Ecogeographic Characterization of Potential Tectona grandis L.f. (Teak) Exploitation Areas in Ecuador

Abstract

Tectona grandis L.f. (teak) is a timber species of exceptional commercial value, widely cultivated in Ecuador for export to international markets. This study aimed to ecogeographically characterise current production and identify zones with high potential for exploitation, using tools from CAPFITOGEN v3.0 and the MaxEnt maximum entropy algorithm, based on data from 1023 plantations. The territory was classified into 26 ecogeographic categories, of which teak is present in 13. Categories 17, 19, and 21 were predominant, collectively accounting for 88.27% of the analysed plantations. Sixteen relevant variables (comprising four climatic, four edaphic, and eight geophysical factors) served as predictors in MaxEnt, with model validation demonstrating strong accuracy (AUC = 0.924). The most influential factors for teak suitability were precipitation seasonality, altitude, annual precipitation and September wind speed. Areas with elevated and high probabilities for teak exploitation were quantified at 6737.83 km² and 10,154.70 km², respectively, with Guayas, Los Ríos, and Manabí provinces showing the most favourable conditions. This integrative framework provides an evidence-based basis for land-use planning and resource management, supporting more sustainable and efficient development of Ecuador’s teak forestry sector.

1. Introduction

Tectona grandis L.f. (teak) is one of the most renowned timber species worldwide, prized for the exceptional hardness, durability, colour, uniformity, and fine grain of its wood [1,2]. Native to Southeast Asia (India, Myanmar, Thailand, and Laos) [3], teak has been successfully introduced to other tropical regions, initially in Nigeria in 1902 [4] and Trinidad and Tobago in 1913, using seeds from Myanmar [5]. Currently, teak is found in 80 countries and is considered a priority species for large-scale cultivation in many of them because of the quality of its wood. The area of teak forests, both natural and planted, has increased, as has timber extraction. The global area of planted teak forests is estimated at 4.854 billion hectares [6], representing an increase of 11.7% compared to the previous estimate of 4.346 billion hectares [7]. In Ecuador, teak was introduced in the 1950s through initiatives led by Nelson Rockefeller aimed at establishing agroforestry systems in Balzar and Quevedo (Littoral region), in conjunction with coffee (Coffea arabica L.) [8]. Over the past decade, teak cultivation has experienced renewed momentum as a component of the National Forest Restoration Plan 2019–2030 (Ministerial Agreement 65), which stipulates an expansion target of 10,000 hectares annually [9]. The estimated area of teak currently cultivated is 75,000 hectares, making it the seventh-largest producer in the world. This area reflects a 66.7% increase over previous estimates [6]. The high commercial value of the product drives this expansion, supported by the rising international demand for premium timber and its potential within sustainable forestry frameworks [10,11]. Moreover, in 2023 alone, Ecuador’s teak exports generated US$50 million [10]. Despite the growing number of plantations, Ecuador’s forest expansion programmes often lack comprehensive planning that accounts for the country’s ecological diversity [12]. It is important to note that the low-lying Littoral region areas are currently the most productive (provinces of Guayas, Manabí, Esmeraldas, El Oro, Los Ríos, Santo Domingo, and Santa Elena) [13]; however, there is no precise delineation of optimal cultivation zones at the national level. There is no official, accurately georeferenced registry of teak plantations that allows for monitoring their development up to the point of felling, despite this being a requirement [9]. The available information is partial and difficult to access. For practical reasons, only plantations requiring a logging licence and transport guides are documented [14]. Consequently, there is insufficient data to effectively utilise a Geographic Information System (GIS) for the analysis and management of these teak plantations. It is important to emphasise that the low-lying areas of the Littoral region are currently the most productive [15]; however, the extent of land utilised for exploitation remains unknown, and there is no precise delineation of the most suitable cultivation areas at the national level [16].

The development of GIS in the late 1950s responded to the need to generate accurate spatial information, driven by technical advances and public interest. In the forestry sector, efficient management supported by the use of GIS depends on the availability of accurate and correctly georeferenced databases, both spatially and temporally [17]. GIS is a key tool for land-use planning and natural resource management [18]. Relevant examples include the ecogeographic map for creating core collections of bean (Phaseolus vulgaris L.) [19]; the analysis of the geographic distribution of wild potatoes (Solanaceae, Solanum sect. Petota) using georeferenced data [20]; the preparation of geospatial databases in Brazil containing edaphic, topographic, and early agricultural use information [21]; the identification of priority areas for in situ conservation of crop wild relatives in South Africa considering multi-provenance databases [22]; conservation and land-use planning in the Kas-Kekova Protected Area of Turkey using biodiversity data [23]; and, more recently, an R package that reconciles taxonomic names and geospatial information to generate summaries of the World Checklist of Vascular Plants [24].

In Ecuador, GIS and predictive models have facilitated the identification of adaptability zones for species such as the capulí (Prunus serotina Ehrh) by selecting 12 influential variables of different nature (climatic, edaphic and geophysical) and 147 occurrence points georeferenced by the germplasm bank of the National Institute of Agricultural Research—INIAP—of Ecuador [25]. Similarly, 22 adaptability categories for melloco (Ullucus tuberosus Caldas) were determined in the Andean region of Ecuador, based on 187 occurrence sites referenced by INIAP and 13 variables [26]. Later, using 195 occurrence sites and 55 variables, 16 ecogeographic categories were identified for the cultivation of cassava (Manihot esculenta Crantz) [27]. Finally, a recent map of the phenotypic diversity of chocho (Lupinus mutabilis Sweet) revealed that areas of high morphological diversity do not entirely coincide with regions of high ecogeographic diversity [28].

Advances in technology have improved access to environmental data and made high-performance, low-cost computing widely available, facilitating the development of predictive models of ecological requirements and species distribution, as exemplified by the algorithm implemented in MaxEnt, which predicts the geographic distribution of species [29]. These models have been widely applied to assess how climate change, one of the main threats to biodiversity, affects species distribution [30,31]. In this regard, using information from a continuous data set of National Forests in China and data from the collection of Chinese fir (Cunninghamia lanceolata (Lamb.) Hook.), the obtained maximum entropy models predicted probable distribution scenarios for the near future, between 2050 and 2070 [32]. Other applications include modeling the ecological niche of, downy oak (Quercus pubescens Willd.) and holm oak (Quercus ilex subsp ilex) has been modelled for the present and the Middle Holocene [33], or the expansion in China in the 2050s to 2090s of Chinese cinnamon (Cinnamomum cassia Nees) [34], the pine caterpillar (Dendrolimus punctatus (Walker, 1855)) and its host Chinese red pine (Pinus massoniana Lamb.) has been predicted [35] and the identification of environmental factors for the distribution of mistletoe (Phoradendron nervosum Oliv.), a species that could affect the health of forests in Ecuador [36].

