Version classiqueVersion mobile

Rural societies in the face of climatic and environmental changes in West Africa

 | 
Benjamin Sultan
, 
Richard Lalou
, 
Mouftaou Amadou Sanni
, 
et al.

Part II. Impacts of the climate and environmental changes

Chapter 6. Between desertification and regreening of the Sahel

What is really happening?

Cécile Dardel, Laurent Kergoat, Pierre Hiernaux, Manuela Grippa et Éric Mougin

Texte intégral

Introduction

1This chapter is aimed at reviewing the ‘desertification’ of the Sahel, a source of polemic for several decades, and its ‘regreening’, a term that appeared with the first satellite observations of plant cover in the 1970s. The debate between the backers of these two diametrically opposed theories is real, and particularly important because this part of the world is known for its high sensitivity to climatic events.

2The specific contribution of this work lies in the combined use of remote sensing data at the scale of the Sahel and covering the last 30 years and analysis of long-term field measurements in Mali and Niger. The combination of these different data sources allows better understanding of the evolution of plant cover in the Sahel over the last three decades and checking of the coherence of the satellite observations performed for this purpose. It is thus seen that regreening is undeniable, in particular at the scale of the Sahel, but that opposite trends can also be seen at smaller scales, which means that caution is required in overall diagnosis of the long-term evolution of plant cover.

3For many years, the desertification of the arid and semi-arid parts of the world has attracted the attention of scientists and also various international organisations operating for the environment, the media and civil society. Indeed, these regions are very sensitive to climatic variations and in particular the variability of precipitation. The Sahelian climate has experienced a succession of wet and drier periods for tens of thousands of years. In addition, the population has increased very strongly for several decades and this has been accompanied by modifications to the environment that are sometimes very marked (cultivated land, land clearance, deforestation, fires, etc.).

4The Sahel region has recently experienced two very serious droughts—in the 1970s and then again in the mid-1980s. These droughts happened during three decades of overall rainfall deficits. The consequences of the drought periods were serious for the population and the environment and contributed to reviving the theory of the desertification of the Sahel. For example, several authors have mentioned a spectacular spread of the Sahara that would soon threaten all African arable land (Hubert, 1920; Lamprey, 1975; Stebbing, 1935). However, for lack of more appropriate resources, the studies published were based essentially on local observations at a given moment. Typically, the state of the vegetation was observed at a specific place and at a precise instant to draw up a diagnosis of local ‘desertification’ or to extrapolate this to plot desertification maps at much larger scales (continental, global) (Uncod, 1977; Unep, 1992). But several authors questioned this vision, in particular by examining the seasonal cycle of vegetation and its inter-annual variation (Boudet, 1979; Jones, 1938) or by calling into question the methods used by the various institutions to spatialise desertification diagnoses (Mabbutt, 1984; Veron et al., 2006).

5In this context, the arrival of satellite remote sensing in the 1980s provided an extremely valuable tool, in particular thanks to the vegetation indexes used for the first monitoring of vegetation at the global scale. One of the most commonly used vegetation indexes is the NDVI (Normalized Difference Vegetation Index), calculated from the light reflected in red and near-infrared wavelengths (see Equation 1), since green vegetation absorbs red wavelengths and reflects near-infrared. Numerous studies have shown that the NDVI is linked to the fraction of active photosynthetic radiation absorbed by plants (fPAR) (Asrar et al., 1985; Myneni et al., 1995; Sellers, 1985), and other vegetation variables (all linked with each other) such as plant cover (fCover: the fraction of the ground covered by green vegetation seen vertically from above the cover) and the leaf area index (LAI: leaf area per unit area of ground). Finally, the integral of the NDVI in time is considered to provide a good assessment of vegetation production, in turn defined as the quantity of plant tissues produced during a given period, generally that of the plant growth season (Monteith, 1972).

Équation 1. NDVI formula in which ρIR = reflectance in near infra-red (PIR) and ρR = reflectance in red wavelengths (R).

