Introduction
The importance of fast-growing (i.e., production cycles of fewer than 10 yr), intensively managed, planted forests has increased globally during recent years due to the growing need for timber and derived goods (Fernández-Moya et al. Reference Fernández-Moya, Alvarado, Fallas, Miguel-Ayans and Marchamalo-Sacristán2017). Forest plantations play a crucial role in reforesting degraded tropical ecosystems, providing a reliable source of high-quality wood and multiple ecosystem services. Tropical timber plantations continue to be dominated by a few tree species grown in monocultures, including species of eucalyptus, acacia, pine, melina, and teak (Günter et al. Reference Günter, Stimm, Weber and Mosandl2011; Pancel and Kohl Reference Pancel and Kohl2016).
Costa Rica has a long-standing tradition of cultivating teak and melina, which represent the largest areas of planted forest in the country, covering approximately 49,000 ha and 14,000 ha, respectively (INEC 2022). Both species are widely used in commercial reforestation projects (Boley et al. 2009), with end uses ranging from pallet production to construction timber for domestic use and export, primarily to markets in Asia (Moya et al. Reference Moya, Bond and Quesada2014). Costa Rica ranks among the top 10 global exporters of teak roundwood and holds the third largest area of planted teak forests in the region, following Ecuador and Panama (Kollert et al. Reference Kollert, Sandeep and Sreelakshmy2024).
Weed management during the establishment and early stages of growth of a plantation is crucially necessary to improve final productivity. In the same way as in agricultural crops, weeds are among the major impediments to the successful establishment, survival, growth, and productivity of forest plantations (Adams et al. Reference Adams, Beadle, Mendham and Smethurst2003; Garau et al. Reference Garau, Ghersa, Lemcoff and Barañao2009; Ladrach Reference Ladrach2010). The composition of weed communities determines the availability of resources such as water and soil nutrients that are essential to tree growth. This composition can be broad, depending on factors such as plantation age, topography, previous land use, initial land preparation, and silvicultural management practices (Zimdahl Reference Zimdahl2018). Generally, in the early stages of crop development, grasses and other herbaceous plants are the main competitors, whereas woody vegetation becomes the primary problem once the plantation canopy closes (Meza et al. Reference Meza, Alfaro, Bedoya, Romero and Valerio2019; Vargas et al. Reference Vargas, Rubilar, Gonzalez-Benecke, Sanchez-Olate and Aracena2018).
In Costa Rica, research has been conducted on the composition of weed communities in crops such as coffee (Mora and Acosta 2001; Ricci et al. Reference Ricci, Virgínio Filho and Costa2008; Rojas et al. Reference Rojas, Rodríguez, Villalobos, Arias and Méndez2003), banana (Agüero-Alvarado et al. Reference Agüero-Alvarado, Rodríguez-Ruiz, González-Lutz, Portuguez-García and Brenes-Prendas2018; Rodríguez and Agüero Reference Rodríguez and Agüero2000), sugarcane (Leon et al. Reference Leon, Agüero and Calderón2017), and pineapple (Brenes-Prendas and Agüero-Alvarado 2007). However, to our knowledge, the composition of weed communities in teak and melina plantations has not been studied, nor has anyone collected data on the weed management practices used to control the most problematic weeds that affect these plantations. To design successful weed management practices, it is first necessary to identify the species associated with plantation species, the impact of the species on production, any beneficial attributes of the species, and effective management of the species. Weed control in plantation forests in Central America has been carried out in several ways, including by use of a machete or shovel weeding and, more recently, herbicides. The main reasons for herbicide use are their rapid implementation, high efficacy, and prolonged effect (Guevara Reference Guevara2011).
Surveys are useful ways for acquiring information about weed management practices and problematic weeds, and have been used to assess farmers’ needs. Although surveys of farmers do not provide a direct quantification of weed abundance and distribution, they are useful in understanding the perceived importance of weeds in each production system and the decisions farmers make regarding their preferred weed management tactics (Sarangi and Jhala Reference Sarangi and Jhala2018). Therefore, the main goal of this study was to characterize the current perceptions of teak and melina farmers on weed problems and herbicide use. The survey hypothesis was that the problematic weed species of forest plantations would be predominantly grass species, they would vary across Costa Rican provinces, and would require region-specific management strategies to effectively mitigate the weed interference challenges. The specific objectives of this survey were to determine 1) the most problematic weed species, 2) current weed management practices, and 3) the herbicides used in teak and melina plantations in Costa Rica.
Materials and Methods
Between January and May 2021, farmers enrolled in the Payment for Environmental Services (PSA) program, reforestation modality (PSA-reforestation) were asked to participate in an email or telephone survey. The PSA-reforestation program is administered by the National Forestry Finance Fund (FONAFIFO by its acronym in Spanish). FONAFIFO’s Environmental Services Department constructed a database of 525 active teak and melina farmers with plantations ranging from 1 to 10 yr old. Using that information, a random sample of 180 farmers was selected, stratified by province. Farmers with current contracts were present in only six of Costa Rica’s seven provinces when the survey was conducted.