Specific ecogeographic variables and producer knowledge influence the distribution and development of agroforestry plantations. In this context, the ecogeographic characterization using the tools developed by CAPFITOGEN [37], combined with predictive models such as MaxEnt [29], offers a key strategy for identifying areas with the greatest potential for the successful and sustainable establishment of these plantations. This approach has been applied in the conservation of wild relatives of priority crops in Norway [38], in the Southern African Development Community [39], and in West Africa [40]. This study aims to fill this gap by employing CAPFITOGEN and MaxEnt modelling, tools not previously combined for forest species analysis in Ecuador, to identify areas with optimal environmental conditions for teak exploitation.

2. Materials and Methods

2.1. Data Sources

Data on the analyzed plantations (Figure 1,

Table 1. Number, area and age of teak plantations analyzed by regions and provinces.

Province/Region	N° of Plantations	Area			Plantation Age (Years)
		Ha	Mean	Min–Max	Average	Min–Max
Littoral Region	984	21,647.53	22.00	0.01–563.0	4.6	1.0–22.0
El Oro	21	523.7	24.94	0.31–267.42	3.6	2.6–5.6
Esmeraldas	48	773.61	16.12	0.12–78.58	4.9	2.5–18
Guayas	350	6647.54	18.99	0.04–500.0	4.4	2.3–18
Los Ríos	251	6875.87	27.39	0.01–563.83	5	1.0–12.0
Manabí	265	3373.14	12.73	0.04–300.0	4.4	2.0–18.0
Santo Domingo	49	3453.08	70.47	0.09–2700.00	5.4	3.3–22.0
Andean Region	20	251.03	12.55	0.49–100.0	4.9	2.0–15.0
Azuay	3	2.13	0.71	0.49–0.96	5.6	5.6–5.6
Bolívar	6	24.5	4.08	0.5–12.0	6	2.0–15.0
Cañar	6	110.11	3.06	4.39–52.15	4	3.4–5.6
Chimborazo	2	9.43	4.72	2.14–7.29	3.6	3.6–3.6
Cotopaxi	2	4.86	2.43	1.92–2.95	5.2	5.2–5.2
Pichincha	1	100	100.00	100.0–100.0	5	5.0–5.0
Amazonian Region	19	1442.6	75.60	0.61–1369.0	7	2.8–14.0
Orellana	5	1390	278.00	3.0–1369.0	9.8	3.0–14.0
Sucumbíos	14	46.4	3.31	0.61–14.24	4.2	2.8–14.0
Total	1023	23,334.38	22.81	0.01–1369.0	5.5	1.0–22.0

) come from the reconciliation of two databases from different sources into a single database with 1023 records. These databases include geographic coordinates, province, plantation age, and area. The first database, comprising 970 plantations, was provided by the Forest Production System (SPF) of the Ecuadorian Ministry of Agriculture (unpublished data) and contains official records of plantations used to obtain forest harvesting permits issued by the Ministry of Environment and Energy. The second database was derived from a previous study that included information on 77 teak forest producers [41]. The 24 duplicate records shared between both databases were removed from the second database, leaving 53 adequate records for inclusion in the final database.

2.2. Study Area

Ecuador, situated on South America’s Pacific coast, borders Colombia to the north and Peru to the east and south. It presents considerable climatic and topographic variability, which contributes to its classification as a megadiverse country [42]. It is home to 14 main ecosystems, of which the following are represented in the study area: coastal rainforest, western dry forest, Amazon rainforest, and Amazon rainforest floodplain [43]. Its territory covers an area of 256,370 km² [44], which is continentally classified into three regions: Littoral, Andean, and Amazonian, ranging from sea level to 5230 m above sea level (m a.s.l.) [43]. This study investigates the altitudinal range from 0 to 600 m a.s.l., because the georeferenced teak plantations were located between 7 and 591 m a.s.l. Furthermore, observations conducted during confirmatory visits to the plantations did not detect higher elevations. These plantations can be found in six provinces of the Littoral region, as well as in the low-lying areas (foothills areas) of six Andean provinces that border the Littoral region, and two provinces in the Amazonian. (Figure 1; Table 1). These areas have average annual temperatures ranging from 23 to 26 °C. In all places where teak is planted in the Littoral region and its neighbouring Andean regions, rainfall is seasonal, with a primary rainy season from February to May. In contrast, in the Amazonian, rainfall is intense and constant throughout the year, with an average annual rainfall of 3000 mm [45].

2.3. Ecogeographic Characterization

For the ecogeographic characterization of the climatic, edaphic, and geophysical conditions of the teak plantations, the CAPFITOGEN version 3.0 toolkit [37] was used (Figure 2). The process begins with the TesTable tool, which verifies the integrity and consistency of the passport data according to the standards required by the tools, detecting inconsistencies or duplicates, thereby preventing the execution of subsequent tools if errors are not corrected. After the database has been cleaned, the GEOQUAL tool is run to assess the georeferencing quality of the passport data. This tool assigns a quality score ranging from 0 to 100 based on criteria proposed by Parra-Quijano [37]. Values close to 100% indicate highly reliable georeferencing, while those below 50% are considered undemanding or of low spatial precision. All records from the 1023 plantations were included in this characterization process, as they had a georeferencing quality above 70%.

The next tool is SelecVar (Figure 2), which extracts 19 climatic variables from WorldClim [46], 19 edaphic variables from HWSD [47], and 31 geophysical variables from the Shuttle Radar Topographic Mission [37,48,49] (Tables S1 and S2). Subsequently, this tool selects a set of non-redundant variables with significant explanatory power, utilising three methodologies: Random Forest (RF), bivariate correlation (BCA), and principal component analysis (PCA). These methodologies determine the contribution of variables that were previously extracted (Tables S3–S11). The Random Forest (RF) (Figures S1–S3) chooses a subset of variables based on the “mean decrease accuracy” index, which is then analyzed using BCA. BCA (redundancy elimination) starts by comparing the least important variable in Random Forest with the most significant variables. If it correlates with any key variable, as defined by the user-specified coefficient and p-value, it is discarded; otherwise, it is retained. This process is repeated for subsequent variables until the specified correlation threshold (r > 0.5) is reached [37]. Finally, PCA is applied to the set of variables chosen by the user, providing information on the eigenvectors and eigenvalues, along with a table showing the position of each RF-BCA selected variable relative to the principal components (5 in the present analysis) (Tables S4, S7 and S10). The minimum distance to consider two plantations as distinct was 1 km to minimize spatial autocorrelation and reduce bias from possible overrepresentation of areas in predictive analysis.

The final tool, ELCmaps (Figure 2), produces a map of ecogeographic categories utilising variables selected by the preceding tool, at a grid resolution of 1 × 1 km (30 arc seconds). It employs the elbow method, which applies the K-means clustering algorithm to determine the optimal cut-off point based on the reduction in the sum of squares within groups [37].

