A Comparison of Two Methodological Approaches for Determining Castor Bean Suitability in Chile

Castor bean (Ricinus communis L.) contains ricinoleic acid, making it one of the world’s most important oil-seeds. There are few studies on this species in Chile, despite its potential as an industrial crop. This study evaluated two methodologies (simplistic and presence-species) for determining the aptitude of land for growing castor beans, both of which use climatic information. The simplistic and presence-species methodologies identified 27.89 and 13.19 million ha, respectively. The most important difference between both methodologies was that the mean minimum annual temperature (TNA) was −8.0 ◦C in the simplistic method, meaning that some areas in the southernmost regions of Chile (Aysén and Magallanes) should be able to grow the plant. Therefore, TNA = 8.0 ◦C was selected, and the zonation by simplistic methodology was updated. Consequently, both zonations showed similar results, although the presence-species method included northern coastlines, precisely where castor bean has been recorded, while the simplistic method did not. Finally, both methodologies determined the best condition to be central-south Chile, between the Maule and Araucanía regions, even though castor bean presence has only been recorded up to the Maule region. These regions have a huge potential to establish castor beans, but more information about agronomic practices is necessary for its development in Chile.


Introduction
Castor bean (Ricinus communis L.) is extensively distributed in tropical conditions, even though it has also been introduced under subtropical conditions [1]. This plant has a perennial behavior in nature, but can sometimes be treated as an annual crop when cultivated, especially in semi-arid and arid conditions [2]. In Chile, its presence has been registered between the Arica y Parinacota and Maule regions [3]. Further, it has recently been reported that the oil content in Chilean castor bean accessions ranged from 45.7 to 54.0% and that it was greater than 50% in seven Chilean accessions [3]. Despite what is already known, there is little information about castor bean in the country, especially from an agronomic viewpoint.
On the other hand, it is a very important industrial crop due to its unique nonedible seed-oil [4,5] and the fact that it can grow in marginal land, where crops for food and feed production usually grow poorly [5]. Castor bean seeds contain 28-59% seed-oil, and ricinoleic acid (12-hydroxy-9-cisoctadecenoic acid) is its most abundant fatty acid (79-92%) [6,7]. Castor oil production reaches 1.8 million Table 1. Parameters for determining agro-climatic zones, according to the simplistic methodology of Falasca et al. [23].

Determining Land Aptitude for Castor Bean with the Presence-Species Method
The presence-species methodology, developed by Román-Figueroa et al. [24], was employed as a second approximation for determining castor bean plant suitability in Chile. Briefly, an international bibliographic review was made to determine places where castor bean plants grow, and geographic coordinates and climatic information were registered. Monthly average minimum temperature ( • C), monthly average maximum temperature ( • C), relative average monthly humidity (%), and monthly precipitation (mm) were considered and registered. Moreover, water-deficit (WD), degree days (DD), and potential evapotranspiration (ETp) were calculated as derived variables. More details on how WD, DD, and ETp were determined in the presence-species methodology can be found in recent literature [24] and [27].
Adaptable ranges for castor bean were subsequently determined for each climatic variable by scatterplots between the variables. Hydric zoning used WD because it includes precipitation and ETp, whereas thermal zoning considered maximum temperature of the warmest month (TMX), minimum temperature of the coldest month (TNJ), and DD variables.
Finally, climatic zoning was determined according to the land evaluation theory of Rossiter [28], where land suitability is obtained from land characteristics and crop requirements; thus, land is evaluated according to qualitative or quantitative approach [21,22,28]. Land suitability, at the national Agronomy 2020, 10, 1259 4 of 15 level, was determined through a qualitative approach, selecting factors associated with the climatic requirements of castor bean and their respective aptitude levels. Climatic factors were categorized according to the following levels of restriction: without restriction, mild restriction, moderate restriction, and restricted. The information was processed using ArcGIS ® 10 (Esri, Redland, CA, USA, http://www.esri.com/software/arcgis) and the method of decision rules using Boolean operators based on top-down logic.
Both simplistic and presence-species zonation methodologies were complemented with current land uses for the determination of limitations in land aptitude for castor bean growth, according to agro-climatic zonation methodology [28]. Limitations keep restriction levels of climatic factors or remove the land aptitude for crop establishment [27]. Urban areas, forests, wetlands, water bodies, snow, and glaciers, protected wildlife areas, and agricultural lands were considered as limitations removing the land aptitude in those uses.

