Hydrology, erosion and nutrient cycling in a forest ecosystem in south Cameroon
I II HYDROLOGY, EROSION AND NUTRIENT CYCLING IN A FOREST ECOSYSTEM IN SOUTH CAMEROON
J.C. Ntonga, M.J. Waterloo and A.B. Ayangma,
Tropenbos-Cameroon Documents 10
The Tropenbos-Cameroon Programme, Kribi, Cameroon
2002
III ABSTRACT
J.C. Ntonga, M.J. Waterloo and A.B. Ayangma, 2002. Hydrology, erosion and nutrient cycling in a forest ecosystem in South Cameroon. The Tropenbos-Cameroon Programme, Kribi. Tropenbos-Cameroon Documents 10, XVIII + 60 pp.; 32 fig.; 9 tab.; 68 ref.; 0 Annexes Rainfall, water and sediment yields were measured and the evaporation quantified in three catchments (2.7-7.7 km2) covered with undisturbed rain forest, selectively logged forest and forest - shifting cultivation. The nutrient balance in the hydrological cycle was also quantified in an undisturbed forest area. Annual evaporation rates were similar in spite of the differences between land uses in the areas and in the regional variation in annual rainfall. On the other hand, the sediment yield depended strongly on the land uses practices. The results of the study of the nutrient budget in four compartments in the hydrological cycle showed that the throughfall and litter percolate constitute the main store of nutrients in the water cycle. In spite of the low nutrient input by rainfall, the losses by the outflow are low; this pattern has been described as an auto-sustained rainforest ecosystem. Keywords: catchment hydrology, undisturbed forest, selective logging, shifting cultivation, water yield, sediment yield, nutrient cycling, nutrient balance, Cameroon.
EUROPEAN COMMISSION Directorate General for Development Development Policy Sustainable Development and Natural Resources
L’Institut de Recherches Géologiques et Minières
The Tropenbos-Cameroon Programme is a research programme executed under the joint responsibility of the Ministry of Environment and Forests of the Republic of Cameroon and Tropenbos International.
This study was achieved with the financial contribution of the European Union Actions in Favour of Tropical Forests in Developing Countries Budget Line (b7-6201). The author is solely responsible for all opinions expressed in this document, and does not necessarily reflect that of the European Union and other donors.
© 2002 Tropenbos International P.O. Box 232, NL-6700 AE Wageningen (the Netherlands). Phone: +31 317 495500; fax +31 317 495520; e-mail: [email protected]
ISSN 1566-2152
No part of this publication may be reproduced or published in any form or by any means, or stored in a database or retrieval system, without the written permission of Tropenbos International.
Tropenbos International assumes no liability for any losses resulting from the use of this document.
IV TABLE OF CONTENTS
PREFACE ...... IX SUMMARY ...... XI RESUME...... XIII 1. INTRODUCTION...... 1 1.1. Background and justification ...... 1 1.2. Objectives of the study ...... 2 1.3. Outline ...... 2 2. PHYSICAL ENVIRONMENT...... 4 2.1. Location, topography and drainage system...... 4 2.2. Climate and hydrology...... 4 2.3. Geology and soils...... 8 2.4. Vegetation and land use...... 9 2.5. Study catchments ...... 10 2.5.1. The Songkwé catchment...... 10 2.5.2. The Biboo - Minwo catchment...... 11 2.5.3. The Nyangong catchment...... 11 3. METHODS ...... 13 3.1. Hydrology and erosion...... 13 3.1.1. Field data acquisition...... 13 3.1.2. Laboratory procedures and methods...... 14 3.1.3. Modelling procedure ...... 15 3.2. Nutrient cycle study ...... 15 3.2.1. Field data acquisition...... 15 3.2.2. Laboratory procedures and methods...... 16 4. HYDROLOGY ...... 17 4.1. regional rainfall distribution ...... 17 4.2. Rainfall interception ...... 19 4.3. Electrical conductivity and pH of stream water...... 19 4.4. Soil hydrological characteristics...... 22 4.4.1. Infiltration rate...... 22 4.4.2. Soil cohesion and compaction ...... 22 4.4.3. Bulk density...... 23 4.4.4. Aggregate stability...... 23 4.4.5. Soil roughness ...... 23 4.4.6. Moisture content of litter layer and biomass ...... 24 4.5. Quantification of the water balance components...... 24 4.6. Catchment sediment yield...... 31 4.7. Hydrological modelling ...... 34 4.8. Meteorology...... 40 5. NUTRIENT CYCLING IN THE HYDROLOGICAL CYCLE ...... 43 5.1. Introduction...... 43 5.2. Sampling and analysis ...... 43 5.2.1. Methodology on water sampling and chemical analyses...... 43 5.2.2. Results of the analyses...... 44 5.3. Nutrient balance in the hydrological cycle ...... 50 5.3.1. Introduction ...... 50 5.3.2. Nutrient input by rainfall ...... 52 5.3.3. Nutrient output in river flow...... 52 5.3.4. Nutrient fluxes in throughfall and in litter percolates...... 52 5.3.5. Export or accumulation of major nutriments...... 54 6. DISCUSSION AND CONCLUSIONS...... 55
V LIST OF TABLES
Table 3.1: Geographical locations and elevations of the rain gauges in the TCP area and of those in Kribi and Ebolowa...... 13 Table 4.1: Quantification of the water balance components over the period 01-01-1996 until 25- 11-1996 for the Songkwé (undisturbed forest) and over a 5-year period (1996, 1997, 1998, 1999 and 2000) for the Biboo - Minwo (selectively logged forest) and Nyangong (forest - shifting cultivation) catchments...... 30 Table 4.2: Sediment concentrations and yields for the three study catchments during the period 09/02/96 until 21/04/01 ...... 33 Table 4.3: Summary of micrometeorological data collected above a secondary rain forest near Adjap between August and November 1996...... 40 Table 5.1: Repartition of water samples over the sampling period in the various compartments ...... 43 Table 5.2: Mean electric conductivity (EC in µS.cm-1) and concentrations (in mgl-1) of Ca2+, 2+ + + - - Mg , K , Na , organic C, total N, total P, NO3 and Cl in rainfall (RF), throughfall (TF), litter percolate (LP) and in river flow (RvF) from May 12, 2000 to April 11, 2001 in the Biboo - Minwo Catchment and the one-ha Ecol2 plot...... 45 Table 5.3: Amount in mm of rainfall (P_Minwo, RF_corr), throughfall (TF), litter percolate (LP) and river flow (RvF) in the Biboo - Minwo catchment from May 12, 2000 to April 11, 2001...... 51 Table 5.4: Nutrient balance and Flows in kg.ha-1 for the major elements studied in the Biboo – Minwo catchment. Rainfall (RF), throughfall (TF), Litter percolate (LP) and River flow (RvF) are considered...... 53 Table 5.5: Amount of nutrient outflows and fluxes in throughfall and in litter percolates (in kg.ha-1) from the Biboo – Minwo catchment and for the sampling period (May 12, 2000 until April 11, 2001)...... 53
VI LIST OF FIGURES
Figure 2.1: Map of the Tropenbos-Cameroon Programme research area, showing the main rivers, the infrastructure and the location of the profile A-B...... 5 Source: Waterloo et al., 2000...... 5 Figure 2.2: Topographical map of the Tropenbos-Cameroon Programme research area6 Source: Waterloo et al., 2000...... 6 Figure 2.3: Slope map of the Tropenbos-Cameroon Programme research area...... 7 Figure 2.5: Topographic maps of the Biboo - Minwo, Nyangong and Songkwé catchments, scale 1:50.000...... 12 Figure 4.1: Mean isohyetic map (mm) of the Kribi – Ebolowa region in south Cameroon for the period 01 January 1996 until 31 December 2000...... 17 Figure 4.2: Mean annual rainfall at the key sites of the TCP area (Kribi, Lolodorf, Ebimimbang, Ebom, Nyangong, Mvié and Ebolowa) in 1996, 1997, 1998, 1999 and 2000...... 18 Figure 4.3: Variation in daily rainfall totals observed at Ebom in 2000...... 18 Figure 4.4: Plots of rainfall vs. throughfall at a primary forest site (a) and a cassava field (b). The difference in rainfall interception characteristics is illustrated in (c), where the regression lines for the two vegetation types are shown...... 20 Figure 4.5: Plot of the EC vs. discharge for the Biboo-Minwo (b), Nyangong (n) and Songkwé (s) catchments...... 21 Figure 4.7: Daily rainfall, discharge and sediment concentration for the Biboo – Minwo catchment (year 1999 and period from 03/01/98 until 28/12/99 for sediment concentration)...... 26 Figure 4.8: Daily rainfall, discharge and sediment concentration for the Nyangong catchment (10/04/98 – 31/12/00)...... 28 Figure 4.9: Daily rainfall, discharge and sediment concentration patterns for the Songkwé catchment (1995 – 1996) ...... 29 Figure 4.10:Water balance components for the TCP catchments (1996 – 2000) ...... 31 Figure 4.11:Sediment rating curves for the Biboo – Minwo, Nyangong and Songkwé catchments...... 32 Figure 4.12:Digital elevation model of the Biboo – Minwo catchment area...... 34 Figure 4.13:Map showing the location and width of roads and skid tracks in the Biboo – Minwo catchment...... 35 Figure 4.14:Average rainfall intensities during the simulation period on 23 March 1996 in the Biboo – Minwo catchment...... 36 Figure 4.15:Measured hydrograph and sediment concentrations for selectively logged Biboo – Minwo catchment during a 48.1 mm rainfall event on 23 March 1996. .. 36 Figure 4.16:Simulated hydrographs and sediment concentrations for the forested and selectively logged Biboo – Minwo catchment during a 48.1 mm rainfall event on 23 March 1996...... 37 Figure 4.17:Erosion map of the selectively logged Biboo – Minwo catchment for a 48.1 mm rainfall event on March 1996...... 38 Figure 4.18:Erosion map of the forested Biboo – Minwo catchment for a 48.1 mm rainfall event on March 1996...... 39 Figure 4.19:Map of Biboo – Minwo catchment showing cells where the erosion increased by more than 1000 kg ha-1 after the construction of skid track...... 39 Figure 4.20:Monthly frequency distributions of the wind direction above secondary forest at Adjap...... 41
VII Figure 4.21:Partitioning of available energy over the latent and sensible heat fluxes on 7 August 1996 at Adjap...... 42 Figure 5.2a: Mean concentrations (in mg.l-1) of Ca2+, Mg2+, K+ and Na+ within the sampling period in the various compartments of the forest ecosystem in the hydrological cycle in the Biboo – Minwo catchment...... 46 - - Figure 5.2b:Mean concentrations of organic C, total N, total P, NO3 and Cl within the sampling period in the various compartments of the forest ecosystem in the hydrological cycle in the Biboo – Minwo catchment...... 47 Figure 5.3a: Average values (with the corresponding standard errors) of EC, pH and 2+ 2+ + + - - concentrations of Ca , Mg , K , Na , organic C, total N, total P, NO3 and Cl for the period from 27/07/00 to 30/11/00...... 48 Figure 5.3b:Average values (with the corresponding standard errors) of EC, pH and concentrations of Ca2+, Mg2+, K+, Na+, organic C, total N, total P, NO3- and Cl- for the period from 06/02/01 to 29/03/01...... 49 Figure 5.4: Amount of rainfall, throughfall, litter percolate and river flow (in mm) recorded at the Biboo - Minwo catchment and the one- ha Ecol2 plot during the water sampling period for the nutrient cycling study (12/05/00 – 11/04/01)...... 51 Figure 5.5: Relation between litter percolate (LP) and throughfall (TF) from May 12, 2000 until April 11, 2001 in the one-ha Ecol2 plot...... 52
VIII PREFACE
About Tropenbos The Tropenbos Foundation was established in 1988 by the Government of the Netherlands with the objectives to contribute to the conservation and wise use of tropical rain forest by generating knowledge and developing methodologies, and to involve and strengthen local research institutions and capacity in relation to tropical rain forests. In 2001, the Foundation changed its name to Tropenbos International.
The Tropenbos Programme carries out research on moist tropical forestland at various locations around the world. Present and former major research sites are in Colombia, Guyana, Indonesia, Ivory Coast, Ghana, Vietnam and Cameroon. At the different locations, research programmes follow an interdisciplinary and common overall approach, with the aim to exchange data and to make results mutually comparable.
About the Tropenbos-Cameroon Programme The Tropenbos-Cameroon Programme (TCP) was established in 1992 by the Cameroonian Ministry of Environment and Forests (MINEF) and the Tropenbos Foundation. The general objective of the TCP is to develop methods and strategies for natural forest management directed at sustainable production of timber and other forest products and services. These methods have to be ecologically sound, socially acceptable and economically viable (Foahom and Jonkers, 1992). TCP consists of several interrelated projects in the fields of ecology, forestry, economy, social sciences, agronomy and soil science.
In 1997, the European Commission (EC) agreed to co-finance five TCP projects, and this component is entitled "Assessment of ecological and economic prospects and limitations for sustainable management of natural forest in Cameroon" (contract B7-6201/96/11/VIII/FOR). The present study is part of this EC-funded component and was financed by the EC, Tropenbos International and the implementing agencies the ‘Institut de Recherches Géologiques et Minières’ (IRGM), Alterra Green World Research and Wageningen University (WU). It is partially based on hydrological data gathered by the Lu1-project, financed by ITTO and CfC.
In 2002, the Tropenbos-Cameroon Programme will be terminated after the final reports of the EC funded component have been published.
Acknowledgements On behalf of the authors, Dr A.D. Mvondo Ze, Dr P. Schmidt and the scientific coordinators Dr B. Foahom and Dr W.B.J. Jonkers are gratefully acknowledged for reading the drafts of this document and commenting on it.
The IRGM is thanked for making a researcher available for this study. The European Commission is thanked for its financial support, which made the realisation of this study.
The authors are grateful for the support from the TCP management team (Dr O. Eyog Matig, Dr M. Tchatat, Mr J.P. Tsimi Mendouga, Ir W.F. van Driel and Dr P. Schmidt), the scientific coordinators Dr B. Foahom and Dr W.B.J. Jonkers, and the TCP administrative staff members (Johan Verhoef, Martin Zogo, Hannah Mokome, Ernest Ola’a and Aristide Ntonga). A word of thanks is due to the students Nangmo Yves Nestor, Fuego Olivier, Mekok Robert, Nkate Simon Petigiakwang, Menno Ruppert, Kooh Jean Simon and Anye Christopher Awason.
IX Special thanks are due to the Lu1 and Ecol2 field assistant Marcel Mva and the rainfall and water level observers Bile Sylvain Bruno, Lambo Jean Paul, Lontchi Charles, Mba Michel, Mva Jean Jacques, Mvondo Jean Louis, Ngbwa Rivel Nestor, Nguiamba Isaac, Njock Siméon, Nna Jean Claude, Nyangono Réné, Oba Joseph, Oyono Fernand, Touandop Augustin and Zamedjo Didier. Last, but not least, the authors are very grateful for the hospitality and cooperation provided by the population of the research area.
B. Foahom and W.B.J. Jonkers Scientific coordinators Tropenbos-Cameroon Programme
X SUMMARY
A 5-year hydrological study has been carried out in south Cameroon by Alterra (formerly the DLO Winand Staring Centre for Integrated Land, Soil and Water Research, SC-DLO) and the “Centre de Recherches Hydrologiques” of the “Institut de Recherches Géologiques et Minières”. The study aimed to assess the impact of land use changes on the hydrology and erosion in rain forest (August 1995 – March 2001) and to quantify basic ecological processes, such as biological production, water and nutrient cycling and accumulation of phytomass and nutrients (May 2000 – April 2001). The results of the present study aims to contribute to the development by the Tropenbos-Cameroon Programme of a land management plan promoting sustainable land use.
The spatial distribution of rainfall was not uniform over the research area due to orographic effects. The central part of the area -, where the elevation increases from about 100 m a.s.l. to 600 m a.s.l. - received a distinctly higher rainfall (2115-2458 mm) than the western lowlands (1816 mm) and eastern uplands (1983 mm) between September 1995 and March 2001. The interception loss in the rain forest canopy at Biboo - Minwo amounted to 35% of incident rainfall, whereas that by a cassava field near Ebom was less at 15% of incident rainfall.
Two types of river water could be distinguished based on electrical conductivity (EC) data collected from 30 streams. The first type was observed in the western lowlands and had an EC of 40-80 µS cm-1. The second type, with a lower EC of 14-35 µS cm-1, was observed in the central part and eastern uplands. The differences in EC could be caused by differences in weathering processes, soils and bedrock.
Agriculture and selective logging both affected soil physical properties. The permeability of the soil was high at 28 m day-1 for undisturbed forest soils in the Biboo - Minwo catchment and decreased sharply to 3 m day-1 upon disturbance of the soil by skidders. Bulk density and soil compaction were lowest in undisturbed forest soils (975-1047 kg m-3) and highest on skid tracks (1297 kg m-3). Top-soil aggregates collected in undisturbed forest were stable, but the stability decreased considerably after disturbance of the soil by skidders. Differences between the forest soils and soils under shifting cultivation were less distinct with somewhat lower cohesion values observed in the latter. The litter layer in undisturbed primary forest at Biboo - Minwo had an average dry mass of 7600 kg ha-1.
The mean annual rainfall and water yield for the Biboo – Minwo and Nyangong catchments varied between 2078-1944 mm and 721-627 mm, respectively. For the Songkwé catchment the values for a one-year observation were respectively 2066 and 920 mm. Mean annual evaporation values were similar in spite of differences in rainfall and land use, ranging from 1358 in the forested Biboo - Minwo catchment to 1316 mm in the Nyangong catchment, which was partly under shifting cultivation. In the present situation with low-intensity land use conversions, the spatial distribution of rainfall, rather than the type of land use, must be considered as the main factor determining the observed variation in annual water yields in the research area.
Sediment concentrations and the annual sediment yields were affected by land use. The lowest concentrations were found for the undisturbed forest cover (Songkwé). The sediment yields were low at all sites, but increased in the order undisturbed forest (31 kg ha-1 for a one-year measuring period) < selectively logged forest (445-637 kg ha-1) < forest - shifting cultivation (702-946 kg ha-1).
A single test run of the results of the modelling of stormflow, erosion and sediment yield from the Biboo - Minwo catchment with the LImburg Soil Erosion Model (LISEM) have been included as a first approximation for the changes following selective logging of undisturbed rain XI forest. As such, the results presented should be considered as indicative only. A 48 mm storm was selected for modelling. The measured discharge, peakflow and sediment yield were 2.3 mm, 1.4 m3 s-1 and 1.4 kg ha-1, respectively. Comparison of the model results for undisturbed forest with those for selectively logged forest indicated that the discharge total, the peakflow and the sediment yield increased significantly after selective logging from 0.4 mm to 1.3 mm, 0.7 m3 s-1 to 2.7 m3 s-1 and from 12.3 kg ha-1 to 39.0 kg ha-1, respectively. The bulk of the increased soil erosion originated from the construction of skidder tracks on slopes exceeding 100. On these slopes, the simulated soil erosion increased with more than 1000 kg ha-1, with a maximum of 140,000 kg ha-1 on a slope of 170, after the construction of the track.
Micrometeorological data were collected from July until November 1996 in a 30 m mast in secondary forest with a height of about 22 m. The data indicated that the daily average temperature, humidity and radiation totals did not show much variation over the four months period. The wind direction was predominantly southwest. A relation was established between the incoming shortwave radiation and the net radiation through regression analysis and this relation compared well with those found for rain forests elsewhere in the humid tropics. The temperature fluctuation energy balance method was used to calculate the evaporation on August 7, 1996 and the result (2.9 mm) compared well with the annual daily average (3.3 mm) obtained with the water balance method for the adjacent Songkwé catchment.
The nutrient balance for the Biboo – Minwo catchment, in an undisturbed forest, was quantified from May 12, 2000 until April 11, 2001 on a basis of input fluxes (atmospheric rainfall), transfer fluxes (throughfall, litter percolate) and output fluxes (solutes in the stream draining the catchment). The nutrient input in rainfall for the major elements were higher that the nutrient output measured in the stream. The sampling of the stream water was easier than those performed in others compartments (e.g. rainfall, throughfall and litter percolate), the sampling being possible at each moment over the sampling period. The values obtained are: 3.3 kg ha-1 Ca, 6.1 kg ha-1 Mg, 3 kg ha-1 K, 8.3 kg ha-1 Na, 1.3 kg ha-1 organic C, 1.3 kg ha-1 total N, 3.2 -1 -1 -1 . -1 kg ha total N, 3.2 kg ha total P, 0.5 kg ha NO3 , 0.6 kg ha Cl. It has been found in the course of this study that the throughfall and litter percolates are the main store of the nutrient of the forest ecosystem in the study area, the values obtained for the both compartments and for the same elements are: 22.9 kg ha-1 Ca, 28.8 kg ha-1 Mg, 93 kg ha-1 K, 14.1 kg ha-1 Na, 4.3 kg ha-1 -1 -1 -1 -1 - - organic C, 4.5 kg ha total N, 3.2 kg ha total N, 10.9 kg ha total P, 1.4 kg ha NO3 , 1.9 kg ha 1 Cl. Forest can be described as being auto-sustainable.
The results presented in this report do not give rise to great concern for the effects of the present low-intensity land use changes on the regional hydrology. However, in view of the increase in sediment concentrations observed after selective logging of rain forest or conversion to agriculture (shifting cultivation), special attention should be given to the protection of village water supply areas to guarantee the water quality. Modelling results indicate that construction of roads and tracks on slopes steeper than 100 should be minimised to prevent excessive erosion and corresponding decreases in the water quality through high sediment concentrations. As the nutrient store is low in the soils, the phytomass, which constitutes the main store, should be prevented for human activities with high disturbances, which could jeopardise the auto- sustainability of the forest.
