Introduction

Theobroma cacao L. (Malvaceae) is a native plant from South America, widely cultivated in the world tropics (Motamayor et al. 2002). The Amazonian region shared between Peru, Ecuador, Colombia and Brazil harbors the highest diversity of the crop and the most ancient evidence of its domestication and consumption (Bustamante et al. 2022; Osorio-Guarín et al. 2017; Thomas et al. 2012). In the Amazonas department, Peru, T. cacao is produced mainly under organic and agroforestry systems, representing the main source of income for 11,623 families (Oliva et al. 2020). In 2016, the cacao produced in the provinces of Bagua and Utcubamba, in Amazonas, became protected by the Peruvian appellation rules under the name "Cacao Amazonas Peru", in recognition for its superior qualities of flavor and aroma (Oliva et al. 2020).

Fungal diseases are the main limiting factor of the crop worldwide (Huda-Shakirah et al. 2022). Currently, the main fungal diseases reported on Peruvian cacao farms are frosty pod rot (caused by Moniliophthora roreri), witches' broom (M. perniciosa), and the emerging thread blight disease (Marasmius infestans) (Díaz-Valderrama et al. 2020; Huaman-Pilco et al. 2023). Additionally, cacao is affected by other yield-reducing pathogens such as Lasiodiplodia spp. (Puig et al. 2021; Rahim et al. 2022). Lasiodiplodia is a cosmopolitan genus of fungal pathogens, with a predominance in tropical climates causing wood diseases (dieback, canker, leaf blight, and root and fruit rot) on a widespread variety of economically important crops (Mohali et al. 2005; Zheng et al. 2021). Its capacity to survive in the soil and vegetative residues makes its phytosanitary management challenging (Picos-Muñoz et al. 2014). In recent years, multiple species of this genus have been reported in different crops (Moreira-Morrillo et al. 2021). These include L. iraniensis described for the first time from diseased plants of Salvadora persica, Juglans spp., Eucalyptus spp., Citrus spp., and Prunus spp. in Iran (Abdollahzadeh et al. 2010). This pathogen was also associated with mandarin (Al-Sadi et al. 2013), mango (Marques et al. 2013; Rodríguez-Gálvez et al. 2017; Sakalidis et al. 2011), Bougainvillea spectabilis (Li et al. 2015), and Anacardium occidentale (Netto et al. 2017). Lasiodiplodia brasiliensis was first described causing papaya stem-end rot (Netto et al. 2014), and L. jatrophicola and L. euphorbicola on nuts in Brazil (Machado et al. 2014). In Italy, L. citricola was associated with trunk diseases in grapevines (Carlucci et al. 2015). The most common species, L. theobromae, is pathogenic on more than 500 hosts (perennial fruit and nut trees, vegetables and ornamentals) causing cankers, necrosis and rotting of stems and roots (Gnanesh et al. 2022).

The germplasm botanical garden of native cacao accessions from Northern Peru, managed by the Instituto de Investigación para el Desarrollo Sustentable de Ceja de Selva (INDES-CES) of the National University Toribio Rodríguez de Mendoza de Amazonas (UNTRM-A) (Oliva-Cruz et al. 2021, 2022) often harbors trees with a high incidence of symptoms of dieback on twigs, fruit rotting, and stem and leaf necrosis, typical of Lasiodiplodia spp. infection, which affects their performance. In this study we aimed to conduct a morphological and molecular identification of the causal agents of these Lasiodiplodia-like symptoms in the INDES-CES cacao germplasm collection.

Materials and methods

Sampling, incidence, and isolation

Between November 2021 and March 2022, we conducted phytosanitary inspections of native cacao accessions in the INDES-CES cacao germplasm collection, located in the Cajaruro district, Utcubamba province, Amazonas department, Peru (E: 792283; N: 9364101; 648 m a.s.l.). The samples were collected in February 2022. A total of 10 accessions (220 plants in total) were evaluated. We collected samples of diseased cacao tissues from plants showing dieback on young stems, and leaf spots (Fig. 1). The incidence of stem dieback was determined using the formula: I (%) = (ni/ N) × 100 (ni: total number of plants affected, N: total number of plants evaluated) posed by Rodríguez-Gálvez et al. (2017).

