Effect of temperature on in vitro germination and growth of Colletotrichum fioriniae, a new emerging pathogen of olive fruits

Abstract Olive anthracnose induced by different Colletotrichum species causes dramatic losses of fruit yield and oil quality. The increasing incidence of Colletotrichum fioriniae (Colletotrichum acutatum species complex) as causal agent of olive anthracnose in Italy, is endorsing new studies on its biology, ecology, and environmental factors such as temperature. Five isolates from different sampling sites in Lazio region (Central Italy) were studied under controlled laboratory conditions aiming to better understand the differences of thermal development among the isolates and to lay the foundations of a future mathematical model able to describe the key aspects of the pathogen's life cycle. The mycelial growth rate and the conidial germination rate were assessed at seven different constant temperatures (5, 10, 15, 20, 25, 30, and 35°C) and fixed relative humidity (100% RH). The obtained dataset was analysed to estimate the parameters of mathematical functions that connect the mycelial growth rate and the spore germination with the environmental temperature. The parameters set provided as the result of this study constitute a key step forward in the biological knowledge of the species and the basis for future formulations of mathematical models that might be the core of decision support systems in an integrated pest management framework.


INTRODUCTION
Olea europaea L. is among the most important tree crops worldwide, with a surface of over 12 million hectares devoted to the production of oil and table olives (https://www.internationaloliveoil.org).The greatest part of the production is concentrated in the Mediterranean basin, its native area, where Spain, Italy, Greece, Tunisia, and Portugal are the main growers (Fraga et al., 2020;Rossini, Bruzzone, et al., 2022).The olive plant is known for its excellent adaptability to different environmental and climatic conditions (Fraga et al., 2020).This feature endorsed the expansion of olive farming in Countries outside the native area, as recently occurred in different areas of south-western Asia, Oceania, South Africa, and the Americas (Mousavi et al., 2019).
This phenomenon is gradually exposing olive plants to new abiotic and biotic stresses that affect the production.Moreover, the expansion of the olive cultivation and the climate change is also endorsing the reemergence of autochthonous diseases that are quickly expanding in areas where historically the outbreaks were contained.It is the case of the recent spread of the olive anthracnose, caused by different fungal species belonging to the genus Colletotrichum, which is leading to severe yield losses and to a strong decrease of the oil quality (Moral et al., 2021).
The species belonging to this genus have different biological behaviours, ranging from endophytic to necrotrophic, and are characterized by a high phenotypic and genotypic plasticity and diversity that explain its variability and aggressiveness (Cacciola et al., 2012;Moral et al., 2010, p. 20;Talhinhas et al., 2011;Talhinhas & Baroncelli, 2021).The ability of this group of fungi to survive and multiply without causing visible symptoms and/or remain in a quiescent state may explain why many producers suffer unexpected preharvest losses by not observing symptoms during the season (Moral et al., 2021).Typical olive anthracnose symptoms consist of depressed, round, and ochre/ brown lesions leading to fruit rot with evident great orange conidial masses.The pathogen can also cause the blight of olive flowers, mainly when mummies remain attached to the plant until the blossoming (Cacciola et al., 2012;Moral & Trapero, 2012), or cause the dieback of the branches because of the production of Aspergillomarasmine-A, a phytotoxin produced in the rotten fruits (Ballio et al., 1969;Moral et al., 2021).The infected fruits mummify and fall on the soil in late autumn or in winter, under low temperature and higher relative humidity (RH) conditions.
To date, 18 Colletotrichum species have been associated with olive anthracnose worldwide (Garcia-Lopez et al., 2023).These species belong to three Colletotrichum complexes, namely Colletotrichum acutatum, Colletotrichum boninense, and Colletotrichum gloeosporioides, but the most recent studies highlighted that the main species responsible of the disease are C. acutatum (sensu stricto), Colletotrichum fioriniae, Colletotrichum godetiae, Colletotrichum nymphaeae, Colletotrichum rhombiforme, and Colletotrichum simmondsii (Garcia-Lopez et al., 2023;Moral et al., 2021).It is worth saying that multiple species can be present at the same time in the same olive growing areas, and in this case, it is typical to find a dominant and a secondary species (Garcia-Lopez et al., 2023).
To date, the knowledge on key biological features of C. fioriniae, such as the thermic requirements for the mycelial growth and conidial germination, is still scarce.This lack of information hampers the formulation of proper control strategies and the development of decision support systems (DSS) based on mathematical models.DSS are gaining more importance in agriculture because of their capability to simulate different scenarios of plant infestations by insect pests or infections by fungal and bacterial pathogens (Capalbo et al., 2017;Knight & Mumford, 1994;Körner et al., 2014;Robinet et al., 2012).Having a reliable DSS would be a valuable tool to prevent or contain the spread of the diseases (Rossi et al., 2019).The first step toward modelling is to analyse, individually, the response of the pathogen to the external environment by accurate laboratory trials, where isolates of the fungus are cultivated under different constant conditions of temperature, humidity, and water activity.
This work aims to investigate the thermal response of different isolates of C. fioriniae collected in an important olive productive area of Central Italy, providing: (i) a detailed analysis, for the different isolates, of the mycelial growth under constant temperature conditions in a laboratory environment, and (ii) a first mathematical interpretation of the life cycle of this relevant pathogen, in particular of the mycelial growth and conidia germination rates.We believe that this set of quantitative information is a key propaedeutic step to extend the biological knowledge and to reach a final effective DSS for this pathogen.Laboratory experiments are, in fact, the first important step to carry out to provide the set of parameters that can be further implemented in more extended models, which, once developed, hold the potential for practical application in open field scenarios.Furthermore, the coexistence of various isolates within a single field highlights the importance of conducting this analysis, given that different responses to temperature fluctuations can be a synonym of shorter or longer risk of infection in function of the field microclimate conditions.

