soil respiration measurements in the kruger national park, south africa

CARBOAFRICA: MIDPROJECT STATUS REPORT

NEWS FROM...News from Ghana

The 62 meter high Eddy Covariance Flux Tower in the Ankasa Conservation Area (Ghana) has been

equipped with the relevant instrumentation byUniversity of Tuscia team. Now it has been workingsince April 2008. In addition to carbon fluxes alsometeorological variables are being measured. Inparticular, a complete micrometeorological stationwas installed to measure solar radiation, airtemperature and humidity, precipitation, soil watercontent, soil temperature and other variables.

An Eddy Covariance System with GILL Wind MasterAnemometer and a closed path IRGA LiCor 7000 wasalso installed to measure in continuum carbon, water,and heat exchange associated to the forestecosystem. A profile of CO2 system, with six points, will be installed and tested during the next expeditionon the end of July 2008.

EDITORIAL

Dear reader,

CarboAfrica is halfway on the difficult, but fascinating path we are going along towards a better comprehension of the Africa role in the global carbon cycle. The project has already produced concrete results and we think it is important to show them, even if still partial and incomplete. Therefore, we have decided to turn the CarboAfrica newsletter N. 5 in a Mid Project Status Report, containing many pieces of information on the first results obtained and 3 short scientific articles. The showed results are just some little "carbon spot" taken in different situation and not linked each other. However, at the end of the project, these single "pixels" will be integrated in a global picture of the carbon cycle in Sub-Saharan Africa!

The CarboAfrica Secretariat

4-Monthly NewsLetter No. 5: May 2008

CONTENTS

CARBOAFRICAQuantification, understanding and prediction of carbon cycle and other GHG gases in Sub Saharan Africa

CARBOAFRICA Newsletter No. 5 : May 2008 1

Figure 1. A moment of the installation of the instruments on theflux tower in the Ankasa Tropical Forest.

Editorial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1NEWS FROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1News from Ghana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1News from Zambia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3TRAINING NEWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4Field Training Workshop on Carbon Cycle Measurements . . . . . . . . . . . .4MEETING NEWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6Climate, carbon and cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6

FIELD ACTIVITIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6Quantifying carbon emissions of African wildfires . . . . . . . . . . . . . . . . . . .6Measuring impacts of logging on C stocks in Bobiri forest, Ghana . . . . . .7APPENDIX - CARBOAFRICA SHORT SCIENTIFIC ARTICLES . . . . . . . . . . . . . . .8Soil respiration in the Kruger National Park, South Africa . . . . . . . . . . . . .8Evaluating NPP and biomass in a Congolese savannah . . . . . . . . . . . . .10Estimations of fine root biomass in eucalyptus plantations . . . . . . . . . . .13Links & Contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

The tower will be managed by University of Tuscia (Italy) in cooperation with the Cape Coast University (Ghana) that will carry out the periodic maintenance of the two systems.A small laboratory is near to be finished just at the base of the tower; it will be of great support to all the research activities that are planned.Associated to the tower site a field campaign for biomass, biodiversity and soil data was carried out.

In particular biomass and biodiversity have beenestimated along 2 transects 1000 m in length and 10m in width. The 2 transects were perpendicular eachother and the tower was at their intersection. The 4branches of the 2 transects were oriented followingN, E, S and W, respectively. Plant height anddiameter were considered: plant diameter wasmeasured at 1.30 m from the soil, or 0.5 m above theexternal roots. The average canopy height wasabove 30 m and the maximum measured treediameter was almost 2 m. Plant species wereidentified: around 180 tree species were identifiedincluding at least one new species for Ghana.

The soil in the surrounding of the eddy covariancetower, located in the Ankasa tropical forest, wassampled to assess its basic properties and the totalcarbon and nitrogen content. In this purpose,following the CarboAfrica soil protocol, a grid of 200x 200 m has been identified so to cover the fetcharea of the eddy covariance measurements. Apreliminary survey of the area was done to identifythe main pedogenic soil horizons by the use of ahand auger. Then the litter layer and the soilsamples were collected down to 50 cm depth. Thesampling points were marked with the use of a GPSand the samples divided by horizons and labelled.


Figure 2. A break at "high altitude" (!) during the instrumentsinstallation.

Figure 3. Measuring LAI by hemispherical photo technique.

Figure 4. Measuring tree diameter in the forest.

Figure 5. Litter layer in the Ankasa Tropical Forest.

Consequently, an area was individuated and the representative soil profile was dug down to 1 m depth, described by genetic horizons according to the USDA (United States Department of Agriculture) soil survey manual. The samples are now being processed for their basic characteristics (texture, pH and exchangeable phosphorous) and for their total amount of organic carbon and nitrogen. The bulk density as well as the field capacity and the wilting point were estimated directly in the field.All the data are still under analysis and in the next newsletter we will show the first results and a preliminary analysis of the Ghana Primary rain forest!In parallel a study on forest degradation has been conducted in the Bobiri forest (see pages 5 and 6 for details).

AcknowledgmentsSpecial thanks to Ms Elisa Grieco, Mr Renato Zompanti, Mr Paolo Stefani, Mr Abdul-Kareem Fuseini, Mr Joseph Appiah, Mr Justice Mensah, Mr Francesco Mazzenga, Mr Tommaso Chiti, Ms Claudia Consalvo, Mr Enrico Vittorini, the Personnel of the Ankasa Conservation Area and the workers, who have contributed to the achievement of these results.

