Abstract

The nutrient release potential of selected types of organic materials available in a coffee farming system was studied with three contrasting soils of Kilimanjaro Region, to broaden the ISFM options. An incubation experiment was conducted at TaCRI Lyamungu screenhouse between June and November 2019. The design was split plot RCD, three soil types Humi-Umbric Nitisols from Lyamungo, Eutric Cambisols from Kilacha and Humi-Rhodic Luvisols from Kikafu Chini as main factors, and different organic additives (cattle manure and leaves of Tithonia diversifolia, Tephrosia vogelii, Vernonia subligera and Adhatoda engleriana) as sub-factors. The organics were dried, crushed, sieved in a 6 mm sieve, mixed with the soils at 2% organic to soil ratio, moistened to field capacity and incubated in 10 litre plastic containers at room temperature. Duplicate soil samples were taken at day 0, 3, 8, 15, 26, 45, 74, 112 and 180 and analyzed for NH₄-N, NO₃-N, and available P. The total amounts of nutrients released during the incubation period were subjected to Analysis of Variance using COSTAT Statistical Software, with means separated by Tukey’s HSD at p ≤ 0.05. The nutrient release trends were more or less the same in all the three soil types. In terms of NH₄-N, NO₃-N and P, soil types, additives and their interactions were very highly significant (p < 0.001). In all the soil types, the four organics are fairly comparable to manure and can be used as its substitute. It was noted that soils differ in their responsiveness to organic treatments. Also unveiled was the potential of the wild plants such as Adhatoda, available in the wilderness around Mt. Kilimanjaro and Usambaras, which could be domesticated and planted at hedgerows for ISFM purposes. It also encourages the use of the semi-domesticated hedgerow plants like Tithonia and Vernonia; and Tephrosia as temporary shade plants.

1. Introduction

Coffee is one of the most traded commodities in the world. Its importance in the Tanzanian economy is well documented by ^{1, 2, 3}, among others. Of the challenges facing the coffee industry for the time being, soil fertility degradation ⁴ was identified by the stakeholders as one of the most limiting factors. Tanzania Coffee Research Institute (TaCRI) aims at promoting integrated soil fertility management (ISFM), which includes use of organic materials in the coffee ecosystems, as complementary to industrial fertilizers, for improved and sustainable productivity. Gumbo ⁵ and Raab ⁶ describe ISFM as the key in raising productivity levels in agricultural systems while maintaining the natural resource base. It aims at replenishing soil nutrient pools, maximizing on-farm recycling of nutrients, reducing nutrient losses to the environment and improving the use efficiency of external inputs.

A number of efforts have been made in other countries to develop coffee ISFM by making use of organic residues around a coffee farm. Examples are ⁷ in India and ⁸ in Zimbabwe. In Tanzania, however, there has not been a clear ISFM strategy in the coffee areas ⁹. The contribution of organic components of the coffee ecosystem has not been thoroughly studied. As a result, farmers apply such materials haphazardly, while others even destroy them by burning ¹⁰. Maro and others ¹¹ initiated such an effort, noting high potential of four green manure plants (Mucuna, Lupine, Canavalia and Crotalaria) in NPK nutrient release. During the study, some wild plant species were identified by farmers as potential for nutrient release. Some of them have been studied to a considerable extent; like Tithonia diversifolia ^{12, 13, 14, 15} and Tephrosia vogelii ¹⁶. There is limited literature about the other two, Adhatoda engleriana Lindau and Vernonia subligera, except that the former is known as a treatment for epilepsy ¹⁷; and the latter known to enhance P availability in soils ¹⁸. This follow-up study aimed to assess the nutrient (N and P) release potential of the four wild plants as they compare among themselves and against cattle manure in order to enrich the farmers’ ISFM options. It also aimed to check if such potential differs significantly among different soils.