‘Regreening’, a term dating back to the first satellite analyses

6The data collected by AVHRR radiometers installed in NOAA satellites and launched in the early 1980s provided the first opportunity to analyse NDVI at the continental stage and at a time step suitable for monitoring vegetation. The datasets drawn from AVHRR data are still among those most frequently used today for the study of long-terms trends in plant cover as they have unequalled historical depth: several sensors launched one after the other have provided a continuous dataset from 1981 to the present day. The GIMMS dataset in particular has been much used; the first version is available for 1981 to 2006 with spatial resolution of 8 km and bi monthly frequency.

7The first analyses of temporal trends of NDVI AVHRR in the Sahel date from the early 1990s as data for 10 years were available. In a study using satellite observations over a long period to examine the hypothesis of the Sahara Desert spreading, Tucker and Nicholson (1999) showed that in fact the “expansion” of the Sahara depends on the inter-annual variability of precipitation. In other words, the hypothesis of an inexorable advance of the Sahara, the supposed sign of the desertification of the Sahel, was invalidated. To go further, the NDVI trends shown are were fact positive and statistically significant trends. These positive trends gave birth to the term ‘regreening’ of the Sahel, defined simply as an increase of the vegetation index in time. As further years of observations were added to satellite records, this regreening of the Sahel was documented in a number of studies (Anyamba and Tucker, 2005; Fensholt et al., 2013; Herrmann et al., 2005; Tucker and Nicholson, 1999).

8A new version of the GIMMS dataset has appeared recently and is called GIMMS-3g (for ‘third generation’). Data are provided for the period running from 1981 to 2011 at a spatial resolution of 1/12° and bi-monthly time frequency. The coherence of this dataset in comparison with other existing NDVI datasets was examined and demonstrated in Dardel et al. (2014 b) and Dardel (2014). Here, we focus only on temporal trends throughout the entire period for which GIMMS data are available, that is to say from 1981 to 2011 (Fig. 1).

9The NDVI GIMMS-3g trends are thus positive in most of the Sahel belt for the whole of the period 1981-2011. Only a few regions like western Niger and central Sudan display statistically significant negative trends. The signature of the regreening of the Sahel is thus clear and confirms recent studies (Fensholt et al., 2012), this time for the entire period for which AVHRR data are available.

10However, as for all data sources, satellite data are subject to a number of limits and to bias, especially in long-term series. It is therefore essential to verify that the signal detected by satellite is truly linked to the evolution of vegetation and not to satellite artefacts. One way of doing this is to compare them to data collected in the field at a spatial scale compatible with satellite measurement. However, such observations are very difficult to make, especially for large areas and for long periods of time, and are thus rare. Within the framework of international projects such as AMMA (Analyse multidisciplinaire de la mousson africaine–Multidisciplinary analysis of the African monsoon) and previously under the aegis of the ILRI (International Livestock Research Institute), plant production measurements have been performed in Mali since 1984 (Hiernaux et al., 2009 b; Mougin et al., 2009) and in Niger since 1994 (Hiernaux and Ayantunde, 2004; Hiernaux et al., 2009 a). In addition, these two study locations are particularly interesting as they display trends with opposite signs: while the Gourma in Mali is ‘regreening’, the Fakara in Niger is marked by strong negative trends (Fig. 2). Comparing the trends found by remote sensing with long-term field observations is thus a real test for satellite archives and for their performance in terms of the monitoring of vegetation.

Figure 1. Temporal trends in NDVI GIMMS3g averaged from July to October for the period 1981 to 2011 in the Sahel belt.
Zones that are not statistically significant (P < 0.05) are greyed.
Source: Dardel et al. (2014 b).

x Observation sites
Figure 2. Temporal trends in NDVI GIMMS3g averaged for the growth season (July to October) from 1981 to 2011: a close-up on the Gourma region in Mali (rectangle on the left) and the Fakara in Niger (square on the right).
The crosses represent pixels at which vegetation sites are present (for the Gourma region only). Source: Dardel et al. (2014 b).

Comparison with in situ measurement networks (Mali, Niger)

The Gourma in Mali

11The vegetation in the Gourma consists mainly of annual herbaceous plants. The soil is thus mainly bare during the dry season, with the exception of a few perennial plants and scattered shrubs. Herbaceous plants grow if there is sufficient rainfall during the growth period, that is to say from June to October (Fig. 3). The ligneous stratum forms less than 5% of the total area, as do cultivated areas. The land is used mainly for grazing.

