Capture of mammal excreta by Nepenthes is an effective heterotrophic nutrition strategy

Abstract

Background and aims

While isotopic enrichment of nitrogen (¹⁵N) and carbon (¹³C) is often used to determine whether carnivorous plant species capture and assimilate nutrients from supplemental sources such as invertebrate prey or mammal excreta (heterotrophic nutrition), little is known about how successful the different strategies deployed by carnivorous plants are at obtaining supplemental nutrition. The collection of mammalian faeces by Nepenthes (tropical pitcher plants) is the result of a highly specialized biological mutualism that results in heterotrophic nitrogen gain; however, it remains unknown how effective this strategy is in comparison to Nepenthes species not known to collect mammalian faeces.

Methods

We examined how isotopic enrichment varied in the diverse genus Nepenthes, among species producing pitchers for invertebrate capture and species exhibiting mutualisms for the collection of mammal excreta. Enrichment factors were calculated from δ¹⁵N and δ¹³C values from eight Nepenthes species and naturally occurring hybrids along with co-occurring reference (non-carnivorous) plants from three mountain massifs in Borneo: Mount Kinabalu, Mount Tambuyukon and Mount Trus Madi.

Results

All Nepenthes examined, except N. edwardsiana, were significantly enriched in ¹⁵N compared to co-occurring non-carnivorous plants, and ¹⁵N enrichment was more than two-fold higher in species with adaptations for the collection of mammal excreta compared with other Nepenthes.

Conclusions

The collection of mammal faeces clearly represents a highly effective strategy for heterotrophic nitrogen gain in Nepenthes. Species with adaptations for capturing mammal excreta occur exclusively at high elevation (i.e. are typically summit-occurring) where previous studies suggest invertebrate prey are less abundant and less frequently captured. As such, we propose this strategy may maximize nutritional return by specializing towards ensuring the collection and retention of few but higher-value N sources in environments where invertebrate prey may be scarce.

INTRODUCTION

Carnivorous plants supplement their nutrient requirements by attracting, capturing and in most cases actively digesting animal prey (primarily invertebrates), and occur most frequently on nutrient-poor substrates in areas with high light and seasonal or perennial moisture (Givnish et al., 1984; Clarke and Moran, 2016). Approximately 860 species of carnivorous plants from 20 genera, 13 families and six orders are currently described, mostly from Lamiales and Caryophyllales (Fleischmann et al., 2018; Cross et al., 2020; Lin et al., 2021). Nepenthaceae (Caryophyllales) is a monotypic carnivorous plant family of over 160 species, and the largest family employing ‘pitcher’ pitfall traps (Clarke et al., 2018a). Nepenthes occur throughout the Palaeotropics, being distributed predominantly within the Malesian and Papuasian biogeographical regions but ranging from Madagascar to New Caledonia and several outlying western Pacific islands (Jebb and Cheek, 1997; Moran et al., 2010). Nepenthes are terrestrial or epiphytic vines or subscandent shrubs that produce a characteristic leaf structure comprising a leaf-like phyllodium (modified leaf base) from the apex of which a tendril (the true petiole) emerges, giving rise to an epiascidiate (tubular, with the abaxial surface on the inside) lamina or ‘pitcher’ (representing the true leaf); these evolved primarily as the means of attracting, capturing and digesting prey, and occur in the form of so-called ‘lower pitchers’ and ‘upper pitchers’ specialized towards the capture of crawling and flying insects, respectively (Kubitzki, 2002; Bauer et al., 2009; Moran et al., 2010).

Pitcher shape, size and colour are highly variable among species of Nepenthes (Moran et al., 2013), but most species produce pitchers with an overhanging lid, a more or less cylindrical peristome surrounding the pitcher mouth (sometimes flattened, markedly flanged or toothed), and interior walls comprising a glandular zone sometimes surmounted by an upper waxy zone. Extrafloral nectaries on the tendril, outer pitcher wall, underside of the lid and between each pair of teeth on the inner edge of the peristome act as prey attractants (Merbach et al., 2001; Kubitzki, 2002; Di Giusto et al., 2008). Prey fall into the pitcher upon encountering anti-adhesive ‘slippery’ surfaces on the peristome or on the pitcher wall (Gaume et al., 2002, 2004; Bohn and Federle, 2004); these surfaces often comprise epicuticular wax crystals which detach and adhere to the tarsi and claws of invertebrate prey (Bohn et al., 2004;Bauer et al., 2009) or comprise microstructures such as trichomes, lunate cells and digestive glands that impair grip (Gaume et al., 2002). Captured animals drown in an enzyme cocktail fluid that also facilitates the decomposition of, and nutrient-acquisition from, prey (Bauer et al., 2009; Buch et al., 2013).