2. Materials and Methods

Soils were obtained from Lyamungo Primary Cooperative Farm, Hai district (-3^o2049; 37^o2438; 1280 metres above sea level - masl), representing Humi-Umbric Nitisols, Mwika South – Kilacha, Moshi District (-3^o3746; 37^o5662; 895 masl), representing Eutric Cambisols, and Kikafu Chini - Longoi, Hai district (-3^o4437; 37^o2914; 791 masl), representing Humi-Rhodic Luvisols. The soil classification is according to the SOTER database for Tanzania, described in ¹⁹. In each site, a pit 1.5m x 1.5m was dug down 50cm and the experimental sample taken as a vertical slice representing the 50-cm profile. Enough soil was transported to the TaCRI Screenhouse, spread on canvas to dry for 2 days with coarse pebbles removed, then stored for the experiment.

Fresh cattle manure was taken from an indoor kraal belonging to one TaCRI staff. It was dried in a well ventilated drying oven at 40°C for 48 hours, then ground, sieved at 6 mm mesh and stored ²⁰. Leaves from the experimental plants – Wild sunflower (Tithonia diversifolia), Fishbean (Tephrosia vogelii), Wild Tobacco (Adhatoda engleriana) and Tugutu (Vernonia subligera) were collected, spread in the open to dry for about 1 week in raised coffee drying beds, then ground in a tissue grinder and sieved through 6 mm mesh.

The test materials were mixed with the soils at 2% organics to soil ratio, moistened to field capacity (FC) and incubated in 10 litre plastic containers arranged in split-plot RCD (5 treatments and 3 replications; on three tables each representing a different soil type) in the screenhouse at room temperature (24°C ± 2) ²¹. Moisture level was maintained around FC by covering with poly-sheet during the day and uncovering at night ²²; together with spraying twice a week with a hand sprayer.

Duplicate soil samples were taken with a soil scoop at day 0, 3, 8, 15, 26, 45, 74, 112 and 180. Fresh soils were used for the determination of mineral nitrogen as suggested by ^{23, 24}. 20g of moist soils in 200 mL of 2M KCl solution was shaken for 40 minutes and filtered through Whatman filter paper no 42. NH₄⁺-N and NO₃-N from soil extracts were measured by steam distillation procedure using MgO and Devarda’s alloy. Available phosphorus was determined by using the same samples, but after the routine drying, grinding and sieving. It was analyzed by using the Bray 1 method ^{20, 25}.

Nutrient release trends were descriptively assessed. Total NH₄-N, NO₃-N and P were calculated and values for Day 0 subtracted from the totals to get the nutrients released only during the time of the experiment. These were exposed to ANOVA under COSTAT Software according to the split plot model suggested by ²⁶ as follows:

(1)

Where µ stand for the general mean, α_i the effect of the main factor (soils), P_k the block effect, D_ik the main factor random error, β_j the effect of the sub-factor (organic additives), (αβ)_ij the interaction effect between main factor and sub-factor, and ε_ijk the sub-factor random error. Then means were separated using the Tukay’s HSD method at p ≤ 0.05. Finally, the normalized difference between the means for each of the test organics and the means from cattle manure (the standard check) was computed using the formula suggested by ²⁷:

(2)

whereby ND represents normalized difference; R_o is the mean release from the test entries, and R_m is the mean release from manure. This way, positive and negative variations could be visualized.

3. Results and Discussion

Nutrient release trends are summarized in Table 1 for NH₄-N, NO₃-N and available P respectively. Peak release was generally between Day 3 and Day 45, accounting for 68-72% of the total NH₄-N released for the Luvisol of Kikafu; 66-74% for the Nitisol of Lyamungo; and 64-67% for the Cambisol of Kilacha. With NO₃-N the same time interval accounted 56-67%, 60-74% and 61-68% of the total NO₃-N released respectively, while with P, the interval accounted for 61-68%, 67-79% and 59-65% of the total P released respectively. This trend is partly in line with ¹¹ who noted peak nutrient release between 8 and 45 days. Amounts initially present (Day 0) were 9-15% of the total NH₄-N for Kikafu and Lyamungo; and 11-14% for Kilacha. NO₃-N was 14-20%, 10-20% and 13-20% respectively, while that of P was 11-13%, 7-14% and 8-11% respectively.