News from Zambia

During the three months of January till March 2008, a team from the Max-Planck Institute for Biogeochemistry in Jena (MPI-BGC), Germany and the Meteorological Department of Zambia, Mongu Division, conducted an intensive field campaign at the Mongu flux tower. First objective was to do an intensive maintenance program on the Eddy-covariance system and thereafter the site characterisation for better understanding of the net ecosystem exchange (NEE) measurements done at the tower. This included the set up of 4 different plots lying on moisture and disturbance gradient (Fig. 6).

During the time, the team managed to finishcharacterising 2 of the 4 plots, including an inventoryof the aboveground biomass and takingapproximately 1500 soil samples for Carbon andNitrogen content analysis in the lab (Fig. 7).

Additionally measurements of soil respiration,photosynthesis and leaf area index were takencovering an overall period of 3 weeks. Where soilrespiration measurements were studied under theaspect of heterogeneity due to differences in groundcover (Fig 8) and the separation between root andmychorizal respiration using nylon mesh bags in anexclusion experiment (Moyano et al., 2007). Photosynthesis measurements were taken at theeight dominating tree species dividing between sunand shade leaves.


Figure 6. Aerial image of the Mongu site, showing the forestreserve and the disturbed surrounding. Small red squares are thetower and the solar station; big red squares show the plots, set upduring the field campaign. The red arrow is show the disturbanceand moisture gradient to the south.

Figure 7. Soil sample collection at the Mongu tower site.

Figure 8. temperature corrected (R20) soil respirationmeasurements against soil water content in one of theundisturbed plots at Mongu flux tower. R2 = 0.24.

During the campaign, the team tried to get in close contact to the surrounding communities, since the rate of deforestation seemed to have increased during the last two years. First analysis of the eddy covariance data shows that the tower measures intact and protected Miombo woodland on the one hand and a highly disturbed and not protected Miombo woodland on the other (Fig 9). First results of the sectoral analysis will be represented at the "Carbon and Communities in Tropical woodlands" conference in Edinburgh, Scotland, June 2008. At least one additional campaign is planned for August and September 2008, including further studies on soil respiration (wetting experiment) and photosynthesis.

AcknowledgmentsWe are very thankful to all helping hands (Maurice R. Muchinda - director of the Meteorological Department Zambia, Mukufute M. Mukelabai - head of the Meteorological department Mongu, Waldemar Ziegler - technician of the MPI-BGC, Olaf Kolle - meteorologist at the MPI-BGC, Liunji Lumbala - field assistant, Anna Kamwanga - field assistant, Kabukabu Muhau - field assistant and Moses S. Namusunga - field assistant) preparing the field trip in Jena and Mongu and the never ending field work under difficult circumstances, including thunderstorms and broken equipment.

BibliographyMoyano FE, Kutsch WL & Schulze ED (2007) Response of mycorrhizal, rhizosphere and soil basal respiration to temperature and photosynthesis in a barley field. Soil Biology & Biochemistry 39: 843-853

InformationLutz Merbold: [email protected]

TRAINING NEWS

Report of the CarboAfrica Field Training Workshop on Carbon Cycle Measurements

Phalaborwa, South Africa, 10-15 March 2008

The second CarboAfrica Field Training Workshopwas held at Phalaborwa, on the outskirts of theKruger National Park, South Africa, 10th - 15th March 2008. The workshop was organized by the Councilfor Scientific and Industrial Research (CSIR),Natural Resources and the Environment. Theobjective of the workshop was to expose trainees toa range of field-based methods relating to carboncycle measurement. This relates to the CarboAfricaWorkpackage 5, which focus on communication andcapacity building, particularly to African institutionsand stakeholders, and the dissemination of data andinformation.One of the objectives of CarboAfrica is to train andequip students and professionals in the Sub-Saharan African region with skills, information andtechnologies to enable them to measure, analyseand understand the carbon cycle and other GreenHouse Gases (GHG). Eddy covariance flux towerhas been set-up at Malopeni (23.832545 S and31.21436 E), inside the Kruger National Park, underthe CarboAfrica project. The Malopeni site waslocated to sample Colophospermum mopanesavanna, a widespread vegetation type in southernAfrica, which characteristically occurs on hot, drysites with poor soil nutrients. The Council for the Scientific and IndustrialResearch (CSIR) released an announcementthrough CarboAfrica network, inviting Africanstudents and professionals to attend the CarboAfricaField Training Workshop. The workshop was basedat the recently-completed node office of the SouthAfrican Earth Observation Network (SAOEN) in thetown of Phalaborwa, on the border of the KrugerNational Park. The training was very practical andhands-on, with participants working in small groups,each with an instructor.


Figure 10.CarboAfrica Field Training Workshop attendance at theSAEON hall, Phalaborwa, Kruger National Park, South Africa.

Figure 9. Deforested Miombo woodland close to the tower. Alllarge trees have already been cut and it is just a question of timetill the small trees are gone as well.

Fourteen participants from all over Africa (in particular from Burkina Faso, Congo, Ghana, Kenya, Lesotho, Senegal, South Africa, Togo, Zambia and Zimbabwe) were selected to attend the workshop. Their registration, travel, accommodation and meals were funded by the Food and Agriculture Organization of the United Nations (FAO). The first day of the workshop was programme briefing and networking. The curriculum was divided into the following three Working Groups (WG): • WG1: Site characterisation - Dr Bob Scholes

(Council for Scientific and Industrial Research (CSIR), Natural Resources and the Environment, South Africa)

• WG2: Ecophysiology measurement and analysis - Dr Werner Kutsch (Max-Planck-Institute of Biogeochemistry)

• WG3: Eddy covariance flux measurements and processing - Lutz Merbold (Max-Planck-Institute of Biogeochemistry, Germany).