The survey consisted of 12 questions divided into three themes or sections (Table 1): 1) general information and plantation establishment information; 2) weed management practices; and 3) problematic weeds. Section 1 consisted of questions about farm location (province and county), total planted area, and plantation establishment techniques. Section 2 asked farmers about their current weed management practices. Finally, in Section 3, respondents were asked to identify (by local common name) and rank the five most problematic weeds based on their own experience and the reason(s) for their choice. The questionnaire was pretested on six people, including farmers, weed scientists, foresters, and field technicians, to assess its readability and its value for meeting the stated objectives. Their comments were reviewed and incorporated into the final version. The final questionnaire was sent by email or completed via phone call.
Table 1. Questionnaire for teak and melina farmers regarding weed management practices and problematic weeds in Costa Rica.a
a Abbreviation: OA, open answer (respondents were asked to list the county in their province, the herbicides they used, and the weeds they considered to be problematic.
b County data were not included in the analysis because sample sizes were too small to allow meaningful comparisons.
c If a farmer had fields with both species they were asked to complete a survey for each species.
Data Analysis
To rank the most problematic weeds, we used the equation proposed by Sarangi and Jhala (Reference Sarangi and Jhala2018; Equation 1), where F is the number of respondents choosing a particular rank (r) for a certain species, X is the problematic points associated with that rank, and n is the total number of responses for that rank including all weed species:
$$RP = \sum\limits_{r = 1}^5 {} {{FX} \over n}$$
Results were tabulated, and a database was created. For each weed documented as being problematic, its family, genus, and species were recorded. Additionally, each weed was grouped into one of the following functional groups: 1) grasses, 2) vines, 3) broadleaf plants, and 4) sedges. Scientific names were assigned based on the common names provided by respondents. When identification was uncertain, photographs were requested, or a field sample was obtained for identification. Common names are reported based on the U.S. Department of Agriculture’s Plants database (USDA-NRCS 2025).
To analyze response patterns for categorical questions, response frequency was estimated as the number of times a category was mentioned divided by the total number of responses for a given question expressed as a percentage. This was calculated at the province level. Data on weed species were then compiled at the national level.
Results and Discussion
A total of 180 surveys were sent to growers in six of the seven provinces of Costa Rica, and 87% were completed and returned for analysis. The largest number of surveys were completed by respondents in Guanacaste and Alajuela provinces, representing 50% and 29%, respectively, of the completed surveys (Figure 1). No teak or melina farmers in Cartago Province were registered with the PSA-reforestation program, and thus no surveys were collected from farmers in that province.
Figure 1. Distribution of respondents in each province of Costa Rica who participated in the survey on weed management in teak and melina plantations.
General Information
The respondents were mainly plantation owners and, to a lesser extent, forestry engineers in charge of forest plantations with reforestation companies. Both owners and professionals oversaw removing weeds from the plantation. Seventy-five percent of survey participants identified as teak farmers, whereas the remaining 25% grew melina. Among the respondents, 43% owned plantations larger than 50 ha, while 35% had plantations ranging from 1 to 20 ha. Furthermore, nearly 50% of respondents indicated that prior to establishing their current plantations, the land was used as forest plantation, while 43% noted the land had been used for pasture and grassland. The remaining 9% of previous land uses included fallow and for growing agricultural crops such as cassava and corn (Table 2). Even though 90% of the respondents had established plantations on pastures or immediately harvested a tree plantation before their current product, only 11% employed soil mechanization techniques such as push/pile debris, subsoiling, and harrowing. Thus, the majority of the plantations were established without using tillage (i.e., no-till), which likely favors perennial species (Scursoni et al. Reference Scursoni, Vera, Oreja, Kruk and De La Fuente2019).
Table 2. Characteristics of plantation ownership, previous land use, and site preparation methods.
Weed Management Practices
The diversity and availability of means of weed control and production practices can greatly influence the efficacy of weed management plans (Sarangi and Jhala Reference Sarangi and Jhala2018). More than 85% of respondents reported employing a combination of manual, mechanical (e.g., weed wacker and/or tractor-mounted mower), and herbicides to control weeds in their plantations. Only 2% mentioned controlling weeds exclusively with herbicides (Figure 2).
Figure 2. Response frequency for weed control tools used in forest plantations (A) and area targeted for control (B) according to perceptions of teak and melina farmers.
