All cultures were grown on half strength Potato Dextrose Agar (19.5 g PDA/Liter adjusted to 1.5% agar; IBI Scientific, USA) with continuous light and replated weekly. After 10-14 days spores were collected by removing a small square of agar 3cm from the leading edge of mycelia and pipetting 0.5 mL ddH₂O to suspend the conidia. The concentration was then adjusted to 5x10⁵ for all experiments. To observe conidiogenesis, the culture was grown on tap water agar with 2% agar and three autoclaved rice grains were placed on the surface of the agar to observe single conidiophores. After 14 days of growth with continuous light of cool white fluorescent 1050 Lumens at room temperature, agar blocks were cut into roughly 2 cm squares approximately 3 cm from the edge of the plate and inverted on a coverslip with 30 µL of water (Hickey and Read, 2009). To maintain hydration of the colony, 20 µL of water was added to the top-facing (opposite face of the agar block from the specimen) side of the agar block every hour of imaging, and 150 µL of water was placed in Microcentrifuge Tube Caps from Fisher Scientific and was kept in proximity to the agar block to avoid desiccation.

Microscopic observations were made using an Olympus FV3000 laser scanning confocal system interfaced with an Olympus IX83 inverted microscope with a Galvanometer scanner and High Sensitivity GaAsP PMT detectors. UAPON 100x OTIRF objective (NA = 1.49) with Olympus oil immersion type-F. The motorized stage enables optical sectioning in the z-axis, and the z-drift compensator allows stable time-lapse images over long time periods. Calcofluor-White (Sigma, USA) was excited at 405 nm and the detection wavelength was captured between 430-470 nm. Concanavalin A (Sigma, USA) was excited at 561 nm and the detection wavelength was captured between 570-670 nm. Images were exported via Olympus CellSens software desktop version 2.3, and images were analyzed with the image software ImageJ version 2.14.0 (Schneider et al., 2012) images were prepared in Photoshop version 25.3.1 20231212.r.241 (Adobe Inc.., San Jose, CA, United States, 2019).

A C. graminicola strain expressing Lifeact::GFP (CgAB221-17) was used to visualize actin (Wang and Shaw, 2016), all other strains are in Table 1 . Calcofluor-White (Sigma, USA) was used to visualize the cell wall at a working concentration of 50 µg/mL from a stock solution of 1 mg/mL. Concanavalin A (Sigma, USA) was used to visualize the adhesive at a final working concentration of 200 µg/mL from a stock concentration of 5 mg/mL. Dyes were then added to the spore solutions to allow for visualization of the cell wall and adhesive. Concanavalin A, a lectin that binds glycoproteins, was chosen to label adhesive due to its strong affinity for carbohydrate-rich secretions typical of fungal adhesins. Lifeact::GFP was used to label the actin array, providing a fluorescent marker for actin dynamics that aligns with adhesive strip formation.

DNA was extracted from cultures after one week of growth. Once the agar with fungal mycelium was cut out of the plate, it was mixed with lysis buffer (0.5M NaCl, 10mM Tris-HCl pH 7.5, 10mM EDTA, 1% SDS), placed in a bead beater, and followed by the phenol-chloroform method as done previously (Sambrook and Russell, 2006; Wang et al., 2016). After the internal transcribed spacer (ITS) region of 18S rDNA was amplified with PCR using universal primers ITS 4 (TCCTCCGCTTATTGATATGC) and ITS 5 (TCCGTAGGTGAACCTGCGG), as well as for actin (act) using act-512F (ATGTGCAAGGCCGGTTTCGC) and act-783R (TACGAGTCCTTCTGGCCCAT). PCR reactions were done with the following settings: Initiation for 5 min at 95°C, denaturation for one minute at 95°C, annealing for 30s at 55°C for ITS and 48°C for act, extension for 1 min at 72°C. The Thermal Cycler was run for 35 cycles and then a final extension cycle was run at 72°C for 5 min. Sequencing performed by Eton Bioscience Inc. Sequence quality was determined using SeqTrace software (Stucky, 2012).