A polar system was adopted as the sample frame for the site survey to quantify, and measure soil and vegetation characteristics at the Malopeni site. The sample lines radiate out from the flux tower along the cardinal directions, for a distance approximately equal to the potential tower footprint (Figure 3). This scheme is not only easier to layout than the more conventional grid, but more appropriate for eddy covariance analysis and interpretation.Soil samples, grass and leaf biomass were collected at each point along the lines (Figure 4). Participants were shown how to use circular quadrat sampling and allometric models to estimate the aboveground woody biomass at the site.Biomass collected from the field was dried and weighed at the SAEON lab and species were identified. The diameter and height of trees were measured. Hemispherical photographs were captured to estimate the canopy cover and leaf area index. Data was analysed in the last day of

workshop. The tree cover is dominated by Colophospermum mopane and Combretum apiculatum, with an average height of 3m. The dominant grass species are Schmidtia pappophoroides and Aristida congesta, with Panicum maximum under the tree canopies.

Dr Werner Kutsh demonstrated ecophysiological measurements and analysis. The measurements were taken on leaves of C. mopane and C. apiculatum. These leaves open their stomata early in the morning and then closed them when the leaf temperature increases. This is a survival mechanism of drought tolerant species to conserve the loss of water and other nutrients. Those species fold their leaves to avoid direct heat from the sun. Lutz Merbold gave hands-on experience on eddy covariance data collection, quality checks,


Figure 12. Polar survey method followed for data collection atMalopeni site.

Figure 13. Quadrat used for sampling leaf and grass biomass,and an easily-seen but biodegradable and non-harmful to wildlifelocation marker made of electrical conduit. The permanent pointmarker is a galvanised washer with the point angel and distancepunched onto it, pegged to the ground with a 15 cm nail.

Figure 11. Dr Bob Scholes describing sampling strategies for thevegetation at Malopeni, Kruger National park.

processing and analysis. Response of participants after the workshop shows that they have gained new skills, knowledge and methods, which they can use in the future for better understanding of carbon cycle.

AcknowledgementsWe thank FAO for financial support; participants of CarboAfrica for data collection; Tony Swemmer of SAEON and Helen De Beer of CSIR for their assistance with logistic arrangements.

InformationRudzani Makhado: [email protected] Scholes: [email protected]

MEETING NEWS

Climate, carbon and cultures: recent research and case studies from West Africa

Saly, Senegal, 14th - 15th February 2008

A symposium was held in Saly, Senegal, 14-15 February 2008, based on climate and carbon-related research and development needs in West Africa. The two-day symposium was organized by the Forestry Technology Centre of Catalonia (CTFC) from Spain, and hosted by the Senegalese Agriculture Research Institute (ISRA) from Senegal. Invited were delegates from four additional West African nations (Côte d'Ivoire, Ghana, Niger, Togo), UNFCCC National Focus Points, and African and European (Spain, Ireland and Italy) researchers and students working in the domains of the terrestrial carbon cycle, climate change adaptation, and agricultural research and development.

Dr. Antonio Bombelli was invited as the keynote speaker to present the objectives of CarboAfrica and the state of European research on GHG in Africa. There were other 12 presentations on research from the domains of forestry, agronomy, terrestrial

ecology, and the social sciences in Africa. In addition, round-table discussions were held as an opportunity for UNFCCC focal points and scientists to discuss the importance, constraints, and research on climate adaptation and terrestrial carbon resources.

FIELD ACTIVITIES

Quantifying carbon emissionsof African wildfires

A consortium of scientists from WP4 (fire-climate-carbon cycle interactions), and contributing scientists have developed and applied a method to quantify carbon emissions from African wildfires. The approach couples a satellite based burned area map (L3JRC) listing burned areas for the years 2001-2006 with a dynamic vegetation model containing a state-of-the-art fire module (LPJ-GUESS-SPITFIRE).The burned area map is a directly sensed (in contrary to earlier approaches using so called hot spots i.e. active fires) map of burned vegetation at a 1 km scale. Carbon emissions depend on fuel availability (an output of the dynamic vegetation model LPJ-GUESS) and fuel combustion completeness as calculated by the fire module (SPITFIRE). The latter strongly depends on the climatic conditions at the time of burning. Here we use a climate data set with one day resolution which has been especially generated for the African continent by members of the CARBOAFRICA consortium by adjusting datasets with a long time-span but known systematic errors with other data sets of better quality which cover only a short-time span, to have a consistent and highly reliable climate data set for joint applications in the CARBOAFRICA project. The study estimated an average area of 195.5±24·104 km2 that was burned annually, releasing an average of 723±70 Tg C to the atmosphere for the time period from 2001 to 2005. We used the L3JRC data set which is the first available multi-annual dataset of burned area. However, to track recent climate or sociological caused changes in burning patterns a continuous time-near product listing burned areas is required using an integrated methodology to observe and understand wildfires.While the emission estimates for the time span of 2001-2006 are submitted, the next step is to look more in detail in driving factors of burned area which is assumed to be mainly human ignited and to develop future scenarios of wildfire emissions based on a model of human ignition instead of remotely sensed burned area.


Figure 14. The participants in the symposium.

Measuring impacts of logging activitieson C stocks in Bobiri forest, Ghana

Degradation of Ghanaian forests occurs over large areas and can significantly contribute to the overall emissions from forest loss. Since the 1940s more than 90% of Ghana's high forests had been exploited and about 80% was converted to agricultural land.