Fig. 1
figure 1

Sample types of diseased cacao tissues associated with Lasiodiplodia spp. in the INDES-CES germplasm botanical garden. A Tree showing terminal branch dieback. B Leaf spot. C Branch showing rot symptoms after pruning

Full size image

Small sections (~ 1 × 1 cm2) of symptomatic tissues were sterilized in 2% sodium hypochlorite (NaClO) for 2 min. They were then washed in sterile distilled water for 1min, three times. The surface-sterilized sample was dried with a sterile paper towel, transferred to potato dextrose agar medium (PDA), and incubated at 28 °C for 3–5 days. Six fungal strains were obtained using the hyphal tip subculture technique (Table 1).

Table 1 KUELAP herbarium voucher numbers of Lasiodiplodia spp. isolates and source
Full size table

Morphological identification

The macro and micro morphological characteristics of isolates were evaluated and compared to characteristics reported in previous relevant literature (Phillips et al. 2013; Sakalidis et al. 2011). Three isolates for each strain were incubated at 28 °C for 12 h day/night, and the colony diameter was measured with a digital caliper every 24 h for three days. The isolates were checked every week for the formation of conidiomata, which were observed and photographed on a stereoscope Nikon SMZ18 (Tokyo, Japan). Conidia morphology (cell wall thickness, form, color and presence or absence of septa) was evaluated on an inverted microscope OLYMPUS DP74 (Tokyo, Japan). Conidia were mounted on slides using lactophenol. We measured at least fifty micro structures. Each isolate was preserved in a metabolically inactive state as dried culture in the KUELAP herbarium of the UNTRM-A (voucher numbers in Table 1).

Molecular identification and phylogenetic analysis

Total genomic DNA was extracted from mycelium growing on PDA using the Wizard® DNA kit (Promega, Madison, WI, USA.) as indicated by the manufacturer. PCR assays were performed to amplify the internal transcribed spacer region (ITS1, 5.8S and ITS2 rDNA regions; ITS), and the partial translation elongation factor 1-α (tef1), β tubulin (tub2) and RNA polymerase II subunit (rpb2) genes. We used the primer pairs ITS1F/ITS4, EF1-688F/EF1-1251R, Bt2a/Bt2b, and rpb2-LasF/rpb2-LasR, respectively, as in Huda-Shakirah et al. (2022). PCR products were sequenced at Macrogen (Seoul, South Korea). Raw sequences were edited and assembled with Sequencher 5.4.6. Cleaned sequences and other sequences from reference specimens (Table 2) were aligned with MUSCLE (Edgar 2004) in MEGA 11.0 (Kumar et al. 2018). We generated a concatenated dataset with the four loci obtained with Seaview (Gouy et al. 2010). A maximum likelihood phylogenetic analysis with the concatenated dataset was performed in the CIPRES gateway (Miller et al. 2010). The phylogenetic tree was mid-point rooted and edited with FigTree 1.4.4. (Rambaut 2018).

Table 2 GenBank accession numbers of DNA sequences of Lasiodiplodia spp. included in the phylogenetic study
Full size table

Pathogenicity tests

Pathogenicity tests were conducted to fulfill Koch's postulate on cacao stems and fruits. Tests on stems were conducted following the mycelial disc method (Puig et al. 2021). Stems of seedlings were superficially disinfected with 70% alcohol. Then, a small wound (5 mm long) was made on each disinfected stem with a sterile scalpel. A disk (5 mm in diameter) of five-day-old isolates grown on PDA at 28 °C were inoculated on two two-month-old cacao seedlings for each Lasiodiplodia strain; an additional seedling was inoculated with a disc without the fungus for each isolate as a control. Finally, the inoculated areas were covered with parafilm. All inoculated plants were maintained at 28 °C in a growth chamber.