Fungal isolation, morphological, and molecular characterization
During the growing season 2022, a survey was carried out in 10 fields located in the Viterbo province (Lazio, Central Italy), in which 210 olive fruits showing typical anthracnose symptoms were collected and further analysed in the laboratory.The external surface of the collected olive fruits was sterilized in a 2% sodium hypochlorite solution for 2 min, and subsequently rinsed three times with sterilized distilled water.The olive fruits were then sectioned in slices of 1 mm of thickness and each slice was subsequently placed in a Petri dish with Potato Dextrose Agar (PDA).The Petri dishes were incubated at 25 C and 100% RH and inspected after 7 and 14 days.Colonies showing morphological traits of Colletotrichum spp.were placed in separate dishes to obtain pure cultures.

Phylogenetic relationship
The sequences obtained were concatenated and compared with the corresponding sequences from C. fioriniae and related species from NCBI, listed in Table S1.The latter include C. nymphaeae, C. simmonsdii, and C. acutatum as part of the Colletotrichum acutatum species complex, and the sequences of the isolates defined by Chen et al. (2022) and Zhang et al. (2023a) as Colletotrichum orientalis and Colletotrichum radermacherae, but deposited as Colletotrichum sp. or as C. fioriniae.The sequences of ITS, TUB2, ACT, CHS-1, and GAPDH were concatenated and aligned to the sequenced amplicons of the five isolates from olive using MUSCLE v3.8.31 (Edgar, 2004).The alignment file was the input for building a maximum likelihood (ML) phylogenetic tree using RAxML-HPC v8.2.12 (Stamatakis, 2014), set with GTRCATI algorithm as a substitution model and 1000 bootstraps.The tree was visualized using FigTree v1.4.4 (http://tree.bio.ed.ac.uk/ software/figtree/) and further edited with Inkscape v 0.92 (https://inkscape.org).

Experimental design for mycelial growth at different constant temperatures
The mycelial growth rate of the selected C. fioriniae isolates was explored at seven constant temperatures, that is, 5, 10, 15, 20, 25, 30, 35 C, and fixed relative humidity of 100% RH, so that we could focus as much as possible on a single, and important, growth factor.We evaluated four replicates per isolate per each constant temperature to avoid errors due to potential anomalies in the growth, and the number of repetitions has been increased by measuring the radius in multiple directions, as detailed below in this section.
A 4 mm ; mycelial plug from 7 days-old actively growing colony margins of each Colletotrichum isolate were placed on the centre of PDA medium (Microbiol ® , Italy) plates.The plates were subsequently transferred in an incubator where temperature and humidity were maintained constant for all the duration of the experimentation.
The mycelial growth was measured with a ruler (±1 mm) after 7 days, a time range low enough to ensure that the fungus did not reach the border of the dishes, following the four orthogonal directions (N-W-S-E) from the centre of the plug (Figure 1).The four directions measured in each dish ulteriorly increase the precision of the measure of the radius.