Another important parameter for weed management is to properly identify and characterize the target area where weeds are more problematic (Garau et al. Reference Garau, Ghersa, Lemcoff and Barañao2009). Plantation age is a determining factor in selecting the area to be controlled. During plantation establishment, weed control at the base of the tree (≤50 cm radius) is the predominant method of vegetation management; however, as the plantation matures, the approach shifts toward strip or total vegetation control, with a corresponding increase in the reliance on herbicide applications (Kogan and Alister Reference Kogan and Alister2010; Murillo et al. 2024). The current survey confirms this observation because more than 50% of the respondents indicated that they control weeds in the entire plantation, including both tree lines and interrow areas.
Herbicides are commonly applied for weed control in intensively managed forest plantations due to their efficiency, cost-effectiveness, labor-saving benefits, and prolonged effect compared to other methods (Miller and Wigley Reference Miller and Wigley2004; Wang et al. Reference Wang, Gao and Bai2001). Among the respondents, 90% reported using a total of nine herbicides with five different mechanisms of action. Among the herbicides mentioned, two specifically control grasses, and no preemergence herbicides were reported. The most frequently mentioned products were glyphosate, metsulfuron methyl, and paraquat (Figure 3; Table 3). This trend was consistent across all six provinces.
Figure 3. Survey responses indicating herbicide use in teak and melina plantations in Costa Rica.
Table 3. Herbicides used in teak and melina plantations in Costa Rica.a
a Abbreviations: HRAC, Herbicide Resistance Action Committee; WSSA, Weed Science Society of America.
b Herbicides are categorized into groups depending on the herbicide’s site of action. Group information is available at https://wssa.net/wp-content/uploads/WSSA-Herbicide-SO_WSSA_20210505.xlsx.
Some farmers reported using a single herbicide, and others said they used two or three herbicides in rotation (Figure 4). Herbicide use varied among respondents, with most reporting using two herbicides in their management plans (45%). Specifically, the most common pair was glyphosate and metsulfuron methyl. Among respondents who reported using a single herbicide, glyphosate was the dominant, far surpassing metsulfuron methyl and paraquat.
Figure 4. Number of farmers reporting using one, two, or three herbicides for weed management in teak and melina plantations in Costa Rica.
Glyphosate is the most widely used herbicide for vegetation management in forest plantations in Costa Rica, primarily due to its broad-spectrum weed control, flexibility in application timing (e.g., plants of different sizes), and low cost. Its widespread use also reflects its effectiveness in perennial crop systems across the country (Kogan and Allister 2010; Ramírez-Muñoz et al. Reference Ramírez-Muñoz, Bravo-Durán and Herrera-Ledezma2017). Glyphosate is sprayed several times during the first 2 to 3 yr after plantation establishment to control weeds that grow understory, both in the alleys and at the base of the tree. In Costa Rica, glyphosate resistance has been documented in two grass species, goosegrass (Heap Reference Heap2025) and arrocillo (Ramírez-Muñoz 2016), both of which are commonly found in forest plantations, raising concerns about the long-term effectiveness of glyphosate-based weed management strategies.
Few respondents reported using five herbicides in their weed management practices. These included glyphosate, paraquat, metsulfuron methyl, and the auxin herbicides (Group 4 herbicides as categorized by the Weed Science Society of America), 2, 4-D, and triclopyr. These herbicides were valued for their efficacy against broadleaf species and the relatively low occurrence of resistance cases (Heap Reference Heap2025). Specifically, triclopyr is often employed for stump removal in harvested plantations where site preparation for a new production cycle is underway.
Problematic Weeds
In this survey, farmers identified a total of 40 weed species distributed across 21 plant families (Table 4). The frequency of these families varied among provinces according to farmer perceptions of their importance. Poaceae was consistently reported to occur in all provinces, ranking first in every region with frequencies ranging from 16% in Limón to 32% in Puntarenas. Other frequently mentioned plant families included Fabaceae, Cyperaceae, and Malvaceae, although their importance varied by province (Table 5). For instance, Fabaceae was frequently reported by respondents in Guanacaste (13%) and Puntarenas (8%), whereas Cyperaceae plants were dominant in Heredia (12%) and Limón (8%). Furthermore, weeds in the Melastomataceae and Dillenaceae families were reported to be significant in Alajuela, with frequencies of 14% and 11%, respectively. This regional vegetation variation is strongly influenced by Costa Rica’s physiography and climatic gradients. The country’s two main mountain ranges run longitudinally through the central region, interacting with prevailing northeastern winds to create diverse ecological zones. As a result, provinces on the Atlantic side of the country are characterized by more humid conditions. For example, annual rainfall in Limón ranges from 2,500 to more than 4,800 mm, and parts of Heredia and northern Alajuela can receive 2,700 to 3,800 mm of rain. Conversely, the Central Pacific (Puntarenas, San José) and northwestern (Guanacaste) regions are drier with distinct wet and dry seasons and annual rain amounts of 1,800 to 2,600 mm (Quintero and Villalobos Reference Quintero and Villalobos2001). During the dry season, from January to March, total rainfall can total less than 6 mm, which favors the presence of drought and fire-tolerant weeds.
