Ecological Modeling of the Potential Distribution of the Mistletoe Phoradendron nervosum (Viscaceae) Parasitism in Ecuador

Abstract

This study characterizes Phoradendron nervosum, a hemiparasitic mistletoe species prevalent in Ecuador, using morphological, molecular, and ecological modeling approaches. Morphological analysis revealed that P. nervosum possesses green-yellowish cylindrical stems, lanceolate leaves with entire margins, and berry-like fruits with mucilaginous pulp. DNA sequencing of the internal transcribed spacer (ITS) region confirmed a 99.43% identity with P. nervosum (GenBank: AH009776.2), supporting the taxonomic classification. A maximum entropy (MaxEnt version 3.4.4) model was developed using 36 occurrence points and 19 bioclimatic variables to assess potential distribution across the Tumbaco region in Ecuador. Key environmental factors influencing the species’ distribution were precipitation during the warmest quarter (BIO_18), temperature seasonality (BIO_4), and mean diurnal temperature range (BIO_2). The model showed good predictive performance (AUC = 0.736), identifying areas with high suitability for P. nervosum, particularly in habitats with adequate water availability and thermal stability. Findings suggest that this mistletoe parasitizes both native and exotic tree species, potentially impacting biodiversity and forest health. This research provides a baseline for monitoring mistletoe spread under climate change scenarios and emphasizes the need for management strategies in agroforestry systems where host trees are vulnerable.

1. Introduction

For centuries, humans have moved plants around the globe from places where they are native to areas where they are not. Some moves have been unintentional, and we refer to these as weeds, invasive plants, introductions, etc. The plants moved intentionally are grown in horticultural, agricultural, or forestry settings, and they include the vast majority of fruit and vegetable crops upon which humans depend. There has been much interest in documenting the susceptibility of trees to disease when they are grown in non-native locations, particularly cases where mistletoes occur on introduced trees [1,2]. Hemiparasitic plants in the genus Phoradendron are obligate pathogens belonging to the family Santalaceae, previously grouped within the family Viscaceae [3]. Commonly known as “nervous mistletoes,” these plants can produce their own energy through photosynthesis while absorbing water and nutrients at the expense of the xylem of the host tree [4]. The fruits of these parasitic plants contain a sticky tissue (viscin) that covers the seeds, which, upon germination, produce a root called a haustorium that penetrates the host plant’s body, causing disturbances that lead to the formation of woody tumors [3]. It is often found growing on tree branches and is characterized by its glossy green leaves and small white or yellow flowers.

Phoradendron is one of the 36 genera of mistletoes listed by Hawksworth [4]. The following eight species occurred on introduced trees, mostly from North America: Phoradendron argentinum Urban (as P. pruinosum Urban), Phoradendron bathyoryctum Eichler (as Phoradendron heironymi Trel.), P. californicum Nutt., P. crassifolium (Pohl) Eichler, P. leucarpum (Raf.) Reveal & M. C. Johnst. (as P. macrophyllum (Engelm.) Cockerell), P. serotinum (Raf.) M. C. Johnst., P. tomentosum (DC.) Engelm., P. villosum (Nutt.) Nutt., P. piperoides (Kunth) Trel., P. quadrangulare (Kunth) Griseb., and P. trinervium (Lam.) Griseb. A survey of trees in Mexico City found that P. velutinum (DC.) Eichler (as well as some Loranthaceae) parasitized 17 host species in ten botanical families [5].

Possibly the first report of Phoradendron on cultivated trees in South America was from Argentina [6], where peach, chinaberry, and pomegranate were parasitized. More recently, 35 host trees near Quito, Ecuador were infected by Phoradendron nervosum Oliv. and P. parietarioides Trel. Severe infestations were observed on Prunus serotina subsp. salicifolia (Kunth) Koehne) where 20 trees harbored 131 individual plants of P. [7]. Another study in a second temperate valley in Pichincha Province detected an increase in P. nervosum on Populus, Mimosa, Callistemon, Hibiscus, Acacia, and Prunus serotina subsp. salicifolia; however, the abundance of mistletoe in this valley was considered low [8]. The presence of nervous mistletoe in Ecuador is ecologically significant, as it can weaken host trees and reduce their growth, potentially leading to negative consequences for biodiversity and, in some cases, even causing the trees’ death [9]. Fortunately, no economic losses attributed to Phoradendron nervosum have been reported in Ecuador to date. However, research has shown that the spread of tree diseases caused by aerial parasitism, for example, is influenced by globalization and climate change. This situation highlights the importance of investigations like the one presented here [10]. From an ecological perspective, in Ecuador, it has been found that P. nervosum infects 21 plant species in Quito [11] and an additional 2 species in nearby areas [7,8], indicating that it has a highly aggressive status [8]. Its wide distribution within a small geographical area raises concerns about its potential phenotypic plasticity, which may allow P. nervosum to adapt to a diverse range of host plants. This situation highlights the importance of both morphological and molecular characterization of the species.

