Trees

DOI 10.1007/s00468-013-0871-3

ORIGINAL PAPER

A tree-ring based comparison of Terminalia superba climate–

growth relationships in West and Central Africa
Maaike De Ridder • Valerie Trouet •
Jan Van den Bulcke • Wannes Hubau •

Joris Van Acker • Hans Beeckman

Received: 27 September 2012 / Revised: 5 February 2013 / Accepted: 22 February 2013

Ó Springer-Verlag Berlin Heidelberg 2013

Abstract Tropical lowland forests are characterized by the Gulf of Guinea and the South Atlantic Ocean during the
humid climate conditions with interannual variations in early rainy season. However, tree growth was influenced by
amount of precipitation, length of dry season, and relative ENSO in both regions. In the Mayombe, La Niña years were
humidity. The African tree species, Terminalia superba associated with stronger tree growth whereas in Ivory Coast,
Engl. & Diels has a large distribution area and potentially El Niño years corresponded with stronger tree growth. The
incorporates these variations in its tree rings. Tree ring presented relation between ENSO, precipitation and tree
analysis was performed on 60 plantation trees (increment growth is original for equatorial African forests, suggesting
cores) and 41 natural trees (stem disks) from Ivory Coast and an influence of global climate variability on tree growth.
the Congolese Mayombe Forest. Natural forests and old
plantations (50–55 years) showed similar growth patterns. Keywords Africa  Dendroclimatology  ENSO  Sea
Regional chronologies were developed for the two sample surface temperature  Tree rings  Tropical forest
regions and showed a long-distance relationship for the
period 1959–2008. Growth in the Mayombe was associated
with early rainy season precipitation, but no relation was Introduction
found between tree growth and precipitation in Ivory Coast.
Congolese trees possibly show a higher climate-sensitivity Tropical climates are characterized by high temperatures,
than Ivorian trees, because precipitation in the Mayombe is humidity and precipitation, light frosts and an average
more limiting, and Congolese T. superba trees are found temperature of at least 18 °C during the coolest month
closer to the margins of their distribution. Likewise, tree (McKnight and Hess 2000). However, tropical climates are
growth in the Mayombe was also influenced by the SSTs of far from homogeneous and large differences in precipita-
tion periodicity, amount of precipitation, air humidity and
annual temperature occur, related to the distance from the
Communicated by M. Buckeridge. equator, ocean and altitude (McKnight and Hess 2000; van
Oldenborgh and Burgers 2005). Within the Köppen–Geiger
M. De Ridder  J. Van den Bulcke  W. Hubau  J. Van Acker
climate classification, these differences result in three
Laboratory of Wood Technology, Faculty of Bioscience
Engineering, Ghent University, Coupure Links 653, tropical climate subtypes: equatorial, monsoon and savan-
9000 Ghent, Belgium nah climates (Köppen and Geiger 1930). Notwithstanding
these climatic differences, several tree species have a broad
M. De Ridder (&)  W. Hubau  H. Beeckman
distribution area within the tropics and are able to grow in
Laboratory for Wood Biology and Xylarium, Royal Museum
for Central Africa, Leuvensesteenweg 13, diverse climate types. Furthermore, they can grow in a
3080 Tervuren, Belgium large range of local site conditions, including dense mixed
e-mail: [email protected] forest stands and widely spaced man-made plantations
(CTFT 1983). Light-demanding species such as Terminalia
V. Trouet
Laboratory of Tree Ring Research, University of Arizona, superba Engl. & Diels (Groulez and Wood 1985) are
105 W Stadium, Tucson, AZ 85721, USA expected to be sensitive to such variability in climate and

123
 Trees

site conditions. The link between light-demanding trees for human welfare and socio-economic development
and significant climate correlations was already presumed (Boninsegna et al. 2009).
in the 1920s (Douglass 1920). The interannual precipitation variability in West Central
Dendrochronology in tropical regions has a history of Africa is extremely complex, contrary to the rest of tropical
more than 100 years (Worbes 2002), but a lot more tree- Africa, where the variability is coherent over very large
ring research was conducted in tropical forests of Latin sectors (Balas et al. 2007). Research on the causes of this
America and Asia than in Africa. Several studies on dif- variability has been focused on teleconnections with trop-
ferent species resulted in tree-ring chronologies that were ical SSTs rather than on land–atmosphere interactions (Joly
verified with radiocarbon dating (Worbes et al. 2003) or et al. 2007). The SSTs of the equatorial Pacific (Balas et al.
with relations to climate (Bräuning et al. 2009; Brienen 2007; Camberlin et al. 2001; Joly et al. 2007; Paeth and
and Zuidema 2005; Dünisch et al. 2003; Pumijumnong Friederichs 2004), the tropical Atlantic in general (Cam-
et al. 1995; Worbes 1999). For the African continent, tree- berlin et al. 2001; Paeth and Friederichs 2004), and the
ring chronologies were mostly developed for semi-arid Benguela Coast including the Gulf of Guinea in particular
savannas (Fichtler et al. 2004; Gebrekistos et al. 2008; (Balas et al. 2007; Paeth and Friederichs 2004; Joly et al.
Gourlay 1995), the miombo woodlands in southern Africa 2007) are among the most important factors governing the
(Stahle et al. 1999; Trouet et al. 2001, 2006, 2010) and interannual variability in western and equatorial Africa.
Ethiopia (Couralet et al. 2005; Sass-Klaassen et al. 2008; The influence of these oceans is seasonally dependent
Wils et al. 2011). However, the number of exactly dated (Balas et al. 2007) and resulting relations also vary
tree-ring chronologies from West and Central African between studies based on the season, the time span, the
species is limited (Couralet et al. 2010; Schöngart et al. study region, and the methods selected for analysis.
2006; Worbes et al. 2003), despite clear potential as Moreover, the abovementioned anomalies of SSTs of
pointed out by the exploratory research by Hummel the equatorial Pacific Ocean are generally used to calculate
(1946) and Détienne and Mariaux (1970, 1975, 1976, ENSO indices (Trenberth 1997). ENSO is the result of the
1977). For the tropics, especially the work of Détienne coupling between oceans and the atmosphere (Nicholson
and Mariaux was important because they showed the and Entekhabi 1987), but most ENSO indices, including
existence of annual tree rings in different climates and the Niño3.4 index, are based on anomalies of SSTs in the
countries by applying cambial wounding, phenological region between 5°N–5°S and 120°–170°W (Trenberth
observations, and monthly diameter measurements. More 1997). Camberlin et al. (2001) studied the gridded response
recently, Tarhule and Hughes (2002) published a list of of African precipitation on Niño3 indices and found no
West African species with dendrochronological potential. consistent large-scale relation in West Africa. Surprisingly,
Only 7 out of more than 70 species were labelled they observed a clear negative relation in March–June and
‘potentially useful’, indicating that potential exists but the August–November in western equatorial Africa [Camer-
choice of study species is limited. oon, Gabon, Republic of Congo and the coastal region of
The majority of dendroclimatological studies in tropical the Democratic Republic of Congo (DRC)], a region which
regions focus on one climate type and compare tree growth had never before been shown to exhibit significant corre-
of different species within this climate, e.g. Devall et al. lations with ENSO.
(1995) and Brienen and Zuidema (2005). Still, when tree The studied species is T. superba, a long-living pioneer
ring analysis is applied and results in long time-series, it tree species, characterized by large buttresses and typically
offers a powerful tool to analyze the responses of trees, found in secondary forests and fallows (Groulez and Wood
species and stands to different and changing climatic 1985; Hawthorne 1995; Humblet 1946; Swaine and
conditions. Detailed knowledge on the growth patterns of Whitmore 1988). This tree species has a large range of
tropical tree species can thus be obtained, that is necessary distribution on both hemispheres (from Sierra Leone to
for reforestation programs and REDD ? files (reduced Angola) (Groulez and Wood 1985). The climate within this
emissions from deforestation and degradation). In addition, region [dry season with \60 mm monthly precipitation in
tree ring time-series enables the development of dated at least 3 months (Worbes 1995)] is believed to provoke
proxies for climate reconstruction (Schöngart et al. 2006; the formation of yearly growth rings. Moreover, T. superba
Therrell et al. 2006) and can extend instrumental climate was also planted over vast areas, including DRC, and the
time-series that are often sparse in tropical regions. The known planting dates allowed for exact dating and confir-
comparison of those climatic reconstructions based on tree mation of the annual character of tree rings in this species
rings with atmospheric circulation patterns, sea surface (De Ridder et al. 2010). The formation of tree rings was
temperatures (SSTs) and the El Niño—Southern Oscilla- already studied by Mariaux (1969), who used periodic
tion (ENSO) provides a useful bridge between past and woundings and dendrometers to monitor tree growth, but
future trends in global climate change and its implications his study was limited to \20 trees younger than 20 years.

