Temperature-Dependent Growth and Spore Germination of Fungi Causing Grapevine Trunk Diseases: Quantitative Analysis of Literature Data

Grapevine trunk diseases (GTDs) are currently considered to be among the most destructive diseases of grapevines worldwide (Fontaine et al. 2016b; Gramaje et al. 2018; Kenfaoui et al. 2022). These diseases are permanent, deep-seated, difficult to diagnose, and difficult to manage (Pascoe and Edwards 2002). GTDs reduce the vineyard lifespan and cause serious economic losses in the wine industry. The economic cost for the replacement of dead grapevines caused by GTD is estimated to be >$1.5 billion per year (Hofstetter et al. 2012).

Three GTD syndromes have been described in published articles: Botryosphaeria dieback (BD), esca complex (EC), and Eutypa dieback (ED) (Bertsch et al. 2013; Fontaine et al. 2016b; Mondello et al. 2018). These syndromes are caused by different fungi, which frequently coexist in affected plants.

The symptoms of BD include dead spurs or buds, stunted or delayed growth, leaf spots, dead arms, shoot dieback, and bunch rot. The main wood symptom is wedge-shaped, perennial cankers, similar to those caused by ED, or circular to nonuniform staining of the affected wood; pycnidia can also be found in dead or cankered wood. BD is caused by fungi belonging to the Botryosphaeriaceae family, and 26 species in the genera Botryosphaeria, Diplodia, Dothiorella, Lasiodiplodia, Neofusicoccum, Neoscytalidium, Phaeobotryosphaeria, and Spencermartinsia have been associated with BD of grapevines (Gramaje et al. 2018; Pitt et al. 2013a, c, 2015; Rolshausen et al. 2013; Úrbez-Torres 2011; Yang et al. 2017). The most common of these species are Botryosphaeria dothidea, Diplodia seriata, D. mutila, Neofusicoccum parvum, and Lasiodiplodia theobromae (Úrbez-Torres 2011).

The EC has five syndromes, including wood streaking disease, Petri disease, grapevine leaf stripe disease (GLSD), white rot, and esca proper (Mondello et al. 2018; Surico 2009). The dark wood streaking is the common internal wood symptom that appears in young plants and is caused by Phaeomoniella chlamydospora and Phaeoacremonium minimum. Petri disease involves a typical decline of young vineyards and is caused by Phaeomoniella chlamydospora, Phaeoacremonium spp., and Cadophora spp.; the typical internal symptoms consist of vascular discoloration, often accompanied by the formation of black tar droplets after wood cutting (Mondello et al. 2018). GLSD is characterized by a tiger stripe pattern on leaves and is often associated with the partial or complete apoplexy of affected plants; it is caused by Phaeomoniella chlamydospore and Phaeoacremonium minimum, which can be found in both young and mature vineyards (Mondello et al. 2018). White decay is caused by basidiomycetes (in Europe, Fomitiporia mediterranea) and does not cause specific aerial symptoms. Esca proper is characterized by the concurrent presence of major vascular pathogens (Phaeomoniella chlamydospore and Phaeoacremonium spp.) together with Basidiomycetes (Fomitiporia spp.), and by the development of foliar symptoms in mature vineyards (Mondello et al. 2018).

Foliar symptoms of ED include stunted shoots with chlorotic leaves that are often cupped and have necrotic margins (Gramaje et al. 2018). Symptoms on the wood consist of cordon dieback with loss of spurs and internal, necrotic, wedge-shaped staining in the cross-section of cordons and trunks. External cankers appear as the dieback progresses and are characterized by flattened areas of the wood with no bark, leading to eventual vine death (Gramaje et al. 2018). Perithecia of the causal fungus develop in the cankered wood and are embedded in the bark. To date, in total, 24 species in the family Diatrypaceae and in the genera Eutypa, Anthostoma, Cryptosphaeria, Cryptovalsa, Diatrype, Diatrypella, and Eutypella have been associated with ED (Luque et al. 2012; Pitt et al. 2013b; Rolshausen et al. 2014; Trouillas et al. 2010). Among these species, Eutypa lata is the most virulent and common (Carter 1991).

GTD fungi share a similar life cycle and some aspects of their epidemiologies (Agustí-Brisach et al. 2019; Billones-baaijens et al. 2017; Díaz and Latorre 2014; González-Domínguez et al. 2020; Halleen et al. 2020; Songy et al. 2019; Úrbez-Torres et al. 2010). Their spores (mainly conidia or ascospores, depending on the fungal species) are produced and released from fruiting bodies on diseased wood; they are then dispersed by rain splashes or by air currents, and are deposited on vines. After the spores germinate and the fungus enters the wood (mainly through pruning wounds), the hyphae grow in and around xylem vessels and parenchyma cells, producing the metabolites (e.g., enzymes and toxins) that contribute to the internal and external disease symptoms (Andolfi et al. 2011; Bertsch et al. 2013; Bruez et al. 2014; Gramaje et al. 2018; Luque et al. 2014).

Spore germination and mycelial growth can be considered the most critical stages of GTD fungi, because they determine the onset of wood colonization. The environmental conditions required for spore germination in some GTD fungi have been investigated in recent years (Amponsah 2010; Díaz and Latorre 2014; Ji et al. 2023; Úrbez-Torres et al. 2010), while the literature on mycelial growth is larger and involves a wide range of fungi (Agustí-Brisach et al. 2019; Amponsah 2010; Mostert et al. 2006; Qiu et al. 2016; Rooney-Latham 2005; Songy et al. 2019; Whiting et al. 2001). Thus, there is a need to take advantage of the substantial amount of published information in order to determine the similarities and differences in the environmental response of mycelial growth and spore germination among the different fungi responsible for GTDs. Systematic literature reviews and meta-analyses have been increasingly used in plant pathology (González-Domínguez et al. 2017, 2019; Ngugi et al. 2011; Philibert et al. 2012; Scherm et al. 2014) as a quantitative statistical analysis of several separate but similar individual experiments or studies in order to test the pooled data for statistical significance (Madden and Paul 2011).

