Holocene environmental and hydrological changes in Congo Basin using using sediment δ13C Bentaleb I*, Nguetsop F, Favier C, Moreau M, Martin C, Bietrix S, Aleman J, Oslisly R, Fontugne M, Bremond L, Ngomanda A, Sebag D, Debret M, Maley J, Willis K, Brncic T, and members of
*Ilham.bentaleb@univ
[email protected]
Introduction Long-term ecological records are essential to understanding past responses of vegetation to climate change and human activity. In the frame of Coforchange, we aim to reconstruct past conditions that prevailed during the Holocene in a region holding currently the world's second largest rainforest. Tropical Moist Forest (TMF) extension in Central Africa has experienced dramatic changes over the last millennia(1), related to variation in rainfall patterns or to anthropogenic pressure. These disturbances are thought to still influence today’s repartition of forest and savannas, and species distributions. Are the light-demanding canopy plants in large TMF a recovery phase from past anthropogenic disturbances or from dry episodes? Lake and peat sediments contain a variety of paleoenvironmental proxies to reconstruct local to continental climate histories. Here we focus on carbon stable isotopes (δ13C) and Carbon/Nitrogen indicators of modern surface soil, plants and 10 sedimentary cores in Central Africa to reconstruct spatial and temporal ecosystem perturbations and savanna corridors.
Material & Methods
δ13C
Principle: in soil organic matter reflect the balance between plants with C3 and C4 photosynthetic pathways (PP). Principle All trees and ~ 50% of grasses possess C3 and C4 PP respectively (2). In tropical environments, most savanna grasses are C4 plants with typical δ13C of -13‰ very different compared with C3 plant (-27‰) (3). Soil δ13C is a potential indicator of C3/C4 balance widely used for the qualitative description of forest savanna dynamics and grassland/woodland boundaries (4). This proxy has already been calibrated quantitatively in savanna against plant biomass (5) to reconstruct past vegetation. Carbon isotope discrimination (∆ ∆13C) is also used as a measure of intrinsic water-use efficiency (WUE photosynthesis divided by stomatal conductance, A/gs) which can be calculated from carbon isotope values. Limitation: Limitation Lake sedimentary OM may results from a complex combination of autochtonous and/or allochtonous material. Hence the δ13C may reflect these diverse sources. Autochthonous OM has low C/N ratios (20 (6). Hence we measured δ13C and C/N ratios of modern surface sediment and plants from CAR, Gabon and Cameroon for calibrating our proxies needed to reconstruct past vegetation changes archived in 15 well dated (7) sedimentary cores (See Fig 1&2) .
2
1
7 4 5
Bangui
3 Matures forests 6
Pioneer forests
Yaoundé Yaoundé
Swampy forests Herbaceous savanas
11 10
Woody savanas Ghana 1-Bosumtwi Benin 2-Sélé Cameroon 3-Barombi Mbo 4-Shum Laka 5-Bambili 6-Ossa 7-a)Mbalang, b)Tizong
Results
M4&T2 cores
9 8
Ouesso
12
Central African Republic 9-core MPLG2 + Kajack Gabon 13 10-Maridor, Oyane 11-Nguène 12-Kamalété Congo 8-a) Mopo bai, b) CORE10, c) Goualougo 13-Kitina 14-Sinnda 15-Ngamakala
14
MPLG2 15 core CORE 10
Brazzaville
Figure 1 : Map of current vegetation and location of the cores (cores of sites 8 and 9 were studied in the frame of Coforchange) Nguetsop et al. In prep.
