Memoire Online - Using tree- ring analysis to study the growth performance from saplings to trees for five savanna species in West Africa

4.3.3. Fire frequency and intensity in Comoé national Park and Upper Aguima Catchment 27

LIST OF FIGURES

Figure 2 Macroscopically polished stem discs observation of investigated species Figure 3 The different parts of wood. Stem disc of D. microcarpum

Figure 10 Cross-dating ring width of Diospyros abyssinica oldest sampled tree cut at two different heights

Figure 11 Cross-dating ring width of Detarium microcarpum oldest sampled tree cut at two different heights

Figure 12 Cross-dating ring width of one of sampled tree of Anogeissus leiocapus cut at two different heights

Figure 17 Illustration of compartmentalisation concept on D. microcarpum stem disc

LIST OF PICTURES

Picture 1 Morphological characteristics of Anogeissus leiocarpa Picture 2 Morphological characteristics of Detarium microcarpum Picture 3 Morphological characteristics of Diospyros abyssinica

Picture 4 Morphological characteristics of Isoberlinia doka Picture 5 Morphological characteristics of Pterocarpus erinaceus Picture 6 Some of the analysed stem discs

LIST OF TABLES

DEDICATION

This work was funded by the BIOTA Africa programme (Biodiversity Monitoring Transect Analysis in Africa). It was mainly realized with the technical support of Institute of Agronomy in the Tropics (University of Göttingen) where I analysed my data, the Laboratory of Applied Ecology (University of Abomey-Calavi, Benin) and the Laboratory of Biosciences (University of Rostock, Germany).

ACKNOWLEDGEMENTS

The above biblical excerpt from the prophet Hosea moved me into action to acknowledge the promoters of this work. My warmest thanks are due in the first place to my professors who have shown the very greatest spirit of co-operation and whose competence, understanding and helping have made my first experience on dating of tropical trees species a successful issue.

In this respect, special mention is made of the invaluable contribution of Professor Brice Sinsin, head of Laboratory of Applied Ecology from the University of Abomey -Calavi (Benin) who played a key role in the achievement of this thesis. I could not have achieved this alone without his help. He accepted me into his lab as a postgraduate student and showed me the way to take. His advices and recommendations enabled me to overcome successfully. I must acknowledge him with deepest gratitude for the excellent advice, encouragement and guidance.

I am also particularly indebted to Professor Stefan Porembski and his assistant Dr. Bettina Orthmann from the University of Rostock (Germany) who accepted me into their laboratory and taught me the methods of tree dating. Dr Bettina was never far from me. First, she facilitated the process of my trip. Secondly, she introduc ed me to the specialist of tree-ring dating, helped me to improve my research proposal and offered me a pleasant working time. Thirdly, she read the entire work and helped me once more to improve it. I miss words to tell her my acknowledgements as she deserves.

I should express my respectful thanks in special place to PD Dr Martin Worbes for his simplicity, helpful and teaching during the two months I worked with him. I got tree-ring dating in his lab.

I could never forget Dr Romain Kakaï Glèlè, Dr Klaus Hennenberg, Miss Sofie Blanchart and a very large number of my graduate colleagues (Promotion 31) for the fruitful exchanges.

Finally, I have to render my sincere thankfulness to my family and Miss Camen T. Adjadémè whose fondness and love helped me to bring to a successful conclusion.

TECHNICAL EXPRESSIONS

Thin layer of meristematic tissue that lies between, and gives rise by Cambium: active division to, secondary xylem on the inside and secondary phloem
on the outside

Compartmentalisation: A dynamic defence process in an injured tree which forms structural and chemical boundaries in order to resist the spread of pathogens

Cross-dating: The procedure of matching variations in ring width or other ring

characteristics among several tree ring series, allowing the identification of the exact year in which each tree ring was formed, or th e geographic origin of the tree

Master fire chronology: Chronologies of fire dates constructed from the fire scar information derived from all samples within a collection area

Outermost ring: The most recently formed tree ring still visible on a wood sample

Parenchyma: Mostly rectangular, storage cells alive in the sapwood, with simple

Parenchyma band: Parenchyma forming one or more tangential lines within a tree ring

Radius: On a stem disc, measuring line between the pith and the outermost ring

ABSTRACT

Dendrochronology was used to study the wood anatomy and growth performance of five West African savanna species namely Anogeissus leiocarpa (DC.) Guill., Perr., Detarium microcarpum Guill., Perr., Diospyros abyssinica (Hiern) F. White, Isoberlinia doka Craib., Stapf., and Pterocarpus erinaceus Poir. This technique was also used to reconstruct the fire past in studied areas that are Comoé National Park in Côte d'Ivoire and Upper Aguima Catchment in Benin. Tree-ring analysis (Dendrochronology) through 72 stem discs collected from two to three different heights of trees and using cross dating approach helped to better understand the biology of wood. Using these data, we categorized the five s pecies in different groups according to their wood anatomy and radial growth performance. Then, we determined, described the inter and intra specific variation of the growth rate. Finally, we compared the two studied areas among their seasonal fire interva l, its frequency and intensity. So, species such as Detarium microcarpum Guill., Perr., and Isoberlinia doka Craib. were identified as fast growing species with brown wood. The rings boundaries were easily distinguished and formed by marginal parenchyma ba nds. Wood of Anogeissus leiocarpa (DC.) Guill., and Pterocarpus erinaceus Poir. were white or yellow in colour. The rings showed a variation in vessels distribution and the growth performances were also good. Diospyros abyssinica (Hiern) F. White was demonstrated like a slow-growing species that wood's colour usually varies between white and yellow with black streaks. The rings are narrow and not identifiable macroscopically. However, they were easily measured with high accuracy. The rings boundaries were presented in single concentric line and characterized by patterns of alternating parenchyma and fibre bands. Finally, both studied sites were seriously disturbed by using of several consecutive fires that caused the injuries in tree and ecosystems.

Key words: Tree-dating, Dendrochronology, wood anatomy, Bush fire, Savanna, Upper Aguima Catchment (Benin), Comoé National Park (Côte d'Ivoire)

RESUME

La présente étude a porté sur l'utilisation de la dendrochronologie (Analyse des cernes d'accroissement) pour comprendre l'anatomie et le rythme de croissance de cinq espèces végétales caractéristiques de la région savanicole de l'Afrique de l'Ouest. Cet outil a été également utilisé à partir de la provenance du matériel d'étude pour retracer l'historique des feux de végétation dans cette région et précisément dans le Parc National de Comoé en Côte d'Ivoire et dans la forêt claire des Monts Couffé du Bénin. Les objectifs poursuivis ont été de documenter en premier lieu l'anatomie du bois pour les espèces Anogeissus leiocarpa (DC.) Guill., Perr., Detarium microcarpum Guill., Perr., Diospyros abyssinica (Hiern) F. White, Isoberlinia doka Craib., Stapf., et Pterocarpus erinaceus Poir. Dans un second temps, il fallait déterminer l'âge des arbres échantillonnés par l'analyse des cernes puis étudier la variabilité des performances de croissance entre ces différentes espèces. Enfin, il était question d'utiliser la méthode pour fournir des informations sur l'écologie et la dynamique des habitats de ces espèces, en particulier, l'impact des feux de végétation sur la biodiversité. La méthode utilisée est basée essentiellement sur l'observation à l'oeil nu, à la loupe, au microscope et au scanner des cernes d'accroissement qui ont été comptés utilisant la technique «cross da ting» via le logiciel TSAPWin à partir des rondins de bois coupés à deux différentes hauteurs. L'anatomie du bois a pu être reconstituée par une combinaison d'outils : microscope associé au Lintab avec le logiciel Leica et parfois le scanner. Les résultats obtenus ont permis de catégoriser les espèces suivant les caractéristiques anatomiques et les performances de croissance. Ainsi les espèces comme D. microcarpum et I. doka sont identifiées comme des espèces à bois bruns présentant des cernes dont les limi tes sont facilement observables à l'oeil nu et caractérisées par une bande de parenchyme marginale. Les arbres de ces espèces ont une croissance relativement bonne. A. leiocarpa et P. erinaceus sont des espèces à bois variant entre le blanc et le jaune et présentant des cernes avec une variation dans la distribution des vaisseaux. Les performances sont relativement assez bonnes. D. abyssinica présente des cernes rétrécies, difficilement identifiables aussi bien à l'oeil nu qu'à la loupe. Cependant, les limit es des cernes ont pu être appréciées au microscope à un fort grossissement (x 40). Elles sont caractérisées par une alternance de bande de parenchymes. Les deux zones d'études ont été enfin reconnues comme des milieux subissant de fortes pressions anthropi ques par une utilisation régulière et fréquente des feux de saison sèche. Ainsi l'utilisation des pare -feux pour préserver les zones à grand risque s'avère nécessaire.