In this work, we conducted a literature review, extracted quantitative data from published articles, and used the pooled data with four objectives: (i) to assess and synthesize findings on the effects of environmental factors on mycelial growth and spore germination of GTD fungi, (ii) to analyze the differences among GTD fungi, (iii) to develop mathematical equations that describe the response of mycelial growth to temperature, and of spore germination to temperature and moisture for GTD fungi, and (iv) to identify knowledge gaps that require further research. To our knowledge, this is the first article in which quantitative data from different individual experiments were pooled and used to study temperature-dependent growth and spore germination of fungi causing GTDs.

Materials and Methods

Systematic literature search

A systematic literature search was conducted by using the following digital databases: Scopus (https://www.scopus.com/), Web of Science (https://www.webofscience.com/), CAB Abstract (https://www.cabdirect.org/cabdirect/search/), and Google Scholar (https://scholar.google.com/). The following keywords were used in the literature search: (i) GTDs OR esca OR each genus listed in Table 1 of Mondello et al. (2018); (ii) germination OR mycelial growth; and (iii) temperature. To be included in this study, articles had to satisfy the following criteria: they (i) had to contain terms related to GTDs in the title, abstract, or authors’ keywords; (ii) had to contain original or fit data on the quantitative effect of temperature on mycelial growth or spore germination for at least one species of GTD fungi, and these data must originate from correct and repeated experiments (e.g., with proper experimental design, with different isolates, or in repeated experimental runs) so as to ensure data robustness; and (iii) had to be published in peer-reviewed scientific journals, conference proceedings, scientific reports, or peer-reviewed theses. The cut-off date for publications to be considered for inclusion in this analysis was 1 June 2022.

Articles found in the databases were first reviewed based on the information in the title and abstract. If the article met the eligibility criteria, it was considered of potential interest; otherwise, it was discarded. A cross-reference-based search of studies was performed to reduce the risk of missing information due to the choice of search terms. Selected articles were read in full, and the quantitative data were extracted.

Data collection

Data on the effect of temperature on mycelial growth and on the effect of temperature and moisture duration on spore germination for each fungal species were retrieved in text, tables, and figures of the selected studies; the GetData Graph Digitizer 2.24 (http://getdata-graph-digitizer.com; accessed on 3 July 2019) was used to obtain precise data from graphs.

To enable the comparison of data retrieved from studies conducted under different experimental conditions (e.g., different growing media, colony ages, incubation conditions and times, and types of measurement), original data were rescaled from 0 to 1. For the studies in which fungal growth or spore germination were investigated at different temperatures, moisture durations, and times, each growth or germination value was divided by the maximum obtained in the experiment. For instance, Tello et al. (2010) reported that Phaeomoniella chlamydospora cultured on potato dextrose agar (PDA) for 2 months at 25°C (which was considered the optimal temperature) and at 15°C had a colony daily growth rate of 1.07 and 0.60 mm/day, respectively; the rescaled values were then calculated as follows: Y₁₅ = 0.60/1.07 = 0.56 and Y₂₅ = 1.07/1.07 = 1. For the studies that included a temperature-dependent equation for fungal growth (Pitt et al. 2013a; Table 1), the equation was solved for discrete values for temperature, and the resulting values rescaled to the maximum. For the studies that only reported the cardinal temperatures (Crous and Gams 2000; Úrbez-Torres et al. 2012; Table 1), minimum temperature (T_min) and maximum temperature (T_max) were set at 0, and optimal temperature (T_opt) was set at 1.

Table 1. Estimated parameters (and standard errors) and goodness-of-fit to real data for equation 1, which predicts the effect of temperature on mycelial growth for grapevine trunk disease (GTD) fungal species and syndromes^a

