A horizon depth as ecological indicator for grassland: Mapping approach by using geophysical methods GRELLIER Séraphine (1) , FLORSCH Nicolas (2) , JANEAU Jean-Louis (1) , LORENTZ Simon (3) , PODWOJEWSKI Pascal (1) References : Wilcox B.P. (2002). Shrub control and streamflow on rangelands: A process based viewpoint. Journal of Range Management 55: 31 8-326. Hibbard K.A., Archer S., Schimel D.S., Valentine D.W. (2001). Biogeochemical changes accompanying woody plant encroachment in a subtropical savanna. Ecology 82: 1999–2011. 1. INTRODUCTION 2. METHODS Study area: Potshini catchment (near Bergville), representative of the KwaZulu-Natal Drakensberg foothills - 28 48' 37" S; 29 21' 19" E. In the Slingram EM38 apparatus, one coil serves as a transmitter and produces an alternative magnetic field in the ground. It induces an electric field ( electric field and magnetic induction). The later leads to a density current where σ is the conductivity. These currents produce a secondary magnetic field which is measured by using the receiving coil. Hence the secondary field reflect the conductivity. 1 IRD c/o School of Bioresources Engineering and Environmental Hydrology (BEEH), Rabie Saunders Building, UKZN, Box X01, Scottsville, 3209, South Africa. 2 UMMISCO/IRD, 32, avenue Henri Varagnat, 93143 Bondy Cedex, France; UPMC, Paris; Dept of Mathematics and Applied Mathematics, UCT, South Africa. 3 School of Bioresources Engineering and Environmental Hydrology (BEEH), Rabie Saunders Building, UKZN, Box X01, Scottsville, 3209, South Africa. Ecological and soil survey: Trees mapping and topography have been realized with DGPS (Leica). Grain size fractions (pipette method) were measured every 5cm until 65cm depth (always after reaching the B horizon). 3. RESULTS and DISCUSSION 4. CONCLUSION 0 10 20 30 40 50 50 60 70 80 90 100 0 10 20 30 40 50 50 60 70 80 90 100 0 10 20 30 40 50 50 60 70 80 90 100 1 2 In terfac e d ep th 2 3 4 5 6 7 8 9 10 11 12 13 14 15 23 25 27 29 31 33 35 37 39 41 43 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 m S/m m S/m m 0 10 20 30 40 50 50 60 70 80 90 100 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 0 10 20 30 40 50 50 60 70 80 90 100 0 10 20 30 40 50 50 60 70 80 90 100 0 10 20 30 40 50 50 60 70 80 90 100 0 10 20 30 40 50 50 60 70 80 90 100 0 10 20 30 40 50 50 60 70 80 90 100 V ertical D ip o le M ode (g ro u n d ed ) app. H o rizo n talD ip o le M ode (g ro u n d ed ) app. V e rtica l D ip o le M ode (0 .5 m ab o v e th e g round) app. 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 12 13.5 14 14.5 15 15.5 16 16.5 17 17.5 18 18.5 19 19.5 20 20.5 21 21.5 22 22.5 23 23.5 24 24.5 m S/m m S/m m S/m F ield d ata R econstructed syn th e tic d ata In verte d param eters In versio n o f E M 38 d a ta to re trie v e th e co n d u c tiv itie s o f A and B h o rizo n s an d th e in te rfa c e d ep th Figure 1 1 2 In terface d ep th Validation by using measurements on a gully face and close electrical sounding. Clay amount and electrical resistivity were both measured on an up-dated face. A p p a re n t re sistivity (O hm.m) D e p th (m) R e m o te effe ct o f th e g u lly fa ce S yn th e tic s o unding d e rive d fro m the re sis tivity lo g R e sis tivity lo g m e a su re d on th e g u lly fa ce In te rp re ta tio n o f th e near g u lly s o unding. E xperim entalS clum berger so u n d in g 4 m e te rs a part fro m th e gully C la y a m o u n t (% ) L o w e r lim it o f th e A h o rizon S im u la tio n o f th e gu lly fa ce T h e d e e p e r la ye r is fic tive and u se d to ta ke in to a cco u n t th e g u lly fa ce 4 m apart to m o d e lth e d e e p a p p a re n t re sistivity 0.1 1.0 10 10 