In Ecuador, there are few references to these plants to the national botanical biodiversity and their impact as agricultural hemiparasites. Therefore, this study aimed to characterize the Phoradendron sp. morphologically and molecularly. As a hemiparasitic species, Phoradendron nervosum is modeled using a maximum entropy approach to assess its spatiotemporal distribution and evaluate its potential impact on regional plant diversity. The results confirm the presence of P. nervosum, characterized by its yellowish-green cylindrical stems, lanceolate leaves with smooth margins, and berry-shaped fruits containing mucilaginous pulp, showing a 99.43% identity in DNA analysis. Additionally, factors such as precipitation during the warmest quarter (BIO_18), temperature seasonality (BIO_4), and mean diurnal temperature range (BIO_2) may significantly influence the potential distribution, more so than the other variables examined. This research establishes a baseline for monitoring mistletoe spread under climate change scenarios and emphasizes the need for management strategies in agroforestry systems where host trees are at risk.

2. Materials and Methods

2.1. Specimen Surveying, Sampling, and Morphological Characterization

Samples of Phoradendron sp. were collected at the Academic Teaching and Experimental Center (CADET) of the Universidad Central del Ecuador-Facultad de Ciencias Agrícolas, located at an altitude of 2463 m above sea level, at a latitude of 00°13′32″ S and a longitude of 78°22′22″ W of the Tumbaco area. Morphological identification was conducted on 10 specimens. The samples were placed between newspaper sheets, labeled, and subsequently transported to the Botanical Department of the National Institute of Biodiversity (INABIO) of the Republic of Ecuador, for further analysis. An inverted microscope (OLYMPUS Corporation, Milan, Italy) and a stereomicroscope (OLYMPUS Corporation, Milan, Italy) were used for the identification process, along with tweezers, a scalpel, a 30 cm ruler, microscope slides, cover slips, and A4 white paper sheets. A Redmi Note 11 phone camera was used to document the samples. Leaves of the hemiparasitic plant were described in detail, including their shape, size, margin (edge), color, and arrangement on the stem. Also, the fruits were examined, describing their shape, size, color, and any notable surface characteristics [12]. Data from the collected specimens were then compared with herbarium specimens from the National Herbarium of INABIO’s Botanical Department. The observations were cross-referenced with scientific literature and herbarium specimens to confirm the identification of the hemiparasitic plant.

2.2. Field Observations

Field observations were conducted across multiple transects within the study area, specifically CADET-Tumbaco, which included an orchard of Prunus persica. In each transect, the presence of Phoradendron was recorded, with particular attention to host association and parasite density.

2.3. Molecular Analysis

The analysis was conducted in the Plant Genetics Laboratory of the Faculty of Agricultural Sciences—CADET. DNA extraction of one specimen and subsequent amplification were performed using the polymerase chain reaction (PCR) technique. Genomic DNA was extracted using the Invitrogen Thermo Fisher Scientific PureLink™ kit, Waltham, MA, USA. DNA was extracted from plant tissue following the recommended protocol.

For amplifying the plant ribosomal DNA region, ITS (internal transcribed spacer), the primers used were ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA) and ITS4 (5′-TCCTCCGCTTATTGATATGC). DNA amplification by PCR was performed using a GoTaq G2 Master Mix in a 50 µL reaction volume, consisting of 25 µL GoTaq G2 M.M., 1 µL reverse primer, 1 µL forward primer, 18 µL distilled water, and 5 µL DNA. PCR conditions in the thermocycler included an initial denaturation at 94 °C for 3 min, followed by 35 cycles consisting of denaturation at 94 °C for 30 s, primer annealing at 50 °C for 30 s, and extension at 72 °C for 1 min. After completing the cycles, a final extension was performed at 72 °C for 5 min, followed by cooling at 4 °C for 5 min. The final products were placed in a 1.5% agarose gel and visualized using 1× TAE buffer.