123
Trees

Therefore, proper cross-dating and comparison with cli- a short period with less precipitation during the rainy
mate was not possible. Worbes et al. (2003) also used season (January–February). The proximity of the ocean,
T. superba discs and wood cores from Cameroon yet manifested by a strong nebulosity, buffers the intensity of
without climate matching. In our study, tree rings of both dry periods. Relative humidity remains more or less
T. superba from old plantations (ca. 50- to 55-year old) and constant throughout the year (84 % in the rainy season,
natural forests in the Mayombe Forest (southern hemi- 82 % in the dry season). The average annual precipitation
sphere) and western Ivory Coast (northern hemisphere) based on precipitation data from 1959 to 1996 is
were analyzed in combination with climate data to analyze 1,168 mm, but some years are particularly dry (Fig. 1). For
the following hypotheses: instance, the annual precipitation in 1972 and 2002 was
\800 mm and the lowest values were registered in 1954,
1. Planted and natural T. superba trees reveal different
1978 and 1997 with annual precipitation values of
growth patterns due to different site conditions and/or
respectively 657, 692 and 626 mm. Temperatures oscillate
different responses to climate.
around 26 °C in the rainy season and drop to a minimum of
2. Trees of T. superba show cambial growth in one
20 °C in the dry season. The soils of the Luki Reserve are
hemisphere, but not in the other hemisphere and vice
classified as orthic Ferralsols, but Tshela is characterized
versa due to the shift of 6 months in the precipitation
by ferric Acrisols (FAO 2008). Most soils are argillaceous
seasonality.
with a pH between 4 and 6 and a C/N between 4 and 9.
3. This means that trees react sensitively to climate at one
In western Ivory Coast (06°070 to 07°150 N, 07°300 to
site, but do not incorporate these climate conditions at
08°150 W), the four study sites (Fig. 1) are located in nat-
the other site and vice versa. Climate conditions can
ural forests. The forest of Scio can be considered as pri-
include regional climate variability, and can be
mary forest whereas the forests of Goya, Bin Houye and
influenced by SST anomalies in the surrounding
Danane are secondary forests, often with cacao plantations
tropical ocean sectors as well as by ENSO events.
in the understory. Unlike in DRC, study sites were wide-
spread and sometimes more than 100 km apart. Moreover,
no central climate station was available and data (including
Materials and methods monthly precipitation and temperature) from six sur-
rounding climate stations (Table 1; Fig. 1) were therefore
Study sites averaged. Ivorian climate data were extracted from the
KNMI explorer (van Oldenborgh and Burgers 2005). In
All study sites belong to the Guineo-Congolian regional this region, the dry season generally lasts for 3 months
centre of endemism (White 1983). The Ivorian study sites (December–February). In July and August, a period of less
are classified as evergreen moist rainforest, but the Con- precipitation is observed. Relative humidity drops from
golese study sites are situated within a drier semi-evergreen approximately 88 % during the rainy season to approxi-
rainforest. mately 70 % during the dry season. The average annual
The three Congolese study sites (Fig. 1) are located at precipitation, based on precipitation data from 1959 to
the southern border of the Mayombe Forest that covers the 1996, is 1,650 mm. Annual mean temperature is 25 °C,
western parts of Gabon, the Republic of Congo, DRC and with a minimum of 18 °C in January and a maximum of
Cabinda (Angola). Two study sites were chosen within the 33 °C in February/March (van Oldenborgh and Burgers
UNESCO Man and Biosphere Reserve of Luki (05°300 to 2005). Study sites were situated between 200 and 370 m
05°450 S and 13°070 to 13°150 E). The first study site was a above sea level. Soils have a pH between 4 and 7, a C/N of
natural forest stand close to the climate station of the 8–12 and are classified as Ferralsols and Acrisols, i.e.
reserve, and the second study site was located at a distance typical acid soils of tropical lowlands (FAO 1986).
of 15 km in a plantation of T. superba, established between
1955 and 1957 in the village of Monzi. Plantations were Sampling strategy
installed on clear-felled areas, at planting distances of
8 9 12 m. After the installation, no structured manage- A total of 41 stem discs were collected from natural forests
ment was carried out. Natural regeneration of T. superba in the Mayombe and western Ivory Coast (one stem disc
and other species took place after a few years (De Ridder per tree). In Monzi, two perpendicular wood cores were
et al. 2010). A third study site was selected in a natural taken per tree, resulting in a total of 120 cores from 60
forest stand near Tshela, about 70 km to the north. All planted trees (Table 2). Cores were taken above the but-
study sites were situated at altitudes \300 m above sea tresses where possible. All samples were air-dried to pre-
level. The region is characterized by a dry season of vent fungal infestation. Wood cores were frozen for
approximately 5 months (May to September/October) and 2 weeks to prevent insect infestation. Stem discs were too

123
 Trees

Fig. 1 Geographical location of all study sites in Ivory Coast and the stations, 1959–1996, extracted from the KNMI explorer (van
Mayombe Forest in the Democratic Republic of Congo (DRC) with a Oldenborgh and Burgers 2005); for temperature data only two
climate diagram for the Mayombe Forest (climate station of Luki, stations, Man and Gagnoa, provided data from 1959 to 1984]. The
1959–1996) and Ivory Coast [regional climate average from 6 climate dashed line represents the equator

large for freezing and therefore only superficially disin- Tree ring analysis
fected before storage in the Tervuren Xylarium. All discs
and cores were sanded with gradually increasing grid from Before measurement, all tree ring boundaries were marked
50 to 600 or 1,200. with pencil under a stereomicroscope or magnifying glass.
On stem discs, three radii were chosen and every tenth ring
was followed along the circumference to check for ring
Table 1 Characteristics of climate stations around Ivorian study sites anomalies. In case of anomalies (false rings and particularly
and the Mayombe (station, precipitation, dry season (\60 mm), dis-
tance to the closest study site) wedging rings), every tree ring in this zone was followed
separately. When unclear, the particular ring num-
Station Region Annual Dry Distance to
ber(s) were archived to facilitate corrections during sub-
precipitation season closest study site
(mm) (km) sequent tree-ring analysis. Ring widths were measured to
the nearest 0.01 mm using a stereo-microscope and a Lintab
Daloa Ivory 1,256 Nov–Feb 140 (Goya) measuring device with TSAP-Win software (Rinn 2003).
Guiglo Coast 1,644 Nov–Feb 52 (Scio) The age of plantation trees was compared with the
Man 1,607 Nov–Feb 58 (Danane) planting period if two conditions were fulfilled: at least one
Soubre 1,401 Dec–Jan 110 (Goya) of the two wood cores included the pith and the trees were
Toulepleu 1,668 Dec–Feb 25 (Bin Houye) sampled within the fixed planting scheme of 8 9 12 m.
Nzerekore Guinea 1,838 Dec–Feb 100 (Danane) Outside of this scheme, T. superba trees could be regen-
Luki Mayombe 1,168 May–Sep 5 (natural forest erated naturally or planted later. If the age corresponded
of Luki)
with the planting period, tree rings were considered annual
The common time span in all stations is from 1959 to 1996 (Stahle 1999).

