Susceptibility of Pinus nigra and Cedrus libani to Turkish Gremmeniella abietina isolates

Gremmeniella abietina causes shoot dieback and stem cankers on conifers throughout Northern hemisphere. The aim of this study was to investigate the virulence of Turkish G. abietina isolates in a field experiment. The lower branches of 15-20-year-old P. nigra and C. libani in a plantation site at 1,050 m a.s.l. in Isparta were inoculated at 1-2-month intervals during September-January. Five isolates obtained from high altitude mountainous forests were used. Each isolate was inoculated into two branches per tree and repeated ten times on both tree species at each inoculation date. The branches were sampled at the end of February, and in August, and lesion lengths in the inner bark measured. The mean lesion length on P. nigra and C libani were 10.6 ± 0.8 and 3.8 ± 0.2 mm in February and 17.6 ± 1.4 and 7.8 ± 0.8 mm in August, respectively. Differences in the mean lesion length between the isolates were small. Nevertheless, there were significant differences between the isolates on P. nigra in November and January inoculations, and on C. libani at all four inoculation times. The mean lesion lengths for all isolates at both sampling dates was the highest (p < 0.01) in December inoculations for both P. nigra (22.0 ± 1.9 February; 32.9 ± 2.9 August) and C. libani (5.6 ± 0.7; 11.3 ± 1.2). There was no difference between the September and January inoculations on P. nigra, despite the almost six-fold difference in incubation period. During the December inoculations, the trees were most likely in winter dormancy, i.e. unable to defend themselves, which would explain the large lesions.

Palabras clave: temperatura; fecha de inoculación; Crecimiento micelial; periodo de incubación.Susceptibility against Turkish Gremmeniella abietina isolates biochemical defence but it works also in environmental conditions that do not predispose the trees for infection (Ranta et al., 2000).The susceptibility of conifers varies at different stages of shoot development and annual rhythm of the host; therefore the timing of inoculations is important (Roll-Hansen, 1964;Yokota et al., 1974;Barklund and Unestam, 1988;Aitken, 1993;Petäistö and Kurkela, 1993).While inoculation with spores is the most successful in the first half of growing season (Petäistö and Kurkela, 1993) the optimal time for inoculations with mycelium is the dormant period of the host (Patton et al., 1984).
The aim of this study was to investigate the susceptibility of P. nigra subsp.pallasiana and C. libani A. Rich.against G. abietina at different times during autumn and winter in an inoculation experiment in field conditions.

Material and Methods
Cultures of the fungus were obtained from pycnidia developed on dead branches of naturally infected Anatolian Black pine (Pinus nigra Arnold ssp.pallasiana) and Scots pine (P.sylvestris) collected from high altitude mountainous areas in the Black Sea region and the Lakes District of Turkey.Five G. abietina isolates obtained from five pycnidia were used in inoculation experiments (Table 1).The isolates were grown on oatmeal agar (OMA, Difco, Difco Laboratories) medium at 18 ± 1 °C in dark.
Field inoculations were repeated four times; at 15 th of September, 8 th of November, 27 th of December 2010 and 27 th of January 2011, in a P. nigra and C. libani plantation site near the campus area of Süleyman Demirel University, Isparta at 1,050 m a.s.l.The density of the stands was approximately 1,300 trees per hectare.The
Recently it has been reported from Lakes District of Turkey on western Taurus Mountains at about 1,700-2,100 m a.s.l. on Pinus nigra J. F. Arnold subsp.pallasiana (Lamb.)Holmboe (Lehtijärvi et al., 2010a,b).It was also observed on P. sylvestris L. at Black Sea Region above 1,800 m a.s.l. on lower branches of young trees.However, to our knowledge, the virulence of the Turkish G. abietina isolates has not been studied.
Susceptibility of conifers to G. abietina is to some degree under genetic control of the host (Roll-Hansen, 1972;Nevalainen and Uotila, 1984;Stephan et al. 1984;Dietrichson and Solheim, 1987;Aitken, 1993;Hansson 1998;Sonesson et al., 2007).However, the role of environmental factors in disease development is essential; severe epidemics occur exclusively under environmental conditions which both predispose the host and favour the spread and survival of the fungus (Donaubauer, 1972;Petäistö and Kurkela, 1993).Although such factors as topography, microclimate, stand structure and nutrient imbalances affect the development of disease locally, widespread G. abietina epidemics usually appear after cool and rainy growing seasons with low light intensities (Petäistö and Kurkela, 1993;Karlman, 2001;Petäistö and Heinonen, 2003;Thomsen, 2009).
Host susceptibility can be tested by inoculating the hosts either with spores or mycelium of the fungus.The first method is more natural as the fungus must be able to both cross the structural barriers and endure biochemical defence inside the living tissues (Patton et al., 1984).The second method tests only the degree of the oculated into two branches per tree and repeated ten times on both tree species at each inoculation date (totally: 5 isolates × 2 branches × 10 trees × 2 tree species × 4 inoculation dates = 800 branches in 80 trees).Branch diameter was measured at the inoculation point and the bark surface disinfected with 70% ethanol.Thereafter the outer bark and phloem were removed with a 4-mm diameter cork borer and inoculated with 4-mm diameter agar plugs cut from the edge of 4-6-week-old G. abietina colonies.Control shoots (2 × 10 × 2 × 4 = 160) were inoculated with sterile OMA.Wounds were tightly wrapped with parafilm.
Half of the inoculated branches (i.e.400 + 80) were sampled at the end of February 2011 and the remaining ones six months later in August, which resulted in incubation periods of variable length (Table 3).The outer bark around the inoculation point was removed with a sterile scalpel and the lesion length measured.Reisolation of G. abietina isolates was attempted from 10% of the inoculated shoots.Small pieces of tissues were cut from the edges of necrotic areas with a sterile scalpel.The pieces were plated onto 2% malt extract agar (MEA) and incubated in the dark at 15 °C for 3 weeks.The mean temperature and relative humidity, as well as degree-days above 5 °C were calculated from data obtained from the Turkish State Meteorological Service for each incubation period starting from the beginning of inoculations till the date of harvesting.
Data were analyzed using the factorial design ANOVA, in which, each host species and each harvest time was analyzed separately.Six levels of isolate factor (GaIPs, GaZPs, GaDgPn1, GaDgPn2, GaDgPn3, and the control) and 4 levels of inoculation time factor (September, November, December and January inoculations) were used.Duncan's multiple range test was used in order to determine the differences between the group means.The analyses were performed using SPSS statistical software (SPSS Inc., Chicago, IL, USA).

