The relationship between vegetation associations        and floods Vegetation   associations   have   privileged   relationships   with    flood   heights,   as   shown   by   Pierre   Hiernaux ,    in   conclusion   of   the statistical    analyses    establishing    the    floristic    profiles    of    these    associations    and    the    statistical    relationships    between    ecological variables and vegetation associations: the flood parameters appear as the most determinant variables (Hiernaux et al . 1980). The   privileged   role   of   the   flood   is   confirmed   by   the   overall   scheme   of   ecological   relationships,   which   results   from   a   factorial analysis   of   the   matrix   of   the   state   of   environment   variables   and   the   presence-absence   of   vegetation   species.   Indeed,   whether   the analysis   focuses   on   all   169   survey   sites   or   on   a   selection   of   the   127   sites   within   the   proper   flood   plains,   the   structure   of   "matrix clouds"   is   characterized   by   the   strong   hierarchy   of   the   first   axes,   the   main   beams   of   the   architecture   of   matrix   clouds   which constitute a synthetic representation of the interrelations – in terms of presence/absence – between taxa and states of the variable. In   addition   to   the   participation   (expressed   in   %)   of   the   first   five   axes   in   the   inertia   of   the   matrix,   the   following   table   shows   the contribution of each of the variables to this participation, axis by axis, and overall for the five axes. Table 1: Features of the factor analysis of species / status variables for 169 readings of the Inner Niger Delta In   the   analysis   of   the   169   survey   sites,   the   first   two   axes   are   closely   linked   to   the   flood   conditions,   which   total   71.6%   and   73.1% of   the   inertia   explained   by   the   first   two   axes.   However,   a   more   detailed   observation   shows   that   on   the   first   axis,   46.3%   of   the   inertia is   explained   by   the   ‘no   flooding’   situation.   In   other   words,   the   first   axis   opposes   the   non-floodable   sites   of   the   Delta   to   all   of   the flood   plain   sites,   the   second   axis   distributes   the   latter   over   a   gradient   of   height,   regularity   and   duration   of   flooding.   (P.   Hiernaux,   The fodder resources map of the routes of the Inner Niger Delta - Notice ,  Bamako: CIPEA-ODEM, 1980, p.23) One   of   the   main   factors   in   the   flood-vegetation   relationship   is   the   flood   height,   which   makes   it   possible   to   classify   the   different vegetation associations by level, each level representing the most frequent flood height for a plant association. Table 9 :  Plant associations  according to the levels or heights of submersion At   level   7,   BP   and   PAK   are   the   two   plant   formations   that   support   –   or   require   –   the   strongest   submersions.   The   first   is   a   low " bourgoutière"  with Vossia cuspidata , the second corresponds to a very deep grove with Acacia kirkii . Level   6   is   occupied   by   B   a   " bourgoutière"    with   Echinochloa   stagnina ,   VB   a   very   low   " vetiveraie ",   OP   a   deep   "orizaie"    and PAM a formation with Mitragina inernis . Level   5   is   occupied   by   an   "eragrostaie"       EOR,      an   "orizaie"   O   and   a   "vétivéraie"   VOR      as   well   as   rice   paddies   from   the   Office du Niger, downstream from Ké Macina. Level   4   is   occupied   by   VSP   and   ESP,   the   "vetiveraies"    and   medium-sized   "eragrostaies"    as   well   as   AC   an   "eragrostaie"    with Andropogon canaliculatus  . Level   3,   which   corresponds   to   a   submersion   between   10   cm   and   30   cm,   is   occupied   by   a   "panicaie"    P      and   a   high   vétiveraie   VH. Level   2   –   the   last   level   of   submersion   –   between   0   and   10   cm,   is   occupied   by AG,   a   savanna   with   Andropogon   gayanus    and   a complex formation called ZB (for zone beaten by maximum floods). Finally   –level   1   –   the   formations   ranging   from   TA   to   TT   are   all   located   on   the   " togge ",   the   Fulani   name   for   exposed   mounds   in the Delta and on the dry edges. Nb   -   PAN,   PAS,   PAR   represent   plant   formations   located   on   plains   where   flood   is   deferred.   The   most   remarkable   characteristic of   these   acacia-based   formations   is   the   bimodal   dimension   of   the   flood:   the   first   mode   is   linked   to   the   run-off   of   the   rains   from   July- August,   and   the   second   to   the   late   arrival   of   the   flood   in   October   or   November.   Finally,   MB,   the   riverbank   mosaic,   is   not   shown   in the   table.   