Experimental design for conidial germination at different constant temperatures
Conidia were scraped from the mycelium and subsequently filtered using four layers of cheesecloth to remove any bigger fragment (Drais et al., 2021).The conidial suspension was then quantified using a hemocytometer and diluted to obtain a final concentration of 10 5 spore/mL.After the quantification, 5 μL of conidial suspension was placed on 4 mm ; Water Agar (Concentration of 12 g/L) plugs placed on a microscope glass slide.The slides were subsequently placed on Petri dishes and incubated at constant temperature conditions of 5, 10, 15, 20, 25, 30, and 35 C and 100% of relative humidity, analogously to the mycelial growth.Four replicates per isolate per each constant temperature were evaluated.
The conidia were considered as 'germinated' once the length of the germ tube was more than one half of the length of the spore, as shown in Figure 2. Germinated conidia were counted at fixed time ranges of 6, 10, 15, 20, 24, and 48 h by staining with lactophenol cotton blue.The percentage of germinated spores was determined by observing random groups of 100 conidia per replicate, so that the results are directly expressed as a percentage.

Mycelial growth rate data analysis
Before the modelling approach, the mycelial growth dataset was analysed in a more classical way using the RStudio software v. 4.3.2(R Core Team, 2018), so that we could highlight eventual differences among the isolates.Notably, we first carried out a wider analysis of the whole dataset, then we focused on the response of the isolates to each constant temperature explored in the experimentation.

Whole dataset
Before the analysis, we first checked the normality of the complete dataset through a Shapiro-Wilk test, using the shapiro.test()function within the basic R environment, and through a visual inspection of the Quantile-Quantile (Q-Q) plot, draw using the qqmath() function within the lattice R package (Sarkar, 2008).Given the non-normal trend of the dataset, the best function to transform the dataset was chosen by using the bestNormalize() function within the R package bestNormalize (Peterson, 2021;Peterson & Cavanaugh, 2020).The transformed dataset was then analysed through a linear model (LM) by using the lmer() function within the R package lme4 (Bates et al., 2015) and considering temperature and isolate as independent variables and the plate and the orthogonal direction (N-W-S-E) of the measured radii as random effects.The LM was subsequently followed by a Bonferroni post hoc test (α = 0.05) to assess the differences among the temperatures and the isolates by using: the emmeans() function within the R package emmeans (Searle et al., 1980), the pairs() function within the R package multicompView, and the cld() function within the R package multcomp (Hothorn et al., 2008).

Temperature-by-temperature analysis
Before this second part, the whole dataset was divided into specific sub-datasets, where the measures of the radii of the different isolates were grouped by temperature.The normality trend of each sub-dataset was checked before the analysis, as well as eventual transformations were carried out in the same way of the general dataset.The analysis procedure was the same as the general dataset, except for the use of a LM considering only the isolate as independent variable and plate and orthogonal direction of the measured radii as random effect.

Modelling the mycelial growth rate in function of temperature
A single value of the radius from each replicate was obtained from the dataset of each isolate by considering the average and the standard deviation of the four radial distances measured after 7 days.The average radius of each Petri dish was initially expressed in millimetres per week (mm/w), but a further conversion was carried out by dividing the mean and the associated standard deviation by 7, so that the unit was mm/day (Drais et al., 2023).The converted dataset of each isolate was subsequently interpolated with the Briére equation (Briere et al., 1999), whose mathematical expression is the following: where T is the temperature of growth, a and m are empirical parameters, and T L and T M are the minimum and maximum thresholds for the mycelial growth, respectively.The parameters and the associated standard errors of the Equation (1) were obtained by least squares fit, while the goodness of fit was evaluated through a χ 2 -test and the coefficient of determination R 2 (Bellocchi et al., 2011;Ikemoto & Kiritani, 2019;Rossini et al., 2019aRossini et al., , 2019b;;Rossini, Bono Rossell o, et al., 2021;Rossini, Severini, et al., 2020;Rossini, Speranza, & Contarini, 2020;Rossini, Virla, et al., 2021).
The Equation ( 1) has a typical increasingdecreasing profile with a maximum, coinciding with the optimal temperature, T opt , for mycelial growth.By placing the first derivative of temperature to zero, namely d dT R T ½ ¼ 0, it is possible to obtain the abscissa of the maximum (Briere et al., 1999;Rossini, Bono Rossell o, et al., 2022;Rossini, Bruzzone, et al., 2022;Rossini, Severini, et al., 2020): additional quantitative information of fundamental importance.It is worth remarking that the least squares fit provides the errors associated with the parameters, so that it is possible to calculate the error associated with T opt by applying the propagation of the uncertainty formula to the Equation (2), as detailed in Rossini, Contarini, et al. (2020).