123
Trees

Table 2 Summary of the details of the sampling campaigns in the how well the finite-sample chronology matches the theo-
Mayombe and Ivory Coast retical population chronology. A value of 0.85 was con-
Study site Region Number of Date sampling sidered as a reasonable limit for a reliable chronology
sampled trees (Wigley et al. 1984).
Luki Mayombe 5 Oct–Nov 2005
Analysis of climate–growth relationships
Monzi 60 Aug 2007
Tshela 7 Jun 2008
A climatological year was considered as a sequence of a
Scio Ivory Coast 8 Jan 2009
dry and a rainy season. As such, climatological years for
Goya 8
the Mayombe Forest were delimited from Mayn-1 to Mayn
Danane 6
where n is the calendar year to which the tree ring is
Bin Houye 7
assigned. For western Ivory Coast, years were delimited
from Decembern-1 to Decembern. First, annual, seasonal
(dry/rainy season) and monthly precipitation data were
correlated with regional tree-ring chronologies. Depending
Temporal correspondence between the various tree ring on the resulting Pearson’s correlations, months were
series was checked by cross-dating (Douglass 1941). grouped to test the influence of the early rainy season,
Cross-dating is a technique that ensures that each individ- transition from dry to rainy season, etc. Precipitation data
ual tree ring is assigned to the exact year of its formation. were derived from the aforementioned climate stations
This is accomplished by matching patterns of wide and (Table 1; Fig. 1). In DRC, the regional chronology (plan-
narrow rings between cores/radii from the same tree. tation and natural forest chronologies combined) was cor-
Subsequently, the tree ring series of each tree are merged related with precipitation data of the Luki station. The
into an average tree ring series and then cross-dated with regional tree-ring chronology for western Ivory Coast was
the other trees of the same site. Successful cross-dating of correlated with the mean precipitation data of six sur-
several trees indicates that there is a common external rounding climate stations. Both growth–precipitation rela-
factor influencing the growth of these trees (Cook and tions were calculated over the common time span limited
Kairiukstis 1990; Worbes 1995). Those trees can then be by the length of the regional chronologies and precipitation
averaged to a site chronology or even to a regional tree-ring time series (1959–1996).
chronology. We conducted spatial correlation analyses to study the
Two statistical parameters are taken into account to influence of SSTs on precipitation and tree growth in the
evaluate the success of cross-dating: the Pearson’s corre- two regions. Spatial correlation maps were generated using
lation (r) and the t value of Baillie–Pilcher (Baillie and the KNMI explorer (van Oldenborgh and Burgers 2005,
Pilcher 1973). Another non-statistical parameter is Gle- Trouet and van Oldenborgh (2013) http://climexp.knmi.nl)
ichläufigkeit (Eckstein and Bauch 1969) or percentage of and were based on gridded 5 9 5 monthly and seasonal
parallel run (ppr), which reflects the percentage of oscil- SST fields (Hadley Centre HadSST3; Kennedy et al. 2011a,
lations in the same direction within the overlapping inter- b). Correlations were calculated over the period from 1959
val. Still, especially for tropical trees, visual control and onwards. These correlation maps give a clear view on
expert knowledge of Terminalia wood anatomy and tree which oceanic regions influence precipitation patterns and
rings are indispensable. The work of Trouet et al. (2010) is tree growth.
referred to for a detailed explanation on the lower thresh- Furthermore, we used the SST indices for the Niño3.4
olds for ppr (C60 %) and t values (C2) in tropical trees region (Trenberth and Stepaniak 2001) to study ENSO
compared to temperate regions. Autocorrelation (AC) was influence on regional precipitation patterns and tree growth.
checked to filter possible age trends and standardize the For this purpose, we calculated spatial correlation maps
tree ring series to enable dendroclimatological analysis between the annual (July–June) Niño 3.4 index and gridded
(Fritts 1976). For site and regional chronologies, series 1 9 1° monthly and seasonal precipitation fields (CRU
were standardized using a one-sided moving average over T.S3.0; Mitchell and Jones 2005). ENSO signals in the
5 years (Baillie and Pilcher 1973). The mean sensitivity regional tree-ring chronologies were analyzed following the
(MS) of these standardized chronologies reflects the sen- method described in Schöngart et al. (2004, 2006). Two
sitivity of tree growth to a common external factor (Fritts sample tests between tree growth in ENSO years and other
1976). For each site chronology, the expressed population years were performed (Schöngart et al. 2004, 2006). In this
signal (EPS) was also calculated (Haneca et al. 2005; analysis, ENSO years were defined by 5-month running
Wigley et al. 1984) to quantify uncertainty due to means of SST anomalies in the region of 5°N–5°S and
decreasing sample replication. As such, the EPS indicates 120°–170°W exceeding ?0.4 °C for 6 or more consecutive