February harvest
The mean lesion lengths on both P. nigra and C. libani in February harvest were significantly different among inoculation dates and isolates as well as their interaction.On the other hand, in August harvest, the mean lesion lengths were significantly different among inoculation dates but not among isolates, and the interaction between isolates and inoculation dates was significant only for P. nigra (p < 0.01) (Table 2).The mean lesion lengths caused by the G. abietina isolates were larger than those measured for control inoculations on both P. nigra and C libani in February harvest (Table 3).There were also statistically significant differences in mean lesion length between the inoculation dates (p < 0.01).Both P. nigra and C. libani were found to be the most susceptible against G. abietina when inoculated in December; the mean lesion length for all isolates was significantly larger in December inoculations for both P. nigra (22.00 ± 1.9SE) and C. libani (5.58 ± 0.72) than at the other inoculation dates.the shortest lesions on P. nigra were found in both September and January inoculations, the shortest ones on C. libani were found only in January inoculations (p < 0.01) (Table 3).The length of the incubation period did not explain the differences in lesion length; there was no statistically significant difference between the September and January inoculations on P. nigra.In contrast to the January inoculations the mean lesion length in September was not different from the control.On C. libani the situation seemed to be the opposite (Table 3).
There were some differences in lesion length between the isolates (p < 0.01), but virulence of the isolates varied in the different host-inoculation date combinations.
The daily mean temperature (°C) and relative humidity (%) between 15 September and 28 February ranged from -3.1 to 20.9 °C and 39.5 to 99.0%, respectively.The daily mean temperatures were above 10 °C until the first week of December.The daily mean temperatures were fluctuating mainly between 0 and 5 °C, or below 0 °C after December and January inoculations (Figure 1).Degree-days (d.d.) above 5 °C decreased rapidly from 741.1 for September inoculations to 9.9 for January inoculations (Table 4).

August harvest
In August harvest the lesion lengths were larger both in G. abietina inoculated branches and control than in February harvest.On P. nigra the lesions sizes in G. abietina inoculated branches were 112, 36, 49 and 150% larger in August harvest in September, November, December and January inoculations, respectively.On C. libani, in turn, the lesion sizes were 30, 79, 103 and 263% larger, respectively.However, also the lesion lengths in controls were larger and the difference in relation to those in the G. abietina inoculated branches smaller than in February.As a result, the differences in lesion lengths between the G. abietina inoculated branches and controls were not significantly different with two exceptions.Results on C. libani were similar: the differences observed in February had disappeared.
On both hosts the lesion lengths were 2-3-fold larger in control inoculations done in December and January than those done in September and November.However, the differences were not statistically significant.