Modelling the conidial germination rate
The experimental setup to assess the conidial germination under different temperatures provided a dataset distributed according to a logistic function.The main assumption is that as time passes, the percentage of germinated conidia reaches 100%, more or less faster depending on temperature.The dataset of each isolate at each constant temperature listed in Experimental design for mycelial growth at different constant temperatures section was interpolated with the following logistic equation describing the germination rate over time (Gabriel y Gal an et al., 2015;Prosser, 1995): The parameters k, G 0 , and r in Equation ( 3) are respectively the carrying capacity, the number of germinated conidia after 6 h (the moment of the first sampling), and the instantaneous germination rate.The experimental set up led us to consider k ¼ 100, with the advantage of estimating only two parameters (G 0 and r) instead of three.As for the Briére equation (Experimental design for conidial germination at different constant temperatures section), the parameters were estimated through a nonlinear least squares regression, while the goodness of fit was evaluated through a χ 2 -test and considering the coefficient of determination R 2 .Before the fitting operation, the dataset of each isolate and for each constant temperature was organized considering the four values measured at each sampling time, for a total of 24 values.
From the step described above we obtained the profiles of G 0 and r over temperature.We can therefore interpolate the instantaneous germination rate over temperature, T , using the following equation (Gougouli & Koutsoumanis, 2012;Omuse et al., 2021;Rosso et al., 1993Rosso et al., , 1995)): where r opt , T max r , T min r , T opt r represent the value of the instantaneous germination rate at the optimal temperature and the maximum, minimum, and optimal temperature for the conidial germination, respectively.The interpolation was carried out through a non-linear fit using the least squares method, but given the reduced number of degrees of freedom, the goodness of fit has been entrusted only on the coefficient of determination R 2 .
Parameters estimation and analysis of the Equations (1-4) The non-linear regressions mentioned in Mycelial growth rate data analysis and Modelling the mycelial growth rate in function of temperature sections were carried out through ad hoc Python scripts publicly available at https:// github.com/lucaros1190/Colletotrichum-Temperature.
The repository also contains the raw dataset, the R script used in Experimental design for conidial germination at different constant temperatures section, and all the additional information to fully reproduce the results of this study.

Morphological characterization of C. fioriniae isolates
Five representative isolates were chosen for downstream analysis, according to the uniqueness of their morphological traits.After 10 days at 25 C, pure cultures of the five isolates showed a grey to orange cottony and aerial mycelium, from pale orange to dark red in reverse (Figure 3).Conidia were hyaline, smooth-walled, aseptate, narrowly elliptical pointed at both ends, measuring 11-21 μm (mean 15.5 μm) Â 3.5-7 μm (mean 4.5 μm), and contained in orange masses.

Molecular and phylogenetic characterization of C. fioriniae isolates
After PCR amplification of the six partial genes and Sanger sequencing, the BLASTn analysis revealed 100% identity of our isolates with the deposited sequences of C. fioriniae.Our sequences were deposited under the Gen-Bank Accession no.ON773228-ON773232 for ITS, ON807567-ON807571 for TUB2; ON791496-ON791500 for ACT, ON791501-ON791505 for CHS-1, ON791 506-ON791510 for HIS3, ON791511-ON791515 for GAPDH, respectively (Table 1).
The phylogeny inferred by ML analysis (Figure S1) indicates that the olive isolates under investigation are included in the C. fioriniae cluster.Colletotrichum orchidophilum CBS-632.80 was here used as an outgroup.

Mycelial growth rate-Experimental data analysis
From the first step of the analysis described in Mycelial growth rate data analysis section, we assessed that the mycelial growth of the overall isolates over temperature was significantly different (LM, p < 0.0001), namely there were significant differences among all the experimental temperatures explored.For the sake of exposition, we hereafter report only the p-values obtained by the analysis in Mycelial growth rate data analysis section, referring the most interested reader to the shared scripts (Parameters estimation and analysis of the Equations (1-4) section) and dataset for further details.Additionally, in case of multiple comparisons, the greater p-value is reported between parentheses after the results, so that the other ones are implicitly meant smaller.An overall idea of the experimental dataset is provided by the boxplots in Figure 4, reporting the raw data F I G U R E 4 Mycelial extension in terms of radial growth (millimetres per day) at different constant temperatures ( C).Each graph reports a boxplot of the raw data corresponding to each isolate considered in this study.