Table 4. Common weed species reported in teak and melina plantations based on perceptions of survey respondents.
Table 5. Ranking and frequency of the most commonly reported plant families by province based on farmers’ perceptions of problematic weeds in teak and melina plantations.
Weed prevalence also reflected these climatic differences. In the humid provinces, vines such as raspa guacal and orchid vine were reported more frequently, while in the drier regions grasses such as jaraguagrass and arrocillo were reported to be more problematic. At the national level, based on farmers’ input, the most problematic weeds in order are 1) raspa guacal, 2) arrocillo, 3) jaraguagrass, 4) coyol palm, and 5) orchid vine (Table 6). No single species was consistently reported in all provinces, yet it is noteworthy that two species in the genus Paspalum were considered to be the most problematic in three provinces. Arrocillo has been documented as being resistant to glyphosate in areas near Costa Rica’s Caribbean coast (Ramirez-Muñoz Reference Ramirez-Muñoz2016). Given that respondents reported glyphosate as the most widely used herbicide and that arrocillo is one of the most problematic weeds, these findings emphasize the urgent need for ongoing monitoring of this species in forest plantations so as to detect and manage emerging cases of resistance.
Table 6. Farmers’ perceptions of the most problematic weeds in teak and melina plantations by province and nationally, ranked 1 to 5, based on farmers’ perceptions.a
a Respondents in San José province reported only weed type and did not identify specific weed species.
b Respondents in Heredia reported only two species.
In Alajuela province, vines were reported to be the most problematic weeds. Volunteer teak and melina trees were also mentioned as being problematic because of their strong regrowth and persistence capability. In established plantations where these species were previously grown, their rapid regrowth can compete with newly planted seedlings. If this is not managed promptly, the competition among planted and regrowing trees may hinder the preferred crop’s growth and, in some cases, lead to the death of newly planted individual trees.
The considerable weed biodiversity present in teak and melina plantations in Costa Rica aligns with reports from previous studies carried out by Agüero-Alvarado et al. (Reference Agüero-Alvarado, Rodríguez-Ruiz, González-Lutz, Portuguez-García and Brenes-Prendas2018) in coffee and banana plantations and by Brenes-Prendas and Agüero-Alvarado (2007) in pineapple plantations. In particular, the predominance of grasses has been reported in studies by Takin et al. (Reference Takin, Affinnih, Falola-Olasunkanmi, Garuba and Osatuyi2019) and Tuffi Santos et al. (Reference Tuffi Santos, Cardoso Filho, Santos Júnior, Sant’Anna-Santos, Felix and Leite2013) that plants in the Poaceae family are the most representative weeds in teak plantations in Nigeria and eucalyptus plantations in Brazil.
This study identified the weed species that pose significant challenges to forestry plantation owners in Costa Rica, with variations observed among provinces. Specific species of grasses and vines present major challenges in forest plantations. Their effect varies across regions, but they are especially aggressive during the early stages of tree establishment, when competition for light, water, and nutrients can substantially inhibit tree growth; specifically by reducing height increment and stem diameter development. When grasses are present, combinations of herbicides and their application in a series can be more effective than mechanical methods alone, leading to better vegetative growth (Domingos and Coelho Reference Domingos and Coelho2018; Silva et al. Reference Silva, Alves and Toledo2012). Conversely, vines are frequently removed mechanically and treated with selective herbicides to effectively suppress their regrowth and to allow tree canopy closure (Paul and Yavitt Reference Paul and Yavitt2011).
The results of the survey underscore the importance of tailoring weed management strategies to address regional differences in weed composition and abundance, thereby enhancing the efficacy of control efforts. Given the limited information on management of tropical weed species and silvicultural practices, more information is needed about the ecology of those weeds and cost-effective control tools need to be developed.
Practical Implications
The findings of this survey provide a valuable characterization of the weed challenges faced by teak and melina plantation owners in Costa Rica, and the need for understanding the composition of the local weed community and to develop specific weed management strategies. Having identified Poaceae species and climbing vines as the most problematic weeds highlights the necessity for targeted control measures, particularly during the early establishment phase of teak and melina trees. Given the high reliance on glyphosate, metsulfuron methyl, and paraquat, there is a pressing need to diversify herbicide options to prevent resistance development and ensure their long-term effectiveness.
Acknowledgments
We thank FONAFIFO and all those who kindly responded to the survey.
Funding
This study received financial support from the Vicerrectoría de Investigación y Extensión del Instituto Tecnológico de Costa Rica, and by Vicerrectoría de Investigación de la Universidad de Costa Rica.
Competing interests
The authors declare they have no competing interests.
Open access
