EFFECT OF SEED AGING ON THE SEED QUALITY AND SEEDLING GROWTH OF TIMOTHY GRASS (Phleum pratense L.) UTICAJ STAROSTI SEMENA NA KLIJAVOST I PORAST KLIJANACA MAČJEG REPA (Phleum pratense L.)

UTICAJ STAROSTI SEMENA NA KLIJAVOST I PORAST KLIJANACA MAČJEG REPA (Phleum pratense L.) Rade STANISAVLJEVIĆ, Dobrivoj POŠTIĆ, Jasmina MILENKOVIĆ, Dragoslav ĐOKIĆ, Dragoljub BEKOVIĆ, Ratibor ŠTRBANOVIĆ, Snežana JOVANOVIĆ, Marijenka TABAKOVIĆ Institute for plant protection and the environment, Teodora Drajzera 9, 11000 Belgrade, Serbia Institute for forage crops, Kruševac, Globoder 37251, Serbia University of Priština, Faculty of Agriculture, 38219 Lešak bb, Serbia Maize research institute Zemun Polje, Slobodana Bajića 1, 11000 Belgrade, Serbia *stanisavljevicrade@gmail.com


INTRODUCTION
Timothy grass is a fodder grass mostly grown as a pure stand or in mixed stands with fodder legumes at higher elevations.It is widespread in cool regions of Europe, Asia, North Africa and North America.In Serbia, Timothy grass is widely cultivated in mountainous areas, favoring damp valleys and mountain meadows of the Arrhenatheretea type.On balance, its cultivation under appropriate climatic conditions results in high yields and high-quality fodder (Knezević and Stanisavljević, 2018).Unlike most fodder grasses, Timothy grass is not susceptible to seed shattering, so Timothy grass seeds can be harvested with low moisture contents, which is of paramount importance to the preservation of seed quality.The methods for preserving the seed quality of fodder grasses depend on the following parameters: species, variety, harvest methods, postharvest treatments, as well as the storage temperature and air humidity (Stanisavljević et al., 2016(Stanisavljević et al., , 2017)).Poor seed germination is often caused by the presence of dormant seeds which fail to germinate despite ideal conditions in the field or laboratory (Adkins et al., 2002).After a certain storage period, fodder and ornamental grass seeds undergo complex physiological and biochemical processes, during which dormant seeds start germinating, i.e. achieve the maximum germination and agricultural use-value.
In many intensively cultivated crops, dormant seeds have been minimized or virtually non-existent (Adkins et al., 2002).However, fodder grasses exhibit no significant difference in the share of dormant seeds between cultivated and wild populations within the same species (Stanisavljević et al., 2012).Under natural conditions, dormant seeds delay germination, awaiting favorable conditions for the survival of the seedlings (Bewley, 1997).However, prolonged storage periods result in seed aging which is associated with reductions in seed germination, seedling growth and the total mass of the root system.The purpose of this paper is to examine the seed quality features of four Timothy grass seed lots (the 'Foka' cultivar), at ages of 3, 15 and 27 months, according to the share of dormant, germinated and dead seeds, as well as abnormal seedlings.Upon assessing seed germination, the stem length (cm), radicle length (cm) and fresh weight (g) of Timothy grass seedlings were measured as well.

MATERIAL AND METHOD
The seed quality features of four Timothy grass seed lots (the 'Foka' cultivar), at ages of 3, 15 and 27 months, were examined in this study according to the share of dormant, germinated and dead seeds, as well as abnormal seedlings.After harvest, the seeds were dried to moisture contents of 10 to 12 %, and placed in two-layer paper bags.The seeds of the following ages were analyzed: three months (AGS1 -sown and harvested in the same year), fifteen months (AGS2 -sown and harvested with one year apart) and twenty seven months (AGS3 -sown and harvested with two years apart).The following parameters of germinated seeds were measured using the method described by Stanisavljević et al. (2011): stem length (cm), radicle length (cm), and fresh seedling weight (g).The seed quality assessment was performed in accordance with the official seed quality rules (Official Gazette of SFRJ, 47/87) and the ISTA Rules (ISTA 2002(ISTA -2018)).The analysis of variance (ANOVA, F test) was applied for examining the seed lot and age parameters.Treatment differences were determined using the Tukey's multiple range tests.The variability for each seed quality feature was expressed by the coefficient of variation (CV, %), whereas the dependence between the features examined was expressed by the simple correlation coefficient (r).The percentages of germination and dormancy obtained were arcsine transformed (√x / 100) before subjected to the analysis of variance (Snedecor and Cochran, 1980).The Minitab 16.1.0program was used for data processing.

RESULTS AND DICUSSION
On the basis of the analysis of variance (F test) performed, the SL seed lot and the SL x AGS interaction were found to exert no significant effects on the seed quality indicators (Tab.1).