Mots clés: Datation, Dendrochronologie, Anatomie du bois, feu de végétation, savanes, Afrique de l'Ouest, Monts Couffés (Bénin), Parc National de Comoé (Côte d'Ivoire)

1. INTRODUCTION

1.1. Background on tree dating methods

Knowledge of the age of trees has a number of implications. Firstly, tree dating, combined with knowledge on stand structure can give information about forest disturbance. Therefore, it is often used to study the forest dynamics (Mundo et al., 2007). It also prevents the loss of genetic diversity and allows foresters to develop sustainable harvest pr actices (Roel, 2005). The knowledge on age and increment growth of the trees also sheds new light on global climate models (Gerhard et al., 2004; Walter, 2004; Bouriaud et al., 2005; Schöngart et al., 2006; Bütgen et al., 2007; Thomas, 2007). To estimate tree a ge, scientists use two major methods. The first is the relative dating that regroups the periodic annual increment (PAI) method and the crown class model (Backer, 2003). These methods are not always accurate tree dating. They are based on diameter growth without taking in consideration that many other factors like soil fertility could influence tree-growth. Therefore, they can overestimate or underestimate the age of trees.

To address this deficiency, dendrochronology also called Tree-ring analysis was proposed in 1901 by Ellicott Douglas (1867-1962). He was the first to remark that each year, trees add a layer of wood to its trunk and branches and then producing annual rings. This approach has a lot of advantages for forest management studies (Brienen , 2005). First, tree-ring analysis gives information on the real age of a tree and the lifetime growth rates and is therefore more effective than relative methods. It can be used to reconstruct past disturbance (Brienen et al., 2007; Patrick et al., 2008). For example, tree-rings are often used to reconstruct fire history from fire scars (Welsberg & swanson, 2001; Guyette & Stambaught, 2004; Van Horne & Fule, 2006; Hall, 2008). Another advantage of the ring analysis is the possibility it offers to quantify variation in growth among individuals over long periods of time (Desta et al., 2003). Tree-ring analysis is also a good complementary tool to permanent sample plot measurements. It needs only short time to provide many d ata on tree life. Finally, tree-ring is used to reconstruct atmospheric gas concentration over the past (Kennichi et al., 2004; XingYun et al., 2006; Kristopher et al., 2007; Louise et al., 2008).

Despite all these advantages, this method has sometime s limited applications in the tropics
because of invisible rings in certain species (Pascale et al., 2004). Also, it is only useful for
trees that are less than 600 years old (Worbes, 2002). However, it has been demonstrated in

recent reports that some trees live more than 1000 years (Miguel & Elena, 1998; Patrut et al., 2007).

Therefore, an alternative absolute dating method like Radiocarbon is necessary. This approach was developed by Willard in 1946. The radiocarbon method is more effective than all other approaches (Ramsey, 2007). The dating method was used to explain some ecological and paleoclimatological phenomena (Anouk et al., 2004; Patrut et al., 2007). But its application requires well equipped laboratory and thus, it is more expensive. According to Worbes (2002), it is better to use dendrochronology when tr ees show visible rings. Many recent studies have also reported the presence of annual ring in tropical species that may allow the use of tree-ring analysis as dating tool (Miguel & Elena, 1998; Worbes, 2002; Dezzeo et al., 2003; Fichtler et al., 2006; Patrut et al., 2007).

In the present study we established the growth performance from sapling s to trees for five savanna species (Isoberlinia doka Craib and Stapf, Pterocarpus erinaceus Poir., Anogeissus leiocarpa (DC.) Guill. & Perr., Detarium microcarpum Guill & Perr., and Diospyros abyssinica (Hiern) F. White) of West Africa. All of these species are found in open natural stands that were subject to periodical bush fire. The main research question in these savanna stands is related to the time the trees species, especially the endangered ones needs to become fire-resistant. All of the five targeted species showed visible annual tree -rings (Nigärd et al., 2004; Tarhule & Leavitt, 2004; Poussart et al., 2006; Shöngart et al., 2006) and allowed the use of tree-ring analysis for the study.

1.2. Interest of the study

Savanna ecosystems of West Africa including those of Benin and Côte d'Ivoire are overexploited. Agriculture by burning, village hunting, overgrazing and fire misuse increased the fragmentation of species habitats with increasing shortage of biodiversity and climate change as direct feedback. Fire misuse profits just within short time only for its users but its disadvantages span over a longer time for all mankind. This threat on sudanian resouces lead to the disappearance of plant and animal sp ecies. The conservation of the endangered species should require a great description and analysis of their habitats. For that purpose, research on the population dynamics of tree-species in their ecological biotop e become important to understand destructive process of anthropogenic pressure on the habitats of species. Tree species like A. leiocarpa, I. doka, P. erinaceus, D. microcarpum, and D. abyssinica are characteristics of the savanna areas from Upper Aguima C atchment (Benin) and Comoé

National Park (Côte d'Ivoire) and some of these species are endemic to the sudanian region depending the phytogeographical classification of White (1983). Therefore, this research aimed to trace the fire history from fire scars o bserved on the tree-rings. Thus, the frequency of fire use in these studied areas and the number of years required for these five species to become fire-resistant will be determined.

1.3. Aims

The study aimed at contributing to a better management and conservation of West African tree species on the basis of their annual growth. Therefore, five specific objectives were defined:

> to determine the growth performances of sampled species in their biotope ; > to synchronize fire scars for each of these species;

1.4. Research questions

> how long do tree species need to reach the height of 1.3 m chosen as reference for dbh measure?

> which one of the studied site experienced more treated by fire year round events?

2. STUDY AREA

The working material (sample tree discs) came from the Comoé National Park (CNP, Côte d'Ivoire) and the Upper Aguima Catchment (UAC, Benin). The CNP is situated between 8° 41'- 8° 44' N and 3° 47'- 3° 51' W. The UAC is located between 9° 12'- 9°15' N and 1° 90- 1° 92 E.

Comoé National Park is comprised in interfluvial peniplain of schist and granite with a mean altitude of 250 m to 300 m. The geological subsurface of Upper Aguima Catchment is granite or gneiss with typical ferralitic soils (Orthmann, 2005).