XII RESUME
Dans le cadre du Programme Tropenbos-Cameroun (PTC), des études hydrologiques ont été menées pendant cinq ans (de 1995 à 2000) par le Centre de Recherches Hydrologiques de l'Institut de Recherches Géologiques et Minières du Cameroun et Alterra, un Institut de recherche des Pays-Bas. Les études ont porté sur un ensemble de bassins versants représentatifs des sites de recherche du PTC localisés à 70 km à l'est de Kribi dans la forêt dense humide sempervirente. L'étude et l'évaluation des impacts des différents modes d'utilisation des terroirs forestiers sur l'hydrologie locale et régionale et ainsi que la quantification des processus écologiques dans un écosystème forestier non perturbé étaient les objectifs du volet hydrologie du Programme. Les résultats obtenus à l'issue de deux projets du PTC (Lu1 et Ecol2) apportent un meilleur éclairage sur les pluies et le ruissellement, l'alimentation des nappes, la qualité des eaux, l'érosion et le transport des matières solides, la quantification du bilan des nutriments dans les différents compartiments de l'écosystème.
La zone de recherche qui couvre une superficie de 2000 km2 est caractérisée par une orographie contrastée. Celle-ci induit une hétérogénéité spatiale de la distribution des pluies. Ainsi la partie centrale de la zone enregistre annuellement plus de précipitations (2115 - 2458 mm) que les zones occidentales de basses altitudes et orientales de hautes altitudes qui reçoivent respectivement 1816 et 1983 millimètres d'eau.
Les mesures de la conductivité électrique (EC) des eaux des rivières et ruisseaux effectuées en différents points de la zone ont permis d'en distinguer deux groupes. Le premier groupe des rivières avec des valeurs de EC entre 40 et 80 µS cm-1 est localisé dans les zones basses à l'ouest de la zone de recherche, tandis que le deuxième avec les faibles conductivités de 14 à 35 µS cm-1 draine l'est de la zone. Les différences de conductivité sont dues à la nature des sols et de la roche mère et ainsi que les processus de lessivage y associés.
L'agriculture itinérante (sur brûlis) et l'exploitation forestière sélective affectent les propriétés physiques des sols. C'est ainsi que la perméabilité (infiltration) est élevée pour les sols de la forêt non perturbée de Biboo - Minwo (28 m/jour) et décroît à 3 m/jour sur les sols du même bassin mais ayant connu plusieurs passages d'engins lourds lors des opérations de débardage des grumes, etc. La densité apparente et la compaction des sols sont également faibles pour les sols en forêt non perturbée (975-1047 kg m3) et élevées pour les pistes de débardage. Les agrégats de la couche superficielle des sols en forêt non perturbée étaient plus stables au test de désagrégation de Imeson et Vis (1984) que ceux des pistes de débardage. Les différences de comportement des sols étaient moins nettes entre les sols de forêt non perturbée et ceux de la forêt sous agriculture. La cohésion des sols était cependant plus faible pour les derniers.
La pluie moyenne annuelle et les lames d'eau écoulées sont respectivement de 2078 et 721 mm pour le bassin versant de Biboo - Minwo et de 1944 et 627 mm pour celui de Nyangong. Pour une année d'observations sur le bassin de la Songkwé à Adjap, ces valeurs ont été de 2066 et 920 mm. L'évaporation moyenne annuelle est restée constante d'une année à l'autre et avec pratiquement les mêmes valeurs sur les bassins versants, 1358 mm pour Biboo - Minwo (bassin forestier) et 1316 mm pour Nyangong (bassin dont une grande partie est le siège des activités agricoles intenses). A l'état actuel de l'intensité observée dans les différents usages des terroirs forestiers, la distribution spatiale de la pluie plutôt que les types d'utilisation des terres demeure le facteur déterminant pour le ruissellement, l'infiltration et l'alimentation des aquifères de la zone de recherche.
La charge en sédiments des cours d'eau et l'érosion spécifique des bassins versants sont affectées par les pratiques d'utilisation des terres forestières. Pour le bassin versant forestier non perturbé de Songkwé, la charge spécifique est de 31 kg ha-1 tandis qu'elle est de 445 à 637 kg ha-1 pour le Biboo - Minwo et de 702 à 946 kg ha-1 pour Nyangong.
XIII Les résultats de la modélisation des processus hydrologiques et de l'érosion avec le modèle LISEM (Limburg Soil Erosion Model) sur le bassin versant de Biboo - Minwo, bien qu'obtenus après un seul test expliquent les changements et impacts qui peuvent suivre la conversion de la forêt naturelle en un autre type d'utilisation. Pour le test du modèle, une averse de 48 mm survenue le 23 mars 1996 avait été sélectionnée. La lame d'eau écoulée, la pointe de la crue et la charge spécifique mesurées ont donné les valeurs suivantes : 2,3 mm, 1,4 m3 s-1 et 1,4 kg ha-1. Deux scénarios du bassin versant de Biboo - Minwo ont été modélisés, la forêt non perturbée et la forêt sous exploitation sélective des grumes. Les résultats de la modélisation montrent une augmentation des valeurs de certains paramètres hydrologiques quand on passe de l'état non perturbé de la forêt à l'état de l'exploitation forestière. La lame d'eau écoulée, la pointe de la crue et la charge spécifique varient respectivement de 0,4 m à 1,3 mm ; de 0,7 à 2,7 m3 s-1 et de 12,3 à 39,0 kg ha-1. Ces résultats indiquent également une augmentation du taux de l'érosion, celle-ci peut être attribuée en priorité aux zones du bassin dont les pentes sont supérieures à 10°. Sur ces pentes, une augmentation de l'érosion spécifique de l'ordre de 1 000 kg ha-1 et un maximum de 140 000 kg ha-1 sur des pentes de 17° après la construction des pistes de débardage ont été obtenus.
Les paramètres météorologiques (pluie, direction et vitesse du vent, température de l'air et du sol, radiation globale et nette, humidité relative, etc.) ont été mesurés pendant une période de 4 mois dans une forêt secondaire près du bassin versant de Songkwé de juillet à novembre 1996. Les résultats obtenus montrent une faible variation des valeurs moyennes journalières de la température, de l'humidité relative et du rayonnement net et global. La direction prédominante du vent est sud-ouest. La valeur de l'évaporation de la journée du 7 août 1996, calculée selon la méthode du bilan énergétique est de 2,9 mm, elle ne diffère pas beaucoup de celle calculée pour le même bassin (par la méthode du bilan hydrologique) qui était de 3,3 mm ce jour.
Le bilan des nutriments pour le bassin versant de Biboo - Minwo en forêt non perturbée a été calculé pour la période du 12 mai 2000 au 11 avril 2001 sur la base des entrées (pluies), flux intra compartiments (pluviolessivats, eaux de percolation sous litière) et des sorties (éléments dissous dans l'eau de rivière). Les concentrations de la majorité des éléments contenus dans la pluie étaient supérieures à celles des éléments drainés par l'eau de la rivière. Pour la période de l'étude, les valeurs sont les suivantes pour les éléments dissous dans la rivière : 3,3 kg.ha-1 Ca ; 6, 1 kg.ha-1 Mg ; 3,0 kg.ha-1 K ; 8, 3 kg.ha-1 Na ; 1,3 kg.ha-1 C organique ; 1,3 kg.ha-1 N total ; 3, -1 -1 - -1 2 kg.ha P total ; 0,5 kg.ha NO3 ; 0,6 kg.ha Cl. La dynamique des flux de nutriments est dominée par les pluviolessivats et les eaux de percolation sous litière, la litière au sol constituant le principal réservoir des nutriments dans le cycle hydrologique de l'écosystème. Pour les deux compartiments réunis (pluviolessivats et les eaux de percolation sous litière), les valeurs calculées sont les suivantes : 22,9 kg.ha-1 Ca ; 28,8 kg.ha-1 Mg ; 93 kg.ha-1 K ; 14,1 kg.ha-1 Na ; -1 -1 -1 -1 - -1 4,3 kg.ha C organique ; 4,5 kg.ha N total ; 10,9 kg.ha P total ; 1,4 kg.ha NO3 ; 1,9 kg.ha Cl.
Les résultats obtenus dans le cadre de l'étude hydrologique montrent que l'intensité actuelle dans les différents modes de l'utilisation des terres forestières n'induirait pas fondamentalement des effets négatifs sur l'écosystème (impacts sur l'hydrologie régionale et dynamique des flux des nutriments). Cependant une attention particulière devrait être apportée à la qualité des eaux utilisées par les populations en évitant par exemple la surcharge et/ou la pollution des eaux potables des rivières, phénomènes observés lors de certaines pratiques dans le mode d'utilisation des terres : exploitation forestière, pêche avec barrages saisonniers, agriculture itinérante, etc. Un essai de simulation numérique des écoulements et de l'érosion a montré que les zones à fortes pentes (excédant 10°) sont à éviter pour les activités d'exploitation des grumes. Les sols du site étant très pauvres en nutriments, la matière végétale sous forme de litière constitue le principal réservoir en nutriments, à cet égard toute action tendant à la diminuer de façon significative et non maîtrisée pourrait perturber durablement les fonctions écologiques de la régénération de la forêt.
XIV 1. INTRODUCTION
1.1. BACKGROUND AND JUSTIFICATION
Due to economic and population pressures, the rain forests in south Cameroon are increasingly being exploited, either for commercial purposes (e.g. selective timber harvesting, oil palm and banana plantations) or for subsistence farming (shifting cultivation). To avoid the biophysical degradation of the area as a result of these activities, calls have been made by the Cameroonian Ministry of Environment and Forests (MINEF) for the development of a land management plan, which should promote and establish sustainable land use in these areas.
As scientific data on which such a management plan could be based were lacking, the Tropenbos-Cameroon Programme was initiated by MINEF and the Tropenbos Foundation in 1992. Research in fourteen interrelated projects of the programme was carried out by the Wageningen University, the “Institut de la Recherche Agronomique pour le Développement (IRAD)”, the DLO Winand Staring Centre for Integrated Land, Soil and Water Research (SC- DLO, now Alterra) and the “Centre de Recherches Hydrologiques” of the “Institut de Recherches Géologiques et Minières (CRH-IRGM)”. Funding for the projects was provided by the International Tropical Timber Organisation (ITTO, Project PD 26/92) through the “Office National de Développement des Forêts (ONADEF)” and by the Tropenbos Foundation in the first phase and by the European Union (EC contract reference number b7- 6201/96/11/VIII/FOR) in the latter phase.
The objective of the Tropenbos Cameroon Programme was to develop strategies and methods to provide a scientific basis for decisions regarding sustainable land use alternatives for the rain forest area. The programme attempts to achieve this goal by combining information obtained from studies of the biophysical environment, with those obtained from studies of the social, political and economical environments. All research activities were carried out in two concessions of the “Houthandel Gebr. Wijma en Zonen B.V.” timber company, which cover 1916 km2 of rain forest in south Cameroon.
Two of the studies contributing to the development of a land management plan are the Forest Land Inventory and Land Evaluation (Lu1) and Functional aspects of the evergreen forest in southern Cameroon (Ecol2) projects.
The Lu1 project was geared towards carrying out a land evaluation study in the Tropenbos Cameroon Programme (TCP) research area, and has provided the biophysical background on which decisions regarding sustainable land use alternatives in the area can be based. Within the framework of this study, a reconnaissance scale (1:100,000) inventory on landforms, soils and vegetation has been carried out and results have been published by van Gemerden and Hazeu (1999). More detailed studies of the relationships between landforms, soils and vegetation have been carried out in four small catchments, which were selected to reflect the physiographical range of the whole study area. The landform, soil and vegetation studies have been complemented by hydrological, ecological and agronomic studies. The results of these small- scale biophysical studies have been combined with the data obtained from the large-scale inventory to define and locate a number of land mapping units. Finally, the land utilisation types (and their specific requirements) have been matched with the qualities of the land mapping units to produce a map depicting land suitability classes for the whole study area (Hazeu et al., 2000) and used for management plans (Fines et al., 2001a; Fines et al., 2001b).
In the tropical rain forests, nutrients are often in short supply and in general it can be said that land use is only sustainable if the nutrient balance is in equilibrium. The importance of nutrient studies in the area research has been emphasised by the TCP. The Ecol2 project which aimed at quantifying some of the key ecological processes in a forest ecosystem started 1n 1998. The 1 results obtained on net primary productivity, phytomass accumulation and nutrient cycling within the area in spite of the short observation period are essential and useful tools for the land use management. The hydrological study was carried out by the CRH-IRGM, in collaboration with Alterra. The present report provides an overview of the hydrological components inside the Lu1 and Ecol2 projects. The study period extends from August 1995 until April 2001.
1.2. OBJECTIVES OF THE STUDY
It is well-known that changes in land use affect the total water yield, the runoff distribution pattern, (Bosch and Hewlett, 1982; Bruijnzeel, 1990, 1993; Malmer, 1993; Fritsch, 1993; Waterloo, 1994; Sahin and Hall, 1996), the sediment yield (Douglas, 1967, 1968; Baharuddin, 1989; Bruijnzeel, 1990; Abdul Rahim and Zulkifli, 1994), as well as the nutrient balance (Bruijnzeel, 1990; Stoorvogel, 1993). Clearfelling of tropical rain forests, whether for timber harvesting or for agricultural purposes, and subsequent burning of the slash generally causes an increase in the annual water yield as a result of a decrease in rainfall interception and transpiration losses. The effect becomes less marked as soon as the area becomes covered by closed-canopy secondary forest vegetation. Changes in the runoff pattern (e.g. decrease in baseflows) have been observed in some cases where the soil had severely been disturbed, limiting percolation to deeper ground water reservoirs during periods with high rainfall (Bruijnzeel, 1990). Changes in water yield or runoff distribution are less obvious in areas where forests are selectively logged, which is the common practice in south Cameroon. In general, no significant changes have been observed when less than 20% of the area was affected by logging (Gilmour, 1977a; Subba Rao et al., 1985; Sahin and Hall, 1996). Sediment yields from areas covered with undisturbed forest are usually lower than those of areas with similar physiography under other land use systems (Douglas, 1967). A conversion from forest to another type of land use therefore usually causes an increase in the river water sediment concentrations, the magnitude of which depends mainly on the intensity of soil disturbance during and after the conversion (Bruijnzeel, 1990). The nutrient cycle in a rainforest has close links with the hydrological cycle, the rainfall inputs being one of the providing systems of nutrients in the ecosystem. In forests growing on nutrient poor soils as in south Cameroon, any interference with forest could have a serious effect on the nutrient budget (UNESCO - MAB, 1994). The impact can be severe where logging is carried out at a high intensity and in uncontrolled manner, because logging removes a proportion of the aboveground nutrients held in the timber. The nutrients contained in the extra soil eroded constitute also for an important part to the nutrient loss of the forest (Bruijnzeel, 1990).
The aim of the hydrological study was to assess the sensitivity of the hydrology of the rain forest area to changes in land use (Dolman and Waterloo, 1995) and to quantify the nutrient balance in the hydrological cycle.
To achieve this goal, studies on rainfall distribution, catchment water use, sediment yield and nutrient cycling were initiated in areas with contrasting land use (undisturbed forest, selectively logged forest and shifting cultivation).
1.3. OUTLINE
This report provides the main results of the hydrological component of the Forest Land Inventory and Land Evaluation (Lu1) and the Functional aspects of the evergreen forest in southern Cameroon (Ecol2) studies of the Tropenbos-Cameroon Programme. Chapters 1 to 3 are parts of the two Lu1 reports (Waterloo et al., 1997, 2000). They provide background information on the Tropenbos-Cameroon Programme and the hydrological study in particular, on the study area and on the methods used for data collection. The hydrology of the area is discussed in detail in Chapter 4, which provides information on the rainfall distribution; soil properties, water and sediment yield for more than five complete years. The results of the
2 hydrological modelling of selective logging in the Biboo - Minwo catchment as well as a brief overview of the micrometeorology are also discussed in Chapter 4. The nutrient cycle in the hydrological cycle (Ecol2 project) is reviewed in Chapter 5. A discussion of the results and conclusions are provided in Chapter 6.
3 2. PHYSICAL ENVIRONMENT
2.1. LOCATION, TOPOGRAPHY AND DRAINAGE SYSTEM1
The Tropenbos-Cameroon Programme area is located in a humid tropical rain forest in South Cameroon (South Province, Departments of Océan and Mvila). The area consists of two concessions (Nos. 1600 and 1790) of the 'Houthandel Gebroeders Wijma & Zonen B.V. (GWZ)' timber company (Wijma-Douala SARL), which cover an area of 1916 km2. The concessions are situated between the villages of Lolodorf (3°14'N, 10°44'E) in the North, Adjap-Essawo (3°02'N, 10°52'E) in the East, Akom II (2°48'N, 10°34'E) in the South and Bipindi (3°04'N, 10°25'E) in the West (see Figure 2.1).
A gradual transition from lowland in the West to upland in the East occurs in the TCP area, as shown in Figures 2.2 and 2.4, where the topography and a WNW-ESE elevation profile have been presented, respectively. The elevation ranges from about 60 m a.s.l. near Bipindi, to up to 1057 m a.s.l. in the Bingalanda mountain range near Nyangong. The topography ranges from undulating to rolling in the lowland area, but changes to steeply dissected in the more mountainous areas in the Southeast. Isolated hills, which are several hundreds of meters higher than the surrounding area and which have slopes exceeding 500, occur in both the lowland and upland areas (van Gemerden and Hazeu, 1999), as is illustrated by the slope map presented in Figure 2.3.
The area is drained by four major rivers, of which the Lokoundjé River is the largest. This river enters the research area at Lolodorf and drains the northern part of the area (Figure 2.1). The second largest river is the Tchangué River, which drains the central part of the area and confluents with the Lokoundjé just south of Bipindi. The southwestern part of the area is drained by the Songkwé River, whereas the Biwomé River drains the south eastern part. The flow direction of the larger rivers is predominantly from the north-northeast to the south- southwest, reflecting the regional pattern of faulting. The drainage system can be considered coarse dendritic to trellised. The stream density, as determined for a combined surface area of 73 km2 (Saa and Biboo-Minwo research areas), is 1.6 km km-2.
2.2. CLIMATE AND HYDROLOGY
The climate is humid tropical with two distinct wet seasons (August - November, March - May) and two dry seasons in a year, associated with the movement of the intertropical convergence zone over the area. The average annual rainfall generally decreases in an eastern direction, ranging from 2836 mm year-1 in Kribi (n= 45 yrs) to 2096 mm in Lolodorf (n= 25 yrs) and 1719 mm in Ebolowa (n= 48 yrs; Olivry, 1986). In Lolodorf, maximum monthly rainfall averages 375 mm in October and 263 mm in April. Dry season values are well below 100 mm month-1. Average rainfall values for a 5-year period (1996, 1997, 1998, 1999 and 2000) which are: 3712 mm in Kribi, 2341 mm in Lolodorf, 1241 mm in Ebolowa are not very different for those calculated by Olivry for long-term period.
1 Many parts of this section are full parts of the section 2 of the previous Lu1 hydrology component reports (Waterloo et al., 1997 and 2000).
4 Figure 2.1: Map of the Tropenbos-Cameroon Programme research area, showing the main rivers, the infrastructure and the location of the profile A-B. Source: Waterloo et al., 2000.
5 Figure 2.2: Topographical map of the Tropenbos-Cameroon Programme research area Source: Waterloo et al., 2000.
6 Figure 2.3: Slope map of the Tropenbos-Cameroon Programme research area. Source Waterloo et al., 2000
1000
Mefak Nyangong 800
Minkan 600
Nyamenkoum 400
Elevation [m a.s.l.] Bipindi Ebimimban
200
TCP area 0 0 20 40 60 80 Distance WNW-ESE [km]
Figure 2.4: WNW-ESE profile (see Figure 2.1 for location of profile A-B) through the Tropenbos Cameroon Programme research area. Source Waterloo et al., 2000
7 The air temperature shows little variation over the year with minimum monthly values of 25.0 0C and 22.9 0C in August and maximum values of 27.5 0C and 25.0 0C in March in Kribi (10 m. a.s.l.) and Ebolowa (628 m a.s.l.), respectively (Olivry, 1986). The relative humidity is high throughout the year, with minimum monthly values varying between 70% and 78% in Kribi and between 62% and 74% in Ebolowa. Wind speeds are generally low, being less than 4 m s-1 for 98% of the time. However, high wind speeds may occur during the passage of squall lines associated with large thunderstorms. The wind direction is predominantly Southwest to West (Olivry, 1986).
The discharge patterns of the rivers of the research area and around strongly reflect the seasonal rainfall pattern discussed above, with maximum values being observed in October and May, and minimum values in February and August. The long-term (1950-1977) monthly average discharge of the Lokoundjé river at Lolodorf (1150 km2) ranged from 8 m3 s-1 in February to 65 m3 s-1 in October. Corresponding values for the Kienké (1435 km2, 1955-1977) and Lobé (2305 km2, 1953-1977) rivers ranged between 16 m3 s-1 and 121 m3 s-1 and 21 m3 s-1 and 284 m3 s-1, respectively (Olivry, 1986). The long-term average annual discharge of the Lokoundjé river at Lolodorf amounted to 773 mm. Higher annual values of 1082 mm and 1397 mm were observed for the Kienké and Lobé rivers, respectively. This may be attributed to higher rainfall inputs received by these more coastal basins, as compared to that of the Lokoundjé basin. Based on data collected at very few rainfall stations, Olivry (1986) estimated the long-term annual rainfall input to the Lokoundjé river basin at 1880 mm, whereas those for the Kienké and Lobé basins were both estimated at 2425 mm. This implies that runoff coefficients are in the range of 41% (Lokoundjé river) to 58% (Lobé river). Long-term annual evaporation values, obtained with the water balance method using the rainfall and runoff values quoted above were rather similar for the three basins, ranging from 1025 mm for the Lobé basin to 1107 mm for the Lokoundjé basin and 1345 mm for the Kienké basin (Olivry, 1986). It should be noted that the errors in the evaporation totals may be considerable due to the uncertainties in the annual rainfall inputs into the basins.