ENVIRONMENTAL MICROBIOLOGY REPORTS
(converted in mm/day) collected for each isolate.Notably, the dataset of all the isolates presents an increasing-decreasing profile, with a well identifiable maximum coinciding with the optimal temperature for the mycelial growth.The absence of outliers suggests that there were no anomalies referable to systematic errors during the experimentation, even if they were considered in the LM analysis by the two random effects (plate and orthogonal direction).
Excluding temperature as a factor, it was assessed that the isolates COL-2, COL-4, and COL-6 showed no overall statistical differences in the mycelial extension rate (LM, p = 1), while the isolates COL-1 and COL-3 were different to each other, and to the rest of the isolates (LM, p < 0.0001).
By grouping the dataset by temperature, instead, it was possible to appreciate the differences among the isolates at the various temperatures with a major detail.As shown in Figure 5, at 5 C the isolates COL-3, COL-4, and COL-6 were statistically different from each other (LM, p < 0.047), while the isolates COL-1 and COL-2 were statistically different only from COL-3 (LM, p < 0.0001).At 10 C the isolates COL-1 and COL-6 did not report statistical differences (LM, p = 1) while being both different from the other isolates (LM, p < 0.0001); COL-2, COL-3, and COL-4 were different from each other as well, (LM, p < 0.001).At 15 C the isolate COL-3 and COL-6 showed the lowest and the highest mycelial extension rate, respectively, statistically different from the other isolates (LM, p < 0.002); the isolates COL-1, COL-2, and COL-4, instead did not show statistical differences (LM, p = 1).The growth at 20 and 25 C showed the same behaviour, with no statistical differences among the isolates COL-1, COL-2, COL-4, and COL-6 (LM, p = 1).The isolate COL-3, instead, was the lowest one in terms of mycelial extension, and it was different from the other isolates (LM, p < 0.0001).At 30 C, COL-1 and COL-3 showed the highest and the lowest mycelial extension rate, respectively, and they were different both from each other and from the other isolates (LM, p < 0.001).Conversely, the isolates COL-2, COL-4, and COL-6 were not different from each other (LM, p > 0.6).At 35 C, the last temperature explored, the isolates COL-3 and COL-1 were the lowest and the highest in terms of mycelial extension rate, respectively, and statistically different from each other F I G U R E 5 Mycelial extension in terms of radial growth (millimetres per week) at different constant temperatures ( C).Each graph reports a boxplot of the dataset grouped by temperatures.Different letters mean statistical differences assessed with a linear model followed by a Bonferroni post hoc test (α = 0.05).
In light of these results, it is noteworthy that both lower and upper temperatures provoke a differentiation among the isolates, highlighting their positive or negative response to temperature variations.Conversely, the lower differences occurred between 20 and 30 C, indicating that this range includes the optimal temperature for mycelial growth.This first analysis of the dataset highlights important aspects that will be further refined in the subsequent section, where we present the results of the model parameters' estimation.

Mycelial growth rate-Mathematical interpretation
According to the experimental and data analysis protocols described in Modelling the mycelial growth rate in function of temperature section, the first result concerns the measurement of the mycelial extension rates over temperature and the subsequent estimation of the parameters of the function (1).A quantitative interpretation of the dataset was provided by the interpolation of the dataset with the Briére Equation ( 1), whose results are listed in Table 2 and graphically represented in Figure 6.The overall optimal temperature for the mycelial growth is around 24 C, with a slightly lower value of T A B L E 2 Best fit parameters (±standard error) of the Briére function ( 1) describing the mycelial extension rate over temperature.