Statistical analyses of the variables were conducted within the RStudio v. 2024.03 environment [50], utilising the “car” [51], “ggplot2” [52], “nortest” [53], and “FSA” [54] packages. In cases where the assumptions of normality (Shapiro–Wilk test) and homogeneity of variances (Levene’s test) were not fulfilled, the non-parametric Kruskal–Wallis test was applied. This was subsequently followed by Dunn’s post hoc test with Holm’s correction for multiple comparisons.

2.4. Predictive Model

The potential geographic distribution of teak plantations was modelled using MaxEnt version 3.4.1 [29,55]. MaxEnt v3.4.4 consistently delivers optimal performance compared to other models [56]. It is a widely used modelling method [40,57] across various fields of biogeography, conservation biology, and ecology, especially where observations have not been systematically collected [58], to predict potential distribution from presence data and environmental variables [59,60]. Spatial autocorrelation and overrepresentation bias were addressed through the removal of duplicate records. Feature types were configured automatically when k-fold cross-validation was selected, and the logistic output type was specified. The program automatically allocated 408 points (39.9%) to the training data and 46 points (4%) to the test data, culminating in a total of 10,281 points—including background and presence points—to ascertain the Maxent distribution. Validation metrics were obtained following 10 replicates [35,61]. The quality of the model was evaluated based on the area under the receiver operating characteristic (ROC) curve (AUC), with an AUC exceeding 0.7 regarded as indicative of acceptable predictive accuracy [62].

Since the prediction model generates a continuous probability gradient for species presence, these probabilities were classified into five equal-width intervals similar to those employed by [32]. The resulting probability categories were labelled as follows: “null” (p = 0–0.2), “low” (p = 0.2–0.4), “moderate” (p = 0.4–0.6), “high” (p = 0.6–0.8), and “very high” (p = 0.8–1.0).

All geospatial data generated during the CAPFITOGEN and MaxEnt operations (Figure 2) were processed in ArcGIS v. 10.8.2 [63] for visualisation and further geospatial analysis.

3. Results

This section presents the results in three subsections. The first subsection outlines the ecogeographic categories and highlights the differences between areas with teak presence in Ecuador. Next, the predictive model identifies the most influential variables that define potential areas, along with the value ranges of these variables that favour the species’ development. Finally, the forecasted potential areas for teak plantations are displayed, accounting for the probability of success across the entire country, its provinces, and ecogeographic categories.

3.1. Ecogeographic Characterization

The environmental variables selected using the SelecVar tool were 16 (

Table 2. Statistical descriptors of the environmental variables that delineated the ecogeographic categories. The variables were organised in descending order based on their coefficient of variation (CV).

Variable	CV	Mean ± S.D.	Min–Max	Q1–Q3
Precipitation of the driest quarter (bio_17)	81.47	340.57 ± 277.45	0–841	38–617
Precipitation seasonality (coefficient of variation) (bio_15)	75.75	55.73 ± 42.22	12–189	18–99
Elevation (alt)	61.61	226.03 ± 139.26	0–600	112–305
August wind speed (wind_8)	54.83	1.62 ± 0.89	0.8–5.3	1–2
July wind speed (wind_7)	54.82	1.58 ± 0.87	0.8–5.3	1–1.9
September wind speed (wind_9)	52.83	1.67 ± 0.88	0.9–5.2	1.1–2
October wind speed (wind_10)	50.82	1.69 ± 0.86	0.9–5	1.1–2
November wind speed (wind_11)	49.8	1.68 ± 0.84	0.9–4.9	1.1–2
Annual wind speed (wind_annual)	49.39	1.56 ± 0.77	0.87–4.75	1.03–1.84
December wind speed (wind_12)	48.53	1.63 ± 0.79	0.9–4.8	1.1–1.9
Annual precipitation (bio_12)	47.06	2243.31 ± 1055.67	103–4374	1231–3125
Cation exchange capacity of soil in cmolc/kg—topsoil (t_cecsol)	46.95	16.1 ± 7.56	4–102.82	10.5–20.33
Available soil water capacity (volumetric fraction) for h2—topsoil (t_awc2)	11.88	10.11 ± 1.2	5.2–20.95	9.45–10.68
Bulk density (fine earth) in kg/m3—topsoil (t_bulk_dens)	7.02	1333.55 ± 93.61	492.33–1579.8	1299.75–1386.9
Saturated water content (volumetric fraction) for tS—topsoil (t_awcts)	6.68	45.37 ± 3.03	38.05–71.57	43.6–46.45
Maximum temperature of the warmest month (bio_5)	2.73	30.54 ± 0.83	26.4–33.1	30.1–31.1

CV: Coefficient of variation. Mean: Mean. SD: Standard deviation. Min: Minimum value. Max: Maximum value. Q1: First quartile. Q3 Third quartile.

): four climatic variables (precipitation of the driest four-month period -bio_17-, seasonality of rainfall -bio_15-, annual precipitation -bio_12-, and maximum temperature of the warmest month -bio_5-), four edaphic variables related to the top soil layer (cation exchange capacity -t_cecsol-, available soil water capacity -t_awc2-, bulk density -t_bulk_dens-, and saturated water content -t_awcts-), and eight geophysical variables (elevation and wind speeds in August, July, September, October, November, as well as annual and December -wind_8, wind_7, wind_9, wind_10, wind_11, wind_annual, and wind_12-).

The territory was classified into 26 ecogeographic categories (Figure 3), notable for their considerable environmental variability, as 12 of the selected variables exhibited coefficients of variation exceeding 45% (Table 2 and Table S12). Among these, precipitation in the driest quarter (bio_17), precipitation seasonality (bio_15), and elevation (alt) showed the highest coefficients of variation. Conversely, three soil variables (available water capacity -t_awc2-, bulk density -t_bulk_dens-, and saturated water content -t_awcts-) and one climatic variable (maximum temperature of the warmest month -bio_5-) displayed the lowest coefficients of variation, all below 12%. The remaining variables had coefficients of variation ranging from 45% to 55% (Table 2 and Table S12).

Among the total ecogeographic categories, 13 contain teak plantations (

Table 3. Area occupied by the resulting ecogeographic categories. Number of farms and corresponding area. Categories with teak presence are marked with an asterisk (*).