It is believed that heterotrophic nutrient-acquisition in Nepenthes, as for other carnivorous plants studied (e.g. Skates et al., 2019), is an evolutionary strategy to improve the competitive ability of these species on nutrient-impoverished substrates, especially at high elevation (Owen, 1999; Bauer et al., 2009). While Nepenthes occur on a variety of nutrient-impoverished soils developed from sandstone, granite or limestone bedrock (Clarke et al., 2016), many species occur in or are restricted to ultramafic soils derived from ultramafic bedrock that is not only severely nutrient-poor, but also contains toxic concentrations of nickel and magnesium (van der Ent et al., 2016). On Borneo, a large island in the Malay Archipelago harbouring the richest morphological diversity of Nepenthes, many species occur in montane habitats on numerous ultramafic mountain ranges, especially in Kinabalu Park (van der Ent et al., 2015a). Kinabalu Park is well known for its exceptional biodiversity, which includes over 5000 plant species in an area of <1200 km² (Beaman, 2005), of which over 2500 plant species are known from ultramafic outcrops (van der Ent et al., 2015b). The Mount Kinabalu massif (4095 m asl) has a granitic summit, but is flanked by numerous low- and high-elevation ultramafic outcrops, whilst Mount Tambuyukon (2574 m asl) located 12 km to the north is entirely ultramafic. It is also a hotspot for Nepenthes with 12 taxa, including the world’s largest, N. rajah (Clarke, 1997). An earlier investigation has shown that Nepenthes are restricted to areas with extremely high insolation and permanently wet soils, brought about by the combined effects of an edaphic filter (ultramafic soils) and a climatic filter (elevation) (van der Ent et al., 2015a). Notably, a number of Bornean Nepenthes, predominantly species occurring at high elevation, have evolved morphological adaptations that facilitate the capture of mammalian faeces.

Clarke et al. (2009) and Chin et al. (2010) provided the first evidence of an extraordinary mutualistic association between three species of Bornean Nepenthes, N. lowii, N. rajah and N. macrophylla, with mountain tree shrews (Tupaia montana, of the mammalian order Scandentia). They established that T. montana visited these species of Nepenthes to feed on carbohydrate-rich secretions produced on the pitcher lids, and habitually defecated into the pitchers. Subsequent studies by Wells et al. (2011) and Greenwood et al. (2011) found evidence of a similar mutualism between N. rajah and the summit rat, Rattus baluensis (Rodentia), and concluded that R. rajah pitchers on Mount Kinabalu were routinely visited by both mammals at a rate of approximately one visit every 4 h. Visitation appears to occur exclusively during daylight hours for T. montana and predominantly at night for R. baluensis, potentially offering greater nutritional reward for N. rajah compared with other species collecting the excreta only of T. montana (e.g. N. lowii; Greenwood et al., 2011). In N. lowii 57–100 % of foliar N derives from scats, and the pitcher modifications that facilitate the association with T. montana have rendered its aerial pitchers largely ineffective for attracting and trapping invertebrate prey (Clarke et al., 2009). It has been hypothesized that T. montana and R. baluensis exclude other small mammals from participating in mutualism with N. rajah by monopolizing the pitcher lid secretion resource during their preferred diel periods of activity (Greenwood et al., 2011). However, while the distribution of T. montana encompasses considerable portions of Borneo where Nepenthes occur, R. baluensis is a range-restricted endemic from Mount Kinabalu and it remains unknown whether other nocturnal small mammals are involved in similar mutualism with Nepenthes. Nepenthes burbidgeae and N. villosa have no apparent association with T. montana, and Chin et al. (2010) argued that the combination of pitcher orifice size, degree of lid reflexion and the vaulted shape of the lid forced visiting tree shrews to straddle the pitchers to feed on the lid gland exudates, and that only Nepenthes species that possessed this combination of characteristics could engage in resource-exchange mutualism with T. montana. Since arthropod diversity and abundance decrease as elevation increases (Collins, 1980; Samson et al., 1997), there may be significant selective pressure for highland Nepenthes to exploit alternative sources of nutrition. Although tree shrew faeces contain less N per unit dry weight than ants (Chin et al., 2010), the steady supply and amount of faecal material represents a substantial contribution to nutrient budget (Clarke et al., 2009).

Two other Bornean Nepenthes are known to engage in biological mutualism related to faeces capture. The lowland species Nepenthes hemsleyana gains an estimated 34 % of its total foliar nitrogen from the urine and faeces of Hardwicke’s woolly bats (Kerivoula hardwickii hardwickii), which habitually roost in its aerial pitchers (Grafe et al., 2011). Nepenthes hemsleyana pitcher occupancy by K. hardwickii was almost 20 % over a 6.5-week period of monitoring, with nearly 30 % of all plants harbouring at least one K. hardwickii individual. Additionally, pitchers of the montane N. macrophylla are apparently frequented by a bird, the mountain blackeye (Chlorocharis emiliae), which have been observed to defecate into the pitchers (Moran et al., 2018). Interestingly, no faecal capture mutualisms have yet been reported for any Nepenthes outside of Borneo.