Discussion
Although both P. nigra and C. libani were found to be susceptible in the February sampling, the symptom severity on black pine was almost 3-fold higher than that on cedar.To our knowledge G. abietina has not been reported to occur on C. libani.Nevertheless, this species was reported to be susceptible against Heterobasidion annosum sensu lato (Fr.)Bref.(Lehtijärvi et al., 2011) and Sphaeropsis sapinea (Fr.)Dyko et Sutton (Doğmuş- Lehtijärvi et al., 2009) in mycelial inoculations -without being their natural host.
The length of the incubation period did not explain the differences in lesion length in the February sampling; there was no statistically significant difference between the September and January inoculations on P. nigra, despite the almost 6-fold difference in incubation period.Although, the difference was larger in the August sampling neither September nor January inoculations differed from the controls.
The temperatures during the first two inoculation dates were within the adequate intervals for plant growth, and therefore the defense mechanisms probably were still active.This could explain the increase in susceptibility of P. nigra and C. libani from September to December inoculation.During December inoculations and onwards, in contrast, the trees were most likely in dormancy, i.e. unable to defend themselves.It is known that both virulence of G. abietina and host resistance are dependent on weather conditions (Roll-Hansen, 1964;Blenis et al., 1984;Barklund and Unestam, 1988;Marosy et al., 1989;Karlman et al., 1994, Terho andUotila, 1999).Although spore dispersal and infection of the host occur during the growing season, G. abietina is able to grow into the living host tissues firstly during host dormancy (Patton et al., 1984).In an inoculation experiment using conidiospores, Marosy et al. (1989) found that a period of at least 44 conducive days, i.e. days with temperatures remaining between -6 and +5 °C, was necessary for symptom development.In the present study, the inoculation of the fungal mycelium into wounds reaching to the phloem bypassed any natural barriers present in the bark.However, once inside the living tissues, growth of the fungus in the inner bark can be expected to be controlled by the same defence mechanisms irrespective of how the fungus entered the host.Therefore, the conducive day concept should apply to both penetration of and growth within the living tissues.Several mycelial inoculation experiments support this view.Roll-Hansen (1964) found out that mycelial inoculations on the stems of Scots pine in late winter gave larger necrosis than those done in early spring when the soil had thawed indicating that the cold weather promoted mycelial colonization of the bark tissues.Terho and Uotila (1999) found an increasing trend from August to October in canker and necrosis lengths on 2-m-tall Scots pines inoculated with mycelium with two weeks intervals.Although the daily mean temperatures during the inoculation period in their study (August-October) were similar to those in November-January in the present study, the pattern in susceptibility of the trees was different.In their study, the longest cankers and lesions were formed on trees which were inoculated the latest, in October, while in the present study the lesion length peaked in December inoculations.The result was the same in both February and August sampling indicating that the reduction in lesion length from December to January inoculations was not because of the short incubation period from January to February in the first sampling.The short lesion lengths in January inoculations indicate a lower susceptibility of the host compared with December inoculations despite the continued low temperatures and probably dormancy as well.By the February sampling the number of conducive days for the December inoculations was 42 (estimated from mean daily temperatures below +5 °C) and for January inoculations 21.After February there was only eight more conducive days.As the lesion lengths for January inoculations in August were not significantly different from controls it seems likely that only the December inoculations were performed within the susceptibility period predicted by the conducive day concept (Marosy et al., 1989).Alternatively, the 3.8fold difference in mean lesion lengths (22 vs. 5.8 mm) could be explained by a 2.6-fold difference in degreedays above 5 °C (26 vs. 9.9).Application of a threshold temperature of 5.9 °C would result in a ratio 3.86, which is almost identical with that of the lesion lengths.If the growth rate of the fungal mycelium within the host tissues during the host dormancy were temperature limited, the colder weather from January inoculations onwards would explain the shorter lesions.However, owing to the fact that G. abietina is able to grow at temperatures down to about -6 C (Marosy et al., 1989) and that the lesion lengths for January inoculations in August sampling did not differ from control, this alternative seems less likely.
The larger lesion lengths in the controls in the August sampling indicate that active defence reactions occurred also because of wounding alone after February sampling when the weather had become warmer.Wounding during the dormancy seemed to be more damaging as indicated by the longer lesions on both tree species in the December and January controls, although the difference was not statistically significant.Ceased fungal growth during the warm weather in summer could explain the smaller relative differences between the lesions on G. abietina inoculated branches and the controls (cf.Patton et al., 1984;Marosy et al., 1989).
The results of the current study clearly showed that the most sensitive period for both host species against G. abietina colonization was December.This was most likely a result of host dormancy and appropriate temperatures for fungal growth within the host tissues.More evenly distributed inoculation intervals and shorter incubation periods could have given more accurate results about when the susceptible period could have started and ended.Although both P. nigra and C. libani were found to be susceptible, the symptom severity on black pine was almost 3-fold higher than that on cedar.Moreover, the lesions were relatively short on C. libani, possibly because it may not be a natural host of G. abietina.The isolates used in the present study were obtained from P. sylvestris and P. nigra, and therefore could be expected to be adapted to infect pines, but not C. libani.

Table 2 .
Two-way ANOVA table for the lesion length for both host species and harvest times

Table 3 .
Mean lesion lengths (mm) on branches inoculated with five Gremmeniella abietina isolates ABAverages in the same column with the same lowercase letter are not significantly different (p > 0.01) according to Duncan's test.Averages in the same row with the same uppercase letter are not significantly different (p > 0.01) according to Duncan's test.* Average lesion length in branches inoculated with G. abietina only (excluding the control inoculations).Susceptibility against Turkish Gremmeniella abietina isolates

Table 4 .
Effective temperature sums (degree-days), mean temperatures and relative humidity during incubations until first sampling in February