African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 7, Num. 4, 1999, pp. 433-439

African Crop Science Journal, Vol. 7. No. 4,  pp. 433-439, 1999                                    

BIOPRODUCTIVITY AND DECOMPOSITION OF WATERHYACINTH IN UGANDA

A. Amoding, R. Muzira, M.A. Bekunda¹ and  P.L. Woomer²
Department of Soil Science, Makerere University,  P.O. Box 7062, Kampala, Uganda
¹Makerere University Agricultural Research Institute, P.O. Box 7062, Kampala, Uganda
²Department of Soil Science, University of Nairobi, P. O. Box 30917, Nairobi, Kenya

Code Number: CS99034

ABSTRACT

The biological invasion of the waterhyacinth (Eichhornia crassipes (Mart.) Solms) into lakes and rivers of East Africa forced the implementation of mechanical harvesting around key harbours and dams, resulting in subsequent difficulties of waste disposal. Utilising these wastes assists in minising the costs of waste management. Estimates of waterhyacinth biomass were made by randomly deploying 1 m² buoyant sampling frames across four sites on the Ugandan shores of Lake Victoria and determining the weight within the frame.  Single plants ranging from 0.3 - 0.9 kg were placed within similar buoyant frames and their productivity monitored over a period of 16 weeks in a sheltered bay at Bugiri and a pond at Kajjansi. Nutrient contents and mineralisation patterns of harvested waterhyacinth wastes were characterised.  Whole chopped plants and tissues were separated into leaves, petioles and roots, placed into litter bags, deployed as surface mulch and recovered over 16 weeks. Fresh biomass at the four sites was between 300 to  610 t ha^-1. Productivity ranged from 58  to  228 t ha^-1 yr^-1 resulting from rapid production of daughter plants (108 to 237 plants m^-2 yr^-1).   Decomposition of the waterhyacinth was rapid but with significant differences between plant tissues. Time to 50% decomposition of whole plants, leaves and roots was21, 31 and 45 days, respectively. Waterhyacinth applied as surface mulch to fields  may offer opportunity as an organic input to soils because of the relatively rich nutrient contents and rapid decay pattern but the large bulk of fresh plants (92% water) may offset these advantages.

Key  Words:  Aquatic systems, decomposition, Eichhornia crassipes, productivity, surface mulch

RÉSUMÉ

L’invasion de l’eau par la jacinthe (Eichhornia crassipes (Mart.) Solms) dans les lacs et rivières de l’Afrique de l’Est a forcé l’application des  recoltes mécaniques au tour d’entrées importantes et de digues résultant aux difficultés conséquentes des ordures.  L’utilisation de ces ordures aide à differer les coéts de contrôle.  Les estimations de la biomasse vivante ont été faite aléatoirement par des cadres flottables d’1m² d’échantillonnage dans 4 sites sur les rives du lac Victoria en Uganda et le poids des  mauvaises herbes a été déterminé à l’interieur du cadre.  Des plantes individuelles d’environ 0.3-0.9 kg ont été placées à l’intérieur de cadres semblables flottables et leur productivité suivie pour une période de plus de 16 semaines sous l’abri de baie à Bugiri et dans un êtang artificiel à Kajjansi. Le contenu  nutritif et la tendance de minéralisation des déchets de la jacinthe recoltéc ont été caracterisés, toutes les plantes découpées et tissus ont été séparés en feuilles, pétioles et racines; placées dans des sacs de détritus à la surface comme fumier et récuperé après 16 semaines.  La biomasse frèche dans les 4 sites a été estimée être de l’ordre de  390 - 610 t ha^-1. La productivité variait de 58 - 228 t ha^-1 yr^-1 venant de la reproduction de  108 - 237 plantes ha^-1 yr^-1.  Ces données suggèrent que la performance était associeée aux differentes qualités de systèmes aquatiques.  La décomposition de la jacinthe était rapide mais avec de differences significatives entre les tissus des plantes.  Le temps de décomposition de 50% de toutes les plantes, feuilles et racines était de21, 31 et 45 jours respectivement. La jacinthe utilisée comme paille de couverture en champs est avantageuse comme une resource organique aux sols à cause du contenu nutritif relativement riche et de sa décomposition rapide.

