The research was carried out in the “Leopoldo Arana Cabrera” Agricultural Unit, in the municipality of Muna, Yucatan, Mexico, located at coordinates 20˚24'52'' north latitude and 89˚44'31'' west longitude in a soil classified as K’ankab lu’umfor the Mayan classification and Luvisol for the World Reference Base of soils (WRB) [9]. Peanuts (Arachis hypogaea) were the previous crop before planting the habanero pepper.

The species, in the field, were identified one week before the establishment of the treatments; and by using three unit squared areas of 50 × 50 cm (0.25 m²), per each replication, the next variables were recorded: coverage, frequency of appearance, abundance and dominance of each species. The Value of Relative Importance (VRI) of the weeds were calculated according to the methodology by Gámez López et al. (2011) [10] for weeds characterization.

The study was carried out from September to October 2021 (Autumn-Winter cycle). The land was prepared by a combination of manual clearing from the third week of September until October 2. A pre-sowing herbicide Paraquat (200 g a∙i∙L⁻¹) of contact action was applied (10 mL of commercial material per liter of water to eliminate the first vegetation. Once the weeds started recovering, the treatments were applied in October 9 in pre-transplantation using manual backpack sprinklers.

The transplant was carried out on October 11, two days after the first herbicide application (DAA), using 12 cm high seedlings of the “Jaguar” variety, 15 treatments with different herbicides were established and a 16^th, treatment as the control with no herbicide application was considered.

The doses used (Table 1) were determined using the sheets of each herbicides published by the manufacturers and those suggested by INIFAP [11]. Legal authorization for Mexico and USA, for peppers cultivation (Capsicum spp.), was also considered. Crop management was carried out according to recommendations of Avilés et al. (2010) [11] for habanero pepper under outdoor conditions of Yucatan, Mexico.

Sixteen treatments with four replications were established, under drip irrigation conditions, and analyzed in a randomized complete block design in experimental unitsof 14.25 m² (1.5 × 9.5 m). The seedlings were transplanted 40 cm apart with a projected population density of 16,750 plants ha⁻¹. Data were subjected to an Analysis of Variance (ANOVA), Mean Comparison Test by Tukey’s method (p ≤ 0.05) using the Statgraphics Centurión program, version 16.1.2.0.

The percentage of coverage was measured visually, adapting the methodology described by Rodríguez et al. (2008) [12] and Gámez López et al. (2011) [10] for

*Weight percentage of main components; yHRAC = Herbicides Resistance Action Committee 2020; zTC = Toxicological Category.

weed populations. In this case, sixteen unit-squared areas of 50 × 50 cm (0.25 m²) were used per treatment (four squares per repetition) at 7 and 14 days after herbicides application (DAA). The data were then, transformed to the Arcsine Root of x for statistical Analysis of Variance (ANOVA) [13].

For phytotoxicity evaluation of herbicides on habanero pepper all plants in the experimental units were taken into account where mortality and the symptomatology of the damage were evaluated using the method proposed by the European Weed Research Society (EWRS) cited by Pérez et al. (2014) [14] (Table 2).

In order to detect any effect on crop growth, the height, from base of the stem to the last apex, of sixteen randomly selected plants (four per replication) was taken twice, in October 11 and October 23 (14 DAT).

Source: Urzúa (2001), cited by Pérez et al. (2014).

A preliminary analysis of profitability per treatment was carried out considering the costs of the products and the application days per hectare as compared to the estimated cost carried out by the producer (combination of manual and chemical control).The cost reduction ($) of each treatment was calculated when comparing the production cost of the producer, as 100%, against the production cost of each treatment.

The farmers usually consider the use of polypropylene tunnels to protect the crop against virus, transmitted by a whitefly, during the first 45 to 60 DAA.