123
 Trees

months (El Niño years) or -0.4 °C for 6 or more consec- Marginal parenchyma bands were generally discontinuous
utive months (La Niña years) (Trenberth 1997). and often hard to distinguish from other bands of paren-
chyma. In some tree rings, the anatomical structure was
Phenological data even inversed: flattened fibres (dark tissue) and banded
parenchyma were found in the earlywood of the tree ring.
Historical phenological data were available for the Con- In summary, tree-ring delineations were more complex on
golese Mayombe study site (Couralet 2010). From 1948 to Ivorian stem discs.
1957, phenological data on 3,750 woody plants were col- Marking tree-ring boundaries was rather straightforward
lected in Biosphere Reserve of Luki. Patterns of defolia- in young trees (\30 years), from both regions. Occasion-
tion, flowering, fruiting and dissemination of fruits were ally, the first rings showed intra-annual density variations
studied, also for T. superba (n = 184 trees). These data but they were not considered problematic on stem discs
were compared to precipitation and tree-ring chronologies because they were not present on the whole circumference.
within the overlapping time span. No phenological data After a period of fast growth with wide tree rings, ring
were collected or available for western Ivory Coast. width drops significantly in the most recent rings. This
decrease usually occurs gradually and could be linked to
the age trend. False rings and double rings were easy to
Results detect but wedging rings formed the main problem in
identifying annual tree rings in T. superba. Wood cores
Tree-ring related wood anatomy, ring anomalies from the planted trees in the Mayombe offered the
and the use of wood cores opportunity to quantify these ring anomalies. A tree-ring
chronology of stem discs from the natural Mayombe forest
Tree rings were distinct and measurements were possible in was developed (see below) and subsequently used to test
the Mayombe as well as in western Ivory Coast. Tree rings through cross-dating if tree-ring series from wood cores
in T. superba are also annual: this was confirmed by nine were accurate enough to construct a tree-ring chronology.
plantation trees that had the pith included in one of the More than 28 % of the measured wood cores showed no
wood cores and were located within the planting schemes. false or wedging tree rings upon comparison with the
Those nine trees had ages comparable to the planting natural forest chronology. Those wood cores formed the
period, with differences between tree-ring dated and known foundation of a plantation chronology that spans from 1959
planting age of up to 3 years. This discrepancy can be to 2007. On average, per tree, 1 false ring (3 %) occurred
attributed to the sampling height and a more or less con- and 0.5 ring (1 %) was at least partially missing. In the
tinuously active cambium during the first years, resulting in
ill-defined or non-detectable tree-rings (Mariaux 1969).
The structure of juvenile rings is somewhat different from
other tree rings in Ivory Coast as well as the Mayombe:
smaller vessels, less parenchyma and lighter fibre tissues.
However, the anatomical structures of tree rings differed
between the two regions. In general, with exclusion of the
first juvenile rings, tree rings in the Mayombe (Fig. 2a)
have mostly (sparse) aliform parenchyma. Banded paren-
chyma is sparse and when it is present, it is located in the
middle and/or at the end of the tree ring. Confluent
parenchyma is regularly present, but connects only a few
vessels. Marginal parenchyma is distinct (Fig. 2a), but is
often only present in most recent rings. The sharp transition
between light and dark tissues (flattened fibres) in partic-
ular substantially simplifies tree-ring delineation. In Ivory
Coast (Fig. 2b), tree rings were more undulating than in the
Mayombe and transitions in colour were not as sharp.
Ivorian T. superba wood contained a lot of parenchyma. In Fig. 2 Transversal view on a Congolese T. superba tree rings and
addition to a small amount of aliform parenchyma, the b Ivorian tree rings, with aliform (1), confluent parenchyma (2) and
narrow bands of parenchyma (3). Tree-ring boundaries are marked.
majority of tree rings contained confluent parenchyma,
The white bar represents 5 mm. In both cases, the cambium side
extending over several vessels. In the middle and at the end corresponds with the upside and the pith side with the underside of the
of tree rings, narrow parenchyma bands can occur. pictures

123
Trees

latter case, the tree ring was visible on one wood core of a Comparing the two regional chronologies, we found a
tree but not on the other one. Analysis of the exact location weak long-distance relationship for the period 1959–2008
of tree-ring anomalies shows that 45 % of the anomalies (ppr = 55 %; r = 0.39; t = 3.7). The ppr is low, but the
were found between the age of 21 and 30 years (0.15 Pearson correlation and t value are significant (p \ 0.01).
anomalies/ring). In sections over 30 years, the number of A visual check confirms this correspondence and revealed
anomalies decreased to 21 % or 0.11 anomalies/ring a short discrepancy only during the 1990s (Fig. 4). In
(31–40 years), thus the most recent formed rings were not addition, there exists a significant positive correlation
necessarily more sensitive to anomalous rings. However, between the precipitation from November to April in both
plantation trees were not older than 55 years and wedging regions (r = 0.42; p \ 0.01).
rings also caused many problems in 9 stem discs
[100 years. Unfortunately, quantifying these wedging Response of T. superba growth to interannual climate
rings was less straightforward due to the low growth rates variability
and therefore, low visibility.
At a local scale, Mayombe tree growth was primarily
Growth patterns in natural and planted African forests positively correlated with precipitation. However, a sig-
nificantly (p \ 0.05) positive correlation was found only
Cross-dating was successful at all sites in the Mayombe for November, at the start of the rainy season and also the
and Ivory Coast and site chronologies with a minimum of month with the maximum precipitation of the year (Figs. 5,
four contributing trees were developed for every site 6). Correlation values for Ivory Coast were much lower,
(Table 4). The t value of the combination Luki–Tshela is and no significant correlations were found.
not significant, but the ppr, inter series correlation and At a regional scale, November precipitation in the
visual control confirm that a preliminary regional chro- Mayombe was used for correlation maps with SSTs and
nology covering the natural Mayombe forests can be was positively influenced by SSTs of the Gulf of Guinea
constructed (Table 3; Fig. 3). The chronology of the and mostly the coastal South Atlantic Ocean during the
plantation (Table 3; Fig. 3) was compared to this pre- months of October–December (Fig. 7a). SSTs during this
liminary regional chronology, showing co-varying growth time of year were also positively correlated with our
patterns (r = 0.89; t value = 4.0; p \ 0.001). The regio- Mayombe tree-ring chronology in a part of the Gulf of
nal chronology for the Mayombe thus includes all tree- Guinea and the South Atlantic Ocean (Fig. 7b).
ring series from the trees that were part of the site At a global scale, early rainy season precipitation in both
chronologies of Monzi, Luki and Tshela. In Ivory Coast, regions was influenced by ENSO, with negative ENSO (La
some sites show satisfactory results visually as well as for Niña) years resulting in wetter than normal early rainy
a minimum of two statistical parameters. Nevertheless, season conditions in the Mayombe region (Fig. 7c). As a
the relations are generally less strong and less significant result, tree growth in the Mayombe was significantly higher
than in the Mayombe (Table 4). Unlike in the Mayombe, in La Niña years than in ‘normal’ years (Table 5). In Ivory
not all trees from the Ivorian site chronologies were used Coast, no significant correlations were found between
to build the regional chronology (Tables 3, 4; Fig. 3). ENSO and precipitation on the sample sites. However,
Instead, all individual Ivorian tree-ring chronologies were positive ENSO years corresponded with higher precipita-
compared to each other, and 22 trees that showed the tion in northern Ivory Coast (Fig. 7c), likely reflected by
strongest common signal were extracted. Those 22 trees stronger tree growth during El Niño years (Table 5).
were used to construct the regional chronology for wes-
tern Ivory Coast.
Table 3 shows that the percentage of trees included in Discussion
the stem disc-based chronologies is higher than in the wood
core-based chronology. In the Mayombe, the mean growth Tree-ring related wood anatomy, ring anomalies
of natural T. superba trees is remarkably higher than for and the use of wood cores
plantation trees (p \ 0.01). For comparable diameters,
T. superba trees from the Mayombe grow faster Mariaux (1969) was the first to describe annual tree rings in
(p \ 0.001) than Ivorian trees. Plantation and natural forest T. superba. His description was based on young trees with
chronologies from the Mayombe region showed high a lot of juvenile wood. Mature tree rings appear to differ
autocorrelations, indicating a clear age trend. The Ivorian substantially. In trees older than 30 years, bands of
chronology has a lower AC and therefore a higher MS. parenchyma were less abundant and the marginal paren-
After standardization, AC was close to zero for all chro- chyma (if present) was not always continuous. However,
nologies. All EPS values approach the threshold of 0.85. the large variations between years that Mariaux (1969)

123
 Trees

Table 3 Site and regional chronology characteristics of natural forests in Ivory Coast (IC) and the Mayombe (n samples, diameter, number of
tree rings, mean growth, MS (mean sensitivity), AC (auto-correlation), time span, Pearson correlation, EPS (expressed population signal)
Natural forest IC Plantations Mayombe Natural forest Mayombe

Total n sampled trees 29 60 12

No. of samples in chronology 22 (81 %) 28 (47 %) 11 (92 %)
Mean diameter (cm) 56 ± 11 41 ± 12a 57 ± 15
Mean no. of tree rings 46 ± 34 40 ± 6 47 ± 38
Time span chronology 113 (1895–2008) 49 (1959–2007) 36 (1973–2008)
Mean growth of chronology (mm) 4.53 ± 1.64 5.45 ± 2.60 7.19 ± 2.67
AC 0.64 0.91 0.84
MS after standardisation 0.26 0.16 0.16
Pearson r 0.14 0.36 0.33
EPS 0.79 0.94 0.84
a
Based on the cumulative tree-ring measurements, not on standing tree measurements