A
C3 Land plants
C4 Land plants
B
C3 Land plants
- 6
C4 Land plants
C 20
-4
30
12
-6 8
Other plants
0
4000
6000
Freshw ater algae
0
Adam awa Plateau-North Cam eroon Cores T2&M4
-26
-30
0
+ -8
8000
2000
4000
6000
Marine algae
Freshw ater algae
Marine algae
-40 -35 -30 -25 -20 -15 -10
-5
-40 -35 -30 -25 -20 -15 -10
2000
-5
4000
6000
8000
Age (yrs cal BP)
δ 13C (per mil)
C/N 30 CAR- MPLG2
20 15 10
-19
D
5 0
2000
CAR- Swamp
4000
6000
Conclusion & perspectives
8000
4
13
-23
Filtered data: C/N>10
3
1,35 1,3 1,25
-27 13 ∆ C (‰)
1,2
-31 0
2000
4000
6000
8000
2
1,15 1,1
1
1,05
CONGO-Swamp
-20
0,95
18
-1 4
C/N
Algae contribution
6
150
1 100
4 2000
4000
6000
0
8000
50
Age Cal BP
SCIENTIFIC PARTNERS
University of Bangui, Central African Republic University of Yaoundé I, Cameroon University Marien Ngouabi, Republic of the Congo CIFOR : Centre for International Forestry Research, Indonesia
200
2
Cu/TiO2
8
∆ C (‰)
10
-32
0
Republic of the Congo
250
3
12
-28
-36
-1 4
0 80,0 70,0
3
60,0 50,0
2
40,0
1
30,0 20,0
0
10,0
-1 1000
0,0
2000
3000
4000
Age Cal BP
5000
6000
Sn/TiO2
∆ C (‰)
Literature cited 1- Maley J. (1996). Proc R Soc Edinb 104B: 31-73 (2) Kelly et al 1998. Geoderma 82: 59-81. (3) Boutton et 1998. Geoderma 82: 5-41. (4) Schwartz et al 1996 Oecologia 106: 516-524. Guillet et al 2001. Journal of Tropical Ecology 17: 809-832. McPherson et al. 1993 (5) Gillson et al 2004. Journal of Vegetation Science 15: 339-350. (6) Meybeck, 1982 Am. J. Sci., 282, 401–450, 1982. (7) More than 50 radiocarbon dates were obtained thanks to ARTEMIS facility (C. Moreau, J-P Dumoulin). (8)Aleman et al. 2011 Journal of Vegetation Science.(9) Nguetsop et al. 2011.Climate of the Past, 7, 1371-1393.
13
CIRAD : French Agricultural Research Centre for International Development, France CNRS : National Centre for Scientific Research, France FRM : Forest Resources Management, France FUSAGx : Gembloux Agricultural University, Belgium IRD : Research Institute for Overseas Development, France JRC : Joint Common Research Centre, Italy University of Oxford, United Kingdom University of Aberdeen, United Kingdom University of Bayreuth, Germany IRET : Research Institute in Tropical Ecology, Gabon MINRST : Ministry of Scientific Research and Technical Innovation, Republic of the Congo CRDPI : Research Centre on Sustainability and Productivity of Industrial Plantations,
0,9
300
14 CORE 10 CONGO
13
13 δδ C (‰)
-24
1
0
16
C) C/N and isotope ratios of 3 sediment cores D) Decrease of the Fe/TiO2: less iron transport, less rain Climate shift towards dryer conditions. In phase with ecosystem water stress. Increase of the Cu and Sn during the dryer episode? Hypothesis: Atmospheric input, higher NE trade winds? Changes in rain belt movements. Higher seasonality of precipitation
12000
25
δ C
A) C/N versus δ13C of plants of Cameroon and modern soil of CAR (8) and Gabon show lower C/N compared with expected terrestrial plants and are well above aquatic plants typical values (Rectangles). C/N>10 may be typical of African TMF ecosystems. Hence relatively low C/N ratios of terrestrial material show intermediate values of C/N (10-20) not necessarily linked to a mixture of aquatic and terrestrial sources (9; Grey triangles: CAR liters ). B) C/N ratios of the lacustrine or swamp sediments are above 10 except CORE 10 (Congo) of which half the samples have C/N10): As soil water becomes less available, the carbon isotope ratio increases (δ13C), discrimination decreases (∆13C), and plant intrinsic water-use efficiency goes up (WUE).
Lakes
-34 0
8000
Age cal BP
Fe/TiO2
Liter from CAR
10
2000
-22
13
20
0 -2
16
δ δ C(‰)
C/N
Orchids
13
C/N
40
∆ C (‰)
2
WUE
50
Liter : Cam eroonGabon
M4-Cam T2_Cam Sele Lake-Bénin Ghana Bosumtwi Barombi-Mbo-Cam CAR-South-MPLG2 Ossa4-Cam N-Congo-Core 10 NGUEN-Gab KAMALETE-Gab
4
DRYER
60
E
Fossil organic matter: matter: sediment core
WETTER
Modern plant and soil
- δ13C useful indicator of water availability to the ecosystem. C/N potential indicator of the water table level (still need calibration) - Responses of ecosystems to the dryer climate (ITCZ shift ) ~ 3Kyrs are spatially variable (Savanna in Northern Cameroon; swamp in CONGO is less humid and a semisemi-deciduous forest might develop). - Current savanna of the Adamawa plateau began about 3 kyr, kyr, open forest in Congo and Benin were present 5Kyrs ago. Some areas did not changed except the last 1000 years (Ossa (Ossa,, CAR). - Not all forest ecosystems are equally resilient to the climatic disturbance: Nguen, Nguen, BarombiMbo forest resisted to environmental changes. - Links with archeology: Bantu expanssion during the drying phase