Some of this ambiguity was evident even prior to restoration. Prior to restoration, most of the water features in the diked dairy ranch were drainage ditches, managed ponds, and small creeks that flowed through the pastures, existing via tidegates or culverts into Lagunitas Creek or portions of Tomales Bay. Before the levees were breached, zooplankton communities or species assemblages did appear to differ between the muted tidal or non-tidal pasturelands and natural marshes, although total numbers or densities of zooplankton were actually relatively similar. This similarity in zooplankton numbers may result from high spatial and temporal variability in samples, which weakens the power of statistical tests. In general, average numbers of zooplankton--which is influenced by extremely small or large samples--appeared higher in Reference Areas (133,742 indiv/m3) than in the dairy (54,222 indiv/m3), while median numbers--the centerpoint of invertebrate densities--actually appeared higher in the Project Area (37,460 indiv/m3) than Reference Areas (19,240 indiv/m3). The number of zooplankton species did not differ statistically between the dairy ranch (8.82) and natural marshes (9.81), but the number of species and the evenness of their distribution--distribution of individuals across species number--did differ, with this Shannon-Weiner Diversity Index being higher in Reference Areas (H=1.44) than in the Project Area (H=1.19; Figure 1 [109 KB PDF]).
While differences were certainly expected between the dairy ranch and natural marshes in species composition, interestingly, there were no strong statistical differences between these areas, although upstream/freshwater areas not surprisingly supported distinct zooplankton and pelagic invertebrate communities. Other statistical techniques such as ordination suggested some separation between the dairy ranch and Reference Areas, however. Reference Areas or natural marshes supported higher numbers of Harpacticoids, Cumaceans, and Nematodes than either the Project Area or dairy ranch or upstream areas. Many of these species live in soils and forage along channel bottoms, feeding on bacteria, micro-organisms, and detritus or undecomposed plant material. There also appeared to be higher numbers of Calanoids, and the calanoid copepod, Eurytemora affinis, in Reference or natural marsh areas. Conversely, insects appeared to be potentially more abundant in the Project Area. Many of these species live in soils and forage near the soil-water interface, feeding on bacteria, micro-organisms, and detritus or undecomposed plant material (Anne Slaughter, SFSU, pers. comm.). Eurytemora, one of the most common species in the Low Salinity Zone of San Francisco Bay during the 1980s prior to large-scale invasion by non-native invertebrates, forages on phytoplankton and some detritus and is considered an excellent food source for fish (A. Slaughter, SFSU, pers. comm.).
With restoration, species composition did appear to shift, although numbers or densities of organisms again remained similar between the different restoration phases. While average densities appeared higher during both Passive (91,575 indiv/m3) and Full (85,856 indiv/m3) Restoration than Pre-Restoration (54,222 indiv/m3), these differences were not significant from a statistical standpoint, perhaps due to high variability in the data. In addition, median densities actually appeared higher during the Pre-Restoration period (37,460 indiv/m3) than in either Passive (11,949 indiv/m3) or Full (16,609 indiv/m3), although, again, these differences were not statistically significant. The number of zooplankton species in the Project Area did appear to increase with restoration: species number did not change much between Pre-Restoration (8.82) and Passive Restoration (7.78) periods, however, it jumped during Full Restoration (11.59). In addition, species diversity in terms of number and evenness also climbed between Pre-Restoration (H=1.19) and Full Restoration (H=1.47) periods.
Restoration substantially affected species composition or assemblages, with substantial changes in the type of organisms present during every treatment period, although the changes are not always consistently interpreted by different statistical approaches (Figure 2 [30 KB PDF]). Based on results from statistical ordination, certain species did appear more associated with particular treatment periods, including Eurytemora affinis with Pre-Restoration; Rotifera, Calanoida, Cyclopoida, Bosmina (Cladoceran), Brachionus sp. (Rotifera), and Diptera with Passive Restoration; and Nematoda, Oligochaeta, Annelida, Harpacticoids, Tisbe furcata (Harpacticoid), Microsetella norvegica (Harpacticoid), and Corophrium sp. (Amphipoda) with Full Restoration (Figure 2). Eurytemora, one of the most common species in the Low Salinity Zone of San Francisco Bay during the 1980s prior to large-scale by non-native invertebrates, forages on phytoplankton and some detritus and makes good food for fish. This and other species common during Passive Restoration are mostly freshwater species that are sometimes found in brackish environments (A. Slaughter, SFSU, pers. comm.). In contrast, species associated with Full Restoration such as Tisbe, Microsetella, and Corophrium are typically associated with brackish to saline environments, although water pH appeared to affect species composition more than salinity in statistical analyses (A. Slaughter, SFSU, pers. comm.).
Based on species composition, restoration appears to have introduced a disturbance factor that has led to dominance by hardy, disturbance-tolerant taxa that tend to feed at lower, microbial trophic levels, while the Pre-Restoration Project Area actually supported a diversity of zooplankton and pelagic invertebrate species that fed at different trophic levels (A. Slaughter, SFSU, pers. comm.). Large die-offs of pasture vegetation once tides were reintroduced generated massive volumes of organic matter and detritus that could support large numbers of lower-trophic level zooplankton and invertebrate taxa.
While zooplankton communities of the former dairy ranch have changed considerably since dairy operation with restoration, the zooplankton community in the restored dairy is still different than that of natural or reference marshes, although some statistical techniques suggest that convergence might be occurring even after only two years (Figure 3 [31 KB PDF]). Some compositional differences between the Project Area and Reference Areas still appeared to exist. The restored Project Area appeared to have higher numbers of Diaptomus sp. (Calanoida) and Gammaridae (Amphipoda), while Reference Areas actually supported a higher abundance of some pollution tolerant invertebrate taxa such as Diptera, Insects, and Oligochaetes. Diaptomus is typically classified as a freshwater organism, while Gammaridae occur in waters with a wide range of salinities (A. Slaughter, SFSU, pers. comm.).
As with pre-restoration, it was difficult to draw any conclusions about whether zooplankton and pelagic invertebrate numbers or densities differed between the restored dairy and natural marshes following restoration due probably, again, to high variability in the data. Median densities actually appeared higher in the Project Area (17,169 indiv/m3) than in reference marshes (9,151 indiv/m3), however, average densities were pretty similar between the two Study Areas (23,677 and 24,044 indiv/m3) for the Project Area and Reference Areas, respectively. The number of zooplankton species remained similar between the restored dairy (11.59) and natural marshes (10.26) after restoration, but, unlike pre-restoration, species diversity and evenness was also similar between the Project Area (H=1.47) and the Reference Areas (H=1.40; Figure 1 [109 KB PDF]).
In summary, then, zooplankton and pelagic invertebrate communities definitely changed with both Passive and Full Restoration of the Project Area. However, how different Project Area assemblages were from natural marshes even prior to restoration is more difficult to ascertain, although some exploratory statistical analyses did suggest some differences. Two years after restoration, zooplankton communities appear to show some convergence between the former dairy and natural marshes, although species assemblages still differ to some degree. From a statistical standpoint, numbers or densities of zooplankton did not appear to differ between Study Areas either before or after restoration or between different periods in the Project Area, however, large variability in numbers or densities between sampling sites may have obscured the seeming differences in densities. Certainly, the number of species between the restored dairy and natural marshes remains very similar, although the species diversity--or number and distribution of species--has improved within the Project Area after restoration such that it is now equivalent to that of Reference Areas.
Untangling the Food Web: Changes in Prey Base Following Restoration
Last updated: February 28, 2015