Heterotrophic planktonic bacteria represent a reasonably large proportion of the living particu-late organic matter in the Hudson River and so might be a significant food resource for some consumers. The debate over whether microbes are "links" versus "sinks" for organic carbon depends to a large extent on bacterial growth efficiencies but will also be a function of the particle-harvesting abilities of the microconsumers in an ecosystem. In the Hudson, as in many other aquatic systems, smallheterotrophic flagellates are the predominant consumers of free-living bacteria (Vaque et al., 1992) and these flagellates are themselves potential prey for larger zooplankton (see Chapter 16). Given the rapid turnover times for bacterial biomass there must be large consumption or other losses, otherwise the cell accumulations would be much greater than the ~50 percent seasonal changes in abundance actually observed. While zebra mussels have proven capable of filtering a large proportion of the river volume per day (Strayer et al., 1999), they do not capture natural bacterial cells efficiently and in fact the abundance of bacteria has increased postzebra mussel (Findlay et al., 1998a) particularly in the upriver stations where zebra mussels are most numerous. These observations together with experiments designed to examine zebra mussel clearance of various grazers suggests zebra mussels have been released from flagellate control because zebra mussels can very effectively clear natural flagellates from the Hudson River's water column (Findlay et al., 1998a). This change implies that HR planktonic bacteria maybe under less grazer control currently than pre- zebra mussel and perhaps their contribution to higher trophic levels has declined since the early 1990s.
Carbon inputs to the Hudson are overwhelmingly dominated by loads of dissolved organic carbon (DOC) and particulate organic carbon (POC) from the catchment. Mean annual inputs are on the order of 600 gC m-2 y-1 and this input is primarily from the upper Hudson drainage basin at a ratio of roughly 2/3 DOC and 1/3 POC (Howarth, Schneider, and Swaney, 1996). Tidal marshes (~4500 ha for entire river) are highly productive with NPP values commonly 1 or more kg carbon/m2/yr and although the net export to the mainstem is uncertain, outwelling of particulate and dissolved organic carbon was estimated as 16 gC/m2/yr (Howarth et al., 1996). Autochthonous carbon inputs include phytoplankton and submersed vegetation, which together currently make up about 20 gC m-2 y-1 (see Chapter 9).
Linking bacteria to potential carbon sources can be examined via correlational analyses and experimental manipulations. In the past, correlations have shown weak associations between bacterial abundance or growth and chlorophyll a (Chl a). Although some of the relationships using the full data set are statistically significant, they account for a small proportion of the variation in bacterial variables. For example, there is a positive association between bacterial production (BP) and Chl a (Fig. 8.5; p < 0.05; r = 0.34) but this might be covariation with temperature rather than evidence for phytoplankton as an important carbon source for planktonic bacterial production. There is no correlation between bacterial abundance and Chl a (p > 0.05; r = -0.06). Considering the nonliving carbon pools, there was no association between bulk DOC (the largest component) and bacterial growth (p > 0.05; r = 0.07). There was a significant positive relationship between DOC and bacterial abundance (p > 0.05; r = 0.37) but this was probably due to temporal covariation, as both cell density and DOC increase seasonally, which could generate a positive association between the
Figure 8.5. Correlation between bacterial growth and planktonic chlorophyll a. The relationship is significant (p < 0.0001) but only explains 32% of the variance in growth rate.
Was this article helpful?