Species, syndromes	Estimated parameters of equation 1						Goodness-of-fit to real data				N	Literature sources
	Tmin	Tmax	Topt	a	b	c	R2	RMSE	CRM	CCC
Botryosphaeria dothidea	3	41	28.3	6.450 ± 0.115	1.990 ± 0.036	1.248 ± 0.050	0.990	0.037	0.013	0.995	4	Pitt et al. (2013a), Qiu et al. (2016), Úrbez-Torres et al. (2006), Yan et al. (2012)
Diplodia corticola	5	37	27	7.267 ± 0.078	2.188 ± 0.024	1.480 ± 0.038	0.996	0.023	0.012	0.998	1	Agustí-Brisach et al. (2019)
Diplodia mutila	4	41	24.5	4.608 ± 0.031	1.243 ± 0.010	2.748 ± 0.061	0.997	0.021	0.017	0.998	5	Baskarathevan et al. (2012), Bellée et al. (2017), Pitt et al. (2013a), Úrbez-Torres et al. (2006), Agustí-Brisach et al. (2019)
Diplodia seriata	3	41	25.5	5.147 ± 0.043	1.444 ± 0.014	2.171 ± 0.053	0.996	0.024	0.010	0.998	8	Giambra (2016), Baskarathevan et al. (2012), Bellée et al. (2017), Pitt et al. (2013a), Qiu et al. (2016), Úrbez-Torres et al. (2006), Amponsah (2010), Agustí-Brisach et al. (2019)
Diplodia spp.	4	36	23.8	5.678 ± 0.207	1.614 ± 0.062	1.201 ± 0.098	0.980	0.076	−0.096	0.978	1	Giambra (2016)
Dothiorella americana	9	36	22.5	4.010 ± 0.039	1.000 ± 0.012	1.529 ± 0.037	0.997	0.022	0.009	0.998	1	Úrbez-Torres et al. (2012)
Dothiorella iberica	2	34	22.6	6.202 ± 0.052	1.801 ± 0.015	0.923 ± 0.015	0.999	0.014	0.001	0.999	1	Pitt et al. (2013a)
Dothiorella viticola	5	35	23.7	5.665 ± 0.054	1.654 ± 0.016	1.016 ± 0.020	0.998	0.017	0.004	0.999	1	Pitt et al. (2013a)
Lasiodiplodia missouriana	9	36	26.3	6.148 ± 0.148	1.788 ± 0.043	0.971 ± 0.047	0.989	0.040	0.009	0.995	1	Úrbez-Torres et al. (2012)
Lasiodiplodia theobromae	3	41	29.6	7.752 ± 0.107	2.346 ± 0.031	0.886 ± 0.023	0.996	0.023	0.005	0.998	3	Pitt et al. (2013a), Qiu et al. (2016), Úrbez-Torres et al. (2006)
Lasiodiplodia viticola	9	37	27.2	6.415 ± 0.111	1.857 ± 0.031	0.937 ± 0.032	0.996	0.026	0.005	0.998	2	Bellée et al. (2017), Úrbez-Torres et al. (2012)
Neofusicoccum australe	5	41	25.9	4.974 ± 0.028	1.377 ± 0.009	2.037 ± 0.032	0.998	0.015	0.006	0.999	4	Pitt et al. (2013a), Úrbez-Torres et al. (2006), Amponsah (2010), Baskarathevan et al. (2012)
Neofusicoccum luteum	5	41	26.1	5.038 ± 0.062	1.415 ± 0.020	1.892 ± 0.064	0.992	0.033	0.018	0.996	3	Úrbez-Torres et al. (2006), Amponsah (2010), Baskarathevan et al. (2012)
Neofusicoccum mediterraneum	5	35	25.9	7.633 ± 0.199	2.309 ± 0.058	1.031 ± 0.056	0.991	0.044	0.041	0.994	1	Agustí-Brisach et al. (2019)
Neofusicoccum parvum	3	41	25.8	5.318 ± 0.053	1.496 ± 0.017	2.247 ± 0.068	0.994	0.030	0.013	0.997	8	Giambra (2016), Bellée et al. (2017), Pitt et al. (2013a), Qiu et al. (2016), Úrbez-Torres et al. (2006), Amponsah (2010), Agustí-Brisach et al. (2019), Baskarathevan et al. (2012)
Neofusicoccum vitifusiforme	9	41	26.3	4.461 ± 0.134	1.183 ± 0.041	1.560 ± 0.120	0.986	0.074	−0.109	0.979	1	Giambra (2016)
Botryosphaeria dieback	2	41	25.3	5.251 ± 0.020	1.478 ± 0.006	2.338 ± 0.027	0.999	0.012	0.002	0.999	45	–
Y′	6	40	24.9	4.607 ± 0.034	1.257 ± 0.011	2.373 ± 0.054	0.996	0.022	0.001	0.998	–	–
Y″	2	41	25.5	5.410 ± 0.021	1.521 ± 0.007	1.827 ± 0.019	0.999	0.011	0.002	1.000	–	–
Phaeomoniella chlamydospora	5	40	24.3	4.597 ± 0.052	1.222 ± 0.017	2.888 ± 0.112	0.993	0.039	−0.039	0.994	7	Crous et al. (1996), Valtaud et al. (2009), Whiting et al. (2001), Díaz and Latorre (2014), Tello et al. (2010), Crous and Gams (2000), Agustí-Brisach et al. (2019)
Phaeoacremonium aleophilum (minimum)	5	40	26.6	5.552 ± 0.059	1.605 ± 0.018	1.430 ± 0.036	0.996	0.024	0.008	0.998	6	Valtaud et al. (2009), Whiting et al. (2001), Rooney-Latham (2005), Mostert et al. (2006), Crous et al. (1996), Agustí-Brisach et al. (2019)
Phaeoacremonium alvesii	10	40	27.9	5.001 ± 0.133	1.478 ± 0.039	0.716 ± 0.034	0.990	0.039	−0.001	0.995	1	Rooney-Latham (2005)
Phaeoacremonium angustius	14	31	24.1	5.289 ± 0.199	1.464 ± 0.060	1.351 ± 0.122	0.987	0.062	−0.123	0.982	1	Mostert et al. (2006)
Phaeoacremonium australiense	14	37	28.4	5.925 ± 0.271	1.688 ± 0.080	1.152 ± 0.116	0.982	0.079	−0.140	0.975	1	Mostert et al. (2006)
Phaeoacremonium austroafricanum	14	31	24.1	5.259 ± 0.187	1.469 ± 0.057	1.352 ± 0.115	0.987	0.058	−0.109	0.984	1	Mostert et al. (2006)
Phaeoacremonium canadense	14	36	28.2	6.337 ± 0.314	1.838 ± 0.092	1.069 ± 0.112	0.980	0.078	−0.136	0.975	1	Úrbez-Torres et al. (2014)
Phaeoacremonium griseorubrum	10	40	28.8	5.661 ± 0.402	1.668 ± 0.112	0.580 ± 0.070	0.963	0.091	−0.105	0.971	1	Rooney-Latham (2005)
Phaeoacremonium inflatipes	5	39	28.1	6.967 ± 0.205	2.132 ± 0.066	1.544 ± 0.114	0.972	0.061	−0.006	0.985	3	Crous et al. (1996), Whiting et al. (2001), Mostert et al. (2006)
Phaeoacremonium iranianum	10	38	28.6	6.554 ± 0.190	1.966 ± 0.060	1.631 ± 0.120	0.973	0.062	−0.005	0.985	2	Mostert et al. (2006), Agustí-Brisach et al. (2019)
Phaeoacremonium krajdenii	14	38	28.7	5.558 ± 0.226	1.571 ± 0.068	1.240 ± 0.114	0.984	0.074	−0.127	0.978	1	Mostert et al. (2006)
Phaeoacremonium mortoniae	10	36	26.4	5.977 ± 0.114	1.703 ± 0.033	1.024 ± 0.040	0.994	0.031	0.007	0.997	2	Rooney-Latham (2005), Mostert et al. (2006)
Phaeoacremonium parasiticum	10	40	29.1	5.959 ± 0.224	1.764 ± 0.063	0.626 ± 0.041	0.983	0.051	−0.001	0.992	1	Rooney-Latham (2005)
Phaeoacremonium roseum	14	36	25	3.963 ± 0.102	1.000 ± 0.032	1.750 ± 0.120	0.990	0.058	−0.100	0.986	1	Úrbez-Torres et al. (2014)
Phaeoacremonium rubrigenum	10	40	27.8	5.329 ± 0.207	1.460 ± 0.055	0.600 ± 0.040	0.983	0.053	−0.004	0.991	1	Rooney-Latham (2005)
Phaeoacremonium scolyti	10	38	26.9	5.358 ± 0.102	1.529 ± 0.032	1.461 ± 0.067	0.988	0.040	0.011	0.994	2	Rooney-Latham (2005), Mostert et al. (2006)
Phaeoacremonium venezuelense	14	41	30	5.190 ± 0.170	1.463 ± 0.053	1.521 ± 0.125	0.973	0.070	−0.049	0.980	1	Mostert et al. (2006)
Phaeoacremonium viticola	14	33	26.9	7.266 ± 0.215	2.092 ± 0.058	0.729 ± 0.040	0.992	0.035	0.007	0.996	1	Mostert et al. (2006)
Fomitiporia mediterranea	10	40	28.8	5.917 ± 0.274	1.688 ± 0.081	1.192 ± 0.123	0.953	0.092	−0.054	0.968	2	Fischer (2002), Agustí-Brisach et al. (2019)
Cadophora luteo-olivacea	3	36	21	4.576 ± 0.061	1.197 ± 0.019	2.023 ± 0.078	0.990	0.039	0.000	0.995	4	Gramaje et al. (2011), Travadon et al. (2015), Rooney-Latham (2005), Agustí-Brisach et al. (2019)
Cadophora melinii	3	35	20.7	4.699 ± 0.102	1.243 ± 0.033	2.366 ± 0.160	0.971	0.065	0.011	0.985	2	Gramaje et al. (2011), Travadon et al. (2015)
Esca complex	3	41	26.5	5.518 ± 0.030	1.618 ± 0.010	2.470 ± 0.042	0.998	0.016	0.003	0.999	42	–
Y′	10	40	25.8	4.091 ± 0.023	1.113 ± 0.008	2.578 ± 0.045	0.998	0.015	0.012	0.999	–	–
Y″	3	41	26.9	5.838 ± 0.028	1.694 ± 0.009	1.817 ± 0.023	0.999	0.013	−0.001	0.999	–	–
Cryptovalsa ampelina	9	36	23.8	4.519 ± 0.057	1.220 ± 0.019	2.107 ± 0.079	0.993	0.031	0.015	0.996	2	Giambra (2016), Agustí-Brisach et al. (2019)
Diatrypella spp.	4	31	22.5	7.212 ± 0.418	2.188 ± 0.121	0.881 ± 0.099	0.973	0.087	−0.127	0.971	1	Giambra (2016)
Eutypa lata	1	35	21.6	5.396 ± 0.080	1.546 ± 0.026	2.021 ± 0.087	0.989	0.037	0.009	0.995	4	Amborabé et al. (2005), Giambra (2016), Agustí-Brisach et al. (2019), Sosnowski et al. (2006)
Eutypa spp.	4	31	21.8	6.328 ± 0.227	1.931 ± 0.069	0.987 ± 0.072	0.981	0.058	−0.050	0.987	2	Giambra (2016)
Eutypella citricola	9	31	24.0	7.053 ± 0.402	2.126 ± 0.118	0.910 ± 0.103	0.977	0.078	−0.127	0.975	1	Giambra (2016)
Eutypella microtheca	14	36	28.1	6.004 ± 0.209	1.770 ± 0.063	1.110 ± 0.084	0.983	0.055	−0.058	0.988	1	Giambra (2016)
Eutypa dieback	1	36	23.3	5.881 ± 0.039	1.753 ± 0.013	2.394 ± 0.049	0.997	0.020	0.007	0.998	11	–
Y′	5	31	24.4	8.257 ± 0.036	2.929 ± 0.012	1.248 ± 0.012	1.000	0.006	0.004	1.000	–	–
Y″	1	36	23.3	6.112 ± 0.054	1.746 ± 0.016	1.526 ± 0.033	0.997	0.022	−0.001	0.998	–	–

