Previous meta-analyses suggested that carnivorous plants – despite access to N, P, and K from prey – have significantly lower leaf concentrations of these nutrients than non-carnivores. Those studies, however, largely compared carnivores in nutrient-poor habitats with non-carnivores in more nutrient-rich sites, so that the differences reported might reflect habitat differences as much as differences in nutrient-capture strategy. Here we examine three carnivorous and 12 non-carnivorous plants in the same nutrient-poor bog to compare their foliar nutrient concentrations, assess their patterns of nutrient limitation using leaf NPK stoichiometry, and estimate %N derived from prey by carnivores using a mixing model for stable N isotopes. We hypothesized that (1) carnivore leaf nutrient concentrations approach or exceed those of non-carnivores in the same nutrient-poor habitat; (2) species in different functional groups show different patterns of stoichiometry and apparent nutrient limitation; and (3) non-carnivores might show evidence of employing other means of nutrient acquisition or conserv­ation to reduce nutrient limitation.

At Fallison Bog in northern Wisconsin, carnivorous plants (Drosera rotundifolia, Sarracenia purpurea, Utricularia macrorhiza) showed significantly lower leaf % C and N:P ratio, higher δ¹⁵N, and no difference from non-carnivores in leaf N, P, K, and δ¹³C. Sedges had significantly lower leaf % P, % C, and N:K ratio, and higher K:P ratio than non-sedges restricted to the Sphagnum mat, and may tap peat N via aerenchyma-facilitated peat oxidation (oxipeditrophy). Evergreen erica­ceous shrubs exhibited significantly higher levels of % C and lower values of d¹⁵N than mat non-ericads. Calla palustris – growing in the nutrient-rich moat at the bog’s upland edge – had very high values of leaf N, K, δ¹⁵N, and N:P ratio, suggesting that it may obtain nutrients from minero­trophic flows from the adjacent uplands and/or rapidly decay­ing peat. Stoichiometric analyses indicated that most species are N-limited. A mixing model applied to δ¹⁵N values for carnivores, non-carnivores, and insects produced an estimate of 50% of leaf N derived from prey for Utricularia, 42% for Sarracenia, and 41% for Drosera.

Leaf samples of the above species were gathered in situ – Utricularia and Nuphar in the bog pond, Calla and Carex pseudocyperus in the moat, and all other species in the grounded and floating mats. We collected samples from 2-6 scattered individuals per species (mean = 3.3 ± 1.1 SD) from three carnivorous and twelve non-carnivorous species (see supplementary Table S1), including almost all the common herbaceous and shrub species present. Leaf samples were transported on ice from the field to a drying oven, and then dried at 70°C for 72 hrs. Insect samples were captured via sweep and aerial netting and then oven dried. Large Odonata and other insect-eating forms were excluded to avoid complications due to the higher δ¹⁵N values expected in such predators; in most instances, these are massive and unlikely to be captured by carnivorous plants. Dried samples were ground in a Wiley mill before being sent for quantifi­ca­tion of δ¹³C, δ¹⁵N, and N, P, and K concentrations to the Nippert lab at Kansas State University.

At KSU, dried leaf samples were further ground into a fine homogenous powder, and 2 mg aliquots were placed into tin capsules. These capsules were loaded into an Elementar Pyrocube Elemental Analyzer (EA), in which the samples were combusted at 950°C. The resulting gas was analyzed for %C and %N, and then injected into an Elementar Geovision Isotope Ratio Mass Spectrometer (IRMS) to measure δ¹⁵N and δ¹³C in the gas. Isotopic data were calibrated with lab standards to convert the raw data to δ¹⁵N (air) and δ¹³C (VPBD). The lab standards are calibrated annually with IAEA standards. The run precision was 0.04 per mil δ¹⁵N and δ¹³C.

To measure P and K concentrations (% dry mass), 0.25 g of prepared tissue was weighed into a 50 ml Kimax digestion tube. In a fume hood, 2 ml of concentrated sulfuric acid was added, then 1 ml 30% hydrogen peroxide. After oxidation subsided, the tube was placed on a digestion block at 285°C for 20 min. After cooling to room temperature, another 1 ml of hydrogen peroxide was added before returning the tube to the digestion block. This was repeated until the acid in the tube was colorless. After cooling, the sample was diluted to 50 ml with ddH₂0 and mixed by inverting the tube twice. P and K concentrations were then measured using inductively coupled plasma spectrometry.