Introduction:

Different phytopathogenic agents have been associated to the gummosis of citrus fruits. Species of the Phytophthora genus are the main ones that have been reported (^{Bright et al., 2004}; ^{Brentu and Vicent, 2015}; Graham and Feichtenberger, 2015). However, the members of the Botryosphaeriaceae family have displayed certain relevance due to their capacity to adapt in order to infect other crops near the original native host fields (Mondragon-Flores et al., 2021).

Gummosis is distributed in all the citrus-producing areas of the world. It has caused losses in Italy (^{Aloi et al., 2021}), Algeria (Linaldeddu et al., 2015; ^{Berraf-Tebbal et al., 2020}), Iran (^{Abdollahzadeh et al., 2010}) and the United States (^{Adesemoye et al., 2014}), and in Mexico, it is distributed in the areas that produce sweet citrus fruits and Mexican and Persian limes in the states of Colima (Rocha-Peña et al., 2003), Morelos (Valle-de la Paz et al., 2019a), Puebla, Veracruz (^{Bautista-Cruz et al., 2019}), Nuevo León and Tamaulipas (Polanco et al., 2019), with important data that range between an incidence of 2 and 14% (Medina-Urrutia et al., 2002).

Symptoms were observed in necrotic lesions in stems and branches, along with the appearance of a gummy exudate, followed by wilting, the yellowing of leaves, defoliation and finally, the partial or complete death of the tree (^{Bautista-Cruz et al., 2019}; ^{Berraf-Tebbal et al., 2020}; ^{Aloi et al., 2021}). The pathogen survives in the stubble of diseased plants, and is spread via trimming tools, but also via rainfall, irrigation, wind and insects (Moreira-Morillo et al., 2021). The disease prevails largely thanks to temperatures between 26 and 32 ^oC and a high relative humidity (80%) (Úrbez-Torres et al., 2010; Picos-Muñoz et al., 2015).

The control of gummosis has become more efficient with the use of preventive chemical fungicides such as Carbendazim (Da Silva Pereira et al., 2011; Valle-de la Paz et al., 2019b), Thiabendazole (Da Silva Pereira et al., 2011; ^{Camacho-Tapia et al., 2021}), Benomyl and copper-based compounds (^{Everett and Timudo-Torrevilla, 2007}; Sáenz et al., 2019; Valle-de la Paz et al., 2019). Antagonistic microorganisms have also been used, such as Trichoderma spp. and Bacillus subtilis (^{Bhuvaneswari and Rao, 2001}; Rusin et al., 2021). However, the use of chemical fungicides must be revised depending on the pathogen under study, as well as avoiding the generation of resistance. Obtaining biological alternatives for control is important for the conservation of biodiversity. This study has been developed with the goals of identifying the pathogens associated to gummosis in citrus fruit orchards in Plan Chac, Sacalum, Yucatan, as well as to evaluate the chemical and biological control alternatives that offer alternatives for the management of the disease.

Materials and Methods

Plant material and isolation of fungi. Plant tissue (bark) and soil samples were used, taken from a semi-commercial orchard with orange (Citrus sinensis), lemon (C. latifolia) and grapefruit trees (C. paradisi) with symptoms of exudate and necrosis. The orchard is located in the town of Plan Chac, Sacalum, Yucatan. The samples were extracted in July and October, 2023. For the isolations of the tissue, 5 mm² were cut, which were disinfected with 1 % sodium hypochlorite per minute and washed three times with sterile distilled water (^{Bautista-Cruz et al., 2019}). The tissues were left to dry in sterile paper towels and then planted in dishes with natural PDA media (200 g potato, 20 g agar-agar, 15 g dextrose in 1000 mL of water) and incubated at 24 °C for 24 h.

Soil samples were taken around the trunks of trees with symptoms and at a depth of 30 cm. For the isolation of the soil fungi, the technique consisted in serial dilutions. Thus, in the first tube, 1 g of soil in 9 mL of sterile distilled water was added; later, 1 mL of that solution (soil-water) was transferred to a series of tubes with the same characteristics (^{Aziz and Zainol, 2018}). Three dilutions were made of each sample (10^-1 to 10^-3). Out of the final solution, 20 µL were transferred into dishes with the natural PDA medium, which were incubated at 24 °C for 24 h and observed under the microscope to detect fungal growth.