12Within the framework of a long-term ecological field study, annual plant production is estimated by measuring the aerial mass of the herbaceous stratum, with measurement in kilograms of dry matter per hectare (kg MS. ha-1). A destructive method was used for the measurements: the grass was cut, dried and weighed for 1 m × 1 m squares chosen at random along transects 1 kilometre long (Hiernaux et al., 2009 b). These transects sampled the variety of the landscape: the different soil types, vegetation densities, etc. Some 40 sites were thus sampled in this way practically every year from 1984 to 2011. When several measurements were performed at the same site in the same year the maximum mass gives the best estimate of annual production.

Figure 3. Annual dynamics of plant cover at Agoufou in Mali in 2005, for a) 8 April, b) 17 June, c) 19 August and d) 28 September. Source: Photos AMMA-CATCH.

13Calculation of the average of the masses measured at the 40 sites sampled from 1984 to 2011 (Fig. 2) gives a time series comparable to the NDVI averaged for the same region (the rectangle shown in the left of Figure 2). The regreening detected using satellite data is therefore confirmed by the field measurements (Fig. 4a). There is good agreement between the two independent datasets (R2 = 0.61, Fig. 4b), proving that NDVI is strongly correlated with the evolution of herbaceous vegetation and that satellites are perfectly suitable for monitoring plant production over long periods of time. Furthermore, these regreening trends demonstrate the resilience of the Sahel ecosystems in the Gourma in Mali in spite of the very severe droughts of the 1970s and 1980s.

14These first results show that the regreening of Sahelian vegetation is a proved phenomenon and that the satellite archives are fully relevant for performing longterm monitoring of vegetation, at least in these semi-arid ecosystems.

Figure 4. a) The temporal evolution of field data (mass of the herbaceous stratum) averaged from the sites in the Gourma in Mali, from 1984 to 2011. b) Correlation between field data and NDVI GIMMS-3g data from 1984 to 2011. Source: adapted from Dardel et al. (2014 b).

The case of Fakara in Niger

15The case of the Fakara is slightly different as this region in south-west Niger has experienced very strong local population growth since the 1950s, combined with a considerable increase in cultivated areas. The proportion of cultivated land surfaces has thus increased from 25% of the total area in 1986 to some 50% in 2010 (Hiernaux and Ayantunde, 2004; Hiernaux et al., 2009 a). As a result, the areas under fallow have decreased considerably; long fallows have been particularly affected. The land use system is agropastoral with a combination of cultivated fields, fallow fields and rough pasture. The pasture areas consist of land on which crops cannot be grown because the loose soil is shallow with rocky outcrops. This land is grazed in all seasons while crop fields are only grazed after the harvest until sowing in the following year. Fallows are generally grazed in all seasons except when they are surrounded by crop fields. The Fakara also features strong inter-annual land use dynamics. Thus a field cultivated in one year may be left fallow in the following year and vice versa. Estimation of plant productivity must take all these parameters into account in the sampling of the sites, making assessment a little more complicated than in the Gourma. As in the Gourma, the variable measured in the field is the aerial mass of the herbaceous stratum measured in 1 m × 1 m squares chosen at random along homogeneous transects for fallow and grazing zones. The zones under crops are sampled by planting hole or plant with density measured in parallel (Hiernaux et al., 2009 a). The division of the landscape in this region makes it necessary to use transects 200 m long in fallows and grazing areas and 100 m for crop fields (in comparison with 1 km for the Gourma sites). Finally, the historic depth in the dataset is a little shorter as the collection of field data started in 1994. However, this means 17 years of comparison with satellite data.

Figure 5. a) The temporal evolution of field data (mass of the herbaceous stratum) averaged from the sites in the Fakara in Niger, from 1994 to 2011.
b) Correlation between field data and NDVI GIMMS-3g data from 1994 to 2011. Source: adapted from Dardel et al. (2014 b).