Mots Clés:  Systèmes aquatique, décomposition, Eichhornia crassipes, productivité, paille de surface

Introduction

Waterhyacinth (Eichhornia crassipes (Mart.) Solms) has invaded many rivers and lakes of East Africa, posing a serious obstacle to the exploitation of these water resources (Twongo et al., 1995; Woomer et al., 1998). Control of the weed in Uganda is focused on two approaches, release of biological control agents and mechanical clearance.  Waterhyacinth is reported to be one of the most productive plants (Gopal, 1987), but this needs to be determined for the waters of Lake Victoria to serve as a baseline for monitoring its control. Woomer (1997) estimated that the waterhyacinth accumulating at the Owen Falls hydroelectric facility at Jinja reaches approxi-mately 900 Mg dry weight (1.8 x 107 kg fresh weight at 95% moisture),from an area of  33 ha, the rate of accumulation depending upon river currents, wind direction, bioproductivity and the rate of mechanical removal. In order to verify these estimates, experiments were conducted to determine bioproductivity of the waterhyacinth in different locations of  Lake Victoria.

Disposing mechanically harvested water-hyacinth poses a problem where land is limited. These wastes are transported as far as 6 km away from the Owen Falls hydroelectric facility near Jinja. Woomer (1997) observed that the waste piles decompose in a complex manner with some remaining more or less intact until covered by weed vegetation or flattened by bulldozers while others degrade uniformly into compost. Concern over delayed decomposition and difficulties in waterhyacinth waste disposal have been raised elsewhere (Solly, 1984; Saustroutomo, 1991). 

Waterhyacinth may be under-utilised as organic inputs to soil.  However, the economic use of these wastes with regard to soil fertility depends upon the nutrient concentration and release rates through decomposition and the costs of transportation and application. Experiments were conducted to examine chemical properties and decay of waterhyacinth under surface mulched conditions using materials collected during aquatic weed control operations in Uganda.

Materials and Methods

Water sampling. Water randomly sampled within  four floating frames was taken to the laboratory and analysed immediately for   pH, NO₃^-- N, NH₄⁺- N, P, Na, K, Ca and Mg (Heffernan, 1985).

Standing biomass.  A 1 m² wooden frame was randomly cast from a small boat and the waterhyacinth and other aquatic weeds within the frames collected and weighed. This was done at Kyondo Bay, Kirinya Bay, Owen Falls dam and Bugiri  on Lake Victoria.  Three replicates were recovered per site.

Productivity studies. Bioproductivity was monitored at two sites; in a man-made pond at Kajjansi and in a sheltered bay at Bugiri  on Lake Victoria. Single waterhyacinth plants were cleaned by removing the dead parts and washing off the sediments, weighed and placed in 1 x 1 x 0.5 m open ended box frames made of wood and galvanised iron sheets and attached to plastic floats. The 0.5 m depth was intended to prevent crossover of daughter plants from adjacent frames. Sixteen  frames were installed at each site allowing  weekly monitoring of daughter plant production and weight gain over a period of 16 weeks. The plants were removed, the number of shoots counted, fresh weights taken and the plants returned to the frames. Destructive sampling of four frames was done every four weeks and a sample of each was taken to the laboratory for further analysis. The experiments were arranged in a completely randomised design at both sites.

Decomposition study.  The decomposition and nutrient release patterns of fresh waterhyacinth plants and plant parts applied as a surface mulch was characterised using the litter bag technique (Anderson and Ingram, 1993) at the Makerere University Agricultural Research Institute  Kabanyolo (MUARIK).  Water hyacinth plants were collected from Owen Falls dam near Jinja, transported to Kabanyolo, washed to remove lake sediments and separated into plant parts; leaves, stems (consisting of petioles, stolons and corms) and roots.  A sub-sample was collected for moisture determination and nutrient analysis.  Samples were placed into a 60º C drying oven until a constant weight was established (4 days for stems), weighed, ground to <2 mm and analysed for N, P, K, Ca and Mg (Okalebo et al., 1993).  One kg of fresh waterhyacinth plants and plant parts were placed into 30 x 25 cm plastic mesh (5 mm) litter bags and deployed as a randomised complete block design with two replicates.  The soil at the site is a Kandiudalfic Eutrudox (Yost and Eswaran, 1990) that was recently tilled.  Litter bags were recovered after 1, 2, 3, 5, 8 and 16 weeks, the contents washed, oven dried, weighed and nutrient contents analysed.