Fig. 4 Standardized regional chronologies of the Mayombe (full line)

and Ivory Coast (dashed line). Comparing both chronologies results
in a t value of 3.7 and a correlation of 0.39 (both p \ 0.01)

Fig. 3 Standardized chronologies for natural forest (dotted line) and

plantations (dashed line) in the Mayombe and the natural forests of
Ivory Coast (full line). The lower part of the graph indicates the inversed, as was the case in some Ivorian tree rings.
number of trees included in the chronologies (sampling depth) Clearly, the anatomical tree-ring structure can differ in
T. superba from different regions.
reported were also found in our study. In some years, there Tree rings were easier to detect in the Mayombe com-
is a vessel-less area at the start of a growing season. Also, pared to Ivory Coast. This is potentially related to a dif-
the normal anatomical structure within a ring can be ference in leaf shedding behaviour in the two regions and

Table 4 Ppr (percentage of parallel run), t value Baillie–Pilcher and Pearson’s correlation coefficient for cross-dating of local chronologies in
the Mayombe and Ivory Coast
Sites Region Overlap (years) n ppr (%) Pearson r t value Baillie–Pilcher

Tshela vs. Luki Mayombe 1981–2006 6 vs. 5 61 0.51** 0.9

Tshela vs. Monzi 1973–2007 6 vs. 28 89 0.87*** 5.6***
Luki vs. Monzi 1981–2006 5 vs. 28 63 0.76*** 0.8
Bin Houye vs. Danane Ivory Coast 1972–2007 4 vs. 6 62 0.21 1.0
Bin Houye vs. Goya 1986–2008 4 vs. 6 71 0.13 2.6*
Bin Houye vs. Scio 1972–2008 4 vs. 7 57 0.24 1.5
Danane vs. Goya 1986–2007 6 vs. 6 55 0.61** 1.7
Danane vs. Scio 1966–2007 6 vs. 7 63 0.45** 3.1**
Scio vs. Goya 1986–2008 7 vs. 6 62 0.09 2.3*
Overlap refers to the time period the two site chronologies have in common and n refers to the number of trees included in the site chronologies
* p \ 0.05, ** p \ 0.01, *** p \ 0.001

123
Trees

Fig. 5 Correlations between a the Mayombe chronology, b the

Ivorian chronology, and the monthly, seasonally and annual precip-
itation. The dotted lines indicate the p \ 0.05 level

Fig. 6 Correlation between the chronology of the Mayombe and

November precipitation. The precipitation curve is a dashed line and
the regional chronology is a full line

the generally drier conditions of the Mayombe. In the

Mayombe, defoliation is simultaneous in the late dry sea-
son (up to 50 % of T. superba trees lose their leaves in
September (Couralet 2010)) and complete while in Guinea
(adjacent to Ivory Coast), the leafless period occurs not
simultaneously in every tree, is shorter in time and less Fig. 7 Correlation maps between a November precipitation, b tree
complete (CTFT 1959). Personal observations confirm that growth, and gridded average SSTs for the months October–Decem-
defoliation in the DRC is complete in June–July for all ber, all for the Mayombe over the period 1959–1996. The correlation
map (1959–2008) between the annual Niño 3.4 time-series and
trees of one whole plateau or valley while in Ivory Coast, it gridded October–November precipitation is shown in (c). The two
differs strongly between individual trees: trees with new study sites are indicated as black stars in (c)
leaves, leafless trees and trees that are still shedding leaves
are found at the same site at the same time. unknown. Our study confirms that wedging mostly occurs
Problems to define tree rings were caused by the first in the outer parts of the tree circumference, when the tree
unclear juvenile tree rings, false rings, double rings and starts to grow slower due to its inherent growth trend and
especially ring wedging (Mariaux 1969). Worbes (2002) forms buttresses (Worbes 2002). Buttresses become larger
also classified Terminalia as a taxon that tends to form when tree diameters increase (De Ridder et al. 2010), and
wedging rings. Ring wedging has been noted in many could also cause wedging. Worbes (2002) related ring
tropical tree species (e.g., Trouet et al. 2010; Tarhule and wedging to trees that grew under poor light conditions
Hughes 2002), but the exact cause for this phenomenon is and under competition. T. superba, abundantly present in

123
 Trees

Table 5 Visualization of two sample tests evaluating the influence of (2006), and higher than the correlations of Trouet et al.
ENSO events on tree growth (2006, 2010). Auto-correlation is high compared to other
t values El Niño Year (?1) La Niña Year (?1) tropical tree species (Bräuning et al. 2009; Couralet et al.
2010; Trouet et al. 2006, 2010) but not unusual because
Mayombe 0.18 -0.32 2.41* -0.94
T. superba has a strong and known age trend, colonizing
Ivory Coast 2.04* -1.00 0.72 -1.50 fallows with fast growth during the first years and strongly
Ring-width indices during the ENSO event and one year after the decreasing in growth after this juvenile phase.
event were compared in the Mayombe region and western Ivory Coast Fifty-year-old plantations and 40- to 50-year-old natural
(both from 1959 to 2008)
forest of the same tree species in the Mayombe enable an
* p \ 0.05
original comparison of growth patterns in natural versus
planted trees. Eshete and Stahl (1999) as well as Worbes
fallows and forest gaps, mostly receives full light during its (1999) used plantation trees but their time-series were too
juvenile phase. Afterwards, when canopy closure occurs or short for chronology development (\25 years). In our
trees of the climax forest reach their full height, competi- study, the overlap between planted and natural trees is
tion for light could cause wedging, especially in the most 36 years, and planted and natural trees showed similar
recent tree rings. To answer Worbes’ hypothesis on com- growth patterns. The studied plantation is a production
petition, more ecological data on T. superba and the forest that is not intensively managed, but nevertheless, the
accompanying tree species need to be collected within the natural forests were more diverse with a more explicit
sampled forests. competition for light and nutrients. The average growth as
Because of the frequent occurrence of wedging rings well as the initial growth (first 20 years) was lower in the
and its interference with cross-dating, Worbes (2002), plantation compared to the natural forest. Two extremely
Trouet et al. (2006), and Brienen and Zuidema (2005) dry years (1956 and 1958) and the use of inferior planting
recommend the use of stem discs rather than wood cores material have been shown to explain the loss of plantation
for tropical tree ring analysis. Our study shows that the trees immediately after the installation (De Ridder et al.
percentage of cross-dated samples in our wood-core based 2010) and have likely also influenced the growth of sur-
plantation chronology is significantly lower than in the viving trees.
stem disc-based chronology of the natural forest. However, In temperate regions, reference chronologies are avail-
there were three main reasons why plantation trees were able from a dense network of dendrochronological records
excluded from the chronology: (1) trees had an eccentric and can be used for provenancing and dating wooden
centre (the orientation of wood cores is difficult in this objects (Haneca et al. 2009). These chronologies can span
case), (2) wood cores were broken, resulting in invisible more than 10,000 years [e.g., oak (Friedrich et al. 2004)].
tree rings due to cracks, or (3) tree ring series from Such long time spans are difficult to accomplish in the
opposing cores from the same tree did not cross-date. Tree tropics, but the development of a regional network of
ring anomalies were the main cause for omitting tree ring species-specific chronologies that include older trees is
series in the latter case. However, when a first chronology feasible. T. superba appears to be a suitable tree species for
can be built from stem discs with accurate mapping and this goal, because it shows distinct tree rings, has a wide
documenting of ring anomalies (especially wedging and distribution range, and displays cross-dating potential
false rings), tree ring series of wood cores can be checked throughout a large section of this range. The longest
and implemented in or omitted from this existing chro- regional chronology here developed only spans 113 years,
nology. In the tropics, Schöngart et al. (2006), Stahle et al. but the oldest sampled tree was 165 years in the Mayombe
(1999) and Worbes (1999) also successfully analyzed tree and 182 years in Goya (Ivory Coast) and nine trees in total
growth using only wood cores and one or a few stem discs. in our dataset covered [100 years. Unfortunately, the
Moreover, sampling wood cores are less destructive and sample replication for such long tree-ring chronologies was
substantially facilitate sample transport. too low to incorporate the oldest trees in the regional
chronologies. Our study shows that wood cores give an
Growth patterns in natural and planted African forests accurate indication of the number of tree rings and thus the
age of a tree and we suggest to first core trees before
In the Luki Reserve (DRC), Couralet et al. (2010) found cutting for chronology development. With additional
between-trees correlations for understory trees that were samples, it should be possible to reconstruct a regional
similar in magnitude to our results. The correlation strength chronology for T. superba with a length of up to 200 years,
in our study was also in accordance with the findings of which would greatly increase the time span covered by
Bräuning et al. (2009) and Brienen and Zuidema (2005), instrumental climate records and thus allow for paleocli-
lower than those of Stahle et al. (1999) and Therrell et al. matic analysis.