Once the ITS and act region were amplified and sequenced, it was aligned in MEGA 11 using the ClustalW algorithm and manually analyzed and adjusted. Sequences were compared via Standard Nucleotide BLAST (Altschul et al., 1990) comparing to GenBank (Benson et al., 2018). Evolutionary analyses were conducted in MEGA11 (Tamura et al., 2021). Maximum likelihood phylogenetic trees were visualized in FigTree v. 1.4.4 (Rambaut, 2018) and edited in Photoshop version 25.3.1 20231212.r.241 (Adobe Inc., San Jose, CA, United States, 2019) for trait mapping. The resulting phylogeny was compared to (Talhinhas and Baroncelli, 2021).

The proportion of cell surface with adhesive coverage was quantified for 50 cells from each of eight species of Colletotrichum shown in Table 1 with 2 biological replicates. From these species there are three conidia morphologies that were observed: falcate with single-sided adhesive, falcate with double sided adhesive, and cylindrical with double sided adhesive. The surface area was computed using the 3D Objects Counter plugin in ImageJ. To facilitate cross-species comparisons, both cell walls and adhesives were considered, with their respective surface areas calculated. The process involved isolating the Calcofluor-White channel (cell wall stain) from the Concanavalin A channel (adhesive), applying the plugin, and determining a threshold to eliminate noise.

The proportions of cell surface occupied by adhesive were transformed with a logit transformation. Pairwise means comparisons of adhesive content for were calculated for Colletotrichum species and for spore shape morphologies using Tukey’s Range Test. Statistical testing was done using the GLM procedure of SAS v9.4 (SAS Institute Inc. Cary, NC).

In order to document the development of the adhesive strip, we used Concanavalin A to label developing conidia, as shown for mature conidia in Figure 1 . Concanavalin A, which binds to glycoproteins, labeled a distinct mucilage secretion at the cell-cell junction found at each individual phialide and its nascent conidium ( Figure 2 ). However, specific binding to the conidial adhesive in acervuli was not observed, likely due to interference from the surrounding mucilage matrix. Therefore, we decided to use the formation of the actin array as a proxy for the formation of the adhesive strip, given the established colocalization between the actin array and adhesive in mature conidia (Vasselli et al., 2022). Therefore, we reasoned that the formation of the actin array would function as an indicator of when the conidium formed the adhesive.

To determine the timing of the actin array development the acervulus was examined using the Lifeact strain of Colletotrichum graminicola (CgAB221-17) (Wang and Shaw, 2016). Conidia that remained attached to their mother cell exhibited no discernible actin array; however, they did retain observable actin patches ( Figure 3 ) ( Supplementary Video 1 ). To determine if actin cables form prior to detachment from the colony, it was necessary to observe nascent conidia emerging from their conidiogenous phialide.

Observations of developing conidia utilizing transmitted light are inhibited by the mucilage secreted around the acervulus that was previously described (Moraes, 1980; Sugui et al., 1998). However, the limitation was overcome with the use of a fluorescent cell wall stain, Calcofluor-White. A total of 4199 conidia were observed that remained attached to their conidiogenous cells. Of these, 99.8% of conidia which remained attached to their conidiogenous cell did not possess any observable actin array or actin cables. In contrast, 1544 conidia that had detached from their conidiogenous cell were observed. Of these, 95.7% of conidia that were detached from their conidiogenous cell displayed a fully formed actin array that ran from tip to tip of the falcate conidium.

This observation led to the hypothesis that the actin array formed only after detachment from the conidiogenous cell. Actin inhibition pharmacological agents were considered but deemed impractical due to the vital roles of the cytoskeleton in conidiogenesis, endocytosis, exocytosis, and hyphal growth (Schultzhaus and Shaw, 2015) (Gabriel et al., 2006; Samaj et al., 2004). However, long-term time-lapsed imaging of a single conidiogenous cell proved useful in determining the timing of the development of the actin array ( Figure 3 ). During the initial stages of development of the conidium, there was no observable actin array (00:00:00). Actin patches were observed in the apex of the conidium as the cell developed. When the conidium reached its entire length and appeared fully formed, the actin array was not detectable (04:00:00). The appearance of the actin array occurs only after the conidium detaches from its conidiogenous cell (06:00:00). Within three more hours, the actin array becomes fully formed and reaches its maximum intensity (08:30:00). The production of the next blastic conidium aids in the detachment of the previous conidium.