The two sites are characterized by alternating rainy and dry season with mean annual rainfall and temperature of about 1150 mm and 26.5°C - 27°C respectively (Orthmann, 2005; Hennenberg, 2005). Dry period occurs from November to February (Figure 1). The CNP is in the borderline between the centre and the Guinea -Congolian / Sudanian transition region (White, 1983 in Schöngart 2006). Its annual rainfall varies from 856 to 1248 mm (F ischer et al., 2002 in Hennenberg, 2005) and the temperature fluctuates from 10°C to 40°C following seasons (Hennenberg, 2005; Schöngart, 2006). The other one site is in Sudanian regional centre of endemism (White, 1983). Its temperature ranged less than 15°C to more than 40°C among periods. Figure 1 shows the rain fall trend per month for the studied areas.

Figure 1: Mean monthly rainfall for studied sites (CNP and UAC). Period from 1960 to 2001

The dry season of about four months is more remarked in Comoé National Park than Upper Aguima Catchment. However, during the rainy season less precipitation can be observed at UAC from May to September than in the CNP (Fig. 1). This could influence differently the plant species growth.

The Upper Aguima Catchment site is located in Sudanian regional centre of endemism following White's phytogeographical classification of Africa. The vegetation of the site is therefore mapped as savanna and open forest. It is characterized as undifferentiate d woodland that included trees with an understory of grasses, shrubs and herbs. The trees are mainly deciduous in the dry season. Typical tree species are A. leiocarpa, Acacia seyal, Kigelia africana and species of Combretum and Terminalia genus. In this ecoregion, the dominance of I. doka was also noted (White, 1983). As far as the Comoé National Park is concerned, it is classified in IUCN category II and located in two different phytogeographical regions. The south-western part of the park is in the Sudanian zone while the northern part is in Guineo -

Congolian/Sudanian regional transition zone. The vegetation of this transition zone is described as mosaic of dry, peripheral, semi -evergreen rainforest and woodland and secondary grassland. Thus, it is noted the presence of some savanna tree species like D. microcarpum, A. leiocarpa, Daniellia oliveri, I. doka, P. erinaceus etc...

3. MATERIAL AND METHODS

3.1 Studied plant species and collection of samples

This research aimed to establish the growth performance for five tree species from four different wood species families (Table 1). The Samples were collected in 2002 in the Comoé National Park (Côte d'Ivoire, Hennenberg, 2005) and in Central Benin (Orthmann, 2005). At CNP, 10, 8 and 8 samples were respectively collected from D. microcarpum, A. leiocarpa and D. abyssinica. A. leiocarpa was collected at the forest border, D. microcarpum in the savanna and D. abyssinica in the forest. In Benin, six (6) samples were also collected from I. doka, P. erinaceus and A. leiocarpa in open mosaic forest.

TABLE 1- The five tree species investigated in two different sites. UAC: Upper Aguima catchment, CNP: Comoé National Park; Cutting levels (level 1: cut at 0,1m - level 2: cut at 1,3m) and N (tree) means the number of sampled trees.

The african birch species named Anogeissus leiocarpa (DC.) Guill. & Perr., belong to the family of Combretaceae and ranged from the borders of the sahara up to the outlier humid tropical forests and is found from Senegal to Cameroon in West Africa. The tree of about 20 m of height with light green foliage grows in dry forest s, fringing forests and semi aris savanna areas. The leaves, bark, roots and seeds serve for traditional tanning and its wood is used for house construction and tool handles because of its resistance against insects (Seed leaflet, 2007). The picture 1 shows some of morphological characteristics of the species.

Picture 1a: Flowers - Picture 1b: aspect of the stem - Picture 1c: the tree (Brunken et al., 2008)

Detarium microcarpum Guill. & Perr. from the family of Caesalpiniaceae typically found in high rainfall savanna areas, dry forests and fallow land, on sandy or iron rich hard soils. It can grow up from 10 m to 25 m in height. D. Microcarpum tree is also found in open savanna as a more stunded tree with smaller fruits as showing the picture 2. It has great fruit-bearing potential and is also used as medicinal plant.

Picture 2a: the fruits - Picture 2b: the aspect of the bark - Picture 2c: the tree (Brunken et al., 2008)

Diospyros abyssinica (Hiern) F. White has small, medium or large tree up to 36 m with a
relatively sparse and shortly branched crown. The leaves are altern ate, glossy dark green and
distinctly red when they are young. Its ecological distribution is known as widespread in

tropical Africa from the Guinea Republic to Eritrea southwards to the FZ area an d Angola. It is known for its medicinal use and used to make pestles. The picture 3 shows at left the leaves and at right its trees in forest stand.

Picture 3: Morphological characteristics of Diospyros abyssinica Picture 3a: a leave - Picture 3b: the tree (Brunken et al., 2008)

Isoberlinia doka Craib Stapf is one of several sudanian species of the family of Caesalpiniaceae found in deciduous woodland, tall grass savanna alway s in stands forming open forests with Isoberlinia tomentosa, Burkea africana, Prosopis africana, Uapaca somon etc... The tree has 10 to 20 m of height and the trunk straight 5 m height. The wood is used in marine construction, naval architecture, in paper products, as pulpwood and for office materials. The aspect of the foliage is presented in figure 4.

Picture 4a: the fruits - Picture 4b: the aspect of the bark (Brunken et al., 2008)

Pterocarpus erinaceus is a tropical African deciduous tree of 12 to 15 m (or more) that can be found in open stands, woodland and savanna. The bark is fissured as showing the picture 5.

Its wood is useful for house construction, in carpentry for doors and windows frames. It makes very good charchoal and decorative things.

Picture 5a: the fruits - Picture 5b: the aspect of the bark - Picture 5c: the tree (Brunken et al., 2008)

The dendrometric measurements such as diameter at 10 cm from grou nd, diameter at breast height (DBH) and the total height were made on some of the sampled trees by Orthmann (2005) and Hennenberg (2005). Their mean values were presented in table 2. The cross-sectional wood discs were cutted at 10 cm from ground level and at breast height (DBH) for all tree samples from Ivory Coast. In Benin, at least one stem disc per tree was cutted.

TABLE 2- Some characteristics of the investigated species. MD (1): mean diameter at 10 cm from ground, MD (2): mean DBH, MNR: mean number of ring of samples cutted at 10 cm from ground, MH: mean height and s.d: standard of deviation.

The stem discs were labelled and conducted in laboratory (University of Göttingen, Germany). It is useful to note that the most of Comoé National Park samples were really cutted at two levels of height (10 cm from ground and 1.3 m). The missing diameter values were added after measurements on these stem discs with electronic forest compass (Pictures 6 and 7).

Picture 6: Some of the analysed stem discs Picture 7: The used electronic forest compass

3.2 Sample processing

The stem discs were polished successively with different bands using a polishing machine (Picture 8). The band of 80 um was first used and followed by grains of 120, 240, 400 and 600 um. After this step, the wood dust was cleaned from sample surface with electrical air pump. These applications helped us to make good macroscopic observation s of the rings boundaries (Figure 1). The different rings of each disc sample were then marked using a pencil. A magnifying glass was sometimes used to identify the ring when the macroscopic observation was not easy.

3.3 Method used to study the wood anatomy

Before describing the wood structure, we defined some criteria that were mostly based on wood structure following cell types: vessel elements and its distribution, disposition of rays and parenchyma cells. For that, we referred to Coster (1927, 1928), Worbes (2002) and Shöngart (2006). Following these criteria, the macroscopic and microscopic observations were done using a scanner and the Leica software with magn ifying glass and microscope.