Data analysis. The results for the bioproductivity studies were assembled within a computer spreadsheet and the data were analysed using summary statistics and Analysis of Variance procedures in Genstat. Decomposition and nutrient mineralisation were analysed by fitting data (observed/initial) over time, expressed as years, to a first order exponential equation (y = exp^(k*time)).  The time to 50% decomposition or mineralisation was calculated by dividing the natural log of 0.5 by the coefficient of exponential decline (k) (Woomer et al., 1999).

Treatment mean values within each measurement were compared by calculating the Least Significant Difference (Little and  Hills, 1975).

Results

Mineral contents of the lake and pond waters are presented in Table 1. Note that the pond water at Kajjansi contained much greater amounts of bases (Na, K, Ca, Mg) and much less soluble phosphorus. 

Table 1. Chemical characteristics of water from the bioproductivity study sites

Standing biomass at the four sites ranged between 300 to 610 t ha^-1  (fresh weight). The highest mass was observed at the Owen Falls dam and the lowest at Bugiri (Table 2).   Results of biomass accumulation (Fig. 1A) and daughter plant production (Fig. 1B) indicate faster growth rates of the waterhyacinth in lake water than in the pond.  While the regression coefficient values for both bioproductivity and propagule production are strong, a distinct pattern of oscillation occurs along the regression lines, particularly at Bugiri Bay. By the 16th week of growth, there was about three times more biomass in the lake water than in the pond water, and twice as many daughter plants had been produced (Fig. 1).

Table 2. Standing biomass of waterhyacinth (kg m-2, fresh weight) on different shores of Lake Victoria

The greatest proportion of plant parts were leaves (Table 3) followed by roots and then stems.  Moisture contents were relatively high in all tissues but leaves were likely to have dried slightly during handling.  N, P and Ca contents tended to be higher in leaves and K and Mg contents higher in stems.  Decomposition and mineralisation rates are presented in Table 4.  ANOVA revealed no significant differences in decay patterns except for K mineralisation, in part because of large variations observed in root data (Table 4).

Table 3. Proportion, moisture and nutrient contents of waterhyacinth and its tissues (standard errors in parentheses)a

^aAfter Woomer et al., 1999

Table 4. Time (days) to 50% decomposition and nutrient mineralisation of fresh waterhyacinth residues (standard errors in parentheses)

Discussion and conclusions

The results of this study demonstrate that waterhyacinth on Lake Victoria has a rapid growth rate (Fig. 1) and achieves biomass accumulation (Table 2) similar to, or greater than, that reported elsewhere in tropical fresh waters (Gopal, 1987).  Within the biomass, considerable nutrients are accumulated but the recovery and utilisation of these nutrients by smallhold farmers in the Lake Victoria Basin is difficult. This difficulty is reduced somewhat by on-going large-scale mechanical clearance in protection of harbours and hydro-electric facilities.  At Owen Falls alone, we estimate that 23.2 t N , 3.5 t P and 52.0 t K accumulates within waterhyacinth covering 33 ha immediately behind the dam.  This biomass has been cleared at regular intervals (about thrice annually) over the past three years and to date, no effort has been made to mobilise this resource for agriculture.  When the estimated 209 t N contained in waterhyacinth and recovered through mechanical clearance at Owen Falls is compared to the 5500 t of nitrogenous fertilisers imported into Uganda between 1995 and 1997 (FAO, 1998) containing an estimated 1100 t N, one can sense that an opportunity for corrective soil fertility re-plenishment (Buresh, 1997) is being lost.     

Our study also indicates that the thickly congested waterhyacinth readily observed from the River Nile’s Owen Falls dam is not particularly representative of much more widespread open water mats and shoreline fringes of Lake Victoria.  A combination of current and predominant wind direction (Woomer, 1997) result in compaction and rapid recolonisation  of waterhyacinth at this site.  However, it is the shoreline fringe that poses an obstacle to small harbours and local commu-nities and it is partial consolation that the control of these weeds requires proportionally less removal of aquatic  biomass.

The pond at Kajjansi had a lower growth rate (Fig. 1) and this may be explained by the fact that phosphorus, important in plant growth, was lower than in lake waters. The results of this study have a direct bearing upon the harvesting and disposal of mechanically cleared waterhyacinth wastes and their potential for utilisation as an organic input to soil.  The rate at which the waterhyacinth plant multiplies (Fig. 1) impacts on the costs involved in harvesting it.  The fact that waterhyacinth grows readily on pond and sheltered water surfaces has far-reaching implications with respect to the potential spread of the weed in Uganda because the large water bodies such as Lakes Victoria and Kyoga have numerous sheltered bays and inlets that are ideal for growth of the weed. Furthermore, large portions of wetlands, which constitute at least 15% of the surface area of Uganda, are suitable for infestation by the weed (Twongo, 1991).