^a Y′ and Y″ refer to the 95% confidence interval for the average of the indicated GTD syndrome. Optimal temperatures (T_opt) (°C) were calculated using equation 3. T_min and T_max indicate minimum and maximum temperatures (°C), respectively. RMSE = root mean square error, CRM = coefficient of residual mass, and CCC = concordance correlation coefficient. Number of cases (N) and literature sources used in the quantitative analysis leading to equation 1 are also indicated.

View as image HTML

Rescaled data were organized into a database containing the following information: (i) the GTD (i.e., BD, EC, and ED) (Mondello et al. 2018), (ii) the fungal species, (iii) the tested temperatures, (iv) the incubation length (for data concerning spore germination dynamics), (v) the corresponding level of relative fungal growth or spore germination, and (vi) additional information (literature source, country of the study, the medium used for fungal growth, and notes).

Data analysis

Rescaled data for mycelial growth were regressed against temperature. Different bell-shaped nonlinear regression equations were compared based on the Akaike’s Information Criterion (AIC); the following $\beta$ equation (Analytis 1977) provided the smallest AIC values and, therefore, was considered the most suitable (Burnham and Anderson 2002):

where Y is the rescaled mycelial growth (on a 0-to-1 scale); T_eq is an equivalent temperature, calculated as (T – T_min)/(T_max – T_min), where T is the temperature regime (°C), and T_min and T_max are minimal and maximal temperatures, respectively, for mycelial growth, which were considered as equation parameters; and a, b, and c are the equation parameters defining the top, symmetry, and size of the unimodal curve, respectively.