16The temporal trends calculated from averaged field data for the Fakara region also confirm the trend observed in the NDVI averaged for the same zone: the average of the masses measured at the different sites displays a negative trend from 1994 to 2011 (Fig. 5a). Agreement between field and remote sensing data is not as good as for the Gourma in Mali (R2 = 0.38, Fig. 5b). This can be explained partially by the greater heterogeneity of the landscape and strong land use dynamics. Nevertheless, the important thing is that there again there is also coherence between the two totally independent datasets. It is obviously important to identify the cause of these trends—that are increasing in the Gourma and in most of the Sahel but decreasing in the Fakara.

Can these trends be explained by the evolution of rainfall?

17In semi-arid regions like the Sahel, the main explanatory factor of plant production is the amount of rainfall received by the ecosystems: in years of rainfall deficit production is very small or nil in extreme cases. The relation between production and rainfall may be modified if other factors disturb the rainfall/production relation (for example, factors related to human activities such as land clearance, deforestation, the formation of soil crust, salinisation, fires, livestock, etc.). Here, we examine the link between plant production and rainfall measured in the field using recording rain gauges in the Gourma in Mali and the Fakara in Niger.

The Gourma, Mali

18Rainfall in the Gourma region increased very strongly from 1984 to 2011, as is shown by the measurements from a network of recording rain gauges covering the region and whose annual cumulation anomalies are shown in Figure 6a. The increase was some 5 mm per year throughout the period 1984-2011. Strong correlation is observed between production of the herbaceous stratum averaged over the region and precipitations: 76% of inter-annual variability of production is accounted for by the evolution of rainfall (Fig. 6b).

19The following question is raised: can the regreening of the Gourma be accounted for by the recovery of rainfall during the same period?

20Rain Use Efficiency (RUE) is frequently used as an indicator to separate the influence of rainfall on plant production from other potential factors. It describes the efficiency of rain use by plants. It is calculated simply as a ratio of annual production to rainfall. Theoretically, if the environment is not changed as time goes by, rain use efficiency for pant growth should remain unchanged (Le Houerou, 1984). In contrast, if ecosystems are changed by a factor of some kind, and especially if they are degraded, rain use efficiency decreases. This decrease in RUE in time is used as an indicator of ecosystem degradation.

Figure 6. a) Anomalies in annual precipitation measured in the Gourma region from 1984 to 2011.
b) Correlation between production measured in the field and precipitation. Source: Dardel et al. (2014 b). Data provided by DMN (Direction de la météorologie nationale) Mali.

21Calculation of RUE trends for the Gourma region as a whole does not show any decrease against time (Fig. 7). On the contrary there is a slight trend for an increase but this is not statistically significant (P = 0.29). Analysis of the RUE at the scale of the region thus confirms that since the early 1980s there has been no degradation of Gourma ecosystems—or at least those that are accessible by satellite observation. However, it is seen in Chapter 9 that these conclusions should be modulated when a finer spatial scale is examined: it is possible, in fact, to detect degradation of the vegetation in a small fraction of the landscape—in certain surface soils—while the overall signal for the region remains that of pronounced greening.

22We can thus affirm today that regreening has occurred in the Gourma and that the direct cause is a partial recovery in precipitation after the major droughts of 1983-1984. At a closer scale, mechanisms that are still not well known may nevertheless affect plant production in part of the landscape but without modifying the overall regreening signature.

Figure 7. Evolution of RUE (Rain Use Efficiency) in time, calculated as the ratio of plant production measured in the field to annual rainfall (cumulated during the growth season, that is to say from July to October).
Source: Dardel et al. (2014 b).

The case of the Fakara in Niger

23The rainfall trend during the past two decades is not as clear for the Fakara region. The slight increase observed between 1990 and 2011 is not statistically significant and the same applies to the slight negative trend observed during the period 1994-2011 (Fig. 8a).

24Correlation between plant production measured in the field and rainfall measured by recording rain gauges is much poorer than for the Gourma region (R2 = 0.07, Fig. 8b).

Figure 8. a) Anomalies in annual precipitation measured in the Fakara region from 1990 to 2011.
b) Correlation between production measured in the field and precipitation from 1994 to 2011.

25However, the rainfall records used here do not correspond totally with the Fakara region where vegetation is monitored as the data used were collected in the Niamey square degree, which is a larger zone.