3.4 Method used to study the growth performance s

The growth performances were mostly based on the age of trees, ring increment and the initial growing in height. They helped to assess the differences between the species. To determine the age of each sampled tree, the rings were easily counted visually. However, all measurements (tree ring counting and ring width measurement) were done with high accuracy using electronic microscope and Lintab (Picture 9). The number of rings identified on wood discs cutted at 10 cm from ground indicates after a cross dating, the age of specified tree. The number of years needed by the targeted trees to reach the breast height (1.3 m) was determined making a difference between the numbers of observed rings on two stems discs (the stem disc cutted at 10 cm from ground and the one cutted at 1.3 m) from a same tree completed by cross dating.

Tree-rings widths were measured using the software TSAPWin by digital measuring device Lintab 5, Factor: 1, 00 and using 1/100 mm as Length unit, COM port 4 and PC Mouse. To compare the variation in growth performances between species, the data on the ages and mean annual ring increment (MRI) were analysed with statistical approach. The test of ANOVA followed by Student-Newman-Keuls was used after a logarithmic transformation to stabilize the variance and normalize the data.

3.5 Cross-dating

Cross dating of time series was used for the verification of the series and the elimination of possible errors and to find the correct dated position in time.

It helped in the elimination of measurement errors, e.g. the removal of «false rings» and the insertion of «missing rings». TSAP-Win offered a combination of both visual (graphical) and statistical cross-dating. Statistical models are excellent tools to find possible matches or to verify the dates of pre-dated time series.

In dendrochronology two main concepts are used to express the quality of accordance between time series: Gleichlaeufigkeit and/or t -values. While the t-statistic is a widely known test for mean difference significance, Gleichlaeufigkeit was developed as a special tool for cross-dating of tree-ring series (Eckstein & Bauch, 1969). These concepts are characterized by a different sensitivity to tree-ring patterns. While Gleichlaeufigkeit represents the overall accordance of two series, t-values are sensitive to extreme values, such as event years. A combination of both is realized in the Cross-Dating Index (CDI) that was used in this study

to obtain the exact year in witch each tree ring was formed and also for matching variation in ring width among several tree rings series. To identify real age of each tree, the cross -dating was done between the two series time of the stem disc cut at 10 cm from ground and the other one felt at 1.3 m of height. However, for samples cutted only in one level of tree (level 1), it was done between different trees of the same species and from the same stand.

3.6 Method used to analyse fire regime

According to Multilingual Glossary of Dendrochronology, fire regime means the combination of fire frequency, intensity, size and seasonality that determine the role of fire in a given ecosystem. When a bush fire passes, the trees of this specified site are injured forming the fire scares that are often observed on cross wood section. In the present study, these fire scares were used to assess the fire interval, its frequency and intensity for comparing both studied sites.

Fire interval means the number of years between two consecutive fi res scars in a specific cross-wood section.

Fire frequency differs from fire interval in the sense that it means the number of fires per year in a given area. To determine this, we examine d fire season by their intra annual position in the fire scars using high accuracy. The scars observed in ring boundary were formed during dry season and those that were observed in ring growth zone occurred during the growth season (rain period).

To assess the intensity of each fire, all the stem discs were examined. For individual tree, a fire is qualified as low intensity when the scar was observed only in sample cut ted at 10 cm from ground. When the same fire was identified from both samples of height (10 cm and 1.3 m) for a same tree, it was considered as a high intensity fire. However, we carefully completed this analysis with the number of tree that recorded the fire and sometimes we analyzed the season of fire that allowed confirming the observations. We must precise that most discs samples were cutted from two levels in Comoé National Park that is not the case in Upper Aguima Catchment where the most stem discs were only collected at 10 cm from ground.

3.7 Nature of data

Quantitative and qualitative data were obtained from the stem discs analysis. The qualitative
data were obtained through macroscopically and micr oscopic appreciation of the wood

colour, the analysis of rings structure, the distribution of vessels and parenchyma that helped to describe the wood anatomy of the species. The position (year) of the identified fire scares on cross wood was also a qualitative data that allowed the determination of the years of bush fire from sampled trees.

The quantitative data were the number of tree rings per stem discs, their widths, the number of scarred trees per species and the number of rings scarred on each stem disc. To do this, it was necessary to prepare the stem discs by their processing.

4. RESULTS

4.1 Wood anatomy of the species

All dendrochronological investigations of the study were based on the hypothesis that wood structure varies considerably following the species. The Figure 2 shows the difference between the species in wood colour, the aspect of the bark and the distinctiveness of the rings boundaries depending of the cross-wood sections which were used for analyses. Specifically, the figure 3 and 4 were chosen for the description of the common legend of the wood anatomically.

Figure 2: Macroscopically polished stem discs observation of investigated species. Fig a. - Fig

b. D. microcarpum (Caesalpiniaceae) - Fig c. D. abyssinica (Ebenaceae) - Fig d. I. doka (Caesalpinaceae) - Fig e. P. erinaceus (Fabaceae).

4.1.1. Anogeissus leiocarpa

The bark of A. leiocarpa is slightly furrowed with dirty white drawing sometimes ashy in colour. The sapwood is yellowish and the difference bet ween heartwood is not so remarkable in young tree.

From pith to cambium, we observed the thin radiu ses which are very near one another. The growth type zone shows variations in the vessel distributio n (Figure 5). The border of ring is presented like single circular line. For the samples of UAC (Benin), it was not always easy to identify macroscopically the rings boundaries when they are narrows. But, all the CNP samples showed distinct tree-rings by visual observation. Only in stem discs from CNP, an alternating of early wood and latewood was identified. The latewood s are very hard and darker. The distinctiveness of ring boundary in A. leiocarpa trees depends also on the ring size and the environmental conditions. A. leiocarpa, samples collected from two different sites showed variation in ring visibility. The samples from CNP have best distinct rings that may be explained by month rainfall diagram. In this area, the dry season occurs from November to February like in Benin site (UAC) but it is more marked. Thus, A. leiocarpa has high tree-ring sensitivity. It is strongly influenced by changes in moisture conditions. A. leiocarpa trees react strongly to environmental factors.

4.1.2. Detarium microcarpum

The outer bark was moderated furrowed and greyish with reddish inner bark. The wood of D. microcarpum is hard and dark brown in colour. The sapwood was lighter than heartwood. The polished stem disc showed good distinct tree-rings (Figure 6).

There are many parallel rays mostly in sapwood which are perpendicular to tree rings. In these rays, we identified radial parenchyma cell s. The small vessels are distributed in all sapwood. No vessel has been noticed in heartwood. The pith is characterized by spongy tissue. Tree-rings are darker in colour and were formed by consecutive vessels in single line for heartwood. The vessels of rings boundaries are bigger. Two successive vessels are separated by one radius. The radiuses are wide and the growth border was characterized by marginal parenchyma bands. In sapwood, we did observe that the growth ring boundary was delimited by single concentric line. The number of vessels in growth ring border decreases from heartwood to sapwood. Sometimes, no vessel was noticed in concentric line.

4.1.3. Diospyros abyssinica

The bark of D. abyssinica is black spotted white in colour. The sapwood is slightly different from heartwood in colour.

The wood colour is usually between white an d yellow and the heartwood presents black streaks (Figure 7). The identification of tree-rings is not always easy macroscopically. They are narrow and are presented like single concentric line. In the growth zones, no vessels were observed. The boundary zones are characterized by patterns of alternating parenchyma and fibre bands.

4.1.4. Isoberlinia doka

I. doka tree has moderately furrowed bark whose colour varies from white to brown. In the studied samples, there were no significant colour differences between heartwood and sapwood. The wood is usually brown (Figure 8).

