Diversity and Allelopathic Potential of Weeds among Panamanian Coffee Crops

1Departamento de Química Orgánica, Facultad de Ciencias Exactas y Tecnología, Universidad de Panamá, Panamá 2Secretaria de Competitividad y Logística, Ministerio de la Presidencia, Palacio de Las Garzas, Ciudad de Panamá, República de Panamá 3Centro de Biodiversidad y Descubrimiento de Drogas, Instituto de Investigación Científica y Servicios de Alta Tecnología, Clayton, Ciudad del Saber, Republica de Panamá


Introduction
The first coffee plantations in Panama sprung up in the early nineteenth century. At present, Panama is a discreet producer of coffee globally due to it is a very small country. However, a mix of favorable environmental conditions (altitude, precipitation, temperature, relative humidity, inter alia) makes the country ideal for the production of high-quality coffee (Cherigo et al., 2012;. The main coffee production area is located in the province of Chiriqui. In this region coffee production is an activity of high economic value. Therefore, the harvest efficiency of this crop is crucial for farmers. Unfortunately, every year coffee production is widely affected by several factors including pests, weeds and climate change, among others. Weeds provoke considerable economic losses for coffee producers by interfering with the growth and development of the coffee plants in the fields. Weeds directly reduce coffee crop yields by competing for nutrients, sunlight, water and negative allelopathic effects. Also, they can contain pests and pathogens which can also decrease the yield and quality of coffee (De Graaff, 1986;Radosevich et al., 1997;Cherigo et al., 2012).
Weed competition against coffee plants could be minimized by reducing or eliminating weeds in the cultivated fields. Most traditional farmers in the Chiriqui Province still work under the paradigm that weeds are plants that only cause damage to crops, so they often eliminate them completely. However, recent studies have shown that different weed species have potentially useful applications for humans. For example, some weeds can protect against other weeds when they are used as cover crops (Altieri, 1995). In addition, there are other weeds with ethno medical uses (Pani et al., 2002;Panda et al., 2014). Therefore, it is important to generate information that would allow farmers to distinguish between potentially useful weeds from those having only negative effects for their rational or systematic elimination (Altieri, 1995;Cherigo et al., 2012). The identification of the allelopathic properties of weeds, especially those related to the inhibition of plant germination and growth (phytotoxins), is of significant relevance because some phytotoxins could be efficient substitutes for highly toxic synthetic herbicides as those currently used in developing countries (Bhadoria, 2011).
We suggest that one of the first steps towards more sustainable crop production is to gain complete knowledge of the plant species diversity within the defined growing area. Therefore, studies to generate information about biological diversity and dynamics of weeds within crop fields are necessary for their rational and systematic control. Examples of this can be seen in studies performed in countries like Brazil (Dedeca, 1959;Ronchi and Silva, 2006), Cuba (Caro et al., 1990), Honduras (Ordoñez et al., 2000), Kenya (Anonymous, 1992), Mexico (Braver, 1957), Nicaragua (Aguilar et al., 2003) and Papua Nueva Guinea (Byrne, 1984), where farmers have developed effective weed control programs from the identification of the different species found in their coffee plantations. These studies highlight that appropriate methods for handling coffee weeds should be based on a better understanding of the biology and population dynamics of different species.
Our research group proceeded to record the different types of weeds that grow in the coffee plantations in Santa Clara (Chiriquí). We also evaluated the phytotoxic potential of the weeds using a standardized method for determining if the identified weeds provoke adverse effects by chemical competition (production of phytotoxic compounds) or/and by competing for nutrients, light and available water in soil (Cherigo et al., 2012;. Finally, we proceeded to evaluate the organic extracts from Blechum pyramidatum and Borreria verticillata in a panel of bioassays to demonstrate to farmers that weeds may also have useful applications to human health. All the information generated will serve to propose to local farmers changing their conventional strategies of elimination of weeds by others approaches systematic and rationally designed.

Study area and Sampling
An important part of our study was to select a representative area that possesses the prevailing growing conditions (altitude, soil, temperature, among others) of the Panamanian coffee fields. We located an appropriate region for this study in the town of Santa Clara. Once we got the required permits (from the owner and the Panamanian government), we proceeded with our study.
The study field was divided into 12 altitudinal floors; each altitudinal section was established by an increase of 50 meters' elevation above sea level (m.a.s.l.). The first altitudinal floor was located at 950 meters and the last at 1500 meters. To facilitate the recording of weeds, we subdivided each altitudinal floor in 3 transects of equal length and every two meters all existing varieties of weeds were collected and recorded. Also, the geographic location information of each collected weed was recorded using Global Positioning System (GPS) coordinates (Table 1).
According to standard protocols, we considered ˝abundant˝ all weeds that were found at least twice in two of three installed transects per floor. All obtained data is in Table 2, abundant weeds were marked with "X," and weeds found only once or absent were marked with "0". The number of altitudinal floors marked with "X" was divided by the total number of transects (12) to generate the weed Abundance Index (AI), which reflects the percentage of weed appearance. To illustrate the diversity of weeds that occurred for each altitudinal floor, we proceeded to generate the Altitudinal Floor Index (AFI), which was calculated by taking the number of species found at each altitudinal floor and dividing it by the total number of species found in the 12 floors (Table 2).

