Sediment remobilisation from decomposing cordgrass ( Spartina anglica ) patches on a wave‐exposed intertidal flat

Abstract

Concerns about the potential adverse effects of Spartina invasion on New Zealand estuaries, such as habitat loss and degradation, have led to herbicide use to control its spread. An experiment was conducted in the Manukau Harbour (New Zealand) to determine sediment remobilisation rates (? E ) from herbicide‐treated S. anglica patches. Paired treatment and control patches (≤25 m diam.) were monitored at two sites 500 m apart with different wave exposures. Above‐ (AGB) and below‐ground (BGB) biomass decomposition and AE were determined from repeat surveys over c. 2.5 years. Work done by waves ( W ) , which integrates wave‐energy dissipation at the bed over time, was used as a surrogate measure of sediment transport potential. Complete AGB loss occurred in both treated patches within 3 months of initial herbicide application, whereas BGB (a matrix of roots and rhizomes) persisted much longer. An exponential decay model ( r 2 = 0.6) fitted to the site‐two treatment data predicted an c. 85% reduction in BGB after 5 years, whereas the BGB changes for the larger and more heterogeneous site‐one treatment were more complex. BGB was a better predictor of average bed elevation than AGB in multiple regression models and, although bed elevation decreased through time and with waves, BGB was positively related to bed elevation at both sites. AE was similar for both treatment patches (c. 0.05 and c. 0.06 m yr ‐1 , P 0.001, r 2 ≥ 0.97) and complete sediment loss was predicted to occur within 6–10 years. The site‐one control patch accumulated sediment (0.01 m yr ‐1 , P 0.001, r 2 = 0.81) whereas the more wave‐exposed site‐two control eroded (0.01 m yr ‐1 , P = 0.0017, r 2 = 0.61). The potential for adverse sediment effects resulting from Spartina control at wave‐exposed sites will be minimal if the treated areas are small relative to the size of the estuary.

Bibliographic Information