COASTAL SALT MARSHES: HYDROLOGY AND SALINITY

the ephemerally connected peatland; at certain times of year, usually during spring snowmelt, a peatland can receive water either as surface or subsurface flow from adjacent hydrological components of a drainage basin. Once the exchange ceases, discharge from the peatland recedes. The spring recession in flow extended over a month in a small southern Ontario cedar swamp seasonally connected to a perched aquifer (Taylor and Pierson 1985), but lasted less than two weeks in a low arctic fen which received water from the overflow of adjacent lakes and streams (Roulet and Woo 1988). An isolated peatland depends on direct precipitation for its water input, and discharge occurs in response to snowmelt and rainfall. A peatland that is effective in sustaining flow i s usually connected to a larger hydrologic system. In some peatlands, discharge can be very large. For example, the mean annual daily runoff from a southern Ontario headwater swamp was46 mm d-’, but the peatland itself had little effect on the magnitude of the flux, which was controlled, rather, by groundwater discharge through the peatland (Roulet 1988). The magnitude and timing of storm runoff from a peatland is a function of the size of the storm input, the antecedent storage capacity of the peatland, and the ability of the peatland to shed water. If a peatland has little available storage capacity, the storm response will be large, whether or not the peatland is hydrologically isolated or connected. An understanding of controls on antecedent storage capacity is important in explaining response to storms. Isolated peatlands have a high degree of variability in response because the antecedent storage capacity depends on the frequency and magnitude of rainfall, which i s itself highly variable. Peak stormflow from a Minnesota peatland was three times larger when the water table was near the peat surface than when it was 15 cm below (Bay 1969). In ephemerally isolated peatlands, response to storms is large when the peatland is still connected to additional sources of water (Taylor and Pierson 1985), but is similar to that of the isolated peatland once the connection is broken. Storm run-off was produced for all rain events in a headwater cedar swamp because continuous groundwater input between storms maintained areas of saturation (Roulet 1988). Groundwater input varied little, and the proportion of rain discharged as storm flow was reasonably constant (e.g., approximately 31 per cent of direct precipitation onto the wetland) (author’s unpublished data). Current runoff theory explains the characteristics of stormflow hydrographs by relating the magnitude and rate of change in storage to the mode of transport of water through a hydrological system. The production of runoff from peatlands should be treated in the same way (e.g., Verry, Brooks, and Barten 1988) to improve the understanding of the hydrological role of peat-covered wetlands.