PCR products were analyzed in a 1.5% agarose gel with 1× TBE buffer. For gel preparation, 0.75 g of agarose was weighed and placed in a 125 mL Erlenmeyer flask. 1× TBE (tris-borate buffer) was added. To this, 50 mL of 1× TBE (tris-borate buffer) was added and mixed. The solution was microwaved for 30 s until the agarose was dissolved, and then it was allowed to cool to room temperature. Subsequently, 4 µL of SYBR Safe was added and gently mixed. The agarose solution was poured into the electrophoresis mold, with a comb placed to form wells, and left to solidify. Once the gel had set, it was submerged in 1× TBE buffer after removing the comb. A solution of 8 µL of DNA mixed with 2 µL of blue juice buffer was prepared and loaded into the wells using a micropipette. The electrophoresis chamber was powered on, applying 140 V for 30 min. Finally, the gel was visualized using a BIORAD photo-documentation system, the samples were labeled, and sent to the BIOGENA (Quito, Ecuador) for sequencing. Phylogenetic analysis was performed using the obtained sequencing data. Groups of the genus Phoradendron were selected from the GenBank database using the BLAST (Basic Local Alignment Search Tool version 2.14.1) online program (https://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 25 April 2025). The sequences obtained in FASTA format were introduced into a matrix, which was then used to align the sequences in MEGA11 software. Once the sequence alignment was completed, an appropriate phylogenetic tree model was selected. The phylogeny was constructed using the neighbor-joining method, yielding percentage similarities that closely matched those found in the BLAST sequence database.

2.4. Maximum Entropy Model

QGIS software (version 3.32.0) was used to delimit the boundaries of the Tumbaco area. To improve the selection of occurrence points, we incorporate data from the Global Biodiversity Information Facility (GBIF) [13] and iDigBio [14], accessing 41 georeferenced occurrences of Phoradendron nervosum. These records were filtered to ensure accuracy, relevance, and proximity to the Tumbaco parish to ensure robust ecological modeling. Building on this dataset, a shapefile (.shp) was created using updated land cover and land use maps from the year 2022, selecting 77 strategic points with X and Y coordinates within the parish. The time frame for data collection on the occurrence of Phoradendron nervosum should align closely with the time scale of environmental variables [15,16]. An Excel file (.xlsx) was then generated, recording each coordinate, where X is measured along the east–west axis and Y along the north–south axis. Subsequently, a ZIP file containing 19 bioclimatic variables for each of the 36 coordinates was downloaded. The variables, with a spatial resolution of 30 s (~1 km²), were obtained from the WorldClim online database “https://www.worldclim.org/data/worldclim21.html (accessed on 13 April 2025)“. Once both files were prepared, the Maxent program was downloaded [16,17], which operates using a JAR file. The program was launched, where a file containing the species’ location data (“coordinates”) was uploaded, along with a directory containing the environmental variables.

The model was trained with the expanded dataset, and parameter tuning was performed to optimize predictive performance. The selection of variables was validated through Jackknife statistical analysis to minimize bias and improve ecological inference. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC). Predictive performance was classified as follows: poor (0.5 ≤ AUC < 0.7); average (0.7 ≤ AUC <0.9); and high (0.9 ≤ AUC ≤ 1). The model achieved an AUC of 0.7, indicating good performance.

3. Results

The Phoradendron reported in this study is P. nervosum, a widespread species found in Central and South America. The species produces recalcitrant epigeal seeds encased in a mucilaginous pulp, contained within spherical, yellowish-orange berries that measure up to 5 cm in diameter. The fruits are arranged in four longitudinal series. Notably, stems and leaves are chlorophyllous and glabrous. Specimens were compared with previously identified species from the National Herbarium (INABIO Botanical Department), confirming their classification as Phoradendron nervosum. It is most similar to P. chrysocladon, P. applantum, and P. peruvianum Eichl. in Ecuador (Figure 1A–F). Observations of Phoradendron seedlings growing and establishing in trees of the genus Prunus, Alnus, and Salix indicate that they can persist under different growth conditions and parasitize a wide range of hosts. Large clumps were frequently observed, often contributing to the collapse of branches in mature trees. Phoradendron nervosum is a member of the Santalaceae. These plants exhibit distinctive morphological characteristics. Specimens were green, slightly yellowish hemiparasitic shrubs with cylindrical, herbaceous stems (Figure 1A,B), internodes ranging from 6 cm to 14 cm in length, and axillary basal cataphylls (Figure 1C). The leaves are lanceolate and simple, measuring 10 × 4 cm (Figure 1D), with an acuminate apex and a poorly differentiated petiole. Based on their margin, they are classified as entire. They are parallel-veined with prominent veins (Figure 1E) and are oppositely arranged on the stem. Trichomes were absent on the stem and leaves, with simple, opposite, or decussate leaves with entire margins and pinnate venation. Phoradendron sp., produced fleshy berry-like fruits ranging from 3 to 6 mm in size (Figure 1F), varying in color from white to reddish, with a mucilaginous pulp (Figure 1G). A cross-section of the stem was performed (Figure 1H,I), revealing an open collateral vascular bundle, where the phloem is located on the outer side, the xylem on the inner side, and the fascicular cambium can be observed between them. The specimens analyzed in this study had an intrafascicular vascular cambium (located between the xylem and phloem). This cambium is arranged in an orderly circular pattern, separated by medullary rays that connect the cortical parenchyma with the medullary parenchyma (Figure 1I).