Tree-rings are darker and wide. They appear like concentric bands. Thus, macroscopically the boundary zones are distinct and are formed by tangential lines within a tree ring. In wood cross section, vessels of fairly uniform are distributed throug hout a growth ring (diffuse-porous). Therefore, Tree-rings are characterized by marginal parenchyma bands. All vessels are housed in storage cells (radial parenchyma) that are lighter and show a parallel disposition to the boundary zone. In each radial par enchyma, we identified one to three vessels. The radial parenchyma was mostly seen in sapwood. The Figure 8 shows visible annual ring in I. doka wood stem disc. In the central core of this stem we noticed a small hole that demonstrates the unlignified cell walls of pith.

4.1.5. Pterocarpus erinaceus

P. erinaceus bark is deeply furrowed with dark mahogany in colour. The wood is yellowish and no difference was noticed between sapwood and heartwood (Figure 9).

The observed structure through high accuracy showed a decreasing of vessels size towards the tree-ring. Then, just after one ring, the vessel is wide and the size decrease s gradually until the next boundary growth, thus a clear distinction of the rings. The boundary growth has circular slightly undulating form. The presence of alternating fibres and parenchyma tissues is remarkable.

4.1.6. Conclusion on wood anatomy

For each investigated species, the wood anatomical structure showed a variati on from one to another. We also identified three different tree-ring structures:

> border of rings presenting variation in vessels distr ibution that was the case of A. leiocarpa (Combretaceae) and P. erinaceus (Fabaceae) species;

> border of rings delimited by marginal parenchyma bands which was represented by D. microcarpum (Caesalpiniaceae) and I. doka (Caesalpiniaceae) species;

> growth ring boundary like alternating bands of fibre and parenchyma cells illustrated by D. abyssinica (Ebenaceae).

The visual analysis demonstrated that there are differences in tree-ring structures among the species. Thus, genetic impact is questionable.

On the other hand, throughout the wood colour, we categorized the targeted species in three major groups:

> light brown to dark brown wood that are the case of D. microcarpum and I. doka. Both species are in the same family of Caesalpiniaceae .

Finally, about the distinctiveness of ring borders, we conclude that I. doka, D. microcarpum, and P. erinaceus have the best distinct rings. As far as the D. abyssinica is concerned, it showed annual tree-rings but the use of high accuracy method is necessary.

4.2. Age and growth performances for investigated species

4.2.1. Age, ring increment and cross-dating

The results of the cross dating process are summarized in table 3 and show ed a great difference between the species depending of the genetics characteristics and the growth conditions of the a tree.

TABLE 3 - Age and MRI of the investigated species. MRI= Mean Ring Increment values. SNK = Student - Newman - Keuls grouping of the species according to age and MRI (logarithmic transformation was applied before the ANOVA test). Values with the same letter are not significantly different. m= mean; s= standard deviation and do: error margin for 95% of probability.

The youngest sampled species was A. leiocarpa from UAC followed by the group of I. doka and P. erinaceus and the oldest was D. abyssinica. The annual radial growth is great for young trees and low for the oldest. These findings proved that the age and the mean annual ring increment varied from one species to another, between trees and also from one stand to another. According to these results, A. leiocarpa trees from UAC, the youngest trees, showed the best mean annual radial growth that was twice more than those of CNP. In fact, the targeted trees of UAC (Benin) were between 5 to 7 years old and all had 1.8 m of height. The diameters at 10 cm from ground vary between 4 cm and 5 cm. Those of CNP were 16 to 19 years old in 2001. Their height varied between 5.6 m to 7.1 m and the diameters varied between 6.5 and 7.5 cm. D. abyssinica, the oldest investigated trees showed the lowest radial growth witch is a characteristic of the species. This variation of annual growth within the same tree is observed in figure 10.

Figure 10: Cross-dating ring width of D. abyssinica oldest sampled tree cut at two different
heights. The yellow curve was obtained from sample cut at 10 cm from ground and the ot her one was for sample cut at 1.3 m from ground. Axe X describes the year and axe Y presents the widths of rings

The information about the lifetime of the oldest sampled tree is present in figure 2. This figure was obtained using microscopic, TSAPWin software in combination with lintab. The raw ri ng widths were cross-dated. This curve shows the variation in annual ring increment from one year to another. The tree germinated in 1969 and formed its first ring in 1970. The highest ring width was observed in 1982 and the lowest was noticed in 1973 that would probably be due to rainfall effect on tree growth in tropical area. T he tree reached the height of 1.3 m in 1984 and proved the low growth rate of D. abyssinica trees. The cross-dating index (CDI) is 23.5 (CDI > 10). This shows that the test of cros s-dating is very significant with the probability of 99 %. The t-value (25.1) is greater than 3 and demonstrates the high similarity between the two curves. Then, the statistic test proved a similar radi al rate growth from ground to 1.3 m in D. abyssinica trees. The same remark was made for D. microcarpum

Figure 11: Cross-dating ring width of Detarium microcarpum oldest sampled tree cut at two
different heights. The yellow curve was obtained from sample cut at 10 cm from ground and the white curve was for sample cut at 1,3 m from ground. X describes the year and axe Y presents the widths of rings

The oldest sample tree of D. microcarpum germinated in 1984. It reached the DBH after eight years of life. The greater ring increment was observed during the first year. The cross-dating test is significant. The similarity between the two curves is also high.

Figure 12: Cross-dating ring width of one of sampled tree of A. leiocarpa cut at two different

heights. The yellow curve was obtained from sample cut at 10 cm from ground and the white curve was for sample cutted at 1.3 m of height. X describes the year and Y presents the width of the rings.

This tree produced its first ring in 1985. It reached the height of 1.3 m in 1992. The cross - dating index is significant and the similarity between the two curves is high.

4.2.2. Growing in height

The height growth was used to estimate the number of years needed by each species to be
outside fire events. Table 2 presents the total height of different investigated trees and the
figure 13 informs about the number of years needed by each species to reach the height of 1.3

Figure 13: Initial height growth for some of investigated species. For all figures, the variable Age of abscise means the number of years. (a) A. leiocarpa - (b): D. Microcarpum - (c): I. doka - (d): P. erinaceus.

In these figures, axes X and Y describe respectively the absolute frequency and the number of years. They show that the height growth varies among species. Some savanna species such as A. leiocarpa, D. microcarpum and I. doka need between 2 to 8 years to reach the height of 1.3

m. However, D. abyssinica needs more time to reach the same height of 1.3 m. In fact, this species grows slowly in wide and also in height. It needs long time to be outside fire events.

4.2.3 Conclusion on tree species growth

A. leiocarpa, D. microcarpum and I. doka can be categorized as fast-growing savanna species while D. abyssinica is a slow-growing species. These observations are in accordance with those of Worbes (1989). If we refer to anatomical characteriza tion described in our results combined with the findings of Worbes (1989), we c ould assume a relationship between wood colour and the growth performances. But, before confirming this hypothesis, further studies on several savanna wood species are necessary.

4.3. Fire regime analysis from tree -rings of investigated species

4.3.1. Fire master chronology in CNP and UAC

The analysis of fire scars on investigated stem discs with dendrochronology techniques helped to have the master fire chronology in both sites as experienced by sampled tree that was presented in Figure 14 and 15.

Forest border (A. leiocarpus, N = 8) Savanna (D. microcarpum, N = 10) Forest (D.abyssinica, N = 8)

Figure 14: Fire history in Comoé National Park as experienced by sampled trees. In Common

Figure 15: Fire history in Upper Aguima Catchment as experienced by sampled trees.

