Figure 9.3. The biomass of phytoplankton in the mid-Hudson region. Shown in A are weekly to bi-weekly data for chlorophyll-a near Kingston, New York (river km 144). In B we show means for the May-October period for each year (hatched bars, errors are SD). The solid line (right-hand Y-axis) shows the water filtration rate of the zebra mussel (Caraco et al., 1997; Strayer, this volume). The filtration rate is expressed as volume per area per time (m3 m-2 d-1), which is the same as md-1.

concentration of chlorophyll-anear Kingston, New York (rkm 144-147) reveal several key features at several time scales about the magnitude and variation of phytoplankton in the Hudson (Fig. 9.3A).

First, there is an obvious seasonal cycle with peak biomass generally occurring in late spring. While the peak values can be quite high (20 to 50 |g liter-1), the average level is moderate or low compared to other rivers and estuaries (discussed below). In many estuaries and lakes rapid phytoplankton growth occurs early in the season leading to a "spring bloom" in February to April. In the Hudson River, the bloom is substantially delayed and rarely, if ever, occurs in the spring. Among all years the mean day of peak phytoplankton biomass would be August 14. The earliest peak we have observed was in 1999 (May 12) and the latest in 2000 (October 25). In most years the peak occurs in mid July (Fig. 9.3A). There is high variance among years of the timing of the rapid growth phase. The day-of-year of peak chlorophyll-a is negatively correlated to the average amount of suspended load in the river (r2 = 0.39; p = 0.01). Suspended load is the major factor controlling light extinction in the river. This correlation, however, explains only a fraction of the variance in the timing of the peak, so other factors are clearly involved.

Second, there are very obvious interannual differences in the magnitude of chlorophyll-« in the Hudson (Fig. 9.3B). The largest is the change from moderately high values to low values before and after 1992, the year the zebra mussel first became established at high numbers in the river (Caraco et al., 1997; Strayer et al., 1999; Strayer, this volume). Prior to 1992 mean growing season (May-October), chlorophyll-« at Kingston averaged 22.1 ± 5.9 |g liter-1. From 1993-2000 the mean was 4.4 = 1.2 |g liter-1. This 80 percent decline in phytoplankton biomass is consistent with the dramatic increase in water filtration brought about by the zebra mussel (Caraco et al., 1997). Prior to the zebra mussel invasion biological filtration of the tidal, freshwater Hudson occurred about once in 50 d, and was largely the result of suspension feeding by cladocerans such as Bosmina and copepods (Caraco et al., 1997; Strayeretal., 1999;Pace and Lonsdale, this volume). The zebra mussel increased biological filtration so that the entire water column turnover time was as short as 1 to 3 d depending on the year (Fig. 9.3B).

As in San Francisco Bay, some of the interannual differences in the Hudson's phytoplankton biomass are related to variation in freshwater discharge among years, which controls the advec-tive loss of phytoplankton (Cloern et al., 1985). Discharge during the growing season varies about three-fold in the Hudson from 100 to nearly 400 m3 s-1 among years. Thus, the residence time of water within the tidal-freshwater region varies with this discharge from about 100 to 25 days. Discharge was negatively correlated to chlorophyll-a during the pre-zebra mussel period (Fig. 9.4; p = 0.002). During the post-zebra mussel period, this relationship has the same trend, but is not significant (Fig. 9.4; p = 0.19). Prior to the zebra mussel invasion, the advective loss term was of comparable magnitude to biological filtration; following the invasionbiologicalfiltrationgreatly exceeds the advective loss term.

0 0

Post a comment