A series of time-specific food webs for the macroinvertebrate riffle community of Duffin Creek, Ontario was constructed using dietary information obtained from the analysis of gut contents. Trophic links were quantified using a dietary index of relative importance. Precision of the analysis was maintained at a high level by: (1) identifying dietary items as accurately as possible via direct gut analysis; (2) identifying web members to the species level, thus avoiding the taxonomic aggregation and lumping of size classes common in food web analyses; and (3) ensuring temporal resolution of the web by determining ontogenic variation in the diets of dominant members of the community.The Duffin Creek webs are heavily detritus-based with a large proportion of top-to-basal, and intermediate-to-basal links. Top-to-basal links, proportions of top and intermediate species, and lower connectance (0.180-0.219) varied temporally. Trophic connectance ranged from 0.090 to 0.109, consistent with values expected for a web consisting largely of specialist feeders. Weak links made up the largest proportion of total links in the webs, whereas very strong links made up the smallest proportion. Omnivory was more common than indicated in other webs and can be attributed to ontogenic diet switching.Comparison of the statistics for a summary web with those generated for the time-specific webs indicated that the total number of links per web, total number of species, number of top and intermediate species, and linkage density were much greater for the summary web. In view of these differences, the importance of temporal resolution when assessing food web structure and dynamics is emphasized. The possibility that some of the observed features in our web are common to other detritus-based webs is considered. Future studies of this calibre are justified.

The relationship between the biology and habitat of larval mosquitoes was studied in a series of artificial ponds of differing characteristics established across a terrestrial vegetation gradient from open fields to deep woods. The ponds were uniformly colonized by two widespread species of Culex, both characteristic of small bodies of water, including artificial, "container" habitats. First-instar larvae of Culex restuans were found within 2 days of filling the ponds with water and four or five generations were produced from May until the end of September. One very long generation occurred in July, which corresponded to maximum larval densities. Overall, there was a strong relationship between larval development time of C. restuans and larval density. A few larvae of Culex pipiens pipiens occurred sporadically throughout the summer, but numbers did not increase until C. restuans populations began to wane in late July. Thereafter two, or possibly three, generations were produced into the autumn. The patterns of colonization, synchrony of life history, and growth of these two species were remarkably consistent amongst the ponds, despite considerable variation in both their physical and biological environments (e.g., over the 2 years of study, conductivity ranged from 20 to 890 μS, pH from 6.4 to 10.7, dissolved oxygen from 0 to 13.5 ppm, and water temperature from 5 to 29.8 °C). Density of C. restuans was related to water temperature and pH at the "open" site and to water temperature, dissolved oxygen, percent algae, and percent detritus at the "edge of the woods" site. At the "deep woods" site, larval numbers were related to temperature, pH, conductivity, and dissolved oxygen. At both the edge and deep wood sites, larval development time of C. restuans increased with mean water temperature to 20 °C. Above this temperature, larval development time tended to decrease.

WOOD, C.M. 1994. Colonization dynamics of algae in small artificial ponds. Can. J. Bot. 72: 1654-1665. The arrival, establishment, and growth of a variety of algae were studied in a series of newly established, small (48.5 L) artificial ponds located at three sites of differing characteristics in southern Ontario, Canada: open field, edge of woods, and deep woods. Algal colonization was most rapid at the open site, within 4 days of the ponds being filled, and was slowest at the deep woods site, presumably because of a filtering effect of the surrounding terrestrial vegetation. The patterns of algal colonization and subsequent population abundances were consistent among the ponds, despite very considerable variation in both their physical and biological environments (e.g., among ponds over the 2 years of study, conductivity ranged from 20 to 890 pS, pH from 6.4 to 10.7, dissolved oxygen from 0 to 13.5 mg . L-', and water temperature from 5 to 29.8"C). The first colonizers were airborne microchlorophytes, particularly a species of Chlorella, which may have been favoured by clear water (high light) conditions, lack of competition, and an ability to tolerate low nutrient levels. Chlorella populations declined within 1 or 2 weeks, depending on site, coincident with an increase in the numbers of a species of Synechococcus, a cyanobacterium. Synechococcus dominated all the pond types at all three sites thereafter, although its abundance was strongly correlated with water temperatures. In their 2nd year, undisturbed mature ponds developed encrusting algae, especially Rhizocloniutn and Oedogonium, on their walls. In years 1 and 2, the densities of Synechococcus in both the new and mature ponds, respectively, were positively correlated with the larval densities of the dominant mosquito species Culex restuans. Synechococcus was the primary item in the diet of these insects.

No abstract