4.3.2. Fire interval

In Comoé National Park, 19 different fire dates were identified. The earliest fire of 1976 was recorded on two D. abyssinica trees (D1-105 and D1-108). The first tree germinated in 1971 and has only one fire scars and the other one germinated in 1974 with two different fire scars. The recent fire in this park was recorded in 2001 on D. microcarpum (DE-103; DE-105) and A. leiocarpa trees (AN-103; AN-105). The highest number of fire scars was observed on D. abyssinica tree (D1-104) that had five different fire scars (1981, 1985, 1987, 1990 and 1 992) followed by one of A. leiocarpa tree (AN-104) which was scarred at four different years (1988, 1991, 1995 and 1999).

Specifically, D. abyssinica trees that were collected from forest area in CNP allowed us to date fire from 1976 to 1997. During this p eriod, the fire interval varied from one to four years. Most of sampled wood of D. abyssinica had their youngest fire scars before 1993. In 1997, only two samples were scarred by fire. After this year, no fire scars was observed on sampled cross wood of this species, indicating that fire experienced by the sampled tree in forest of Comoé National Park was restricted from 1997. This area was probably protected against fire by technical forest management disposition.

About fire history in forest border, we estimated it from A. leiocarpa samples from 1985 to 2001, the year before collecting samples. It was estimated to be between four and six years until 2001. Thus, we conclude that forest border of CNP was always treated by fire. A. leiocarpa was described in CNP by Klaus (2006) as pioneer species. Thus we suppose that the forest border was cleaned for fire breaks building. This observation confirms the hypothesis that from 1997, forest area of CNP was protecte d against fire. Its effectively was proven since 1997.

Concerning savanna area of this park, the fire past was reconstructed from D. microcarpum species. All sampled trees for this species were originated after 1985 and the earliest fire date in this area was from 1992. From this year to 2001, the fire in terval varied between one and five years. That could be a positive management action either for ensuring grazing for herbivorous mammals or as to facilitate tourism. However, the regular use of fire in CNP savanna area could also be due to the destructive pressure of poachers during game searching.

In upper Aguima Catchment, the oldest tree collected from this second studied area was twelve years old in 2002. The fire interval was from 1995 to 2001 with four different dates (1995, 1998, 1999 and 2001). On any of samples, we identified more than one fire scar. However, like all samples were collected in open mosaic forest, we can reconstruct fire history in this area by dating the fire scar for each sa mple as shown in the figure 9. Thus, fire interval in open mosaic forest of Upper Aguima Catchment was estimated between one to three years. However, the all estimated fire interval depend of the sampled tree that could escape fire event during the supposed -non fire period even if bush fire occurs on year round cycle.

4.3.3. Fire frequency and intensity in Comoé National Park and Upper Aguima Catchment

In Upper Aguima Catchment, the fire frequency ranged from 1 to 3. In 1995, 1998 and 2001, the fire was used only during dry season but in 1998, three different fires were used consecutively (dry season, rain season and at the end of rain season). In CNP, the fire frequency ranged from 1 to 2. Most of them occurred during dry season when phytomass are abundant and dry.

The intensity of fire was assessed in CNP and five «high fires date» were identified. These are originated from 1980, 1992, 1993, 1996 and 1997. The fire occurred during dry season when the combustible were available.

4.3.4. Conclusion and discussion on events of fire in savanna areas.

This study showed that the CNP needed more attention to become a real biodiversity conservation area. The frequently use of fire caused a lost of biodiversity.

Figure 17: Illustration of compartmentalisation concept on D. microcarpum stem disc

Concerning trees, when the stem was injured, the tree develops a dynamic defence process which forms structural and chemical boundaries in order to resist the spread of pathogens. In short time, the tree forms compartment and this positive reaction of the tree is named compartmentalisation. The wood formed just after a fire injury showed an anatomical wood which is different from the normal. Most of time, no vessel was observed in this part of wood but a callous margin (edge of callous tissue overgrowing a stem wound) was formed.

5. DISCUSSIONS

Based on the findings of this study and following the wood anatom ical classification of Worbes (1989), Fichtler (2003) and Schöngart (2006) , the presence of annual tree-ring in tropical African species such as A. leiocarpa, D. microcarpum, D. abyssinica, I. doka and P. erinaceus is undoubtable. A dry period of at least two months with less than 50 mm of rain would be required to expect annual rings in tropical tree-species (Worbes, 1999). However, Fichtler et al. (2003) reported that even small annual variation in rainfall occurring under everwet conditions may trigger ring formation. This confirms the role of abiotic factor triggering on tree growth such as the precipitation (Worbes, 1995; Fichtler et al., 2004; Shöngart, 2006). Furthermore early studies and especially those of Coster ( 1927) well illustrated the connection between the formation of annual tree rings and seasonal precipitation. He found that trees of the same species might form clear and annual rings under monsoon climates, while the same species form distinct and irregular rings under almost everwet conditions.

Concerning the growth performance, the results of Worbes (1989) proved that whenever the pioneer species like A. leiocarpa grow faster, those of the understory like D. abyssinica grow slowly and constantly during all life time. He also demonstrated t hat within one species, the mean annual growth can show a variation according to tree age. His findings in 2003 proved also that the lowest values of diameter growth rates are observed for understorey species whereas the highest values were noticed in the main canopy and in emergent species. The young trees of A. leiocarpa grew under good light conditions and showed therefore a good ring increment. This high difference of growth within the same species could be explained by the mean annual rainfall, index of the soil fertility and the variation in disturbance following areas. However, the analysis of these ecologic factors combined with the results about the ring increment of A. leiocarpa from the two studied sites (Tables 2, 3 & 4) proved that in this case, age of trees would be the main variable that caused the variation observed. About D. microcarpum species, the trees grow fast the first year with an annual ring increment of about 8 mm. The rate of growth decreases the next years. In fact, the light is on e of most important limiting factor on tree growth. According to Shöngart (2006), pioneer species are highly light demanding for germination and growth, and are generally short lived. The non -pioneer species have low light requirements and are able to surv ive in the dark forest understorey, growing at low rates. The juvenile trees of A. leiocarpa were found in gap dynamics (canopy

opening) and D. abyssinica trees were observed under closed forest where less than 2 % of light penetrates.

According to Hennenberg et al. (2006), the use of fires in Comoé National Park justifies a dramatic decrease of wild animals. The results of his study proved that fire is more used in savanna area than in closed forest. He demonstrated that the forest boundary was sometimes treated by fire to about 30 m which was the consequence of using of wide forest boundary of 60 m. This helps us to accept our hypothesis about closed forest protecting against fire.

6. CONCLUSION

Dendrochronology was demonstrated in the present study like a vital method of juvenile tree dating. All investigated species are sensible to abiotic factors and especially rainfall that determine the distinctiveness of ring boundary. The oldest sample tree was 33 years old whereas the youngest was 5 years old. The variability in wood anatomy of the studied species could be used to understand the diversity in tropical forest. The events of fire on tropical ecosystems and especially its impact on tree growth w as studied and revealed that West African ecosystems are not always controlled by human pressure. Further research works on these ecosystems with emphasis on endangered species by using tree -ring analysis approach for developing news forest management actions are need.

7. LITERATURE CITED

ADU-BREDU S., FOUA T. B. A., BOUILLET J. P., KOUAME M. M., YAMOAH K. S., & SAINT-ANDRE L., 2008. An explicit stem profile model for forked and un -forked teak (Tectona grandis) trees in West Africa. Forest ecology and management, 255(7): 2189 -2203.