Comparison of relative isotope abundances in plant tissues can be used to describe the nutritive relationships between plants and their environment. The extent to which carnivorous plants are reliant upon or obtain benefit from heterotrophic nutrient acquisition can be determined by applying the theory of trophic isotopic enrichment (Deniro and Epstein, 1981), which states that consumers exhibit greater heavy isotope enrichment compared with their food sources. Due to discrimination against the heavier ¹⁵N isotope during excretion, animal tissues have higher ¹⁵N abundance compared with autotrophic plants (Deniro and Epstein, 1981; Minagawa and Wada, 1984; Peterson and Fry, 1987; Boecklen et al., 2011) and numerous carnivorous plant species that capture and assimilate nutrients from trapped invertebrates exhibit ¹⁵N enrichment that is lower than the invertebrate prey but higher than in co-occurring non-carnivorous plants (e.g. Schulze et al., 1991; Klink et al., 2019; Skates et al., 2019). However, reliance upon heterotrophic nutrition appears variable between carnivorous plant genera, between species within genera, and even among individuals of the same species from different habitats (Skates, 2021). As the nutritional returns from prey capture appear to correlate strongly with increasing investment in carnivory (e.g. Adamec, 2008), it might be expected that Nepenthes investing resources in the production of pitchers specialized for the collection of mammal excreta derive comparatively greater heterotrophic nutritional reward compared with species producing more morphologically ‘typical’ pitchers. However, it remains unknown to what extent the capture of mammalian faeces by Nepenthes represents an effective strategy for obtaining supplementary nutrition.

Understanding the ecological requirements of species, particularly where species are involved in complex biological mutualisms (Cross et al., 2019), is crucial in conservation planning and management for threatened carnivorous plants (Cross et al., 2020). Nepenthes contains more species at risk of extinction than any other carnivorous plant genus, with 40 % of currently recognized species assessed as Critically Endangered, Endangered or Vulnerable against the IUCN 3.1 Red list criteria (IUCN, 2012), and so improving our understanding of their ecological requirements is an urgent priority (Clarke et al., 2018b; Cross et al., 2020). This study examined heterotrophic nitrogen gain in six species and four naturally occurring hybrids of Nepenthes from ultramafic and non-ultramafic locations in Malaysian Borneo on Kinabalu and on Mount Trus Madi, investigating whether Total [N] and Total [C] as well as ¹⁵N and ¹³C enrichment varied among species with and without adaptations that facilitate the collection of mammal excreta.

MATERIALS AND METHODS

Sample collection

Collections of ten Nepenthes taxa (Figs 1 and 2), as well as co-occurring non-carnivorous plants (that serve as ‘nutritional reference plants’), were collected in Malaysian Borneo during March to May in 2012–2014 as described in van der Ent et al. (2015a). Briefly, foliar samples of 300–1000 mg were collected from plants in 38 plots on ultramafic sites inside Kinabalu Park, including around Mount Kinabalu (itself a granite intrusion) at Marai Parai [N. edwardsiana, N. rajah, N. burbidgeae, N. × alisaputrana (= N. burbidgeae × N. rajah)], Layang-Layang (N. villosa, N. × kinabaluensis (= N. rajah × N. villosa)], Mesilau (N. rajah, N. rajah × lowii), Bukit Babi (N. rajah), Bambangan (N. burbidgeae, N. rajah), Bukit Hampuan (N. burbidgeae), and Mount Tambuyukon (N. lowii, N. villosa, N. burbidgeae, N. edwardsiana, N. rajah). Additionally, samples were collected from nine plots on non-ultramafic (i.e. predominantly sedimentary; Acres, 1972) Mount Trus Madi [N. lowii, N. macrophylla, N. × trusmadiensis (= N. lowii × N. macrophylla]). Plots each contained at least five Nepenthes individuals and those on Mount Kinabalu and Mount Tambuyukon provided a representative sample of the different ultramafic vegetation types present on these peaks (e.g. tall lower montane forest, short upper montane forest, sub-alpine scrub). Details of vegetation plots are provided in van der Ent et al. (2016), but in summary Mount Kinabalu and Mount Tambuyukon plots ranged from 1332 to 2950 m above sea level (median 2393 m) and Mount Trus Madi plots ranged from 2351 to 2642 m above sea level (median 2623 m).

Fig. 1.

Nepenthes species sampled in Malaysian Borneo during this study: (A) N. burbidgeae, (B) N. edwardsiana, (C) N. lowii, (D) N. macrophylla, (E) N. rajah and (F) N. villosa. Note the presence of scat in D and E. Photographs by A. van der Ent (A), A. Robinson (B–F).

Fig. 2.

Nepenthes hybrids sampled during this study: (A) N. × alisaputrana, (B) N. × kinabaluensis, (C) N. rajah × lowii and (D) N. × trusmadiensis (note mammalian scat presence). Photographs by A. van der Ent (A), A. Robinson (B, D) and J. Thong (C).

