Ipomoea tricolor (Convolvulaceae) in Turkey: New occurrence record and potential spread areas under current climatic conditions

2.1 General characteristics of the plant and seed germination potential

Surveys were carried out in agricultural and residential areas in the Tokat province of Turkey. Voucher specimens and seeds were collected from four populations in three locations where the plant was detected in the survey area. The plants were pictured on site. The general characteristics of the plant were determined, and a germination trial was carried out to find out the germination potential of the collected seeds. The seeds were manually sacrificed, and 25 of them (five replications) were placed between two layers of moistened filter paper in Petri dishes. The seeds were incubated 24 °C, 12 h light/dark conditions (Hardcastle, 1978).

2.2 Habitat suitability and occurrence records

The occurrence records of species at global scale were retrieved from various databases and search engines. Occurrence records were downloaded from GBIF and cleaned. For data cleaning, first the records having coordinate precision < 1 km were deleted. The records having no coordinates, belonging to other species and duplicate entries were deleted in the second step. Furthermore, Scopus (https://www.scopus.com/home.uri) and Web of Science (https://www.webofscience.com/wos/woscc/basic-search) were explored for occurrence records. The occurrence records were combined to get the final data set for model calibration. The occurrence records in Turkey were added to complete the dataset and the final data included 479 occurrences of the species.

2.3 Climatic data

The high-resolution climate data (∼1 km grids) for current climate (1979–2019) was downloaded from CHELSA (https://chelsa-climate.org/ (Karger et al., 2017) and used in predicting the potential spread areas of the species.

2.4 Modeling approach

The model was calibrated in a hierarchical fashion (Petitpierre et al., 2012). Model calibration was done on a global scale using global and local occurrence records. Habitat suitability was then projected globally and for Turkey. Since the Maxent model requires both occurrences and absences, 10,000 pseudo-absences were included in the model. A 1 km distance around each occurrence was buffered while sampling the pseudo-absences.

The values of 18 environmental variables (Table 1) were extracted for each occurrence/absence record. The model was calibrated with 70 % of the data and evaluated on the remaining 30 % (Elith et al., 2011). The model had 10 replications and results were averaged to get habitat suitability. Bio18 was removed due to high auto collinearity.

2.5 Model validation and statistical analysis

The final fitted model was visually validated with global distribution data. The goodness of fit of the model was evaluated by “area under the receiver operating characteristic curve” (AUC) (Swets, 1988). The AUC measures the overall discriminatory ability of the model by quantifying the probability that the model correctly ranks a random presence locality higher than a random background pixel (Phillips et al., 2006). The AUC ranges between 0.5 (model that performs no better than random) and 1 (model with perfect discrimination).

3.1 Species description and occurrence

During field observations, a plant belonging to the genus Ipomoea L. was first encountered on the edge of an agricultural land (orchard) and in the garden of a house in Kat town in Tokat province, Turkey (Figs. 1–4). The plant was subsequently identified as I. tricolor. Although Uludag et al. (2017) listed I. tricolor among the alien taxa in the flora of Turkey, a detailed literature review found no actual records showing the presence of the species in Turkey (Hançerli et al., 2018; Onen et al., 2021; Özkil and Üremiş, 2020). The general characteristics of the plant and the locations where the plant was observed were not mentioned by Uludag et al. (2017). Hançerli et al. (2018) and Özkil and Üremiş (2020) did not report the presence of I. tricolor in Turkey. Furthermore, the species is not listed in the Plants Data Services in Turkey nor in other databases keeping the record of plant distribution in the country.

Close

Fig. 1

Ipomoea tricolor plants observed during the exploratory surveys in Tokat province, Turkey.

Export to PPT

Close

Fig. 2

Ipomoea tricolor A) flowers, B) fruits and C) seeds obtained from the plants recorded during the exploratory surveys.

Export to PPT

Close

Fig. 3

Ipomoea tricolor plants climbing (A) apple and (B) vine plants on the edge of an orchard.

Export to PPT

Close

Fig. 4

Global occurrence records of Ipomoea tricolor used in the study to predict habitat suitability at global and regional scales (n = 479).

Export to PPT

Detailed surveys were carried out in agricultural and residential areas in the region, and the plant was found at two more locations (in the courtyard of a mosque and in a green area). Local people reported that seeds were collected from the garden of the house where the plant first detected due to its extremely dense bloom and stunning flowers. The coordinates of the occurrence sites are given in Table 2.

Ipomoea species are generally grown as ornamental plants in Turkey (Onen et al., 2021). Their seeds are often circulated throughout the country, are available from local and foreign sources and can also be purchased via internet from ecommerce websites (Onen, 2015; Onen et al., 2021, 2016b). Therefore, the species has possibly been introduced to the region as an ornamental plant and may spread throughout the country.