26As for the Gourma, analysis of RUE trends should make it possible to describe the influence of precipitation on plant production. For this, RUE trends were studied with separate examination of three types of land use: cultivated fields, fallows and rangelands. Only fallows stood out among the three types of land use as it was the only one to feature a decrease of RUE in time (Fig. 9). This means that the evolution of herbaceous plant production in fields and rangelands is clearly accounted for by the simultaneous evolution of precipitation while the areas under fallow seem to display a decrease in plant production that is not explained by the evolution of rainfall. More in-depth analysis (using high spatial resolution images for example) would be necessary to confirm this preliminary diagnosis and then to better understand the cause of the different evolutions. For example, one hypothesis might be that the strong increase in local population and the substantial cultivation of land has diminished soil fertility via the acceleration of fallow/crop rotation mechanisms (with a reduction of the longest fallows) and the cultivation of less fertile land, but this remains to be confirmed.

Figure 9. Evolution of RUE in time (estimated from field measurements of herbaceous plant mass and rainfall) in the Fakara region from 1994 to 2011.

Simulation of herbaceous plant production in the Gourma during the period 1950-2012

27Models are the most appropriate tools for setting the trends observed using in situ measurements and remote sensing data for a longer period, in particular going back to the 1950-1960 wet period. Some of the data accessible are discontinuous, such as aerial photographs, Landsat images and the first field observations (Boudet, 1972) but are often insufficient for reconstructing long trends. The STEP vegetation model (Mougin et al., 1995) was therefore used to simulate the evolution of herbaceous biomass in the Gourma over the period 1950-2012. Rainfall variability is the only variable used to run the model as a series of homogeneous rainfall data measured at the Hombori meteorological station in central Gourma for 1950-2012 is available. The other input parameters (incident short-wave radiation, air temperature, relative humidity and windspeed) were set for each year at their daily values measured at the Agoufou meteorological station in 2006, used within the framework of the AMMA programme. Sensitivity analysis showed that approximately 90% of inter-annual variation of vegetation is covered in this way.

28The production of herbaceous mass simulated by STEP is well correlated overall with precipitations (Fig. 10) but the relation is less linear for rainfall greater than 400-500 mm per year. During the period 1984-2010, correlations with simulated production (R2 = 0.50) were weaker that those calculated using in situ data (R2 = 0.76, Fig. 6b). The difference might result at least partly from the fact that the Hombori is in central Gourma and the simulation does not take into account the more arid locations in the north where the growth of vegetation should be even more dependent on available soil moisture.

29Precipitation measured at Hombori station and used as input for the STEP model and the figures for green biomass given by simulations, together with their trends during different periods, are shown in Figure 11 and summarised in Table 1. The effect of the extremely severe droughts (1984 and 2004) is clearly marked in vegetation productivity, which shows a very pronounced minimum in 1984. The trends drawn from modelling agree closely with those drawn from field observations and remote sensing for the period during which the in situ and NDVI data are available. In particular, a significant regreening signal is observed from 1981-1984 which, as has already been mentioned, is to be ascribed to the recovery of precipitation during this period.

Figure 10. Relation between biomass simulated by STEP (annual maximum) and precipitation (cumulated annual figure).

Figure 11. a) Precipitation and b) green biomass simulated by STEP and their trends during the periods 1950-2012 (black), 1981-2011 (red) and 1984-2011 (blue). All the trends are statistically significant (p < 0.05 in Student’s Test).

30In contrast, examination of a longer period that includes wet years (1950-2012) shows negative trends in both precipitation and biomass simulated by STEP. This is a sign that in parallel with precipitation that is still smaller than during the wet period the ecosystem was unable to attain the strongest biomass levels of 1950-1960.

Tableau 1. Comparison of STEP simulation trends in precipitation and biomass and the trends observed using NDVI observations and in situ herbaceous mass data. All the trends are statistically significant (p < 0.05 using Student’s Test).

Period

Precipitation trends at Hombor met. Station
(mm/period)

Trends in biomass simullated by STEP
(kg DM.ha-1/period)

Trends in biomass observed by remote sensing
(NDVI units/period) Source: Dardelet al. (2014 a)

Trends in biomass measured in situ (kg DM.ha-1/period) Source: Dardelet al. (2014 a)

1950-2011

- 167.4

- 821

1981-2011

+ 134.2

+ 552

+ 0.032

1984-2011

+ 146.2

+ 654

+ 0.05

+ 626

What is the future for Sahelian systems?