Sampling comprised one leaf sample per biological replicate (i.e. one individual Nepenthes or reference plant), with five biological replicates per species per site. Leaves were stored individually in paper bags within a zip-lock bag containing silica desiccant beads for 5–8 d during fieldwork (silica gel completely desiccates small leaf samples and fully arrests metabolism/degradation within 24 h), prior to being dried for 48 h at 70 °C and stored in a desiccator containing silica desiccant beads. Reference plants included only species with no (known) specialized strategies for nutrient-acquisition (i.e. excluding N₂-fixing legumes, orchids and parasitic species; Lambers and Oliveira, 2019), and were selected if they co-occurred with and exhibited the same or a similar growth habit and life history to Nepenthes (i.e. herbaceous rosetted or climbing perennials). It was not possible to collect mammalian faeces for analysis.

Stable isotope abundance analysis and calibration, and calculation of isotopic enrichment

Dried samples were ground using ball mills (MM2 and MM400; Retsch, Haan, Germany), with samples placed into grinding cylinders with steel balls, clamped in the ball mill, and ground for ~5 min until rendered a homogeneous fine powder. Following grinding, samples were stored in paper envelopes and dried overnight at 105 °C before being cooled to room temperature and stored in a desiccator containing silica desiccant beads prior to stable isotope analyses. Ground samples were weighed into tin capsules (SA76980702; IVA Analysentechnik GmbH & Co. KG, Meerbusch, Germany) using a micro-analysis scale (CPA2P; Sartorius, Göttingen, Germany), with desired weight being 2.5–3.5 mg of ground sample. Capsules were sealed using steel forceps, sorted into a storage palette and stored in a desiccator containing silica desiccant beads.

To enable comparison of stable isotope data to results from other laboratories, isotope abundance data were referred to primary standards [Vienna Pee Dee Belemnite (V-PDB) for ¹³C and atmospheric air for ¹⁵N], by calibration of standard gases (CO₂ for ¹³C and N₂ for ¹⁵N; Rießner Gase, Lichtenfels, Germany) with secondary standards including IAEA-CH6 (sucrose) and NBS 19 (limestone) for ¹³C as well as IAEA-N1 and IAEA-N2 (ammonium sulphate) for ¹⁵N. Secondary standards were sourced from the International Atomic Energy Agency (IAEA, Vienna, Austria). Acetanilide (C-content: 71.09 %, N-content: 10.36 %; Merck KGaA, Darmstadt, Germany) in amounts of 0.4–1.6 mg, measured regularly throughout analyses, was used to control the reproducibility of the isotope abundance measurements (±0.2 ‰ or better) and to create regression lines for the calculation of Total [C] and Total [N] in plant samples (Gebauer and Schulze, 1991). An additional blank value was measured to prevent errors due to leaks by analysing empty tin capsules at the beginning of each measurement cycle.

Samples were analysed at the Laboratory of Isotope Biogeochemistry, a central facility of the Bayreuth Center of Ecology and Environmental Research (BayCEER) at the University of Bayreuth, Germany. The C- and N-isotope abundances and tissue N or C content (i.e. total concentration in mmol/g_dw, hereafter referred to as Total [C] and [N]) of ground samples were determined simultaneously using elemental analysis/isotope ratio mass spectrometry (EA-IRMS] coupling, comprising an NA 1108 elemental analyser (CE Instruments, Milan, Italy), a ConFlo III interface (Thermo Fisher Scientific, Bremen, Germany) and a delta S isotope ratio mass spectrometer (Finnigan MAT, Bremen, Germany).

Isotopic composition was analysed by determination of relative nitrogen and carbon isotope abundances rather than absolute values. Relative isotopic abundance was determined by comparing samples with predetermined standards and denoted as δ values. To determine isotopic enrichment in Nepenthes samples, an enrichment factor ε (sensuPreiss and Gebauer, 2008) was calculated using reference plant samples collected from the same sites as Nepenthes samples. Reference samples were labelled with co-occurring Nepenthes allowing for clear assignment of reference samples to each Nepenthes sample. Only reference samples were used for the calculation of ε¹⁵N and ε¹³C values, with ε calculated as $\epsilon {\mathrm{X}}_{heavy}\left(Nepenthes\right)\left[‰\right]=\mathrm{ẟ}{\mathrm{X}}_{heavy}\left(Nepenthessample\right)-\overline{\mathrm{ẟ}{\mathrm{X}}_{heavy}\left(Reference\right)}$ where $\overline{\mathrm{ẟ}{\mathrm{X}}_{heavy}\left(Reference\right)}$ was defined as the mean value of reference plants assigned to the Nepenthes sample and calculated as $\overline{\mathrm{ẟ}{\mathrm{X}}_{heavy}\left(Reference\right)}=\frac{\sum _{i=1}^n\mathrm{ẟ}{\mathrm{X}}_{heavy}}{n}.$

The enrichment factor of reference samples, which is zero by definition, was calculated analogously to include scattering of the values into the analysis as $\epsilon {\mathrm{X}}_{heavy}\left(Reference\right)\left[‰\right]=\mathrm{ẟ}{\mathrm{X}}_{heavy}\left(Referencesample\right)-\overline{\mathrm{ẟ}{\mathrm{X}}_{heavy}\left(Reference\right)}.$ Mean enrichment factors, as well as mean Total [C] and Total [N] (mmol g d.wt⁻¹), were calculated both within each site and across all sites for each Nepenthes species and for reference plants, as well as for all Nepenthes species at each site. Due to a lack of reference samples collected at Bukit Babi, enrichment factors were not calculated for Nepenthes samples at this site. Raw data for Total [C], Total [N] and enrichment factors is provided in the Supplementary Data (Appendix S1).