BAKER P. J., 2003. Tree age estimation for the tropics: A test from the southern Appalachians. Ecological Applications ISSN 1051-0761 CODEN ECAPE7; 13(6):1718- 1732.

BAKER P. J., PALMER G. J. & D'ARRIGO R., 2008. The dendrochronology of Callitris intratropica in northern Australia: annual ring structure, chronology development and climate correlations. Australian Journal of Botany; 56(4):311-320.

BOURIAUD O., LEBAN J.-M., BERT D. & DELEUZE C., 2005. Intra-annual variations in climate influence growth and wood density of Norway spruce. Tree Physiology 25: 651- 660.

BRIENEN R. J.W., 2005. Tree rings in the tropics: a study on growth and ages of Bolivian rain forest trees. PhD Thesis, Universiteit Utrecht, 139p. PROMAB Scientific Series 10.

BRIENEN R. J.W., LEBRILA E., BREUGUEL V. M. & MARTINEZ-RAMOS M., 2007. Dating tropical trees in secondary forests along chronsequences. ESA/SER Joint Meeting. (10): 72-102.

BRUNKEN U., SCHMIDT M., DRESSLER S., JANSSEN T., THIOMBIANO A., & ZIZKA G., 2008. West African plants. A photo Guide. w ww. Westafricanplants.senchenberg.de. -forschungsinstitut senchenberg, Frankfurt / Main, Germany.

COPENHEAVER A. C., POKORSKI A. E., C URRIE J. E. & ABRAMS D. M., 2006. Causation of false ring formation in Pinus banksiana: A comparison of age, canopy class, climate and growth rate. Forest Ecology and Management, 236 (2-3):348-355.

COSTER C, 1927. Zur Anatomie und Physiologie der Zuwachszonen und Jahresringbildung in den Tropen I. Ann. Jard. Bot. Buitenzorg, 37: 105-119.

CULLEN E. L., ADAMS A. M., ANDERSON J. M. & GRIERSON F. P., 2008. Analyses of 13C and 18O in tree rings of Callitris columellaris provide evidence of a change in stomatal control of photosynthesis inresponse to regional changes in climate. Tree Physiology 28: 1525-1533.

DESTA F. F., RAY R. H. J. & COLBERT J., 2003. Influence of topographic aspect, precipitation and drought on radial growth of four major tree species in an Appalachian watershed. Forest Ecology and Management, 177(1-3):409-425.

DEZZEO N., WORBES M., ISHII I., & HERRERA R., 2003. Annual tree rings revealed by radiocarbon dating in seasonally flooded forest of the Mapire River, a tributary of the lower Orinoco River, Venezuela. Vegetatio 168(1):165-175.

ENGLAND R. J. & ATTIWILL M. P., 2007. Changes in sapwood permeability and anatomy with tree age and height in the broad -leaved evergreen species Eucalyptus regnans. Tree Physiology 27: 1113-1124.

FICHTLER E., CLARK A. D. & WORBES M., 2006. Age and Long-term Growth of Trees in an Old-growth Tropical Rain Forest, Based on Analyses of Tree Rings and ¹⁴C . Biotropica, 35(3):306 - 317.

GUYETTE P. R. & STAMBAUGH C. M., 2004. Post-oak fire scars as a function of diameter, growth and tree age. Forest Ecology and Management, 198(1-3):183-192.

HALL L. R. M., 2008. The fire history of a 416-years-old western larch tree in southeastern Britisch Columbia. BC Journal of Ecosystems and Management 9(2): 5 -10.

HENNENBERG K. J., 2005. Vegetation ecology of forest -savanna ecotones in the comoe National Park (Ivory Coast): Border and ecotone detection, core area analysis and ecotone dynamics. Ph-D dissertation, University of Rostock (Germany). 108 pages.

HENNENBERG K. F., FISHER F., KOUADIO K., GOETZE D., ORTHMANN B., LINSENMAIR K. E., JELTSCH F. & POREMBSKI S., 2006. Phytomass and fire occurrence along forest-savanna transects in the Comoé National Park, Ivory Coast. Journal of Tropical Ecology 22: 303-311.

KENICHI Y., SACHIE I., TAKAYUKI K., EITARO F., YUTAKA M., MITSUTOSHI K., HIROYUKI T., TAKAYOSHI K. & RYO F., 2004. Effects of elevated CO2 concentration on growth, annual ring structure and photosynthesis in Larix kaempferi seedlings. Tree Physiology 24: 941-949.

Martínez-Ramos M. & Alvarez-Buylla E. R., 1998. How old are tropical rain forest trees? Trends in plant science 3: 400-405.

MUNDO A. I, BARRERA D. M. & ROIG A. F., 2007. Testing the utility of Nothofagus pumilio for dating a snow avalanche in Tierra del Fuego, Argentina. Dendrochronologia, 25(1):19-28.

NYGÄRD R., SAWADOGO L. & ELFVING B., 2004. Wood-fuel yields in short-rotation coppice growth in the north Sudan savanna in Burkina Faso. Forest Ecology and Management, 18 (1-3):77-85.

NOVAK K., CHERUBINI P., SAURER M., FUHRER J., SKELLY M. J., KRÄUCHI N. & SCHAUB M., 2007. Ozone air pollution effects on tree -ring growth, 13C, visible foliar injury and leaf gas exchange in three ozone -sensitive woody plant species. Tree Physiology 27: 941-949.

ORTHMANN B., 2005. Vegetation ecology of woodland-savanna mosaic in central Benin (west Africa): ecosystem analysis with a focus on the impact of selective logging. Ph-D dissertation, University of Rostock (Germany). 137 pages.

PATRUT A., VON REDEN K. F., LOWY D. A., ALBERTS A. H., POHLMAN J. W., WITTMANN R., GERLACH D., XU L. & MITCHELL C. S., 2007. Radiocarbon dating of a very large African baobab. Tree Physiology 27:1569-1574.

POUSSART P. F., EVANS N. M. & SCHRAG P. D., 2004. Resolving seasonality in tropical trees: multi-decade, high-resolution oxygen and carbon isotope records from Indonesia and Thailand. Earth and Planetary Science Letters, 218(3-4):301-316

POUSSAART P.M., MYNENI S.C.B. & LANZIROTTI A. 2006. Tree-ring dating, age-dating trees by counting annual rings. Geophysical Research Letters 3:L17711.

PRIYA P. B. & BHAT K. M., 1997. False ring formation in teak (Tectona grandis L.f.) and the influence of environmental factors. Wood Science Division, Kerala Forest Research Institute Peechi, 680 653 India .

RAMSEY C. B, 2007. Radiocarbon dating: Revolution in understanding. Archaeometry, 50(2):249 - 275.

SANTOSH K. S., AMALAVA B. & VANDANA C., 2007. Reconstruction of June- September precipitation based on tree -ring data of teak (Tectona grandis L.) from Hoshangabad, Madhya Pradesh, India. Dendrochronologia, 25(1):57-64.

SAURER M., CHERUBINI P., BONANI G. & SIEGWOLF R., 2003. Tracing carbon uptake from a natural CO2 spring into tree rings: an isotope approach. Tree Physiology 23: 997-1004.

SHÖNGART J., ORTHMANN B., HENNENBERG K. J., POREMBSKI S. & WORBES M., 2006. Climate-growth relationships of tropical tree species in west Africa and their potential for climate reconstruction. Global Change Biology, 12:1139-1150.

TARHULE A. & LEAVITT W. S., 2004. Comparison of stable-carbon isotope composition in the growth rings of Isorberlinia doka, Daniella oliveri, and Tamarindus indica and West African climate. Dendrochronologia, 22(1):61-70.

