Sclerified Cork Outperforms the Exodermis: Root Water Permeability Decreases in the Soil‐To‐Canopy Transition of the Aroid Vine Epipremnum aureum
André Mantovani, Yago Chagas Groba
TL;DR
Aroid vines adapt to canopy life by replacing root coverings to reduce water loss, enabling survival in atmospheric conditions.
Contribution
The study reveals that sclerified cork replaces exodermis in aerial roots of Epipremnum aureum, reducing water permeability during canopy transition.
Findings
Aerial roots of Epipremnum aureum develop a sclerified cork, reducing water loss compared to terrestrial roots.
Exodermis is replaced by cork in non-contact regions of aerial roots, while contact regions retain exodermis and root hairs.
Ligno-suberized tissues in aerial roots show structural differences between host-contact and atmospheric regions.
Abstract
The aroid vine Epipremnum aureum undergoes changes in habitat and growth axis direction from terrestrial (plagiotropic) to canopy (orthotropic) conditions. Since aerial roots connected to the forest soil are vital for water and nutrient uptake in these vines, we hypothesize that morphophysiological acclimation occurs, enabling root survival under atmospheric conditions. Root morpho‐anatomy, water balance, water absorption assayed via fluorescent tracer, and photochemical activity measured by chlorophyll fluorescence were analyzed. Gentle mechanical abrasion was applied to remove root coverings and compare water retention capacity in intact versus abraded roots. During the soil‐to‐canopy transitions, the root surface shifted from smooth, light brown in terrestrial roots to rough, dark brown surface in aerial roots, a change resulting from exodermis replacement by a sclerified cork in…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7
FIGURE 8
FIGURE 9| Plagiotropic | Orthotropic at 1.5 m in height | Orthotropic at 6 m in height | |||
|---|---|---|---|---|---|
| Terrestrial | Aerial roots | Aerial roots | |||
| Aerial phase | Terrestrial phase | Aerial phase | Terrestrial phase | ||
| External diameter (mm)–root intact | 2.20 ± 0.27a | 3.72 ± 0.78a | 4.56 ± 0.72ab | 6.45 ± 1.20c | 8.80 ± 1.03d |
| External diameter (mm)–root abraded | 1.91 ± 0.30a | 3.67 ± 0.18ab | 5.25 ± 1.13bc | 5.07 ± 1.32c | 10.13 ± 2.16d |
| Cortex width (mm) | 1.62 ± 0.21a | 2.00 ± 0.62a | 3.21 ± 0.61b | 3.60 ± 0.58b | 5.36 ± 0.91c |
| Stele diameter (mm) | 0.58 ± 0.28a | 1.63 ± 0.56 | 1.34 ± 0.20b | 3.08 ± 0.47c | 3.44 ± 0.19c |
| Covering tissue thickness (μm) | 42.68 ± 5.12a | 93.11 ± 13.93b | 47.02 ± 6.60a | 134.78 ± 15.93c | 54.63 ± 7.77a |
| Xylem vessel diameter (μm) | 70.36 ± 4.76a | 117.11 ± 55.67b | 147.69 ± 14.15b | 214.28 ± 31.16c | 280.58 ± 18.65d |
| Source | d.f | SS | MS | F |
|
|---|---|---|---|---|---|
| Growth axis‐size | 4 | 6.28 × 10−6 | 1.57 × 10 −6 | 4.16 | 0.006 |
| Covering | 1 | 1.95 × 10−5 | 1.95 × 10−5 | 51.60 | < 0.001 |
| Interaction | 4 | 1.42 × 10−5 | 3.56 × 10−6 | 9.43 | < 0.001 |
| Residuals | 50 | 1.89 × 10−5 | 3.77 × 10−7 |
- —PIBIC ‐ CNPq10.13039/501100003593
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsHorticultural and Viticultural Research · Plant nutrient uptake and metabolism · Plant Surface Properties and Treatments
Introduction
1
The Araceae family exhibits many growth habits, presenting submerged aquatic and terrestrial plants, saxicolous plants, hemiepiphytes, and sensu strictu epiphytes (Mayo et al. 1997). The survival and performance of these plants in such different environments is favored by morphophysiological strategies, especially at the leaf and stem levels (Filartiga et al. 2014; Brito et al. 2025). The aroid vines represent another habit in the family, which largely characterizes neotropical rainforest communities (Putz and Mooney 1991). Notably, aroid vines undergo marked habitat transitions during their life cycle. The initial development of aroid vines occurs horizontally, with plagiotropic growth on the soil surface supported by terrestrial roots, until they find a host tree (i.e., phorophyte). Once climbing begins, growth becomes orthotropic, and the individuals adhere to the phorophyte trunk while developing vertically toward the canopy. During this phase, aerial roots originating on the orthotropic shoot grow downward until they reach the forest soil, ensuring that access to this primary source of water and nutrients is never lost (Lee and Richards 1991; Zotz et al. 2021).
In aroid vines, this habitat transition from terrestrial to canopies is followed by allometric changes, such as increased internode thickness, lengthened leaf petioles, and increased leaf area, suggesting an improvement in plant growth (Ray 1992; Niklas 1994; Brito et al. 2022; Zotz et al. 2001). The soil to canopy transition imposes abiotic changes and may be stressful for these plants, for example, by drought, photodamage and herbivory (Freiberg 1997; Filartiga et al. 2014). In response to these new environmental conditions, aroid vines present different strategies at the leaf level that minimize the effects of excess light and dehydration, such as greater photochemical capacity, thicker mesophyll, greater succulence, and greater resistance to water loss through the cuticle (Mantovani 1999; Mantovani et al. 2017; Mantuano et al. 2021).
Morphological changes are not limited to leaves but can also be observed in the roots of aroid vines (French 1987). Individuals with plagiotropic growth develop thin, short roots, while orthotropic individuals have an aerial root system made up of two types of roots: anchor roots, which are short and branched and responsible for adhering to the host; and feeder roots, which are long and unbranched and connect to the soil to obtain water and nutrients (Lee and Richards 1991). Filartiga et al. (2018, 2021) showed that these root modifications in the aroid vine Rhodospatha oblongata Poepp. involved not only morphological changes, such as shape and size, but also anatomical and physiological changes. For example, the replacement of the original covering tissue by a new one that is more resistant to water loss, with an increased water transport capacity through the xylem. French (1987) states that sclerotic hypodermis in aerial roots of Philodendron Schott can disappear after entering the soil. Similarly, Schreiber and Franke (2011) and Suresh et al. (2022) attributed the very low water permeance of the aerial roots of Monstera deliciosa Liebm. to the presence of waxes besides suberin in their covering tissues. These initial data suggest that aerial roots of aroid vines become more efficient in the water balance throughout the transition from the soil to the canopy. More research is needed to validate this as a water conservation strategy for the aerial roots of aroid vines.