2.3. GEOLOGY AND SOILS
The TCP research area is located on the Precambrian shield, which is the most extensive geological formation in Cameroon. The shield consists mainly of metamorphic rocks (gneisses, micaschists, quartzites) and old volcanic intrusions (diorites, gabbro) (Franqueville, 1973). The geology of the TCP area reflects that of the overall Precambrian shield. The rocks are mostly acid gneisses, the composition varying between light-coloured quartzites, quartz-biotite- muscovite gneisses and granite gneisses to darker coloured pyroxene-rich gneisses. Basic ferro- magnesian amphibolite, diorite or gabbro intrusions occur locally within the more acid rock formations. These intrusions are mostly oriented in a NE-SW direction in discontinuous bands along the main faults (Bilong, 1992). Four erosional planes can be distinguished, with elevations increasing from 50-100 m a.s.l. in the western part to 200-300 m and 400-500 m in the central part and 600-800 m a.s.l. in the eastern part (Figure 2.2).
Based on drainage characteristics and texture, four soil types have been distinguished in the area (van Gemerden and Hazeu, 1999). Poorly drained soils were commonly found in the river valleys and adjacent swamp areas throughout the research area. These soils were characterised by a thin, sometimes peaty, A-horizon and a mottled B-horizon showing phases of oxidation- reduction (gley). The texture was often sandy to gravelly with clay interlayers. Following the FAO classification system, such soils were classified as Gleysols or Fluvisols (van Gemerden and Hazeu, 1999).
The moderately well to well-drained soils were subdivided into sandy soils (Ebimimbang soil type), clayey soils (Ebom soil type) and very clayey soils (Nyangong soil type).
8 The Ebimimbang soil type occurred mainly in the lowland area (50-350 m a.s.l.) between the Lokoundjé and Tchangué rivers near Ebimimbang. It was a yellowish-brown sandy clay loam soil with a sandy (60-90% sand) A-horizon and a soil depth rarely exceeding 2 m. It was the most nutrient-rich soil type in the area and has been classified as Acri-xanthic, Acri-plinthic or Plinthic Ferralsol. The Ebom soil type had a higher percentage of clay (20-60%) with the minimum clay content in the upper horizon. It was a yellowish-brown to strong-brown soil developed on gneisses and had a depth exceeding 1 m. This soil type was common in the central part of the area at elevations between 350 m and 600 m a.s.l. and has been classified as a Xanthic or Plinthic Ferralsol.
The Nyangong soil type was a deep to very deep clay soil with a sandy clay loam to sandy clay topsoil (clay content of 35-80%, with the lower value observed in the A-horizons). The colour was yellowish to strong-brown. This soil has developed on fine grained gneisses and was deep to very deep. It has only been observed in the eastern part of the area at elevations above 500 m a.s.l. and has been classified as a Xanthic Ferralsol. More detailed descriptions of the landforms and soil types have been presented in van Gemerden and Hazeu (1999).
2.4. VEGETATION AND LAND USE
The vegetation in the TCP area consists mainly of humid tropical rain forest. Selective logging of the more accessible forests has presumably not affected their structure and floristic composition to a large extent because logging intensities remained invariably low.
The western and central lowlands (elevation lower than 700 m a.s.l.) are covered by evergreen lowland rain forest. The canopy of the lowland forest can be subdivided into four structurally different levels, which usually show a gradual transition from one level to the next. The crowns of emergent trees, often surpassing 60 m in height, form the highest level and cover 20-30% of the soil surface. Mature trees with their canopy level at 25-40 m form the second layer and their crowns cover 60-80% of the surface. Shrub and herb layers may reach heights of 3-6 m and 1 m, respectively, and their foliage covers 40-60 % of the surface. Climbers (lianas) are abundant in the canopy and in gaps where light conditions are favourable to their growth. The lowland rain forest can be characterised by the presence of Dialium spp., Calpocalyx dinklagei, Hymenostegia afzelii and Saccoglottis gabonensis (van Gemerden and Hazeu, 1999) and is relatively rich in commercial timber tree species, of which Lophira alata (Azobé), Cynometra hankei and Saccoglottis gabonensis are the most important. A separate swamp-type of vegetation is common in the relatively broad valley bottoms where soils are close to saturation during most of the year (Foahom and Jonkers, 1992).
The eastern, more mountainous part of the TCP area (700-1050 m a.s.l.) is covered by sub-montane forest, which is characterised by a low, irregular canopy at 15 to 20 m height, occasionally reaching a height of 35 m, with a foliage cover of 70-80%. The shrub layer (3-7 m), which in some instances replaces the tree layer as canopy, varies from closed to very open. The herb layer is closed and may reach heights of up to one meter. The canopy is often climber infested and the presence of epiphytic mosses gives these forests the appearance of 'cloud forests'. The broad-leaved trees, probably both evergreen and deciduous, are branched at low heights and have an irregular stem shape. The forest can be characterised by the presence of Drypetes spp., Anisophyllea polyneura, Maranthes glabra and Scorodophloeus zenkeri. These sub-montane forests are not of great interest for commercial timber exploitation because of their inaccessibility (steep slopes) and the irregular shape of the stems of the few timber species (van Gemerden and Hazeu, 1999). Hunting and gathering of non-timber forest products are therefore the most important human activities in these forests.
The more easily accessible lowland forests have been selectively logged by any of the various timber companies that have operated in the area over the past decades, or constitute of old regrowth on abandoned fields used in the shifting cultivation system by the local population for
9 their subsistence. The latter type of forest is usually located in a zone of several kilometres wide along the main roads and villages. More detailed descriptions of the various vegetation types and land use practices have been presented by van Gemerden and Hazeu (1999). Selective logging has been practised for decades in areas of which the slopes did not exceed 200. The intensity of logging is rather low at about 1 tree ha-1. The extraction is restricted to the bole of the tree (with an average diameter of 1.16 m), and the slash is left in situ to decompose. The most important commercial tree species extracted from the area are Lophira alata (Azobé, over 60% of the extracted volume), Erythrophleum ivorensis (Tali), Pterocarpus soyauxii (Padouk), Distemonanthus benthamianus (Movingui) and few redwood species. The trees are felled with chain saws, after which the bole is separated from the slash. The extraction is mechanised and is done by teams consisting of two skidders (Caterpillar 528) and one bulldozer (Caterpillar D7). The damage afflicted to the forest during logging and timber extraction (gaps, tracks and roads) covers roughly 7% of the area logged (pers. comm. G.J.R. van Leersum, 1996). These gaps, skid tracks, and to a lesser extent the landings, are rapidly invaded by vines, climbers and pioneer tree species such as Musanga cecropioides and, although at a much lower intensity as the former, by Lophira alata.
Agricultural activities are generally restricted to the lower slopes and valley bottoms where shifting cultivation practises have lead to the presence of fields, weed infested thickets and secondary forest vegetation along the main roads and around villages. In the shifting cultivation system, the farmer clears small plots (usually not larger than two hectares at a time) of secondary forest, or occasionally virgin rain forest, from most of the vegetation. This is followed by burning of the slash at the end of the dry season and planting at the onset of the wet season. The usual practice is to plant a combination of crops at the same time. The most important food crops are cassava, coco-yam, banana and plantain, peanuts, maize and yams. Cocoa has traditionally been the most important commercial crop. More recently however, large commercial plantations of oil palm and plantain have been established in the area. The fallow succession consists mainly of Chromolaena odorata in fields of up to five years old, Musanga cecropioides and other fast-growing pioneer trees in seven to nine years old fields, followed by a dominance of slower growing tree species in older fields. The length of the fallow period varies greatly, ranging from a minimum of three years on more fertile sites to up to twenty years on less productive sites, with an average of about twelve years (pers. comm. L. Nounamo, 1996).
2.5. STUDY CATCHMENTS
Hydrological measurements were made in three key catchments (Songkwé, Nyangong and Biboo-Minwo catchments). These catchments were selected after an extensive field survey of the soils, vegetation and hydrology had been completed by the land evaluation study team (see van Gemerden and Hazeu, 1999), to assure the representativity of the sites with respect to the larger project area. Topographic maps of the three catchments are presented in Figure 2.5. In addition to these catchments, a fourth catchment (Saa catchment) was selected by the land evaluation team for detailed soil and vegetation studies to cover the full range of soils and vegetation types in the TCP area. A short description of the physiography of the catchments is given below.
2.5.1. The Songkwé catchment The Songkwé catchment is located northwest of the village of Adjap-Mvié and covers a surface area of 2.7 km2. The elevation ranges from 310 m at the discharge measurement site (2055.16'N, 10032.60'E) to just over 440 m at the top of the catchment. Slopes are fairly steep, with maximum values of over 220 on the hills forming the catchment boundary. The rock consists mainly of light-coloured gneiss and the soil has tentatively been classified as belonging to the Ebom soil type (Section 2.3).
10 The vegetation consists for a large part of primary or old-secondary rain forest and has a species composition typical for the lowland rain forest in the area. The height of emergent trees is well over 40 m, even in the area affected by shifting cultivation, where some of the large trees have not been cut to provide shading for the crops. Few small agricultural fields are presently in use, and these are mostly located along the road in the southernmost part of the catchment area. Their influence on the hydrology of the catchment is negligible.
2.5.2. The Biboo - Minwo catchment The Biboo - Minwo catchment is located 7 km northeast of Ebom and covers a surface area of 7.7 km2. The elevation ranges from about 430 m at the discharge measurement site (3006.48'N, 10044.32'E) to 719 m on top of a hill in the centre of the basin. Slopes are moderately steep to steep, with maximum values of up to 500. Soils have tentatively been classified as belonging to the Ebom soil type, although the Nyangong soil type may also be found on the hill slopes in the catchment. Rock outcrops are common on hills and in the river valleys. A 100*100 m plot of undisturbed rainforest close to the discharge station was selected for the ecological measurements (see Ibrahima et al., 2002).
The vegetation consists of primary forest on the hills, where the steepness of the slopes does not allow for timber harvesting or shifting cultivation, and of primary or old-secondary forest in less accidented terrain. The forest has a species composition typical for the transition between lowland rain forest and lower montane rain forest. Less accidented areas (37% of the total catchment area) were selectively logged between May 1995 (eastern part of the catchment) and August 1996 (southwestern part). The logging intensity in these areas was about one tree per ha, resulting in a value of 0.4 trees per ha for the whole catchment. Trees were felled with a chain saw and the boles were extracted using a team consisting of two Caterpillar skidders (528) and one bulldozer (D7). The skid track density in the selectively logged area was 3.9 km km-2. Skid tracks covered 2% of the logged area, but the area covered by tracks showing severe disturbance and compaction of the soil was significantly smaller at 0.6% of the logged area, or 0.2% of the total catchment area. There was no indication of recent agricultural activity in the area.
2.5.3. The Nyangong catchment The Nyangong catchment is located south of Nyangong village and drains part of the Bingalanda mountain complex. The basin area is 6.8 km2 and the elevation ranges from 550 m near the discharge measurement site (2058.11'N, 10045.18'E) to about 1010 m. Slopes are steep, often reaching values of above 450. Rock outcrops are abundant. The rock consists of fine- grained gneisses with intrusions of quartzite. The soil may be classified as belonging to the Nyangong soil type.
The main vegetation types are primary and old-secondary rain forest on the steeper slopes, and lower montane forest on the hilltops above 800 m a.s.l. About 35% of the catchment area is affected by agriculture (pers. comm. M. Yemefack, 1996), but this area includes abandoned fields, which are covered with young secondary forest. The area presently under agriculture may cover less than 10% of the catchment area. Shifting cultivation and cocoa plantations are commonly found along the larger river valleys, where slopes are less steep. However, several agricultural fields have recently been established within primary forest on steep hill slopes at higher elevations (above 700 m a.s.l.).
11 Figure 2.5: Topographic maps of the Biboo - Minwo, Nyangong and Songkwé catchments, scale 1:50.000. Source Waterloo et al., 2000.
12 3. METHODS
3.1. HYDROLOGY AND EROSION
3.1.1. Field data acquisition The regional rainfall pattern has been established from daily data collected at eleven rainfall stations. Nine rain gauges (Productive Alternatives Inc., capacity 230 mm) were installed in August 1995 at Lolodorf, Bipindi, Ebimimbang, Ebom, Melan, Minkan, Nyangong, Mvié and Akom II. All rainfall gauges were installed with their collecting surfaces at heights between 1.6 and 1.8 m. Daily rainfall totals collected by the bureaux of the “Direction de la Météorologie Nationale du Cameroun” at Kribi and Ebolowa have been used to provide additional information. An overview of the location of all rainfall stations and their elevations is given in Table 3.1.
Table 3.1: Geographical locations and elevations of the rain gauges in the TCP area and of those in Kribi and Ebolowa. Location Latitude North Longitude East Elevation (m a.s.l.) Kribi 2057.11' 9054.53' 10 Ebolowa 2054.00' 11010.00' 628 Lolodorf 3014.17' 10043.55' 480 Bipindi 3004.64' 10024.65' 70 Ebimimbang 3002.67' 10028.25' 100 Ebom 3004.73' 10041.24' 410 Melan 3003.66' 10048.25' 610 Minkan 2059.12' 10039.68' 400 Nyangong 2058.11' 10045.18' 550 Mvié 2054.22' 10033.31' 420 Akom II 2048.88' 10033.04' 400
Throughfall measurements were made in rain forest in the Biboo - Minwo catchment (from April 4, 1996 to November 1997 [Lu1 project] and during the nutrient cycle study from May 12, 2000 to April 11, 2001 [Ecol2 project]), as well as in a cassava field (from April 10, 1996 to November 1997 [Lu1 project]) near the village of Ebom. In the forest 36 custom-made throughfall gauges (orifice 56 cm2) were installed in a 25x25 m grid, whereas 24 gauges were placed in the cassava field. Rain gauges were located in sufficiently large gaps at both sites to measure the above-canopy rainfall. The rainfall and throughfall measurements were made on a daily basis during the week. Less regular measurements were made during the weekends due to lack of transport to the sites. During the Ecol2 project, 40 gauges (116.9 cm2 of surface reception funnel draining water into a plastic vessel of 5 l) were used for the throughfall measurements.
The electrical conductivity (EC) of river water was determined using an Eijkelkamp EC-meter capable of measuring ranges from 0-30 µS cm-1 to 0-10,000 µS cm-1 at a temperature of 25 0C (Lu1 study). Water level measurements were made at the outlets of the Songkwé, Biboo-Minwo and Nyangong catchments using Hydrotrack Well Sensor dataloggers with staff gauge references. The sensors in the Minwo and Songkwé rivers had a resolution of 1.8 cm (range 0-5 m), whereas that at Nyangong had a resolution of 1.0 cm (range 0-3 m). The water level data were collected at 30-minute intervals. From November 1995 until March 31, 2001, as a backup, local observers have been employed to record the water levels twice daily and more frequently during stormflows. Later on in the course of Lu1 and Ecol2 studies, to supply for the continuous water recording, mechanical OTT water level recorders were installed at Nyangong and Biboo – Minwo for the replacement of the default Hydrotrack Well Sensor dataloggers.
Stage - discharge relations at the three measurement sites were established from water level readings and corresponding discharge measurements using the salt dilution technique at low discharges (Q < 0.5 m3 s-1) and the velocity - area method (Qualimetrics model 6660 digital
13 water current meter, and OTT C31 water current meter) at higher discharges. From November 23, 1995 until March 31, 2001, 93 stage – discharge measurements have been performed for the Biboo and 70 for the Nyangong river. The discharge rating curves for the three catchments under study were obtained using the TIDHYP software. TIDHYP (Traitement Informatique des Données Hydropluviométriques or Computerised Analysis of Hydrological Data) is software developed by the Centre de Recherches Hydrologiques of IRGM (Yaoundé).
Water samples were collected regularly at the discharge measurement sites in 1000-ml plastic containers. In addition, an automatic water sampler (ISCO) was installed in the Biboo - Minwo catchment to sample storm runoff automatically. The location and width of roads skid tracks and landings were mapped using a global positioning system in combination with a compass, topofil and tape measure. Four classes of disturbance to the forest and soil have been recognised. The main and permanent access roads to a logging site and the landings were defined as Class 1 of disturbance (severely compacted surfaces, sometimes gravelled, no vegetation). The degree of damage to the soil on skid tracks was assessed visually at the time of mapping and these soils were classified as being severely damaged (Class 2, A-horizon removed, B- or C-horizon exposed and severely compacted, no vegetation cover left) or moderately damaged (Class 3, A-horizon still present but compacted, some vegetation left). The fourth class of disturbance consisted of tracks where the vegetation had been damaged but the soil remained relatively undisturbed during the (single) passage of the skidder.
Measurements of the infiltration rates of water into the soil were made using an Eijkelkamp double-ring infiltrometer system with a diameter of the inner ring of 30 cm. The saturated permeability was determined from repeated measurements of the rate of descent of the water level in the inner ring until a constant value was obtained.
Soil cohesion was determined with a pocket tor vane with a blade width of 25 mm. The average of 20 measurements was used as the soil cohesion value of the sample area. An indication of the compaction of the soil at the soil surface was obtained using a pocket penetrometer. Measurements were made on forest soils, as well as on soils under shifting cultivation and on skid roads and tracks. The soil roughness (or random roughness) was determined using a 1 m rod with 47 holes, spaced at a distance of 2 cm, through which the micro-relief of the soil could be determined using a pin and ruler. The standard deviation of the 47 measurements was taken to represent the soil roughness (pers. comm. A.P.J. de Roo, 1995).
A micrometeorological set-up, which measures wind speed at 30.03 m (Vector A100M), wind direction at 30.36 m (Vector W200P), temperature and relative humidity at 29.87 m (Vaisala HMP35A), thermocouple temperature at 30.01 m (custom built), global radiation at 29.92 m (Kipp & Zn. CM6B), net radiation at 28.10 m (R.E.B.S. Q7), rainfall at 29.52 m (ARG 100) and soil temperatures (PB107 probes, at 2 and 10 cm depth) was installed in secondary forest on a former field in the Songkwé catchment. The height of the vegetation around the mast was about 22 m. All data were measured at 30-second intervals from June 28 until November 1996 by a Campbell Scientific CR10 datalogger and averaged to provide half-hourly values. The thermocouple temperature was measured at 0.125-second intervals and averaged over 5-minute intervals. Rainfall totals were also measured at 5-minute intervals.
3.1.2. Laboratory procedures and methods Water samples were analysed for sediment concentrations using a Sartorius filter unit connected to a water-jet vacuum pump. Pre-weighted cellulose nitrate filters (Sartorius A.G., order no. 13906-50-ACN) with a diameter of 50 mm and a pore diameter of 0.47 µm were used to collect sediment particles in a predetermined volume of water sample (0.5-1.0 l). The filters were subsequently dried at a temperature of 105 0C and reweighed on a Sartorius balance (model BP-
14 110), capable of weighing to the nearest milligram. The discharge - concentration data are presented in Waterloo et al., 1997; 2000.
The pH of a selected number of samples was determined using an Aqualytic pH-17 pH meter (Ingold electrode), which was calibrated using buffer solutions of pH = 4.00 and p H= 7.00. Soil aggregate stability was determined following the drop-test method of Low (1954) and Imeson and Vis (1984). The test involved counting the number of drops necessary to break down a moist soil aggregate with a diameter of 4.0-4.8 mm, such that it passed completely through a sieve with a maze width of 3.0 mm. The drops of water (weight 0.1 g) were supplied by a reservoir connected to a nozzle with PVC tubing, and were allowed to fall on the aggregates from a height of 1.0 m through a PVC pipe (15 cm diameter). The test was repeated on 20 aggregates and the median value was taken to represent the aggregate stability.
3.1.3. Modelling procedure The LImburg Soil Erosion Model (LISEM) was selected to simulate the hydrological behaviour and erosion in the catchments for various land use scenarios. LISEM is a physically based distributed-area model developed by the Department of Physical Geography from the University of Utrecht. It is completely incorporated in a raster geographical information system (PCRASTER). Detailed descriptions of the model equations, capabilities and data requirements have been presented by Roo et al. (1994; 1995).
Because the model has been developed and tested for agricultural land on löss soils in a temperate climate, rather than for tropical rain forest areas, some of the equations describing the hydrological processes may need to be modified (e.g. rainfall interception, flow transport capacity). Field data collected during the present study will be used to test the model and to develop new equations when necessary. In addition, these data will be used to calibrate the model with respect to the predicted water yield and runoff distribution, as well as with respect to the sediment yield.
Digital Elevation Models (DEMs) have been developed for the Saa and Biboo - Minwo study catchments using digitised contour lines (40 m interval) from the 1:100,000 topographical map from the research area. The maps were digitised in vector format using ARC/INFO and subsequently rasterised. The SURFER program was then used to develop a DEM using a geostatistical interpolation method (kriging) to interpolate between data points. The resulting raster maps (cell size 25x25 m) were converted to the CSF file format for further processing with the PCRASTER geographical information system developed by the Department of Physical Geography of the University of Utrecht (van Deursen and Wesseling, 1992). Maps of the soil characteristics, land use, areas covered by the various rain gauges, roads and tracks, streams etc. have been digitised for the same catchments and have also been converted to the CSF file format. As such complete sets of maps in digital form exist for these areas.
When properly calibrated, the LISEM model will be run using extreme rainfall events as input because a very large proportion of the erosion takes place during these events (Bruijnzeel, 1990). The hydrological information obtained from the forested Songkwé catchment was foreseen to form the baseline against which the effects of land use changes on the hydrology, as evident in the Biboo - Minwo (selective logging) and Nyangong (shifting cultivation) basins, could be measured. But the data collection period reduced to 6 months only could not allow fulfilling this objective.