31This work shows that there has been an overall regreening of plant cover throughout the Sahelian region in the last 30 years although there are a few regions where plant cover trends are negative, as in the Fakara in Niger and the central parts of Sudan. Satellite observations were found to be a robust method for detecting changes in plant cover over long periods of time as they match the changes observed in the fields in two regions with opposite trend signs (the Gourma in Mali and the Fakara in Niger).

32However, low-resolution satellite observations such as AVHRR data are not sufficient for gaining understanding of all the mechanisms involved, as for example in the Fakara in Niger or at a scale of less than one AVHRR pixel. More in-depth analysis of these changes requires the use of other facilities such as very high spatial resolution satellite images and above all networks of field observations that provide information about numerous parameters that are difficult to observe from space—composition of the flora, grazing pressure, etc.

33In examination of the production trends imposed by climatic trends, it is important to bear in mind that they are closely dependent on the spatial and temporal scales considered, especially for diagnosis of the ‘state of health’ of vegetation in comparison with a reference situation (e.g. the beginning of the period for trend studies). We have shown that the regreening observed in the Gourma region over the past 30 years has been caused mainly by the recovery of precipitation during this period, and this is probably the case for the whole of the Sahel. However, it is probable that a small part of the landscape (such as the surface soils in the Gourma, see Chapter 9) may be subject to degradation via mechanisms that are still poorly known. These contrasted trends are not in opposition with each other as they involve different scale factors.

34The simulations performed using the STEP model allowed us to reconstitute the evolution of herbaceous plant cover for more distant periods—in this case since the 1950s. The results of these simulations show that the vegetation has not recovered to levels as high as those before the 1970s and 1980s droughts, in parallel with precipitations that are also not at the same level as in 1950-1960.

35In this context, one might legitimately wonder about the evolution of precipitation in the coming decades. For this, only modelling can give a few lines of approach but climate models diverge and do not give either the trend or intensity of the changes to be expected with any certainty. The regreening observed today must therefore not mask stronger variability or lead to forgetting the possibility that this may be just a transitory state of the vegetation in a context of change or of climatic variability in which forecasting is delicate, especially as regards precipitation.

Bibliographie

References

Anyamba A., Tucker C. J., 2005
Analysis of Sahelian vegetation dynamics using NOAA-AVHRR NDVI data from 1981-2003.
Journal of Arid Environments, 63: 596-614.

Asrar G., Kanemasu E. T., Jackson R. D., Pinter P. J., 1985
Estimation of Total above-ground Phytomass Production Using Remotely Sensed Data.
Remote Sensing of Environment, 17: 211-220.

Boudet G., 1972
Désertification de l’Afrique tropicale sèche.
Adansonia. Série, 2 (12) : 505-524.

Boudet G., 1979
Quelques observations sur les fluctuations du couvert végétal sahélien au Gourma malien et leurs conséquences pour une stratégie de gestion sylvopastorale.
Bois et Forêts des Tropiques, 184 : 14.

Dardel C., 2014
Entre désertification et reverdissement du Sahel: diagnostic des observations spatiales et in situ.
Thèse doct., univ. Paul Sabatier, Toulouse III, 199 p.

Dardel C., Kergoat L., Hiernaux P., Grippa M., Mougin E., Ciais P., Nguyen C. C,. 2014 a
Rain-Use-Efficiency: What it Tells us about the Conflicting Sahel Greening and Sahelian Paradox. Remote Sensing, 6: 3446-3474.

Dardel C., Kergoat L., Hiernaux P., Mougin E., Grippa M., Tucker C. J., 2014 b
Re-greening Sahel: 30 years of remote sensing data and field observations (Mali, Niger). Remote Sensing of Environment, 140: 350-364.