The aroid vine Epipremnum aureum (Linden and André) GS Bunting initially follows plagiotropic growth in the soil (Mantovani et al. 2017). After transitioning to orthotropic growth, the plant develops significantly larger leaves on the upper parts of the host tree (Brito et al. 2022) with higher photosynthetic capacity and dehydration resistance (Mantovani et al. 2017; Brito et al. 2025). In E. aureum , morphophysiological changes in the leaves are effective in minimizing the impacts of excess light and drought on the canopy (Mantovani et al. 2017). These strategies were triggered by multiple factors acting simultaneously, including the orthotropic stem growth axis, contact with the host, and high light (Brito et al. 2022). However, no studies have determined whether these strategies, induced or not by multiple factors, extend to the roots of this species when developed in an aerial environment.
Field observations of E. aureum aerial roots indicate modifications in external morphology, specifically in color, texture, and size (Brito et al. 2022). We hypothesized that the survival of E. aureum during the soil‐to‐canopy transition depends on a fundamental modification of its root surface tissue. We predicted that roots developing in the drier canopy atmosphere would form a more protective covering tissue, decreasing root water loss and thus facilitating the plant's establishment in the new habitat. We also investigated whether root permeance could be influenced by the interaction of multiple factors, as previously observed in the leaves and stems of E. aureum .
Materials and Methods
2
Study Area and Model Species
2.1
This study was conducted in a forest near the arboretum of the Rio de Janeiro Botanical Garden. The climate in the region is of the Am type (sensu Köeppen 1948) with an average temperature of 29°C in summer and 22°C in winter and an average annual rainfall of 1075 mm (Mantovani and Pereira 2005).
The terrestrial roots of the plagiotropic individuals of Epipremnum aureum are light brown, approximately 15–30 cm long, and may be ramified (Figure 1A). Soon after ascending the host, the orthotropic individuals of E. aureum begin to produce two types of aerial roots, anchor roots and feeder roots (Figure 1B). Anchor roots are short and ramified, never reaching the soil. Feeder roots are long and connect orthotropic individuals to the forest soil below (Figure 1C). Feeder roots are dark brown in orthotropic individuals beginning the ascension (Figure 1C,D), with smooth surfaces and sparse small lenticels. Higher orthotropic individuals (Figure 1E,F) produce larger leaves (Figure 1G) and thicker feeder roots covered by a rough surface with several large lenticels (Figure 1H). Anchor and feeder roots of E. aureum always grow adhered to the host trunk (Figure 1H,I). This host contact region appeared right above the root apex and remains along the root length. The aerial root surface developed trichomes when adhered to the host, while on the opposite side exposed to the air, the root surface was always glabrous (Figure 1I). Although the roots of plagiotropic individuals are exclusively terrestrial, the feeder roots of orthotropic individuals are born as aerial roots and develop in atmospheric conditions for several meters until penetrating the soil, acquiring a contiguous terrestrial phase (Figure 1F). During the terrestrial phase, the feeder roots became smooth in surface texture, larger and without lenticels. As the color, shape, and size of the feeder roots change as long as E. aureum individuals increase in height (Figure 1), we compared the morphology, anatomy, and physiology of the roots of E. aureum under five developmental conditions: (i) terrestrial roots of plagiotropic individuals, (ii) aerial feeding roots of orthotropic individuals growing at 1.5 m, (iii) its contiguous terrestrial phase, (iv) aerial feeding roots of orthotropic individuals growing at 6 m in height, and (v) its contiguous terrestrial phase.
Morphology of Epipremnum aureum during soil to canopy transition under field conditions. (A) Plagiotropic individual showing terrestrial roots. (B) One orthotropic individual initiated ascent toward the canopy. This individual developed a new aerial feeder root (arrow), a structure that establishes a connection between the shoot positioned on the host trunk and the soil below. (C) Aerial roots from an orthotropic individual 1.5 m in height, detailing small anchor roots () and one feeder root with a manually peeled window exposing the green cortex (arrow). Lenticels are rare in this stage. (D) Habitus of orthotropic individuals 1.5 m in height. (E) An orthotropic individual ascending the canopy at 6 m in height. Note the thick aerial roots covered by lenticels (arrow). (F) Aerial feeder root establishing soil contact (dashed line indicates excavated soil), constituting the terrestrial phase of the aerial root. Note how lenticels present in the aerial phase (arrow) disappear while the root thickens in the contiguous terrestrial phase (arrowhead). (G) Leaf area increase comparing plagiotropic (smaller leaf) to orthotropic individuals 6 m in height (larger leaf). (H) Aerial root system from an orthotropic individual 6 m in height, detailing anchor roots () and one feeder root with a peeled window exposing the green cortex (arrow). Note abundant lenticels. (I) Aerial feeder root from an orthotropic individual 1.5 m in height, showing root hairs only on the host‐contact side (arrowhead). The arrow indicates a peeled window. Scale bars: (A, B, D, G) (=10 cm); (C, E, F, H) (=10 mm), I (=5 cm).
Root Morphology
2.2
The roots of plagiotropic and orthotropic individuals were analyzed regarding their respective morphology, anatomy, and physiology (Figure 2). In the field, 18 roots were chosen from 18 individuals, with 1 root per individual. Using razor blades, these roots were sectioned near their junction with the stem, packed in dark plastic bags and immediately taken to the laboratory. Of these, six represented the terrestrial phase of plagiotropic individuals, six represented orthotropic individuals 1.5 m in height, and six represented orthotropic individuals 6 m in height. In the laboratory, the roots were cleaned with distilled water to remove soil and other debris. For the roots of orthotropic individuals (both 1.5 and 6 m in height), each collected root was divided into its aerial and contiguous terrestrial phase, that is, from the same specimen, the aerial root that penetrated the soil and became subterranean. This division was always performed far from the apical meristem to obtain root specimens in their mature form. Thus, five root types were compared: plagiotropic terrestrial phase, orthotropic aerial phase at 1.5 m and its terrestrial phase, and orthotropic aerial phase at 6 m and its terrestrial phase. A total of six samples were obtained per root type, totaling 30 samples (five root types × six samples). The final objective was to create two groups of roots for each root sample collected: one group with the original covering tissue (called intact) and one group without the covering tissue (called abraded; Figure 2). To do this, from each of the 30 root samples, two contiguous sections of identical length around 5 cm were obtained. One half was left intact, and the other was superficially abraded (Supporting Information S1, video). After testing different materials, a small metallic coin was identified as the most suitable tool for abrading the roots of plagiotropic and orthotropic individuals 1.5 m in height. For the roots of orthotropic individuals 6 m in height, 80‐mesh sandpaper was chosen due to the greater superficial mechanical resistance. Observations under a stereoscope and microscope visually identified the effectiveness of the technique (Supporting Information S2), where the covering tissue was removed without intense damage to the cortical tissues (Figure 2). From each root sample, another root sub‐sample that was 5 cm long was preserved in 70% ethanol for anatomical analysis.
Schematic diagram explaining the experimental protocol used to analyze root morphophysiological changes during the habitat transition from the soil to canopy of the aroid vine Epipremnum aureum . Note that aerial roots of orthotropic individuals were studied during the ‘aerial phase’ (i.e., atmospheric) and ‘terrestrial phase’ (i.e., subterranean phase of aerial roots after penetrating below soil). ‘Abraded’ roots mean intact root specimens that had the original covering detached using the abrasion technique (Supporting Information S1). In graphs presented throughout the article, terrestrial and aerial roots are consistently color‐coded in brown and green, respectively. Arrows indicate the statistical design represented by paired comparisons—red arrows show comparisons between intact × abraded roots and aerial × terrestrial roots of orthotropic individuals, while blue arrows show all groups comparisons. Plant groups increase in size but are not drawn to scale.