Young leaves from a Phoradendron sp. specimen were used for DNA analysis. Molecular characterization focused on the internal transcribed spacer (ITS) region. The ITS-DNA sequence was uploaded to CodonCode Aligner for alignment and trimming, resulting in a final sequence of 729 base pairs. Molecular characterization results indicated a 99.43% identity match with the hemiparasitic plant Phoradendron nervosum (GenBank: AH009776.2) via BLAST analysis, corroborating the morphological identification. Additionally, sequences showed 96.55% compatibility with Phoradendron cf. tonduzii (GenBank: AF178736.1) and Phoradendron trianae (GenBank: LT599672.1). Phylogenetic analysis shown in Figure 2 shows the evolutionary relationships among different species of Phoradendron and related taxa, based on ribosomal RNA gene sequences (including ITS and 5.8S rRNA). The phylogeny was performed using the neighbor-joining (NJ) model, and reveals a well-defined hierarchical structure, grouping the strains into clades that help validate the relationships between the isolates and confirm the exact species under study. Clade 1 includes eight isolates of Phoradendron sp., based on the internal transcribed spacer 1 58.S ribosomal RNA, and internal transcribed spacer 2 complete sequences. Within this clade, P. vernicosum, P. tamaulipense, and P. serotinum cluster together, sharing a node with 82% bootstrap value. The isolate of the studied species, Phoradendron-DCH-UCE-FAG, is positioned within a subclade of Clade 1, showing a close genetic relationship with Phoradendron nervosum, supported by a 100% bootstrap value. Phoradendron crassifolicum was close with a 93% bootstrap value. P. nervosum and P. crassifolicum were highly related and shared similar anatomical features, such as leaves with prominent venation, small inflorescences arranged in racemes, and a hemiparasitic lifestyle that depends significantly on their hosts. Both species are native to tropical and subtropical regions, with a higher presence in Central America [15]. The tree was rooted to the isolate of Phoradendron leucarpum, which, despite sharing general characteristics of the genus, exhibits notable morphological differences compared to the studied species, being a more divergent/outlying taxon.

Phoradendron nervosum parasitism affects the fitness of its hosts (Prunus persica).

The hemiparasitic plant P. nervosum was detected in a peach plot evaluated at CADET-Tumbaco, with a maximum incidence of 70%. Incidence was assessed monthly over a period of four months, by counting the number of infected trees per row. A total of 20 rows of peach trees were evaluated, with incidence levels recorded individually for each row. The highest incidence occurred in rows located in the central and southern areas of the plot. P. nervosum exerts a significant physiological impact on its host by producing haustorial structures that penetrate the vascular tissue, leading to direct interference with nutrient and water transport (Figure 3).

Maximum Entropy Model

After clipping the 19 bioclimatic layers, a database was created with 77 coordinate points. A land cover and land use map was used to identify 36 strategic points (Figure 4) corresponding to agricultural land, shrub vegetation, and herbaceous vegetation. Given that the species is a hemiparasitic plant, these locations were selected based on the presence of suitable host plants, from which the species partially derives its nutrients.

For the modeling process, 36 occurrence coordinates of the species were recorded within the boundary topography of the Tumbaco parish. Environmental layer data from 19 bioclimatic layers were uploaded (

Table 1. Significance of bioclimatic variables in the predictive model.

