Proposal for the restoration of floodplain forests in the Middle Ebro River (NE Spain) based on greenhouse experiments and monitoring of seed dispersal

Background/Question/Methods

River regulation has altered the flow patterns of most of the rivers worldwide, thus disrupting the flow-dependent life-cycle of dominant riparian trees. The effects of dams, diversions and dikes on modifying the timing, magnitude and frequency of peak flows and altering the geomorphic dynamics have been proposed as causes of systematic regeneration failure of phreatophytes and overall floodplain forest decline. In the case of the Middle Ebro River (NE Spain), geomorphologic dynamics are highly constrained by regulation and the release of catastrophic floods (necessary for phreatophytes recruitment at the catchment scale) is not accepted by the human populations. Populus alba is an autochthonous species dominating the floodplain forests across the Mediterranean region. In the Ebro River, its sexual regeneration is considered to be very affected by flow regulation. To promote early stages of P.alba forests, we propose a restoration strategy based on locally flooding an abandoned cropland within a particular river meander (Rincón Falso meander), by means of an inflow-outflow screw gate system at both upstream and downstream old and unused dikes. The timing and rate of drawdown of the proposed flood is defined according to previous greenhouse experiments examining the survival and growth of P. alba seedlings, seed viability and longevity, and the field monitoring of the seed dispersal in 2006, 2007 and 2008.

Results/Conclusions

Survival was significantly greater in two treatments consisting in permanent saturation and 1cm·day^-1 water table decline (87 and 88%) than in a third treatment draining 2.5cm·day^-1 (58%) and two last treatments: 5cm·day^-1 and immediate drainage (25 and 22%), using a sandy substrate similar to that found in the field. The root and shoot growth were reduced both under the saturated and immediate drainage conditions compared to the declining treatments (i.e., 1, 2.5 and 5cm·day^-1). The seed dispersal period lasted 6-8 weeks and peaked in mid-April. The cumulative seed rain density averaged 31,491 seeds·m^-2, reflecting that seed production is not limiting P.alba recruitment in the study area. Initial seed germinability was high (92%) and seed longevity was long (half viability period of 30 days). Therefore, we recommend the creation of barren (nursery) sites in the Rincón Falso meander by mechanically disturbing the substrate and releasing a controlled flood in mid-April, followed by water table decline smoother than 5cm·day^-1 to enhance the initial establishment of P.alba seedlings as a pilot strategy for the restoration of the Ebro River floodplain forests.