3.2. NUTRIENT CYCLE STUDY
3.2.1. Field data acquisition To quantify the internal and external nutrient cycles, a 100*100 m plot has been chosen in the Biboo – Minwo catchment, in an undisturbed forest near the hydrological station. The terrain is
15 slightly sloping. The chemical composition of throughfall and litter percolate have been determined in the 100*100 m, whereas the nutrient status of the atmospheric inputs and the river was successively quantified in the former landing park in I2 and at the outlet of the Biboo – Minwo catchment.
The amount of atmospheric rainfall (input in the ecosystem) has been measured with a standard totalling rain gauge (Productive Alternatives Inc., capacity 230 mm), whereas water for chemical analysis has been collected with a funnel (12.2 cm diameter) and a collector vessel (plastic, 25 l). The funnel was covered with wire mesh on the top to prevent entry to coarse debris. Filter wool was placed in the narrow end of the funnel to prevent entry of fine biomass and soil particles and insects. To store the water in the dark between samplings, a shelter has been constructed for the plastic collector. Both the standard totalling rain gauge and the rain gauge for chemistry were placed in the I2 landing, near the hydrological station.
Throughfall was determined using 40 rain gauges (surface reception 116.9 cm2) and plastic collectors (5 l) equipped with wool filters in the funnels to avoid contamination from forest debris. To minimise errors related to the variability of the canopy, the gauges were relocated after each sampling and the filter wool replaced. The gauges were emptied every week (and after each storm event) and stored in a dark emplacement in the hydrological camp until the final sampling at the end of each sampling period.
Litter percolates were collected over the period of sampling using 10 percolation plates (surface reception of 126.68 cm2) inserted between the litter layer and the mineral soil. The gauges consisted of a plastic tray, covered with nylon wire mesh and equipped with wool filters to avoid entrance of debris into the collector. The plates are connected (tubing) to collecting plastic vessels (2 l) placed in the soil pits down slope of the trays. The gauges were maintained in a fix position over the sampling period. The quantification of the nutrient output from the river was performed at the outlet of the Biboo – Minwo catchment using plastic collectors of 500 ml for the sampling. Base flow was sampled weekly and kept in dark in the hydrological camp until the final sampling. Due to the short period of sampling, only two storm flows have been sampled completely.
The period of sampling extended from May 12, 2000 until April 11, 2001. For the different compartments, some chemical properties of water and the following nutrients have been 2+ 2+ + + - - quantified: EC, pH, Ca , Mg , K , Na , organic C, total N, total P, NO3 , Cl .
3.2.2. Laboratory procedures and methods At the end of each period of sampling (one month in general), samples were brought to the TCP laboratory in Kribi where they were processed and stored in the fridge at 4° C. The final samples retained for the nutrient cycle study were then sent to the “Laboratoire des Sols et de l’Environnement” of the University of Dschang. The analysis methodology is discussed in the relevant chapters.
16 4. HYDROLOGY
4.1. REGIONAL RAINFALL DISTRIBUTION
To obtain information on the rainfall distribution within the TCP area, daily rainfall observations collected between September 1995 and April 2001 at the eleven rainfall stations have been analysed. Daily rainfall data of the TCP area are therefore available for five complete years (1996, 1997, 1998, 1999 and 2000).
Over the period of data collection the variation of rainfall in the TCP area was considerable, both in space and time. Annual rainfall was invariably highest at the coastal station of Kribi (always above 3000 mm with a maximum of 4581 mm in 1998) and decreased to values between 1500-2500 mm in the research area at Ebimimbang (minimum of 1501 mm in 1997), Akom II (minimum of 1484 mm in 2000) and at Ebolowa (minimum 1487 mm in 1998). The village of Ebom showed very little variation of mean annual rainfall within the 5-year period.
The isohyetic map for the mean year is shown in Figure 4.1. The variation in rainfall amounts for the key sites of the TCP area over the 5-year period is shown in Figure 4.2. Both Figures confirm the findings in Waterloo et al. (1997, 2000) that there is a distinct spatial rainfall pattern. For the mean year, the lowland (Ebimimbang) and upland (Nyangong) areas received less rainfall (1816 mm and 1983 mm, respectively) than the lowland - upland transition zone along the SSW-NNE axis Lolodorf - Ebom - Minkan - Mvié (2115-2458 mm). This suggests that orographic effects caused by the lowland-upland transition (Figure 2.2) may contribute significantly to the rainfall in this zone. The seasonal variation in 2000 at Ebom is presented in Figure 4.3. The daily rainfall totals collected during that year in Ebom are similar to those calculated for the mean year and showed a clear seasonal trend with high rainfall during October - November and April - May, and much lower rainfall from December - March.
Lolodorf
3.20 Bipindi Ebom Ebimimbang Melan Minkan 3.00 Nyangong Kribi Mvie Ebolowa
Northern latitude (°) latitude Northern Akom II 2.80
10.00 10.20 10.40 10.60 10.80 11.00 11.20 Eastern longitude (°)
Figure 4.1: Mean isohyetic map (mm) of the Kribi – Ebolowa region in south Cameroon for the period 01 January 1996 until 31 December 2000.
17 Water balance components for the TCP catchments in 1998
3000
2000 Rainfall Disc h a rg e Evaporation 1000
0 Biboo - Minw o N yangong
Figure 4.2: Mean annual rainfall at the key sites of the TCP area (Kribi, Lolodorf, Ebimimbang, Ebom, Nyangong, Mvié and Ebolowa) in 1996, 1997, 1998, 1999 and 2000.
100,0 ) 80,0
60,0
40,0
20,0 Daily rainfall (mm 0,0
Date (January - December 2000)
. Figure 4.3: Variation in daily rainfall totals observed at Ebom in 2000. The daily rainfall values recorded in 2000 are similar to those of the mean year (1996 – 2000).
Daily rainfall data may be used to calculate an antecedent precipitation index (API). This index can be considered as an indicator for the variation of the wetness of the soil in time and may be calculated from the equation given below using daily rainfall records (van de Griend, 1979 in Waterloo et al., 1997 and 2000).
n t =API ∑ t KP 1=t where: Pt = precipitation on the day before the calculation date n = number of days used in the calculation K = a constant (range: 0.80-0.98)
Waterloo et al., (1997 and 2000) have calculated API values for Ebom and Lolodorf using a seven days time period and a constant of 0.8. A good knowledge of the API specially for the logging process will have impact on the efficiency of the latter, because soil moisture conditions affect the efficiency of the logging operation and the degree of damage afflicted to the soil, API values may be used to serve as indicators for the efficiency of the skidders in hauling logs from
18 the forest under varying weather and soil moisture conditions (pers. comm. G.J.R. van Leersum, 1996). Forestry research on this topic should indicate at which API level logging operations cease to be feasible in an economic way, as well as in an environmental way.
4.2. RAINFALL INTERCEPTION
Conversion of one vegetation type to another type with a different canopy structure usually results in changes in the amount of water reaching the soil surface due to changes in the interception loss in the canopy of the vegetation (Bonell and Balek, 1993). In south Cameroon, forest land is usually cleared for shifting cultivation and a study has been initiated to determine the effects of a conversion from primary rain forest to a cassava crop, which is the dominant crop type in the Ebom area. The effects of such a conversion on the interception of rainfall were determined by comparing rainfall and throughfall measurements made in primary forest in the Biboo - Minwo catchment with those made in a cassava field south of Ebom. The results of this study have been discussed in more detail by Ruppert (1996).
At the forest site in the Biboo – Minwo catchment near the hydrological camp, incident rainfall and throughfall measurements were performed during two distinct periods. The first period (April 11 until July 9, 1996) was related to the Lu1 project whereas the second (May 12, 2000 until April 11, 2001) took place in the framework of Ecol2 project (nutrient cycling in the hydrological cycle). For the first period the results obtained for the primary forest amounted to 683 mm (36 measurements) for the incident rainfall and 436 mm for the corresponding throughfall. With an approximation for the stemflow during this period of 1.5% of the above- canopy rainfall (11 mm) the interception loss could be calculated at 238 mm, or 35% of incident rainfall. Rainfall and throughfall totals in the cassava field over the same period were somewhat lower at 498 mm and 413 mm, respectively. As reported in Waterloo et al. (1997; 2000) the interception loss for this crop was 76 mm, or 15% of incident rainfall (stemflow again set at 1.5% of total rainfall). Plots of rainfall vs. throughfall for the forest vegetation and for the cassava field are shown in Figure 4.4a and 4.4b, respectively. Regression lines were calculated for both rainfall - throughfall data sets and these lines are shown in Figure 4.4c, together with the regression equations. A comparison of the interception losses shows that a conversion from forest to crop types similar to cassava will result in an increase in the amount of water reaching the soil by some 20% of incident rainfall. When such a conversion is made on a large scale, the catchment water yield may be expected to change accordingly, as long as all other environmental factors (e.g. soil permeability) remain the same. The Figure 4.4 shows the different results for both sites during the first period.
The results of incident rainfall and throughfall data during the second period will be discussed in chapter 5 of this report.
4.3. ELECTRICAL CONDUCTIVITY AND PH OF STREAM WATER
The chemical composition of water in small streams is strongly influenced by the chemical composition of the precipitation and the rate and intensity of the chemical processes in the soil (biological activity, buffering reactions, precipitation or dissolution of minerals, etc.) and rock (weathering of minerals) from which the water is derived (Duchaufour, 1982). The chemical composition of rainfall in the area is low, as indicated by the low EC-value of 5 µS cm-1 (measured on two occasions at Kribi and Ebom in the beginning of the Lu1 study and later, during the Ecol2 study from May 12, 2000 until April 11, 2001). The low EC value may be considered fairly uniform over the area, as such differences in the chemical composition of stream water in the area will reflect differences in soil and rock, rather than variations in the chemical composition of the precipitation. Because the EC is a very good indicator of the sum of anions and cations up to EC-values of about 2000 µS cm-1 (Appelo and Postma, 1993), measurements of the variation of the EC of stream water may be used as an indicator for the rate and intensity of these chemical processes. 19 80
70 Rain forest Biboo - Minwo Rainfall: 772.6 mm 60 Throughfall: 491.1 mm
50
40
30 Throughfall [mm] 20
10
0 0 20 40 60 80 Rainfall [mm]
80
70 Rain forest Biboo - Minwo Rainfall: 772.6 mm 60 Throughfall: 491.1 mm
50
40
30 Throughfall [mm] 20
10
0 0 20 40 60 80 Rainfall [mm]
80
Cassava field 60 Tf = -0.77 + 0.89 * P, r2= 0.99
40 1:1 line Throughfall [mm] 20 Rain forest Tf = -0.64 + 0.667 * P, r2= 0.94
0 0 20 40 60 80 Rainfall [mm]
Figure 4.4: Plots of rainfall vs. throughfall at a primary forest site (a) and a cassava field (b). The difference in rainfall interception characteristics is illustrated in (c), where the regression lines for the two vegetation types are shown. Source Waterloo et al., 2000.
During the Lu1 study (Waterloo et al., 1997 and 2000), EC measurements were taken from 30 streams, draining areas varying in size from less than 1 km2 to over 3000 km2 (Lokoundjé River at Bipindi). Regular measurements in the three research catchments indicated that the EC showed little variation in time (less than 5 µS cm-1 variation over the period of a year). Measured EC values ranged from 14 µS cm-1 in two catchments near Lolodorf and Nyangong, to 76 µS cm-1 in the Saa catchment near Ebimimbang.
20 Based on the EC data, Waterloo et al. (1997, 2000) distinguished two types of river water. The first type is characterised by relatively high EC values (range: 40-80 µS cm-1) and has only been observed in the area west of the line Bipindi - Ebimimbang -Mimfombo - Adjap - Mvié. The higher EC values encountered in this area support the findings of the soil scientist that the nutrient status of the soil (Ebimimbang soil type) in this area is higher than that in the soils in other parts of the TCP research area (van Gemerden and Hazeu, 1999).
The second type of river water is characterised by low EC values (range: 14-35 µS cm-1). This type is typical for most parts of the area, as well as for large areas upstream of the TCP research area as indicated by the EC values of 20 µS cm-1 and 24 µS cm-1 observed for the large Tchengué and Lokoundjé (at Lolodorf) river systems. The small range in EC for this type of water suggests that the intensity of chemical processes in the soil and rock is quite uniform throughout the area. The intensity of chemical processes in the soil and in the bedrock may be considered as very low because the EC value of this type of stream water is close to that of the rainfall (EC of about 5 µS cm-1).
All study catchments were located in the area with the second type of surface water (EC range 17-25 µS cm-1, n = 3) but detailed mapping of the landforms, soils and vegetation was also carried out in the Saa catchment near Ebimimbang where the EC at the basin outlet measured 75 µS cm-1. During the Ecol2 study, 66 measurements of EC river water samples at the outlet of the Biboo – Minwo catchment were performed, the results were similar to those described above (see section 5). A decrease of the EC was observed with increasing discharge in all study catchments, which may be attributed to dilution of the baseflow with rainwater with a lower EC. The steepest decrease was observed in the Songkwé catchment, as can be seen in Figure 4.5.
26 s s b 24 bs s b s b 22 sb sb b b n n b bs bb b 20
EC [uS cm-1] sb Songkwe b nnn n b Biboo - Minwo 18 b n nn n n Nyangong 16 0 20000 40000 60000 80000 100000 Discharge [m3 day-1]
Figure 4.5: Plot of the EC vs. discharge for the Biboo-Minwo (b), Nyangong (n) and Songkwé (s) catchments. Source Waterloo et al., 2000
The pH of river water collected at the outlet points of the small-scale study catchments was measured on several occasions between 26 February and 4 April 1996. The water was slightly acidic with pH values of 5.53±0.14 for the Nyangong river (n = 6), 5.55±0.13 for the Songkwé river (n = 8) and at 5.60±0.26 for the Biboo river (n = 6). The pH values for the three catchments were not significantly different. Further pH rainfall and pH river measurements were performed during the nutrient cycling study between May 12, 2000 to April 11, 2001. The results are discussed in section 5 and are not very different from those obtained during the first study.
21 4.4. SOIL HYDROLOGICAL CHARACTERISTICS2
One of the aims of the hydrological component of the Lu1 project was to determine the hydrological controls in the current land use systems encountered in the Tropenbos region and their vulnerability to changes in land use, particularly with respects to erosion hazards (Dolman and Waterloo, 1995). As such, a model (LISEM) was selected, which needed a set of various parameters to be run properly. All the measurements of soil parameters required were performed during the Lu1 study.
The LISEM model requires the spatial distribution of a number of soil parameters as input for the prediction of the hydrological behaviour and erosion patterns in an area. Measurements of such soil properties have been initiated in the Nyangong and Biboo-Minwo catchments and to a much lesser extent in the Saa catchment.
In the Nyangong catchment a survey has been carried out to see if differences existed in the cohesion and compaction of forest soils on the one hand, and of soils under shifting cultivation (fields, fallow areas) on the other hand. In the Biboo-Minwo catchment, the effects of logging on soil properties such as bulk density, compaction, cohesion and permeability were measured by comparing measurements made on various classes of skid tracks with those made in adjacent forest areas. The results of these studies have been discussed in more detail in Nangmo (1996), and only a short overview will be presented below.
4.4.1. Infiltration rate Measurements of infiltration rates were made on undisturbed forest soils and on adjacent soils, which were severely disturbed by the passage of skidders during timber harvesting. The infiltration rate in the forest soil was high, averaging 28±23 m d-1 (range: 8-84 m d-1, n = 10). Much lower rates were observed on the skid tracks where the infiltration rates ranged from 0.1- 8.4 m d-1, with an average of 3.1±3.3 m d-1 (n = 7). The differences between the means of the undisturbed and disturbed soils were significant at a confidence level of 95% (Student's t - test).
4.4.2. Soil cohesion and compaction The cohesion of the topsoil is one of the parameters in the LISEM model, which determines the detachment of sediment by overland flow (de Roo et al., 1995). This parameter has therefore been measured with a pocket tor vane in the Nyangong, Biboo - Minwo and Saa catchments. In the Nyangong catchment, measurements of soil cohesion were made in primary or old- secondary forests, in soils in young (5-20 years old) secondary forests, in agricultural fields and in swampy valleys. The cohesion was highest in the old-secondary or primary forests (0.30±0.08 kg m-2, n = 59) due to the abundance of organic matter in the topsoil (root mat). The cohesion of the soils in the agricultural fields was significantly lower (95% confidence level) at 0.17±0.05 kg cm-2 (n = 8), whereas intermediate values were observed for the young secondary forests (0.23±0.06 kg cm-2, n = 8). The lowest values were observed in the swampy valley bottoms where the average soil cohesion was 0.12±0.01 kg cm-2 (n = 3). In the old-secondary or primary forests, the cohesion increased with slope steepness, ranging from 0.28±0.09 kg cm-2 on slopes of 0-20%, to 0.34±0.08 kg cm-2 on slopes steeper than 40%. Evidence of natural erosion (sheet wash) exposing a less well-developed A-horizon, was observed for the latter type of soils.
The compaction of the soil at the surface was measured with a pocket penetrometer for various land use types and slope classes in the Nyangong catchment and showed a pattern similar to that of the soil cohesion. The compaction was highest for the forest soils at 0.41±0.19 kg cm-2 (n = 59), which could be due to the presence of a well-developed litter layer and root mat. The
2 The following sections of this chapter are integral parts of two scientific reports issued earlier in 1997 and 2000 (Waterloo et al., 1997 and 2000) 22 compaction of the forest soils showed an increase from 0.39±0.20 kg cm-2 (n = 25) for soils on slopes less than 20%, to 0.46±0.21 kg cm-2 (n = 17) for soils on slopes exceeding 40%.
The compaction of the soils in agricultural fields (root crops, peanuts, plantain, banana) where root mats were usually not observed, was significantly lower (95% confidence level) than that of the forest soils at 0.22±0.14 kg cm-2 (n = 8). This suggests that the conversion from forest to agriculture results in a lower cohesion and compaction of the soil due to decomposition of the soil organic matter. Intermediate values (0.26±0.16 kg cm-2, n = 8) were again observed for young (5-20 years old) secondary regrowth on former agricultural fields, suggesting that a recovery of the soil cohesion and compaction occurs, although at a slow rate.
The cohesion of forest soil in the Minwo - Biboo catchment was similar to that of the forest soils in the Nyangong catchment with an average of 0.33±0.05 kg cm-2 (n = 5). Much higher cohesion values (0.53±0.19 kg cm-2, n = 6) were found in areas where the soil had severely been disturbed by skidders during log extraction.
The compaction at the soil surface under forest in the Biboo - Minwo catchment was 0.35±0.23 kg cm2 (n = 5), which is not significantly different from the value observed for the forest soils in the Nyangong catchment. The disturbance of the soil as a result of the construction of skid tracks and roads resulted in a high compaction value of 0.95±0.72 kg m-2 (n = 6) for the areas affected.
The degree of soil cohesion and compaction depended on the number of passes made by the skidder and the highest values were observed on the main skid track (0.80 kg cm-2 and 2.35 kg m-2, respectively), whereas those of a soil disturbed by a single skidder pass were similar to the forest values at 0.23 kg cm-2 and 0.29 kg m-2, respectively.
4.4.3. Bulk density The bulk densities of the surface horizons of forest soils in the Biboo - Minwo and Nyangong catchments were 975±96 kg m-3 (n = 4) and 1047±66 kg m-3 (n = 8), respectively. The bulk density of the topsoil of skid tracks in the former catchment was significantly higher at 1297±354 kg m-3. In a separate study, Voeten (1996) observed an average bulk density of 968±178 kg m-3 (n = 31, range 627-1335 kg m-3) for undisturbed forest soils in a different part of the Biboo - Minwo catchment and a significantly higher (99% confidence level) value of 1241±271 kg m-3 (n = 46) for skid tracks.
4.4.4. Aggregate stability The stability of soil aggregates is an important parameter determining the rate of splash detachment in the LISEM erosion model (de Roo et al., 1995). Aggregates collected from the clayey A-horizon in the forests of the Nyangong and Biboo - Minwo catchments were very stable, as they did not break upon the impact of over 200 drops of water. Soil aggregates collected on frequently used skid tracks in the Biboo - Minwo catchment were less stable with median values of 30 and 63 drops. Lower values were also observed for the sandy soils under agriculture in the Ebimimbang area (Saa catchment), where the median values ranged between 20 and 118 drops (n = 7). The aggregate stability of soils under different land use types (e.g. shifting cultivation) in the Nyangong catchment has yet to be determined.
4.4.5. Soil roughness The soil roughness, or random roughness, is a measure of the micro-relief and is used by the LISEM model to simulate surface storage of water in depressions (Onstad, 1984). At present, four measurements have been made on forest soils at the discharge-gauging site in the Biboo -
23 Minwo catchment. The roughness of these soils varied between 1.1 and 3.4 cm and averaged of 2.0 cm.
4.4.6. Moisture content of litter layer and biomass The biomass and moisture content of the litter layer (L+F layers) was determined from 0.25 m2 samples collected between 20 June and 29 November 1996 in a primary rain forest plot in the Biboo - Minwo basin. The litter biomass over this period averaged 7602±2352 kg ha-1 (n = 92) and contained between 0.3 mm and 1.5 mm of moisture. Monthly averages of the litter biomass decreased from 7978±2149 kg ha-1 (n = 20) in July to 5932±1553 kg ha-1 (n = 17) during the dry season in August and increased again gradually during the wet season to 8584±2632 kg ha-1 (n = 19) in November. A detailed quantification of the biomass and moisture of the litterfall and litter layer has been performed in the vegetation Ecol2 component between January 1999 and June 2001 (see Ibrahima et al., 2002).