Variable	Contribution	Importance of the Permutation
BIO_18	62.6	26.3
BIO_2	18	18.7
BIO_4	16.2	44.4
BIO_12	2.3	5
BIO_17	0.5	2.2
BIO_3	0.2	0
BIO_11	0.1	0
BIO_1	0	0.2
BIO_14	0	1.7
BIO_6	0	0.2
BIO_10	0	0.9
BIO_8	0	0.2
BIO_5	0	0.1
BIO_15	0	0
BIO_16	0	0
BIO_13	0	0
BIO_7	0	0
BIO_9	0	0
BIO_19	0	0

). The parameters were adjusted using a cumulative output format. The most critical variables for the model were the following: BIO_18 (precipitation of warmest quarter (mm)), BIO_4 (temperature seasonality (standard deviation × 100)), and BIO_2 (mean diurnal range (mean of monthly (max temp—min temp))). The identification of species dispersion was based on significant environmental variables relevant to its reproduction. The Jackknife statistical test was used to assess estimation bias. This test identifies the ecological requirements of the species and, based on these factors, determines the optimal spatial distribution of Phoradendron nervosum.

In Figure 5, the omission rate and predicted area of the cumulative threshold can be observed. The black line represents the expected omission according to a perfectly fitted model, serving as a reference. The blue line indicates the omission of the samples, while the red line represents the fraction of the background area that the model predicts as a suitable habitat for the species. This can be considered an ideal model because the blue line remains below the black line, indicating that it is a highly accurate presence-only model for the species. However, this also suggests a limited generalization capacity. The progressively decreasing red line confirms that the model is functioning correctly. As the line descends smoothly, it shows that the model is gradually reducing the predicted area as stricter thresholds are applied. At the beginning, the red line starts at a high value (1.0) when the cumulative threshold is low (near 0). This is because the model initially predicts a large portion of the area as suitable habitat. As the cumulative threshold increases, the red line decreases gradually, indicating that the model becomes more restrictive, retaining only areas with higher suitability values as suitable habitats. Finally, the red line approaches 0 when the threshold is high (100), meaning that only a few areas are considered suitable for the species’ survival. There are no clear signs of overfitting, indicating that the model has strong predictive and generalization capabilities.

Figure 6 shows the receiver operating characteristic (ROC) curve, which represents sensitivity as a function of the false positive rate and evaluates the predictive capacity of the model. The figure explains that the X-axis (specificity—fractional predicted area) indicates the proportion of false positives, while the Y-axis (sensitivity—omission rate) represents the proportion of actual species presences. From the ROC curve, we observe an area under the curve (AUC) value of 0.736, indicating a good predictive model with acceptable performance, though not completely precise. Generally, the minimum AUC value is 0.5, and predictive ability is considered strong as the value approaches 1. A higher AUC means the model can more accurately distinguish between suitable and unsuitable areas for the presence of Phoradendron nervosum.

Figure 7 displays a graphical representation of the distribution model for Phoradendron nervosum within the topographic boundary of Tumbaco. Warmer colors indicate excellent predicted conditions or high suitability for the species’ presence. The color bar represents the logarithmic probability scale of species presence: dark red (100–50): Very high probability of presence; orange and yellow (25–3.1): Intermediate probability; green (0.39–0.1): Low probability; and white (0.001–0.00001): Extremely low or no probability. Each cell in the map corresponds to a geographic area. White zones represent urban areas with infrastructure and no vegetation cover, where no data is available. Red and orange zones indicate shrubland, grasslands, forest plantations, and agricultural land—areas where the species is most likely to thrive. Green zones represent transitional areas with suboptimal conditions, while blue and white zones are populated areas where the species cannot survive.The model’s predictive performance was primarily influenced by three bioclimatic variables: BIO_18 (62.6% contribution): Precipitation of the warmest quarter—the most significant variable, indicating that water availability during the hottest season strongly affects the species’ distribution, BIO_2 (18% contribution): Mean diurnal temperature range—the second most influential factor, suggesting sensitivity to daily temperature fluctuations, and BIO_4 (16.2% contribution, highest permutation importance: 44.4): Temperature seasonality—while contributing less than the others, its exclusion significantly impacts model accuracy, highlighting the importance of thermal stability for the species. BIO_18 confirms that P. nervosum is highly dependent on water availability during peak heat, and BIO_2 suggests adaptation to daily thermal variations, likely influencing its microhabitat preferences.

Precipitation during the warmest quarter is the most influential bioclimatic variable affecting the presence of Phoradendron nervosum. This hemiparasitic species obtains its water from the host’s xylem, making its water balance entirely dependent on the host’s water absorption capacity. This critical dependency explains why P. nervosum preferentially selects hosts with deep root systems capable of accessing groundwater, ensuring sustained hydration even during dry periods.