Fensholt R., Langanke T., Rasmussen K., Reenberg A., Prince S. D., Tucker C., Scholes R. J., Le Q. B., Bondeau A., Eastman R., Epstein H., Gaughan A. E., Hellden U., Mbow C., Olsson L., Paruelo J., Schweitzer C., Seaquist J., Wessels K., 2012
Greenness in semi-arid areas across the globe 1981-2007-an Earth Observing Satellite-based analysis of trends and drivers.
Remote Sensing of Environment, 121: 144-158.

Fensholt R., Rasmussen K., Kaspersen P., Huber S., Horion S., Swinnen E., 2013
Assessing Land Degradation/Recovery in the African Sahel from Long-Term Earth Observation Based Primary Productivity and Precipitation Relationships.
Remote Sensing, 5: 664-686.

Herrmann S. M., Anyamba A., Tucker C. J., 2005
Recent trends in vegetation dynamics in the African Sahel and their relationship to climate. Global Environmental Change-Human and Policy Dimensions, 15: 394-404.

Hiernaux P., Ayantunde A., 2004
«The Fakara: a semi-arid agro-ecosystem under stress». In: Report of research activities, First phase (July 2002-June 2004) of the DMP-GEF Program (GEF/2711-02-4516), Nairobi (Kenya), Ilri.

Hiernaux P., Ayantunde A., Kalilou A., Mougin E., Gerard B., Baup F., Grippa M., Djaby B., 2009 a
Trends in productivity of crops, fallow and rangelands in Southwest Niger: Impact of land use, management and variable rainfall.
Journal of Hydrology, 375: 65-77.

Hiernaux P., Mougin E., Diarra L., Soumaguel N., Lavenu F., Tracol Y., Diawara M., 2009 b
Sahelian rangeland response to changes in rainfall over two decades in the Gourma region, Mali.
Journal of Hydrology, 375: 114-127.

Hubert H., 1920
Le dessèchement progressif en Afrique occidentale. Bulletin du Comité d’études historiques et scientifiques de l’Afrique occidentale française, 3 : 401-467.

Jones B., 1938
Desiccation and the West African Colonies.
Geographical Journal, 91: 401-423.

Lamprey H. F., 1975
«Report on the desert encroachment reconnaissance in Northern Sudan: 21 October to 10 November 1975». In: Technical report: 1-7, Paris/Nairobi, Unesco/Unep.

Le Houerou H. N., 1984
Rain Use Efficiency-a Unifying Concept in Arid-Land Ecology. Journal of Arid Environments, 7: 213-247.

Mabbutt J. A., 1984
A new global assessment of the status and trends of desertification. Environmental Conservation, 11: 103-113.

Monteith J. L., 1972
Solar-radiation and productivity in tropical ecosystems. Journal of Applied Ecology, 9: 747-766.

Mougin E., Loseen D., Rambal S., Gaston A., Hiernaux P., 1995
A regional Sahelian grassland model to be coupled with multispectral satellite data. 1. Model description and validation.
Remote Sensing of Environment, 52: 181-193.

Mougin E., Hiernaux P., Kergoat L., Grippa M., de Rosnay P., Timouk F., Le Dantec V., Demarez V., Lavenu F., Arjounin M., Lebel T., Soumaguel N., Ceschia E., Mougenot B., Baup F., Frappart F., Frison P. L., Gardelle J., Gruhier C., Jarlan L., Mangiarotti S., Sanou B., Tracol Y., Guichard F., Trichon V., Diarra L., Soumare A., Koite M., Dembele F., Lloyd C., Hanan N. P., Damesin C., Delon C., Serca D., Galy-Lacaux C., Seghieri J., Becerra S., Dia H., Gangneron F., Mazzega P., 2009
The AMMA-CATCH Gourma observatory site in Mali: Relating climatic variations to changes in vegetation, surface hydrology, fluxes and natural resources.
Journal of Hydrology, 375: 14-33.

Myneni R. B., Hall F. G., Sellers P. J., Marshak A. L., 1995
The Interpretation of Spectral Vegetation Indexes. IEEE Transactions on Geoscience and Remote Sensing, 33: 481-486.

Sellers P. J., 1985
Canopy reflectance, Photosynthesis and Transpiration.
International Journal of Remote Sensing, 6: 1335-1372.

Stebbing E. P., 1935
The Encroaching Sahara: The Threat to the West African Colonies.
Geographical Journal, 85: 506-524.