Rescaled data for spore germination were fit by using a three-dimensional surface model, which was previously used to describe the effect of temperature and wetness duration on infection of grape berries by Coniella diplodiella (Ji et al. 2021) as follows:

where Z is the rescaled spore germination; T_eq is an equivalent of temperature, calculated as in equation 1, in which T is the mean temperature during the moist incubation period (°C); and MD is the duration of the moist incubation period (hours). The parameters a, b, and c are as described before; d estimates the intrinsic rate of increase of the response variable with respect to MD; and e is the intrinsic rate of acceleration.

Parameters of the equations were estimated using the nonlinear regression procedure of Origin 8 Pro (OriginLab Corporation, MicroCal) for single fungal species and for all of the species of each GTD syndrome. In the latter case, to account for the variability in the temperature–moisture response of each species, equations for any species in a GTD syndrome were first solved at 1°C intervals between 0 and 40°C; thereafter, the average, standard deviation, and 95% confidence interval were calculated over the species for any temperature. Finally, equation 1 was fit to the average and its 95% confidence interval (Rossi et al. 2001).

The equations were then evaluated for goodness-of-fit based on the adjusted R², the magnitude of the standard errors of the parameters, the root mean square error (RMSE), the coefficient of residual mass (CRM), and the concordance correlation coefficient (CCC). RMSE is the measure of the average distance between the real data and the fitted line (Nash and Sutcliffe 1970). CRM represents the tendency of the model to overestimate (CRM < 0) or underestimate (CRM > 0) the measurements, with a value of 0 indicating a perfect fit of predicted and observed data (Nash and Sutcliffe 1970). CCC estimates the concordance between continuous, approximately normally distributed variables, and ranges from −1 to 1, with a value of 1 indicating a perfect fit of predicted and observed data (Lin 1989; Madden et al. 2007).

Extrapolation of relevant data

Equations 1 and 2 were used to extrapolate relevant data concerning the temperature and moisture requirements of the GTD fungi. The T_opt for mycelial growth and spore germination was calculated according to Analytis (1977) based on the estimated values of b, T_min, and T_max in equations 1 and 2, as follows:

Equation 2 was used to calculate the number of moist hours required for the germination of 50% of the spores (i.e., Z = 0.5) at any temperature, and to draw the corresponding contour line in a graph; the 95% confidence interval of the line was calculated for the three GTD syndromes as described before.

Results

Database overview

In total, 207 articles were obtained through the literature search. After selection based on a reading of the abstract and keywords and then of the full text, 22 of the 207 articles were considered useful for the quantitative analysis; 5 articles were added based on the listed references of the selected articles. Therefore, in total, 27 articles were considered. These were published between 1957 and 2022 (59% were published after 2009), and described research conducted in 12 countries (Supplementary Tables S1 and S2).

The 27 articles included 116 cases, where a “case” refers to a dataset corresponding to a specific fungal species in an experiment. Among the 116 cases, 98 documented mycelial growth on PDA (59 cases) or on other media; for example, malt extract agar and grapevine extract media (39 cases; Supplementary Table S1). The remaining 18 cases documented spore germination on PDA or water agar (Supplementary Table S2).

Mycelial growth was reported for a total of 43 fungal species, which all belong to phylum Ascomycetes, with the exception of F. mediterranea; the names of the species as indicated in the original article are listed in Supplementary Table S1. Among the 43 species, 16 were associated with BD (BD accounted for 46% of the 98 cases), 21 were associated with the EC (EC accounted for 43% of the 98 cases), and 6 were associated with ED (ED accounted for 11% of 98 cases). Most species were represented by only one case but others were represented by multiple cases; for instance, for N. parvum and D. seriata were each represented by eight cases, Phaeomoniella chlamydospora by seven cases, Phaeoacremonium minimum by six cases, and D. mutila by five cases (Supplementary Table S1). Spore germination was reported for 14 species, and this included 10 species associated with BD (two cases each for D. mutila and N. parvum, and one case each for the others), 3 species associated with EC (two cases for Phaeomoniella chlamydospora, and one case each for the others), and 1 species associated with ED (two cases for E. lata; Supplementary Table S2).

Effect of temperature on mycelial growth

All of the GTD pathogens grew at 10 to 35°C (Fig. 1); only a few species grew at <5°C, and these were E. lata (Amborabé et al. 2005) and B. dothidea, D. seriata, L. theobromae, and N. parvum (Qiu et al. 2016). In all, 8 of the 16 species belonging to family Botryosphaeriaceae (Giambra 2016; Pitt et al. 2013a; Úrbez-Torres et al. 2006) and 7 of 21 species belonging to the EC (Crous et al. 1996; Mostert et al. 2006; Whiting et al. 2001) grew at temperatures >35°C (Fig. 1). The optimal temperatures retrieved in the literature were reported either as single temperatures or as a range of temperatures, and we calculated ranges when there were multiple studies for a species (Fig. 1). The lowest T_opt was 21.5°C for E. lata (Agustí-Brisach et al. 2019), and the highest T_opt was 35.0°C for Phaeoacremonium parasiticum and P. minimum (Crous et al. 1996; Rooney-Latham 2005; Supplementary Table S3). Considering the GTD syndromes, the average of T_opts for species associated with BD, EC, or ED were 26.4°C (with 24.9 to 28.0°C as the interquartile range), 26.1°C (24.0 to 29.2°C), or 24.0 (22.2 to 25.3°C), respectively.

Download as PowerPoint

Equation 1 fit the data on mycelial growth of the 43 fungal species, with CCC values ranging from 0.968 to 0.999 (indicating high accuracy), RMSE ranging from 0.014 to 0.092 (indicating small differences between real data and the fitted line), and only a slight tendency toward underestimation (CRM from 0.001 to 0.041) or overestimation (CRM from −0.140 to > −0.001; Table 1). Equation 1 also provided a good fit of the data on the mycelial growth of the three GTD syndromes, with CCC values ≥ 0.998, RMSE ≤ 0.020, and CRM from 0.002 to 0.007 (Table 1). T_opt values were higher for BD (average of 25.3°C, with 24.9 to 25.5°C as the 95% confidence interval) and EC (26.5°C, 25.8 to 26.9°C) than for ED (23.3°C, 23.3 to 24.4°C; Fig. 2; Table 1); T_min and T_max were also lower for ED than for BD or EC. The 95% confidence intervals for the temperature response curves accounted for the variability in the within-syndrome, single-species responses (Fig. 2).