Improving forest management in Ghana has received particular attention through the Reducing Emissions from Deforestation and forest Degradation (REDD) mechanism and the EU Forest Law Enforcement Governance (FLEGT) program. Ghana is a pilot country under the World Bank Forest Carbon Partnership Facility (FCPF) and the FLEGT. However, developing carbon forestry projects that focus on decreasing forest degradation faces several scientific constraints: (1) estimation of C stocks, (2) identifying forest changes using remote sensing, (3) identifying baseline scenario and (4) fulfilling the additionality criteria of REDD project.The aim of our research was to account for the impact of selective logging on C balance. The objectives were: (1) to assess live and dead biomass before and after logging, (2) to build coefficient of logging impacts based on timber volume extracted and canopy opening, (3) to predict C balance for a project time period and (4) to scale-up local field measurement to the forest.

The study was carried out in three compartments of the Bobiri forest, which falls within the South-east subtype of the moist semi deciduous forest (latitude 6°44'N and longitude 1°22'W). The climate is semi-humid tropical. The Bobiri forest was used as an experimental area since the 40's to improve logging activities and timber yields. It resulted in various types of forest structure, from secondary forest to natural untouched forest. Forest inventories were conducted in protected forest and secondary forest compartments. Aboveground biomass of tree, liana and palm tree was estimated. Sample plots of 1 ha were selected and centred on trees to be logged. After logging, the various impacts of logging on biomass were estimated. The various types of logging impacts were roads, logging bays, skid trails and tree falls. Damaged and dead tree were inventoried and tagged for each of those damage type within a 150 ha compartment. Canopy structure was estimated using LAI (Leaf Area Index) and was measured before and after logging. A relation between canopy opening, ground damages and C losses was established for the various types of impact and used to estimate the total losses at the scale of the forest.

AcknowledgmentsThis experiment was implemented by Marion Chesnes, Matieu Henry, Stephen Adu Bredu and Laurent Saint-André, in partnership among the Forest Research Institute of Ghana (FORIG), the French Agricultural Research Centre for International Development (CIRAD), UPR80 - Ecosystèmes de Plantations, and the University of Tuscia (Italy).

InformationMatieu Henry: [email protected]

Figure 16. Conceptual diagram of the C stock variation afterlogging.

Figure 17. Dr. Stephen Adu Bredu (FORIG) measuring thediameter of a liana with a collaborator.


Figure 15. A skidder entering in the forest to extract a logged tree.

C st

ocks

Years

Cinitial Closses Crecovery

Logging

CarboAfrica Short Scientific Articles

Soil respiration measurements in the Kruger National Park, South Africa

R. Makhado (Natural Resources and the Environment, CSIR, Pretoria, South Africa; [email protected], [email protected])

1. Introduction

CarboAfrica also aims to provide analysis of CO2 fluxes from soil, fire, atmosphere and other ecological variables. This article report on the progress made on CO2 fluxes from soil at the Kruger National Pak. Soil respiration measurement is an ongoing activity undertaken by the Council for the Scientific and Industrial Research (CSIR) since 2000 at Skukuza, Kruger National Park.

2. Objectives

The purpose of soil respiration research at Skukuza is to understand the effect of soil temperature, soil moisture, rain and camopy on the rate of soil respiration.

3. Material and Methods

Respiration measurements are taken in the broadleafed Combretum savanna on sandy soil, and the fineleafed Acacia savanna on clayey soil. A LICOR Infrared Gas Analyzer System was used to measure CO2 emission from the soil under the chambers in the background (Fig. 1). Musa Mavundla, MSc student from University of the

Witwatersrand, measured soil respiration at Skukuza bi-weekly from May 2000 to April 2001 using a PP Systems instrument and an array of 100 mm diameter collars by 70 mm high.Dr Werner Kutsch of the Max-Planck-Institute for Biogeochemistry added to our collection of respiration data during the period he spent at the site in 2004. Dr Jason Neff of the University of Colorado installed a Viasala CO2 probe in two locations in the site for near-real time soil respiration monitoring. He also draws soil atmosphere samples for C-isotope analysis, to partition soil respiration between tree- and grass-derived sources. In December 2006, the CSIR collaborated with the South African National Biodiversity Institute (SANBI), for further respiration measurements, and in 2007 with the Second University of Naples, Italy.

We find it important in these savannas to stratify the measurements between under-canopy locations, with a much higher respiration rate, and between-canopy locations (Fig. 2). We often keep a set of collars as controls, and simulate rainfall by adding 20 mm of deionised water to another set.

The aims of the 2007 experiment were to test the following hypotheses:1. What is the effect of the first pulse of rain on soil respiration rate (as opposed to subsequent rainfall events),

and2. What is the effect of soil temperature, moisture and tree canopies on soil respiration rates?

Figure 1. The LICOR Infrared Gas Analyzer system for measures of soil CO2 emission.


Appendix - Soil Respiration

4. Results

Part of the results of these researches were presented at the 6th Annual Science Networking Meeting, 2008, Skukuza, Kruger National Park. We find that the soil moisture content in the savannas is generally very low (1% - 17%), and the soil temperatures can be high (11oC - 45oC). Major finding from the study are:

1. Respiration rates increase linearly with increasing soil moisture,

2. Soil respiration reached a maximum at 28 oC at this site, and decline above that level.

Moreover, the following experiments show that tree canopies, in addition to soil temperature and moisture, have an effect on soil respiration rate. Respiration rate

positively correlates with soil temperature and moisture, particularly under the canopies. This is a result of the fact that tree canopies reduce the rate of loss of soil moisture through evaporation, a less extreme soil temperature and a higher soil organic matter content.