Statistical analysis

Each sampled Nepenthes was categorized as either having evidence of adaptations for the collection of mammal excreta from the scientific literature (N. lowii, N. macrophylla, N. rajah, N. × trusmadiensis; sensuClarke et al., 2010, 2018a; Moran et al., 2012) or as not having a known biological mutualism (N. burbidgeae, N. edwardsiana, N. villosa, N. × alisaputrana, N. × kinabaluensis, N. rajah × lowii). As data violated assumptions of normality and homogeneity of variances, non-parametric independent-samples Kruskal–Wallis H tests (SPSS Statistics 27, IBM, USA) were used to determine if there were differences in Total [C], Total [N] and enrichment factors (ε¹⁵N and ε¹³C) between Nepenthes and reference plants, between different species of Nepenthes, and between Nepenthes with and without biological mutualisms. A two-tailed bivariate Pearson correlation was used to assess the relationship between elevation and enrichment factors (ε¹⁵N and ε¹³C) and Total [C] and Total [N] for Nepenthes and non-carnivorous reference plants.

RESULTS

Total [N], Total [C], and enrichment factors ε¹⁵N and ε³C in Nepenthes and reference plants

Mean Total [N] was consistently 20–25 % higher for Nepenthes samples (0.8 ± 0.02 mmol g d. wt⁻¹) compared with reference plant samples (0.6 ± 0.02 mmol g d. wt⁻¹), χ²(1) = 25.046, P < 0.001. An exception to this otherwise global trend was N. burbidgeae at a single site, Bambangan (Fig. 3), where Total [N] was markedly lower than the global mean for this species (0.5 ± 0.04 compared with 0.7 ± 0.04 mmol g d. wt⁻¹) and was similar to co-occurring reference plants. In contrast to Total [N], mean Total [C] was not significantly different between reference plants and Nepenthes samples (Fig. 3) (χ²(1) = 0.196, P = 0.658). Total [C] varied from 30.1 to 49.1 mmol g d. wt⁻¹ in reference plants (mean 42.5 ± 0.29 mmol g d. wt⁻¹), and ranged from 41.3 ± 0.18 mmol g d. wt⁻¹ for N. burbidgeae to 45.6 ± 0.34 mmol g d. wt⁻¹ for N. × trusmadiensis (mean 43.0 ± 0.15 mmol g d. wt⁻¹ for all Nepenthes samples). For Nepenthes samples, increasing elevation was positively associated with Total [N] (R² = 0.027, t = 2.147, P = 0.033), Total [C] (R² = 0.035, t = 2.479, P = 0.014), and enrichment factors ε¹⁵N (R² = 0.030, t = 2.298, P = 0.023) and ε¹³C (R² = 0.056, t = −3.146, P = 0.002). For reference samples, the same association was evident for Total [C] (R² = 0.118, t = 3.979, P < 0.001), but not for Total [N] (R² = 0.001, t = −0.421, P = 0.675), ε¹⁵N (R² = 0.001, t = −0.404, P = 0.687) or ε¹³C (R² = 0.000, t = 0.049, P = 0.961).

Fig. 3.

Mean leaf Total [N] and Total [C] for ten Bornean Nepenthes taxa (black symbols) and for co-occurring reference plants (white symbols) from seven sites in Borneo.

Nepenthes samples were significantly different from reference plant samples in both ε¹⁵N (mean 3.8 ± 0.19 %o) and ε¹³C (−0.8 ± 0.18 %o) (χ²(1) = 88.428, P < 0.001 and χ²(1) = 4.306, P = 0.038, respectively). Mean ε¹⁵N was significantly higher than reference plants for N. villosa (1.7 ± 0.45 %o), N. burbidgeae (4.3 ± 0.20 %o), N. lowii (4.5 ± 0.56 %o), N. × alisaputrana (4.6 ± 0.57 %o), N. rajah (4.9 ± 0.38 %o), N. macrophylla (5.1 ± 0.10 %o), N. × trusmadiensis (6.1 ± 0.42 %o) and N. rajah × lowii (6.5 ± 0.03 %o), but not for N. edwardsiana which had the lowest mean ε¹⁵N of sampled Nepenthes at 0.3 ± 0.18 %o (Fig. 4). Mean ε¹³C in Nepenthes samples was lower than or similar to that of reference plants in all taxa except N. rajah and N. × trusmadiensis, and was lowest in N. edwardsiana (Fig. 4),

Fig. 4.