100 1000 T ru e re s istivity (O hm .m ) S p a cin g (m) 10 100 1000 0.01 0.1 1.0 10 E lectrica l sounding R e sistivity log Bayesian method: The Bayesian inverse computation is based on: While data are Gaussian-like, one makes use of: Relative sensibility of the 3 modes as a function of depth (layered medium) Red: VDM Blue: HDM (normalized) Brown: VDM at 50 cm height (normalized) Slingram method: The ground integrating probe response depends whether the dipoles are handled vertically (VDM) or horizontally (HDM): this provide two independent measurements. A third one is obtained by hanging the device 50cm above the ground. But using three three measures, one can retrieve the three parameters of a two layer shallow sub-surface: the two conductivities and the depth of the interface. Principle of the EM38 EM38 survey, vertical position on the ground (VDM). a prio riin fo rm ation o n data a prio riinform ation o n pa ram e te rs a p o sterio ri in form ation o n p ara m ete rs p h ysica llaw be tw e e n data a n d p a re a m e te rs [d =G (m )] fro m th is p d f, co m p u tatio n of m a rg in al p d f, m ean s an d m o m en ts fo r a ll p ara m e te rs (co n d u c tiv ity o f th e tw o la yers an d th e firs t la y e r th ickness) Figure 2 control name EM38 Sounding point 1 25 26 point 2 31 42 point 3 49 43 point 4 23 23 Depth (cm) of transition of A and B horizons on 4 control points. All controls are well validated except point 2, which can be explained by a non-two layers structure at this point: heterogeneity of the grassland appears here. 0 10 20 30 40 50 50 60 70 80 90 100 1308 1308.5 1309 1309.5 1310 1310.5 1311 1311.5 1312 1312.5 1313 1313.5 1314 1314.5 1315 1315.5 1316 T re e s lo ca tio n (d ot) an d to p o g ra p h y (c o lor) m T rees d en s ity p e r o ne h u n d re d sq u are m e te rs n u m b er/are L e ft figu re : co rre la tio ns b e tw e e n tre e s de n s ity a n d co n d u ctivity. T o p o g ra p h y (rig h t) an d d e n sity o f tre e s ( c e n te r). O n ly o n th e ha lf u p p e r rig h t p a rt th e tre e s h a ve grow n. 0 5 10 15 20 25 30 35 40 45 50 50 55 60 65 70 75 80 85 90 95 100 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 seedlings b ig trees alltrees F ig u re 3 20 24 28 32 36 40 44 -5 0 5 10 15 20 25 H ig h cond uctivity (m o re cla ye y ) = => lo w tre e d ensity L o w condu ctivity (le s s c la ye y ) = => h ig h tre e d e nsity 2 T h e re is n o h ig h tre e d en sity w h e re h ig h cla y a m o u n t o ccurs Very low conductivity ==> low tree density There is no high tree density where high and very low clay amount occurs The expanding grasslands in Southern Africa and all around contribute to agro-pastoral activities and to the evolution of the ecological quality of soils. In this context, grasslands are sometimes invaded by trees, which have eventually a strong impact on the ecosystemss (Hibbard et.al., 2001, Wilcox et al. 2002). The mapping of the A horizon, which is involved in the resistance to erosion, could be a relevant indicator of the determinisms and interactions contributing to asses soil quality and soil evolution in the landscape. Objectives here: to understand the potential involving role of the A and B horizons in the presence of invading trees (Acacia Sieberiana) by using geophysical methods (Slingram). The transition between the A and the B horizon (depth < 0.6 m) appears stiff in term of conductivity contrast (A being rather resistive while B is clayey and conductive). As a very robust inversion procedure, the Bayesian approach is efficient to map the A horizon thickness and both layer conductivities, and reveals large clay variations in the B horizon.