4.5. QUANTIFICATION OF THE WATER BALANCE COMPONENTS
Several methods and equations can be used to obtain the long-term evaporation rates (or runoff deficit) for a basin. One among those methods commonly used is the water balance method. For a given year (water year), the period considered for the calculation is the one beginning by the absolute low water and ending by the next absolute low water. The water balance equation reads: ∆ L-S-Q-P=E where: E = evaporation (mm) P = mean aerial precipitation (mm) Q = discharge (mm) ∆S = change in soil moisture and ground water storage (mm) L = leakage into or out from the catchment (mm)
As the basement of the catchments consists of solid gneiss and the discharge measurement sites were located on bedrock, leakage through the river bedding and to regional ground water flows through faults is presumably negligible and has been ignored in the water balance calculations. Prior to the actual report, components of the water balance for the study catchments were quantified using rainfall and discharge data for the period 27 November 1995 until 26 November 1997. Since then, rainfall and discharge data have been recorded until April 2001 except for the Songkwé catchment for which the observations stopped in 1997. A 5-year period pattern of daily rainfall – discharge data and periodic sediment transport is available for the Biboo – Minwo and Nyangong catchments for 1996-2000. Rainfall totals for the different catchments were estimated from observations made at Mvié and Adjap (Songkwé catchment), at three stations in the Nyangong catchment and at Ebom, Lolodorf and Minwo (Biboo - Minwo catchment).
Water yields were calculated by converting water level readings (H, in cm) to discharge values (Q, in m3 s-1) using discharge-rating curves (see Waterloo et al., 2000). The actual discharge rating curves take in account the whole period of water level – discharge observations from November 11, 1995 until March 31, 2001 for the Biboo – Minwo and Nyangong catchments.
24 Discharge - rating curve, Biboo - Minwo catchment (23/11/95 - 31/03/01)
15 ) 1 .s- 3 10
5 Discharge (m 0 0 50 100 150 Stage (cm)
Discharge - rating curve, Nyangong (23/11/95 - 31/03/01)
4
3 .s-1) 3 2
Discharge (m 1
0 0 20406080
Stage (cm)
Discharge - rating curve, Songkwé (23/11/95 - 02/10/96)
0,3 0,25 ) 1 .s-
3 0,2 0,15 0,1 Discharge (m 0,05 0 0 102030405060
Stage (cm)
Figure 4.6: Discharge rating curves for the Biboo – Minwo, Nyangong catchments (23/11/95 – 31/03/01) and for the Songkwé catchment (1995 – 1996).
For the Biboo - Minwo, the measurements ranged for the water level – discharge from 14 cm and 0.021 m3 s-1 to 129 cm and 8.01 m3 s-1, whereas for the Nyangong they ranged from 13 cm and 0.012 m3 s-1 to 71 cm and 1.81 m3 s-1. The highest water levels recorded were 139 cm on October 26, 1999 for Biboo – Minwo and 103 cm on February 2, 2000 for Nyangong.
25 100,0 ) 80,0
60,0
40,0
20,0 Daily rainfall (mm
0,0 1-janv 2-m ars 1-m ai 30-juin 29-août 28-oct 27-déc Date (year 1999)
6,000
4,000
2,000 Discharge (m3/s)
0,000 1-janv 2-m ars 1-m ai 30-juin 29-août 28-oct 27-déc Date (year 1999)
0,600
0,400
(kg/m3) 0,200
Sediment concentration 0,000 3-janv 4-févr 15-avr 17-m ai 30-m ai11-juin 27-juin 18-oct23-févr11-avr 3-juil Date (03/01/98 - 28/12/99)
Figure 4.7: Daily rainfall, discharge and sediment concentration for the Biboo – Minwo catchment (year 1999 and period from 03/01/98 until 28/12/99 for sediment concentration).
The rating curves were obtained using TIDHYP software developed by the “Centre de Recherches Hydrologiques” of IRGM (Yaoundé, Cameroon). TIDHYP (Traitement Informatique des Données Hydropluviométriques) stands for Computerised Analysis of Hydrological Data. The equations used have of general form of:
n Q = Q0(H-H0) ,
Where Q0 and n are constants, which can be determined on scaled logarithmic paper, H0 is a hydraulic constant of the river control channel. The new regression equations are given below, whereas the rating curves and water level - discharge data points are presented in Figure 4.6.
26 Songkwé River: -13 (H = 20-30 cm), H0 = 0, Q0 = 16.74*10 , n = 7.30131 Q = 16.74*10-13*H7.30131 -4 (H = 30-150), H0 = 0, Q0 =34.00*10 , n = 1 Q = 34.00*10-4*H1 N = 16 (number of discharge measurements), the average deviation of measured discharges from the fitted curve is 0.1426479
Biboo river: -5 (H= 1-40 cm), H0 = 0.0, Q0 = 21.91*10 , n = 1,74650 Q = 21.91*10-5*H1,74650 -7 (H = 40-72 cm), H0 = 11, Q0 = 46.69*10 , n = 3,05707 Q = 46.69*10-7*(H-11)3,05707 -7 (H = 72-200 cm), H0 = 0.0, Q0 = 13.49*10 , n = 3.22915 Q = 13.49*10-7*H3.22915 N = 93 (number of discharge measurements), the average deviation of measured discharges from the fitted curve is 0.3563092
Nyangong river: -8 (H= 12-200 cm), H0 = -5, Q0 = 60.86*10 , n = 3.47547 Q = 60.86*10-8*(H+5)3.47547 N = 70, (number of discharge measurements), the average deviation of measured discharges from the fitted curve is 1.032687.
There was little difference with the discharges calculated by Waterloo et al. (1997, 2000) using the older rating curves obtained in the Lu1 study. As the TCP catchments form now a part of the Cameroon hydrological network, comparisons between the various stations (over 400 in Cameroon) are easier as uniform methods of calculation are used. This justifies the use of the TIDHYP package for hydrological calculations within the Cameroon hydrological network. Daily rainfall totals, discharges and sediment concentrations for specific period for each catchment are shown in Figures 4.7, 4.8 and 4.9.
In the long-term water balance calculations, changes in the soil moisture and ground water storage were neglected, because the effects on calculated evaporation may be considered small over periods longer than a year. The results of the water balance calculations for a 5-year period (1996- 2000) for the TCP catchments are presented in Table 4.1 and in Figure 4.10.
27 150,0 )
100,0
50,0 Daily rainfall (mm
0,0 10-avr-98 07-oct-98 05-avr-99 02-oct-99 30-m ars-00 26-sept-00 Date (10/04/98 - 31/12/00)
4,000
3,000
2,000
1,000 Discharge (m3/s)
0,000 10-avr-98 07-oct-98 05-avr-99 02-oct-99 30-m ars-00 26-sept-00 Date (10/04/98 - 31/12/00)
0,200 n
0,100 (kg/m3)
Sediment concentratio 0,000 10-avr-98 26-avr-98 20-m ai-98 15-août-98 07-sept-00 05-oct-00 17-déc-00 Date (10/04/98 - 31/12/00)
Figure 4.8: Daily rainfall, discharge and sediment concentration for the Nyangong catchment (10/04/98 – 31/12/00)
28 100,0
) 80,0
60,0
40,0
Rainfall (mm 20,0
0,0 11-août- 10-oct-95 9-déc-95 7-févr-96 7-avr-96 6-juin-96 5-août-96 4-oct-96 3-déc-96 95 Date (1995 - 1996)
0,300
0,200
0,100 Discharge (m3/s)
0,000 23-nov-95 22-janv-96 22-m ars-96 21-m ai-96 20-juil-96 18-sept-96 17-nov-96 Date (1995 - 1996)
0,150 n
0,100
(kg/m3) 0,050
Sediment concentratio 0,000 09-févr-96 20-avr-96 04-juin-96 28-sept-96 19-nov-96 Date (1995 - 1996)
Figure 4.9: Daily rainfall, discharge and sediment concentration patterns for the Songkwé catchment (1995 – 1996)
29 Table 4.1: Quantification of the water balance components over the period 01-01-1996 until 25-11-1996 for the Songkwé (undisturbed forest) and over a 5-year period (1996, 1997, 1998, 1999 and 2000) for the Biboo - Minwo (selectively logged forest) and Nyangong (forest - shifting cultivation) catchments.
Year Component of Biboo – Minwo Nyangong Songkwé water balance 1996 Rainfall (mm) 2113 1815 2066 Discharge (mm) 824 422 920 Evaporation (mm) 1289 1393 1146 Runoff Coeff. 0.39 0.23 0.45 1997 Rainfall (mm) 1834 1701 Discharge (mm) 417 490 Evaporation (mm) 1417 1207 Runoff Coeff. 0.23 0.29 1998 Rainfall (mm) 2083 2022 Discharge (mm) 839 749 Evaporation (mm) 1244 1273 Runoff Coeff. 0.40 0.37 1999 Rainfall(mm) 2387 2335 Discharge (mm) 913 712 Evaporation (mm) 1475 1623 Runoff Coeff. 0.38 0.30 2000 Rainfall (mm) 1974 1847 Discharge (mm) 610 762 Evaporation (mm) 1364 1085 Runoff Coeff. 0.31 0.41
The components of the water balance for the Biboo – Minwo and Nyangong catchments may be considered good estimations because the major storms events have all been recorded with a good accuracy, especially for the years 1998-2000. The annual rainfall at the Biboo - Minwo catchment was consistently higher than at Nyangong. This was also reflected in the annual discharges of these catchments. The Table 4.1 clearly shows that differences in observed water yields between the study catchments can be explained by differences in rainfall recorded for each basin. As such, evaporation rates remained fairly constant over the study period and differences between years were negligible.
The conversion from rain forest to selectively logged forest or to shifting cultivation generally causes a decrease in the evaporation due to decreases in the canopy interception loss and, to a lesser extent, in the transpiration rate (Bonell and Balek, 1993; Bruijnzeel, 1990). The values given in Table 4.1 indicate that the evaporation was lowest in the undisturbed forest and highest in the shifting cultivation. The low evaporation value recorded in 1996 for Biboo – Minwo can be attributed to the less frequent sampling of discharge during this period due to which the discharge may have been overestimated. The errors in the evaporation totals may be as large as 20% due to errors in the measured rainfall (about 10%) and discharge totals (about 15%) and to a much lesser extent due to errors introduced by neglecting changes in soil moisture and ground water storage and leakage (Lee, 1970). The impact of differences in land use on the catchment water yield and evaporation was not significant as it fell well within the range of the measurement errors. This is not surprising as most of the catchment areas remained permanently under forest.
The evaporation totals presented in Table 4.1 were similar to those given by Olivry (1986, Section 2.2) for much larger basins in south Cameroon and to those published by Seyler et al., 1993; 1115-1322 mm yr-1) for the Ngoko River basin in southeast Cameroon, where rainfall was somewhat lower (1460-1689 mm). The values are also well within the range published by Bruijnzeel (1990) for rain forests in the humid tropics (900-1400 mm yr-1).
30 Water balance components for TCP catchments in 1996 Water balance components for the TCP catchments in 1997
3000
2000
2000 Rainfall (mm) Discharge (mm) E vapo ration (mm) 1000 1000 Rainfall (mm) Discharge (mm) Evaporation (mm) 0 0 Biboo - Minw o N yangong S ongkw é Biboo - Minw o N yangong
Water balance components for the TCP catchments in Water balance components for the TCP catchments in 1998 1999
3000 3000
2000 2000 Rainfall Rainfall Discharge Discharge Evaporation Evaporation 1000 1000
0 0 Biboo - Minw o N yangong Biboo - Minw o N yangong
Water balance components for the TCP catchments in 2000
3000
2000 Rainfall Discharge Evaporation 1000
0 Biboo - Minw o N yangong
Figure 4.10: Water balance components for the TCP catchments (1996 – 2000)
4.6. CATCHMENT SEDIMENT YIELD
In the framework of the Lu1 project, preliminary sediment rating curves were developed for each of the three basins, relating the sediment concentration, measured at the outlet of the basin (C, in kg m-3), to the corresponding discharge (Q, in m3 d-1). Since then and with the additional data recorded within the Ecol2 project, new rating curves have been established using linear regression analyses on the data presented in Waterloo et al., 1997; 2000. The actual rating curves take in account the slight changes of the former discharge rating curves (Figure 4.6). The sediment concentration (C, in kg m-3) and discharge (Q, in m3 s-1) data collection extended for the Songkwé catchment from February 9, 1996 until December 12, 1996 and from February 8, 1996 until April 21, 2001 for the Biboo – Minwo and Nyangong catchments.
31 B iboo river, M inwo (08/02/96 - 21/04/01)
) 10,000 -3
1,000
0,100
0,010
0,001
Sediment concentration (kg.m 0,000 0,010 0,100 1,000 10,000 100,000
C = 0,0595Q 0,8435 Discharge (m3.s-1) R 2 = 0,6964
N yangong river, Nyangong (08/02/96 - 21/04/01)
10,000 ) -3
1,000
0,100
0,010 Sediment concentration (kg.m 0,001 0,001 0,010 0,100 1,000 10,000 100,000 C = 0,0494Q 0,4345 -1 R 2 = 0,2928 Discharge (m3.s )
Songkwé river, Adjap (09/02/96 - 12/12/96)
1,000 ) -3
0,100
0,010
0,001
Sediment concentration (kg.m 0,000 0,010 0,100 1,000
-1 C = 0,0328Q 0,7446 Discharge (m3.s ) R 2 = 0,1124
Figure 4.11: Sediment rating curves for the Biboo – Minwo, Nyangong and Songkwé catchments.
32 The new sediment rating equations are presented below, whereas the rating curves are shown in Figure 4.11.
Songkwé river (Q = 0.027-0.303 m3 s-1) C = 0.0328*Q0.7446 r2 = 0.11, n = 49 Biboo river (Q= 0.019-9.268 m3 s-1): C = 0.0595 * Q0.8435 r2 = 0.70, n = 389 Nyangong river (Q= 0.012-7.103 m3 s-1) C = 0.0494 * Q0.4345 r2 = 0.29, n = 257
The sediment load was calculated by multiplying the discharge with the corresponding sediment concentration and the total sediment yield was then determined for the length of the Lu1 and Ecol2 projects research period using the equations above. Sediment concentrations showed an exponential increase with discharge at all sites. The observed ranges as well as the sediment yield are given in Table 4.2.
Table 4.2: Sediment concentrations and yields for the three study catchments during the period 09/02/96 until 21/04/01 Songkwé Biboo - Minwo Nyangong Year 1996 Concentration Range (mg l-1) 1 – 150 Sediment yield (kg ha-1 yr-1) 31 552 702 Year 1997 Sediment yield (kg ha-1 yr-1 342 751 Year 1998 Sediment yield (kg ha-1 yr-1 576 874 Year 1999 Sediment yield (kg ha-1 yr-1 637 946 Year 2000 Sediment yield (kg ha-1 yr-1 445 943 01/01/01 - 01/04/01 Sediment yield (kg ha-1 yr-1) 40 157
The highest sediment concentration (908 mg/l) was measured at the outlet of Biboo – Minwo in October 2000; this value may be attributed to the highest daily rainfall recorded that year in that part of the TCP area. Nevertheless, high sediment concentrations were also measured at Nyangong, which is under agricultural activities along the main river channel, as well as due to fishing activities in the river for which small earthen dams are constructed in the main channel during dry period. In addition, some of the sediment may have been derived from natural erosion of the steep slopes under the undisturbed rain forest vegetation.
The values indicate that the overall sediment yield was highest in the Nyangong catchment, which may be attributed to erosion of agricultural land (Nounamo and Yemefack, 2000). The sediment yield in the Biboo - Minwo catchment was much lower over the whole study period, in spite of the fact that the eastern part of the area had been logged in May 1995, whereas the western part was logged during the Lu1 study period (February - March, 1996) and in spite of the high sediment concentrations recorded during periods of high rainfall. The values for Biboo – Minwo decreased in time as the selectively logged forest recovered from the logging impact.
Very low sediment yields were calculated for the Songkwé catchment, which was in line with the absence of any human activity in the area during the data collection period. Small-scale studies on catchment sediment yields have not been carried out earlier in the rain forests of south Cameroon. However, sediment yields from the large Ngoko River drainage system (67,000 km2) have been determined between 1988 and 1992 at Moloundou, southeast Cameroon. There, the sediment yield amounted to 86 kg ha-1 in 1989-1990. In 1992, however, the yield had increased to 140 kg ha-1, which was attributed to a 30% increase in logging activities and associated road construction, as well as to agricultural activity in the region (Seyler et al., 1993; ORSTOM, 1995).
Although the sediment yields from Biboo - Minwo and Nyangong catchments were influenced by logging and shifting cultivation, the values remained very low in comparison to those observed elsewhere in the humid tropics. Published values of surface erosion and sediment yield range from 30 to 6200 kg ha-1 yr-1 for natural forests in tectonically stable areas (Douglas, 1967;
33 Wiersum, 1984) to up to 40,000 kg ha-1 yr-1 for forests in tectonically active zones during wet years (Dickinson et al., 1990).
4.7. HYDROLOGICAL MODELLING3
Discharge, sediment yield and erosion in the selectively logged Biboo - Minwo catchment have been simulated using a 48.1 mm storm (catchment average), which occurred in the afternoon of March 23, 1996. The simulations were made for two scenarios, i.e. for undisturbed forest and for selectively logged forest. The results were compared to the measured hydrograph and sediment yield. Detailed soil and vegetation maps were not available for this area and the model input was therefore estimated from the few and incomplete soil and vegetation data presented in the previous sections. Furthermore, the model has not been calibrated sufficiently and the results of these test-runs should therefore not be considered as final.
A digital elevation model (DEM) of the area was developed using topographical information obtained from the 1:100,000 map. This map is not very detailed as contour intervals are at a spacing of 40 m. Aerial photographs were therefore used to provide additional information on the topography. The DEM was developed with a statistical interpolation program (Kriging interpolation method). The result is shown in Figure 4.12.
Figure 4.12: Digital elevation model of the Biboo – Minwo catchment area. Source Waterloo et al.,2000.
A list of input maps, soil tables and the various settings used in the simulations are presented in Appendix 7. The SWATRE sub-model of LISEM was used to describe the infiltration process. This sub-model allows for the simulation of the effects of roads and skid tracks (Figure 4.13) on the hydrology, erosion and sediment yield of the catchment area.