Anatomy
2.3
Cross‐sections from each root specimen were obtained freehand from the material preserved in 70% ethanol. Sections obtained by razor blades were clarified with standard techniques and then stained using a combined Safranin and Astra Blue staining solution (Johansen 1940; Ruzin 1999). The material was then photographed under a BX 50 optical microscope (Olympus) and under a binocular SZ61 stereomicroscope with micrometric ruler (Olympus) for larger samples to quantify the following parameters: external diameter, cortex width, stele diameter, covering tissue thickness and diameter of the five largest xylem vessels (Figure 2) using the Image Pro‐Plus program (Media Cybernetics).
Histochemical tests were performed on transverse root sections analyzed by optical and fluorescence microscopy (Ruzin 1999; Piccinini et al. 2024). For optical microscopy, lignin was detected by its red coloration in safranin‐astra blue staining (Vazquez‐Cooz and Meyer 2002) and confirmed with 2% phloroglucinol (Jensen 1962). Suberin deposits were identified by red staining with Sudan IV 0.1% and Sudan Red 7B 0.1% in ethanol 50% for 30 min (Brundrett et al. 1998). For visualization of the Casparian bands in the hypodermis, samples were stained with 0.1% Berberine Hemisulfate and counterstained with 0.5% Aniline Blue for 30 min before being observed under fluorescence microscopy (Brundrett et al. 1988). Lignin was identified with 0.1% Basic Fuchsin (BF) in 50% ethanol (30 min; Sexauer et al. 2021), suberin lamellae with 0.01% Fluorol Yellow 088 (FY 088) in 100% ethanol (10 min; Yamauchi et al. 2025), and cellulose walls with 0.01% Calcofluor White (CW, 10 min; Ruzin 1999). Co‐localization of the results obtained by FY 088 + BF (Kitin et al. 2020) was visualized by overlaying isolated images obtained from the two different filter sets below. Triple staining was performed according to established methods (Ursache et al. 2018; Lux et al. 2005) using the sequential application BF‐FY‐CW (Sexauer et al. 2021) with thorough washing in 50% ethanol between each step (Yamauchi et al. 2025). Unless otherwise specified above, all dye solutions were prepared in distilled water at weight/volume (w/v) concentrations. Sections were analyzed using an Olympus BX50 epifluorescence microscope equipped with two filter sets. The Ultraviolet (UV) filter set (U‐MWU2; Excitation = 330–385, Dicroic mirror = 400, Emission = LP 420) was used to visualize Berberine Hemisulphate, Fluorol Yellow 088 (FY), and Calcofluor White, while the Green filter set (U‐MWG2; EX = 510–550, DM = 570 nm, EM = LP 590) was used for Basic Fuchsin (BF). Photomicrographs were obtained with a Coolsnap digital camera. Cell and tissue classification followed Schreiber and Franke (2011) for exodermis and Keating (2002) and French (1987, 1997) for nomenclature in Araceae.
Root Water Loss, Succulence, and Density
2.4
Root water loss was measured by gravimetry (Taleisnik et al. 1999; Filartiga et al. 2021) along transpiration kinetic curves (Suresh et al. 2022). Prior to the measurements, the collected root samples were first hydrated in closed Petri dishes with moistened filter papers at 7°C overnight. They were then superficially dried with filter paper and weighed on a precision balance (Mettler Toledo, 0.1 mg precision) to obtain the maximum fresh weight (MFW). A thin film of petroleum jelly was applied to the cut ends of the excised root segments to ensure water loss prevention (Perumalla et al. 1990; Seago Jr. et al. 2000; Soukup et al. 2002), and the host contact region was placed in tight contact with glass slides, leaving the root side originally exposed to the atmosphere free. Each root sample was placed on an individual glass slide and was not touched during the experiment. Roots were then dried on the laboratory bench to determine their fresh weight (FW) at 4‐h intervals in a 24‐h period. The air temperature and humidity during this period were respectively 25°C–27°C and 70%–75%, determined using a digital thermohygrometer MTH‐1300 (Minipa, Brazil, 1°C and 5% precision). After the 24‐h period, the samples were dried in an oven at 60°C until a constant weight (DW). The root water loss (RWL, %) at each interval was expressed as a percentage using the formula: ((MFW − FW) × 100)/(MFW − DW) (Mantovani 1999).
Each root sample was considered as a cylinder, allowing the calculation of its lateral surface area (A = 2 × 3.14 × radius × length; m^2^), volume (V = 3.14 × radius^2^ × length; m^3^), and surface‐to‐volume ratio (S/V; m^−1^). These measurements enabled the determination of succulence and root mass per area (RMA) using the following formulas: Succulence = (MFW—DW)/A (g m^−2^); RMA = DW/A (g m^−2^) (Mantovani 1999).
Root Permeance
2.5
Using the transpiration kinetic curves described above, permeance (P) was calculated according to the equation P = F (A × Δc)^−1^ (Suresh et al. 2022). Here, F represents the water loss (g s^−1^) during the first 4 h period, A is the lateral surface area (m^2^), and Δc is the biophysical driving force for transpiration (g m^−3^). The 4 h period was chosen because terrestrial and abraded roots lost 30%–40% of their initial water content within this time, a critical dehydration threshold associated with membrane damage and loss of rehydration capacity (Trifilò et al. 2023). Moreover, transpiration curves exhibited an asymptotic behavior, with a progressive reduction in slope after 4 h in terrestrial and abraded roots and after 8–12 h in most other root types. This early time point provides a basis for comparison across root types and with other studies, for example, the 6 h experiment in Suresh et al. (2022), and aligns with physiologically meaningful dehydration limits.
Considering a liquid phase (root cortex surrounded by a compact hypodermal‐epidermal structure, 23°C) to vapor phase (external air, 23°C and 73% RH) gradient (Grünhofer and Schreiber 2023), a water‐based Δc equals ρliquid (9.97 × 10^5^ g m^−3^)—ρvapor (15.15 g m^−3^) ≈9.96 × 10^5^ g m^−3^. To enable direct comparison with previously published water permeance data for Monstera deliciosa aerial roots (Suresh et al. 2022), obtained using a different technique that employs enzymatically isolated covering tissue to quantify water loss, we also measured the water permeance of intact aerial and contiguous terrestrial roots of M. deliciosa from the same study area used for E. aureum .