Total [N], Total [C], and enrichment factors (ε¹³C and ε¹⁵N) for ten Bornean Nepenthes taxa with (white symbols) and without (black symbols) known biological mutualisms for the collection of mammal excreta, compared with values for co-occurring reference plants. Dashed lines indicate mean reference values.

Total [N], Total [C], and enrichment factors ε¹⁵N and ε¹³C were significantly different between Nepenthes with and without adaptations for the capture of mammalian excreta (Table 1). Compared to species with no known biological mutualism, Nepenthes with reported mutualisms for the collection of mammal excreta had ~1.3 times higher mean Total [N] (0.9 ± 0.04 compared with 0.7 ± 0.02 mmol g d. wt⁻¹ N) and almost two-fold higher mean ε¹⁵N (4.9 ± 0.25 compared with 2.9 ± 0.24 %o), as well as higher mean Total [C] (43.6 ± 0.20 compared with 42.4 ± 0.21 mmol g d. wt⁻¹) and markedly higher mean ε¹³C (−0.2 ± 0.22 compared with −1.3 ± 0.26 %o). Species with biological mutualisms were significantly more likely to have been collected from sites at higher elevation (2251 ± 36.1 m) than species without mutualisms (2051 ± 51.8 m) (χ²(1) = 8.857, P = 0.003).

Table 1.

Results of independent-samples Kruskal–Wallis H tests examining differences in Total [N], Total [C] and enrichment factors (ε¹⁵N and ε¹³C) between Nepenthes and non-carnivorous reference plants occurring on ultramafic and non-ultramafic soils, and between Nepenthes species and naturally occurring hybrids with and without known biological mutualisms for the collection of mammal excreta

Factor	Variable	Sample	χ2	P
Underlying geology	Total [N]	Nepenthes   Reference plants	7.696 1.024	0.006 0.312
	Total [C]	Nepenthes   Reference plants	47.826 19.823	<0.001 <0.001
	ε15N	Nepenthes   Reference plants	4.231 0.129	0.040 0.720
	ε13C	Nepenthes   Reference plants	0.181 0.182	0.671 0.670
Biological mutualism	Total [N]	Nepenthes	20.769	<0.001
	Total [C]	Nepenthes	13.906	<0.001
	ε15N	Nepenthes	33.931	<0.001
	ε13C	Nepenthes	10.557	0.001

Relationship between enrichment factors, Total [N] and Total [C]

Pearson correlation indicated a significant positive relationship between Total [N] and Total [C] for Nepenthes samples (n = 170, R² = 0.374, P < 0.001), and a weak positive relationship between Total [N] and ε¹⁵N (n = 170, R² = 0.165, P = 0.031). A positive correlation was also evident between ε¹⁵N and ε¹³C (n = 170, R² = 0.366, P < 0.001) (Fig. 5). Similar relationships were not evident for reference plant samples (Fig. 5); instead, in reference plant samples Total [N] was negatively correlated with ε¹³C (n = 120, R² = −0.227, P = 0.013) and Total [C] was negatively correlated with ε¹⁵N (n = 120, R² = −0.213, P = 0.019).

Fig. 5.

Significant correlative relationships between Total [N], Total [C] and enrichment factors (ε¹³C and ε¹⁵N) for ten Bornean Nepenthes taxa and co-occurring reference plants.

DISCUSSION

We present clear evidence of heterotrophic N gain in Bornean Nepenthes, and provide compelling data to indicate that the collection of mammalian excreta represents an effective heterotrophic nutrition strategy within this genus. All taxa examined, except for N. edwardsiana, were significantly enriched in ¹⁵N. The theory of trophic isotopic enrichment states that consumers generally exhibit greater enrichment in heavier isotopes than the organisms they consume (Deniro and Epstein, 1981). Although we did not examine δ¹⁵N in the insect prey or mammalian faeces captured by Nepenthes, previous studies indicate that carnivorous plants deriving heterotrophic N gain from insect prey commonly exhibit ε¹⁵N values that are elevated compared with non-carnivorous companion plants and intermediate between these and captured insect prey (e.g. Schulze et al., 1991; Skates et al., 2019). As such, the ¹⁵N enrichment observed for Nepenthes indicates they have some level of reliance upon heterotrophic means for N nutrition, rather than relying solely upon absorption from soil. Our data indicate there is marked variation in ε¹⁵N among different Nepenthes taxa, supporting a hypothesis that different pitcher morphologies and ecological strategies deployed by Nepenthes enhance heterotrophic N gain.