Isolate Briére function's parameters
Goodness of fit and optimal temperature Note: Additional information is reported by the coefficient of determination R 2 , by the number of degrees of freedom (NDF), and χ 2 value.Figure 3 shows the raw dataset utilized for parameter estimation, while Figure 4 provides a graphical representation of the best fit functions.Optimal temperature T opt was calculated through Equation ( 2).The (*) symbol above the χ 2 values mean a significance of p < 0.01 to the χ 2 -test.
F I G U R E 6 Best fit functions (1) describing the mycelial growth rate over temperature of each isolate considered in this study.The corresponding parameters and their standard errors are listed in Table 2.
T opt ¼ 23 AE 1 ð Þ C calculated for the isolate COL-6 and a slightly higher value of T opt = (25 ± 4) C for the isolate COL-1.Besides the variations, the optimal temperatures for the mycelial growth are all in accordance with each other, considering the associated standard errors.
A higher variability among the isolates was observed in the minimum temperatures for the mycelial growth, T L .The values, in fact, range from T L = (À10 ± 4) C calculated for COL-1, to T L = (2 ± 1) C of COL-3 while the other isolates are around a T L value of À3 C. The maximum temperatures for the mycelial growth showed a lower variability, with values included between 35.4 and 36 C.These results are in line with the analysis of the dataset described in Mycelial growth rate data analysis section and provide a higher level of definition.

Conidial germination rate
The data collected from this part of the experimentation allowed us to estimate the parameters of the logistic Equation ( 3), reported as a supplementary material with standard errors and goodness of fit values (see the repository link).As described in Modelling the conidial germination rate section, the Equation (3) provided the instantaneous and the initial germination rate over temperature for each of the isolates considered in this study.
The instantaneous germination rate over temperature was described by the function ( 4), whose parameters, specific for each isolate considered in this study, are listed in Table 3, and that are graphically represented in Figure 7.The optimal temperature for the germination, T opt r , was the same (19 C) for all the isolates, in the limits of the standard errors.This result may indicate a similarity between the isolates, at least in terms of adaptation to the environment, however the main T A B L E 3 Best fit parameters (AE standard errors) estimated for the function (4) describing the instantaneous germination rate over temperature, r T ½ .

Isolate
Best fit parameters values

Goodness of fit parameters
COL-1 r opt ¼ 0:37 AE 0:06 Note: Additional information is reported by the coefficient of determination R 2 and by the number of degrees of freedom (NDF).
F I G R E 7 Best fit functions describing the instantaneous germination rate r T ½ of the isolates considered in this study, and mathematically described by the Equation ( 4).The best fit values and their standard errors are instead listed in Table 3.
differences assessed can be found by looking at the values of the conidial germination rate at the optimal temperature, r opt , (Table 3).These values showed an overall variability, whose extremes are represented by the slower isolate COL-3 (r opt ¼ 0:13 AE 0:03) and the faster one COL-6 (r opt ¼ 0:5 AE 0:1).The minimum temperatures for the conidial germination were all concentrated between 1 and 2 C, with higher uncertainties, if compared with the other parameters.This is likely due to the fitting algorithm, which may have higher difficulties in fitting values next to zero.The upper limits for the conidial germination were concentrated among 34 and 36 C, with COL-1 and COL-4 the least and COL-3 the most tolerant to higher temperatures.
According to the values listed in Table 3, COL-3 is the isolate whose conidia germinate slower at the optimal temperature, but at the same time it covers a higher thermal spectrum, between 2 and 3 C wider with respect to the other isolates.Analogously, COL-6 seems to be more tolerant to higher temperatures (just 1 C less than COL-3), but the germination rate at the optimal temperature is the highest one, suggesting a prompt response in starting the infection on new plants.
The initial germination rates over temperature, instead, are graphically represented in Figure 8.Even if there is no further interpolation of these data, as we did for the instantaneous germination rates r T ½ , an analysis of the percentage of germinated conidia after 6 h provides an idea of the timing for an eventual control action for given environmental conditions.As observable in Figure 8, there are different scenarii among the isolates: for instance, COL-2 and COL-3 seem to have an increasing-decreasing profile where at 25 C the 63% and 74% of the conidia are already germinated, respectively.On the other hand, COL-1, COL-4, and COL-6 reach percentages of 95%, 85%, and 83% of Initial germination rate G 0 over temperature for the isolates considered in this study.
the conidia germinated at 30 C, suggesting a higher potential of infection at higher temperatures.