Fourteen dominant weed species were detected in the original vegetation: from which 85.7% (12) are broad-leaved species such as: Xtees (Amaranthus dubius), k’iix tees (Amaranthus spinosus), k’an tumbuub (Bidens pilosa), pants’ iil (Boerthavia erecta), kidney beans (Crotalaria incana), xanamucuy (Euphorbia hyssopifolia), xuul (Lonchocarpus rugosus), fern grass (Parthenium hysterophorus), jabín (Piscidia piscipula), purslane (Portulaca oleracea), chi’chi’bej (Sida glabra) and sac xiw (Waltheria americana); while only 14.3% (2) are narrow-leaved ones, such as: nutsedge (Cyperus ligularis), and guinea grass (Megathyrsus maximus).The predominant species due to their higher Relative Importance Values (RIVs) were Euphorbia hyssopifolia (Euphorbiaceae), Megathyrsus maximus (Poaceae) and Parthenium hysterophorus (Asteraceae) with 61.8%, 52.4% and 33.7%, respectively (Figure 1).

According to the ANOVA, there was a highly significant difference between treatments at 7 and 14 DAT. At seven DAT, Tukey’s test (p ≤ 0.05) identified the Clomazone herbicide as the only one statistically similar to the Weedy-Control treatment with coverage of 2.78% and 3.75% respectively. On the other hand, the other treatments were statistically different with values ranging from 0.28 to 2.22%, highlighting Pendimethalin, Glyphosate, Natural Herbicide 1, Ammonium Glufosinate, Bensulide, Paraquat, Carfentrazone Ethyl, Oxadiazon and Metolachlor. Particularly Pendimethalin and Ammonium Glufosinate recorded the lowest coverage values with 0.28 and 0.31%, respectively (Table 3).

At 14 DAT, weed coverage in all herbicide-based treatments were statistically different as compared to the weedy control, with values ranging from 0.25 to 8.13%, highlighting Pendimethalin (0.25%), Glufosinate ammonium (1.13%), Trifluralin (1.5%) and Chlorthal dimethyl (1.88%) with the lowest values. With the exception of Pendimethalin, all those herbicides were recommended by Avilés et al. (2010) [10] for good weed control in habanero pepper in Yucatan as an alternative to reduce the weed control costs (Table 3).

Note: Different letters mean Statistical Significant Differences (p < 0.05, Tukey).

The only phytotoxic effects, in the habanero pepper, were detected when using the herbicide Clomazone at 14 DAT with 1.3% of affected plants. The percentage of the canopy affected per plant was 28.0%. The visible symptom was noted with a marked decolorating of young leaves close to the apical bud. However, no mortality was observed in any plant.

No statistical differences were found between treatments at any measured growing stage. Height was not affected, at all, during the first two weeks of plant development (Table 4). The highest height was observed in plots with Pendimethalin (14.9 cm), Fomesafen (14.6 cm), Glyphosate (14.5 cm), Clomazone (14.4 cm) and the Weedy Control (14.3 cm), while the plants with the lowest height were those treated with Carfentrazone ethyl (12.6 cm). Although Clamozone provoked symptoms of toxicity, the results presented here indicate that growth was not affected by this herbicide and by any other.

*No Significant Differences (NS).

Table 5 describes the unit costs of herbicides and treatments as of September 2021, according to the doses per hectare used; all compared to the estimated costs of the combined control used by the producer in the first 30 days after transplanting.

It is observed that the cheapest treatments with respect to the producer’s practice were the herbicides Paraquat ($1987.00) and Glyphosate ($2210.20), both showing the lowest unit costs with a cost reduction between 73.4% and 70.4% respectively. The crop system of the producer includes the use of polypropylene tunnels to protect the crop, for the first 45 to 60 DAA, against viral infections transmitted by the whitefly Bemisia tabaci Genn. In that way, it is not possible to apply these herbicides from outside the tunnels without damaging the crop.

On the other hand, except for the herbicide Bensulide, which was the treatment with the highest cost ($20321.61) and Chlorthal dimethyl, which only reduced the cost by 7.1%, all the other treatments were also highly profitable. The cost reduction ranged between 45.5% (Ammonium Glufosinate) and 68.7% (Ethalfluralin) (Table 5).

*Total cost includes herbicides and labors of application. Calculated in Mexican pesos; **Cost includes manual control of weeds, use of polypropylene tunnels and one application de Glyphosatein dose of 4 L∙ha⁻¹. Costs at September 2021.