Water Absorption Capacity
2.6
The water absorption capacity of root hairs from host‐attached aerial roots of E. aureum was investigated using the fluorescent apoplastic tracer Lucifer Yellow carboxy hydrazide (LYCH; 0.5 μg ml^−1^; Bederska et al. 2012; Filartiga et al. 2021). Two 1‐m‐long root specimens were collected from hosts with detachable bark and immediately transported to the laboratory. From each specimen, two contiguous 10‐cm segments were prepared. The extremities of all segments were sealed with petroleum jelly to prevent tracer access via cut ends (Soukup et al. 2002). Only a 1 cm length of the aerial root side with hairs and bark connections intact was placed in contact with the fluorescent tracer. Subsequently, one segment was exposed to the fluorescent tracer while the other served as a control (distilled water under the same conditions), both for 60 min. After this period, the root pieces were washed with distilled water, hand‐sectioned, and analyzed by confocal microscopy (405 nm excitation, 520–550 nm emission; Burrows et al. 2013).
Chlorophyll Fluorescence
2.7
Rapid light curves (RLCs) of aerial feeder roots from six E. aureum individuals were generated in situ using a pulse‐amplitude‐modulated fluorometer (MINI‐PAM; Walz) to relate apparent electron transport rate (ETR) to light intensity. Following established protocols (Chomicki et al. 2014; Griffiths et al. 2008), light‐adapted roots were exposed to eight 10‐s intervals of increasing light (PPFD). At each step, a saturating pulse was applied to calculate the effective quantum yield (ΦPSII) as (Fm′ − F)/Fm′ (Genty et al. 1989). The ETR was derived as ΦPSII × PPFD × 0.5 × 0.93 (White and Critchley 1999), where the ETR factor of 0.93 was adopted from a study on similar non‐leaf tissues instead of the standard leaf value of 0.84 (Filartiga et al. 2021).
Statistical Analysis
2.8
The roots of plagiotropic and orthotropic individuals were compared for morpho‐physiological traits. Comparative analyses were primarily conducted using one‐way analysis of variance (ANOVA; Zar 1996). Due to its integrative capacity for representing the combined effects of root size and covering tissues on water loss, a two‐way ANOVA was applied to the permeance data obtained from plagiotropic and orthotropic roots. This test was designed to evaluate whether a multifactorial interaction between growth orientation, size classes (categorized as plagiotropic, orthotropic 1.5 m in height, and orthotropic 6 m in height), and the status of covering tissues (intact vs. abraded) significantly affected the root water loss in E. aureum during its ascent into the canopy. After that, Tukey's post hoc test with Bonferroni correction was applied to normally distributed data, while the Kruskal‐Wallis test was used for non‐parametric data (Zar 1996).
Finally, the paired Student's t‐test was used to evaluate the effect of abrasion on external root diameter, as size reduction could influence water loss and permeance. The same paired analysis was also applied to compare aerial feeder roots with their contiguous terrestrial phases regarding changes in water loss. Given the asymptotic nature of root water loss curves, values at both 12 and 24 h were compared for each root condition (aerial vs. terrestrial; intact vs. abraded). Paired analysis was also applied to the corresponding permeance values following abrasion. The Jamovi software (2.3.28; The Jamovi Project 2025) was used for this purpose at a significance level of p < 0.05 (Zar 1996).
Results
3
Root Morphology and Anatomy
3.1
In cross‐sections, the abrasion technique resulted in gentle excoriation of the root surface without destroying the cortex (Figure 3A–D), which was confirmed by qualitative and quantitative anatomy (Supporting Information S2, Table 1). Mature terrestrial roots of plagiotropic individuals presented a simple monolayered epidermis with trichomes covering all root surfaces (Supporting Information S2). Below the epidermis, a uniseriate exodermis was present, characterized by slightly thickened anticlinal and external periclinal walls (Figure 4). The exodermis and an epidermis with trichomes were present in the terrestrial roots of plagiotropic individuals, in the aerial roots of orthotropic individuals that were in contact with the host (Figure 3E,F), and in aerial roots that had entered the terrestrial phase after penetrating the forest soil (Supporting Information S2).
Morphology and anatomy of Epipremnum aureum roots during habitat transition. (A, B) Intact aerial root showing dark brown surface with lenticels (white arrow). (B) Anatomy of intact aerial root confirms a sclerified cork (black arrow) as the covering tissue. (C) Morphology of the root shown in (A) after abrasion, demonstrating technique efficacy. (D) Anatomy of the root shown in (B); abrasion preserves meristematic cells (). (E) Transverse section of a feeder root. Profuse root hairs are restricted to the host contact region (HC). On the ‘atmospheric’ root region (ATM, black arrow), the outer cortex with sclereids forms the new covering tissue. The dashed line delimits the transition zone between the host‐contact region and the atmospheric region. (F) Detail of the one‐layer exodermis covered by an epidermis with trichomes. In aerial roots, trichomes are restricted to host‐contact zones (yellow arrow), whereas they encircle terrestrial roots entirely. (G) Detail of the ATM–HC transition zone, showing the replacement of the trichome‐bearing epidermis and exodermis by a sclerified cork (red arrow). (H) Detail of the sclereids distributed along a one‐layered tissue (black arrow). Debris of the exodermis and cortical tissues are seen above the sclerified cork (). (I) A two‐layered sclerified cork covering an aerial root. The inner layer compacts the outer meristematic cells, demonstrating continuous production of this new covering. Yellow bars indicate the xylem vessel diameter (E) and the covering tissue thickness (F, H, I). Scale bars: (A, C, E) (=2 mm); (B, D, F–I) (= 100 μm).
Fluorescence microscopy of Epipremnum aureum roots along a soil‐to‐canopy transition. Terrestrial roots (T; A–D) and aerial roots in host contact (HC; E–H) and atmospheric (ATM; I‐L) regions are detailed. Berberine‐Aniline Blue staining (A, E) indicate (arrows) Casparian strips in exodermis of terrestrial roots and host contact regions of aerial roots. The presence of Casparian bands are inconclusive in the sclerified cork cells (arrowhead) facing atmospheric region (I). Fluorol Yellow 088 (FY 088) identifies suberin lamellae in the exodermis (B–F) and cork cells (J). Basic Fuchsin (BF) identifies lignin in the epidermis‐exodermis (C, G) and sclereids (K). Co‐localization of ligno‐suberized structures (FY 088 + BF) is shown (D, H). (L) Triple staining (FY 088 + BF + Calcofluor White) also identifies the cellulosic cell walls beneath the ligno‐suberized structures, aiding in the spatial localization of the cortex in the other figures. Scale bar = 20 μm.