Factors influencing ε¹⁵N enrichment in Nepenthes

Total soil [N] is universally low (0.1–0.3 %) in both ultramafic and non-ultramafic habitats where sampled Nepenthes occur in Borneo (van der Ent et al., 2016), and carnivory probably provides an advantage by increasing access to Total [N] in these N-limited environments. Floristic diversity declines along an increasing elevation gradient in these ecosystems, and substrates become increasingly C-rich and acidic (van der Ent et al., 2016). As N is poorly retained in soils at low pH (Lambers et al., 2008), N-limitation is probably greatest at higher elevation and thus carnivory can represent an effective competitive strategy in these habitats. Since Nepenthes sampled from higher elevation in the present study were enriched in ¹⁵N compared with taxa from lower elevation, suggesting greater heterotrophic N gains, we argue there has been pressure for Nepenthes to evolve specialized mechanisms for heterotrophic nutrition at higher elevation. The most extreme examples of morphological and ecological specialization in Nepenthes include the strongly flanged peristomes of species such as N. edwardsiana, N. hamata and N. villosa (‘lobster-pot’ peristomes; Fig. 1B, D, F); mammalian faeces collection by N. rajah, N. macrophylla and N. lowii (toilet pitchers; Fig. 1C–E); the strongly infundibular (funnel-shaped), sticky surface in the upper pitchers of N. dubia, N. eymae, N. flava, N. inermis, N. jamban, N. pitopangii and N. undulatifolia (fly-paper traps); the wet season formation of a liquid bilayer in N. attenboroughii pitchers, with a prey-retaining, jelly-like fluid in the lower half and water above (prey retention matrix); the light traps of N. aristolochioides and presumably N. klossii; and the wide, ruff-like peristomes of N. jacquelineae and N. platychila (McPherson, 2009; Robinson et al. (2009); McPherson and Robinson, 2012; Baj et al., 2022).

Evidence in the literature indicates that microclimatic conditions at different elevations influence the properties of insect communities (e.g. Wolda, 1987; Plant et al., 2018). This presumably alters the spectra of potential prey available to Nepenthes and hence could influence the type and number of prey captured, as well as the relative importance of more diverse (i.e. not prey-derived) N-sources. Although few studies have examined prey spectra in Nepenthes, Formicidae (ants) generally comprise the vast majority (up to 100 %) of prey captured by most taxa in lowland habitats (<500 m), particularly in ‘lower pitchers’, whereas ‘upper pitchers’ predominantly trap flying prey even when positioned close to the ground (Jebb, 1989; Moran, 1996; Adam, 1997; Rembold et al., 2010). However, ants become increasingly less abundant as captured prey with increasing elevation for Bornean Nepenthes and were totally absent from taxa sampled at higher elevation (>2300 m) on Mt Kinabalu (Adam, 1997), potentially because perpetually moist, mossy forest at these elevations is unfavourable habitat to ants (Hotta, 1989). Nepenthes occurring at higher elevations capture fewer total but a greater diversity of prey, often from insect orders in which individual prey items are physically larger such as Chilopoda (centipedes), Coleoptera (beetles), Dermaptera (earwigs) and Dictyoptera [particularly Blattaria (cockroaches]) (Adam, 1997). Larger prey potentially represent higher value captures in terms of Total [N] per capture.

We hypothesize that Nepenthes have become specialized for contrasting solutions to nutrient deficiency at high elevation, and that specialization evident in Nepenthes from high elevations arose from the benefit of diverse trapping regimes that facilitate nutritional returns through (1) the attraction, capture and/or retention of specific prey groups; (2) the capture and retention of prey under environmental conditions that would render typical pitcher morphology less effective; or (3) the attraction and retention of nutrient-rich animal by-products (i.e. faeces). None of these are mutually exclusive with respect to incidental prey capture. The many examples of different species with different trapping regimes stably coexisting – often literally side by side – as components of the climax vegetation on various peaks across Malesia may reflect both the directions that this selection pressure may take and its long-term, success-driven stability. In addition to the syntopic occurrence of species with different trapping strategies highlighted in the present study, N. izumiae (conventional pitfall trap), N. jacquelineae (wide peristome) and N. dubia (flypaper) co-occur in West Sumatra (A. Robinson and Stefan Matschke, pers. observ.); N. dubia (flypaper), N. jamban (flypaper) and N. lingulata (conventional) co-occur in North Sumatra (A. Robinson, pers. observ.); and N. hamata (flanged peristome), N. nigra (conventional), N. glabrata (conventional) and N. eymae (flypaper) co-occur on the summit of Mt Tambusisi, Central Sulawesi, also with N. tentaculata and N. maxima (both conventional) at mid-montane elevations (A. Robinson, pers. observ.; see also Cheek et al., 2019 for formal classification of functional pitcher types). The common co-occurrence of different trapping regimes in this manner may imply niche segregation on the basis of prey specialization in Nepenthes, as has been previously demonstrated for coexisting Drosera from monsoon tropical northern Australia (Krueger et al., 2020).