123
Trees

Our regional chronologies for the Mayombe and Ivory Wood 1985). In addition, our climate–growth relations
Coast were located more than 2,600 km apart, overlapped were only tested over a rather short but common time span
in time for 50 years (1959–2008), and showed a weak to enable comparison between the two regions
synchronization. Fichtler et al. (2004) did not find a match (1959–1996). We correlated growth and precipitation from
between two sites in southern Africa that were 900 km 1959 to 2006 with data from the Luki climate station
apart and used spectral analysis to reveal similar long-term (Mayombe) to illustrate the influence of longer climate
oscillation patterns. Trouet et al. (2010) cross-dated chro- series. In addition to a stronger relation with November
nologies from five Miombo woodland sites in Zambia over precipitation (r = 0.42; p \ 0.01), summed October–
a distance of more than 1,000 km and found long-distance December precipitation also showed a strong positive
growth correlations between some sites. To our knowledge, relation with tree growth (r = 0.37; p \ 0.01). So, more
our study is the first to indicate long-distance growth cor- and stronger climate correlations in the Mayombe were
respondence in rainforests. One hypothesis for these syn- found if the most recent precipitation data were included.
chronized growth variations could be the presence of a This difference could not be tested in Ivory Coast (shorter
common growing season as Mariaux (1969) confirmed precipitation time series, more stations with lacking data),
with dendrometer studies. On a limited number of trees, he but we hypothesize that stronger relations with precipita-
found that tree growth in both the Mayombe and central tion in Ivory Coast could come out if longer precipitation
Ivory Coast mainly occurred during October–December. time series, including the most recent years, were available.
Tree growth and November–April precipitation are not This is indirectly confirmed by the ENSO study (see
directly related in either of the regions, but both tree below), which was run for the two regions from 1959 to
growth and November–April precipitation showed similar 2008, resulting in a significant relation between tree growth
variability. Other relations between tree growth and cli- and ENSO climate indices in both regions.
mate are discussed in detail in the next paragraph. Based on the local relation between climate and growth
in the Mayombe, the regional influence of SSTs on tree
Response of T. superba growth to interannual climate growth and precipitation was analyzed with correlation
variability maps, in order to provide a clear view on the influencing
oceanic regions. The South Atlantic Ocean, including parts
In the Mayombe, early rainy season precipitation is an of the Gulf of Guinea, appears the most important driver
important driver of tree growth. This matches with the for the Mayombe precipitation as well as tree growth.
abovementioned dendrometer measurements in the May- Balas et al. (2007) and Paeth and Friederichs (2004) also
ombe Forest that show the strongest growth during the confirmed the influence of the Gulf of Guinea and the
onset of the rainy season (Mariaux 1969). A positive tropical Atlantic Ocean on precipitation in western Central
response to the first rains was also found in one understory Africa and indicated the ENSO mode as another important
species [Aidia ochroleuca (K. Schum.)] in the Mayombe factor. In addition to precipitation, the SSTs of the Gulf of
(Couralet et al. 2010) and tropical forests in Bolivia Guinea and the South Atlantic Ocean in October–Decem-
(Brienen and Zuidema 2005) and Thailand (Pumijumnong ber influenced tree growth in the Mayombe. This was also
et al. 1995). In Ivory Coast, there is no important influence the case in Benin with significant correlations between tree
of precipitation on tree growth. growth and May–August SSTs from the Gulf of Guinea for
Thus, no comparable relation appears to exist between four tree species (p \ 0.05): Afzelia, Pterocarpus, Dan-
tree growth and precipitation in the two regions, despite iellia, and Isoberlinia (Schöngart et al. 2006).
long-distance connections for growth on one hand and Finally, anomalous ENSO conditions were associated
November–April precipitation on the other hand. Differ- with growth effects in both Ivory Coast and the Mayombe,
ences in response to climate could be related to climate even though local precipitation patterns did not influence
sensitivity of the trees or the use of short precipitation time tree growth in both regions. So far, the effect of ENSO on
series. Regarding climate sensitivity, western Ivory Coast African tree growth in two different hemispheres has not
is located in a wetter region than the Mayombe. This could been encountered in literature. Because the influence of
cause a stronger climate signal in the Mayombe following ENSO on precipitation occurred during the growing sea-
the principle of limiting factors (Fritts and Swetnam 1989). son, tree growth could be influenced by ENSO. Schöngart
Also, trees near the margins of their climatic distribution et al. (2006) only found an influence of ENSO on precip-
have a higher climate-sensitivity (Cook and Kairiukstis itation during the dry season and subsequently, no rela-
1990). The microclimate in the Congolese Mayombe still tionship with tree growth. In general, the Mayombe region
enables the growth of tropical trees while in fact, annual appears to be more sensitive to climate variability than
precipitation is too limited for their growth (optimal annual Ivory Coast, with higher October–November precipitation
precipitation for T. superba is 1,500 mm; Groulez and and higher growth clearly observed during La Niña years.