Below the exodermis, terrestrial and aerial roots differed in cortex anatomy. Near the apex of aerial feeder roots, the epidermis and exodermis were unique covering tissues present just after maturation of the protodermis (Figure 3E). However, it was soon replaced by a new covering tissue, but only in the root region exposed to the atmosphere, that is, not adhering to the host: Figure 3E,G show a meristematic tissue with a few sclereids appearing below the exodermis. This meristematic tissue produced sclereids in the outer region in one to three layers that soon shed the epidermis plus exodermis (Figure 3H,I). Consequently, the aerial feeder roots exposed to the atmosphere were covered by a new tissue, hereinafter called the sclerified cork. At the same root level but in the root region in contact with the host, this did not occur, as no meristematic tissue and sclerified cork appeared. The root region in contact with the host maintained the epidermis with trichomes and the exodermis (Figure 3E). Consequently, a dimorphism between the host region with exodermis and the aerial region with sclereids was always observed in aerial roots under atmospheric conditions (Figure 3E–G). The thickness of the root covering tissue increased 4‐fold along the vertical ascent of E. aureum , ranging from 35 to over 145 μm (Figure 3H,I; Table 1). No statistical differences were detected for this parameter when comparing the root exodermis of plagiotropic and orthotropic (aerial roots in terrestrial phase) individuals, although it is thinner than the sclerified cork of orthotropic individuals (Table 1). Along the soil to canopy transition, the external root diameter of E. aureum increased 3–4 times from the plagiotropic to larger orthotropic individuals, reaching around 6–8 mm. This increase was maintained after root abrasion. Both the cortex and the stele also increased in a similar manner but in different proportions (Table 1). Comparing terrestrial roots of plagiotropic individuals with aerial feeder roots of orthotropic individuals 6 m in height, the cortex only increased around 2‐fold compared to the 5‐fold increase for the stele (Supporting Information S3). However, when aerial and the terrestrial phases of the same aerial feeder roots were compared, the greater diameter increase found during the terrestrial phase was generated by an increase in the cortex instead of the stele dimensions (Supporting Information S3).
The largest xylem vessels increased in diameter as the external root diameter increased, ranging from about 70 μm in plagiotropic individuals to over 200 μm in orthotropic individuals at 6 m height. In other words, along the soil to canopy transition, the roots of E. aureum increased in size, generating a greater capacity for storing and transporting water to leaves (Table 1).
Regardless of plant orientation (plagiotropic or orthotropic) or growth phase (terrestrial or aerial), histochemical tests revealed a consistent anatomical pattern: the presence of a ligno‐suberized covering tissue on the roots (Figure 4A–L). Identified by Casparian bands (Figure 4A), a monolayered exodermis was present in terrestrial roots and in aerial roots that were in contact with the host (Figure 4E). In the root region exposed to the atmosphere, the presence of Casparian bands in cork cells was inconclusive (Figure 4I). The exodermal cells were characterized by thin anticlinal and outer periclinal walls covered by suberin lamellae (Figure 4B,F). Lignin is also present in all cell walls of the exodermis, being less evident in the internal periclinal wall (Figure 4C,G). A thick lignified cap detected by histochemical tests in optical microscopy (Supporting Information S4) becomes clearer when compared with Figure 4B–D.
The same ligno‐suberized composition was observed in the sclerified cork exposed to the atmosphere, but with a different tissue arrangement. The meristematic cells (Figure 4I–L) had thin, suberized walls (Figure 4J), while the thick, lignified sclereids were located above them (Figure 4K). This ligno‐suberized pattern was confirmed by double and triple fluorescence staining (Figure 4D,H,L) and by histochemical tests using optical microscopy (Supporting Information S4).
Root Water Loss, Succulence, and Density
3.2
Root water loss in Epipremnum aureum decreased significantly as plants ascended host trees (Figure 5). This transition occurred abruptly upon the initial meter of ascent in orthotropic individuals, with no further significant changes observed at greater heights. Intact terrestrial roots of plagiotropic individuals exhibited high water loss with RWL values reaching 75%–85% within 24 h (Figure 5A). In contrast, intact aerial roots of orthotropic individuals consistently showed low RWL (~30%), independently of plant height (Figure 5B,D).
Root water loss in Epipremnum aureum along a soil‐to‐canopy transition. (A) Terrestrial roots of plagiotropic individuals. (B‐C) Orthotropic individuals 1.5 m height: Aerial feeder roots (B) and their contiguous terrestrial phase (C). (D–E) Orthotropic individuals 6 m height: Aerial feeder roots (D) and their contiguous terrestrial phase (E). Blue stars indicate significant differences in 12 and 24 h RWL values between intact roots (closed symbols) and abraded roots (open symbols) within each panel. Arrows (closed for intact roots, open for abraded roots) denote significant differences in 12 and 24 h RWL values between aerial and corresponding terrestrial phases of feeder roots in orthotropic individuals at 1.5 m (B vs. C) and 6 m (D vs. E) in height. Green lines represent aerial root phases; brown lines represent terrestrial phases. Data are presented as mean ± standard deviation. All symbols indicate statistical significance (p < 0.05) based on paired Tukey's test. Data are the mean ± standard deviation (n = 6).
A similar gradient was observed among terrestrial roots: RWL after 24 h decreased from 83% in plagiotropic individuals (Figure 5A) to 60% and 30% in orthotropic individuals at 1.5 m and 6 m in height, respectively (Figure 5C,E). When comparing intact aerial (Figure 5B) and terrestrial phases (Figure 5C) within the same orthotropic individuals 1.5 m in height, water loss was two times higher in terrestrial roots. This difference was absent in 6 m tall individuals, indicating that maximum resistance in aerial roots was achieved early during ascent, whereas resistance in terrestrial roots continued to increase with plant height (Figure 5D,E).
Removal of the root covering by abrasion led to a rapid and significant increase in water loss. In plagiotropic individuals, abraded roots reached 98% RWL within 12 h, compared to 68% in intact roots (Figure 5A). A similar pattern was observed in orthotropic roots, regardless of developmental phase. While intact aerial roots maintained low RWL (~30%) after 24 h, abrasion caused a marked increase (Figure 5B,C): abraded aerial roots of 1.5 and 6 m tall plants reached ~92% and ~70% RWL at only 12 h, respectively, converging to 97% after 24 h. A notable exception was observed in abraded aerial roots from the terrestrial phase of 6 m tall orthotropic individuals (Figure 5E). Although these roots showed higher RWL than their intact counterparts (54% vs. 36%), their final water loss remained considerably lower than that of other abraded roots (~97%).
Root succulence, mass per area (RMA), and surface‐to‐volume ratio (S/V) varied significantly in Epipremnum aureum roots across the soil‐to‐canopy transition (Figure 6). Aerial roots of orthotropic individuals showed higher succulence and RMA, alongside lower S/V ratios than terrestrial roots of plagiotropic individuals. These results indicate that E. aureum roots develop greater water storage capacity and a thicker structure while reducing their surface‐to‐volume ratio during canopy ascent.
Changes in succulence, root mass per area (RMA), and surface‐to‐volume ratio (S/V) in Epipremnum aureum roots along a soil‐to‐canopy transition. Terrestrial and aerial roots from plagiotropic and orthotropic individuals are compared. Lowercase letters above bars and data points indicate significant statistical differences (p < 0.05) according to one‐way ANOVA followed by Tukey's test. Data are the mean ± standard deviation (n = 6).
Root Permeance
3.3
Root permeance was significantly influenced by the interaction of multiple factors (Table 2). Although the individual factors (growth axis, size, and tissue abrasion) showed significant effects, their interaction was significant (F 4, 50 = 9.43, p < 0.001). This demonstrates that the effect of tissue abrasion on root permeance in Epipremnum aureum is not uniform, but is substantially modulated by the growth axis (plagiotropic versus orthotropic). The post hoc analysis of the two‐way ANOVA indicates that the permeance of intact aerial roots (from 3 to 5 × 10^−9^ m s^−1^) of Epipremnum aureum is significantly lower than that of respective terrestrial roots (around 10^−8^ m s^−1^; Figure 7). For comparison, the permeance of aerial roots (1.5 to 2.6 × 10^−9^ m s^−1^) and its contiguous terrestrial phase (6.2 to 7.8 × 10^−9^ m s^−1^) of M. deliciosa followed the same pattern. This difference for E. aureum is likely induced by the development of the new covering tissue (sclerified cork) in aerial roots, because whenever this tissue is absent, either in terrestrial roots or in abraded roots, permeance remains unchanged. No significant correlation was observed between the thickness of the covering tissue and root permeance for E. aureum (Supporting Information S5).
