Figure 1.
Association between Nepenthes bicalcarata pitcher plants, Camponotus schmitzi ants, and fly larvae that develop in the pitchers.
A. Pair of N. bicalcarata upper pitchers. B. C. schmitzi workers retrieve a drowned cockroach from the fluid inside a pitcher. C. examples of the rich dipteran infauna of N. bicalcarata; from left to right; top row: Toxorhynchites sp., Tripteroides sp., Culex morphospecies 1, Culex morphospecies 2, Uranotaenia sp.; bottom row: Wilhelmina nepenthicola, large puparium (cf. Phoridae) hanging under the pitcher rim, large culicid pupa at fluid surface (other pupae, culicid larvae and Polypedilum (Chironomidae) larvae living in protective cases are also visible).
Figure 2.
Natural δ15N for the components of the investigated food-web.
Included are non-carnivorous plants (“non-CPs”, n = 7), N. bicalcarata plants (highlighted in grey; climbers without ants, n = 15; climbers with ants, n = 17; rosettes without ants, n = 15; rosettes with ants, n = 11), prey insects (n = 6), pitcher detritus (n = 10; one outlier with δ15N = 11.42 ‰ not shown), ‘other’ dipteran larvae (n = 10), ‘predatory’ dipteran larvae (n = 5) and C. schmitzi ants (n = 5). The dotted horizontal line highlights the level of prey insects. Boxplot shows medians, interquartile ranges, and the largest and smallest values that are not outliers (outliers shown as circles).
Figure 3.
Flux of nitrogen from ant colony to pitcher plant.
Bars indicate the change in 15N abundance in the leaves of a N. bicalcarata plant two weeks after feeding a 15N pulse to the symbiotic C. schmitzi colony. Leaf node 1 bears the youngest leaf. The pictogrammes under the graph explain the structure of the host plant, its ant colony and mark where tracer was fed.
Table 1.
Adult infauna emerged from N. bicalcarata pitchers. Unless otherwise stated, values are total numbers of individuals.
Figure 4.
Effect of C.schmitzi presence on numbers of surviving Aedes sp. pupae and successfully emerging mosquitoes.
The experiment started with 20 living pupae in each pitcher. Boxplot shows medians, interquartile ranges, and the largest and smallest values that are not outliers (outliers shown as circles).
Table 2.
Samples taken for the analysis of the natural abundance of 15N in N. bicalcarata phytotelm food webs.
Figure 5.
Schematic of “emergence trap” for pitcher plant infauna.
The lower bottle containing the pitcher is completely darkened, causing emerging dipterans to move upwards towards the light shining through the neck. Tanglefoot® prevents crawling insects from entering the collection bottle. The fish trap-like design prevents escape from the collection bottle. Emerging insects reaching the collection bottle were killed when falling into a 1.5% CuSO4 solution. Traps were stabilised with a pole. Whole, living pitchers could be used without significant interference. Although C. schmitzi could freely pass through a small gap in the bottle along the pitcher tendril, the trap was mosquito-tight (without ants, all the 20 inserted individuals were always recovered; n = 10 traps).