. The Canadian field-naturalist. 0 ,1^—E«„, Jii«sa^,^-» Opio. Rubus Strept Opiopanax. OpIo. Rubus Strept. Streptopus 100 80 60 40 20 Oplo. Rubus Strept. Figure 1. Mean percentage of seed removal, for each of the three species, observed in each treatment (clean seeds vs. seeds in pseudofeces) and in each microsite (under Opiopanax, Rubus, or Streptopus plants) during the fruiting season of 1994. the other two received trays with the other species of seed. After 4-5 days, seed removal was checked in each plot (recording also the number of broken seeds present), and the experiment was repeated at
. The Canadian field-naturalist. 0 ,1^—E«„, Jii«sa^,^-» Opio. Rubus Strept Opiopanax. OpIo. Rubus Strept. Streptopus 100 80 60 40 20 Oplo. Rubus Strept. Figure 1. Mean percentage of seed removal, for each of the three species, observed in each treatment (clean seeds vs. seeds in pseudofeces) and in each microsite (under Opiopanax, Rubus, or Streptopus plants) during the fruiting season of 1994. the other two received trays with the other species of seed. After 4-5 days, seed removal was checked in each plot (recording also the number of broken seeds present), and the experiment was repeated at the next set of consecutive plots along the transect. for a total of three trials per site. A total of 15 plots (repHcates) were chosen in three localities (2M, TR and SC) whereas only 10 plots (two replicates) could be used in the other locality (HR). Experiments lasted about three weeks, from 18 July through 9 August. Because in Sheep Creek we could not locate enough Streptopus individuals in each plot, we decided to use the closest area (TR) where this species is abundant to place the trays in order to have an orthogonal experimental design. Considering all areas together, a total of 120 trays, half with clean seeds and half with pseudofeces, were placed for each species. Data were analysed by a fully crossed ANOVA with three fixed effects: species, microsite (plant species underneath which seeds were placed) and treatment (clean seeds and pseudofeces). Data from all sites were pooled after knowing that no differ- ences existed among them. The proportion of seeds removed from each tray was the dependent variable, which was normalized with the angular transforma- tion before doing the analyses. SYSTAT for Windows (1992) was the statistical package used for all analyses. Results Seeds were removed significantly more often from the dishes than from the pseudofeces (Table 1), and this was evident in all three species of seed (Figure 1). The effect of clean seeds was stro
Size: 1865px × 1340px
Photo credit: © Book Worm / Alamy / Afripics
License: Licensed
Model Released: No
Keywords: ., bookauthorottawafi, bookcentury1900, bookdecade1910, bookyear1919