3 As no further work has been done in the hydrological modelling, this section is an integral part of two scientific reports issued in the framework of Lu1 project by Waterloo et al. (1997, 2000). 34 5DLQIDOOLQWHQVLWLHVGXULQJWKHVLPXODWLRQSHULRGRQ0DUFKDUHVKRZQLQ)LJXUH5DLQIDOO VWDUWHG DW K W PLQXWHV DQG HQGHG DW K W PLQXWHV 7KH PHDVXUHG UXQRIIIRUWKHPPVWRUPZDVPPRYHUWKH PLQXWHVLPXODWLRQSHULRG7KHPHDVXUHG SHDNIORZDPRXQWHGWRPVDQGWKHWLPHWRSHDNZDV PLQXWHV7KHVHGLPHQW\LHOG FDOFXODWHG IURP WKH GLVFKDUJH XVLQJ WKH DSSURSULDWH VHGLPHQW UDWLQJ FXUYH 6HFWLRQ LQ :DWHUORRHWDO DQG DPRXQWHGWRNJRUNJKD7KHPD[LPXPVHGLPHQW FRQFHQWUDWLRQZDVJO7KHPHDVXUHGK\GURJUDSKDQGVHGLPHQWFRQFHQWUDWLRQVDUHVKRZQ LQ)LJXUH
7KH VLPXODWHG ZDWHU \LHOGIRU WKH VHOHFWLYHO\ ORJJHG VFHQDULR DPRXQWHG WR PP ZKLFK LV VRPHZKDWORZHUWKDQWKHPHDVXUHGZDWHU\LHOG7KHUHVXOWVRIWKHVLPXODWLRQVIRUWKHIRUHVWHG DQG VHOHFWLYHO\ ORJJHG VFHQDULRV DUH VKRZQ LQ )LJXUH 7KH VKDSH RI WKH VLPXODWHG K\GURJUDSKIRUWKHVHOHFWLYHO\ORJJHGVFHQDULRZKLFKVKRZVDSHDNDWPVDQGKDVDPXFK VKRUWHU WLPHWRSHDN RI PLQXWHV ZDV GLVWLQFWO\ GLIIHUHQW IURP WKDW RI WKH PHDVXUHG K\GURJUDSK IRU WKH VHOHFWLYHO\ ORJJHG IRUHVW )LJXUH ZKLFK ZDV OHVV SURQRXQFHG )XUWKHUPRUHWKHVLPXODWHGVHGLPHQW\LHOGZDVPXFKWRRKLJKDWNJRU NJKDZKLFK PD\RQO\SDUWO\EHGXHWRHUURUVLQWKHPRGHOOLQJRIWKHUXQRII SHDNIORZPXFKWRRKLJK 7KH SHDNVHGLPHQWFRQFHQWUDWLRQVLPXODWHGE\/,6(0DPRXQWHGWRJOZKLFKZDVWLPHV KLJKHUWKDQWKHDFWXDOSHDNVHGLPHQWFRQFHQWUDWLRQ
)LJXUH 0DSVKRZLQJWKHORFDWLRQDQGZLGWKRIURDGVDQGVNLGWUDFNVLQWKH%LERR±0LQZRFDWFKPHQW 6RXUFH:DWHUORRHWDO
)LJXUH $YHUDJHUDLQIDOOLQWHQVLWLHVGXULQJWKHVLPXODWLRQSHULRGRQ0DUFK LQWKH%LERR±0LQZRFDWFKPHQW 6RXUFH:DWHUORRHWDO
)LJXUH 0HDVXUHGK\GURJUDSKDQGVHGLPHQWFRQFHQWUDWLRQVIRUVHOHFWLYHO\ORJJHG%LERR±0LQZRFDWFKPHQWGXULQJD PPUDLQIDOOHYHQWRQ0DUFK
)LJXUH 6LPXODWHG K\GURJUDSKV DQG VHGLPHQW FRQFHQWUDWLRQV IRU WKH IRUHVWHG DQG VHOHFWLYHO\ ORJJHG %LERR ± 0LQZR FDWFKPHQWGXULQJDPPUDLQIDOOHYHQWRQ0DUFK 6RXUFH:DWHUORRHWDO
7KHVLPXODWHGZDWHU\LHOGIRUWKHIRUHVWHGFDWFKPHQWVFHQDULRZDVPPZKLFKZDVPXFK ORZHU WKDQ WKDW SUHGLFWHG IRU WKH VHOHFWLYHO\ ORJJHG IRUHVW VFHQDULR GXH WR D KLJKHU WRWDO LQILOWUDWLRQ PPYVPP DQGDVOLJKWO\KLJKHULQWHUFHSWLRQORVV PPYVPP 7KHSHDNGLVFKDUJHZDVPXFKORZHUDWPVEXWWKHWLPHWRSHDNUHPDLQHGWKHVDPHDW PLQXWHV7KHSUHGLFWHGVHGLPHQW\LHOGZDVPXFKORZHUWKDQWKDWSUHGLFWHGIRUWKHVHOHFWLYHO\ ORJJHGVFHQDULRDW NJRUNJKD
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
)LJXUH (URVLRQPDSRIWKHVHOHFWLYHO\ORJJHG%LERR±0LQZRFDWFKPHQWIRUDPPUDLQIDOOHYHQWRQ0DUFK 6RXUFH:DWHUORRHWDO
)LJXUH (URVLRQPDSRIWKHIRUHVWHG%LERR±0LQZRFDWFKPHQWIRUDPPUDLQIDOOHYHQWRQ0DUFK 6RXUFH:DWHUORRHWDO
)LJXUH 0DSRI%LERR±0LQZRFDWFKPHQWVKRZLQJFHOOVZKHUHWKHHURVLRQLQFUHDVHGE\PRUHWKDQNJKDDIWHUWKH FRQVWUXFWLRQRIVNLGWUDFN 6RXUFH:DWHUORRHWDO
0(7(252/2*<
$ PLFURPHWHRURORJLFDO VHWXS UHFRUGLQJ DLU WHPSHUDWXUH UHODWLYH KXPLGLW\ JOREDO DQG QHW UDGLDWLRQZLQGGLUHFWLRQDQGVRLOWHPSHUDWXUHVKDVEHHQIXQFWLRQLQJIURP-XO\XQWLO1RYHPEHU LQ WKH 6RQJNZp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
7DEOH 6XPPDU\ RI PLFURPHWHRURORJLFDO GDWD FROOHFWHG DERYH D VHFRQGDU\ UDLQ IRUHVW QHDU $GMDS EHWZHHQ $XJXVW DQG 1RYHPEHU
7KHYDOXHVTXRWHGIRUWKHJOREDODQGQHWUDGLDWLRQ 5JDQG5Q DUHDYHUDJHGDLO\WRWDOV6RXUFH:DWHUORRHWDO $XJXVW 6HSWHPEHU 2FWREHU 1RYHPEHU 7DYJ (& 7PLQ (& 7PD[ (& 5+DYJ 5J 0-P G 5Q 0-P G
$IUHTXHQF\GLVWULEXWLRQRIWKHZLQGGLUHFWLRQRYHUWKHIRXUPRQWKVLVVKRZQLQ)LJXUH 7KHZLQGGLUHFWLRQZDVSUHGRPLQDQWO\VRXWKZHVWZLWKRIDOOKDOIKRXUO\ZLQGGLUHFWLRQ UHDGLQJV Q IDOOLQJZLWKLQWKHUDQJHRI1(DQGIDOOLQJZLWKLQWKHUDQJH 1($PXFKVPDOOHUSHDNZDVREVHUYHGIRUWKHUDQJHRI1 (ZKLFKFRQVLVWHG RIRIDOOUHDGLQJV 7RDOORZIRUWKHFDOFXODWLRQRIQHWUDGLDWLRQIURPPHDVXUHGLQFRPLQJVKRUWZDYHUDGLDWLRQIRU IRUHVWHGDUHDVHOVHZKHUHLQVRXWK&DPHURRQOLQHDUUHJUHVVLRQDQDO\VLVZDVXVHGWRHVWDEOLVKD
UHODWLRQVKLSEHWZHHQKDOIKRXUO\YDOXHVRIWKHJOREDO 5J DQGQHWUDGLDWLRQ 5Q 7KHIROORZLQJ HTXDWLRQZDVIRXQG
5Q B B x 5JU Q
7KLVHTXDWLRQLVVLPLODUWRWKDWREVHUYHGIRUWURSLFDOUDLQIRUHVWLQ$PD]RQLD %DVWDEOH HW DO 7KHSDUWLWLRQLQJRIWKHDYDLODEOHVRODUHQHUJ\ LHQHWUDGLDWLRQ RYHUWKHVHQVLEOH + DQGODWHQW /( KHDWIOX[HVFDQEHGHWHUPLQHGIURPWKHQHWUDGLDWLRQDQGWKHUPRFRXSOHWHPSHUDWXUHDQG VWDQGDUG GHYLDWLRQ GDWD XVLQJ WKH WHPSHUDWXUH IOXFWXDWLRQ HQHUJ\ EDODQFH PHWKRG 7LOOPDQ GH %UXLQ HW DO :DWHUORR 8QGHU QRQDGYHFWLYH FRQGLWLRQV WKH HQHUJ\ EDODQFHRIDIRUHVWVXUIDFHPD\EHZULWWHQDV
5Q -* 3YHJ O(+
$VQRIXUWKHUGDWDKDVEHHQFROOHFWHGEH\RQG'HFHPEHU LQWKHIUDPHZRUNRIWKHPLFURPHWHRURORJLFDOGDWD FROOHFWLRQWKLVVHFWLRQLVDQLQWHJUDOSDUWRIWZRVFLHQWLILFUHSRUWVLVVXHGLQWKHIUDPHZRUNRI/XSURMHFWE\:DWHUORR HWDO ZKHUH 5Q LV WKH QHW UDGLDWLRQ * WKH VRLO KHDW IOX[ - DQG 3YHJ WKH SK\VLFDO DQG ELRFKHPLFDO VWRUDJH RI HQHUJ\ LQ WKH YHJHWDWLRQ + WKH VHQVLEOH KHDW IOX[ DQG O( WKH ODWHQW KHDW IOX[ DOO IOX[HVLQ:P 7XUEXOHQWPRWLRQVRIWKHDLUDUHUHVSRQVLEOHIRUPRVWRIWKHYHUWLFDOWUDQVIHURI KHDW 7KRP 7KHVWDQGDUGGHYLDWLRQRIKLJKIUHTXHQF\WHPSHUDWXUHPHDVXUHPHQWVPD\ VHUYH DV D PHDVXUH RI WKH LQWHQVLW\ RI WHPSHUDWXUH IOXFWXDWLRQV FDXVHG E\ WXUEXOHQFH DQG 7LOOPDQ REWDLQHG WKH IROORZLQJ HTXDWLRQ UHODWLQJ WKH WHPSHUDWXUH 7 DQG LWV VWDQGDUG
GHYLDWLRQ V7 WRWKHODWHQWKHDWIOX[+
J + U FK ] G V V S 7 7
,QWKLVHTXDWLRQWKHFRQVWDQWKVLVHTXDOWR :LMQJDDUGDQG&RWH ULVWKHGHQVLW\RI DLU NJP FSWKHVSHFLILFKHDWRIDLUDWFRQVWDQWSUHVVXUH -NJ . ]WKHWKHUPRFRXSOHKHLJKW P GWKHGLVSODFHPHQWOHQJWK P DQGJWKHDFFHOHUDWLRQGXHWRJUDYLW\ PV 7KHVRLOKHDW IOX[DQGWKHSK\VLFDODQGELRFKHPLFDOVWRUDJHRIHQHUJ\DUHXVXDOO\PXFKVPDOOHUWKDQWKHRWKHU WHUPVLQWKHHTXDWLRQDQGPD\WKHUHIRUHVDIHO\EHQHJOHFWHG7KHVHQVLEOHKHDWIOX[FDQWKHQEH HYDOXDWHGE\LQVHUWLQJWKHYDOXHRI+FDOFXODWHGZLWKWKHWHPSHUDWXUHIOXFWXDWLRQPHWKRGLQWKH VLPSOLILHGHQHUJ\EDODQFHHTXDWLRQ
$VDQH[DPSOHWKHSDUWLWLRQLQJRIWKHDYDLODEOHVRODUHQHUJ\ QHWUDGLDWLRQ RYHUWKH ODWHQWDQGVHQVLEOHKHDWIOX[HVRQ$XJXVW -XOLDQGD\ LVVKRZQLQ)LJXUH 7KHGD\WLPHHYDSRUDWLRQRQWKLVGD\DPRXQWHGWR PPZKLFKLVIDLUO\FORVHWRWKH HYDSRUDWLRQUDWHFDOFXODWHGZLWKWKHZDWHUEDODQFHPHWKRGLQ6HFWLRQIRUWKH 6RQJNZpFDWFKPHQW PPGLQ:DWHUORRHWDO DQG
)LJXUH 0RQWKO\IUHTXHQF\GLVWULEXWLRQVRIWKHZLQGGLUHFWLRQDERYHVHFRQGDU\IRUHVWDW$GMDS 6RXUFH:DWHUORRHWDO
)LJXUH 3DUWLWLRQLQJRIDYDLODEOHHQHUJ\RYHUWKHODWHQWDQGVHQVLEOHKHDWIOX[HVRQ$XJXVW DW$GMDS 6RXUFH:DWHUORRHWDO
5. NUTRIENT CYCLING IN THE HYDROLOGICAL CYCLE
5.1. INTRODUCTION
The functional aspects of the evergreen forest in Southern Cameroon were studied in the Ecol2 project. The general objective of this project was to quantify the key ecological processes, such as primary productivity, phytomass accumulation and nutrient cycling and the impact of human disturbances (e.g. logging, shifting cultivation, etc.) on these processes.
To preserve some areas while maintaining a sustained development in others, a deep knowledge of rain forest structure and functioning is necessary. The study of the biogeochemical cycle is one among many scientific branches that has had wide but insufficient attention due to the complexity of the ecosystem.
Many authors have described the nutrient cycle in the tropical forest ecosystem and the literature on the subject is widely known. In a humid tropical forest, the vegetation gets its nutrient supply principally from decomposing organic matter, air and soil minerals. The expected nutrient supply from the soil is in general very low due to the strongly weathered status of the soil, as it is the case of the soils of the study area with a low nutrient content (van Gemerden and Hazeu, 1999).
Two nutrient cycles are the main supply of nutrients to the ecosystem. One is in the biological cycle and is the focus of the vegetation component of the Ecol2 project (see Ibrahima et al., 2002). The other one is the hydrological cycle. The nutrient cycle in the hydrological cycle is considered by some authors (Jordan, 1982; Bruijnzeel, 1990; Forti and Neal, 1992; Lesak, 1993; van Dam, 2001) as a major factor in nutrient cycling in a humid tropical forest ecosystem. In this chapter some important compartments of this cycle will be studied: the quantity and the quality (ionic content) of the rainfall and the throughfall that crosses the aerial phytomass, of the percolate through the litter layer (litter percolate) and of the stream flow (river flow). The water movement in the root zone, draining to the stream flow (river flow) is not studied.
5.2. SAMPLING AND ANALYSIS
5.2.1. Methodology on water sampling and chemical analyses As discussed in Chapter 3 on methodology, the sampling in the compartments of the hydrological cycle studied, i.e. rainfall, throughfall, litter percolate and river flow was performed weekly. The sampling took place from May 12, 2000 until April 11, 2001. A total of 169 water samples were collected and analysed, but only the analytical results of 158 of them have been retained for the study, their division over the different compartments is given in Table 5.1.
Table 5.1: Repartition of water samples over the sampling period in the various compartments Sampling period Total number Number of Number of Number of litter Number of river of samples rainfall samples throughfall percolates samples flow samples samples 12/05/00 – 08/06/00 18 0 10 6 2 08/06/00 – 29/06/00 9 0 3 3 3 29/06/00 – 27/07/00 9 0 2 3 4 27/07/00 – 31/08/00 14 2 4 3 5 31/08/00 – 26/10/00 16 4 4 4 4 26/10/00 – 30/11/00 14 3 3 3 5 30/11/00 – 04/01/01 5 0 0 1 4 04/01/01 –06/02/01 4 0 0 0 4 06/02/01 – 01/03/01 12 1 2 2 7 01/03/01 – 29/03/01 52 9 9 9 25 29/03/01 – 11/04/01 5 0 1 1 3 Total 158 19 38 35 66 43 After each weekly sampling process the quantities of water (in mm) collected in each compartment were temporarily stored in the dark at the hydrological camp in four separate 25 l plastics reservoirs which had been thoroughly rinsed with de-ionised water prior to sampling. No additive “biocides” were used for the preservation of the samples, only manipulative “biocides” such as dark storage and refrigeration (see Galloway and Liken, 1978). The storage period of the water samples at the hydrological camp was a function of the logistic constraints (availability of a car, accessibility to the study site, etc). Two sub-samples of 500 ml each (depending of the quantity of water collected in the compartment) were taken to the TCP laboratory in Kribi where they were processed and stored at 4 °C. Only one sub-sample was sent to the laboratory for chemical analyses, the remaining sample being kept in the TCP laboratory for an eventual control in another laboratory. A total number of 169 water samples have been analysed in the Laboratory of Soil and Environment of the Faculty of Agronomy of the University of Dschang (Cameroon). For each water sample, the following physico-chemical 2+ 2+ + + - characteristics were analysed: EC, pH, Ca , Mg , K , Na , organic C, total N, total P, NO3 and Cl-. Due to the malfunctioning of the Portable pH/Con 10 meter (OATON 35630-00) and the HCO3 test kit, even EC and pH were determined at University of Dschang after storage. HCO3 (alkalinity) couldn’t be determined. Chemical analyses of water were performed after Rodier (1982) and Pauwels et al. (1992). Total Nitrogen was determined by the distillation method of Kjeldahl with mixture of strong acids. Nitrates in solution were determined by the steam distillation with devarda alloy following Kjeldahl. Organic Carbon was determined with a T.O.C (total organic carbon) apparatus after digestion with hydrogen peroxide (for very low C content samples) and with potassium dichromate oxidation for samples with a higher C content. The major cations in solution, Ca2+, Mg2+, K+ and Na+ were analysed after digestion with concentrated perchloric acid and measured by atomic absorption spectrophotometer. No verification of the results of chemical analyses was made in another laboratory.
5.2.2. Results of the analyses For the whole sampling period (from May 12, 2000 to April 11, 2001) and for the samples analysed, the analytical results showed a great variation of the electrical conductivity from 4.9 to 296 µS.cm-1. These values confirm the low solutes content observed in the research area (van Gemerden and Hazeu, 1999) and in others rainforest ecosystems in Cameroon (Ndam Ngoupayou, 1997; Sigha-Nkamdjou, 1993). The pH values were low (3.3) for most samples, while a few others had higher pH values (6 to 7.9). Similar pH values were found by Ndam Ngoupayou (1997) and Sigha-Nkamdjou (1993) in others Cameroon rainforests especially for rainfall and throughfall. The evolution of the solutes content is linked to the season. Low values were observed with the decreasing of the rainy season. In the cases, the ionic content of the litter percolate is higher than of the other compartments; the solute content status of the macro- nutriment is as follows: litter percolate (LP) > throughfall (TF) > river flow (RvF) > rainfall (RF). The results of the analyses of the samples retained for the study are shown in the Table 5.2.
Table 5.2 gives for each compartment, the sampling period of reference and the corresponding mean values of EC, pH, and concentrations of Ca2+,Mg2+, K+, Na+, organic C, total N, total P, - - NO3 and Cl . Those values are also illustrated in the Figures 5.1, 5.2 (a) and 5.2 (b).
44 -1 -1 2+ 2+ + + - - Table 5.2: Mean electric conductivity (EC in µS.cm ) and concentrations (in mgl ) of Ca , Mg , K , Na , organic C, total N, total P, NO3 and Cl in rainfall (RF), throughfall (TF), litter percolate (LP) and in river flow (RvF) from May 12, 2000 to April 11, 2001 in the Biboo - Minwo Catchment and the one-ha Ecol2 plot. Element Compart- 12/05/00 – 08/06/00 - 29/06/00 - 27/07/00 - 31/08/00 - 26/10/00 - 30/11/00 - 04/01/01 - 06/02/01 - 01/03/01 - 29/03/01 - ment 08/06/00 29/06/00 27/07/00 31/08/00 26/10/00 30/11/00 04/01/01 06/02/01 01/03/01 29/03/01 11/04/01 RF 9.7 5.3 4.9 10.0 7.2 EC TF 22.8 32.5 36.1 69.8 20.4 21.7 96.0 40.1 36.0 LP 44.8 296 84.7 68.1 44.6 28.3 42.0 197.0 76.7 75.0 RvF 18.7 21.5 21.4 24.1 19.8 20.0 25.3 27.5 28.1 26.6 27.0 RF 5.4 5.0 5.1 5.9 5.5 PH TF 4.8 6.9 5.5 5.5 5.0 5.2 6.3 5.8 5.5 LP 5.1 6.8 6.4 6.1 5.7 5.6 6.0 4.5 6.5 6.6 RvF 5.2 6.2 6.4 6.2 6.2 6.1 6.3 6.7 6.6 6.4 6.8 RF 1.07 0.53 0.33 0.23 0.07 Ca2+ TF 2.39 2.59 0.83 2.06 1.04 0.97 0.79 0.17 0.06 LP 6.47 2.81 3.13 4.21 2.47 2.55 1.68 2.50 0.45 0.28 RvF 1.55 0.77 0.62 0.56 0.47 0.51 0.60 0.68 0.33 0.10 0.03 RF 0.06 0.03 0.10 0.50 0.02 Mg2+ TF 9.03 0.81 0.35 1.44 0.34 0.27 0.36 0.5 0.03 LP 9.40 2.51 2.34 5.42 0.90 0.85 0.48 1.32 0.23 0.39 RvF 8.81 0.49 0.58 0.60 0.44 0.51 0.50 0.58 0.93 0.08 0.10 RF 0.01 0.04 0.01 0.01 0.01 K+ TF 12.60 7.58 8.41 17.13 4.19 4.76 0.24 0.14 0.12 LP 21.87 8.58 8.65 19.89 9.39 3.95 3.66 0.30 0.25 0.23 RvF 4.70 0.08 0.08 0.19 0.29 0.04 0.21 0.17 0.15 0.03 0.03 RF 0.15 0.10 0.14 0.01 0.49 Na+ TF 3.51 2.31 0.34 0.28 0.07 0.05 0.02 0.77 0.04 LP 2.88 19.35 2.59 0.53 0.13 0.06 0.13 0.03 0.72 0.06 RvF 2.09 1.43 0.90 0.90 1.26 1.47 2.76 3.22 0.83 0.69 0.87 RF 0.02 0.01 0.07 0.54 0.30 Organic C TF 0.39 0.22 0.46 0.30 0.19 0.14 0.56 0.33 0.12 LP 1.04 0.59 0.60 0.74 0.34 0.34 0.42 0.22 0.26 0.38 RvF 0.21 0.45 0.40 0.26 0.21 0.20 0.14 0.19 0.25 0.29 0.16 RF 0.32 0.24 0.23 0.34 0.29 Total N TF 0.28 0.28 0.41 0.35 0.26 0.26 0.26 0.27 0.30 LP 0.69 0.29 0.23 0.40 0.27 0.22 0.21 0.24 0.29 0.21 RvF 0.10 0.27 0.23 0.36 0.24 0.26 0.07 0.09 0.29 0.28 0.27 RF 0.46 0.70 1.08 1.24 0.73 Total P TF 0.12 1.08 1.24 0.86 0.84 0.60 0.76 0.87 0.60 LP 0.37 0.97 0.71 0.84 0.88 0.76 0.12 1.03 0.82 0.60 RvF 0.06 0.95 0.88 0.66 0.55 0.75 0.09 0.10 0.55 0.72 1.19 RF 0.10 0.15 0.10 0.18 0.12 - NO3 TF 0.09 0.07 0.12 0.11 0.09 0.17 0.12 0.08 LP 0.16 0.09 0.19 0.12 0.12 0.10 0.11 0.15 RvF 0.11 0.08 0.13 0.08 0.15 0.12 0.10 0.16 RF 0.17 0.12 0.11 0.12 0.17 Cl- TF 0.21 0.12 0.08 0.15 0.10 0.11 0.14 0.10 0.14 LP 0.24 0.17 0.16 0.17 0.15 0.19 0.12 0.13 0.12 0.20 RvF 0.11 0.10 0.10 0.14 0.10 0.15 0.14 0.15 0.14 0.13 0.18
45 RF pH TF LP RvF 8 6 4
pH units 2 0 1234567891011
EC RF TF
400 LP 300 RvF 200
µS.cm-1 100 0 1234567891011
Figure 5.1: Mean electrical conductivity (EC) and mean pH of the water samples in the various compartments of the forest ecosystem in the hydrological cycle in the Biboo – Minwo catchment Standard errors are also shown in the figure. For the sampling period, the different numbers in x-axes correspond to: 1 = 12/05/00 – 08/06/00 2 = 08/06/00 – 29/06/00 3 = 29/06/00 – 27/07/00 4 = 27/07/00 – 31/08/00 5 = 31/08/00 – 26/10/00 6 = 26/10/00 – 30/11/00 7 = 30/11/00 – 04/01/01 8 = 04/01/01 – 06/02/01 9 = 06/02/01 – 01/03/01 10 = 01/03/01 – 29/03/01 11 = 29/03/01 – 11/04/01. RF, TF, LP and RvF correspond respectively to rainfall, throughfall, litter percolate and river flow.