DISCUSSION AND CONCLUSION
Colletotrichum fioriniae is a widespread fungal species known for causing anthracnose in several crops.Belonging to the C. acutatum species complex, it was initially recognized as C. acutatum var.fioriniae (Marcelino et al., 2008) and the year after upgraded to C. fioriniae as a separated species by Shivas and Tan (2009).In a more recent paper, two subclades of C. fioriniae were proposed as separated species, namely C. fioriniae and C. orientalis (Chen et al., 2022), as well as a third group of closely related isolates were assigned to the new species C. radermacherae (Zhang et al., 2023a).In our phylogenetic analysis, the isolates are clustered both with C. fioriniae and C. orientalis (indicated as Colletotrichum sp. as deposited on the NCBI GenBank database).However, the very few nucleotide polymorphisms existing along the entire concatenated 5 gene sequences (7 SNPs out of 1600 bp) of these isolates raise uncertainties about their taxonomic arrangement in different species instead of being considered as representatives of intraspecific genetic variability, as originally proposed by Damm et al. (2012) and recently assessed by Zhang et al., 2023b.Such questions are not extraordinary in fungal taxonomy (Hil ario et al., 2021) and they have already interested other species complexes of the Colletotrichum genus, requiring specific coalescent approaches to be solved (Liu et al., 2016).With this premise, all the five isolates characterized in this study were classified, and their sequences were deposited, as C. fioriniae.
Here, we provided a set of information on the thermal response of the mycelial growth and the conidial germination rate of these five isolates, together with a set of biological parameters included in mathematical functions that provide a more refined description of the results obtained by laboratory experiment.Given the entity of the damages that this pathogen is causing on olive cultivations, a mathematical interpretation of its life cycle with respect to the effect of the external environment is of great help.This study is propaedeutic to further develop specific DSS that may be of great help to predict and effectively constrain the incidence of the olive anthracnose and will be fundamental for planning strategic field interventions to control this pathogen.Targeted interventions when environmental conditions are favourable to the development of the pathogen can lead to a significant reduction in disease incidence, a better management of costs, and an optimization of the use of plant protection products.It is worth remarking that we explored only the dependence on temperature, neglecting the role of the water lato sensu.While we recognize that parameters such as relative humidity, rain, leaf wetness or dew are important for the development of fungi, we are deferring these aspects to future studies, in order to concentrate our conclusion on individual parameters.Conversely, the choice for this work was the comparison of the thermal response of different isolates that coexist within the same olive productive area, an interesting aspect from a phytopathological point of view.As outlined in Introduction section, the prevalence of multiple isolates, even on the same plant, is a common occurrence for this pathogenic agent (Garcia-Lopez et al., 2023), and only simple and wellfocused laboratory experiments can provide clearer indications about their coexistence and interaction.
Our results showed different scenarii for the Colletotrichum isolates explored, presenting an overall variability of the parameters that may correspond to an earlier or delayed infection depending on the environmental conditions.Moreover, the differences assessed among the isolates, above all to low and high temperatures, leave us to suppose that the primary and secondary isolate responsible of olive anthracnose may change from field to field depending on the microclimate variations.In most of the cases, Colletotrichum spp. is present in the environment in crop residues and in mummified fruits, the main source of inoculum for the following season (Cacciola et al., 2012).The primary inoculum is caused by fresh conidia produced in acervuli at the beginning of the growing season as soon as the environmental conditions are favourable.According to the existing literature, there is production and germination of conidia once relative humidity values overcome 95% and temperature ranges between 10 and 30 C, while the formation of the appressoria reaches maximum levels as the RH approaches 100% (Estrada et al., 2000;Moral & Trapero, 2012).Our study quantitatively completes this information, providing a profile of the instantaneous germination rate at different constant temperatures.It is worth noting however that while the Briére Equation (1) correctly described the mycelial growth rates over temperature, the values related to the minimum thresholds T L should be regarded more as a theoretical limit.The reason is behind the mathematical structure of the Briére equation, that makes the least square fitting operation challenging for this specific parameter (Jin et al., 2022).A consequence is the higher relative error associated to T L , if compared with the other ones.In contrast to the Briére (1), the Rosso Equation (4) had a lower performance in terms of data fitting.The reason can be the high variability of the dataset of the conidial germination, probably due to the fast germination that occurs at temperatures near the optimum.The reactivity of the isolates in conidial germination was quantitatively described by the parameter G 0 of the logistic function (3), that clearly shows the different behaviour.At optimal temperatures, the overall isolates showed high germination rates after just 6 h, with differences mostly concentrated on high temperatures more than lower ones.As a possible consequence, we may say that while the germination of the conidia during the spring and autumn is more or less similar, during the summer, when warmer temperatures happen, there are some isolates more performant than others, representing a higher risk for the ripening fruits.