Land Aptitude Determination Using the Simplistic Method
Chile has a total surface area of 75.29 million ha. From climatic-zonation using the simplistic methodology, it was found that 27.89 million ha (37.04% of the country's area) has some level of aptitude for castor bean growth ( Figure 1A). This area is localized between the Coquimbo and Magallanes regions, excluding the northernmost regions of Chile. Therefore, 37.04% of the national territory is suitable for growing castor bean. The southernmost regions of Chile have the greatest area that is suitable-23.65%, 14.91%, and 14.82% of the total surface area with aptitude registered in the Magallanes, Los Lagos, and Aysén regions, respectively-but the aptitude level in these regions was marginal due to temperature (4 in Table 2). The Los Ríos region had the greatest area with aptitude-93.34% of the Los Ríos region's total territory-according to the regional surface, although all land with aptitude was marginal due to temperature ( Table 2).
Agronomy 2020, 10, x FOR PEER REVIEW 4 of 15 evaluated according to qualitative or quantitative approach [21,22,28]. Land suitability, at the national level, was determined through a qualitative approach, selecting factors associated with the climatic requirements of castor bean and their respective aptitude levels. Climatic factors were categorized according to the following levels of restriction: without restriction, mild restriction, moderate restriction, and restricted. The information was processed using ArcGIS ® 10 (Esri, Redland, CA, USA, http://www.esri.com/software/arcgis) and the method of decision rules using Boolean operators based on top-down logic.
Both simplistic and presence-species zonation methodologies were complemented with current land uses for the determination of limitations in land aptitude for castor bean growth, according to agro-climatic zonation methodology [28]. Limitations keep restriction levels of climatic factors or remove the land aptitude for crop establishment [27]. Urban areas, forests, wetlands, water bodies, snow, and glaciers, protected wildlife areas, and agricultural lands were considered as limitations removing the land aptitude in those uses.

Land Aptitude Determination Using the Simplistic Method
Chile has a total surface area of 75.29 million ha. From climatic-zonation using the simplistic methodology, it was found that 27.89 million ha (37.04% of the country's area) has some level of aptitude for castor bean growth ( Figure 1A). This area is localized between the Coquimbo and Magallanes regions, excluding the northernmost regions of Chile. Therefore, 37.04% of the national territory is suitable for growing castor bean. The southernmost regions of Chile have the greatest area that is suitable-23.65%, 14.91%, and 14.82% of the total surface area with aptitude registered in the Magallanes, Los Lagos, and Aysén regions, respectively-but the aptitude level in these regions was marginal due to temperature (4 in Table 2). The Los Ríos region had the greatest area with aptitude-93.34% of the Los Ríos region's total territory-according to the regional surface, although all land with aptitude was marginal due to temperature ( Table 2).   There was 20.1 million ha of suitable land that was marginal due to temperature, i.e., 72.09% of the total land with some aptitude ( Table 2). On the other hand, there were 3.75 million ha (13.44%) of suitable land with a humid regime (1 in Table 2), and it was concentrated between the O'Higgins and Aracaunía regions ( Figure 1A). This category included the best conditions according to this zonation method because optimal and very suitable aptitude were not registered in Chile. In addition, no land with frost restriction was registered in Chile (5 and 7 in Table 2).

Land Aptitude Determination using the Presence-Species Method
There were 47 places where castor bean presence was registered and climatic information was available (Table A1). Climatic information (thermal and hydric) was obtained for each place. This information was used to establish the aptitude ranges in which castor bean can grow (Table 3). Table 3. Thermal and hydric critical ranges for castor bean adoption in Chile.

Parameters
Aptitude Ranges Climatic-zonation found that 13.19 million ha in Chile (17.52% of the country's area) had some aptitude for castor bean production ( Figure 1B). There was no ideal area for castor bean production in Chile, as all areas had some thermic and hydric restrictions ( Figure 1B; Table 4). The Maule, Araucanía, and Coquimbo regions had the highest land concentration with some aptitude level, with 14.70%, 14.69%, and 13.87%, respectively, of the total land having climatic aptitude. However, the aptitude level was different in each of these regions. Coquimbo mainly included land with mild thermic and moderate hydric restrictions (10 in Table 4). In contrast, in the Maule region, the majority of land had moderate thermic and hydric restrictions (8 in Table 4), while that in the Araucanía regions had moderate thermic and mild hydric restrictions (9 in Table 4). On the other hand, the Biobío and Araucanía regions had the best condition for castor bean production in Chile. Both regions registered 122,586 ha (0.93% of the total land with some aptitude level) with mild thermic restrictions and without hydric restrictions (Table 4). Thirty point nine percent of the total land with some aptitude level had moderate thermic and hydric restrictions, and this was concentrated in the central valley of Chile ( Figure 1B).