RF RF 2+ 2+ Ca TF Mg TF LP LP RvF RvF 10,0 10,00 8,0 8,00 -1 6,0 -1 6,00 4,0 4,00 mg.l mg.l 2,0 2,00 0,0 0,00 1234567891011 1234567891011
RF RF + + K TF Na TF LP LP RvF RvF 30,00 30 -1
20,00 -1 20 mg.l 10,00 mg.l 10
0,00 0 1234567891011 1234567891011
Figure 5.2a: Mean concentrations (in mg.l-1) of Ca2+, Mg2+, K+ and Na+ within the sampling period in the various compartments of the forest ecosystem in the hydrological cycle in the Biboo – Minwo catchment Standard errors are also shown in the figure. For legends see Figure 5.1
46 RF RF organic C TF total N TF LP LP RvF RvF 1,50 1,00 0,80
-1 1,00 0,60 -1 0,40
mg.l 0,50
mg.l 0,20 0,00 0,00 1234567891011 1234567891011
total P NO3-
2,00 0,25 0,20 1,50 RF RF TF 0,15 TF 1,00 LP LP 0,10 mg.l-1 mg.l-1 RvF RvF 0,50 0,05 0,00 0,00 1234567891011 1234567891011 Sampling period Sampling period
Cl-
0,30 0,25 RF 0,20 TF 0,15 LP
mg.l-1 0,10 RvF 0,05 0,00 1234567891011 Sampling period
- - Figure 5.2b: Mean concentrations of organic C, total N, total P, NO3 and Cl within the sampling period in the various compartments of the forest ecosystem in the hydrological cycle in the Biboo – Minwo catchment Standard errors are also shown in the figure. For legends see Figure 5.1.
5.2.2.1. Rainfall The atmospheric input of chemicals consists of wet and dry deposition (Galloway and Likens, 1978). The dry deposition (solid particles, aerosols) suspended in the atmosphere and on the canopy is rained-out during the wet period. The device used to collect rainfall water for chemical analyses didn’t separate wet deposition from dry deposition, rainfall samples collected were bulked samples. The ionic content of the rainfall water depends also on the chemical sources of the rain water, maritime or terrestrial. Additional,- but to a lesser extent - this ionic content depends too on eventual fire events in the vegetation (Forti and Neal, 1992).
Only 19 rainwater samples distributed over five sampling periods were retained for this study. This low number compared to the samples collected in other compartments is due to two main reasons. Some rainfall water samples with high EC values (more than 80 µS.cm-1) were considered as contaminated (vegetation, animal and human contamination, etc.) and disregarded. This situation occurred during the first sampling periods (12/05/00 – 27/07/00) probably due to the weak ability of the field labour in the sampling process. Another reason could be the spatial heterogeneity of the rainfall events over the Biboo – Minwo catchment. For some rainfall events, low quantities of rainfall water were observed as compared to those
47 collected for throughfall in the one-ha Ecol2 plot during the same rainfall event. The rapid recovery of vegetation in the former landing of the logging company Wijma-Douala SARL where the rain gauge was placed could prevent rainfall reaching the device.
For the analyses of the rainfall water, two sampling periods characterised by a complete sampling in all the compartments (i.e. rainfall, throughfall, litter percolate, river flow) have been considered. The average values (with the corresponding standard errors) of EC, pH and concentrations of chemicals studied are shown in Figure 5.3a for the period 27/07/00 – 30/11/00 and in Figure 5.3b for the period 06/02/01 – 29/03/01. )
-1 70,00 7,00 60,00 6,00 50,00 RF 5,00 RF 40,00 TF 4,00 TF 30,00 LP 3,00 LP 20,00 RvF Acidity 2,00 RvF 10,00 1,00 0,00 0,00 Conductivity (µS.cm EC pH Sampling period (27/07/00 - 30/11/00) Sampling period (27/07/00 - 30/11/00) ) -1 )
12,00 -1 1,00 10,00 RF 0,80 8,00 RF TF 6,00 0,60 TF LP 4,00 LP RvF 0,40 2,00 RvF 0,00 0,20
Concentration (mg.l Ca2+ Mg2+ K+ Na+ organic 0,00 Concentration (mg.l C total N total P NO3- Cl- Sampling period (27/07/00 - 30/11/00) Sampling period (27/07/00 - 30/11/00)
Figure 5.3a: Average values (with the corresponding standard errors) of EC, pH and concentrations of Ca2+, Mg2+, K+, Na+, - - organic C, total N, total P, NO3 and Cl for the period from 27/07/00 to 30/11/00. For legends see Figure 5.1.
The rainfall water of the Biboo – Minwo catchment compared to the water of the others compartments are characterised for both periods by a very low value of EC and solute content, especially for all the cations, Ca2+, Mg2+, K+, and Na+. Those values are well correlated with Organic Carbon, which presence in the rainfall water seems to be linked to the presence of the cations. Potassium in the rainfall in this area is near the detection limit. The solute content for total N, total P, NO3- and Cl- remains constant and at the same level for all the water samples of all compartments. The low cations content in the rainfall water could be related to the probable dominance of terrestrial sources of rainfall (with chemicals produced by biological processes) in the TCP research area.
48 )
-1 120,00 7,00 100,00 6,00 80,00 RF 5,00 RF TF 4,00 TF 60,00 LP 3,00 LP
40,00 Acidity RvF 2,00 RvF 20,00 1,00 0,00 0,00 Conductivity (µS.cm EC pH Sampling period (06/02/01 - 29/03/01) Sampling period (06/02/01 - 29/03/01) ) -1 )
1,20 -1 1,00 1,00 RF 0,80 0,80 RF TF 0,60 0,60 TF LP 0,40 LP RvF 0,40 0,20 RvF 0,00 0,20
Concentration (mg.l Ca2+ Mg2+ K+ Na+ organic 0,00 Concentration (mg.l C total N total P NO3- Cl- Sampling period (06/02/01 - 29/03/01) Sampling period (06/02/01 - 29/03/01)
Figure 5.3b: Average values (with the corresponding standard errors) of EC, pH and concentrations of Ca2+, Mg2+, K+, Na+, organic C, total N, total P, NO3- and Cl- for the period from 06/02/01 to 29/03/01. For legends see Figure 5.1.
5.2.2.2. Throughfall Throughfall (and stemflow) provide important fluxes for the internal transport and dynamics of nutrients in tropical rainforest. Three processes govern the chemistry of throughfall: concentration due to the evaporation from the wet canopy, washout of the dry deposition over the vegetation, leaching of the nutrients from the vegetation (Forti, 1989 cited in Forti and Neal, 1992). Figures 5.2a and 5.2b show a relative increase of EC compared to the rainfall water. The mean values of the solute content of the throughfall for the major elements, Ca2+, Mg2+, K+, and Na+ have also increased. As observed in others studies (Bruijnzeel, 1990; Forti and Neal, 1992; Stoorvogel, 1993; Waterloo, 1994), throughfall is enriched with the macro-nutriments after the transfer of the rainfall from the canopy to this compartment (washout of dust, vegetal and animal debris, decomposition, etc.). The largest increase with respect to atmospheric inputs was observed for K. This has been reported in the literature by Parker, 1983; Forti and Neal, 1992 and Burghouts, 1993 (cited in Waterloo, 1994). As for the other compartments, the solute - - content for total N, total P, NO3 and Cl remains constant and at the same – low - level for both periods.
5.2.2.3. Litter percolate The composition of the litter aboveground (leaves, twigs, small branches, fruits, etc.) and the decomposition process influence the composition of the litter percolate water. All the values of the chemicals with those of EC and pH are higher than the ones measured in the others - compartments except of total Nitrogen, total Phosphorus and NO3 . The higher values of EC indicate higher solute content of the major elements (Ca2+, Mg2+, K+, Na+). The leaves in the litter layer enrich this compartment in K+. The litter percolate compartment may contribute much to the nutrient status of the forest ecosystem. Litter percolates are brown and have higher cations and organic carbon contents than the others compartments. This status indicates clearly that a large amount of the nutrients in the Biboo – Minwo forest ecosystem available for the vegetation is stored in the litter layer. Van Gemerden and Hazeu (1999) while studying the TCP soil nutrient content concluded that “a large part of the nutrients is stored in the biomass of the
49 vegetation”. This compartment is the main store of the nutrient of the forest ecosystem in the study area. No other study of ionic content of the litter percolates has been made before in Cameroon rainforests, the comparison with other similar sites is therefore very difficult.
5.2.2.4. River flow The nutrient losses from the rainforest ecosystem are primarily associated with drainage waters although these losses may be minimal due to mechanisms of nutrient conservation. This minimisation of losses is well illustrated in this study. For the two sampling periods for which the chemical analyses are available for the all the compartments, the losses from the catchment to drainage through the river flow represent only 28% of the total atmospheric inputs 2+ 2+ + + - of Ca , Mg , K , and Na . For organic C, total N, total P, NO3- and Cl , the output values are the same as the rainfall input. This kind of nutrient cycle has been described in other studies as a closed cycle (Turkey, 1974 in Forti and Neal, 1992).
The low level of solutes in the surface water results from one of three conditions, or from a combination of those: • Insufficient nutrients are available in the soil to allow significant transfer to the river or to the plants; • The residence time of nutrients in the system is short compared with the time taken for mineral equilibrium to occur; weathering and ionic exchange are thus insignificant; • The residence time of nutrients in the system is long because the turnover rate is slow: the replenishment rate is small.
The low values of EC, pH and chemicals observed for the water river samples have been recorded in other streams studied in rainforest ecosystems in Cameroon (Sigha-Nkamdjou, 1993; Ndam Ngoupayou, 1997). The same EC values were recorded at the same site through the Lu1 project.
For the Biboo – Minwo rainforest ecosystem, the nutrient output through the river water is low. The retention of ions within the catchment is effective and well demonstrated by the low surface water ionic content compared with the total amount reaching the soil through throughfall and litter percolate. A similar catchment behaviour has been observed in one of the Amazonian rain forest studies and was described as a conservative mechanism by Franken et al. (1985) and Forti (1989, cited in Forti and Neal, 1992).
5.3. NUTRIENT BALANCE IN THE HYDROLOGICAL CYCLE
5.3.1. Introduction The nutrient balance (input and output) depends strongly on amounts of rainfall entering and stream flow leaving the ecosystem (Likens et al., 1977). To have a representative nutrient budget, it is important that the sampling period is comparable to the long-term hydrological cycle in the area in term of rainfall amounts, its division over the compartments, and drainage. That has not been the case for this study. The following nutrient budget is only an estimation made for that specific sampling period (May 12, 2000 – April 11, 2001). Table 5.3 and Figure 5.4 give for each compartment (i.e. rainfall, throughfall, litter percolate, river flow) and for each sampling period (11 sampling periods) the amount of water collected in mm. Also in Table 5.3, the rainfall amounts in the Biboo – Minwo catchment (P_Minwo) are given. Values of bulk precipitation entering the system are corrected values (RF_corr) (see Section 5.4.1 below). A relation has been established between the litter percolate and throughfall amounts over the study period (see Figure 5.5). Table 5.4 shows the nutrients inputs in kg.ha-1 for each compartment and for each sampling period.
50 Table 5.3: Amount in mm of rainfall (P_Minwo, RF_corr), throughfall (TF), litter percolate (LP) and river flow (RvF) in the Biboo - Minwo catchment from May 12, 2000 to April 11, 2001. Values of RF_corr correspond to rainfall corrected values, see Section 5.3.2.
Sampling period P_Minwo RF_corr TF LP RvF 1. 12/05/00 – 08/06/00 313.2 489.4 219.2 34.9 41.7 2. 08/06/00 – 29/06/00 61.4 188.0 84.2 9.0 27.6 3. 29/06/00 – 27/07/00 47.8 64.0 28.6 3.1 20.1 4. 27/07/00 – 31/08/00 129.0 68.9 30.8 1.5 16.8 5. 31/08/00 – 26/10/00 742.4 1341.5 601.1 110.2 274.4 6. 26/10/00 – 30/11/00 288.6 247.6 110.9 28.0 107.7 7. 30/11/00 – 04/11/00 0.4 0.9 0.4 1.7 35.2 8. 04/11/00 – 06/02/01 27.6 0.0 0.0 0.0 14.9 9. 06/02/01 – 01/03/01 74.6 122.1 54.7 14.5 7.8 10. 01/03/01 – 29/03/01 210.2 222.1 99.5 28.4 14.5 11. 29/03/01 – 11/04/01 100.8 76.5 34.2 8.8 8.1 Total amount of water 1996.0 2821.0 1263.6 239.9 568.8 m 1600
1400
1200
1000
800
600
400 RF_corr
Throughfall
200
Litterpercolate
River flow 0 Rainfall, Throughfall, Litterpercolate, and River flow in m
1. 12/05/002. - 08/06/003. - 29/06/004. - 27/07/005. - 31/08/006. - 26/10/007. - 30/11/008. - 04/01/019. - 06/02/01 - 01/03/01 10. 02/03/0111. - 30/03/0129/03/01 - 11/04/01 Sampling period
Figure 5.4: Amount of rainfall, throughfall, litter percolate and river flow (in mm) recorded at the Biboo - Minwo catchment and the one- ha Ecol2 plot during the water sampling period for the nutrient cycling study (12/05/00 – 11/04/01). The rainfall (RF_corr) values are estimated values obtained by the regression equation between throughfall and rainfall after intensive measurements in the same site in 1996 (Ruppert, 1996, Waterloo et al., 1997). For legends see Figure 5.1
51 m 120,0 100,0 80,0 60,0 40,0 20,0 0,0 Litter percolate in m 0,0 100,0 200,0 300,0 400,0 500,0 600,0 700,0 Throughfall in mm
Figure 5.5: Relation between litter percolate (LP) and throughfall (TF) from May 12, 2000 until April 11, 2001 in the one-ha Ecol2 plot. Regression: LP = 0.1811TF +1.0119 (R2 = 0.973; n = 11).
5.3.2. Nutrient input by rainfall The rainfall nutrient inputs for each sampling period were calculated by multiplying the corresponding rainfall amounts (Table 5.3) by the mean concentrations of chemicals given in Table 5.2. As discussed in Section 5.2.2.1, the amounts of rainfall recorded at the chemistry rain gauge were not satisfactory due to the reasons given. We observed for some rainfall events, higher quantities of throughfall in the one-ha Ecol2 plot than those recorded for the same event in the rain gauge for rainfall inputs. A study conducted by the Lu1 hydrology team from 11 April until 9 July 1996 on the rainfall interception (36 measurements) in the actual site of the Biboo - Minwo forest lead to a regression equation between the throughfall (TF) and the rainfall (RF) (Waterloo et al., 1997). For the calculation of the amount of rainfall, we used that regression equation which reads: for the rain forest Biboo – Minwo, TF = 0.64 + 0.667*RF.
The nutrient inputs in rainfall (Table 5.4) for the major elements are in general higher than the outputs in the river flow. In particular for Ca2+, the atmospheric input is the main source for the ecosystem.
5.3.3. Nutrient output in river flow Sampling river water is more easy than sampling rainfall water; the river water samples for our ecosystem were available all over the sampling period and processed with a minimum of contamination even during storms. This explains the complete sampling of river water all over the sampling period. A total of 66 samples have been collected. Table 5.4 gives the amount of nutrient outflow (RvF). The amount of nutrients in the river flow for a given sampling period was obtained by multiplying the corresponding discharge by the concentrations of elements. The nutrients outflow in kg.ha-1 from the Biboo – Minwo catchment and for the sampling period (12 May 2000 until 11 April 2001) are given in Table 5.5. .
5.3.4. Nutrient fluxes in throughfall and in litter percolates The ion concentrations increased for the relatively mobile ions (K, Mg and Ca), particularly K, which is linked with the vegetal cycle. As a result of dry deposition on the forest floor, evaporation of water from the forest floor, and - most important - leaching of constituents from decomposing litter, throughfall and stemflow become enriched upon reaching the forest floor and while percolating through the litter layer. The flows of nutrients in both compartments constitute the main store of nutrients. In comparison with the values observed for the stream outflow, the values recorded for both compartments combined are given in Table 5.5.
52 Table 5.4: Nutrient balance and Flows in kg.ha-1 for the major elements studied in the Biboo – Minwo catchment. Rainfall (RF), throughfall (TF), Litter percolate (LP) and River flow (RvF) are considered. Study period: May 12, 2000 – April 11, 2001.
2+ 2+ + + - - Sampling Compart- Ca Mg K Na Organic C total N total P NO3 Cl Period ment 12/05/00 RF - TF 5.23 19.79 27.62 7.69 0.86 0.62 0.27 0.46 08/06/00 LP 2.26 3.28 7.63 1.00 0.36 0.24 0.13 0.08 RvF 0.64 3.67 1.96 0.87 0.09 0.04 0.03 0.04 08/06/00 RF - TF 2.18 0.68 6.38 1.94 0.19 0.24 0.91 0.08 0.10 29/06/00 LP 0.25 0.23 0.77 1.75 0.05 0.03 0.09 0.01 0.02 RvF 0.21 0.14 0.02 0.39 0.12 0.07 0.26 0.03 0.03 29/06/00 RF - TF 0.24 0.10 2.41 0.10 0.13 0.12 0.35 0.02 0.02 27/07/00 LP 0.10 0.07 0.26 0.08 0.02 0.01 0.02 0.00 0.00 RvF 0.12 0.12 0.02 0.18 0.08 0.05 0.18 0.02 0.02 27/07/00 RF 0.73 0.04 0.01 0.10 0.01 0.22 0.32 0.07 0.11 - TF 0.63 0.44 5.28 0.08 0.09 0.11 0.27 0.04 0.04 31/08/00 LP 0.06 0.08 0.29 0.01 0.01 0.01 0.01 0.00 0.00 RvF 0.09 0.10 0.03 0.15 0.04 0.06 0.11 0.02 0.02 31/08/00 RF 7.08 0.37 0.47 1.38 0.13 3.22 9.32 1.98 1.58 - TF 6.27 2.06 25.17 0.42 1.11 1.58 5.05 0.65 0.59 26/10/00 LP 2.72 0.99 10.34 0.14 0.37 0.29 0.97 0.13 0.17 RvF 1.30 1.20 0.80 3.45 0.56 0.67 1.50 0.22 0.26 26/10/00 RF 0.81 0.26 0.03 0.35 0.18 0.58 2.67 0.24 0.28 - TF 1.08 0.30 5.28 0.06 0.15 0.29 0.67 0.10 0.12 30/11/00 LP 0.71 0.24 1.10 0.02 0.10 0.06 0.21 0.03 0.05 RvF 0.55 0.54 0.04 1.58 0.22 0.28 0.80 0.16 0.16 30/11/00 RF - TF 04/01/01 LP 0.03 0.01 0.06 0.00 0.01 0.00 0.00 0.00 RvF 0.21 0.18 0.07 0.97 0.05 0.02 0.03 0.05 04/01/01 RF - TF 06/02/01 LP RvF 0.10 0.09 0.02 0.48 0.03 0.01 0.01 0.02 06/02/01 RF 0.28 0.61 0.01 0.01 0.66 0.42 1.51 0.22 0.15 - TF 0.43 0.20 0.13 0.01 0.31 0.14 0.42 0.09 0.07 01/03/01 LP 0.36 0.19 0.04 0.00 0.03 0.03 0.15 0.01 0.02 RvF 0.03 0.07 0.01 0.06 0.02 0.02 0.04 0.01 0.01 01/03/01 RF 0.15 0.05 0.03 1.09 0.67 0.64 1.62 0.26 0.38 - TF 0.17 0.05 0.14 0.77 0.32 0.27 0.86 0.12 0.10 29/03/01 LP 0.13 0.06 0.07 0.21 0.07 0.08 0.23 0.03 0.03 RvF 0.01 0.01 0.00 0.10 0.04 0.04 0.10 0.01 0.02 29/03/01 RF - TF 0.02 0.01 0.04 0.01 0.04 0.10 0.21 0.03 0.05 11/04/01 LP 0.02 0.03 0.02 0.01 0.03 0.02 0.05 0.01 0.02 RvF 0.00 0.01 0.00 0.07 0.01 0.02 0.10 0.01 0.01
Table 5.5: Amount of nutrient outflows and fluxes in throughfall and in litter percolates (in kg.ha-1) from the Biboo – Minwo catchment and for the sampling period (May 12, 2000 until April 11, 2001)
- Compartment Ca Mg K Na Organic Total N Total P NO3 Cl Carbon River Flow 3.3 6.1 3.0 8.3 1.3 1.3 3.2 0.5 0.6 Throughfall + 22.9 28.8 93 14.1 4.3 4.5 10.9 1.4 1.9 Litter Percolate
53 5.3.5. Export or accumulation of major nutriments Rainwater produces the nutrient input to the rainforest ecosystem, as the soils of the study area are chemically poor. By crossing through the aerial phytomass and after the leaching of the vegetation and washing of the deposited particles from the vegetation, throughfall is enriched chemically and play with litter percolate an important part in the nutrient transfer process from the vegetation to the soil. We have noticed only a small loss of chemicals by drainage from the stream flow due to the nutrient conservation processes in the vegetation and the soil. This pattern has been described as an auto-sustained rainforest ecosystem.
54 6. DISCUSSION AND CONCLUSIONS
The results presented in this report are based on hydrological (and soil) data collected between August 1995 and April 2001. These results complete the earlier data given in Waterloo et al., (1997; 2000). The results of the nutrient cycling study, carried out between May 2000 and April 2001, are presented here. The results of the hydrological modelling presented in chapter 4 are based on one test run, therefore they should be viewed with caution.
The distribution of rainfall over the TCP area over a 5-year study period (see Figures 4.1 and 4.2) showed a clear pattern. Relatively low rainfall totals were observed in the lowland (1816 mm) and highland (1983 mm) areas, whereas distinctly higher rainfall totals (2115-2458 mm) were observed in the lowland - highland transition zone. The rainfall distribution pattern may be related to the change in the topography, which favours orographic rainfall in the central zone of the area where the elevation changes from less than 200 m a.s.l. to up to 1000 m a.s.l. over a distance of less than 20 km.