Purifying and identifying isolates. The isolates were purified using hyphal tip, planted in a PDA medium (BD BIOXON^®, Cuautitlán Izcalli, Mexico) and kept for a 7 to 14-day day period at 24 °C in an incubator (BINDER, Model BD53- UL, Tuttlingen, Germany). The morphological characterization of the purified isolations was carried out with the following culture media: PDA with antibiotics (39 g, 200 mg of streptomycin and 1000 mL of water) and Clover Leaf Agar (CLA) (20 g agar, 1000 mL water, five pieces of carnation leaf per dish). In the PDA medium, the pigmentation and growth rates of the cultures were determined. To study the morphology of the Fusarium genus, the CLA medium and moist chambers (Petri dish, wet paper and pieces of aluminum) which were evaluated after five days for the measurement of phyalids. In addition, macro and microconidia and chlamydospores were measured. For Pestalotia, conidia were measured, and for Lasiodiplodia, conidia, mature conidia and pycnidia were measured. After the incubation period, semipermanent glycerin preparations at 50% were prepared and examined with an optical microscope (Leica DM500, Heerbrugg, Germany) with a digital camera installed, with a 40x lens and the images were processed using the LAS EZ Software (version 3.4; Leica Microsystems, Germany).

Pathogenicity test. Based on the highest incidence found in the isolates, four C. sinensis plants were inoculated with L. pseudotheobromae in sour orange C. aurantium. The surface of the healthy plant tissue was disinfected with 70 % alcohol and a lesion was made in which a disk of active, 5-day old L. pseudotheobomae growth in PDA (5 mm) was placed (^{Bautista-Cruz et al., 2019}; ^{Berraf-Tebbal et al., 2020}). The lesions were healed using wet cotton and Parafilm paper (Bemis, U.S.A.). The control plants received a sterile PDA disk. The plants were evaluated under microtunnel conditions for 21 days, with a mean temperature of 24 °C (max. 47 °C, man. 11 °C) and a relative humidity of 97 % (max. 100 %, min. 85 %). Once the period is finished, the presence of signs and symptoms of gummosis was found. The leaves and stems with lesions were used, and the fungi were reisolated in PDA medium. When the culture was developed, the morphological and cultural characterization was carried out to confirm the presence of structures that coincide with the original isolation.

In vitro evaluation of chemical fungicides and antagonists. The technique consisted in the planting of the isolates in a PDA medium with fungicides at different concentrations (^{Dhingra and Sinclair, 1995}). A 5 mm inoculant dick was extracted from the edge of the pure culture of the fungus in question. Four chemical fungicides were evaluated (Benomyl, Mancozeb, Carbendazim, Thiabendazole) and a biological product based on B. subtilis, strain QST 713. For the three isolations, seven concentrations were used with three repetitions, including the control (Table 1). The concentrations were calculated according to the volume of PDA medium and active ingredient. They were incubated at a temperature of 24°C. The inhibition percentages of the mycelial growth (IPMG) and the effective concentration (EC) were obtained. Once the mycelial growth of the control occupied the total area of the Petri dish, the growth percentage of each one of the treatments, was calculated using the formula by ^{Arce-Araya et al., (2019}): % fungal growth = [(Fungi diameter in PDA without fungicide - Fungi diameter in PDA with fungicide) / (Fungi diameter in PDA without fungicide) * 100]. The EC that inhibits 50 % of the mycelial growth (EC₅₀) was calculated using the GraphPad Prism Software (version 8.0.1; GraphPad, San Diego, CA).

Dual in vitro Trichoderma test. In order to evaluate the antagonistic ability of Trichoderma, the dual culture technique, described by Morton and Stroube (1955), was used. The test was carried out with the phytopathogenic agents obtained, and with two antagonistic agents: Trichoderma harzianum and T. viride, which belong to the INIFAP collection. The test consisted in placing a mycelium disk of each fungus (5 mm) on an edge of the Petri dish and on the opposite edge, a disk of the antagonistic agent (5 mm) was placed, with a separation of 2 cm. For each treatment, three repetitions and one control were included. They were kept at a temperature of 24 °C. Upon the termination of the period, the % of radial inhibition (PRI) was calculated using the formula by Osorio et al. (2016): PRI= [(Fungi diameter without Trichoderma-Fungi diameter with Trichoderma) / (Fungi diameter without Trichoderma) * 100].

Statistical analysis. The experimental design used was completely randomized. An analysis of variance was carried out along with a comparison of means using Tukey’s test (P ≤ 0.05). The results were analyzed using the SAS statistical software (version 9.0; SAS Institute, Cary, N.C.).

