Coupling of boreal forests and lakes: Effects of conifer pollen on littoral communities

Conifer pollen deposition is a visually striking spring event in boreal lakes, representing a potentially major allochthonous input of limiting nutrients. We conducted a lake survey and mesocosm experiment at the Experimental Lakes Area in northwestern Ontario to test the hypothesis that jack pine (Pinus banksiana) pollen inputs subsidize littoral nutrient levels and stimulate algal growth and zooplankton abundance. A series of floating litterfall collectors were deployed along transects that span a 0.27-km2 headwater lake (Lake 373) and monitored after ice-out to quantify pollen deposition over a 45-d period. Lake 373 (L373) received 11–56 mmol P m–2 d–1 from pollen, or an annual total of about 10 kg of P. These data were used to determine pollen amendment levels (ambient, 33, 103) for an experiment involving 18, 1-m3 littoral mesocosms distributed over three lakes (L239, L373, and L442). Pollen amendments significantly increased total phytoplankton and herbivorous zooplankton biomass, resulting in greater abundance of inedible filamentous green algae and large diatoms. Pollen also exerted a positive lake-specific effect. Periphyton biomass also increased in response to pollen additions, especially filamentous green algae and diatoms in L239 and L442. Conifer pollen subsidizes nutrient levels and promotes production in small boreal lakes. Terrestrial environments are hypothesized to strongly affect most lakes because they are relatively small ecosystems (,1 km2 surface area, ,10 m mean depth; Wetzel 1990). However, despite being considered an important future direction of limnology, land–water linkages within lakes remain poorly understood (Schindler and Scheuerell 2002). In particular, allochthonous inputs (e.g., pollen, insects, leaves, soil, and wood) from relatively productive terrestrial environments are expected to heavily subsidize small oligotrophic boreal lakes (Wehr et al. 1998; Doskey and Talbot 2000; Pace et al. 2004). For example, large amounts of wind-dispersed, nutrientrich conifer pollen are deposited annually onto small northern lakes during late May to early June (Lee et al. 1996; Doskey and Talbot 2000). Yet, pollen is not considered as a major allochthonous input because most studies that have examined the importance of litterfall to lakes were performed either after the period of maximal pollen deposition (e.g., Cole et al. 1990) or in lakes that have small catchment area : volume ratios (e.g., Lake Tahoe; Richerson et al. 1970). In contrast, Lee and Booth (2003) suggested that the rapid decomposition and enzymatic release of nutrients from conifer pollen grains contribute substantial amounts of nutrients to whole boreal catchments. The productive capacity of small oligotrophic boreal lakes is expectedly enhanced by inputs of conifer pollen because ,60% of its total phosphorus is released in a soluble reactive form (Doskey and Ugoagwu 1989). In addition, pollen can also be an important source of organic carbon to small unproductive lakes (Doskey and Talbot 2000). Although pollen grains contribute substantial amounts of sedimentary organic carbon to lakes because of the recalcitrant nature of their exines (Doskey and Ugoagwu 1989; Doskey and Talbot 2000), they might also supply dissolved organic carbon (DOC) to lakes because of their labile organic content and external polysaccharide layer. Therefore, pollen-derived phosphorus and organic carbon might subsidize boreal lakes that are net heterotrophic systems, relying heavily on allochthonous carbon inputs (del Giorgio et al. 1999). The main goals of our study were to quantify conifer pollen deposition in boreal lakes and to determine its effect on littoral communities as it accumulates along sheltered shorelines. Unlike previous survey studies (Doskey and Ugoagwu 1989; Doskey and Talbot 2000), we experimentally tested for the response of extant biological communities to pollen. Survey data from a headwater lake were used to quantify pollen deposition and establish pollen amendment levels for a mesocosm experiment. We hypothesized that P-rich pollen inputs enhance benthic and planktonic algal abundance, thereby increasing the potential production of herbivorous zooplankton and, in particular, P-limited cladocerans (Brett et al. 2000). Acknowledgments Tyler Cobb and Angela Strecker provided field assistance. This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC), Canadian Foundation for Innovation, and Environment Canada Science Horizons Program to R.D.V., and a NSERC PGS-D scholarship to M.D.G. Additional financial support for this project was obtained by M.D.G. through governmental and private contracts for phycological analyses. Fisheries and Oceans, Canada, provided support to M.T. and to the Experimental Lakes Area where this study took place. Limnol. Oceanogr., 51(3), 2006, 1524–1529 E 2006, by the American Society of Limnology and Oceanography, Inc.