The rainfall input data collected for a 5-year period for the Biboo – Minwo (forested, selective logging) and Nyangong catchments (forested, shifting cultivation) didn’t differ considerably, average values amounted respectively to 2078 and 1944 mm. In spite of the land uses observed in each catchment, water yield and evaporation showed similar average values, which were 721 and 627 mm for the water yield, 1358 and 1316 mm for the evaporation. As such, the effect of differences in precipitation on the water yield was much more pronounced than that of differences in land use. The effects of small-scale land use conversions on the catchment water yield may therefore be considered negligible (i.e. within the range of measurement errors) at the present intensity of logging or agriculture.
Rainfall interception by primary rain forest in the central part of the research area studied during a period of 3-months amounted to 247 mm or 35% of the incident rainfall (683 mm) (Lu1 project). This is a rather high value as compared to those observed in other studies in tropical rain forest, where the interception loss generally varies between 15 and 30% of total rainfall (Bruijnzeel, 1990). Rainfall interception by a cassava crop was with 85 mm or 15% of the incident rainfall (498 mm) much lower than that by the forest vegetation. As the interception of rainfall by the vegetation canopy is one of the parameters affecting the total evaporation from a catchment, and therefore indirectly the catchment water yield, the large-scale conversion of rain forest to a cassava (or comparable) crop type will result in a higher water yield. In this case, the amount of water reaching the soil would have increased by 136 mm over the 3-month period. In South Cameroon, agricultural activities are presently limited to a small percentage of the total area (see Nounamo and Yemefack, 2000). Effects of selective logging and shifting cultivation – on the scale as carried out in the TCP research area during the study period - on rainfall interception may therefore be considered small.
The surface water in the TCP area could be classified on the basis of its electrical conductivity and two types of water have been recognised. The first type had a relatively high EC (35-75 µS cm-1) and was observed only in the western and southwestern parts of the area where soils were known to be relatively nutrient-rich (van Gemerden and Hazeu, 1999). The second type of water covered the central and eastern parts of the TCP area, as well as vast areas to the south and to the east of the TCP area. This type of surface water had a low to very low EC, ranging from 14 µS cm-1 to 30 µS cm-1. The differences in surface water conductivity may be caused by local differences in the mineralogical composition of the rock, or by differences in the chemical weathering processes in the soil and rock, which occur at a lower rate or intensity in the latter area than those in the former zone where EC values are higher. In spite of the - rather small - observed differences in the surface water conductivity, the TCP area may presumably be considered homogeneous with respect to the chemical composition of the surface water. The Biboo - Minwo water samples analysed showed a low level of chemicals (Ecol2 project). 55 Physical properties of the soil (permeability, soil cohesion, aggregate stability) were studied in the Nyangong (forest - shifting cultivation) and Biboo - Minwo (forest - selective logging) catchments. (Lu1 project). The properties of the soils under undisturbed rain forest were distinctly different from those of soils on skid tracks made in the process of selective logging, or from those of soils under agriculture. The permeability of the forest soils was high, but decreased sharply upon disturbance and compaction by skidders during log extraction. The soil cohesion and compaction was higher in forest soils than in soils under shifting cultivation due to the presence of a root mat in the former. The highest cohesion and compaction values, however, were observed in heavily disturbed soils on skid tracks. The bulk density of forest soils was low due to the presence of organic material and large pores. Compaction of the soil by skidders resulted in a significant increase of the bulk density. Soil aggregate stability measurements showed that the clayey forest soils at Biboo - Minwo and Nyangong were very stable. The more sandy soil of a third (the Saa) catchment was less stable. The aggregate stability decreased after the soil had been disturbed by skidders.
The results of the soil studies suggest that the susceptibility of the soil to erosion increase when forestland is converted to another type of land use.
Sediment yields over the 5-year period were highest in the Nyangong catchment (702-946 kg ha-1) where agriculture and fishing activities were common practice along the main river, and lowest for the selectively logged forest Biboo – Minwo catchment (445-637 kg ha-1). In the forested Songkwé catchment, 31 kg ha-1 for a one-year measuring period was recorded. Those values - compared to the other data collected in the tropics - don’t give rise to great concern.
A single test-run of the LISEM model on the Minwo - Biboo area for forested and selectively logged scenarios with rough estimates for the soil and vegetation parameters showed that the simulated hydrograph and sediment yield were very different from the measured hydrograph and sediment yield. The model did predict that roads and skid tracks constructed during timber extraction are the main sources of sediment in the Biboo - Minwo catchment and that severe erosion may be expected when these roads or skid tracks are constructed on slopes exceeding 7.50-100.
The meteorological set-up in the Songkwé catchment - in spite of the short period of functioning - has provided half-hourly averages and standard deviations of global radiation, net radiation, wind direction, temperature, relative humidity measured at a height of 26-30 m, those of soil temperatures at -2 and -10 cm, and wind speed data will be available in the near future. The daily evaporation rates calculated on the basis of different data collected provided reasonable results. With a long-term observation, the modelling of the hydrology may be possible.
The results presented in this report indicate that changes in land use do not have a large impact on the hydrology or water quality at catchment scale (area > 3 km2). The impacts of selective logging and shifting cultivation on the vegetation, soil and water quality are, however, clearly visible on a smaller scale (e.g. at a scale of several hectares). Improvements in the techniques of selective logging and in the method of timber extraction in particular (see Bibani and Jonkers, 2001; van Leersum et al., 2001, Parren and Bongers, 2001), and in the agricultural system (Nounamo and Yemefack, 2000) should make it possible to reduce the damage to the soil and vegetation to some extent, which will further minimise the impact of the land use activities on the hydrology. A first step could be the implementation of a logging code (e.g. Gilmour, 1977b) as part of the management plan, which should provide guidelines for the protection of the drainage system through improved road and track construction, river crossings, implementation of buffer zones around the rivers, etc. (see also Fines et al. 2001a, b). Special attention should be given to village water supply areas, where human activity (e.g. selective logging or agriculture) should be minimised or prohibited to guarantee a continuous supply of water of good quality.
56 As reported by Bruijnzeel (1990), a number of nutrient balances in the tropical rainforests have been studied. For almost all studies however, the length of the data collection and the methodology used are different. The comparison with other rainforests ecosystems is therefor difficult. Nevertheless, on the basis of the results obtained during a 12-month period, some conclusions can be drawn. Values expressed in kg ha-1 per period of measurements have been obtained respectively for the river flow and for both the throughfall and litter percolate compartment (see Table 5.5). It seems that, for the major elements, the nutrient import in the ecosystem by rainfall is in general higher than the export by the river water stream. Throughfall and litter percolates constitute the most important store of the nutrients in the hydrological cycle in the TCP research area. The soils of the area studied have been classified as nutrient-poor soils (van Gemerden and Hazeu, 1999). A large part of the nutrients in the forest ecosystem available for the vegetation is stored in the hydrological cycle (this study), in the vegetation and the litter layer (see Ibrahima et al., 2002). The nutrient pattern described for the Biboo – Minwo catchment is characteristic for an auto-sustained rainforest. . However, uncontrolled human interference’s (excessive logging, shifting cultivation, road building and forest fires, causing erosion and redistribution of top soils and leaching) can disturb the productivity of the ecosystem for many years. This means that much care and well-conducted operations and activities are requested during logging operations.
57 REFERENCES
Abdul Rahim, N. and Zulkifli, Y. (1994). Hydrological response to selective logging in Peninsular Malaysia and its implications on watershed management. In: Proceedings of the International Symposium on Forest Hydrology, Tokyo, Japan. Appelo, C.A. and Postma, D. (1993). Geochemistry, groundwater and pollution. Balkema, Rotterdam, the Netherlands. Baharuddin, K. (1989). Effect of logging on sediment yield in a hill-dipterocarp forest in Peninsular Malaysia. Journal of Tropical Forest Science 1 (1), 56-66. Bastable, H.G., Shuttleworth, W.J., Dallarossa, R.L.G., Fisch, G. and Nobre, C. (1993). Observations of climate, albedo and surface radiation over cleared and undisturbed Amazonian rain forest. International Journal of Climatology 13, 783-796. Bibani Mbarga, R. and Jonkers, W.B.J. (2001). Silvicultural monitoring in permanent sample plots in Ebom forest, Southern Cameroon. In: Jonkers, W.B.J., Foahom, B. and Schmidt, P. (eds.). Seminar proceedings 'Sustainable management of African rain forest', part II: symposium. Tropenbos International, Wageningen, the Netherlands. Bilong, P. (1992). Caractères des sols ferralitiques à plinthite et à pétroplinthite développés sur roches acides dans la zone forestière du Sud Cameroun. Cah. ORSTOM, ser. Pédol. Vol. 27 (2), 203-224. Bonell, M. and Balek, J. (1993). Recent scientific developments and research needs in hydrological processes of the humid tropics. In: Bonell, M., Hufschmidt, M.M. and Gladwell, J.S. (eds.). Hydrology and water management in the humid tropics. UNESCO and Cambridge University Press, Cambridge, UK. Bosch, J.M. and Hewlett, J.D. (1982). A review of catchments experiments to determine the effect of vegetation changes on water yield and evapotranspiration. Journal of Hydrology 55, 3-23. de Bruijn, H.A.R., Kohsiek, W. and van den Hurk, B.J.J.M. (1993). A verification of some methods to determine the fluxes of momentum, sensible heat and water vapour using standard deviation and structure parameters of scalar meteorological quantities. Boundary Layer Meteorology 63, 231-257. Bruijnzeel, L.A. (1990). Hydrology of moist tropical forests and effects of conversion: a state of knowledge review. IHP-UNESCO Publications, Paris, France. 224 p. Bruijnzeel, L.A. (1993). Land use and hydrology in warm humid regions, where do we stand? IAHS Publ. 216, 3-34. Bruijnzeel, L.A. and Critchley, W.R.S. (1994). Environmental impacts of logging moist tropical forests. UNESCO-IHP Humid Tropics Programme Series N° 7. IHP-UNESCO, Paris, France van Dam, O. (2001). Forest filled with gaps. Effects of gap size on water and nutrient cycling in tropical rain forest: a study in Guyana. Tropenbos – Guyana Series 10. Tropenbos – Guyana Programme, Georgetown, Guyana. van Deursen, W.P.A. and Wesseling, C.G. (1992). The PCRASTER package. Department of Physical Geography, University of Utrecht, Utrecht, the Netherlands. Dickinson, A., Amphlett, M.B. and Bolton, P. (1990). Sediment discharge measurements Magat catchment: summary report 1986-1988. Hydraulics Research Report OD 122, Wallingford, United Kingdom Dolman, A.J. and Waterloo, M.J. (1995). Hydrology mission Tropenbos Cameroon Programme. International activities report 49. Winand Staring Centre, Wageningen, the Netherlands. Douglas, I. (1967). Natural and man-made erosion in the humid-tropics of Australia, Malaysia and Singapore. IAHS Publ. 75, 31-39. Douglas, I. (1968). Erosion in the Sungai Gombak catchment, Selangor, Malaysia. Journal of Tropical Geography 26, 1-16. Duchaufour, P. (1982). Pedology: pedogenesis and classification. George Allen & Unwin, London, United Kingdom.
58 Fines, J.P., Lescuyer, G. and Tchatat, M. (2001a). Master management plan for the Tropenbos- Cameroon research site. Pre-Mediation version. Tropenbos-Cameroon Documents 5. The Tropenbos-Cameroon Programme, Kribi, Cameroon. Fines, J.P., Ngibaot, F. and Ngono G. (2001b). A conceptual forest management plan for a medium size forest in southern Cameroon. Tropenbos-Cameroon Documents 6. The Tropenbos-Cameroon Programme, Kribi, Cameroon. Foahom, B. and Jonkers, W.B.J. (1992). A programme for the Tropenbos research in Cameroon: final report Tropenbos-Cameroon Programme (Phase 1). First revision. The Tropenbos Foundation, Wageningen, the Netherlands. Forti, M.C. and Neal, C. (1992). Hydrochemical cycles in tropical rainforests: an overview with emphasis on Central Amazonia. Journal of Hydrology 134 (1192), 103-115. Franken, W., Leopoldo, P.R. and Bergamin, H. (1985). Nutrient flow through natural waters in “terra firme” forest in Central Amazonia. Turrialba. 35, 383-393. Franqueville, A. (1973). Atlas régional sud-ouest 1 - République Unie du Cameroun. ORSTOM, Yaoundé, Cameroun. Fritsch, J.M. (1993). The hydrological effects of clearing tropical rain forest and of the implementation of alternative land uses. IAHS Publ. 216, 53-66. Galloway, J.N. and Likens, G.E. (1978). The collection of precipitation for chemical analysis. Tellus 30, 71-82. van Gemerden, B.S. and Hazeu, G.W. (1999). Landscape ecological survey (1:100,000) of the Bipindi - Akom II - Lolodorf region, southwest Cameroon. Tropenbos-Cameroon Documents 1. The Tropenbos-Cameroon Programme, Kribi, Cameroon. 164 p. Gilmour, D.A. (1977a). Effect of rain forest logging and clearing on water yield and quality in a high rainfall zone of North-east Queensland. Paper presented at the Hydrology Symposium of the Institution of Engineers, Brisbane, Australia. Gilmour, D.A. (1977b). Logging and the environment, with particular reference to soil and stream protection in tropical rain forest situations. Guidelines for Watershed Management, F.A.O. Conservation Guide 1. FAO, Rome, Italy. van de Griend, A.A. (1979). Modelling catchment response and runoff analysis. Lecture notes. Institute of Earth Sciences, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands. Hazeu, G.W., van Gemerden, B.S.,. Hommel, P.W.F.M. and van Kekem, A.J. (2000). Biophysical suitability of forest land in the Bipindi – Akom II – Lolodorf region, south Cameroon. Tropenbos – Cameroon Documents 4. The Tropenbos-Cameroon Programme, Kribi, Cameroon. Ibrahima, A., Schmidt, P., Ketner, P. and Mohren, G.J.M. (2002). Phytomasse et cycle des nutriments dans la forêt dense humide du sud Cameroun. Tropenbos-Cameroun Documents 9. The Tropenbos-Cameroon Programme, Kribi, Cameroun. Imeson, A.C. and Vis, M. (1984). Assessing soil aggregate stability by water-drop impact and ultrasonic dispersion. Geoderma 34, 185-200. Jordan, C.F. (1982). The nutrient balance of an Amazonian rain forest. Ecology 63 (3), 647-654. Lee, R. (1970). Theoretical estimates versus forest water yield. Water Resources Research 6, 1327-1334. Leersum, G.J.R. van, Ngibaot, F., Laan, E.A. and Jonkers, W.B.J. (2001). Logging methods applied in South Cameroon and ways for their improvement. In: Jonkers, W.B.J., Foahom, B. and Schmidt, P. (eds.). Seminar proceedings 'Sustainable management of African rain forest', part II: symposium. Tropenbos International, Wageningen, the Netherlands. Lesak, F.W. (1993). Export of nutrients and major ionic solutes from a rain forest catchment in the Central Amazon basin. Water Resources Research 29 (3), 743-758. Likens, G.E., Bormann, F.H., Pierce, R..S., Eaton, J.S. and Johnson, D.H. (1977). Biogeochemistry of a forested ecosystem. Springer Verlag, New York, USA. Low, A.J. (1954). The study of soil structure in the field and the laboratory. Journal of Soil Science 5, 1-57. Malmer, A. (1993). Dynamics of hydrology and nutrient losses as a response to establishment of forest plantation: a case study on tropical rainforest land in Sabah, Malaysia. PhD
59 dissertation. Swedish University of Agricultural Sciences, Faculty of Forestry, Umea, Sweden. Nangmo, Y.N. (1996). Etude de l’érosion hydrique du sol en relation avec les modes d'utilisation des terres dans le Sud Cameroun : cas des bassins versants de Nyangong et de Biboo - Minwo. National Institute for Rural Development. University of Dschang, Dschang, Cameroon. Ndam Ngoupayou, J.R. (1997). Bilans hydrogéochimiques sous forêt tropicale humide en Afrique : du bassin expérimental de Nsimi – Zoétélé aux réseaux hydrographiques du Nyong et de la Sanaga (Sud-Cameroun). Th. Doc. Université Pierre et Marie Curie, Paris VI, Paris, France. Nounamo, L. and Yemefack, M. (2000). Shifting cultivation in the evergreen forest of southern Cameroon: farming systems and soil degradation. Tropenbos Cameroon Reports 00-2. The Tropenbos-Cameroon Programme, Kribi, Cameroon. Olivry, J.C. (1986). Fleuves et rivières du Cameroun. Collection Monographies Hydrologiques de l'ORSTOM 9. ORSTOM, Paris, France. Onstad, C.A. (1984). Depressional storage on tilled surfaces. Transactions of the ASAE. p. 729- 732. ORSTOM (1995). Inventaire des formations superficielles des plateaux du Sud Cameroun. Institut Français de Recherche Scientifique pour le Développement en Coopération, Paris, France. Parker, G.G. (1983). Throughfall and stemflow in the forest nutrient cycle. Advances in Ecological Research 13, 57-133. Parren, M.P.E. and Bongers, F.J.J.M. (2001). Liana diversity and the effects of climber cutting in southern Cameroon. In: Jonkers, W.B.J., Foahom, B. and Schmidt, P. (eds.). Seminar proceedings 'Sustainable management of African rain forest', part II: symposium. Tropenbos International, Wageningen, the Netherlands. Pauwels, J.M., van Ranst, E., Verloo, M. and Mvondo Ze, A. (1992). Manuel de Laboratoire de Pédologie. Publications agricoles n° 28. Coopération belgo-camerounaise, AGCD- Bruxelles. Rodier, J. (1982). Analyse de l’eau. Masson et Cie. Paris,. France. de Roo, A.P.J., Wesseling, C.G., Cremers, N.D.T.H., Offermans, R.J.E., Ritsema, C.J. and van Oostindië, K. (1994). LISEM: a new physically-based hydrologic and soil erosion model in a GIS environment, theory and implementation. IAHS Publ. 224, 439-448. de Roo, A.P.J., Wesseling, C.G., Jetten, V.G. and Ritsema, C.J. (1995). Limburg soil erosion model. A user manual. Department of Physical Geography, University of Utrecht, Utrecht, the Netherlands. Ruppert, M. (1996). Quantification and conceptualisation of the components of the hydrological cycle in rain forest and shifting cultivation areas in South Cameroon. MSc. thesis. Department of Physical Geography, Faculty of Geographical Science, University of Utrecht, Utrecht, the Netherlands. Sahin, V. and Hall, M.J. (1996). The effects of afforestation and deforestation on water yields. Journal of Hydrology 178, 293-309. Seyler, P., Olivry, J.C. and Sigha Nkamdjou, L. (1993). Hydrochemistry of the Ngoko River, Cameroon: chemical balances in a rain forest equatorial basin. IAHS publication 216, 87- 105. Sigha-Nkamdjou, L. (1993). Caractérisation et fonctionnement hydrochimique d’un bassin versant en milieu forestier équatorial humide : l’exemple de la Ngoko à Moloun dou (Sud- Est du Cameroun). Th. Doc. Sc.. Université Paris XI Orsay, Paris, France. Stoorvogel, J.J. (1993). Gross inputs and outputs of nutrients in undisturbed forest, Taï area, Côte d’Ivoire. Tropenbos Series 5. The Tropenbos Foundation, Wageningen, the Netherlands. Subba Rao, B.K., Ramola, B.C. and Sharda, V.N. (1985). Hydrologic response of a forested mountain watershed to thinning. Indian Forester 111, 681-690.
60 Thom, A.S. (1975). Momentum, mass and heat exchange of plant communities. In: Monteith, J.L. (ed.). Vegetation and the atmosphere: volume 1, principles. Academic Press, London, United Kingdom. Tillman, J.E. (1972). The indirect determination of stability, heat and momentum fluxes in the atmosphere boundary layer from simple scalar variables during dry unstable conditions. Journal of Applied Meteorology 11, 783-792. UNESCO – MAB (1994). Environmental impacts of logging moist tropical forests. IHP Humid Tropics Programme, Series n° 7. Unesco - MAB Programme, Paris, 48 pp. Voeten, J.G.W.F. (1996). Effects of logging on soil physical properties in South Cameroon. MSc. thesis. Wageningen Agricultural University, Wageningen, the Netherlands. Waterloo, M.J. (1994). Water and nutrient dynamics of Pinus caribaea plantation forests on former grassland soils in Viti levu, Fiji. PhD dissertation. Vrije Universiteit van Amsterdam, Amsterdam, the Netherlands. Waterloo, M.J., Ntonga, J.C., Dolman, A.J. and Ayangma, A.B. (1997). Impact of land use change on the hydrology and erosion of rain forest in south Cameroon. DLO Winang Staring Center Report 134. Winand Staring Center, Wageningen, the Netherlands. Waterloo, M.J., Ntonga, J.C., Dolman, A.J. and Ayangma, A.B. (2000). Impact of shifting cultivation and selective logging on the hydrology and erosion of rain forest land use in South Cameroon. 2nd revised edition. Tropenbos Cameroon Documents 3. The Tropenbos Cameroon Programme, Kribi, Cameroon. Wiersum, K.F. (1984). Surface erosion under various tropical agroforestry systems. In: O'Loughlin, C.L. and Pearce, A.J. (eds.). Proceedings symposium on effects of forest land use on erosion and Slope Stability. IUFRO, Vienna, Austria and East-West Centre, Honolulu, USA. Wijngaard, J.C. and Cote, O.R. (1971). The budgets of turbulent kinetic energy and temperature variance in the atmospheric surface layer. Journal of Atmospheric Sciences 28, 190-201.
Unpublished sources van Leersum, G.J.R. 1996. Tropenbos-Cameroon Programme, B.P. 219, Kribi, Cameroon. Nounamo, L. 1996. Tropenbos-Cameroon Programme, B.P. 219, Kribi, Cameroon. de Roo, A.P.J. 1995. Department of Physical Geography. P.O. Box 80115, 3508 TC, Utrecht, the Netherlands. Yemefack, M. 1996. Tropenbos-Cameroon Programme, B.P. 219, Kribi, Cameroon.
61