranged from 0.76 to 0.84 g cm
-3, which is considered a good bulk density for mineral soil
(Hossain et al., 2015). The improvement of the soil might be contributed by the fallen leaves. The dry weight fallen leaves of Mangu and UJ-5 were 2.85 and 2.23 ton ha
-1, which might be contributing to the increase of soil organic carbon by 0.14% and 0.11%. Such a contribution had been reported by
Suwarto and Abrori (2018).
Based on the harmoniously increasing growth along with the cassava growth,
P. javanica was considered tolerant to shading. This LCC produce more organic matter and could conserve soil moisture content better. Therefore,
P. javanica could be considered as a suitable legume cover crop under cassava stand to improve soil quality. Previous study indicated that
P. javanica provided vegetative cover to reduce soil and nutrient loss by erosion
(Baligar et al., 2020), and to reduce weed growth
(Mauro et al., 2015).