Taxonomic Identification
The identification of weeds was pursued using taxonomic keys from the Flora of Panama and the Mesoamerican Flora. Weed identification was also confirmed employing herbarium specimens from the University of Panama. Finally, nomenclature of all collected species was verified using the TROPICOS database from Missouri Botanical Garden and Vascular Plants of Panama catalog.

Plant Material for Phytotoxic Assays
Approximately 100 grams of the fresh aerial parts of each of the different coffee weeds species were randomly collected in the sampling area. Samples were stored in black bags and immediately transported to the laboratory. The samples were rinsed with water and placed to dry at room temperature. Once dried, the samples were pulverized and stored for further processing.

Organic Extracts Preparation
To obtain a maximum amount of organic constituents the extract from each plant was prepared five times by a maceration process using a mixture of ethyl acetatemethanol (1:1). The mixture of solvents was concentrated using a rotary evaporator (Laborota 4010, Germany) until a semisolid paste of crude extract was obtained. This extract was stored at -4°C until further use (Cherigo et al., 2012;.

Inhibition of Radical Elongation of Amaranthus Hypochondriacus
The growth inhibitory activity of the extracts on seedlings of A. hypochondriacus was evaluated using the Petri dish radicle elongation and germination bioassay (Mata et al., 1998;Cherigo et al., 2012).
Briefly, organic extracts were dissolved in ethyl acetate and we prepared dilutions with final concentrations of 10, 100 and 1000 µg mL −1 . One mL of the test solution was added to a Petri dish containing a filter paper disc; then the solvent was evaporated in an extraction hood. After complete evaporation of the solvent, 3 mL −1 of distilled water and ten Amaranthus hypochondriacus seeds were added to the filter paper. This procedure was performed in triplicate (30 seeds for each evaluated extract). Petri dishes were incubated in the dark at 28°C for 24 h. After this time, the radicular growth of each seed was measured. Petri Dishes with 2,4-D were used as a positive control. As a negative control, we used Petri dishes with treatment but without extract and also Petri dishes with seeds without treatment.
The results were analyzed by ANOVA (p<0.05) and IC 50 values were calculated by Probit analysis based on the percent of radicular growth or germination inhibition.

Malaria and α-Glucosidase Bioassays
Malaria bioassays were performed as previously reported by us using chloroquine as a positive control (Cherigo et al. 2012). α-glucosidase enzyme bioassays were conducted following the protocol of Lopez et al. (2015) using acarbose as a positive control.

Sampling and Identification of Weeds
A total of 42 different species of weeds were collected and identified among the evaluated crops. Table 1 lists all species together with their location or GPS coordinates.
The abundance of identified species along marked altitudinal levels is detailed in Table 2. The most abundant species -Emilia sonchifolia, Impatiens walleriana, Elephantopus mollis, Hyptis capitata, Lactuca graminifolia, Rumex crispus and Spermacocce sp.-were found in more than 7 altitudinal floors (≥ 66% presence). Among these, the first two species are the most abundant weeds with an abundance index of 0.83. Four species -Borreria verticillata, Polygala longicaulis, Pseudelephantopus spicatus and Sonchus asper-were present in 58% of the sampled floors.

Allelopathic Potential
IC 50 values for the 42 evaluated extracts are shown in Table 3. The degree of extract phytotoxicity was determined based on the range in which the IC 50 values were encountered using the following categories: very phytotoxic (IC 50 <100 µg mL −1 ), phytotoxic (100< IC 50 <500 µg mL −1 ), moderately phytotoxic (500< IC 50 <1000) and non-phytotoxic (IC 50 >1000 µg mL −1 ). In practical terms, IC 50 values within the very phytotoxic category or very close to this range are those that could cause a markedly adverse effect on coffee plants.

Identification of Alternative Biological Properties in Coffee Weeds
To achieve this objective, we selected the two weeds which had generated the largest quantities of organic extracts, B. verticillata and B. pyramidatum. In order to get evidence of useful applications of both weeds, we evaluated their extracts utilizing a panel of bioassays that included antiparasitic (Leishmania donovani, Plasmodium falciparum and Tripanosoma cruzi), anticancer (MCF-7 cell) and hypoglycemic (αglucosidase inhibition) activities. As a result, B. verticillata sample showed antiparasitic activity (65% growth inhibition at 10 µg mL −1 ) against Plasmodium falciparum. On the other hand, the organic extract from B. pyramidatum showed moderate inhibition (63% of inhibition) of the α-glucosidase enzyme. Both samples were totally inactive in all the other bioassays.