It   constitutes   an   alternation   of   channels   and   rims   of   banks   or   levees,   well-represented   in   the   major   bed   of   the   Niger,   the Bani   or   the   large   tributaries.   It   always   represents   a   strong   gradient   of   submersion   going   from   level   6   to   level   2,   juxtaposed   over   short distances (a few tens of meters). The definition of a "maximum reference flood" Pierre   Hiernaux   relates   the   different   states   of   the   variables   describing   the   conditions   of   the   flood   (its   height,   regularity, duration,   speed   and   the   dates   of   the   rise   and   fall   of   water…)   with   the   data   available   in   the   Delta   in   1980.   The   question   is   not   to determine   the   regimes   of   each   plain   or   of   each   basin   taken   in   isolation,   but   to   define   a   series   of   parameters   which   are   most frequently   reached,   so   as   to   make   it   possible   to   establish   a   relationship   between   plant   formations   and   the   course   of   a   flood,   which,   by analogy   to   the   relation   existing   between   the   climate   and   plant   formations,   can   be   considered   as   the   "maximum   reference   flood". Pierre   Hiernaux   calculated   the   submersion   depths   normally   attached   to   the   flooded   formations.   These   submergence   depths   are measured from a 0 cm point of reference, which corresponds to the level most regularly reached on the gauge. The   comparison   of   the   series   leads   us   to   consider   that   the   value   which   can   be   considered   as   the   reference   value   corresponds to   the   mean   maximum   of   floods   minus   the   standard   deviation.   P.A.   Gosseye   in   (S.   Cissé   and   P.A.   Gosseye,   Competition   for   limited resources:   the   case   of   the   fifth   region   of   Mali.   Report   1:   Natural   resources   and   population.    CABO,   Wageningen,   Netherlands   - ESPR,   Mopti,   Mali.   1990,   106   p   +   appendices)   discusses   at   length   (pp.   48   and   49)   this   notion   of   a   reference   flood.   In   agreement with   Pierre   Hiernaux   and   with   our   own   previous   work   (J.   Marie,   Search   for   a   solution   to   the   problems   of   breeding   in   the   interior delta   of   Niger   in   Mali,    vol.5,   summary   report.,   Addis   Ababa   ,   CIPEA   /   ODEM,   1983,   151   p.   (1983).   He   admits,   that   for   the   Mopti station,   the   series   of   floods   that   can   be   described   as   normal   and   regular   corresponds   to   the   years   1944-1968,   and   that   the   average   ten- day   maximum   for   this   period   (686   cm),   minus   the   standard   deviation   (26   cm),   is   the   value   most   frequently   reached   or   exceeded. This   value   therefore   corresponds   to   the   660   cm   level   on   the   Mopti      gauge   (reached   or   exceeded   in   84%   of   cases)   and   establishes,   for the station of Mopti, a relation with the 0 cm reference for submergence heights. My   own   calculations   relating   to   daily   values   ​​(and   no   longer   ten-day   ones)   give   very   similar   results:   average   1943-1968:   693   cm; standard   deviation   28   cm,   i.e.   a   reference   value   of   665   cm   reached   in   82%   of   the   cases. We   will   ultimately   retain   the   value   of   660 cm   for   the   reference   station   of   Mopti,   a   value   which   we   assume   to   be   representative   of   normal   or   regular   floods,   and   which   we therefore   assimilate   to   the   reference   0   cm   of   the   height   of   submersion   of   the   various   flooded   plant   formations,   which   corresponds to the altitude of 267.20 m. This   notion   of   "maximum   reference   flood",   which   suggests   an   equilibrium   relationship   between   floods,   varying   from   year to   year,   and   plant   formations   can   be   discussed.   Significant   changes   in   floods   should   translate   into   changes   in   plant   formations. The   field   work   continued   until   1985   (with   a   series   of   very   low   floods:   551   cm   in   1982,   502   cm   in   1983,   440   cm   in   1984)   shows very   large   variations   in   forage   production,   and   limited   modifications   in   the   floristic   composition   of   certain   formations,   but   does not   call   into   question   the   staggering   of   the   vegetation   composed   of   perennial   grasses   and   the   general   pattern   of   the   levels   of submersion and their relationship with a reference flood. The   conclusions   of   the   return   to   the   field   of   Pierre   Hiernaux   and   Mathew   Turner   in   2014   leading   to   new   vegetation   surveys confirm this very great overall stability in the distribution of plant formations.(see page 43) Table 3 : Relationship between plant associations, submergence heights and flood at the Mopti gauge