These observations support the conclusion that the actin array is formed after detachment of the newly formed conidium. Here, we use the presence of the actin array as a proxy for the formation of the adhesive strip, as previously described by the colocalization of the adhesive and the actin array (Vasselli et al., 2022). The presence of large amounts of gelatinous secretion around all acervuli of C. graminicola (Mondal and Parbery, 2005; Sugui et al., 1998) precludes a similar analysis of the development of the conidium adhesive because Concanavalin A also reacts with this matrix ( Figure 2 ), making observation of the development of the adhesive strip impossible using Concanavalin A.

Comparison of sequence between Internal Transcribe Region (ITS4/ITS5), as well as the actin gene (act-512F/act-783R) for each of the isolates, revealed a sequence match that was 96.96-100% identity to accessions in GenBank ( Table 1 ). An important observation is that in the BLAST analysis, the top results for C. truncatum (UvTX1) and C. truncatum (UvTX2) resulted in a 99.65% and 99.83% identity, respectively, with their ITS sequences to C. jasminigenum, a species synonymized with C. truncatum according to (Talhinhas and Baroncelli, 2021). Additionally, the act sequences of C. truncatum (UvTX1) and C. truncatum (UvTX2) results had a 99.6% and 99.2% identity, respectively, with C. truncatum. Each of the isolates belonged to five distinct species complexes of Colletotrichum as described by (Talhinhas and Baroncelli, 2021). These include the graminicola complex, the dematium complex, the spaethianum complex, the truncatum complex, and the gloeosporioides complex. To best represent the distribution of spore morphology and adhesive distribution, a maximum likelihood phylogenetic tree was generated ( Figure 4 ). The topology of the tree matched that described by (Talhinhas and Baroncelli, 2021). In constructing the phylogeny presented in Figure 4 , this study employs a more limited dataset, encompassing only eight species, and a narrower set of genes, in comparison to a comprehensive work conducted by (Talhinhas and Baroncelli, 2021). This phylogenetic tree, though narrower in scope, offers a supplementary perspective to the existing literature.

The observed species manifest variations in conidia morphology and adhesive distribution ( Figure 5 ). A consistent observation across all species, with the exception of C. truncatum, is the presence of an asymmetric adhesive distribution. Remarkably, C. truncatum deviates from this pattern by exhibiting adhesives on both sides of the conidia, despite its overall similarity in morphology to other falcate-shaped conidia members. Species such as C. theobrominicola, C. gloeosporioides, and C. siamense, display cylindrical-shaped conidia. Despite this disparity in shape, these spores exhibit an asymmetrical adhesive, aligning with the prevalent trend observed in most Colletotrichum species. In Figure 5 , this adhesive asymmetry is visible in the ‘on side’ view of the spores, where the adhesive appears on only one side of the conidium, despite seeming more evenly distributed in the overlay image due to the adhesive’s contact with the coverslip. These findings suggest that adhesive asymmetry in cylindrical and falcate spores may be an adaptive feature facilitating targeted attachment and improved dispersal across varying host surfaces. Such asymmetry could enhance spore adhesion efficiency, positioning Colletotrichum species for successful colonization in diverse environments.

For a quantitative analysis of the variations in conidia adhesion, the surface area of the adhesive was calculated ( Figure 6 ). Species with falcate conidia, C. sublineola, C. graminicola, C. circinans, and C. incanum as a group, significantly differ from other spore morphologies in proportional adhesive surface area ( Figure 7 ). However, within the falcate group, species exhibit variability in adhesive proportions, as shown by distinct statistical groupings ( Figure 6 ). Similarly, C. theobrominicola, C. gloeosporioides, and C. siamense share the cylindrical conidia morphology and, as a group, significantly differ from other spore morphologies in proportional adhesive surface area ( Figure 7 ). Within the cylindrical group, species exhibit distinct statistical groupings, reflecting variability in adhesive proportions ( Figure 6 ). C. truncatum represents a unique conidia morphology among the species tested, because although it is falcate in shape, it possesses adhesive strip on both sides of the spore. The two strains of C. truncatum, C. truncatum (UvTX1) and C. truncatum (UvTX2), which were isolated from different host plants, were not significantly different from each other in adhesive surface area proportion. However, these strains still formed distinct groupings compared to most other species. When modeled as a function of spore shape, the proportion of adhesive surface significantly differed between all conidia morphologies (p<0.05) ( Figure 7 ).