Results and Discussion

Morphological characterization of fungus associated with gummosis. Cultures were obtained from the plant samples that correspond to only one fungus with the following characteristics: A whit, circular growth and a filamentous edge with a spongy texture that turned military green and finally black (Figure 1a). Picnidia, 133 µm wide x 212.69 µm long appeared, with conidia that measured (17.75-) 25.30

Figure 1 Front and back of isolates in PDA culture medium of fungi related to the gummosis of citrus fruits. A) Six-day old Lasiodiplodia pseudotheobromae culture, obtained from tissue from the cortex. B) Six-day old Fusarium solani culture. C) Six-day old Pestalotia spp. culture, both obtained from the soil. (-34.97) µm long x (11.37-) 13.30 (-18.08) µm wide, sub-ovoidal to ellipsoidal in shape, with a truncated apex and base (Figure 2).  

The conidia are initially hyaline and aspetated, and then brown and septated with a longitudinal striated. In addition, other structures such as conidiogenous cells and paraphyses appeared. These characteristics correspond to those described by Phillips et al. (2013) and Liang et al. (2020) for L. pseudotheobromae.

Figure 2 Microscopic morphology of fruiting Lasiodiplodia pseudotheobromae bodies. A, B and C) Globose pycnidial conidiomata, D) mature septate conidia with longitudinal striations, E) conidiogenous cells and paraphyses and F) immature aseptate conidia. 

From the soil samples, cultures were obtained with a circular, cream-colored growth, filamentous edge and flat texture that turned light brown (Figure 1b). The slightly curved, triseptate macroconidia that measured (28.99-) 35.02 (-42.76) µm in length by (4.18-) 5.18 (-6.99) µm in width. The microconidia with or without a septum, (6.86-) 10.02 (-14.08) µm in length by (5.24-) 3.75 (-2.41) µm in width. The long and cylindrical phialides measuring 97.30 µm, and double-wall chlamidospores, terminal or intercalary measuring 8.69 µm in diameter (Figure 3). The above description coincides with F. solani (Leslie and Summerell, 2006). In addition, a third, white culture with circular rings was formed, cottonlike in texture

Figure 3 Microscopic morphology of Fusarium solani associated to the gummosis in citrus fruits. 

A) Phialides and conidia grouped into flower heads, B) Mature septated conidia, C) Mature microconidia. and filamentous edges (Figure 1c). Presence of acervuli measuring 506 µm and curved conidia measuring (13.19-) 21.61 (-28.07) µm in length (5.93-) 8.01 (-9.97) µm in width, with five cells of hyaline ends (terminal and apical). The terminal end presented three appendages that measured 4.96-30.3 µm, 5.33-31.06 µm and 5.28-29.81 µm. On the other hand, the apical end of just one short, 2.86-7.83 µm long appendage. This description coincides with the Pestalotia genus, with no possible distinction of a species, according to Guba (1961) and ^{Barnett and Hunter (1998}).

Isolate pathogenicity test. On day 21, necrotized lesions were observed with amber-colored exudates in stems, as well as the presence of pycnidia in stems and leaves (Figure 4). Similar to findings by ^{Bautista-Cruz et al. (2019}) and Berraf- Tebbal et al. (2020). The causal agent of the gummosis disease was confirmed to correspond to L. pseudotheobromae, using Koch’s postulates.

Effectiveness of fungicides. Our results helped identify the fungicide with the highest effectiveness in the reduction of the mycelial growth of F. solani, Pestalotia spp. and L. pseudotheobromae. For F. solani, the analysis of variance displayed significant differences between Carbendazim, Thiabendazole and Serenade (P ≤ 0.05). The Thiabendazole was the most effective chamical fungicide, since minimum doses were able to inhibit F. solani (Table 2, Figure 5). The benzimidazoles such as Carbendazim and Thiabendazole have been effective to control the Fusarium genus (^{Agrios, 2005}). Zárate-Ramos et al. (2022) determined that Thiabendazole

Figure 4 Symptoms and signs caused by Lasiodiplodia pseudotheobromae on citrus fruits, A) stem necrosis, B) presence of pycnidia on stems, and C) production of exudate or gum. 

was effective with an EC50 of 14.50 mg L^-1 against F. incarnatum, while Medina- Osti et al. (2022) reported that Thiabendazole was effective at an EC₅₀ of 7.2 mg L^-1 against F. sacchari. In addition, the, biological product Serenade was effective (Table 2, Figure 5). The bacteria of the Bacillus genus have been reported to produce antimicrobial compounds against Fusarium (Mardanova et al., 2017). In the scae of F. solani, to inhibit more than 50 % of the fungus, a concentration of 0.0497 mg L^-1 was needed (Table 2). Carbendazim proved to be efficient to inhibit