Changes in water loss permeance of Epipremnum aureum roots along a soil‐to‐canopy transition. Red stars indicate significant differences (paired t‐test, p < 0.05) between intact and abraded roots, found only in aerial roots. Blue (Tukey's test, p < 0.05) and red (paired t‐test, p < 0.05) arrows denote significant differences among intact terrestrial and aerial roots of plagiotropic and orthotropic individuals. For comparison, permeance values of terrestrial (brown arrow) and aerial (green arrow) roots of Monstera deliciosa are shown at left, confirming the pattern of lower permeance in aerial roots (Box plot data, n = 6).
Water Absorption
3.4
The results of the LYCH experiment are shown in Figure 8. Control samples without fluorochrome application exhibited only the autofluorescence of suberin and lignin. After 60 min, the green–yellow fluorescent water tracer was clearly detected in epidermal, exodermal, and cortical cells, and had also reached the stele.
Water absorption by the aerial roots of Epipremnum aureum , as evidenced by LYCH application using confocal microscopy. (A, B) Control specimen without fluorochrome. (C, D) Specimen with fluorochrome. (A) Host‐connected side showing autofluorescence emitted by trichomes and the exodermis. (B) Detail of the exodermis; note the absence of fluorochrome impregnation in the cortical cells. (C) The host‐connected side shows intense fluorochrome impregnation in the host bark, root trichomes, exodermis, cortical areas (asterisk), and stele. Note cortical areas without (arrow) fluorochrome impregnation. (D) Detailed view of the host‐connected side. Cortical cells with (asterisk) and without (arrow) fluorochrome impregnation are shown. Scale bar = 1 mm (A, C); 20 μm (B, D).
Chlorophyll Fluorescence Analysis
3.5
Maximum ETR values for the aerial roots of E. aureum were around 23.5 ± 10.1 μmol m^−2^ s^−1^ (mean ± standard deviation) obtained with 500–600 μmol m^−2^ s^−1^ of PPFD (Figure 9). Three of the six root samples studied showed a decrease in ETR values just after the peak value was reached (raw data available as Supporting Information S6). For comparison, one small leaf of a terrestrial plagiotropic individual and one large leaf of an orthotropic individual of E. aureum were evaluated to determine the maximum ETR values. The peak ETR value for small leaves was 44.3 μmol m^−2^ s^−1^ obtained at 500 μmol m^−2^ s^−1^ of PPFD, and for the larger leaf, the peak ETR value was 58.8 μmol m^−2^ s^−1^ obtained at 1000 μmol m^−2^ s^−1^ of PPFD.
Electron transport rates (ETR) of Epipremnum aureum aerial roots obtained through light response curves using a chlorophyll fluorescence meter. (A) One E. aureum aerial root (closed triangle) is compared with one small (plagiotropic individual; closed circle) and one large leaf (orthotropic individual at 1.5 m in height; open circle). (B) Auto‐fluorescence of an aerial root exposed to the atmospheric region. Note the sclerified cork above chloroplasts positioned in the outer cortex. Scale bar = 20 μm.
Discussion
4
This study investigated the morphophysiological changes of E. aureum roots during the transition from the terrestrial habitat to the canopy. Specifically, we tested the hypothesis that root covering tissue is functionally modified in its structure to decrease water loss in the drier aerial environment. Our key findings confirm this hypothesis: we demonstrated a fundamental shift in the root surface tissue, from a less protective exodermis in terrestrial roots to a sclerified covering tissue in aerial roots. This structural modification was functionally significant, reducing root water loss and confirming its importance for the survival of aroid vines in the canopy.
Exodermis is a uni‐ to multi‐stratified hypodermis (Peterson and Perumalla 1990; Meyer et al. 2011) characterized by Caspary bands (Evert 2006; Schreiber and Franke 2011). In aroids, an exodermis has been detected in Anthurium, Arisaema, Dieffenbachia, Philodendron, and Syngonium (Perumalla et al. 1990) and now in the Epipremnum genus. Due to the main deposition of lignin and suberin over the outer periclinal and anticlinall walls, the exodermis in E. aureum resembles the polar thickening type (sensu Cantó‐Pastor et al. 2025). Ligno‐suberized thickenings of the exodermis can decrease hydraulic apoplastic conductance to the cortex (North and Nobel 1998; Grünhofer and Schreiber 2023). While evidence on the water permeability of the exodermis is sometimes controversial at the species level (Taleisnik et al. 1999), many studies support its role in protecting roots against water loss to the soil (Enstone et al. 2003; Liu and Kreszies 2023; Cantó‐Pastor et al. 2025). Under drought conditions, the role of the exodermis in limiting water loss can be reinforced by increased suberization (Perumalla and Peterson 1985; Shiono and Matsuura 2024) or by the co‐occurrence of a subjacent hypodermis (French 1987; Lima et al. 2024). When these improvements are not enough to prevent excessive water dehydration, loss of peripheral tissue and replacement by a more efficient tissue occurs.
The loss of peripheral root tissue in angiosperms may be limited to the epidermis (McKenzie and Peterson 1995), making the exodermis a superficial tissue (Reinhardt and Rost 1995; Yang et al. 2014), or it may occur internally, leading to partial (Stevens et al. 1997; Meyer et al. 2011) or even total loss of the cortex when the endoderm is kept as the covering tissue (Kauff et al. 2000). Root tissue losses are commonly replaced by a more resistant tissue in the inner part of the cortex (Strock and Lynch 2020; Lynch et al. 2021). This is the case for the aerial roots of E. aureum under the atmospheric conditions of the canopy, in which a higher efficiency in limiting water loss is necessary.
Aroid vines experience physiological challenges during the transition from soil to canopy habitats, where their leaves, stems, and roots are exposed to markedly different microenvironmental conditions (Freiberg 1997; Mantovani 1999). Among these constraints, drought at canopies is one of the most limiting factors (Zotz 2016). This constraint is partially mitigated for the aroid vines by the production of aerial roots, which originate under atmospheric conditions yet maintain hydraulic continuity several meters below with the forest soil (Zotz et al. 2021). Although this reconnection guarantees access to soil water and nutrients, the maintenance of roots in the atmosphere demands specialized structural and physiological adjustments.
Grünhofer and Schreiber (2023) emphasized the severity of atmospheric exposure by showing that roots and leaves subjected to typical canopy conditions (25°C, 30% RH) experience a water potential gradient around −160 MPa, over two orders of magnitude steeper than the −1.5 MPa associated with the permanent wilting point of plants growing in soil. Roots developing under such extreme gradients must therefore evolve distinct morpho‐physiological strategies to avoid dehydration, as root original configurations optimized for soil environments are inadequate for this purpose (Filartiga et al. 2021).