Categories	Study Area		Studied Farms
	Number of Grid Cells (km2)	Percentage of the Number of Grid Cells	Number	Percentage	Area (ha)	Percentage
1 *	74,378	43.15	18	1.76	1427.40	6.12
2	2	0.00	0	0.00	0.00	0.00
3 *	13,414	7.78	1	0.10	9.00	0.04
4	1994	1.16	0	0.00	0.00	0.00
5	416	0.24	0	0.00	0.00	0.00
6	265	0.15	0	0.00	0.00	0.00
7 *	1881	1.09	1	0.10	1.95	0.01
8	84	0.05	0	0.00	0.00	0.00
9	161	0.09	0	0.00	0.00	0.00
10	680	0.39	0	0.00	0.00	0.00
11	197	0.11	0	0.00	0.00	0.00
12	1918	1.11	0	0.00	0.00	0.00
13	85	0.05	0	0.00	0.00	0.00
14	2870	1.67	0	0.00	0.00	0.00
15	1760	1.02	0	0.00	0.00	0.00
16 *	1822	1.06	30	2.93	281.47	1.21
17 *	6828	3.96	174	17.01	9091.72	38.96
18 *	365	0.21	14	1.37	24.06	0.10
19 *	25,169	14.60	597	58.36	8975.52	38.46
20	411	0.24	0	0.00	0.00	0.00
21 *	13,829	8.02	132	12.90	2436.70	10.44
22 *	1589	0.92	1	0.10	7.84	0.03
23 *	17,056	9.90	41	4.01	823.38	3.53
24 *	1553	0.90	3	0.29	85.12	0.36
25 *	2937	1.70	7	0.68	63.38	0.27
26 *	704	0.41	4	0.39	106.85	0.46
Total	172,368	100	1023	100	23,334.38	100.00

). Six of these represent 87.41% of Ecuador’s classified continental territory and account for 94.13% of the plantations analysed (963 plantations). Category 1 accounts for 43.15% of the total and extends exclusively to the Amazonian provinces (Sucumbíos, Orellana, Pastaza, Napo, and Morona Santiago). Category 3, covering 7.78% of the total area, is distributed across the provinces of the Littoral region (Esmeraldas, Santo Domingo), the foothills of the Andean region (Pichincha), and the Amazon region (Sucumbíos, Orellana, Napo, Pastaza, and Morona Santiago). Category 17, which accounts for 3.96%, is also distributed among provinces of the Littoral region (Santo Domingo and Los Ríos). Category 19, with 14.60%, is distributed throughout the provinces of the Littoral region (Esmeraldas, Manabí, Los Ríos, and Guayas). Category 21, with 8.02% of the area, is fragmented and found in the foothills of the Andean provinces (Cotopaxi, Bolívar, Chimborazo, Cañar) and in other parts of the provinces of the Littoral region (Esmeraldas, Manabí, Los Ríos, Guayas, El Oro). Finally, Category 23, with 9.0%, extends along the western margin of several provinces in the Littoral region (Esmeraldas, Manabí, Santa Elena, El Oro). The remaining seven categories with teak present represent 6.30% of the continental territory and 5.87% of the plantations (60). Among the categories mentioned, categories 17, 19, and 21 have the highest number of plantations: 174, 597, and 132, respectively (Table 3). These three categories account for 88.27% of the sampled plantations.

The significant variable values within the ecogeographic categories where teak is found that we would highlight are as follows: Category 1: Very high annual rainfall (3119 mm), low seasonality (18.3%), and warm temperatures (30.7 °C). Soil shows moderate bulk density (1.35 g/cm³) and cation exchange capacity (10.4 cmol/kg); water capacity is low (9.7%). The area sits at a moderate elevation (298.7 m) with light winds averaging 1.04 m/s. Category 3: High rainfall (3043 mm) with greater seasonal variation (26%) and warm temperatures (30.6 °C). Soil here has lower density (1.24 g/cm³) but higher cation exchange (14.8 cmol/kg) and water retention (11.0%), with a much higher saturated water content (47.9%). Slightly lower elevation (257.5 m) and slightly stronger winds (1.1 m/s annually). Category 17: Moderate rainfall (2584 mm) but much higher seasonality (84.2%); warmest month at 30.3 °C. Soil has a moderately high bulk density (1.26 g/cm³), a high cation exchange capacity (16.8 cmol/kg), and a high water content (47.6%). Lower elevation (209 m) and moderate winds (1.3 m/s annually). Category 19: Moderate annual rainfall (1363 mm) with very high seasonality (102.8%), the warmest month (30.7 °C), dense (1.40 g/cm³) and nutrient-rich soil (19.1 cmolc/kg) but lower water content (9.8%). The area is at a lower elevation (138 m) with brisk winds (annual average of 1.7 m/s). Category 20: Slightly more rainfall (1647 mm), high seasonal variation (97.21%), warm temperatures (30.5 °C), and soil with slightly lower bulk density (1.29 g/cm³) but more cation exchange (22.93 cmolc/kg) and water capacity (10.7%). Elevation is a bit higher (152 m), and wind is moderate (1.6 m/s). Category 23: Much drier (578 mm annual rain), extreme seasonality (113.8%), very little rain in the driest quarter (20.9 mm), but soil remains dense (1.4 g/cm³) and nutrient-rich (17.2 cmol/kg). Sits at 169 m elevation with the strongest winds (2.5 m/s annually).

The Kruskal–Wallis test revealed statistically significant differences (p ≤ 0.001) for all variables evaluated across the 13 ecogeographic categories where teak presence, indicating pronounced environmental heterogeneity among them. However, Dunn’s post hoc analysis with Holm correction showed that, despite these overall differences, some variables display similar values between specific categories, i.e., they did not present statistically significant differences (p ≥ 0.05), suggesting the existence of shared environmental conditions in specific categories (Table S13).

3.2. Predictive Model Validation and Importance of the Variables

The MaxEnt model showed a strong fit, with excellent accuracy (AUC), achieving a training AUC of 0.933 (range across 10 replicates: 0.931–0.935) and a test AUC of 0.924 (range across 10 replicates: 0.907–0.945), with a standard deviation of 0.012. The “true skill statistic” (TSS) was 0.773 (Table S14). The model identifies reliable potential areas for teak cultivation in Ecuador under current environmental conditions (Figure 4a).

Considering the percentage contribution and importance after permuting the predictor variables (

Table 4. Percentage contributions and permutation importance of climatic, geographical and geophysical variables in the predictive model.

Variable	Percent Contribution	Permutation Importance
Precipitation seasonality (bio_15)	42	37.8
Elevation (alt)	34.4	35.2
Annual precipitation (bio_12)	8.4	7.2
September wind speed (wind_9)	5.3	4.5
July wind speed (wind_7)	1.9	2.1
Precipitation of the driest quarter (bio_17)	1.8	1.2
August wind speed (wind_8)	1.6	5.4
Maximum temperature of the warmest month (bio_5)	1.1	0.2
Annual wind speed (wind_annual)	1	1
December wind speed (wind_12)	0.6	1.3
Cation exchange capacity of soil in cmolc/kg—topsoil (t_cecsol)	0.5	0.6
November wind speed (wind_11)	0.5	0
October wind speed (wind_10)	0.4	2.9
Available soil water capacity (volumetric fraction) for h2—topsoil (t_awc2)	0.3	0.3
Bulk density (fine earth) in kg/m3—topsoil (t_bulk_dens)	0.1	0.3
Saturated water content (volumetric fraction) for tS—topsoil (t_awcts)	0.1	0.1

), we highlight four of the 16 variables included in the predictive model: precipitation seasonality (42.0%), altitude (34.4%), annual precipitation (8.4%), and wind speed in September (5.3%). However, other variables showed permutational importance in the Jackknife analysis, such as wind speeds in August and October (Table 4, Figure 4b, Figures S4 and S5).