Pathogenicity tests were also conducted on healthy and mature harvested cacao fruits. Three fruits were used for each of the six strains, two of them were inoculated and one was used as a control. Whole fruits were surface-disinfected in 2% NaClO for 2 min, then washed in sterile distilled water. They were then dried and placed on sterilized Petri dishes within clean and NaClO-disinfected plastic containers. Two UV-sterilized sheets of paper towel humidified with sterile water were placed next to Petri dishes to maintain the relative humidity high. Fruits were wounded with a sterile scalpel to a depth of approximately 5 mm, and mycelium discs (5 mm in diameter) were placed on the wounds. No-fungus PDA disks were used on control fruits. The containers were sealed with stretch film and incubated at 28 °C for 8 days. Pathogenicity tests on seedlings and fruits were replicated twice. When the disease symptoms appeared, the pathogen was re-isolated using the same isolation procedure described earlier. A morphological characterization of the new isolate was performed to confirm that it was the one originally inoculated.

Results

Incidence evaluation

A total of 220 plants in a plot of approximately 0.25 ha were evaluated within the INDES-CES germplasm collection (Table 3). Regressive stem dieback of the cacao crop was detected in all accessions. Accession INDES-50 had the highest incidence of dieback and leaf spot, with 63.6% of plants affected, followed by INDES-67 (59.1%), INDES-54 (57.6%) and INDES-53 (54.5%). The other accessions showed incidence percentages below 50% namely INDES-55 (22.7%), INDES-24 (24.2%), INDES-31 (24.2%), INDES-83 (27.3%), INDES-27 (36.4%), INDES-65 (36.4%).

Table 3 Evaluation of the incidence of dieback in the INDES-CES cacao germplasm collection
Full size table

Morphological characterization of the strains

All six strains had the same growth rate with a mean of 23.47 ± 0.186 (standard error) mm/day at 28 °C after 3 days in PDA medium. We observed dense light-gray mycelium on the first days of growing, turning dark-gray as the days progressed (Fig. 2). After 40–60 days, conidiomata were observed in the deep black colony.

Fig. 2
figure 2

Colony and conidia characteristics of Lasiodiplodia spp. A–C Lasiodiplodia theobromae INDES-JHP40 macro and micromorphological characteristics: a seven-day old colony in PDA culture medium; B hyaline and aseptate immature conidia; C mature conidia. D–F Lasiodiplodia iraniensis INDES-JHP61 macro and micromorphological characteristics: D seven-day old colony in PDA culture medium; E immature conidia; F mature conidia. Photographs of conidia were taken at × 100 (oil immersion). Scale bar = 20 μm

Full size image

All strains showed morphological characteristics typical of Lasiodiplodia spp. (Phillips et al. 2008, 2013), such as, slowly maturing conidia, subovoid to ellipsoid ovoid in shape, with a broadly rounded apex and a truncated base tapering towards the base. Immature conidia were initially double-layered, hyaline and unicellular (Fig. 2). Mature conidia became dark-reddish brown with a central septum, a thick cell wall, and forming longitudinal striations (Fig. 2). Mean length and width of mature conidia did not differ among strains INDES-JHP23 (24.75 × 13.81 μm), INDES-JHP40 (24.98 × 14.08 μm), INDES-JHP57 (24.76 × 13.75 μm), INDES-JHP60 (25.17 × 14.29 μm), INDES-JHP61 (23.16 × 13.31 μm), INDES-JHP62 (25.11 × 13.98 μm) on average.

Molecular identification and phylogenetic analysis of the strains

Phylogenetic analysis with the multi-locus concatenated data set (ITS, tef1, tub2, and rpb2) identified the species of the six strains obtained from T. cacao symptomatic tissues at the INDES-CES germplasm collection. Isolates INDES-JHP23, INDES-JHP40, INDES-JHP57, INDES-JHP60, INDES-JHP62 grouped with the type and reference specimens of L. theobromae with a strong bootstrap support of 76%, while isolate INDES-JHP61 clustered with L. iraniensis type and reference specimens with a support of 100% (Fig. 3).