The general characteristics of the plant were determined from the voucher specimen collected from the region. It has been observed that the plant is a fast-growing vining and/or twining annual species (Wood et al., 2020). The stems are robust, 4–6 mm broad and up to 3.5–4 m in length, and glabrous. The plant has a deep-going taproot with dense fibrous roots. The leaves are alternate. The leaf blades are petiolate heart-shaped, cordate with rather angular, apex acuminate, 3.5–13.5 cm in length, and 3–14 cm in width. Leaves are entire, glabrous, and green. Auricles are rather angular, apex acuminate. Petioles are 1.5–19 cm long.

Inflorescence is pedunculate (few-flowered axillary cymes). Peduncles are 5–21 cm in length. The large showy flowers are trumpet-shaped (salverform). Sepals are dark green with white margin, oblong-deltoid. Corolla is 5-lobed, 5–7.5 cm long, and vivid bright blue in color with white tube and yellowish throat (Figs. 2 and 3). Sepals are dark green with white margin, oblong-deltoid, glabrous, inner slightly longer than the outer. Pedicels are 2–3 cm. Ovaries are 4-celled, with 1 ovule in each cell. The plant is hermaphrodite and is flowering from July to mid-December. The flowers open in the morning and then fade and drop by evening (Fig. 3A). Fruits are in capsular form, ovoid, glabrous, 9–11 × 5–6 mm, light to dark brown, and contain 1–4 seeds (Fig. 3B). The seeds are reddish-brown to blackish in color, 4.5–7 × 3–4.5 mm, minutely tomentose. Hilum wide is 0.5–1 mm (Fig. 2C). The general descriptions of plants from Tokat populations agreed with the existing definitions in the literature (Park et al., 2004; Wood et al., 2015).

Among the Ipomoea spp. found in Turkey, I. tricolor can be confused with I. purpurea (L.) Roth. However, I. tricolor is completely glabrous, and corolla 5.5–7 cm, blue with yellowish throat and white tube. Sepals are lanceolate, acute with white margins. I. tricolor leaves are entire. But all vegetative parts are softly pubescent in I. purpurea. I. purpureasepals are oblong or lanceolate, corollas are 4–5 cm and usually pink (rarely white or blue). I. purpurealeaves are entire or 3–5-lobed (Wood et al., 2020).

It has been reported that I. tricolor can produce many capsules and accordingly seeds (Fig. 3) like other Ipomoea species (Crowley and Buchanan, 1982; Onen et al., 2021; Thullen and Keeley, 1983). Besides the average seed germination rate of I. tricolor seeds in the current study was quite high (94.67 ± 3.27 %) from Petri dish experiment. Previous studies also revealed that the seeds of the genus have a high germination potential (Egley, 1990; Hardcastle, 1978; Onen et al., 2021). Therefore, it is a common annual species that regenerates via seeds in successive years, in the USA and particularly in cooler regions than its tropical origin of Mexico (Wood et al., 2020).

I. tricolor has allelopathic potential (Anaya et al., 1995), and is one of the Ipomoea species distributed around agro-ecosystems or disturbed habitats near settlements oftropical regions in almost all countries of the Americas, and clearly escapes from garden (Wood et al., 2020). In conclusion, the plant can easily escape, establish and naturalize into new areas, eventually becoming weedy.

3.2 Habitat suitability under current climatic conditions

The AUC values of the model were 0.94, indicating that habitat suitability was predicted with high accuracy. Thus, the results of the model are considered reliable for designing action plans to halt the further spread of the species (Fig. 5).

Close

Fig. 5

Curves for AUC under the receiver operating characteristic indicating habitat suitability of Ipomoea tricolor under current climatic conditions.

Export to PPT

The contributions of the bioclimatic variables used to predict the habitat suitability of I. tricolor under current climatic conditions was evaluated by Jackknife analysis and permutations. The results indicated that mean diurnal range, temperature seasonality, minimum temperature of the coldest month, and mean temperature of coldest and the warmest quarter had the highest contribution towards the model under current climatic conditions (Table 3). These results indicatedthat temperature is the main driver for habitat suitability of I. tricolor under current climatic conditions (Table 3). Temperature was the most influential environmental variable affecting the distribution of the species. These results indicated that temperature changes will exert strong impacts on the further spread of the species in Turkey.

The model predicted that the areas neighboring to the current distribution range of the species have high suitability for the spread of the species at global scale (Fig. 6). Interestingly, a large portion of Turkey has been predicted as suitable for the spread of the species (Fig. 7). The Black Sea and partly central Anatolia regions were highly suitable, and some areas in Thrace, Aegean, and Mediterranean regions moderately and/or less suitable, whereas the other areas of the country were predicted as unsuitable for the spread of the species.

Close

Fig. 6

Predicted global habitat suitability of Ipomoea tricolor under current climatic conditions.

Export to PPT

Close

Fig. 7

Predicted habitat suitability of Ipomoea tricolor in Turkey under current climatic conditions.

Export to PPT