The probability of presence of the species, with values greater than 50% (Figure 5), is associated with the ranges of the following variables: precipitation seasonality (9.8–53.7%) (Figure 5a), elevation (25–131 m a.s.l.) (Figure 5b), annual precipitation (1089–2518 mm) (Figure 5c) and September wind speed (1.4–1.7 m s⁻¹) (Figure 5d).

3.3. Identification of Potential Areas for Teak Exploitation in Ecuador

The habitat distribution model provides a framework for identifying zones with varying probabilities of success in teak timber exploitation across Ecuador (Figure 6,

Table 5. Area in km² of the different probability levels by province and region.

Province/Probability	Null	Low	Moderate	Elevated	High	Total
Littoral region	30,360.00	12,065.04	11,565.20	9576.21	6600.43	70,166.90
El Oro	4577.26	699.52	387.91	66.82	0.00	5731.51
Esmeraldas	9425.72	3286.10	2591.42	583.03	0.00	15,886.27
Guayas	5470.19	1805.00	2099.25	2866.64	3029.22	15,270.30
Los Ríos	359.57	1041.79	1289.69	1746.85	2765.29	7203.20
Manabí	5615.92	4079.47	4437.82	4002.36	797.57	18,933.13
Santo Domingo	1898.45	824.10	513.22	202.03	8.35	3446.15
Santa Elena	3012.89	329.06	245.90	108.50	0.00	3696.35
Andean region	57,282.29	1313.45	795.88	578.49	137.39	60,107.49
Azuay	7985.57	132.88	168.42	38.79	0.00	8325.67
Bolívar	3442.78	148.52	181.40	121.98	50.18	3944.86
Cañar	2726.90	51.37	101.35	260.57	6.77	3146.95
Carchi	3528.73	0.00	0.00	0.00	0.00	3528.73
Cotopaxi	5634.76	139.93	149.84	104.15	80.45	6109.12
Chimborazo	6469.30	7.81	4.13	19.38	0.00	6500.62
Imbabura	4574.16	9.31	0.00	0.00	0.00	4583.48
Loja	10,553.53	472.74	18.14	0.00	0.00	11,044.40
Pichincha	8980.77	350.89	172.59	33.62	0.00	9537.87
Tungurahua	3385.78	0.00	0.00	0.00	0.00	3385.78
Amazonian region	115,788.93	692.81	8.35	0.00	0.00	116,490.08
Morona Santiago	24,012.17	0.00	0.00	0.00	0.00	24,012.17
Napo	12,542.42	0.00	0.00	0.00	0.00	12,542.42
Pastaza	29,628.77	0.00	0.00	0.00	0.00	29,628.77
Zamora Chinchipe	10,532.66	0.00	0.00	0.00	0.00	10,532.66
Sucumbíos	17,787.62	317.86	0.00	0.00	0.00	18,105.48
Orellana	21,285.29	374.94	8.35	0.00	0.00	21,668.59
Total	203,431.20	14,071.29	12,369.43	10,154.70	6737.83	246,764.47

). Analysing 1023 teak plantations reveals that they currently encompass 23,334.38 hectares (23.33 km²). Nevertheless, model predictions indicate significant potential for future expansion. The Littoral region is identified as the most promising area for the development of teak plantations, with an anticipated total of 16,176.64 km². This considerably surpasses the potential of the Andean region, which is limited to 715.88 km², and the Amazonian region, where no comparable areas have been identified. Collectively, these figures amount to 16,892.53 km², accounting for 6.85% of the mainland territory of Ecuador.

Provinces with predicted potential areas (Table 5), in decreasing order, are Guayas (5895.86 km²), Manabí (4799.93 km²), Los Ríos (4512.14 km²), Esmeraldas (583.03 km²), Cañar (267.34 km²), Santo Domingo (210.38 km²), Cotopaxi (184.60 km²), Bolívar (172.16 km²), Santa Elena (108.50 km²), El Oro (66.82 km²), Azuay (38.79 km²), Pichincha (33.62 km²) and Chimborazo (19.38 km²). It is noteworthy that all areas predicted to have high and elevated levels are located far from the coastline (Figure 6).

Areas forecasted as having moderate level (Table 5) cover 12,369.43 km², representing 5.01% of the Ecuador’s continental area, which mainly belongs to the Littoral region (11,565.20 km²), followed by the Andean region (795.88 km²) and slightly in the Amazonian region (8.35 km²).

The integration of ecogeographic characterisation with the potential distribution prediction model (

Table 6. Ecogeographic categories with their probability levels and proportion of territory for the presence of teak. Categories with teak presence are marked with an asterisk (*).

Probability/Categories	Null		Low		Moderate		Elevated		High		Total
	Cells	%	Cells	%	Cells	%	Cells	%	Cells	%
1 *	73,594.00	61.44	766.00	4.78	12.00	0.09	6.00	0.05	0.00	0.00	74,378.00
2	2.00	0.00		0.00	0.00	0.00	0.00	0.00	0.00	0.00	2.00
3 *	12,237.00	10.22	672.00	4.19	328.00	2.34	177.00	1.47	0.00	0.00	13,414.00
4	1744.00	1.46	195.00	1.22	49.00	0.35	6.00	0.05	0.00	0.00	1994.00
5	339.00	0.28	49.00	0.31	25.00	0.18	3.00	0.02	0.00	0.00	416.00
6	240.00	0.20	25.00	0.16	0.00	0.00	0.00	0.00	0.00	0.00	265.00
7 *	1634.00	1.36	101.00	0.63	134.00	0.96	12.00	0.10	0.00	0.00	1881.00
8	75.00	0.06	9.00	0.06	0.00	0.00	0.00	0.00	0.00	0.00	84.00
9	161.00	0.13	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	161.00
10	680.00	0.57	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	680.00
11	197.00	0.16	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	197.00
12	1918.00	1.60	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	1918.00
13	85.00	0.07	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	85.00
14	2870.00	2.40	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	2870.00
15	1760.00	1.47	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	1760.00
16 *	68.00	0.06	137.00	0.85	237.00	1.69	433.00	3.59	1007.00	13.75	1882.00
17 *	499.00	0.42	1701.00	10.61	1917.00	13.67	1508.00	12.49	1203.00	16.42	6828.00
18 *	107.00	0.09	109.00	0.68	115.00	0.82	25.00	0.21	9.00	0.12	365.00
19 *	2351.00	1.96	4591.00	28.63	6225.00	44.38	7954.00	65.89	4048.00	55.26	25,169.00
20	212.00	0.18	143.00	0.89	56.00	0.40	0.00	0.00		0.00	411.00
21 *	3684.00	3.08	3764.00	23.47	3536.00	25.21	1813.00	15.02	1032.00	14.09	13,829.00
22 *	896.00	0.75	553.00	3.45	113.00	0.81	3.00	0.02	24.00	0.33	1589.00
23 *	10,484.00	8.75	2311.00	14.41	915.00	6.52	116.00	0.96	3.00	0.04	13,829.00
24 *	1489.00	1.24	91.00	0.57	9.00	0.06	0.00	0.00		0.00	1589.00
25 *	1850.00	1.54	716.00	4.46	355.00	2.53	16.00	0.13		0.00	2937.00
26 *	599.00	0.50	105.00	0.65	0.00	0.00	0.00	0.00		0.00	704.00
Total	119,775.00	100.00	16,038.00	100.00	14,026.00	100.00	12,072.00	100.00	7326.00	100.00	169,237.00