Collection of mammalian faeces

The collection of mammalian faeces by Nepenthes clearly represents a highly effective strategy for heterotrophic nitrogen gain. Species with known biological mutualism for faeces collection, N. lowii, N. rajah, N. macrophylla and N. × trusmadiensis, all exhibited relatively high ε¹⁵N values. Nepenthes lowii is distinctive among Nepenthes in its extreme pitcher dimorphism; it produces fairly typical ‘lower’ pitchers in its rosette stage – these are generally ovoid-ventricose below, cylindrical to narrowly infundibular above, rest on supporting moss or tree limbs, and trap crawling arthropods – while its ‘upper’ pitchers are extremely globose below, dramatically constricted at the midsection and then widely flared above, appearing somewhat akin to a free-hanging, aerial toilet bowl. Almost perpendicular to the pitcher opening is a large, vaulted lid bearing prominent bristles amongst which a fatty white secretion is produced. This secretion represents a valuable food for montane tree shrews (T. montana), which position themselves directly over the pitcher orifice whilst feeding such that their excreta are neatly captured (Chin et al., 2010). Indeed, Nepenthes lowii derives 57–100 % of its nitrogen from vertebrate faeces, a source unavailable to most other Nepenthes (Clarke et al., 2009). Disparity in N. lowii enrichment is apparent between samples collected from different mountains in this study; N. lowii is more abundant on Mt Trus Madi (where samples had greater enrichment in ¹⁵N), and T. montana are more frequently observed visiting its pitchers there, compared with on Mt Tambuyukon where N. lowii is less abundant (e.g. only two individuals occur directly on the summit trail; A. Robinson, pers. observ.).

Nepenthes rajah also attracts T. montana, which defecate into its pitchers while consuming nectar from the large and deeply concave pitcher lid (Wells et al., 2011; Moran et al., 2012). Faeces of other small mammals, such as the nocturnal summit rat, Rattus baluensis, have also been found inside the pitchers (Wells et al., 2011); and mammalian faeces account for 57–100 % of the foliar nitrogen in N. rajah (Clarke et al., 2009). However, recent studies of N. rajah pitchers at Mesilau on Mt Kinabalu indicate the species also captures large numbers of invertebrate prey (79 % Formicidae) and, very rarely, vertebrate prey (Chin et al., 2010). Mammalian scats have also been found inside the pitchers of N. macrophylla, a species producing large pitchers with concave, reflexed lids that position visiting mammals in a manner allowing for faeces capture and which is believed to employ visual cues to attract T. montana (Chin et al., 2010; Moran et al., 2012). Future studies should examine the nutritional reward of mammalian scat capture in Nepenthes, and the prevalence of this mutualism across the genus, in greater detail. For example, ε¹⁵N values were very high in sampled N. × alisaputrana (the natural hybrid between N rajah and N. burbidgeae); although N. × alisaputrana is not known to utilize a mutualism for faeces capture and ε¹⁵N values were also high for the non-faecal capturing parent N. burbidgeae, N. × alisaputrana is similar to the other parent species N. rajah in both morphology and habitat and thus may possess a yet-undocumented mutualism.

Factors influencing ¹³C enrichment in Nepenthes

All sampled Nepenthes taxa were similar to or depleted in ε¹³C compared with reference plants, except for slight elevation in N. rajah and N. × trusmadiensis. Our data provide no evidence for a significant C gain from animal prey or mammalian faeces in Nepenthes. In principle, ¹³C enrichment might be expected for carnivorous plants given the additional C input from captured prey. However, ¹³C abundance in plant tissues is influenced by photosynthetic pathways and transpiration, as well as rates of soil respiration and C input source if C inputs other than photosynthesis are present (Klink et al., 2019), potentially obscuring identification of C gains from insect capture. Klink et al. (2019) proposed that depletion in ¹³C in carnivorous Pinguicula compared with autotrophic reference plants reflected the physiological advantage of growth in sunny, moist habitat where photosynthesis was efficient and outweighed by a higher transpiration rate resulting from increased leaf surface covered by sticky hairs. It may be insightful for future studies to further examine ¹³C enrichment in relation to cost–benefit models of carnivory for Nepenthes in order to elucidate the degree to which investment in carnivory influences or is influenced by physiological performance and how this varies among species and between different habitats. Furthermore, hydrogen stable isotope analysis (δ²H) in addition to C and N isotopes may be a useful tool to elucidate heterotrophic C gains from animal prey or mammalian faeces in Nepenthes (Gebauer et al., 2016; Schiebold et al., 2018; Cormier et al., 2019, 2022).

SUPPLEMENTARY DATA

Supplementary data are available online at https://academic.oup.com/aob and consist of the following. Appendix S1: Raw data for Total [N], Total [C], and enrichment factors ε¹⁵N and ε¹³C for Bornean Nepenthes and non-carnivorous reference plants occurring on ultramafic and non-ultramafic soils as studied.

FUNDING

We thank the International Carnivorous Plant Society for providing funding for stable isotope analyses, through a grant awarded to L.M.S. A.T.C. was supported by the Research Fellowship in Restoration Ecology jointly funded by the Ecological Health Network and Curtin University. L.M.S. was supported by an Australian Government Research Training Program (RTP) Scholarship.

CONFLICT OF INTEREST

The authors declare they have no conflicts of interest