5. Next Steps

Currently, the CSIR is collaborating with the University of Tuscia, Italy and the Max-Planck-Institute for Biogeochemistry, Germany in our mission to understand the effect of meteorological and physiological driving factors on soil respiration in the savannas. The research will increase our understanding on the diurnal and seasonal variation of soil respiration, the relative contribution of root and micro-organisms to the emissions, and the effect of soil temperature and moisture on respiration rate.

6. Acknowledgments,

The South African National Parks (SANPark) is thanked for providing field assistance and for allowing us to undertake our research at the Kruger National Park.

Figure 2. Collars placed under and between tree canopies in the Combretum savanna site.

Appendix - Soil Respiration


Evaluating NPP and biomass increment in a savannah of the Coastal plains of Congo

A. de Grandcourt & L. Saint-André(CIRAD - Centre de Coopération Internationale en Recherche Agronomique pour le Développement, France; UR2PI - Unité de Recherche sur la Productivité des Plantations Industrielles, Congo; [email protected], [email protected])

1. Introduction

In the context of the CDM, afforestation has been suggested as a way to simultaneously sequester carbon, increase wood and paper supplies, and diversify rural incomes. In Congo, the focus of much of the research on this land-use change has been on the biogeochemical cycles: clonal plantations of Eucalyptus have been introduced for 30 years on savannah soils of the coastal plains of Congo and the impact of these stands on soil macrofauna (Mboukou-Kimbatsa et al., 1998), understorey floristic composition (Loumeto and Huttel, 1997), soil organic matter characteristics (Trouvé et al., 1994; Bernhard-Reversat et al., 2001; Landais, 2003, D'Annunzio et al. 2008), and nitrogen mineralization (Nzila et al., 2002) have been investigated. Studies on biogeochemical cycles of nutrients and water had been compared in a clonal Eucalyptus stand and an adjacent savannah ecosystem over more than three years (Kondi site, Laclau et al. 2003 a et b, Laclau et al. 2005). The dynamics of biomass and nutrient accumulation in the plants were quantified in the two ecosystems (Laclau et al., 2000, 2002). To complete this comprehensive study on land use change, an eddy-flux tower was transferred from the plantations to a savannah (Tchizalamou site). In addition to the CO2 balances measured by the flux towers, measurements of CO2, N2O and CH4 soil-atmosphere exchanges were also conducted on this site. The structure of the savannah and its dynamic (species composition, above- and below-ground biomass, canopy height, coverage, leaf area index (LAI), root growth and turn-over) were also quantified. In this paper, we focus on the species composition and on the dynamics of above- and below-ground biomass.

2. Objectives

This study started in September 2006 and ended in May 2008 (the site will be afforested in October 2008). This work aimed at:

1. studying the time courses of above- and below-ground biomass and necromass in the savannah ecosystem2. understanding which factor is driving the vegetation dynamics in this ecosystem3. understanding the consequences of these dynamics on the C and water budget4. assessing NPP independently from eddy-flux measurements


Sixteen (7m x 7m) plots were selected randomly in a one ha area around the flux tower (Fig. 1). Each plot contained 16 sub-plots of 1 m2 (systematic pattern) where biomass was measured alternatively (one subplot is selected randomly at each date of sampling). A total of 15 field campaigns were done since September 2006 (every 6 weeks) corresponding respectively to 3 and 6 measurement points in dry and wet season. Aboveground biomass and necromass was totally harvested in the subplots. Species were considered individually. Belowground biomass was assessed from 4 auger cores (8 cm diameter). Soil samples were taken down to 0.7 m deep (4 layers; 0-10 cm, 10-30 cm, 30-50 cm and 50-70 cm) and bulked. Species were not considered independently for the belowground biomass.

1 2

3 4 A 5

6 7 8

9 10

11

12

B

13 C 14 D

15

16

thermocouplesTDR

hut

solar panels

flux towerprobes tower

rhizotrons

rhizotrons

rhizotrons

1'

2'

3'

4'

5'

NFigure 1. Experimental design of Tchizalamou's site. Plots labelled 1 to 16 are used for biomass assessments.

Appendix - NPP and Biomass


4. Results

Above-ground biomass dynamic is linked to the rainfall, especially during the dry season and at the beginning of the rainy season (Fig. 2). However, once the ears are mature, the vegetation, made of 90% of grasses, stagnates and dries out. During this transition, the above-ground necromass grows up and reaches its maximum at the beginning of the dry season (June). In 2007-2008, the small dry season was particularly long (from mid-December to mid-April) whereas it was missing in 2006-2007. An early increase of the necromass was then observed: 0.5t ha-1 in January 2008 instead of 0.2 t ha-1 in January 2007. On the contrary, the biomass was about 1.5 t ha-1 in January 2008 while it reached 1.8 t ha-1 in January 2007. When the fire occurs, 3.54 t ha-1 of dry matter were volatilised.

The dominating specie in our site is Loudetia simplex, followed by Ctenium newtonii. The proportion of each species varies over the year showing the succession of their biological cycles (Fig. 3). Eulophia cucullata, an orchid, blossoms at first on the savannah during the dry season,

whereas the other herbaceous are still small (October). In November-December, Loudetia arundinacea is the first of the Poacea to blossom. Around March, all the plants have achieved there reproduction cycle and the seeds have been spread over.