Figure 5 In vitro effect of fungicides against fungi associated to the gummosis of citrus fruits in five days. Growth of Fusarium solani: A) Medium with Thiabendazole at 0.5 mg L^-1. B) Medium with Carbendazim at 0.8 mg L^-1. C) Medium with Benomyl at 50 mg L^-1. D) Control. Growth of Lasiodiplodia pseudotheobromae: E) Medium with Thiabendazole at 0.01 mg L^-1. F) Medium with Benomyl at 0.1 mg L^-1. G) Medium with Mancozeb at 50 mg L^-1 H) Control. Growth of Pestalotia spp.: I) Medium with Thiabendazole at 0.01 mg L^-1. J) Medium with Benomyl al 1 mg L^-1. K) Medium with Mancozeb at 100 mg L^-1. L) Control. 

its mycelial growth (Table 2, Figure 5). In general, Carbendazim has displayed good results against F. oxysporum in tomato (Jahanshir and Dzhalilov, 2010) and against F. solani in chili pepper (Madhavi and Bhattiprolu, 2011). González-Oviedo et al. (2022) reported that strains of F. oxysporum, from vanilla, were sensitive to Benomyl and Carbendazim. However, Benomyl was less effective, since the highest percentage of inhibition was displayed at high elevations (Table 2, Figure 5). Applying Benomyl on the soil restricted the colonization of F. oxysporum in melon (Maraite and Meyer, 1971). Benomyl has been shown to reduce the germination rate of the Fusarium conidia (^{Decallonne and Meyer, 1972}). Romero- Velázquez et al. (2015) determined that an EC50 of 0.01 mg L^-1 was presented by B. subtilis against Fusarium in chayote. On the other hand, Zarate-Ramos et al. (2022) indicated that B. subtilis was efficient with an EC50 0.00014 mg L^-1, and completely inhibited Fusarium at 0.01, 0.05 and 1 mg L^-1, On the other hand, the analysis of variance (P ≤ 0.05) for Pestalotia spp. proved that there are significant differences between Benomyl and Mancozeb. Thiabendazole proved to be the most efficient in all the concentrations evaluated (Table 3). Hernández-Ceja et al. (2021) reported that Thiabendazole, starting at 5 mg mL^-1 inhibited 100% of the fungi associated with the regressive death of the cranberry: Pestalotiopsis clavispora, Colletotrichum gloeosporioides and L. pseudotheobromae. With Serenade, an optimum value was obtained of EC₅₀ 0.04870 mg L^-1. Monroy and Lizarazo (2010) found no antagonistic properties against Pestalotia spp., with the cultivation of the fungus with Pseudomonas fluorescens and B. subtilis bacteria at a concentration of 10⁶ UFC/mL⁸ in PDA. Mancozeb was the least effective (Table 3). The 100 % inhibition of the fungus was achieved with the use of the highest dose evaluated. In strawberry, ^{Ara et al. (2017}) and Rajnish & Gauta, (2022) managed to completely inhibit Pestalotia spp. at concentrations of 250, 500 and 1000 mg L^-1. On the other hand, Carbendazim and Mancozeb were the ones to control P. anacardii in mango (Patil et al., 2019).

In L. pseudotheobromae, the analysis of variance indicates that there were no significant differences on the EC₅₀ between Benomyl, Serenade and Thiabendazole, although they do differ statistically with Mancozeb (P ≤ 0.05). Thiabendazole was effective, since it displayed a lower value for EC₅₀ (Table 4, Figure 5) and it displayed a greater inhibition, starting at 0.01 mg×L^-1 (98.55%). Benzimidazoles have been effective against L. theobromae. Out of 120 L. theobromae isolates, 91.6% were reported to be sensitive to benzimidazoles with EC₅₀ values ranging from 0.36 to 1.27 µg mL^-1 for Thiabendazole (^{Da Silva et al., 2012}). Under field conditions, ^{Camacho-Tapia et al. (2021}) showed that the use of Thiabendazole provided good control over gummosis in lemon trees. Control with Benomyl was