Vegetative reproduction under the climatic conditions on Ugandan waters was very rapid. The doubling time of 4 -7 days observed in the study is close to that of 6.0 - 15 days (Lindsey and Hirt, 1999) and 5-15 days (Twongo, 1991). However, on River Zaire, up to 1200 daughter plants  developed from 2 plants within only four months (Ivens, 1993). In  our study,  the highest number of daughter plants from 1 plant was approximately 91 over a period of 16 weeks. This difference may be explained by the fact that the 1 m² boxes limited the growth of the plants due to competition for space and as a result the production of plant propagules was restricted.

In terms of biomass accumulation, in our study, the highest growth rate was 228 t ha^-1 yr^-1 which is higher than 173, 123 and 106 t ha^-1 yr^-1 observed in Florida, Guyana and Indonesia, respectively (Woomer, 1997). However, in contrast, Lindsey and Hirt (1999) reported a growth rate of between 0.1 and 0.5 kg fresh weight m-2 day-1 corresponding to between 400-1700 t ha^-1 yr^-1.

The extremely high growth rate makes it a very important nutrient sink.  It is a unique plant in its capacity to absorb chemical species from solution (Gopal, 1987). In this study, it was observed that on average, a plant in the pond absorbed 44.7 kg N ha^-1, 6.4 kg P ha^-1 and 174.0 kg K ha^-1 by doubling itself in a week’s time.  The plant in the lake absorbed 99.2 kg N ha^-1, 7.7 kg P ha^-1 and 182.3 kg K ha^-1 in the same time period. Therefore, based on the above information, this ability could be harnessed to recycle chemical nutrients (N, P, and K) that have been eroded into the lake from the surrounding agricultural lands.

The results of this study also shed light on the disposal of waterhyacinth wastes because this material has potential to decompose rapidly (Table 4) but fails to do so under many situations (Solly, 1984; Saustroutomo, 1991; Woomer, 1997).  The current method of waterhyacinth recovery which also collects lake sediments, and disposal site handling, which compacts a mixture of plant waste, sediments and soil, often results in much slower decomposition than is possible.  At sites where the accumulation of waste mounds presents a problem, alternative handling procedures may be developed to accelerate waterhyacinth decay.

Because of its favourable nutrient contents (Table 3) and rapid decomposition (Table 4), fresh waterhyacinth may be usefully applied to soil as a surface mulch.  It is unlikely that further benefit would be obtained from incorporation of waterhyacinth into the soil but at the same time, few additional benefits associated with mulching,  such as soil surface protection from erosion and moisture conservation (Brady, 1990) are unlikely to become realised owing to its rapid decompo-sition.  Guar et al. (1989) observed similar decay of waterhyacinth.  Decay rates for many commonly available crop residues such as banana trash, maize stover and coffee husks are considerably longer (Bwamiki et al., 1998; Lekasi et al., 1999).  The large proportion of water in the fresh residues suggests that farmers should be advised to partially dry the material prior to routine field operations and also suggests that it is not economical to haul fresh hyacinth for long distances for use as a soil input.  Drying waterhyacinth results in accelerated decomposition and N mineralisation (Woomer et al., 1999). At the same time, waterhyacinth wastes must be transported with caution as much of its weedy invasion has resulted from movement by humans (Gopal, 1987).  Nonetheless, poor soil fertility is common in the Lake Victoria Basin (Woomer et al., 1998) including areas nearby those experiencing problems in waterhyacinth waste disposal and these materials could be utilised to our advantage as soil inputs, particularly in combatting banana yield decline (Bekunda and Woomer, 1996).  However, alternative uses of waterhyacinth for fibre, livestock feed, potting soil and biofuel have been reported (Thyagarajan, 1984; Gopal, 1987; Woomer, 1997; Lindsey and Hirt, 1999; Woomer et al., 1999) and it is premature to recommend one over another until economic comparisons are made.

  Acknowledgements

Funds for this research were provided through a grant by the Rockefeller Foundation’s Forum for Agricultural Resource Husbandry, for which the authors are grateful

References

Anderson, J.M. and Ingram, J. S. I. 1993. Tropical Soil Biology and Fertility: A Handbook of Methods. CAB International; Wallingford, U.K.