) indicates that Category 19 encompasses 55.26% of the area with a high likelihood of teak presence, followed by Category 17 (16.42%) and Category 21 (14.09%). Notably, Category 16, which initially occupies a limited area and supports only a few plantations, accounts for 3.75%. Among the categories with a high probability of teak occurrence, Category 19 constitutes the largest proportion at 65.89%, followed by Category 21 at 15.02% and Category 17 at 12.49%. Furthermore, Category 1, which constitutes a significant portion of the studied area at 43.49%, exhibits no high suitability, only a very low suitability of 0.05%for the exploitation of this timber resource (Table 6).

The ecogeographic characteristics of territories with a high likelihood (Table S16) of teak presence include a precipitation seasonality of 27.0%, ranging from 1 to 79%; an average elevation of 66.4 m a.s.l., with values between 16 and 189 metres; September wind speeds of 1.59 m/s, from 1.30 to 2.30 m/s; and annual rainfall of 1707.6 mm, spanning 746 to 2665 mm. Conversely, areas with a moderate probability (Table S16) exhibit a precipitation seasonality of 32.1%, ranging from 1 to 86%; an elevation of 96 m a.s.l. (ranging from 10 to 315 metres); September wind speeds averaging 1.75 m/s, with a range of 1.30 to 2.60 m/s; and annual rainfall of 1535.1 mm, between 663 and 2830 mm. The variability in these ranges, even at their mean values, remains relatively small, indicating these conditions could be considered optimal for forest exploitation success rates exceeding 60% for this timber resource.

4. Discussion

Georeferenced information on teak plantations in Ecuador, utilising tools such as CAPFITOGEN, provided comprehensive access to a wide range of environmental variables, including climatic, edaphic, and geophysical factors. However, not all variables significantly contribute to the study’s objective: to identify optimal areas for exploitation in Ecuador. In this regard, the importance of proper variable selection has been documented by [37], who show that this step can modify the prediction of area occupancy by up to 2000% and estimators of extinction risk in plant species by 50%. Therefore, the tool used offered an advantage by integrating information from three primary sources and applying robust statistical analysis for variable selection.

The presence of teak within specific ecogeographic categories signifies conducive conditions for its proliferation. However, the MaxEnt model has identified areas of establishment with varying degrees of success. This model exhibited high predictive accuracy, aligning with findings from studies conducted in India, Myanmar, Laos, and Thailand [64,65]. Similar outcomes have been documented in research involving other forest species such as mahogany (Swietenia macrophylla King) [66], capuli (Prunus serotina Ehrh. subsp. capuli (Cav.) McVaugh) [25], and yagual (Polylepis incana Kunth) [67].

Precipitation seasonality is essential in managing teak plantations because it influences the timing of silvicultural activities throughout the year [3]. During the rainy season, vegetative growth accelerates, necessitating the application of fertilisers and weed control to optimise resource availability and minimise competition for nutrients. Conversely, in the dry season, conditions are more suitable for pruning and thinning, as the risk of pest and disease spread through tree wounds decreases [68].

Beyond rainfall volume, the importance of precipitation seasonality has been emphasized, noting that a dry season of 3 to 5 months is considered necessary [69], as the species requires prolonged dry periods for wood hardening [70]. These conditions do not occur in the Ecuadorian Amazon, where the rainy season persists year-round, with a peak in rainfall in April and May [45]. These precipitation thresholds explain why the Amazon region was excluded from high-suitability predictions, as its annual rainfall exceeds the teak optimum range (2500–3484 mm).

Altitude is closely linked to environmental factors such as temperature and nutrient availability, which tend to decrease with increasing altitude [71]. It could explain the absence of teak above 600 m a.s.l., as our field observations suggest that the species thrives in flat and gently rolling terrain with stable temperatures. The predictive model indicates that the optimal growth zones, at a high probability (level 5, p > 0.8), may extend up to 315 m a.s.l. These values coincide with the altitudinal range reported in Ecuador, approximately 300 m a.s.l. [8]. In other Central American countries, such as Costa Rica and Panama, teak plantations are usually located below 300 m a.s.l. [72] but can reach up to 600 m a.s.l. [73]. However, these figures show a narrower range than its native range, which extends from sea level to 1200 m [74].

Although the optimal altitude values for teak growth have been established between 25 and 131 m above sea level (Figure 5), its cultivation could be limited by temporary waterlogging after the rainiest periods (winter). These circumstances have been observed in the provinces of Guayas and Los Ríos [44], located below 20 m above sea level, where no teak plantations have been recorded, since teak does not tolerate waterlogging [75,76].

Altitude has been recognised as a significant variable in modelling studies of the distribution of teak and other forest species, as is the case for the model under climate change-induced scenarios for teak [64]. Similarly, it has been utilised to predict the distribution of the lacquer tree (Toxicodendron vernicifluum (Stokes) F. A. Barkley) [61] and the potential distribution of the Chinese fir (Cunninghamia lanceolata (Lamb. Hook) [32].

The extracted annual precipitation values correspond closely with previous modelling studies on teak within its native regions [3,64]. This attribute is also a significant consideration in research on other tree species, such as studies on the adaptability of Japanese larch (Larix kaempferi (Lamb.) Carrière) [77] and Delavay soapberry (Sapindus delavayi (Franch.) Radlk.) [78]. The model delineated high-probability zones as those areas experiencing a mean annual precipitation of 1707.6 mm (ranging from 756 to 2655 mm), figures which approximate those reported in Ecuador and suggest that the species prospers within precipitation ranges of 1500 to 3000 mm [8], or, in its native habitat, between 1200 and 2500 mm [70]. Nevertheless, the documented variation range for Costa Rica is broader, spanning 800–3689 mm [69]. Consistent occurrence of teak in certain areas within the Ecuadorian Amazon implies that its growth may be constrained by excessive annual rainfall, a critical factor influencing its development.