This necessity is evident in the adjustments described by Suresh et al. (2022) for aerial roots, including those of Monstera deliciosa . Using enzymatically isolated covering tissues, the authors demonstrated that the capacity of aerial roots to limit water loss can approach that of leaves. The periderm of M. deliciosa exhibits permeance values between 0.03 and 1.27 × 10^−9^ m s^−1^, comparable to leaf cuticular permeance and far lower than those of a barrier‐free surface (74 ± 5.6 × 10^−9^ m s^−1^). Whole‐root measurements in the present study revealed similarly low permeance values (minimum of 1.5 × 10^−9^ m s^−1^) for aerial roots of M. deliciosa . Suresh et al. (2022) also showed that in the herbaceous species Clivia miniata (Lindl.) Regel, aerial roots exhibit permeance values three to four times lower than terrestrial roots (16 vs. 50 × 10^−9^ m s^−1^), indicating that only the terrestrial roots approach freely transpiring conditions. A comparable pattern was observed in the aroid vines studied here: aerial roots of E. aureum and M. deliciosa exhibited permeance values three to four times lower than their terrestrial counterparts.
A key insight from Suresh et al. (2022) is that wax molecules embedded within suberin constitute the primary restriction to root water loss. Upon wax extraction using chloroform, permeance increased substantially—five‐fold in M. deliciosa and 54‐fold in C. miniata —bringing aerial root permeance values close to those of a barrier‐free surface. However, no effect was detected on the permeance of terrestrial roots in C. miniata . This response parallels the effect of abrasion in the present study, an effect that removed the covering tissue and increased permeance values of aerial roots up to fourfold, while leaving terrestrial roots unaffected.
Together, these observations indicate that exposure to the atmosphere induces rapid morpho‐physiological reactions by the aroid aerial roots. In E. aureum , the abrupt decline in permeance shortly after individuals begin climbing suggests early establishment of a very efficient barrier to water loss, without further reductions at greater heights, even though conditions become progressively drier toward the upper canopy (Mantovani et al. 2017). The role of wax sorption into suberin in generating this reduction in aerial roots of E. aureum merits further investigation. Nonetheless, our data provide an initial indication: permeance remained constant despite a threefold increase in covering tissue thickness (from 60 to 170 μm in aerial roots of orthotropic individuals at 1.5 and 6 m height, respectively), implying that structural thickening alone does not drive permeability changes.
The loss of the original covering tissue has been documented in a few Araceae genera (e.g., Monstera; Hinchee 1981, Philodendron; Ferreira et al. 2020, and Rhodospatha; Filartiga et al. 2021). Nevertheless, its functional significance has not been directly assessed in aerial aroid roots (French 1997). By evaluating intact aerial roots and their corresponding terrestrial counterparts in R. oblongata, Filartiga et al. (2021) demonstrated that aerial roots covered by a newly formed, thicker ligno‐suberized envelope displayed greater resistance to desiccation than terrestrial roots retaining only a thin exodermis. In E. aureum , this newly developed protective layer consists of robust layers of sclereids, whereas in R. oblongata it is composed of tabular cells with markedly thickened outer transverse walls (Filartiga et al. 2021) and in M. deliciosa by a periderm‐like structure (Suresh et al. 2022). This structural divergence indicates that aerial roots of aroid vines use distinct covering tissues that converge functionally to reduce water loss, reinforcing the phylogenetic and taxonomic relevance of this morphophysiological trait within the family (French 1987; Keating 2002; Croat and Ortiz 2020).
The replacement of the original covering tissue in R. oblongata (Filartiga et al. 2021) and E. aureum , represented by a ligno‐suberized layer, occurs exclusively on the root atmospheric region and does not develop on the host‐contact region, where the original exodermis plus epidermis and trichomes are retained. This results in a functional dimorphism within the same root level, consistently characterized by a ligno‐suberized protective structure but exhibiting distinct cell types, sizes, and organization. Comparable within‐root dimorphism induced by microenvironmental constraints between the host‐contact and atmospheric regions has been documented in Orchidaceae (Chomicki et al. 2014) and Cyclanthaceae (Wilder 2011), but only recently reported for Araceae by Filartiga et al. (2021). In both R. oblongata and E. aureum , the root region exposed to air limits water loss, whereas the host‐contacting surface remains capable of water uptake. Evidently, each process—water inflow and outflow—is tightly regulated by structural mechanisms, including the activity of water pores (Cruz et al. 1992; Steudle et al. 1993; Taleisnik et al. 1999; Ranathunge et al. 2011; Bederska et al. 2012), the diffusion behavior of suberin within cell walls (molecular sieving in Grünhofer and Schreiber 2023), and aquaporin function (Grondin et al. 2016). Such regulation is crucial to facilitate water absorption during transpiring periods while preventing excessive loss under drought conditions (Zotz and Winkler 2013).
Lignin and suberin occur both in the sclerified cork of the atmospheric regions and in the exodermis in the host‐contact regions. A reduction in water inflow—without complete blockage—caused by ligno‐suberized apoplastic barriers has been described, for example in Agave deserti Engelm. and Helianthus annuus L. (Meyer et al. 2009), and also in Asclepiadaceae species (Perumalla et al. 1990). Consequently, E. aureum roots can take advantage of stemflow when adhering to the trunk, thereby optimizing water and nutrient acquisition, including enhanced phosphorus uptake (Benzing 1990). The improved water balance of R. oblongata (Filartiga et al. 2014) and E. aureum is reinforced by the high axial hydraulic efficiency of their aerial feeder roots while they grow toward the canopy, resembling the hydraulic capacity of lianas (Filartiga et al. 2018). In E. aureum, orthotropic individuals 6 m in height develop enlarged steles containing numerous wide xylem vessels—some exceeding 300 μm in diameter—thereby improving water transport. Conversely, the terrestrial portions of E. aureum roots tend to be thicker and more succulent than their aerial counterparts. The same pattern has been documented for the aroid vine R. oblongata (Filartiga et al. 2021), supporting the idea that the replacement of the original covering tissue alters the water relations (e.g., decreasing succulence) of the aerial roots.
Based on chlorophyll fluorescence data, the aerial roots of E. aureum individuals potentially perform photosynthesis (Natale et al. 2024). An improved carbon budget—induced by internal CO_2_ refixation or even net photosynthesis (Aschan and Pfanz 2003; Wittmann and Pfanz 2014) in aerial aroid roots—helps explain the fast growth of these structures in the canopy, for example, 29 cm/week in Monstera deliciosa (Hinchee 1981; Patiño et al. 1999; Meyer and Zotz 2004). However, the ETR values in the aerial roots of aroid vines are usually lower than those recorded in leaves (Mantovani et al. 2017; Filartiga et al. 2021), a pattern confirmed in the present study. Even so, low‐level photosynthesis here could be relevant to guarantee oxygen supply in the tight cortex. This role may be supported by ligno‐suberified covering tissue in E. aureum , which can limit O_2_ loss to the atmosphere (Aschan and Pfanz 2003; Lendzian 2006; Wittmann and Pfanz 2014), a mechanism also suggested for orchid roots (Brunello et al. 2024). Further investigations are needed to assess whether ligno‐suberized covering tissues help improve the metabolic functioning of aerial roots in aroid vines.