Download as PowerPoint

Effect of temperature and moist period on spore germination

Like mycelial growth, spores of all species germinated between 10 and 35°C (Fig. 3), except for the ascospores of E. lata, which germinated at temperatures <2°C but not at temperatures >32°C (Carter 1957; Ji et al. 2023). Spore germination at >35°C occurred for 8 of the 10 species causing BD (the exceptions were N. australe and N. luteum) (Amponsah 2010) and also for Phaeoacremonium minimum (Ji et al. 2023; Supplementary Table S4). Most of the species in BD had T_opts for spore germination between 22.6 and 32.4°C, with the exception of L. theobromae (33.0 to 37.0°C; Fig. 3). The three species in the EC had optima between 23.0 and 29.8°C, and E. lata had an optimum between 22.7 and 24.9°C (Fig. 3). Conidia of all of the species in BD germinated with 2 to 3 h of moisture (Amponsah 2010; Úrbez-Torres et al. 2010); germination of E. lata ascospores required at least 12 h of moisture (Carter 1957; Ji et al. 2023); and conidia of Cadophora luteo-olivacea, Phaeoacremonium minimum, and Phaeomoniella chlamydospora required at least 6, 12, and 24 h of moisture, respectively (Díaz and Latorre 2014; Ji et al. 2023; data not shown).

Download as PowerPoint

Equation 2 described the effect of temperature and MD on spore germination of the individual species, with CCC ranging between 0.888 and 0.992, RMSE between 0.039 and 0.155, and with a slight tendency toward underestimation (CRM from 0.006 to 0.160) or overestimation (CRM from −0.071 to −0.006; Table 2). T_min estimated by equation 2 ranged from 4.0 to 5.0°C, while T_max was more variable, ranging from 35.0 to 45.0°C (Table 2). Equation 2 also provided a good fit of the data on spore germination for the three GTD syndromes (Fig. 4), with CCC values ≥ 0.924, RMSE ≤ 0.122, and CRM from −0.083 to 0.025 (Table 2). Calculated T_opt values were higher for BD (27.1°C) and EC (26.5°C) than for ED (25.1°C); T_min and T_max showed similar tendencies (Table 2).

Table 2. Estimated parameters (and standard errors) and goodness-of-fit to real data for equation 2, which predicts the effect of temperature and moisture duration on spore germination for grapevine trunk disease (GTD) fungal species and syndromes^a

Species, syndromes	Estimated parameters of equation 2								Goodness-of-fit to real data				N	Literature sources
	Tmin	Tmax	Topt	a	b	c	d	e	R2	RMSE	CRM	CCC
Botryosphaeria dothidea	5	40.5	25.7	6.677 ± 1.081	1.396 ± 0.171	0.410 ± 0.071	0.320 ± 0.033	1.610 ± 0.363	0.869	0.145	−0.010	0.928	1	Úrbez-Torres et al. (2010)
Diplodia corticola	5	40	29.5	7.846 ± 3.576	2.330 ± 0.998	0.184 ± 0.128	0.256 ± 0.022	1.995 ± 0.410	0.886	0.133	0.009	0.940	1	Úrbez-Torres et al. (2010)
D. mutila	5	40	28.8	6.305 ± 0.707	2.116 ± 0.259	0.724 ± 0.183	0.275 ± 0.013	4.705 ± 1.202	0.853	0.128	−0.051	0.921	2	Amponsah (2010), Úrbez-Torres et al. (2010)
D. seriata	5	43	27.7	6.865 ± 1.381	1.487 ± 0.219	0.347 ± 0.092	0.337 ± 0.046	1.163 ± 0.304	0.838	0.149	−0.006	0.909	1	Úrbez-Torres et al. (2010)
Dothiorella iberica	4.5	35.5	22.5	5.936 ± 0.565	1.380 ± 0.116	0.799 ± 0.122	0.119 ± 0.009	2.138 ± 0.292	0.939	0.093	−0.037	0.967	1	Úrbez-Torres et al. (2010)
Lasiodiplodia theobromae	5	45	33.9	9.779 ± 1.306	2.615 ± 0.290	0.449 ± 0.089	0.248 ± 0.014	2.025 ± 0.273	0.948	0.087	−0.015	0.972	1	Úrbez-Torres et al. (2010)
Neofusicoccum australe	5	35	26.3	8.107 ± 0.346	2.452 ± 0.111	4.228 ± 0.592	0.327 ± 0.088	0.518 ± 0.137	0.988	0.039	0.087	0.992	1	Amponsah (2010)
N. luteum	5	35	26.7	8.717 ± 0.207	2.631 ± 0.063	4.647 ± 0.461	0.246 ± 0.036	0.638 ± 0.110	0.968	0.073	0.160	0.973	1	Amponsah (2010)
N. parvum	4	40.5	28.8	6.868 ± 0.614	2.116 ± 0.183	1.292 ± 0.266	0.391 ± 0.063	1.403 ± 0.612	0.792	0.155	−0.018	0.888	2	Amponsah (2010), Úrbez-Torres et al. (2010)
Spencermartinsia viticola	5	40	25	5.349 ± 0.489	1.339 ± 0.121	0.793 ± 0.148	0.216 ± 0.017	1.908 ± 0.370	0.906	0.116	0.006	0.951	1	Úrbez-Torres et al. (2010)
Botryosphaeria dieback	4.5	42	27.1	5.383 ± 0.436	1.524 ± 0.114	0.916 ± 0.130	0.277 ± 0.020	2.285 ± 0.489	0.863	0.122	−0.083	0.924	12	–
Phaeomoniella chlamydospora	5	40	29.3	8.618 ± 0.614	2.284 ± 0.150	1.142 ± 0.180	0.037 ± 0.003	5.315 ± 3.803	0.947	0.066	−0.032	0.972	2	Ji et al. (2022), Díaz and Latorre (2014)
Phaeoacremonium minimum	5	41	29.7	8.907 ± 1.361	2.176 ± 0.244	0.703 ± 0.136	0.057 ± 0.006	2.011 ± 0.377	0.873	0.136	−0.071	0.928	1	Ji et al. (2022)
Cadophora luteo-olivacea	5	40	22.2	4.396 ± 0.337	0.967 ± 0.081	1.359 ± 0.235	0.058 ± 0.006	2.029 ± 0.345	0.891	0.123	−0.036	0.941	1	Ji et al. (2022)
Esca complex	5	41	26.5	5.603 ± 0.414	1.480 ± 0.090	1.090 ± 0.134	0.041 ± 0.004	1.798 ± 0.236	0.959	0.067	0.025	0.973	4	–
Eutypa lata; Eutypa dieback	5	35	25.1	7.081 ± 0.576	2.038 ± 0.163	0.976 ± 0.177	0.056 ± 0.003	2.835 ± 0.560	0.906	0.116	−0.014	0.950	2	Ji et al. (2022), Carter (1957)