The lack of significant difference in adhesive surface area between the two strains of C. truncatum suggests that adhesive properties may not vary significantly within a species, even when strains originate from different hosts. However, the grouping of C. truncatum strains with other species suggests that broader morphological or ecological factors may also influence adhesive surface area proportions. Further studies could investigate the interplay between species-specific traits and environmental factors to better understand the drivers of adhesive variation within and across Colletotrichum species.

The observed dynamics in actin array formation and adhesive distribution across various Colletotrichum species present intriguing insights into the intricate processes governing conidial development and attachment. This investigation into the formation of the actin array during conidial development uncovered distinct patterns. Conidia detached from their conidiogenous cells consistently displayed an observable asymmetric actin array. Notably, incipient conidia still attached to their phialidic cell did not exhibit any observable actin array. This consistent pattern, observed in 99.8% of conidia while connected to their conidiogenous cell, suggests a crucial link between actin array formation and the detachment phase. Furthermore, detached conidia displayed the actin array in 95.7% of cases. One reason that some conidia do not possess the actin array when detached could be that some are forcibly detached as a consequence of mounting the slide, causing the cell to die before it could close the septum of the new conidium. The development of the actin array was clarified through time-lapsed imaging, unveiling a sequence of events. During initial development, no observable actin array was detected. Actin patches emerged in the apex as the conidium matured, yet the complete formation of the actin array became apparent only after detachment. This process underscores the speed of a pivotal event in the conidial life cycle. A reasonable assumption would be that the adhesive is secreted after the formation of the co-localized actin array ( Figure 1 ). However, the feasibility of directly assessing this hypothesis is impeded by the inherent limitation that the Con A dye, employed for visualizing the adhesive, intricately interacts with the mucilage matrix, rendering precise discrimination unattainable ( Figure 2 ).

In this study, five distinct species complexes within the Colletotrichum genus were observed. Despite the existence of fifteen distinct species complexes, our study underscores the pervasive conservation of the asymmetric adhesive in the majority of the observed complexes. The observed variations in both conidial and adhesive morphologies suggest a potential correlation with host-specific differences and diverse leaf morphologies. Future research should investigate not only the adhesive localization within the remaining species complexes of Colletotrichum but also within other fungal genera. Such endeavors promise to enrich our understanding of adhesive mechanisms and their evolutionary implications across diverse fungal taxa. Such investigations hold the potential to enhance our understanding of adhesive mechanisms and their evolutionary implications.

Conidia morphology in the genus Colletotrichum can be generally described by two morphological traits: spore shape, and adhesive deposition. Colletotrichum conidia can be either falcate or cylindrical in shape, with an intriguing discovery of asymmetric adhesive distribution unique to this genus. While the precise implications remain uncertain, there is speculation that this phenomenon may contribute to enhanced spore dispersal. Remarkably, C. truncatum conidia are falcate shaped, but the adhesive is deposited on both sides of the cell. These morphological differences between species present potential lines of inquiry host range modeling as a function of conidial adhesive distribution and shape. Further research should explore adhesive localization in additional species complexes within Colletotrichum. While the precise interplay between the asymmetric adhesive and its consequential impact on spore distribution remains an ongoing investigation, the influence of conidia shape on spore distribution emerges as another intriguing aspect for exploration. The elucidation of these dynamics not only underscores the complexity of adhesion but also calls for a comprehensive understanding of their ecological implications. This correlation prompts contemplation on the potential adaptive significance of adhesive abundance in Colletotrichum species, which could thereby enrich our understanding of host-pathogen interactions within this genus. The morphological diversity and adhesive variations observed across Colletotrichum species may reflect evolutionary adaptations to different ecological niches and host types. The observed asymmetry in adhesive distribution could be a strategy for selective attachment on surfaces that maximizes spore survival and dispersal efficiency. These morphological traits may be particularly advantageous in environments with variable humidity or exposure, where selective adhesion confers ecological benefits. Further studies investigating environmental factors affecting adhesive patterns could provide valuable insights into the ecological roles of these traits within the genus.