Fig. 3
figure 3

Phylogenetic tree based on maximum likelihood analysis inferred from a multi-locus concatenated alignment of internal transcribed spacer (ITS) sequences, the partial translation elongation factor 1-α gene (tef1), β-tubulin genes (tub2) and RNA polymerase II subunit (rpb2). Numbers on nodes represent bootstrap values (only values greater than 50% are indicated). Isolates sequenced in this study are shown in bold. Type specimens are marked with an asterisk. Lasiodiplodia crassispora was included as outgroup

Full size image

Pathogenicity tests of the strains

Both L. theobromae and L. iraniensis isolates were pathogenic to cacao stems and fruits, inducing lesions of similar appearance (Fig. 4). Stems inoculated with L. theobromae and L. iraniensis showed visible lesions four days after inoculation. Symptoms appeared as a dark brown lesion, progressing longitudinally from the inoculation sites. The infected tissue turned brown and rotted, subsequently causing dieback after four weeks. Then, black conidiomata were observed on the diseased area (Supplementary Figure S1). These symptoms and signs were the same as those observed in the field. Control seedlings did not develop any symptoms. Additionally, symptoms such as black lesions were observed on fruits inoculated with L. theobromae and L. iraniensis after two days of inoculation, covering the entire surface as the days progressed after 30 days (Supplementary Figure S2). At this point, a large amount of gray mycelium was observed on the entire surface of the fruit. The fungus was then reisolated from infected stems and pods, showing macro and micro morphological characteristics identical to those observed in the field, fulfilling Koch's postulate.

Fig. 4
figure 4

Pathogenicity of Lasiodiplodia spp. on cacao seedlings and fruits. A, B Stem rot caused by L. theobromae INDES-JHP23 and L. iraniensis INDES-JHP61, respectively, four weeks after inoculation. C Control plant. Red arrows show the point of inoculation. D, E Fruit rot caused by L. theobromae INDES-JHP62 and L. iraniensis INDE-JHP61, respectively, three days after inoculation; the dark brown zone shows that the fungi are still growing and colonizing healthy fruits. F Control fruit inoculated with a disk of PDA without the pathogen showing no visible symptoms

Full size image

Discussion

In the present study, L. theobromae and L. iraniensis were identified as the causal agents of cacao young stems dieback and fruit rot in the INDES-CES native cacao germplasm collection. In recent years, the severity and damage of Lasiodiplodia spp. have increased, causing various phytosanitary problems in different crops including cacao (Gnanesh et al. 2022; Pereira et al. 2006; Pisco-Ortiz et al. 2024). The symptoms observed at the INDES-CES germplasm collection had been commonly associated with Phytophthora damage due to their similar symptoms on fruits and shoots by local cacao producers. Previous investigations described that Lasiodiplodia species transitioned from endophytic to opportunistic pathogens, and they are now considered a threat to different crops of agricultural interest including cacao (Ali et al. 2020; Salvatore et al. 2020). Also, it has been shown that plants subjected to biotic and abiotic stresses are more susceptible to fungi of the genus Lasiodiplodia (Moreira-Morrillo et al. 2021; Pereira et al. 2006). In addition to this, mechanical pruning in cacao is a recommended management practice, but the wounds caused by this activity provide an access point for Lasiodiplodia infection.

Stem dieback and leaf spot were detected in all cacao accessions evaluated at the INDES-CES germplasm collection. However, different levels of incidence were observed. The clones with the highest incidence (50–65%) were INDES-53, INDES-54, INDES-50, INDES-67. On the other hand, those that showed lower incidence (20–40%) were INDES-24, INDES-27, INDES-31, INDES-55, INDES-65, INDES-83. In recent months, the increase in the incidence of Lasiodiplodia diseases is causing concern. Pruning wounds not only provide an entry point for pathogens, but are also a source of stress to the plant (Adu-Acheampong et al. 2012), which may be the cause of the increased incidence of the disease (Rodríguez-Gálvez et al. 2017). Rains and high humidity also play a role on Lasiodiplodia diseases as they favor the production of fungal spores which can be disseminated by raindrops and wind (Vásquez-López et al. 2008). The incidence of Lasiodiplodia spp. is also influenced by temperature above 30 °C, water stress and low levels of plant nutrition (Gunamalai et al. 2023).