Soils, additives and their interactions were very highly significant (p<0.001) but replications were not. Lack of significance with replications in a controlled experiment such as this is not abnormal because the set conditions are the same. This was also noted by ²⁸ in their work on somatic embryogenesis. The model was also very highly significant with R² = 0.897 and CV = 4.08%. Mean rankings for soils (Figure 2) followed a decreasing order Cambisol > Nitisol > Luvisol; while the rankings for additives (Figure 3) followed the order Tephrosia > Tithonia > Adhatoda > Vernonia > manure.

As with the NH₄-N, soils, additives and their interactions were very highly significant (p<0.001) but replications were not. The model was also very highly significant with R² = 0.983 and CV = 3.01%. Mean rankings for soils (Figure 2) followed a decreasing order Nitisol > Luvisol > Cambisol; while the rankings for additives (Figure 3) followed the order manure > Adhatoda > Vernonia > Tithonia > Tephrosia.

Even in this case, soils, additives and their interactions were very highly significant (p<0.001) but replications were not. The model was also very highly significant with R² = 0.998 and CV = 2.86%. Mean rankings for soils (Figure 2) followed a decreasing order Luvisol > Cambisol >>> Nitisol; while the rankings for additives (Figure 3) followed the order Tephrosia > Adhatoda > Vernonia > Tithonia > manure. Kaloi and others ²⁹ noted that P release increased with less clay and vice-versa. And the fact that Nitisols are known to have more clay than the other soils can partially explain its lower ranking.

The relative difference in nutrient release, between the tested organics and the manure check is given in Figure 4. There were positive variations in NH₄-N release from Tephrosia, Tithonia and Adhatoda. Vernonia did not record any variation. Positive variation was also noted in P release from all entries, with Tephrosia recording the widest difference of about 0.3, followed by Adhatoda (about 0.16). NO₃-N had negative variation in all cases, with Tephrosia and Tithonia recording a variation of -0.16 and -0.14 respectively. None of the variations reached 0.5 in any direction, implying that the organics are fairly comparable to manure and can be used as its substitute. Compared to ¹¹, the organics used in this study fall into the lower category (which also included coffee leaves, pulp and husks).

Organic matter decomposition (and therefore, nutrient release) is controlled by many factors, including the quality of the litter itself and the conditions within the soil environment ^{30, 31}. The influence of litter quality is emphasized by ³² who noted that plant (species) diversity influences many fundamental ecosystem functions including nutrient dynamics during litter breakdown. As such, the very highly significant difference among the tested species (Figure 3) is justified.

Other influential factors, as noted by ³³ on decomposition of coarse wood, are location, soil texture (as also noted by ²⁹), clay mineralogy and temperature. Soils in this work differed in altitude of origin, with Lyamungo Nitisols substantially higher (1280 masl) than the Kilacha Cambisols (895 masl) and Kikafu Luvisol (791 masl). The former is in a denudation zone while the rest are typically in a deposition zone. In terms of soil texture, ¹⁹ described Nitisols as clayey soils with a nitic subsurface horizon of high aggregate stability. Cambisols are medium to fine textured soils without much illuviated clay or organic matter, while Luvisols have illuviated clays resulting from ages of deposition of materials from elsewhere. In this case however, no particular trend was noted that can be directly linked to soil texture or clay mineralogy. Temperature at origin is obviously lowest at Lyamungo, increasing with decreasing altitude. On the other hand, as opposed to field-based mesocosms ³³ and litterbags ³⁰, this work was a more controlled experiment in a screenhouse where room temperature (24°C ± 2) was maintained throughout.