Bekunda, M.A. and  Woomer, P.L. 1996. Organic resource management in banana - based cropping systems of the Lake Victoria Basin, Uganda. Agriculture Ecosystems and Environment 59:171-180.

Brady, N.C. 1990. The Nature and Properties of Soils. 10th ed. Macmillan Publishing  Company, New York. 621pp.

Buresh, R.J., Sanchez, P.A. and Calhoun, F. 1997.  Replenishing Soil Fertility in Africa.  Special Publication number 51.  Soil Science Society of America, Madison, USA.

Bwamiki, D.P., Zake, J.Y.K., Bekunda, M.A. and Woomer, P.L. 1998.  The use of coffee husks as an organic amendment to improve soil fertility in Ugandan banana production.  In: Carbon and Nutrient Dynamics in Natural and Agricultural Tropical Ecosystems. Bergstrom, L. and Kirchmann, H. (Eds.), pp. 115-121.  CAB International Wallingford, U.K.

Food and Agriculture Organization of the United Nations (FAO). 1998.  FAO Yearbook: Fertilizer. Vol. 47, FAO, Rome.

Gopal, B. 1987. Water Hyacinth. Elsevier Science Publishers, Amsterdam.

Guar, S., Singal, P.K. and Hasija, S.K. 1989. Process of decomposition in Eichhornia crassipes (Mart) Solms:  Early decomposition in different plant parts and effect of site variation. Journal of Environmental Biology 10:23-33.  

Heffernan, B. 1985.  A Handbook of Methods of Inorganic Chemical Analysis for Forest Soils, Foliage and Water.  CSIRO Division of Forest Research, Canberra, Australia. 281 pp.

Ivens, G.W. 1993. East African Weeds and Their Control. Oxford University Press, Nairobi. 289 pp.

Lekasi, J.K., Bekunda, M.A., Tenywa, J.S. and  Woomer, P.L. 1999. Decomposition of crop residues in banana based cropping systems of Uganda. Biological Agriculture and Horticulture17:1-10.   

Lindsey, K. and Hirt, H.M. 1999. Use Water Hyacinth. Anamed, Winnenden, Germany. 114 pp.

Little, T. M. and Hills, F.J. 1978. Agricultural Experimentation: Design and Analysis. John  Wiley and Sons; New York.

Okalebo, J.R., Gatua, K.W. and Woomer, P.L. 1993. Laboratory Methods of Plant and Soil Analysis: A Working Manual.TSBF Program; Nairobi. 88pp.

Sastroutomo, S.S. 1991. Decay rates and nutrient release from decomposing aquatic weeds. Biotrop Special Publication 40:105 - 115.

Solly, R.K. 1984. Integrated rural development with water hyacinth. In: Proceedings of the International Conference on Water Hyacinth Thyagarajan, G.  (Ed.), pp. 70 - 78. United Nations Environment Program, Nairobi.

Thyagarajan, G. 1984. Proceedings of the International Conference on Water Hyacinth. UNEP; Nairobi.

Twongo, T. 1991. Implications of the water hyacinth infestation in Uganda for fisheries, with particular reference to Lake Kyoga. In: Conference on Water Hyacinth in Uganda, 22-23 October 1991, Kampala.  Thopmpson,  K. (Ed.), pp. 19-23.  FAO-TCP/UGA/9153/A.

Twongo, T., Bugenyi, F.W.B. and Wanda, F. 1995. The potential for further proliferation of water hyacinth in Lakes Victoria, Kyoga and Kwania and some urgent aspects for research. African Journal of Tropical Hydrobiology and Fisheries 6:1 -10.

Woomer, P.L. 1997. Managing waterhyacinth invasion through integrated control and utilisation: perspectives for Lake Victoria. African Crop Science Journal 5:309 - 324.

Woomer, P.L., Bekunda, M.A., Karanja, N.K., Moorehouse, T. and Okalebo, J.R. 1998. Agricultural resource management in East Africa. Nature and Resources  34:22 -33.

Woomer, P.L., Muzira, R., Bwamiki, D., Mutetikka, D., Amoding, A. and Bekunda, M.A. 1999.  Biological management of water hyacinth waste in Uganda.  Biological Agriculture and Horticulture 17:181-196.

Yost, D. and Eswaran, H. 1990. Major Land Resource Areas of Uganda. World Soil Resources, Soil Conservation Service, USDA, Washington D.C.

 ©1999, African Crop Science Society

The following images related to this document are available:

Photo images