The roots are the most important part of the total biomass. The ratio Root/Shoot ranges from 1.7 at the maximum vegetation (in April) to nearly 40 after the destruction by the fire of the aerial part (Fig. 4)

The root biomass increases at the end of the wet season and this increase corresponds to the occurrence of above-ground necromass (Fig. 5). We can then suppose a mechanism of C storage in the root system just before the

dry season. This can save biomass during fire events and it ensures a rapid re-growth at the beginning of the next wet season.

Fire

0

0.5

1

1.5

2

2.5

3

3.5

01/06/2006

31/08/2006

30/11/2006

01/03/2007

31/05/200

7

30/08/2007

29/11/200

7

28/02/2008

date

abov

egro

und

mas

s(T/

ha)

0

500

1000

1500

2000

2500

cum

ulat

ive

rain

stil

l1 ju

ne

aboveground biomass

abovegroundnecromasscumulative rain

Figure 2. Time course of above-ground biomass and necromass (t ha-1) and of the cumulated rainfall from the 1st of June of each year (n=16, vertical bars represent standard deviation).

Figure 3. Proportion (%) of the above-ground biomass (leaves, stubbles, flower and seeds for the Poacea, or stem, leaves, flower and fruits for the dicotyledons) of each species (n=16). The absence of Loudecia arundinacea in April and June 2007 is probably due to confusions with Loudetia simplex. As a matter of fact, the disappearance of the sexual organs made the determination harder. The problem was solved in 2008.

0102030405060708090

100

19/09

/2006

12/12

/2006

06/03

/2007

06/06

/2007

03/09

/2007

04/12

/2007

% li

ving

bio

mas

s

LoudetiaarundinaceaLoudetia simplex

EulophiacucullataJoncacées

Eriosema

Ctenium newtonii

bolbustylislanicepsPoacée ind.

0

5

10

15

20

25

30

35

40

45

19/09

/2006

30/10

/06

12/12

/06

22/01

/07

06/03

/07

15/04

/07

06/06

/07

17/07

/07

03/09

/07

15/10

/07

04/12

/07

16/01

/08

date

R/S

Figure 4. Time course of the root/shoot ratio (n=16, vertical bars represent one standard deviation).

Figure 5. Time course of root biomass and necromass (t ha-1) in the 4 soil layers (n=16, vertical bars represent one standard deviation). Empty circles represent the above-ground necromass in t ha-1.

0

2

4

6

8

10

12

18/09

/2006

31/10

/2006

11/12

/2006

22/01

/2007

06/03

/2007

16/04

/2007

04/06

/2007

16/07

/2007

03/09

/2007

18/10

/2007

04/12

/2007

15/01

/2008

date

Roo

t mas

s (li

ve +

dea

d) (T

/ha)

0-1010-3030-5050-70aboveground necromass

0

0.5

1

1.5

2

2.5

abov

egro

und

necr

omas

s(T/

ha)



5. ConclusionsThe figures obtained for the Tchizalamou site are similar to those obtained by Laclau et al. (2002) in the nearby savannah of Kondi which was dominated by Loudetia arundinacea. Above-ground biomass was respectively of 2.2 t ha-1 and 3.8 t ha-1 for Tchizalamou and Kondi sites, while total biomass amounted respectively to 4.6 t ha-1 and 5.3 t ha-1. The difference is exactly the reverse for belowground biomass which ranged between 6 and 8 t ha-1 for Kondi and between 4 and 11 t ha-1 for Tchizalamou. Root/shoot ratios are also higher in Tchizalamou (1.7 to 40) than in Kondi (1.4 to 14.4). All these figures tend to indicate that the savannah in Tchizalamou is less fertile than in Kondi. Comparison with all savannahs studied in CarboAfrica will also be performed. To get the Net primary Production, studies on the root-turn over and on the root dynamics using rhizontrons are ongoing. Assuming that litter falls are null in such site (which is burnt every year), NPP will be calculated as the sum of biomass increment and root turn over and will be used in the carbon budget of this site.



Estimations of fine root biomass in eucalyptus plantations. A comparison of four sampling methods

J. Levillain, C. Jourdan & L. Saint-André(CIRAD - Centre de Coopération Internationale en Recherche Agronomique pour le Développement, France; UR2PI - Unité de Recherche sur la Productivité des Plantations Industrielles, Congo; [email protected], [email protected])

1. Introduction

The belowground biomass is an important part of the biosphere and may account up to approximately 30% of the aboveground biomass (Grier et al, 1981; van Noordwijk et al., 1996). Although tree roots are an important part of forest biomass both, their mass and distribution in the soil profile, are still very difficult to assess. Various methods, both direct and indirect, have been used to measure root biomass, but until now no technique has been accepted as a method of reference (Fabiao et al. 1995; Misra et al. 1998, Vogt et al. 1998) and uncertainty is often due to the different excavation methods used (Hoffman et al, 2001). Consequently, root biomass data are still scarse, limiting our capacity to fully characterize forested ecosystems and accurately assess forest C stocks (Resh, 2003).

2. Objectives

In this context, we focused our study on four direct methods of root excavation to assess medium and fine root biomass estimation. The main objective was to assess the accuracy of these four methods in relation with the labour time and costs.