Table 1. Analysis of variance for dormant seeds (DS), germinated seeds (GS), dead seeds (MS), abnormal seedlings (AS), stem length (S), radicle length (R) fresh seedling weight (FS). Sources of variation: seed age (AGS), seed lot (SL).
Source According to Baskin and Baskin (2004), the dormancy of seeds is classified as primary and secondary: the primary seed dormancy is inherent only in the harvested seed and develops during seeding, whereas the secondary seed dormancy is the consequence of the environmental influence.The authors describe the possibility of having adequate and secondary dormancy in the same seed.A number of studies on fodder grass have argued that the appearance of seed dormancy is pronounced (Adkins et al., 2002), as confirmed by Stanisavljević et al. (2016Stanisavljević et al. ( , 2017)).According to Hill and Watkin (1975), a total of 30 to 40 % of dormant seeds were found in Timothy grass immediately after harvest.The same authors associate the presence of dormant seeds with seed moisture contents during harvest.However, Havstad and Aamlid, (2013) indicate that Timothy grass seeds can be dormant for several months, regardless of seed moisture contents during harvest.Upon comparing a number of fodder grasses, Stanisavljević et al. (2011) suggest that dormant seeds were present in Timothy grass to a lesser extent than in orchard grass and Italian rye-grass.The dormancy of seeds is undesirable in fodder and ornamental grasses because it impedes fast and uniform germination, especially if the grass is mixed with fodder legumes.Three months after harvest, a total of 8 % of dormant seeds were present in the AGS1 seed lot.By the next fall, the seeding percent of the AGS2 dormant seeds was 7 % lower than that recorded in the AGS1 seed lot.In the AGS1 seed lot, a germination rate of 90 % was recorded, which is 3 % higher than that recorded in the AGS2 seed lot.A statistically significant difference (P > 0.05) in germination rates was not found between AGS1 and AGS2.This suggests a high utilization value of Timothy grass at seed ages ranging from three to fifteen months.
A significant (P ≤ 0.05) decrease in germination (79%) was recorded in the AGS3 seed lot, as well as increases in dead seeds (13%) and abnormal seedlings (8%), compared to both AGS1 and AGS2.This indicates that the aging process of the seed was in progress, adversely affecting the seed use-value.According to the official seed quality rules in Serbia (Official Gazette SFRJ 47/87), seeds meet the criteria of germination even after 27 months of storage, which is the final acceptable age limit for germinating seeds.Timothy grass is often grown in mixed stands with fodder legumes, characterized by the complementary and rapid initial growth of Timothy grass and fodder legume seedlings which allows for a more successful competition with weeds (Vučković 2004).In the present study, the differences between the seedling quality parameters of AGS1 and AGS2 were found to be statistically non-significant.However, the differences between the seedling quality parameters of AGS1 and AGS2, on one hand, and the seedling quality parameters of AGS3, on the other, were found to be statistically significant (P <0.05) (Table 3).
The coefficients of correlation (r) obtained emphasize the dependence between the seed and seedling quality parameters at three seed ages and four seed lots under consideration (Table 4).A negative correlation was found between the germinated and dormant seeds tested (r = -999, P <0.001), as well as between the dead seed and the radicle length of seedlings (r = -999, P <0.001).Furthermore, a negative correlation was also found between the dead seed, the stem length and the fresh seedling weight (by r = -996, P <0.01) for seeds at three months of age (AGS1).Conversely, a positive correlation was established between the germination rate and the stem length (r = -991, P <0.01), as well as between the stem length and the radicle length (r = -995, P <0.01).A positive correlation was also found between the germination rate and the radicle length, as well as the germination rate and the fresh seedling weight (P <0.05) for seeds at all ages under consideration (AGS1, AGS2, AGS3; Table 4).A similar dependence between the seed quality parameters and the seedling growth of fodder grasses has been reported in the literature.In the instance of English ryegrass, the dependence between the germination rate and the stem length was r = 0.989, between the germination rate and the radicle length was r = 0.988, and between the stem length and the radicle length was r = 0.978 (Bukvić et al 2018).In the case of Italian ryegrass, the dependence between the germination rate and the stem length was r = 0.604 to r = 0.806, depending on the variety (Velijević et al., 2018).In the instance of tall fescue, the dependence between the germination rate and the stem length was r = 0.389, between the germination rate and the root length was r = 0.395, and the germination rate and the fresh seedling weight r = 0.305.In the case of red fescue, the dependence between the germination rate and the stem length was r = 0.245, between the germination rate and the radicle length was r = 0.235, between the germination rate and the fresh seedling weight was r = 0.288, between the germination rate and the stem length was r = 0.299, between the germination rate and the radicle length was r = 0.326, as well as between the germination rate and the fresh seedling weight was r = 0.313 (Stanisavljević et al. 2014).

CONCLUSION
The Timothy grass seeds examined at three and fifteen months of age exhibited the highest germination rates, i.e. 90 % and 87 % respectively.As a result of seed aging, the germination rate amounted to 79 % over the period of twenty-seven month, which is close to a minimum seed germination rate of 75 % according to the official seed quality policy in Serbia.The seeds at all three ages under consideration were found to exhibit high germination rates accompanied by strong seedlings, as indicated by the correlation coefficients obtained between the germination rates and the stem lengths.

Table 2 .
Effect of the seed age on the share of dormant seeds (DS), germinated seeds (GS), dead seeds (MS) and abnormal seedlings (AS) Tukey's multiple range test) for the column values.Values are the mean ± the standard error of the mean.

Table 3
a, b, …x (different lower-case letters) significant effect (P < 0.05; Tukey's multiple range test) for the column values.Values are the mean ± the standard error of the mean.

Table 4 .
Simple correlation coefficients (r)between germination (G), dormant seeds (DS), germinated seeds (GS), dead seeds (MS), abnormal seedlings (AS) and seedling quality parameters (namely stem length (S), radicle length (R), and fresh seedling weight (FS)) after three, fifteen and twenty seven months of storage (AGS1, AGS2, AGS, respectively) (n=4) ns Not significant F tests at the P > 0.05, *Significant F tests at the P < 0.05 level of significance, **Significant F tests at the P < 0.01 level of significance, ***Significant F tests at the P < 0.001 level of significance