Land Aptitude with Current Land Use as Limitations
Land aptitude in agro-climatic zonation, considering current land uses as limitations, found 4.29 and 4.59 million ha with some level aptitude under the simplistic and presence-species methods, respectively ( Figure 2; Table 5). The simplistic method showed a reduction of 84.62% in the total land with aptitude in comparison with climatic zonation ( Figure 1A; Table 2), while the presence-species method showed a reduction of 65.20% in the total land with aptitude in comparison with climatic zonation ( Figure 1B; Table 4).   Land suitable from agro-climatic zonation with the simplistic method was concentrated in the Los Lagos region with 1.01 million ha (23.44% of land with some aptitude level), followed by Los Ríos and Coquimbo regions with 493,132 ha (11.47%) and 460,712 ha (10.72%), respectively (Simplistic methodology in Table 5). A total area of 360,067 ha was registered with the best conditions for castor bean adaptation in the O'Higgins, Maule, Ñuble, Biobío, and Araucanía regions (1 in Table 5).
Land suitable from agro-climatic zonation with the presence-species method was concentrated in the Coquimbo region with 1.61 million ha (35.06% of land with some aptitude level), followed by Atacama and Los Ríos regions with 442,574 ha (9.62%) and 388,190 ha (8.44%), respectively (Presence-species methodology in Table 5). A total area of 11,851 ha was registered with the best conditions for castor bean adaptation in the Biobío and Araucanía regions (14 in Table 5).

Land Aptitude with −8.0 • C as TNA
Land that is suitable for castor bean production based on the simplistic methodology reached the southernmost regions of Chile (Aysén and Magallanes regions) (Figures 1A and 2A) because land with frost restriction was not registered (Table 5). These regions have the lowest crop production in the country [29] due to their polar climate [30], with its coldest month reaching as low as −5.0 and 5.0 • C [25]. Land in both regions is mainly used for silvopastoral systems [31] because it is extremely difficult to establish crops there, and, therefore, it would probably be difficult to establish castor bean plants there. The simplistic method applied by Falasca et al. [23] considered an annual mean minimum temperature of −8.0 • C, but there is no castor bean germplasm in Europe that could grow under these conditions. However, castor bean has only been registered in Mediterranean countries, such as Spain [16], Greece [4,5], and Italy [4], the winter minimum temperatures of which are 0.0-10.0 • C in Greece [32] and −3.0-11.0 • C in Spain [33]. It was found that castor bean plants can survive chilling stress, i.e., low non-freezing temperatures [7,34], but there is no information about their behavior below 0.0 • C.
Agro-climatic zoning was determined using the simplistic methodology, and the zonation parameters were changed to make TNA = 8.0 • C (Table 1) according to Patanè et al. [1], who determined that 8.0 • C could be a base temperature for castor bean growth. The southernmost regions, Aysén and Magallanes, did not show land aptitude (Figure 3), in contrast to the original zonation results ( Figure 1A); only the coastal zone in the Los Ríos and Los Lagos regions showed some level of aptitude (marginal area by temperature), probably because the sea coast influence acts as a thermic regulator that avoids very low minimum temperatures (frost) and high fluctuations in the daily temperature [35].
Agronomy 2020, 10, x FOR PEER REVIEW 9 of 15 the country [29] due to their polar climate [30], with its coldest month reaching as low as −5.0 and 5.0 °C [25]. Land in both regions is mainly used for silvopastoral systems [31] because it is extremely difficult to establish crops there, and, therefore, it would probably be difficult to establish castor bean plants there. The simplistic method applied by Falasca et al. [23] considered an annual mean minimum temperature of −8.0 °C, but there is no castor bean germplasm in Europe that could grow under these conditions. However, castor bean has only been registered in Mediterranean countries, such as Spain [16], Greece [4,5], and Italy [4], the winter minimum temperatures of which are 0.0-10.0 °C in Greece [32] and −3.0-11.0 °C in Spain [33]. It was found that castor bean plants can survive chilling stress, i.e., low non-freezing temperatures [7,34], but there is no information about their behavior below 0.0 °C. Agro-climatic zoning was determined using the simplistic methodology, and the zonation parameters were changed to make TNA = 8.0 °C (Table 1) according to Patanè et al. [1], who determined that 8.0 °C could be a base temperature for castor bean growth. The southernmost regions, Aysén and Magallanes, did not show land aptitude (Figure 3), in contrast to the original zonation results ( Figure 1A); only the coastal zone in the Los Ríos and Los Lagos regions showed some level of aptitude (marginal area by temperature), probably because the sea coast influence acts as a thermic regulator that avoids very low minimum temperatures (frost) and high fluctuations in the daily temperature [35].  Land that was appropriate for castor bean growth decreased by 53.21% when TNA was set to 8.0 • C, decreasing from 4.29 million ha (Simplistic methodology in Table 5) to 2.01 million ha (Table 6). This is mainly because this change excluded the southernmost regions and the inland and highland between the O'Higgins and Los Lagos regions (Figures 1A and 3). The largest change that occurred when the TNA was changed to 8.0 • C was observed on marginal land, where land for castor bean growth decreased in area by 81.51%, from 2.61 million ha to 483,093 ha. An area of 320,419 ha of land was made marginal by frost, and 421,982 ha became marginal by both frost and water deficit (5 and 7 in Table 6), including land registered in the interior valleys between the Coquimbo and Araucanía regions (Figure 3). Frost is one of the most consequential abiotic stresses in Chile, registering economic and productive losses of up to US $354 million [36]. Therefore, castor bean adaptation in Chile must be conditioned for frost occurrence probability, especially because of its tropical origin [7] and that these plants are normally sensitive to cold, i.e., they do not tolerate temperatures under 0 • C [37]. Table 6. Regional surface area with agro-climatic aptitude for castor bean production as a national oil-seed crop (ha), according to a slightly altered simplistic methodology (TNA = 8.0 • C). Chilean total area: 75.29 million ha.