Light, gravity, and mechanical contact interact synergistically promoting a large increase in the size of E. aureum shoots and leaves along the soil‐to‐canopy transition (Steinitz et al. 1992; Brito et al. 2022). This size increase is accompanied by the production of more and thicker aerial roots (Filartiga et al. 2014), suggesting that these roots are also affected by multifactorial environmental interactions: either directly through the interaction between root photo‐ and gravitropism, or indirectly via root–shoot hormonal crosstalk (van Gelderen et al. 2018). The reorientation of the shoot axis from plagiotropic to orthotropic, together with mechanical host contact along the soil‐to‐canopy transition, modifies the functional morphology of roots in E. aureum . Two distinct aerial root types are produced: short, branched anchorage roots and long, unbranched feeder roots, orthogonally oriented despite originating from the same orthotropic shoot. Moreover, spatial functional dimorphism demonstrates that these roots are subject to different constraints even at the cellular scale. While the exodermis is maintained in the host‐contact region, it is replaced by a sclerified cork layer with reduced permeability under atmospheric drought (Mantovani et al. 2017). These adjustments in type, size, number, and morpho‐physiology illustrate that different stimuli and constraints are integrated into divergent functional morphologies within the roots of aroid vines. Given that aerial feeder roots are the main conduit for water and nutrient transport from forest soil to leaves in aroid vines, these changes are essential for the persistence of E. aureum in the canopy (Filartiga et al. 2018, 2021).
The distinctive traits of aroid vines—including large leaves, fast growth, high xylem hydraulic conductivity, drought‐resistant aerial roots, and ease of cultivation—combined with the fact that some aroid species are cultivated for food, for example, taro, yam, and so forth, (FAO 2003), further suggest their potential as model systems for genetic research and as promising candidates for vertical farming and urban agriculture. Together, these findings highlight the functional significance of root plasticity in aroid vines and point to their broader ecological and applied relevance.
Conclusions
5
The central hypothesis of this study, that root covering tissue is morphophysiologically modified to decrease water loss during the soil‐canopy transition, was confirmed. Our findings advance the understanding of aroid vine adaptation to the challenging canopy environment. We demonstrated a fundamental structural shift at the root level where the original exodermis is shed and replaced by a highly protective sclerified cork, notably without changing the original ligno‐suberized design. This structural change resulted in a marked reduction in permeance to water loss, a root functional trait influenced by a multifactorial interaction of stem orientation, environment, and plant size. We established a clear functional dimorphism in the roots: the sclerified cork prevents water loss in the atmospheric region, while the specialized, trichome‐rich area in contact with the host bark is able to absorb water from stem flow.
While these findings validate this root‐level adaptation as essential for the survival of E. aureum in the canopy, they open opportunities for further investigation on aroid vines. For example, future studies should deeply investigate the contribution of photosynthesis to aerial root maintenance; assess the absorptive capacity of the anchor roots; and test if similar functional structural changes in root anatomy are also present across the aerial roots of aroid vines, hemiepiphytes, and epiphytes. Ultimately, these investigations will provide a more comprehensive explanation of the morphophysiological improvements at the root level that facilitated the successful transition and survival of Araceae in the potentially stressful canopy environments.
Author Contributions
A.M. conceived and directed the research, designed the experiments, assisted with laboratory work, analyzed the data, wrote and revised the manuscript. Y.C.G. collected the material, coordinated the laboratory workflow, developed and implemented the abrasion technique used in this study, performed most of the experimental analyses, and discussed the data as part of graduate training. Both authors approved the manuscript.
Funding
The second author, Yago Chagas Groba, received a Scientific Initiation Scholarship (PIBIC–CNPq) from the Brazilian National Council for Scientific and Technological Development during the development of this study.
Disclosure
During the preparation of this manuscript, the authors used generative AI tools (ChatGPT) exclusively to improve the clarity and quality of the English language. The authors reviewed and revised the text and take full responsibility for the final content of the publication. Generative AI was not used to generate scientific ideas, influence the interpretation of results, modify figures or data visualizations, or alter the structure of the manuscript. No AI‐assisted technologies were used in the writing process.
Supporting information
Supporting Information: S1. Video with student demonstrating the abrasion technique. One aerial root specimen of an orthotropic individual 1.5 m in height was used.
Supporting Information: S2. Morphology and anatomy of intact and abraded roots of Epipremnum aureum .
Supporting Information: S3. Anatomy of the cortex and stele regions from the roots of Epipremnum aureum .
Supporting Information: S4. Histochemical detection of lignin and suberin.
Supporting Information: S5. Relationship between thickness of covering tissues and root permeance to water loss for Epipremnum aureum along a soil to canopy transition.
Data S1: Supplementary Information.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Aschan, G. , and H. Pfanz . 2003. “Non‐Foliar Photosynthesis—A Strategy of Additional Carbon Acquisition.” Flora 198: 81–97.
- 2Bederska, M. , W. Borucki , and E. Znojek . 2012. “Movement of Fluorescent Dyes Lucifer Yellow (LYCH) and Carboxyfluorescein (CF) in Medicago truncatula Gaertn. Roots and Root Nodules.” Symbiosis 58: 183–190.23482425 10.1007/s 13199-013-0221-7PMC 3589619 · doi ↗ · pubmed ↗
- 3Benzing, D. H. 1990. Vascular Epiphytes. General Biology and Related Biota. Cambridge University Press.
- 4Brito, C. , D. Mantuano , K. L. G. De Toni , and A. Mantovani . 2022. “Untangling Leaf Expansion Triggers: A New Experimental Study With Epipremnum aureum (Araceae).” Flora 295: 152139.
- 5Brito, C. , D. Mantuano , K. L. G. De Toni , and A. Mantovani . 2025. “Increasing Leaf Sizes of the Vine Epipremnum aureum (Araceae): Photosynthesis and Respiration.” Peer J 13: e 19214.40196307 10.7717/peerj.19214 PMC 11974542 · doi ↗ · pubmed ↗
- 6Brundrett, M. C. , D. E. Enstone , and C. A. Peterson . 1988. “A Berberine–Aniline Blue Fluorescent Staining Procedure for Suberin, Lignin, and Callose in Plant Tissue.” Protoplasma 146: 133–142.
- 7Brundrett, M. C. , B. Kendrick , and C. A. Peterson . 1998. “Efficient Lipid Staining in Plant Material With Sudan Red 7B or Fluoral Yellow 088 in Polyethylene Glycol–Glycerol.” Biotechnic & Histochemistry 66: 111–116.10.3109/105202991091105621716161 · doi ↗ · pubmed ↗
- 8Brunello, L. , E. Polverini , G. Lauria , et al. 2024. “Root Photosynthesis Prevents Hypoxia in the Epiphytic Orchid Phalaenopsis .” Functional Plant Biology 51: FP 23227.38442921 10.1071/FP 23227 · doi ↗ · pubmed ↗