Symptoms observed in the field, as well as those obtained from pathogenicity tests in this study, revealed substantial similarity to those of dieback diseases caused by Lasiodiplodia spp. These included a fast-growing, cream-white colony in the first few days, turning dark to black as the days progressed; conidia were dark brown, striated, ellipsoidal and uniseptate (Mbenoun et al. 2008). However, the fungus may show aseptate conidia when young, but these conidia may develop septa as they mature (Burgess et al. 2006; Phillips et al. 2013). In addition, the microscopic morphological characteristics of the conidia were consistent with those reported for this species in previous studies (Coutinho et al. 2017; Huda-Shakirah et al. 2022). The shape and color of mature conidia, as well as the presence of septa and longitudinal striae were important features for the identification of Lasiodiplodia spp. However, even though L. theobromae has slightly larger conidia than L. iraniensis (23.6–28.8 × 13–15.4 μm vs 22.51–26.09 × 12.75–14.97 μm, respectively), it is not sufficient to distinguish them (Abdollahzadeh et al. 2010; Marques et al. 2013). Here, we also found L. theobromae (23.4–26.8 × 12.8–15.2 μm; n = 50) has slightly larger conidia than L. iraniensis (22.2–25.4 × 11.9–14.2 μm; n = 50). However, given that morphological characters are insufficient to identify Lasiodiplodia species, molecular phylogeny has become an important tool for species identification (Marques et al. 2013; Pavlic et al. 2004). Other studies showed that phylogenetic analysis with a single locus, such as ITS, is unable to determine species in the genus Lasiodiplodia, so additional loci are required (Alves et al. 2005; Ismail et al. 2012). In recent studies, taxonomists have frequently used highly conserved protein-coding genes such as tef1, tub2, and ITS to construct species-resolving phylogenies (Phillips et al. 2019; Slippers et al. 2014).

The phylogenetic analysis with the concatenated dataset of ITS, tub2, rpb2 and tef1 sequences clearly placed our isolates within the L. theobromae and L. iraniensis species clusters with reference specimens from previously published studies (Netto et al. 2014; Phillips et al. 2008, 2013). Lasiodiplodia theobromae was the most frequent species in this study, as also found in previous studies (Marques et al. 2013; Netto et al. 2014). This confirms the wide distribution of L. theobromae throughout the INDES-CES cacao germplasm bank in Amazonas, Peru. Only one strain of L. iraniensis was found. This species was first described in Iran and can infect different hosts, such as Salvadora persica, Juglans sp. Citrus sp. and Mangifera indica (Abdollahzadeh et al. 2010). It has also been reported in Brazil on Mangifera indica (Al-Sadi et al. 2013; Marques et al. 2013), Bougainvillea spectabilis (Li et al. 2015), Anacardium occidentale (Netto et al. 2017), and coffee (Ramos et al. 2023).

Finally, the pathogenicity of L. theobromae and L. iranensis was confirmed after inoculation of cacao stems and fruits. Even though L. iraniensis had a lower prevalence, it was a species that showed the same aggressiveness as L. theobromae during pathogenicity tests. Therefore, both species are a threat to this crop. These findings are relevant for management strategies since the disease significantly reduces cacao production upon favorable conditions such as intense rainfall, prolonged drought and the presence of wounds on the plant. We therefore report for the first time L. theobromae and L. iraniensis are the causal agents of dieback and leaf spots in cacao plantations from Northern Peru. This research may also be useful for future studies and could help to find effective management strategies of this disease that represents a threat to cacao cultivation in the Amazonas department, Peru.