THOMAS D. S., MONTAGU K. D. & CONROY J. P., 2007. Temperature effects on wood anatomy, wood density, photosynthesis and biomass partitioning of Eucalyptus grandis seedlings. Tree Physiology 27:251-260.

BÜNTGEN U., FRANK D. C., KACZKA R. J., VERSTEGE A., ZWIJACZ-KOZICA T. & ESPER J., 2007. Growth responses to climate in a multi -species tree-ring network in the Western Carpathian Tatra Mountains, Poland and Slo vakia. Tree Physiology 27:689-702.

VAN HORNE L. M. & FULE Z. P., 2006. Comparing methods of reconstructing fire history using fire scars in a southwestern United States Ponderosa pine forest. Can J. For. Res.36: 855-867.

VERHEYDEN A., KAIRO J. G., BEECKMAN H. & KOEDAM N., 2004. Growth Rings, Growth Ring Formation and Age Determination in the Mangrove Rhizophora mucronata. Annals of Botany, 94: 59-66. www.aob.oupjournals.org.

VERHEYDEN A., ROGGEMAN M., BOUILLON S., ELSKENS M., BEECKMAN H. & KOEDAM N., 2005. Comparison between ä¹³C of á-cellulose and bulk wood in the mangrove tree Rhizophora mucronata: Implications for dendrochemistry. ELSEVIER, Chemical Geology 219:275- 282.

WALTER O., 2004. Influence of climate on radial growth of Pinus cembra within the alpine timberline ecotone. Tree Physiology 24: 291-301.

WELSBERG J. P. & SWANSON J. F., 2001. Fire dating from tree-rings in western cascades Douglas fire Forests: An Error Analysis. Northwest Science, 75( 2).

WORBES M., 2002. One hundred years of tree-ring research in the tropics - a brief history and an outlook to future challenges. Dendrochronologia, 20 (1-2): 217-231.

WORBES M., STASCHEL R., ROLOFF A., & JUNK W. J., 2003. Tree ring analysis reveals age structure, dynamics and wood production of a natural forest stand in Cameroon. Forest ecology and management, 173(1-3):105-123.

ZHAO X. Y., QIAN J. L., WANG J., HE Q. Y., WANG Z. L. & CHEN C. Z., 2006. Using a Tree Ring ä¹³C Annual Series to Reconstruct Atmospheric CO2 Concentration over the Past 300 Years. Pedosphere, 16(3):371-379.

APPENDICES

Appendix 2: The output of ANOVA test about the chapter of growth performance

	Trees		Level		76		79		80		81		83		85		87		88		89		90		91		92		93		95		96		97		98		99		2001		Tot
CNP	DE-101		0,1m																																*		*						2
CNP			1,3																																								0
CNP	DE-102		0,1m																														*		*								2
CNP			1,3																																								0
CNP	DE-103		0,1m																														*								*		2
CNP			1,3																										*														1
CNP	DE-104		0,1m																																								0
CNP			1,3																																								0
CNP	DE-105		0,1m																								*										*				*		3
CNP			1,3																																								0
CNP	DE-106		0,1m																														*										1
CNP			1,3																														*										1
CNP	DE-107		0,1m																								*																1
CNP			1,3																														*										1
CNP	DE-108		0,1m																														*										1
CNP			1,3																														*										1
CNP	DE-109		0,1m																																								0
CNP			1,3																																								0
CNP	DE-110		0,1m																																								0
CNP			1,3																																								0
CNP	D1-101		0,1m						*		*																																2
CNP			1,3						*																																		1
CNP	D1-102		0,1m										*																														1
CNP			1,3																																*								1
CNP	D1-103		0,1m																		*																						1
CNP			1,3																										*						*								2
CNP	D1-104		0,1m								*				*		*						*				*																5
CNP			1,3																																								0
CNP	D1-105		0,1m		*																																						1
CNP			1,3																																								0
CNP	D1-106		0,1m										*																														1
CNP				1,3																																								0
CNP		D1-107		0,1m																		*																						1
CNP				1,3																																								0
CNP		D1-108		0,1m		*		*																																				2
CNP				1,3																																								0
CNP		AN-101		0,1m																																*								1
CNP				1,3																																								0
CNP		AN-102		0,1m																																				*				1
CNP				1,3																																								0
CNP		AN-103		0,1m																		*												*								*		3
CNP				1,3																																								0
CNP		AN-104		0,1m																*						*						*								*				4
CNP				1,3																																								0
CNP		AN-105		0,1m																												*										*		2
CNP				1,3																								*																1
CNP		AN-106		0,1m																												*												1
CNP				1,3																																								0
CNP		AN-107		0,1m												*																												1
CNP				1,3																																								0
CNP		AN-108		0,1m																																								0
CNP				1,3																																								0
Total of scarred trees						2		1		1		2		2		2		1		1		3		1		1		4		2		3		6		5		2		2		4
						+		+		++				+		+		+		+		+		+		+		++		++		+		++		++		+		+		+

Postgraduate Student in Lab of Applied Ecology, FSA-UAC, Benin Researcher in BIOTA-West Africa Programme

Wood anatomy, Wood physiology, Tree dating methods with special emphasis on dendrochronology approach, Tropical Ecosystems (Forest and Savanna) analyses, Tropical ecology, Biodiversity conservation, Tropical silviculture, Dynamic of Forest, Forest Management, Monitoring approach, Cartography and SIG

2008 : Postgraduate student in Natural Resource management (FSA-UAC, Benin)

2006-2007 : Graduate student in Natural Resource management (FSA-UAC, Benin) 2002-2006 : Graduate student in Agronomy (FSA-UAC)

+ 2008 : Certificate of Deutsch learning with very good distinction (Göttingen, Germany) + 2007 : Engineer Agronome Degree in Natural Resource Management (FSA / UAC)

+ 2006 : Degree of General Agronomy (Faculty of Agricultural Sciences / University of Abomey - Calavi, Benin)

+ 2002 : A-Levels, option biology with good distinction (Benin) + 1999 : Brevet of studies of first college round (Benin)

+ 1st september to 31 October : Intern student into Dendrochronology Laboratory of
Institute of Agronomy in the Tropics (Göttingen, Germany). Application on Using Tree rings
analysis to study the growth performance from sapling to wood of five savanna species (Topic of M -Sc
Thesis)

+ 3rd to 31 August 2008 : Initiation to Deutsch speaking in International Summer Course (University of Göttingen, Germany)

+ March 2004 : Methods of Animation and Intervention in Rural Area (Village of Zoungbomey, Department of Ouémé): Study the structure and functioning of local organizations

+ March 2003 : Discovery of Rural Area (Domanouhoué Village, Town of Aplahoué,

Department of Couffo, Benin): Identification of rural activities and logic analysis peasant

+ November 2002 : Discovery of Rural Area. Immersion phase (Village of Dossou somè, Department of Atlantic, Benin): Making contact with the rural environment

+ English: Scientific tongue, Good master in speaking, in writing and in reading

+ Excel, X-act, SAS, Minitab for Data treatment and Statistical Analysis
·
· ArcView for Cartography and SIG

+ Prof. Dr. Ir Brice A. SINSIN : Head of Laboratory of Applied Ecology and Vice deputy of research in UAC

+ Dr Bettina ORTHMANN : Specialist in wood Biology, University of Rostock, Germany

Using tree- ring analysis to study the growth performance from saplings to trees for five savanna species in West Africa

TABLE OF CONTENTS