^a Optimal temperatures (T_opt) (°C) were calculated using equation 3. T_min and T_max indicate minimum and maximum temperatures (°C), respectively. RMSE = root mean square error, CRM = coefficient of residual mass, and CCC = concordance correlation coefficient. Number of cases (N) and literature sources used in the quantitative analysis leading to equation 2 are also indicated.

View as image HTML

Download as PowerPoint

Based on the extrapolation of spore germination data, 50% spore germination of BD fungi required 3.0 to 3.4 moist hours over a wide range of temperatures (20 to 35°C), and the length of the required moist period increased significantly at < 20°C and >35°C; for instance, 7.1 moist hours are required for 50% germination of BD fungi at 10°C (Fig. 5A). For EC fungi (Fig. 5B) and especially for ED fungi (Fig. 5C), the temperature ranges at which the number of moist hours required for 50% germination did not change were narrower than for BD fungi. At any temperature, the duration of the moist period required to cause infection was shorter for BD fungi than for EC or ED fungi (Fig. 5). The 95% confidence intervals of the moisture requirement curves were wider for EC fungi (Fig. 5B) than BD fungi (Fig. 5A), due to the greater variability in the single-species response of EC fungi; no confidence interval was calculated for ED fungi, because data were available only for E. lata (Fig. 5C).

Download as PowerPoint

Discussion

In the present study, we conducted a systematic literature review and a quantitative analysis of the literature data to collect, mobilize, and analyze available information on the effect of temperature on mycelial growth and on the effects of temperature and moist hours on spore germination of GTD fungi. A systematic approach to literature searches and analyses allows researchers to precisely locate and integrate available data (Mulrow 1994); a systematic approach also reduces errors, limits research bias, and improves information exchange (Candel 2014; Madden and Paul 2011; Scherm et al. 2014). Since the last century, several studies have investigated the temperature requirements of several fungi causing GTDs; in our literature search, we found >200 articles whose titles, abstracts, or keywords mentioned the effects of temperature on mycelial growth or spore germination. Only 27 of these articles, however, contained quantitative data that could be used in our analysis.

In a previous work, Songy et al. (2019) summarized cardinal temperatures for the mycelial growth of 15 GTD fungal species, based on a total of 33 cases. Our work is an improvement of the latter study, because our database covers 43 species and 98 cases; in addition, we developed equations accounting for the quantitative response of mycelial growth to temperature. In our analysis, the effect of water availability, another variable that influences fungal growth (Ma et al. 2001; Ragazzi et al. 1999; Whiting et al. 2001), was disregarded because GTD fungi live in the apoplast of the plant vascular system (Larignon and Dubos 2000; Larignon et al. 2009; Mugnai et al. 1999) and, therefore, they obtain a free water supply from the host. In addition, these fungi are known to have a low sensitivity to water potential; although they have optima ranging from −0.3 to −1.75 MPa, they show sustained growth until the water potential drops to from −3.0 to −7.9 MPa, depending on the species (Ma et al. 2001; Ragazzi et al. 1999; Whiting et al. 2001). Therefore, we inferred that temperature is the most important physical factor influencing fungal growth in the vine wood.

We also integrated the information on the effect of temperature and moisture on spore germination for the first time. We did this for 14 fungal species (based on 18 cases), and also developed corresponding equations. Our database shows, however, that the number of articles dealing with the effects of temperature on the mycelial growth and spore germination of GTD fungi are very few compared with the number of articles published on GTD. For instance, in a search on the Web of Science database for articles mentioning “grapevine trunk diseases OR GTD” up to August 2022, we retrieved 2,956 articles.

In the published literature, spore germination of GTD fungi has been considered even less often than mycelial growth (i.e., the number of cases are approximately five times higher for mycelial growth than for spore germination: 98 cases versus 18 cases). As a consequence, germination data are missing for some important GTD fungi. For example, we retrieved data on spore germination for Phaeomoniella chlamydospore, Phaeoacremonium minimum, and C. luteo-olivacea but not for other Phaeoacremonium spp., which are commonly associated with EC in different areas (Mondello et al. 2018). The failure to report the effects of environment on the germination of spores of GTD fungi has at least three explanations. First, the fungi that cause GTDs are commonly thought to readily germinate over a wide range of weather conditions (Carter 1957; Úrbez-Torres et al. 2010). Our analysis demonstrated, however, that this is not completely true, and that differences exist among fungal species for cardinal temperatures and moisture requirements. Dothiorella iberica, for instance, is estimated to have a narrower temperature range (4.5 to 35.5°C) than L. theobromae (5.0 to 45.0°C) and to have a lower T_opt than L. theobromae (22.5 versus 33.9°C). The length of the moist incubation required for 50% germination of conidia at the optimum temperature was also significantly longer for D. iberica than for L. theobromae (6.5 versus 3.2 h), even though both fungi belong to BD. Therefore, the latter species can develop over a wider range of climatic conditions and prefers hotter climates than the former; this may influence the geographic distribution of the species, their adaptation to climate change, and the frequency of their infections. A second explanation for why researchers have seldom assessed the effects of the environment on the germination of the spores of GTD fungi is that spores of some GTD fungi are difficult to obtain. For instance, ascospores of E. lata are produced in perithecia, which form only on affected branches that have died some years before (Carter 1957); in addition, basidiospores are also difficult to obtain and rarely germinate under laboratory conditions (Fischer 2002). A third explanation is that studies on spore germination require more time and are more difficult than studies on mycelial growth, so that the temperature response of mycelial growth is frequently used as an indicator of the temperature requirements of a fungus (Songy et al. 2019). This assumption, however, may be misleading in some cases. In the case of E. lata, for example, the estimated cardinal temperatures were wider for mycelial growth (1.0 to 35.0°C) than for spore germination (5.0 to 35.0°C), and the T_opt was lower for mycelial growth than for spore germination (21.6 versus 25.1°C). Because spore germination is the first step of wood colonization by GTD fungi, these discrepancies between temperature requirements for spore germination and mycelial growth may lead to incorrect estimates of disease risk. Therefore, we suggest that additional research is needed on the effects of environment on the spore germination for GTD fungi.