Tucker C. J., Nicholson S. E., 1999
Variations in the size of the Sahara Desert from 1980 to 1997.
Ambio, 28: 587-591.

Uncod, 1977
Proceedings of the Desertification Conference (Nairobi, UNEP).
London, Pergamon Press.

Unep, 1992
World Atlas of Desertification. London, Edward Arnold, 69 p.

Veron S. R., Paruelo J. M., Oesterheld M., 2006
Assessing desertification.
Journal of Arid Environments, 66: 751-763.

Table des illustrations

Légende Équation 1. NDVI formula in which ρIR = reflectance in near infra-red (PIR) and ρR = reflectance in red wavelengths (R).
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-1.jpg
Fichier image/jpeg, 21k
Légende Figure 1. Temporal trends in NDVI GIMMS3g averaged from July to October for the period 1981 to 2011 in the Sahel belt.Zones that are not statistically significant (P < 0.05) are greyed.Source: Dardel et al. (2014 b).
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-2.jpg
Fichier image/jpeg, 149k
Légende x Observation sitesFigure 2. Temporal trends in NDVI GIMMS3g averaged for the growth season (July to October) from 1981 to 2011: a close-up on the Gourma region in Mali (rectangle on the left) and the Fakara in Niger (square on the right).The crosses represent pixels at which vegetation sites are present (for the Gourma region only). Source: Dardel et al. (2014 b).
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-3.jpg
Fichier image/jpeg, 280k
Légende Figure 3. Annual dynamics of plant cover at Agoufou in Mali in 2005, for a) 8 April, b) 17 June, c) 19 August and d) 28 September. Source: Photos AMMA-CATCH.
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-4.jpg
Fichier image/jpeg, 289k
Légende Figure 4. a) The temporal evolution of field data (mass of the herbaceous stratum) averaged from the sites in the Gourma in Mali, from 1984 to 2011. b) Correlation between field data and NDVI GIMMS-3g data from 1984 to 2011. Source: adapted from Dardel et al. (2014 b).
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-5.jpg
Fichier image/jpeg, 115k
Légende Figure 5. a) The temporal evolution of field data (mass of the herbaceous stratum) averaged from the sites in the Fakara in Niger, from 1994 to 2011.b) Correlation between field data and NDVI GIMMS-3g data from 1994 to 2011. Source: adapted from Dardel et al. (2014 b).
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-6.jpg
Fichier image/jpeg, 104k
Légende Figure 6. a) Anomalies in annual precipitation measured in the Gourma region from 1984 to 2011.b) Correlation between production measured in the field and precipitation. Source: Dardel et al. (2014 b). Data provided by DMN (Direction de la météorologie nationale) Mali.
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-7.jpg
Fichier image/jpeg, 103k
Légende Figure 7. Evolution of RUE (Rain Use Efficiency) in time, calculated as the ratio of plant production measured in the field to annual rainfall (cumulated during the growth season, that is to say from July to October).Source: Dardel et al. (2014 b).
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-8.jpg
Fichier image/jpeg, 67k
Légende Figure 8. a) Anomalies in annual precipitation measured in the Fakara region from 1990 to 2011.b) Correlation between production measured in the field and precipitation from 1994 to 2011.
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-9.jpg
Fichier image/jpeg, 101k
Légende Figure 9. Evolution of RUE in time (estimated from field measurements of herbaceous plant mass and rainfall) in the Fakara region from 1994 to 2011.
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-10.jpg
Fichier image/jpeg, 156k
Légende Figure 10. Relation between biomass simulated by STEP (annual maximum) and precipitation (cumulated annual figure).
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-11.jpg
Fichier image/jpeg, 81k
Légende Figure 11. a) Precipitation and b) green biomass simulated by STEP and their trends during the periods 1950-2012 (black), 1981-2011 (red) and 1984-2011 (blue). All the trends are statistically significant (p < 0.05 in Student’s Test).
URL http://books.openedition.org/irdeditions/docannexe/image/12340/img-12.jpg
Fichier image/jpeg, 158k

Auteurs

Le texte et les autres éléments (illustrations, fichiers annexes importés) sont sous Licence OpenEdition Books, sauf mention contraire.

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search