^zMeans with the same letter in the column are statistically equal, according to Tukey’s test (P ≤ 0.05). ** = Undefined (GraphPad Prism does not show a complete trust interval). Value of P = 0.0004. achieved at low concentrations (Table 4, Figure 5); of Lasiodiplodia was achieved with an application of 0.1 mg L^-1 (86.58 %) and complete inhibition, at 10 mg×L^-1 (99.56 %). This is similar to results by Da Silva Pereira et al. (2021), in which the EC of Benomyl for L. theobromae was between 0.002 and 1.75 µg mL^-1. In addition, some species of Botryosphaeriaceae were controlled with Benomyl (EC ranging from 0.36 to 0.55 µg mL^-1) (^{Bester et al., 2007}). The application of Benomyl and copper oxychloride-based compounds against L. theobromae is effective in different phenological stages of the crop (Sáenz et al., 2019). Serenade was the least effective, with an EC value of 0.0528 mg L^-1. Mancozeb managed to control L. pseudotheobromae at high concentrations of 50, 100, 500 mg L^-1, compared to the evaluated fungicides. This was confirmed by ^{Dianda et al. (2020)} and Sultana and Ghaffar (2010), since Lasiodiplodia was completely inhibited at 100 and 500 mg L^-1. Mancozeb, along with Carbendazim, has allowed for broader control (Jadeja and Bhatt, 2010; Valle- de la Paz et al., 2019b). On the other hand, Sultana and Ghaffar (2010) obtained good results with the application of B. subtilis in pre- and post-emergence control of L. theobromae.

Effect of Trichoderma. Both T. harzianum and T. viride inhibited the development of fungi related to the gummosis of citrus fruits; however, according to the ANOVA, T. harzianum caused the greatest inhibition. Its effectiveness as a biological agent control against fungi, nematodes and insects has been reported (Ferreira and Musumeci, 2021). In F. solani, the ANOVA indicates that T. harzianum (61.08%) displayed significant differences in comparison with T. viride (22.17%) (P ≤ 0.05) (Table 5, Figure 6). Likewise, ^{Fernández and Suárez (2009)} reported that applying T. harzianum inhibited over 50 % of Fusarium in passionfruit. In eggplants, Ganesh and Dwivedi (2019) reported that around 20 % of Fusarium was inhibited

^zMeans with the same letter in the column do not differ statistically according to Tukey’s test (P ≤ 0.05). LSD; Least significant difference. with T. viride. Madhavi and Bhattiprolu (2011) indicate that the integration of different treatments such as soaking seedlings with Carbendazim, the addition of vermicompost, soaking with fungicide and the application of T. viride is efficient for the control of the wilting disease by Fusarium in chili pepper. For Pestalotia, the analysis of variance (P ≤ 0.05) indicates that T. harzianum (62.93%) displayed significant differences in comparison with T. viride (53.78%) (Table 5, Figure 6). In mango, the fungus was inhibited with the application of Trichoderma (72.88%) (^{Bhuvaneswari and Rao, 2001}). In the case of Lasiodiplodia, the ANOVA (P < 0.05) indicates that T. harzianum (35.64%) did not display significant differences in comparison with T. viride (25.45%) (Table 5, Figure 6). ^{Boat et al. (2022}) determined that T. harzianum reduced L. theobromae by 64.1%, whereas Bhuvaneswari and Rao (2001) reported that 62.41% of the fungal mycelial growth was reduced. Likewise, Da Silva et al. (2022) showed that Trichoderma reduced the growth of F. solani (34%) and L. theobromae (89%) related to Nopalea cochinillifera. In Morelos, L. citricola was found to be sensitive to the evaluated doses of T. harzianum (0.55, 0.39 and 0.19 g/100 mL) (Valle-de la Paz, 2019a).

Figure 6 Effect of Trichoderma spp. In the development of fungi associated to the gummosis of citrus fruits. A) Fusarium solani vs T. harzianum. B) F. solani vs T. viride. C) Pestalotia spp vs T. harzianum. D) Pestalotia vs T. viride. E) Lasiodiplodia pseudotheobromae vs T. harzianum. F) L. pseudotheobromae vs T. viride. 

Conclusions

The morphological identification and pathogenicity tests showed that the disease of gummosis in citrus fruits from the town of Plan Chac in Sacalum, Yucatan, is caused by L. pseudotheobromae. In addition, other phytopathogenic fungi were identified: Pestalotia spp. and F. solani. The treatment with Thiabendazole, B. subtilis strain QST 713 and T. harzianum control F. solani, L. pseudotheobromae and Pestalotia spp. efficiently. The management of gummosis fruits can be carried out in combination with the use of biological agents, thus reducing the use of chemical control and avoiding a possible long resistance with its long-term use.