An increase in our understanding of the epidemiological parameters of GTD fungi and of how such parameters are affected by the environmental relationships may improve disease management at different levels. For instance, Songy et al. (2019) stated that studies on temperature requirements have clear implications for the geographic distribution of GTD fungi. Our results strongly support this proposition. Our analysis shows, for example, that L. theobromae can grow over a wide temperature range (4 to 40°C) and prefers high temperatures (T_opt = 30.5°C; Fig. 1), which helps explain why it is the prevalent GTD fungus in the warmest regions of Australia (Pitt et al. 2010), Mexico (Úrbez-Torres et al. 2008), California (Úrbez-Torres et al. 2006), and Arizona (Úrbez-Torres et al. 2007). In contrast, Dothiorella spp. (T_opt = 22.7°C) and E. lata (T_opt = 22.8°C; Fig. 1) prefer lower temperatures and have been commonly isolated from the cooler regions of New South Wales, South Australia, and New York (Carter 1957; Pearson 1980; Pitt et al. 2010; Wunderlich et al. 2009). Other species such as Diplodia seriata (T_opt = 26.2°C) and Phaeomoniella chlamydospora (T_opt = 25.2°C; Fig. 1) occur in almost all grape-growing countries, suggesting that they can grow over a wide range of climatic conditions. Because GTD fungi differ in geographic distribution, knowledge of their environmental requirements can also support climate matching analysis and risk assessment for quarantine purposes (EFSA PLH Panel et al. 2018) in such a way as to limit the introduction of certain GTD fungi via propagation material (Gramaje and Armengol 2011) into areas where they are known to be absent (Sutherst et al. 2011). Qiu et al. (2016), for instance, used the climatic prediction software CLIMEX and the optimal growth temperatures of four fungi (B. dothidea, D. seriata, L. theobromae, and N. parvum) to study their possible distribution in Australia if introduced.

In this work, we developed mathematical equations describing the effect of temperature on mycelial growth and the effects of temperature and MD on spore germination; high concordance between observed and fitted data indicated that our equations reliably represent the considered epidemiological parameters. Therefore, these equations could be used to develop a mechanistic, weather-driven model in order to estimate the risk that GTD fungi germinate on wounds and colonize the wood; such information is needed to develop tactical disease management. For modelling purposes, these equations should be combined with other epidemiological parameters (Claverie et al. 2020) such as the presence of fungal inoculum and the level of wound susceptibility.

To our knowledge, such a model has not yet been developed for GTDs. Some previous studies have focused on empirical relationships between temperature and rain and the appearance of foliar symptoms; most of these studies indicate that the occurrence of foliar symptoms increases with rainfall in spring or summer (Braccini et al. 2005; Calzarano et al. 2018; Dubos 2002; Guérin-Dubrana et al. 2013; Larignon et al. 2009; Marchi et al. 2006; Mugnai et al. 1999; Sosnowski et al. 2007; Surico et al. 2000). This was recently confirmed by a principal component analysis, which showed that, of all of the meteorological variables, effective water had the highest correlation with symptom occurrence over a 3-year period (Calvo-Garrido et al. 2021). In the latter study, temperature was also found to be inversely correlated with symptom appearance, indicating a reduction in symptom expression with higher temperatures, even though the effect of temperature cannot be separated from the availability of effective water (Calvo-Garrido et al. 2021). Expression of external symptoms, however, occurs after long and variable incubation periods (Gramaje et al. 2018) and, therefore, climatic conditions leading to symptom outbreak are not related to conditions favoring infection.

A mechanistic model has been recently developed for Phomopsis cane and leaf spot (PCLS) of grapevines caused by Diaporthe ampelina (González-Domínguez et al. 2022), the prevalent pathogen causing Phomopsis dieback (Fontaine et al. 2016a; Gramaje et al. 2018). By considering the following processes, this model was able to correctly predict PCLS epidemics: (i) the overwintering and maturation of pycnidia on affected canes, (ii) dispersal of α conidia to shoots and leaves, (iii) infection, and (iv) onset of disease symptoms. Even though the epidemiology of GTDs is more complex than that of PCLS, mechanistic models have been demonstrated to be sufficiently flexible to accommodate multiple epidemiological processes in complex systems (De Wolf and Isard 2007; Rossi et al. 2015). Liang (2022), for example, developed a general model for anthracnose diseases caused by different Colletotrichum spp. on different host plants; the general model can be adapted to each host–pathogen combination by specific parameterization of the model equations. Similarly, a general model for GTDs could be adapted for each pathogen by using species-specific calibration, as shown in this article, or syndrome-specific calibration when the species affecting the vineyard are unknown.

Acknowledgments

We thank Carlos Agustí-Brisach and Antonio Trapero (Universidad de Córdoba) for providing the original data of mycelial growth cited as Agustí-Brisach et al. (2019) in Table 1.