Considering the minimum temperature values for the germination, we can say that the conidia present in the environment can germinate, in Central Italy, even at the end of the winter, when the level of relative humidity and water availability in sensu latu is often high.If we consider, in addition, the temperature-dependent mycelial growth rate, we can conclude that after the germination, the mycelium is capable of growing at the same conditions, even if slowly.
The considerations discussed in the previous paragraphs are in line with the existing literature and justify why the symptoms become more visible as the fruit ripening is in an advancing state.Early in the season, the conidial germination and the subsequent mycelial growth are gradual, reaching a peak in late May.This fact is justified by our results, given that the optimal temperature is of $19 and 23 C for conidial germination and mycelial growth, respectively.These temperatures, on average, are prevalent during mid-spring and autumn.While the higher summer temperatures may slow down the development of the fungus, it can anyway reach and affect the small olives.At this point, the dry condition of the summer stops the growth of the pathogen until its end, when fruits are ripening, and the water level increases in the environment.It is however worth saying that in case of high availability of water either through rainfall or irrigation, the probability of germination of the conidia is high and reaches percentages up to 90% after 6 h as in the isolates COL-1, COL-4, and COL-6.
Besides this general overview, our study highlighted the differences between the isolates as well: for instance, the slowest conidial germination and mycelial growth rates have been observed for the isolate COL-3, setting it apart from the other isolates, which showed similar patterns.These differences may be due to the expression of some specific genes, as reported by (Li et al., 2021), responsible for producing specific proteins such as the Heterotrimeric G protein.This protein plays an important role in signal transduction in filamentous fungi, and the deletion the CgGa1 subunit negatively affects growth, asexual and sexual sporulation, appressorium formation, penetration, and pathogenicity of C. gloeosporioides.
If we look at the differences assessed on the mycelial growth, we can assert that except for the isolate COL-3, all the others have almost the same behaviour within the optimal thermal range (between 20 and 30 C), or at least that the differences between the isolates is minimal.The higher variability that occurs going toward the lower and the upper thermal limits, is instead worthy of further investigation given the role that it may have on primary infections or the mycelium's survival during summer/winter.
Colletotrichum fioriniae is an important hemibiotrophic pathogen of the C. acutatum species complex that is present on different host plants.For instance, in South Korea, it is often reported on peach (Lee et al., 2018), in China it is recurrently present on Vaccinium corymbosum (Castro et al., 2023) and kiwifruit (Xu et al., 2023), and in the United States it is has been reported on grapevine (Nigar et al., 2023).The wide number of hosts and the increasing spread in different parts of the world demonstrates how Colletotrichum spp.can easily adapt to different plant species and disparate environmental conditions.This is confirmed in Italy where, besides olive trees, C. fioriniae was reported as causal agent of post-harvest bitter rot of apple (Carneiro & Baric, 2021), of anthracnose on ornamental plants (Guarnaccia et al., 2021), walnut (Luongo et al., 2022) and beech (Giubilei et al., 2023).It is worth saying that before Italy, where it was first detected in Calabria and Lazio regions (Riolo & Cacciola, 2022), this pathogen has been found on olive cultivations in other areas worldwide, such as Australia (2009), Uruguay (2010), Portugal (2021), andCalifornia (2017;Moral et al., 2021).
The spread of this pathogen, as well as its adaptability, is endorsing control actions devoted to constraining the epidemics and development of accurate DSS based on mathematical models.Similar approaches have been carried out for several insect pests and pathogens (e.g., Donatelli et al., 2017;Gonz alez-Domínguez et al., 2014, 2020;Pfab et al., 2018;Rossi et al., 2003;Rossi, Caffi, Bugiani, et al., 2008;Rossi, Caffi, Giosuè, & Bugiani, 2008;Tonnang et al., 2017), but for each specific study case the estimation of the biological parameters was a fundamental preliminary step.Different authors already tried to develop models for different isolates or species of Colletotrichum (e.g., Perfect et al., 1999;Romero et al., 2022;Salotti et al., 2022Salotti et al., , 2023) ) and our work is an additional piece of information for further comparisons and to understand, from a quantitative point of view, the differences in terms of environments.

F
I G U R E 1 Measurement of mycelial growth.After 7 days the length of the radius was marked on the orthogonal axis (A) and subsequently measured with a ruler (±1 mm) (B).The example in the picture is specific for the isolate COL-1 at 25 C. F I G U R E 2 Count of the germinated and non-germinated conidia after 15 h at 25 C. Conidia were considered as germinated once the length of the germ tube was more than one half of the length of the spore, as indicated by the green arrows.Red arrows indicate examples of non-germinated conidia.Measurements were carried out at 40Â magnification.