Kwabiah and others ¹² worked on Tithonia diversifolia, along with five similar species Sesbania sesbani, Croton megalocarpus, Calliandra calothyrsus, Lantana camara and Senna spectabilis; all branded as agroforestry plant species. In that study, Tithonia topped the list in N and P release. As regards Vernonia subligera, ¹⁸ noted its beneficial effects on beans when applied in combination with Minjingu rock phosphate or triple superphosphate. Both species have been domesticated as hedgerow plants. Adhatoda engleriana is a wilder species found in evergreen Tropical forests, but can be domesticated as hedgerow plant too. Only Tephrosia vogelii can be interplanted with coffee as temporary shade during young age (0-3 years). It compares fairly well with the best-bet temporary shade plant Sesbania ³⁴, and was also one of the recommended temporary shade plants in the Climate-Smart Toolbox ³⁵. Its difference from the other three is that it does not regrow when cut, therefore not very suitable for hedgerows.

Of the three soil types involved in this study, only the Nitisol falls within the coffee belt. The other two are outside the coffee belt, but there are such soils growing coffee elsewhere. For Nitisols, Tithonia is the best bet hedgerow plant. The hedgerows can be periodically clipped and the clippings spread into the fields as mulch. We would not recommend immediate domestication of Adhatoda for hedgerows yet – we need to explore it further, in terms of compatibility with coffee. In the meantime, for those near the forests they may clip it right there and carry the clippings to their fields to spread as mulch. The hedgerow plant of choice in Cambisols (Bukoba, Muleba, parts of Kibondo, Uvinza and Mpanda) and Luvisols (parts of Longido, Mwanga, Same and Lushoto) could be Vernonia. The best in all soils was Tephrosia, which we can recommend as temporary shade for young coffee in its first 2-3 years.

4. Conclusion

In this work we explored the nitrogen and phosphorus release potential of different organic additives (cattle manure and leaves of Tithonia diversifolia, Tephrosia vogelii, Vernonia subligera and Adhatoda engleriana) when applied to three soil types Humi-Umbric Nitisols from Lyamungo, Eutric Cambisols from Kilacha and Humi-Rhodic Luvisols from Kikafu Chini, in order to broaden the ISFM options for coffee in Tanzania. The nutrient release trends were more or less the same in all the three soil types. Peak nutrient release was between Day 3 and Day 45, whereby over 50% of the available N and P were released in all cases. In terms of NH₄-N, NO₃-N and P, soil types, additives and their interactions were very highly significant. Cambisol was the best soil in NH₄-N release while Tephrosia was the best additive. Nitisol was the best soil in NO₃-N release while manure was the best additive. Luvisol was the best soil in P release while Tephrosia again was the best additive.

Comparing the relative difference in nutrient release between each entry and the manure check, there were positive variations in NH₄-N and P; and negative variations in NO₃-N. In any case, the variation did not exceed 0.5 in absolute terms. The implication here is that, in all the three soil types, all the four organics are fairly comparable to manure and can be used as its substitute. This study has shown that soils differ in their responsiveness to organic treatments. It has also shown the potential of the wild plants such as Adhatoda, available mostly in the wilderness around Mt. Kilimanjaro and the Usambaras, which could be domesticated and planted at hedgerows for ISFM purposes. It also encourages the use of the semi-domesticated hedgerow plants like Tithonia and Vernonia; and Tephrosia as temporary shade plants.

Acknowledgements

The authors wish to acknowledge the Government of Tanzania, the European Commission and the coffee stakeholders for financially sponsoring this work. Support staff at TaCRI Headquarters (the GAPs Research Programme) are thanked for their moral encouragement and facilitation in data collection.

Statement of Competing Interests

The authors wish to declare no competing interests in this work.

Abbreviations

Acronym Long form/description

ANOVA Analysis of variance

CV Coefficient of variability

FC Field capacity

GAPs Good agricultural practices

ND Normalized difference

NSS National Soil Services centre

SOTER Soils and terrain database of the world

TaCRI Tanzania Coffee Research Institute

References