123
 Trees

In Ivory Coast, higher precipitation during El Niño years Niño years and, on the other hand, higher growth at the
slightly northward of the sample sites was found during the southern hemisphere sites during La Niña years.
growing season, probably resulting in higher growth during Importantly, if length or quality of precipitation data
El Niño years. In literature, most studies in West and hampers proper correlation with tree ring data, comparison
Central Africa are focused on the relation between pre- with large-scale climate factors as SSTs is valuable and can
cipitation and ENSO. Camberlin et al. (2001) found a pinpoint at important influences on precipitation patterns
negative correlation between ENSO and precipitation in and tree growth. Therefore, in tropical regions, where cli-
August–November for Cameroon, Gabon, the Republic of mate data are often scarce, comparison with these SSTs
Congo and coastal DRC, confirming the higher precipita- and derived ENSO indices offers interesting possibilities
tion during La Niña years in the Mayombe, as did Paeth for future dendroclimatological studies in the tropics, even
and Friederichs (2004). For Ivory Coast, Paeth and in equatorial Africa.
Friederichs (2004) did not observe a consistent relationship
for West Africa between ENSO and precipitation. How- Acknowledgments This research project is funded by a PhD grant
(M. De Ridder) of the Flemish Interuniversity Council (VLIR). The
ever, we compared the actual zones of higher precipitation fieldwork in Ivory Coast was supported by a grant from the King
in October–November during El Niño years in Fig. 7c with Leopold III Fund for Nature Exploration and Conservation and the
the precipitation response map of Camberlin et al. (2001), Congolese fieldwork was possible with the help of a grant from VLIR.
Fig. 4, resulting in a response type with some dry months We are indebted to the Special Research Fund of Ghent University for
financing the PhD project of W. Hubau. We would also like to thank
during El Niño years but with wet conditions in October– the teams of WWF DRC, WWF Belgium, Soforma, the ERAIFT
November. Still, this type of response in October– (École Régionale post-universitaire d’Aménagement et de gestion
November is not as strong as the response in the Mayombe, Intégrés des Forêts et Territoires tropicaux), Thanry and Bomaco for
both for precipitation as well as growth. Again, the prin- their financial support and their guidance throughout the fieldwork.
Special thanks goes out to Guy Bayens for all possible help on
ciple of limiting factors (Fritts and Swetnam 1989) and the organizing the Ivory Coast fieldwork and to Laurent Nsenga, Geert
presence of T. superba trees near the margins of their cli- Lejeune, Bruno Pérodeau and Prof. Shango Mutambwe, whose efforts
matic distribution (Cook and Kairiukstis 1990), could were indispensable for the success of the field campaigns in the DRC.
explain the higher climate-sensitivity in the Mayombe. Also a warm thank you to the local crew who guided us through the
forests of West and Central Africa.
From the abovementioned studies on precipitation and
SSTs, it is evident that results are often linked to the
methods used and the area and season studied. The diver-
sity in results underlines the complexity of our study region References
in its response to seasonal and interannual SST changes
Baillie MGL, Pilcher JR (1973) A simple program for tree-ring
and ENSO events. Furthermore, Camberlin et al. (2001) research. Tree Ring Bull 33:7–14
did not observe consistent relationships with any of the Balas N, Nicholson SE, Klotter D (2007) The relationship of rainfall
SST indices in some parts of Africa, suggesting that pre- variability in West Central Africa to sea-surface temperature
cipitation variability responds to other oceanic patterns, or fluctuations. Int J Climatol 27:1335–1349
Boninsegna JA, Argollo J, Aravena JC, Barichivich J, Christie D,
only to local/regional atmospheric dynamics. Ferrero ME, Lara A, Le Quesne C, Luckman BH, Masiokas M,
Morales M, Oliveira JM, Roig F, Srur A, Villalba R (2009)
Conclusion and perspectives Dendroclimatological reconstructions in South America: a
review. Paleogeogr Paleoclimatol Paleoecol 281:210–228
Bräuning A, Volland-Voigt F, Burchardt I, Ganzhi O, Nauss T, Peters
Although planted and natural forests with T. superba have T (2009) Climatic control of radial growth of Cedrela montana
distinct differences in site conditions such as structure and in a humid mountain rainforest in southern Ecuador. Erdkunde
light competition, both forest types showed significantly 63:337–345
similar growth patterns and responses to climate. Cambial Brienen RJW, Zuidema PA (2005) Relating tree growth to rainfall in
Bolivian rain forests: a test for six species using tree ring
growth probably occurs at the same time in planted and analysis. Oecologia 146:1–12
natural forests and was influenced by the onset of the rainy Camberlin P, Janicot S, Poccard I (2001) Seasonality and atmospheric
season in the Mayombe. Unfortunately, precipitation did dynamics of the teleconnection between African rainfall and
not correlate with the Ivorian regional chronology. This is tropical sea-surface temperature: Atlantic vs ENSO. Int J
Climatol 21:973–1005
possibly the result of limited precipitation data, rendering Cook ER, Kairiukstis LA (1990) Methods of dendrochronology.
the hypothesis on cambial growth shifts with precipitation Applications in the environmental sciences. Kluwer Academic
inconclusive. However, regional climate variability was Publishers, Dordrecht
influenced by SST anomalies in the Mayombe whereas Couralet C (2010) Community dynamics, phenology and growth of
tropical trees in the rain forest Reserve of Luki, Democratic
ENSO indices controlled climate variability over both Republic of Congo. PhD thesis, Ghent University
hemispheres in October–November with, on the one hand, Couralet C, Sass-Klaassen U, Sterck FJ, Bekele T, Zuidema PA
higher growth at the northern hemisphere sites during El (2005) Combining dendrochronology and matrix modelling in