Discussion
This study is the first to identify weeds that grow in some of Panama's fields of coffee crop production. It is important to point out that the total number of different species detected was lower than that reported for other countries, although the type of weeds is similar. The number of weeds per altitudinal floor declined strongly with increasing altitude. At floors ≤ 1200 m.a.s.l. (six lower floors) there were 147 records from weeds marked as abundant in this study while there were only 79 records at sites <1200 m.a.s.l. (six upper floors). These findings were corroborated by the analysis of AFI index, where it was observed that the greatest AFIs were produced in the altitudinal floors between 1050 and 1100 meters' elevation (AFI values of 0.57 and 0.78, respectively). At the highest floors, the AFI index value tended to decrease finishing with a value of 0.14 in the last altitudinal floor. Among the less abundant species, Browallia americana, Centradenia inaequilateralis, Cuphea sp. and Sida acuta were recorded only in the six upper floors while Solanum nigrescens, Tradescantia commelinoides, Scrophullariaceae sp., Oxalis sp., Mimosa pudica, Cuphea calophylla, Commelina diffusa, Blechum pyramidatum, Bidens pilosa and Achyranthes aspera were recorded only in the six lower floors.
The phytotoxic activity of organic extracts from weeds was evaluated against the seed germination and initial radical elongation of dicotyledonous species Amaranthus hypochondriacus (Amaranthaceae). Several species of the genus Amaranthus are broadleaved weeds that are commonly associated with coffee plantations (Njoroge, 1994). On the other hand, Amaranthus is one of the model genus widely standardized for phytotoxicity studies of plant extracts (Anaya et al., 1990;Mata et al., 1998;Cherigo et al., 2012). Therefore, this species was considered adequate as target weed for our allelopathic evaluation. In addition, the seeds of this weed present a fast, uniform and consistent germination, which facilitates the implementation of this bioassay. The crude ethyl acetate-MeOH (1:1) extracts from all coffee weeds were prepared. This solvent mixture was used to obtain the greater amount of both hydrophilic and lipophilic plant components. The phytotoxic activity identification of weeds extracts could be beneficial in two different ways: (1) in the identification of the weeds that can cause more chemical damage to commercial plants and (2) in the determination for phytotoxins with potential use as selective bioherbicide mainly against certain specific types of weeds. According to IC 50 values, Emilia sonchifolia and Hyptis capitata are the weeds that could cause damage to coffee plants in the plantations due to their allelopathic potential. It is highly likely that the other 40 species only can interfere with coffee trees by competition for nutrients and water present in the soil.
In traditional agricultural practices, weeds are seen as entirely undesirable agents and farmers want to eliminate them at any cost, but farmers do not know that even though weeds can cause significant economic losses in their crops, they can also have beneficial applications for humans. We just have to remember that plants have also been prolific producers of pharmacological and agrochemical important metabolites (Khalid et al., 2002;Cragg and Newman, 2013;Pino et al., 2013;Atanasov et al., 2015). For this reason, searching for biological properties in weeds can also allow us to detect potential beneficial applications for humans. So, we proceeded to evaluate the organic extracts from B. verticillata and B. pyramidatum in a panel of bioassays that included antiparasitic (Leishmania donovani, Plasmodium falciparum and Tripanosoma cruzi), anticancer (MCF-7 cell) and hypoglycemic (α-glucosidase inhibition) activities, to get evidence of useful applications of this weed. B. verticillata sample showed selective activity (65% growth inhibition at 10 µg mL −1 ) against Plasmodium falciparum. Before this study, the antimalarial activity of this plant has never been reported and no antiprotozoal metabolites from B. verticillata have been published. On the other hand, the organic extract from B. pyramidatum showed 63% of inhibition of the α-glucosidase enzyme. αglucosidase hydrolyzes starch and disaccharides to release glucose and high levels of sugar are related to diabetes mellitus. Inhibition of this enzyme is one of the key mechanisms to regulate blood sugar levels, so this plant could be an attractive hypoglycemic agent.

Conclusion
Weed diversity found in Panamanian coffee crops is less than that reported in coffee production in other countries. Forty-two different species were found, most of which have been previously reported as common weeds in coffee plantations. Most collected weeds had a relatively high abundance and their IA was above 0.33, although there were also a few weed species that had an IA equal to or below 0.33 (with very low abundance). The most abundant weeds belong to Asteraceae family, particularly Emilia sonchifolia and Impatiens walleriana. In the allelopathic bioassays, only eight of the forty-two species showed marked phytotoxicity. In fact, only Emilia sonchifolia and Hyptis capitata showed values of IC 50 s that could cause a marked in vivo negative allelopathic effect against coffee plants. This suggests that the weeds identified in this study are more likely to cause more damage by their physical competition than by their negative allelopathic effects. Biological evaluations of B. pyramidatum and B. verticillata extracts showed that these weeds possess hypoglycemic and antimalaric activities. The information generated in this study offers valuable tools that will help us to propose to the farmers a rational, systematic and selective elimination of weeds because, as we have shown, weeds are not completely harmful to coffee plants and they can have useful applications for the human health.