In the current study, in addition to the three most significant variables identified through ecogeographic characterisation, the prediction model incorporated September wind speed as a pertinent contributing factor. Teak exhibits high wind resistance due to its extensive taproot system [79]; however, this does not imply that the species’ yield is unaffected by wind. Recent research on other forest species has demonstrated that wind speed can adversely affect wood volume [80] and potentially influence tree mortality rates [81]. Consequently, this variable which, according to [45], signifies a transition period between dry and rainy seasons during the last four months of the year, could impose limitations on teak productivity in certain regions of Ecuador.

Wind has been recognised as a variable to consider in analyses, particularly during the dry season from July to December. As an abiotic factor, it significantly influences the morphology and physiology of trees, including teak [82,83]. Studies conducted across various forest species have demonstrated that wind exposure can modify tree structure by increasing stem diameter and secondary growth, while reducing height [82,84] or diminishing leaf size [85,86]. Moreover, it promotes the development of dense, asymmetrical roots [87], thereby enhancing mechanical stability under prolonged loads. However, strong winds can also affect soil properties, reduce fertility, and damage plantations [88,89]. The recorded wind speeds in the area are insufficient to induce structural damage. However, at specific intervals following planting, they may impact physiological responses to water stress, thereby influencing growth parameters by reducing chlorophyll content, stomatal conductance, and leaf water content [90].

The MaxEnt model identified four significant contributing variables: precipitation seasonality, altitude, annual precipitation and September wind speed. Three of these variables, except for wind speed in September, were also recognised as significant in previous models examining teak distribution in Asia [64,65,91] in the context of climate change impact assessments. However, this finding contrasts with the research conducted in Central India by [56], which identified isothermality and temperature seasonality as the most significant variables.

The predictive model identified several regions of high ecological suitability within the provinces of Guayas, Los Ríos, and Manabí, corresponding to categories 17, 19, and 21 as determined through ecogeographic characterization, where most of plantations are presently located. The insights derived from this analysis serve as a valuable reference for making well-informed decisions regarding the introduction and establishment of teak in Ecuador. This approach contrasts with previous land use initiatives that lacked a solid foundation, such as the initiative involving the introduction of abaca (Musa textiles Née) in the southern part of the Ecuadorian Littoral region, specifically in El Oro, Guayas, and Los Ríos, which was later successful in the provinces of Santo Domingo and Esmeraldas [92].

As a comparatively recent species introduced within the nation, teak has undergone swift expansion in its cultivated regions, increasing by nearly 20,000 hectares between 2011 and 2015, primarily attributable to governmental economic incentives [93]. Nevertheless, its prolonged harvesting cycle, which begins after 20 years [94], is regarded as a limiting factor for wider adoption. This is in contrast to the most extensively cultivated food crop, cacao (Theobroma cacao L.), which covers approximately 609,800 hectares [95] and reaches market readiness after two years [96]. Similarly, the forest species balsa (Ochroma pyramidale (Cav.) Urb.) encompasses approximately 15,000 hectares and has a growth cycle of three to five years [97]. According to anecdotal reports, some producers perceive the areas designated for teak plantations as residual land on their farms, serving as a form of long-term savings for the producer’s old age (“un ahorro para la vejez”). Alternatively, others regard it as a primary economic activity.

Although the model identifies 673,783 ha as areas with a high probability of developing teak, only approximately 50,000 ha are currently documented [10]. The final decision on which crop to establish does not rely solely on the land’s agroecological capacity; in practice, market demand and secure commercial agreements are prioritized. Profitability and marketability directly influence this decision [98]. Furthermore, due to phytosanitary issues and adverse environmental factors, such as pests, diseases, and climatic phenomena, which affect crop sustainability [99], decreased productive areas have been reported. In addition to these unfavourable conditions, factors such as price volatility and international competition [99,100] contribute to the abandonment of certain production areas despite their agroecological potential.

Identifying areas with high and moderate potential for teak exploitation can be a valuable tool for land-use planning, especially when issuing permits for establishing and utilising forest resources with introduced species, as seen in this case. This approach would not only increase the supply of cultivated timber but also enhance foresters’ profitability and help reduce pressure on native forests, given the country’s high deforestation rate [101]. Under the Ecuadorian legal framework, the Ministry of Environment, Water, and Ecological Transition (MAATE) is responsible for regulating these activities. According to Ministerial Agreement No. 139, the state is tasked with planning national development, ensuring an equitable distribution of resources, and safeguarding natural heritage as part of its commitment to “Buen Vivir” (Good Living) [14].

Regarding environmental services, teak plantations can help mitigate climate change in several ways. In Ecuador, ref. [102] demonstrated that teak plantations have a greater capacity for carbon sequestration than degraded grasslands, emphasising their importance as an alternative land use with additional environmental benefits. Similarly, studies conducted in India by [103,104] also highlight the potential of this species to store significant amounts of carbon over time. However, it has also been observed that monocultures of teak plantations tend to perform less effectively than mixed, natural, or comparable forest systems in terms of long-term soil carbon storage and maintaining biodiversity and resilience [102,105,106,107,108,109]. Therefore, their value for mitigating climate change is context-dependent and may be limited or counterbalanced by broader environmental factors.

Teak in Ecuador has become an important source of forestry income in recent years, alongside other species [110,111], mainly due to the 2019–2030 forest restoration programme [9]. However, the rapid expansion of plantations in recent years, sometimes without adequate planning, conflicts with the presence of other crops and contributes to deforestation [111]. A comparison of the potential teak cultivation map presented in this study (Figure 6) with land use maps [112] shows overlaps with maize crops in the provinces of Manabí, Guayas, and Los Ríos; cocoa in Manabí, Los Ríos, southern Santo Domingo, and southeastern Guayas; and bananas in Los Ríos and southeastern Guayas. Proper regulation for the coexistence of different agroforestry systems, which may have incompatible and unequal interests among producers, requires well-grounded resource management to maintain a balance between the economy and natural conservation [113,114].

The broad designation of the Amazon region as “the lungs of the world” [115] is not trivial. The potential of the Ecuadorian Amazon region for teak production is negligible or, at best, low, even though some areas have been classified as Category 3 by their ecogeographical characteristics. The region’s potential lies in its native timber species, not in introduced ones. The management of existing plantations should consider Close-to-Nature practices to minimize the adverse effects of monoculture forest management [116] and maintain the ecosystem’s connectivity [117].

Based on the climatic, edaphic, and geophysical requirements identified in this study, which ecogeographically characterized the Ecuadorian territory and model the distribution of teak, we recommend further research that includes climate change scenarios and evaluation trials in the three main ecogeographic categories, which coincide with the two areas with the highest probabilities of success for the target species. These trials would allow us to assess teak’s response to phytosanitary and productivity challenges and to optimize its management and utilisation in the local context. Ultimately, the aim would be to establish strategic guidelines that enable the development of optimal cultivation zones, anticipate the impacts of climate change, validate comprehensive development programmes, and plan land use.