The experiment lasted between mid-April and June 2007 in a 6-year old eucalypt stand (close to the end of the rotation for these plantations). The stand density was 800 trees per ha. The four sampling methods are based on root excavation. Soil samples are taken down to 1 m deep (3 layers; 0-10 cm (H1), 10-50 cm (H2) and 50-100 cm (H3)) around each tree in an elementary space defined by the Voronoï diagram; this design can be adapted to any tree spatial distribution. It consists in decomposing a metric space (S) determined by a discrete set of points. In the simplest and most common case, the Voronoi diagram for S is the partition of the plane which associates a region V(p) with each point p from S in such a way that all points in V(p) are closer to p than any other point from S. Extraction methods differ mainly in the sampled volume (auger core, monolith or large volume samples) from which roots are physically separated dried to constant weight and finally weighted (Fig. 1):• Method 1: Auger sampling method with 8 cm diameter auger core.• Method 2: Monolith sampling method with 25 x 25 cm quadrat• Method 3: Simplified Voronoï trench excavation, which is the excavation of half a Voronoï quarter• Method 4: Full Voronoï trench excavation, excavation of a Voronoï quarter.

A

B

D

C

E

Method n°2 : Monolith method

A

B

D

C

E

Method n°3 : simplified Voronoï trench

Method n°1 : auger core

method

Method 4: full Voronoï trench

A

B

D

C

E

Method n°2 : Monolith method

A

B

D

C

E

Method n°3 : simplified Voronoï trench

Method n°1 : auger core

method

Method 4: full Voronoï trench

Figure 1. Experimental design around one tree.

Appendix - Root Biomass


Precision of the four methods is given by the confidence interval at 95% (IC95) assuming a Gaussian distribution ( , where σ is the standard deviation and n, the number of samples). The same formula gives also N10% which represents the number of samples to be collected in order to get 10% precision for fine root biomass estimations (Chave et al. 2003). Considering that the full Voronoï trench excavation is the reference value, accuracy of auger cores and monolith methods is calculated as the difference between these methods and the reference one. Sampling and sieving times are reported for each method and transformed in man/day (on a basis of 6 hours of field work /day). Differences between methods are tested with an analysis of variance (proc GLM SAS) associated with Bonferroni’s test for means comparisons.

4. Results

Estimation of fine root biomassThe total fine root biomass was 1.74 t ha-1 with large differences among soil depths (Tab. 1).

Roots density within the soilFor each horizon, we calculated a root density index from the root biomass divided by the thickness of the soil horizon. Fine roots were concentred in 1) root mat and 2) in the first 10 cm of soil. This result was consistent with previous studies demonstrating on an other clone planted in Congo that the surface soil (0-25 cm) was highly prospected by fine roots (Bouillet et al. 2002) (Fig. 2).

Variability of resultsIn the sandy soils of Congo (more than 90% of sand), there is no limitation to root prospection. However, we observed a great spatial variability of fine root biomass as demonstrated in figures 3 and 4. This variability tended to decrease with increasing soil depth.

Structure of the variabilityThere were no significant differences between the four methods in estimating mean fine root biomass, meaning that auger core and monolith methods are as accurate as the full Voronoï trench excavation method. The variability was mainly driven by a tree effect (p<0.0001) for superficial horizon (H1 and H2, 0-50 cm) and a "distance to the tree" effect (p<0.0001) for the deeper horizon (H3, 50-100 cm). In order to sample most of fine roots variability, it is then of major importance to refine basal area classes (for example 6 classes) and to increase the number of sampled trees

nIC σ98.195 =

Table 1. Fine root biomass as a function of soil depth (H1: 0-10, H2: 10-50 et H3: 50-100 cm).

Horizon Mean estimation (t ha-1)

H0 0.56

H1 0.38

H2 0.49

H3 0.31

Total (H1, H2, H3) 1.18

Total + H0 1.74

Figure 2. Roots density as a function of soil depth.

Fine root biomass estimations variability for auger core method

0

5

10

15

20

25

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2ton/ha

Nb

of s

ampl

es

H1 (0-10cm)H2 (10-50cm)H3 (50-100cm)

Fine root biomass estimation variability for monolith method

0

5

10

15

20

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6ton/ha

Nb

sam

ples

H1 (0-10 cm)H2 (10-50cm)H3 (50-100)

0

0.05

0.1

0.15

0.2

0.25

-80 -70 -60 -50 -40 -30 -20 -10 0 10

roots density monolithroots density Auger Coreroots density Voronoï

Soil prospection sampling (cm)

roots density ( t/ha/cm)

Figure 3. Comparison of estimations variability's distribution between horizons for auger core method.

Figure 4. comparison of estimations variability's distribution between horizons for monolith method.



by basal area classes (for example 4 trees/class). Excavation samples should also cover the whole range of distance to the trees.

Cost benefits approachAs a result, it is necessary to sample 312 auger cores (N10%) which is equivalent to 24 trees, and 13 samples by tree to achieve 10% precision on the mean value of fine root biomass for the soil surface layer (0-10 cm). These figures fall down to about 150 auger cores for the 10-50 cm and 50-100 cm layers. We need 6 men/day to sample and sieve soil samples and to achieve this task (Tab. 2). The use of 25 x 25 cm2 monoliths reduces the total number of samples by 100% for H1 (compared to auger cores) but the time required for sieving operations increases by 30% (Tab. 3). Finally, the Voronoï method is the least efficient one (in terms of cost/accuracy) because it requires 100, 50 and 20 repetitions for H1, H2, and H3 and a total of 74 men/day for sampling and sieving operations (Tab. 4).

5. ConclusionsWithin the framework of CDM, this study proposed a methodology to find the optimum sampling design for fine roots biomass assessment in tree plantations. This methodology, based on an elementary space defined by the Voronoï diagram, can be implemented in every forest. It should be applied prior to any belowground C stock assessment in order to find the best choice in terms of sample size and labour time. In Eucalyptus plantations in Congo, fine root biomass was highly variable, with a marked tree effect on superficial horizon (0-50cm) and a "distance to the tree" effect on deeper horizon (50-100 cm). There were no significant differences between the four methods in estimating the total fine root biomass. In these plantations, the best sampling design was then the following : from a stand inventory, 6 basal area classes should be defined and 4 trees/class should be selected (for example by tree height sorting within each basal area class), then 13 auger core samples / tree should be collected in order to get a precision of 10% on the total fine root biomass.