123
Trees

demographic studies: an evaluation for Juniperus procera in Haneca K, Cufar K, Beeckman H (2009) Oaks, tree-rings and wooden
Ethiopia. For Ecol Manage 216:317–330 cultural heritage: a review of the main characteristics and
Couralet C, Sterck FJ, Sass-Klaassen U, Van Acker J, Beeckman H applications of oak dendrochronology in Europe. J Archaeol Sci
(2010) Species-specific growth responses to climate variations in 36:1–11
understory trees of a Central African rain forest. Biotropica Hawthorne WD (1995) Ecological profiles of Ghanaian forest trees.
42:503–511 Tropical forestry papers 29, University of Oxford
CTFT (1959) Monographie du limba (Terminalia superba Engler et Humblet P (1946) Aménagement des forêts climatiques tropicales au
Diels). CTFT, Paris Mayumbe. Bull Agric Congo Belge 37:15–87
CTFT (1983) Bois tropicaux (5ième edition). CTFT, Nogent-sur- Hummel FC (1946) The formation of growth rings in Entandro-
Marne phragma macrophyllum A. Chev. and Khaya grandifolia C. DC.
De Ridder M, Hubau W, Van den Bulcke J, Van Acker J, Beeckman Int For Rev 25:103–107
H (2010) The potential of plantations of Terminalia superba Joly M, Voldoire A, Douville H, Terray P, Royer J-F (2007) African
Engl. & Diels for wood and biomass production (Mayombe monsoon teleconnections with tropical SSTs: validation and
Forest, Democratic Republic of Congo). Ann For Sci 67:501. evolution in a set of IPCC4 simulations. Clim Dyn 29:1–20
doi:10.1051/forest/2010003 Kennedy JJ, Rayner NA, Smith RO, Saunby M, Parker DE (2011a)
Détienne P, Mariaux A (1970) La périodicité de formation des cernes Reassessing biases and other uncertainties in sea-surface temper-
dans le bois d’okoumé. Bois For Trop 131:37–50 ature observations since 1850 part 1: measurement and sampling
Détienne P, Mariaux A (1975) Nature et periodicité des cernes dans le errors. J Geophys Res 116:D14103. doi:10.1029/2010JD01521
bois de Niangon. Bois For Trop 159:29–37 Kennedy JJ, Rayner NA, Smith RO, Saunby M, Parker DE (2011b)
Détienne P, Mariaux A (1976) Nature et périodicité des cernes dans le Reassessing biases and other uncertainties in sea-surface temper-
bois de samba. Bois For Trop 169:29–35 ature observations since 1850 part 2: biases and homogenisation.
Détienne P, Mariaux A (1977) Nature et périodicité des cernes dans J Geophys Res 116:D14104. doi:10.1029/2010JD015220
les bois rouges de méliacées africaines. Bois For Trop 175:52–61 Köppen WP, Geiger R (1930) Handbuch der Klimatologie. Verlag
Devall MS, Parresol BR, Wright JS (1995) Dendroecological analysis von Gebrüder Borntraeger, Berlin
of Cordia alliodora, Pseudobombax septenatum and Annona Mariaux A (1969) La périodicité des cernes d’accroissement dans le
spraguei in Central Panama. IAWA J 16:411–424 bois de Terminalia superba. Bois For Trop 128:39–54
Douglass AE (1920) Evidence of climatic effects in the annual rings McKnight TL, Hess D (2000) Climate zones and types: the Köppen
of trees. Ecology 1:24–32 system. Physical geography: a landscape appreciation. Prentice
Douglass AE (1941) Crossdating in dendrochronology. J For Hall, Upper Saddle River
39:825–831 Mitchell TD, Jones PD (2005) An improved method of constructing a
Dünisch O, Montoia VR, Bauch J (2003) Dendroecological investi- database of monthly climate observations and associated high-
gations on Swietenia macrophylla King and Cedrela odorata L. resolution grids. Int J Climatol 25:693–712
(Meliaceae) in the central Amazon. Trees 17:244–250 Nicholson SE, Entekhabi D (1987) Rainfall variability in equatorial
Eckstein D, Bauch J (1969) Beitrag zur Rationalisierung eines and southern Africa: relationships with sea surface temperatures
dendrochronologischen Verfahrens und zur Analyse seiner along the southwestern coast of Africa. J Clim Appl Meteorol
Aussagesicherheit. Forstwiss Centralb 88:230–250 26:561–578
Eshete G, Stahl G (1999) Tree rings as indicators of growth Paeth H, Friederichs P (2004) Seasonality and time scales in the
periodicity of acacias in the Rift Valley of Ethiopia. For Ecol relationship between global SST and African rainfall. Clim Dyn
Manag 116:107–117 23:815–837
FAO (1986) Atlas of African agriculture. FAO, Rome Pumijumnong N, Eckstein D, Sass U (1995) Tree-ring research on
FAO (2008) Digital soil map of the world. FAO, Rome Tectona grandis in northern Thailand. IAWA J 16:385–392
Fichtler E, Trouet V, Beeckman H, Coppin P, Worbes M (2004) Rinn F (2003) TSAP-WinTM user reference. Rinntech, Heidelberg
Climatic signals in tree rings of Burkea africana and Pterocar- Sass-Klaassen U, Couralet C, Sahle Y, Sterck FJ (2008) Juniper from
pus angolensis from semiarid forests in Namibia. Trees Ethiopia contains a large-scale precipitation signal. Int J Plant
18:442–451 Sci 169:1057–1065
Friedrich M, Remmele S, Kromer B, Hofmann J, Spurk M, Kaiser Schöngart J, Junk WJ, Piedade MTF, Ayres JM, Hütterman A,
KF, Orcel C, Küppers M (2004) The 12,460-year Hohenheim Worbes M (2004) Teleconnection between tree growth in the
oak and pine tree-ring chronology from Central Europe—a Amazonian floodplains and the El Niño-Southern Oscillation
unique annual record for radiocarbon calibration and paleoen- effect. Glob Change Biol 10:683–692
vironment reconstructions. Radiocarbon 46:1111–1122 Schöngart J, Orthmann B, Hennenberg KJ, Porembski S, Worbes M
Fritts HC (1976) Tree rings and climate. Academic Press, London (2006) Climate–growth relationships of tropical tree species in
Fritts HC, Swetnam TW (1989) Dendroecology: a tool for evaluating West Africa and their potential for climate reconstruction. Glob
variations in past and present forest environments. Adv Ecol Res Change Biol 12:1139–1150
19:111–188 Stahle DW (1999) Useful strategies for the development of tropical
Gebrekistos A, Mitlöhner R, Teketay D, Worbes M (2008) Climate– tree-ring chronologies. IAWA J 20:249–253
growth relationships of the dominant tree species from semi-arid Stahle DW, Mushove PT, Cleaveland MK, Roig F, Haynes GA
savanna woodland in Ethiopia. Trees 22:631–641 (1999) Management implications of annual growth rings in
Gourlay ID (1995) The definition of seasonal growth zones in some Pterocarpus angolensis from Zimbabwe. For Ecol Manag
African Acacia species—a review. IAWA J 16:353–359 124:217–229
Groulez J, Wood PJ (1985) A monograph on Terminalia superba. Swaine MD, Whitmore TC (1988) On the definition of ecological
Centre Technique Forestier Tropical & Commonwealth Forestry species groups in tropical rain forests. Vegetatio 75:81–86
Institute, Nogent-sur-Marne Tarhule A, Hughes MK (2002) Tree-ring research in semi-arid West
Haneca K, Boeren I, Van Acker J, Beeckman H (2005) Dendrochro- Africa: need and potential. Tree Ring Res 58:31–46
nology in suboptimal conditions: tree rings from medieval oak Therrell MD, Stahle DW, Ries LP, Shugart HH (2006) Tree-ring
from Flanders (Belgium) as dating tools and archives of past reconstructed rainfall variability in Zimbabwe. Clim Dyn
forest management. Veg Hist Archaeobot 15:137–144 26:677–685

123
 Trees

Trenberth KE (1997) The definition of El Niño. Bull Am Meteorol Wils THG, Sass-Klaassen U, Eshetu Z, Braüning A, Gebrekirstos A,
Soc 78:2771–2777 Couralet C, Robertson I, Touchan R, Koprowski M, Conway D,
Trenberth KE, Stepaniak DP (2001) Indices of El Niño evolution. Briffa K, Beeckman H (2011) Dendrochronology in the dry
J Clim 14:1697–1701 tropics: the Ethiopian case. Trees 25:345–354
Trouet V, Haneca K, Coppin P, Beeckman H (2001) Tree ring Worbes M (1995) How to measure growth dynamics in tropical trees.
analysis of Brachystegia spiciformis and Isoberlinia tomentosa: A review. IAWA J 16:337–351
evaluation of the ENSO-signal in the miombo-woodland of Worbes M (1999) Annual growth rings, rainfall-dependent growth
eastern Africa. IAWA J 22:385–399 and long-term growth patterns of tropical trees from the Caparo
Trouet V, Coppin P, Beeckman H (2006) Annual growth ring patterns Forest Reserve in Venezuela. J Ecol 87:391–403
in Brachystegia spiciformis reveal influence of precipitation on Worbes M (2002) One hundred years of tree-ring research in the
tree growth. Biotropica 38:375–382 tropics - A brief history and an outlook on future challenges.
Trouet V, Esper J, Beeckman H (2010) Climate/growth relationships Dendrochronologia 20:217–231
of Brachystegia spiciformis from the miombo woodland in south Worbes M, Staschel R, Roloff A, Junk WJ (2003) Tree ring analysis
central Africa. Dendrochronologia 28:161–171 reveals age structure, dynamics and wood production of a natural
van Oldenborgh GJ, Burgers G (2005) Searching for decadal forest stand in Cameroon. For Ecol Manage 173:105–123
variations in ENSO precipitation teleconnections. Geophys Res Trouet V, Van Oldenborgh GJ (2013) KNMI Climate Explorer: a
Lett 32:L15701 web-based research tool for high-resolution paleoclimatology.
White F (1983) The vegetation of Africa. UNESCO, Switzerland Tree Ring Res 69:3–13
Wigley TML, Briffa KR, Jones PD (1984) On the average value of
correlated time series, with application in dendroclimatology and
hydrometeorology. J Clim Appl Meteorol 23:201–213

123