Methods:
Experimental design
We used a 72-hour pulse-chase experiment with isotopically enriched amino acids as our organic nitrogen (ON) source and ammonium nitrate (15NH415NO3) as our inorganic nitrogen (IN) source to determine if pitcher plants can acquire ON directly and to compare ON and IN uptake under different conditions.
At each site, 125 mature individuals were selected with at least 3 live (no sign of senescence) mature pitchers (firm and open). Five of these plants were not manipulated with 15N and were harvested at the end of the experiment to determine baseline 15N and 13C natural abundance. The remaining 120 plants were randomly assigned to one of six treatment groups: U-Gly (98 atom% U-13C-15N-glycine), U-Phe (98 atom% U-13C-15N-phenylalanine), I15N (98 atom% 15NH415NO3), U-Gly plus unlabeled phenylalanine and unlabeled NH4NO3 (hereafter U-Gly+), U-Phe plus unlabeled glycine and unlabeled NH4NO3 (hereafter U-Phe+), or I15N plus unlabeled glycine and phenylalanine (hereafter I15N+). Plants within treatment groups were randomly assigned to one of two harvests (3- or 72-hr) and one of two food webs treatments (with complete food webs or food webs without the macroinvertebrate larvae [midges Metriocnemus knabi + mosquitoes Wyeomyia smithii). Larvae of the sarcophagid fly Fletcherimyia fletcheri were occasionally found in the pitchers and were excluded from all experimental pitchers. There were N = 5 pitchers for each treatment at each site.
Any liquid in the pitchers, along with the food web, was removed from all experimental pitchers the day before the pulse-chase experiment began; the liquid removed (pitcher “liquor”) was kept for the food web manipulations. Following food web removal in the field, pitchers were rinsed with distilled water to remove as much detritus and as many microbes as possible and the pitcher opening was blocked with a fine nylon mesh to limit subsequent entry of animals and prey. In the laboratory, all living midge and mosquito larvae were removed from liquid collected from each pitcher and kept alive overnight in a solution of pitcher liquor.
The next day, the largest pitcher on each plant was fed with one of the 15N treatments. We fed each manipulated pitcher with a 0.8 mM 15N solution (2 ml for Fort Albany and 9 ml for the larger pitchers at Tom Swamp) and an equal amount of pitcher liquor, resulting in pitchers filled to approximately three-quarters of their volume. Thus, all experimental pitchers contained an enriched (15N) nutrient solution along with the microbial component of the food web (supplied in the pitcher liquor). Pitchers at Fort Albany were fed 0.022 mg N, whereas the larger pitchers at Tom Swamp were fed 0.101 mg N. When we added only single forms of N (i.e., the U-Gly, U-Phe, and I15N treatments), all N added to the pitchers was enriched in 15N. When we added three forms of N (the U-Gly+, U-Phe+, and I15N+ treatments), only one-third of the N added to each pitcher was enriched in 15N; the remaining two-thirds was comprised of equal amounts of the other two forms as unlabeled N.
For the complete food web treatments, we put invertebrate larvae into the pitchers immediately after we added the 15N solution. We added two midge and two mosquito larvae in each pitcher in the complete food web treatment at Fort Albany and nine midge and nine mosquito larvae in each pitcher in the complete food web treatment at Tom Swamp (i.e., 1 midge + 1 mosquito larva per ml of pitcher liquor). Unfed (control) pitchers for which we measured natural abundance of 13C and 15N also had complete food webs (pitcher liquor + midge + mosquito larvae).
Harvest
Target pitchers were cut-off 3 or 72 hr after feeding with a stainless steel razor blade that was rinsed in 50% ethanol between cuttings. Pitcher liquid was transferred to a sealed sterile plastic tube and the pitcher was placed in a zip-lock plastic bag. Both were stored in a cooler with cold packs and shipped immediately to the laboratory for processing. Pitchers were cut open longitudinally, washed thoroughly with tap water, then rinsed with 0.5 mM CaCl2 to remove any amino acids from the surface, and finally rinsed three times with distilled-deionized water and transferred to paper bags. Midge and mosquito larvae were removed from the pitcher liquid with an eye dropper, transferred through three sequential baths of distilled-deionized water and stored in new sterile vials. Because of the small mass of larvae in each pitcher, larvae from the five replicates of each harvest x treatment combination were pooled into one composite larval sample. Plant and invertebrate samples were then oven-dried at 65 deg C for 48 h and then weighed.
Isotopic analyses
Each entire pitcher or composite larval sample was ground to a fine powder in a stainless steel capsule with a stainless steel ball using a Wig-L-Bug mixer (Bratt Technologies, LLC., East Orange, New Jersey, USA). A 4-mg subsample of plant tissue or a 1-mg subsample of larva was then placed into an 8 × 5 mm tin capsule (Elemental Microanalysis Mason, Ohio, USA) and combusted in a Costech ECS4010 Elemental Analyzer and DeltaPlus XP mass spectrometer at the University of New Hampshire to measure 13C/12C, %C, 15N/14N and %N concurrently.