Horizon Fine roots data

Total (t/ha1)

Nb samples IC95 IC95 %

MeansMeans

/10 N10% Time/ man/day

H1Means 0.42 103 0.07 17% 0.042 312

3Table 2. Accuracy and labour time for auger cores.Ecartype 0.37

H2Means 0.55 100 0.06 12% 0.055 134

2Ecartype 0.32

H3Means 0.35 99 0.04 12% 0.035 150

1Ecartype 0.21

Total 1.31 Total 6

H1Means 0.43 64 0.06 14% 0.043 127

3Table 3. Accuracy and labour time for the monolith method.Ecartype 0.25

H2Means 0.45 64 0.06 14% 0.045 125

3Ecartype 0.25

H3Means 0.32 61 0.05 14% 0.032 125

2Ecartype 0.18

Total 1.20 Total 8

H1Means 0.40 16 0.10 26% 0.040 106

11Table 4a. Accuracy and labour time for the Voronoï simplified trench excavation.Ecartype 0.21

H2Means 0.50 16 0.09 18% 0.050 54

7Ecartype 0.19

H3Means 0.29 16 0.03 11% 0.029 20

1Ecartype 0.06

Total 1.19 Total 19

H1Means 0.40 8 0.14 36% 0.040 103

42Table 4b. Accuracy and labour time for the Voronoï full trench excavation.Ecartype 0.20

H2Means 0.50 8 0.13 25% 0.050 50

26Ecartype 0.18

H3Means 0.29 8 0.05 16% 0.029 20

6Ecartype 0.06

Total 1.19 Total 74




CarboafricaBibliographic Archive

CarboAfrica aims to create an archive of a comprehensive bibliography of papers related to Africa, carbon cycle, GHG and Climate Change. Therefore, please send to [email protected] any document, publication, and presentation relevant to the topics mentioned above and CarboAfrica in particular. Then we will put them in the website http://www.carboafrica.net as downloadable documents, or just as references. In any case please let us know if there are any intellectual property rights and/or citation rules to be respected.

Related LinksACE – African Carbon Exchange (ACE) Project http://www.nrel.colostate.edu/projects/ace

AfDevInfo - African Development Information Serviceswww.afdevinfo.com/htmlreports/newsletter_7.aspx

AMMA - African Monsoon Multidisciplinary Analysis http://amma.mediasfrance.org

CARBOEUROPE (Integrated Project CarboEurope-IP,Assessment of the European Terrestrial Carbon Balance) http://www.carboeurope.org

Climate Change and Africawww.climate.org/CI/africa.shtml

EO-LANDEG (Earth Observation initiative in a former homelandof South Africa in support of EU activities in land degradation andintegrated catchments management) http://www.eolandeg.com

ESASTAP - European South Africa Science and TechnologyAdvancement Programme http://www.esastap.org.za/esastap/home/index.php

European Commission - Evaluating protected areas in Africawww-tem.jrc.it/PA/index.html

FIRMS - Fire Information for Resource Management System http://maps.geog.umd.edu/firms

FLUXNET (Integrating Worldwide CO2 Flux Measurements)http://www.fluxnet.ornl.gov/fluxnet/index.cfm

GCP - Global Carbon Project http://www.globalcarbonproject.org

ILEAPS - Integrated Land Ecosystem-Atmosphere ProcessesStudy http://www.atm.helsinki.fi/ILEAPS/

Marien Ngouabi University - University of Brazzaville, Congohttp://www.univ-mngb.net/

NEPAD - New Partnership for Africa's Developmenthttp://www.nepad.org/

PASS - The Pan African START Secretariatpass-africa.org/index.html ROSELT - Réseau d'Observatoires de Surveillance Ecologiqueà Long Termemdweb.roselt-oss.org/index.php?la=eng

SAFARI 2000 Projectdaac.ornl.gov/S2K/safari.html

TCO - Terrestrial Carbon Observation http://www.fao.org/gtos/TCO.html

TroFCCA - Tropical Forest and Climate Change Adaptationhttp://www.cifor.cgiar.org/trofcca

Join the CarboAfrica mailing list server! You will be informed on important incoming events and receive project news, documents and the periodic 4-monthly newsletters.

http://www.carboafrica.net/listserver_en.asp

Please circulate this issue to any interested people.

CarboAfrica NewsLetter N. 5This bulletin is published by FAO (The Food and Agriculture Organization of United Nations) every four months in portable document format (PDF) and distributed free of charge by e-mail and through the CarboAfrica Web site (www.carboafrica.net). If you need printed hard copies please ask by email: [email protected]

AcknowledgmentsThe CarboAfrica NewsLetter N. 4 has been prepared by Antonio Bombelli.Thanks to Claudia Consalvo, Tommaso Chiti, Elisa Grieco, Matieu Henry, Veiko Lehsten, Rudzani Makhado, Lutz Merbold, Bob Scholes, Maria Teresa Sebastià, Reuben Sessa and Paolo Stefani, for their inputs.

ContactsCARBOAFRICA is an international project funded by the European Commission and coordinated by Prof. Riccardo Valentini, University of Tuscia (ITALY).

For any further information: www.carboafrica.netAntonio Bombelli (Project Manager): [email protected]

soil respiration measurements in the kruger national park, south africa

Documents