5. Conclusions

This study has identified optimal areas for teak cultivation and provided a basis for decision-making in the forest management of T. grandis in Ecuador by integrating CAPFITOGEN tool results as input data into the MaxEnt species distribution model. This integration introduces a methodological approach that has not been used before in the analysis of Ecuadorian forest species, enabling a detailed characterisation of ecogeographic variability and the creation of spatial projections that enrich our understanding of teak’s ecological requirements. High-quality georeferencing (90.83% average accuracy) offered a reliable foundation for spatial analyses, while the MaxEnt model (AUC = 0.924) showed excellent predictive capacity for identifying suitable planting areas.

The study revealed that Ecuador’s territory can be classified into 26 distinct ecogeographic categories below 600 m a.s.l., with teak naturally occurring in 13 of these categories. However, three categories (17, 19, and 21) appeared as predominant, collectively supporting 88.27% of existing plantations and representing the most favourable conditions for teak cultivation. These categories are characterised by specific environmental parameters: precipitation seasonality (6–56%), elevation (16–189 m a.s.l.), September wind speed (1.3–1.7 m s⁻¹), and annual precipitation (1089–2518 mm). The identification of precipitation seasonality (42.0% contribution) and elevation (34.4% contribution) as the most influential variables coincides with teak’s physiological requirements for alternating wet and dry periods, which are essential for the development of optimal wood quality. This finding confirms the species’ current distribution patterns and explains why Amazonian regions, despite having suitable temperatures, are unsuitable due to excessive year-round rainfall (exceeding 2500–3484 mm annually).

The predictive model revealed substantial, yet unexplored, potential for teak cultivation in Ecuador. While current plantations occupy approximately 233.34 km², the model identifies 16,892.53 km² (6.85% of Ecuador’s continental territory) as suitable for teak exploitation, with 6737.83 km² classified as high-probability areas and 10,154.70 km² as elevated-probability areas. The Littoral region emerges as the most promising, with a predicted potential area of 16,176.64 km², significantly exceeding the Andean region’s 715.88 km². Province-wise analysis indicates that Guayas (5895.86 km²), Manabí (4799.93 km²), and Los Ríos (4512.14 km²) present the greatest expansion opportunities, aligning with current successful plantation distributions and confirming the model’s practical validity.

Identifying ecologically suitable areas is a critical initial step in strategic forest management, but climate change poses fundamental challenges to the long-term viability of plantations. Precipitation seasonality, identified as the dominant driver of teak suitability (42% model contribution), is predicted to shift significantly across Ecuador under future climate scenarios [118,119]. By 2050 and 2070, models predict that seasonal rainfall patterns will either intensify or weaken in specific regions, potentially expanding suitable zones in some areas while rendering conditions marginal in currently optimal plantations [119].

Future research should employ ensemble climate projection models to evaluate teak suitability under 2050 and 2070 climate conditions, following methodological approaches demonstrated for Ecuadorian forest ecosystems [120]. Such projections, forecasted for 2050 and 2070, should utilize CMIP6 general circulation models under moderate (RCP 4.5) and high-emission (RCP 8.5) scenarios to assess directional shifts in precipitation seasonality, temperature ranges, and optimal elevation zones. Spatial analysis integrating climate projections with current ecogeographic characterization will enable (1) identification of climate-resilient teak plantation zones; (2) anticipation of pressure zones where current suitability may decline; and (3) development of adaptive management strategies incorporating species diversification or assisted migration of teak germplasm adapted to future climatic conditions. The Amazon region, despite current marginality for teak cultivation, merits specific attention given evidence of greater climatic stability [121] for native forest species through 2070, suggesting potential value in integrating native timber species into restoration activities in this region rather than pursuing large-scale teak expansion [122].

While ecological suitability provides necessary guidance, the economic sustainability of teak expansion depends on management approaches that enhance rather than diminish ecosystem services. Ecuador’s experience with fine flavour cocoa production demonstrates that agroforestry systems integrating native shade trees are not only economically competitive but command market premiums [123,124]. Applied to teak plantations in the high-priority ecogeographic categories (17, 19, 21), shade-tree integration strategies could generate multiple products (teak timber, shade-tree fruit/timber, honey, etc.) while sequestering an estimated additional 0.5–2 Mg C ha⁻¹ year⁻¹ above teak monoculture [124]. Silvicultural research integrating native timber and nitrogen-fixing species already successful in Ecuadorian agroforestry systems should evaluate their compatibility with teak cultivation, establishing the technical and economic foundation for biodiversity-enhanced plantation design [116]. Such integration would align teak expansion with Ecuador’s commitment to ecosystem services provision while maintaining farmer profitability.

Beyond Ecuador’s borders, the CAPFITOGEN-MaxEnt integration addresses a widespread methodological gap in Amazon regional forest management. Peru, Colombia, and Bolivia collectively manage approximately 100,000 hectares of teak plantations while facing similar land-use conflicts between timber production, agricultural expansion, and native forest conservation [125,126]. The ecogeographic variables that determine teak suitability in Ecuador precipitation seasonality, elevation, water availability, wind exposure) reflect the species’ physiological requirements that persist across national borders. The global availability of climatic, topographic, and soil data enables the direct application of this framework in neighbouring countries’ land-use planning. Additionally, a comparative regional analysis could reveal whether teak suitability patterns change systematically across the Amazon in response to latitude and longitude gradients. Such expansion would support Amazonian subregional coordination on climate-vulnerable species management and aid in the development of regionally adapted teak germplasm strategies across Peru, Colombia, and Ecuador. Moreover, identifying precipitation seasonality as the primary driver of teak distribution has broad implications for other drought-tolerant timber species currently cultivated across the Amazon, suggesting that this methodological approach could help optimize plantation placement for various commercial forest species across the region.

On the other hand, the Littoral region, which has been extensively deforested for agriculture and livestock, shows both ecological suitability and established economic success for developing teak plantations. The strategic allocation of teak expansion to 16,892 km² of suitable areas in the coastal region thus represents an efficient use of resources in line with biogeographical analysis: it preserves ecosystem service values, alleviates pressure on Amazon conservation through market-driven alternatives in already modified landscapes, and maximizes the economic profitability of forestry investments. Biodiversity conservation and timber production can be achieved through complementary, rather than competitive, land-use strategies.

The substantial gap between current cultivation (approximately 50,000 ha) and identified potential (673,783 ha) represents a significant economic opportunity. Given that Ecuador’s teak sector generated US$200 million between 2019 and 2023 from relatively limited production, strategic expansion into identified optimal zones could substantially increase economic returns while contributing to the national reforestation objectives outlined in the National Forest Restoration Plan 2019–2030.

The integration of detailed ecogeographic characterisation with predictive models provides government agencies, particularly the Ministry of Environment, Water, and Ecological Transition (MAATE), with robust tools for sustainable forest management decisions that balance economic development with environmental conservation objectives.