Comparison of the Two Methodologies
Both methodologies found that hydric conditions had a large influence in Chile. Consequently, the coastal zone (determined by the presence-species methodology) was the only area in the northernmost region that showed aptitude ( Figure 2B). This can be explained by the low rains that were recorded in these regions, which resulted in 3 to 250 mm year −1 between the Arica y Parinacota region and the southern Coquimbo region [23]. On the other hand, the only hydric parameter that the simplistic methodology considered was precipitation, while the presence-species methodology considered WD, that included ETp and Pp. The WD is lower along the Chilean coast than inland due to the marine influence [25].
Furthermore, the land with the best conditions based on the presence-species methodology was between the Biobío and Araucanía regions ( Figure 2B), whereas, according to the simplistic methodology, that was between the Maule and Araucanía regions ( Figure 3). Therefore, both methods determined similar conditions as the best for castor bean adaptation. Castor bean is found as far as the Maule region [3], but it has not been recorded between the Ñuble and Araucanía regions. Therefore, there is an opportunity to establish castor bean in Chile as an industrial crop. However, precautions should be taken when growing castor bean in the best conditions for its establishment because it is not a native species, and, therefore, runs the risk of becoming invasive [38].
Currently, castor bean is registered between the Arica y Parinacota and del Maule regions [3]. Therefore, it could probably grow under mild thermic restrictions and moderate hydric restrictions ( Figure 2B) along the coastal line due to the humidity coming from the ocean in the northernmost regions and the fact that castor bean has been recorded in this area [3]. On the other hand, castor bean has also been registered in the Arica y Parinacota region and between the Atacama and O'Higgins regions in the inland area, but it is always associated with a water body or in farmland borders [3,39] (Figure 4).
Currently, castor bean is registered between the Arica y Parinacota and del Maule regions [3]. Therefore, it could probably grow under mild thermic restrictions and moderate hydric restrictions ( Figure 2B) along the coastal line due to the humidity coming from the ocean in the northernmost regions and the fact that castor bean has been recorded in this area [3]. On the other hand, castor bean has also been registered in the Arica y Parinacota region and between the Atacama and O'Higgins regions in the inland area, but it is always associated with a water body or in farmland borders [3,39] ( Figure 4).

Conclusions
The results obtained by the presence-species methodology corresponded well with the actual castor bean distribution in Chile, especially along the coastline between the Arica y Parinacota and Coquimbo regions. While both simplistic (with TNA = 8.0 °C) and presence-species methodologies showed similar behavior in the inland area, the main barrier for castor bean adaptation in that zone is the water regimen; however, castor bean could grow with irrigation.
Finally, both methodologies concluded that the most suitable land is between the Maule and Araucanía regions, although castor bean has not been registered in those regions. The simplistic methodology determined that 164,052 ha of suitable land with humid regime aptitude was registered between the Maule and Araucanía regions. With the presence-species methodology, 11,851 ha with mild thermic restriction and without hydric restriction was registered between the Biobío and Araucanía regions.
We conclude that both methodologies could be applied for castor bean determination, although the simplistic method, employed with TNA = 8.0 °C, showed high correspondence with its actual distribution in Chile.

Conclusions
The results obtained by the presence-species methodology corresponded well with the actual castor bean distribution in Chile, especially along the coastline between the Arica y Parinacota and Coquimbo regions. While both simplistic (with TNA = 8.0 • C) and presence-species methodologies showed similar behavior in the inland area, the main barrier for castor bean adaptation in that zone is the water regimen; however, castor bean could grow with irrigation.
Finally, both methodologies concluded that the most suitable land is between the Maule and Araucanía regions, although castor bean has not been registered in those regions. The simplistic methodology determined that 164,052 ha of suitable land with humid regime aptitude was registered between the Maule and Araucanía regions. With the presence-species methodology, 11,851 ha with mild thermic restriction and without hydric restriction was registered between the Biobío and Araucanía regions.
We conclude that both methodologies could be applied for castor bean determination, although the simplistic method, employed with TNA = 8.0 • C, showed high correspondence with its actual distribution in Chile.

Conflicts of Interest:
The authors declare no conflict of interest.