Thermal stability proves equally vital for hemiparasitic species, which have evolved sophisticated adaptation mechanisms. These include stomatal regulation to minimize water loss under high temperatures, and morphological adjustments such as developing smaller leaves to reduce transpiration rates. Remarkably, some hemiparasites can even trigger physiological responses in their host plants that enhance joint resistance to extreme temperature conditions.

4. Discussion

P. nervosum imposes substantial hydraulic, nutritional, and carbon-balance stresses on different hosts, leading to marked reductions in growth and reproductive output—clear evidence that the parasite compromises host fitness and that epidemics are apparent.

Examples of epidemics devastating native trees usually focus upon introduced pathogenic fungi such as Dutch elm disease (Ophiostoma ulmi) [18], Chestnut Blight (Cryphonectria parasitica) [19,20], Sudden Oak Death (Phytophthora ramorum) [21,22,23], and Pine Pitch Canker (Fusarium circinatum) [23,24,25]. Mistletoes have not been the causal agent of a massive buildup of parasite populations. When forestry and crop plants are grown outside their native habitats, they encounter different abiotic and biotic stresses [26,27]. Epidemics are generally assumed to occur because the host tree has not evolved or developed in tandem with the pathogen and thus lacks adequate resistance. All four scenarios in Figure 8 represent ecological imbalances that can result in epidemics. Even with native trees, when they are grown as monocultures, i.e., stands of the same age and genotype, they are particularly susceptible to mistletoe infestation, including native mistletoes.

Most mistletoes exist as native parasites on native trees (A). In some cases, these combinations can result in pathology, such as Arceuthobium (dwarf mistletoe) parasitizing various conifer hosts (Abies, Larix, Picea, Pinus, Pseudotsuga, Tsuga, etc.). Over 250 species of the genus Phoradendron occur throughout the New World, and the number of native host species parasitized is too long to enumerate here. The next scenario involves a non-native mistletoe parasitizing a native tree (B). An example of this is Viscum album L., the European mistletoe, which was introduced purposely to California [2,5] and now parasitizes a number of both native and non-native host trees (C). The last scenario involves native mistletoe parasites occurring on non-native trees (D).

There has been considerable interest in documenting the susceptibility of trees to disease when they are grown in non-native locations, particularly in cases where mistletoes occur on introduced trees [1,2,5,6]. The list compiled by Hawksworth (1974) [6] only included situations where the host tree was being grown outside its natural range and then was parasitized by a native mistletoe.

Species distribution modeling (MaxEnt) identified three critical climatic variables influencing the species’ range within Tumbaco’s topographic limits: precipitation of the warmest quarter, temperature seasonality, and diurnal temperature range. Ecologically, this model helped pinpoint key habitats for the species’ survival and assess how its distribution may shift under climate change scenarios. In Ecuador, 40 species of Phoradendron occur [5] of them native. The most widespread species across Central and South America are P. quadrangulare (Kunth) Griseb. and P. berteroanum (DC.) Griseb., followed by P. piperoides (Kunth) Trel., P. dipterum Eichl., and P. chrysocladon. Twelve species are endemic to Ecuador. As stated in the Introduction, P. nervosum and P. parietarioides have both been documented as parasites of introduced trees in Ecuador. Because of the complex taxonomy of Phoradendron, it is likely that other species also occur on cultivated trees but have gone undetected. The example described here shows that Phoradendron nervosum can infect both native and non-native species. For us, it is unclear to assume that P. nervosum in Tumbaco parish affects introduced trees differently from native ones, but some indirect evidence of this is presented as the case of study in the Prunus orchard. Phoradendron-DCH-UCE-FAG has an epidemic effect on Prunus persica (introduced in the Andean valleys). It appears that incidences of mistletoe parasitizing exotic hosts are increasing in Ecuador, although no empirical data have been assembled to test this hypothesis. This and other studies that monitor the rate of mistletoe growth and host effects, under varying environmental conditions (e.g., drought), are needed. In addition, studies on the relationship between Phoradendron occurrence and percent of canopy closure or canopy dieback are called for. In cases of branch and tree collapse, the degree of host canopy shading by massive Phoradendron shrubs may contribute to this phenomenon. Quantifying light availability at the host tree level is necessary to test this hypothesis. Although the seasonality of temperature (BIO-4) showed a high permutation importance, seasonal fluctuations in the study region are relatively small. This result suggests that BIO-4 may be interacting with other climatic variables affecting mistletoe distribution. Finally, this research provides a baseline for monitoring mistletoe spread under climate change scenarios and emphasizes the need for management strategies in agroforestry systems where host trees are vulnerable.