| Sažetak | U svim europskim zemljama, prije stavljanja nove sorte na tržište, državne institucije zahtijevaju procjenu gospodarske vrijednosti sorte u službenim sortnim pokusima (VCU pokusima). Metodologija sortnih pokusa u Republici Hrvatskoj nije se značajnije mijenjala od uspostave sustava 1991. godine. Stoga su se u ovom istraživanju analizirali podatci iz VCU pokusa: ozime pšenice, ozimog ječma, kukuruza, ozime uljane repice, suncokreta i šećerne repe, provedenih u razdoblju od 2001. godine do 2010. godine odnosno od 2001. godine do 2019. godine, u svrhu ocjene pogodnosti postojeće metodologije za procjenu vrijednosti sorata, odnosno definiranja prijedloga za moguća poboljšanja. U istraživanju su se razmatrala tri različita aspekta sortnih pokusa: odnos različitih izvora varijabilnosti obzirom na znatno manji broj pokusnih lokacija u odnosu na druge europske zemlje, povećanje učinkovitosti pokusa koje se može ostvariti uz pomoć izmjena u dizajnu i analizi pokusa, i utjecaju klimatskih čimbenika na dugoročne trendove u ostvarenim prinosima kukuruza različite skupine zriobe. Svi pokusi su postavljeni prema slučajnom bloknom rasporedu s četiri ponavljanja. Pokusi su svaku vrstu svake godine bili postavljeni na 3-5 od slijedećih sedam lokacija: Lovas, Osijek, Beli Manastir, Kutjevo, Nova Gradiška, Zagreb i Koprivnica. Veličina osnovne parcele je 10 m2 za ozimu pšenicu, ozimi ječam i ozimu uljanu repicu; 11,2 m2 za kukuruz i suncokret i 8 m2 za šećernu repu. Prinosi zrna ozime pšenice, ozimog ječma i kukuruza su preračunani na osnovi 14 % vlage u zrnu, dok je prinos zrna za ozimu uljanu repicu i suncokret preračunan na osnovi 9 % vlage. Procjena komponenti varijance za razdoblje od 2001. do 2010. godine provela se u dva koraka: u prvom su se analizirali pojedinačni pokusi, te procijenili učinci (prosjeci) sorata koji su se objedinili u setove podataka (za sve lokacije i godine), na temelju kojih su se u drugom koraku procijenile komponente varijance za svaku biljnu vrstu. Analiza podataka iz prvog koraka u procjeni komponenata varijance koristila se i kao osnova za procjenu potencijalnog dobitka učinkovitosti dizajna nepotpunih blokova primjenom postblockinga, odnosno modela prostorne analize. Utjecaj temperaturnog stresa na prinos kukuruza se procijenio korištenjem sume stresnih toplinskih jedinica (SDD) i modelom jednostavne linearne regresije procijenili su se regresijski koeficijenti prinosa zrna kukuruza (t/ha) i stresnih toplinskih jedinica (SDD) kroz sve lokacije po FAO skupinama.
Procjenom udjela i odnosa pojedinih komponenti varijance prinosa u ukupnoj fenotipskoj varijanci utvrđeno je da okolišna komponenta fenotipske varijance prinosa predstavlja dominantan dio ukupne fenotipske varijance za prinos kod svih kultura, a odnos između pojedinih komponenti u većini pokusa je bio vrlo sličan. Kod pšenice, ječma i kukuruza najveći koeficijent varijacije za prinos je opažen za interakciju lokacije i godine (od 11,79 za kukuruz FAO 400 do 18,38 za pšenicu), a kod uljane repice, suncokreta i šećerne repe najveći koeficijent varijacije za prinos je opažen za godinu. Analiza varijance pojedinačnih VCU pokusa tijekom desetogodišnjeg razdoblja (od 2001. do 2010. godine) je pokazala statistički značajan učinak genotipa za većinu lokacija za sve ispitivane usjeve. Kombinirana analiza varijance VCU pokusa po godinama tijekom desetogodišnjeg razdoblja (od 2001. do 2010. godine) je za većinu godina pokazala statistički značajne učinke genotipa, lokacije i interakcije genotipa i lokacije.
Relativna učinkovitost modela nepotpunih blokova primjenom postblockinga u odnosu na slučajni blokni raspored bila je veća kod usjeva strnih žitarica kod svih tipova podskupova u odnosu na uljarice i šećernu repu.
Regresijska analiza je uglavnom detektirala statistički značajne negativne koeficijente regresije srednjih vrijednosti prinosa i sume stresnih toplinskih jedinica. Promjene okolišnih uvjeta nisu imale različit utjecaj na genotipove kukuruza različitih skupina zriobe. |
| Sažetak (engleski) | Introduction: Assessment of the value for cultivation and use (VCU) of a new cultivar, essential for its official registration, is done through a series of trials carried out over a 2–3 year period and across many locations. In a set of multienvironment VCU trials, evaluation of new genotypes can be a laborious task due to the presence of genotype by environment interactions, which can hide their true genetic value. In an attempt to reveal the true genetic value of new cultivars, a good starting point is investigation of the importance of various genetic and environmental sources of variation, which can be done by estimating relative magnitude of corresponding variance components within the mixed model framework. Genotype × location × year (G × L × Y) data set for seven crops taken from the 10-year period 2001 – 2010 was used in the present study to estimate the variance components for main effects and their interactions in Croatian VCU trials. Depending on the crop, the most important and least important components were Y or L × Y, and L or G × L, respectively. Genotypic effect was relatively small, ranging from 2,1 to 13,4 % of the total variation.
In all European countries, before a new crop cultivar is released to the market, government authorities usually require cultivar (genotype) evaluation in official registration trials to assess its value for cultivation and use (VCU). Usually, the series of VCU trials extends over two or three years and many locations. In multi-environment VCU trials, genetic value of the new genotypes is hidden by variation caused by genotype by environment interaction effects. This can be investigated by considering variance component estimates. Based on the analysis of the relative magnitude of the variance components, it is possible to classify and select superior plant material more precisely by determining how much each variance component contributes to the total phenotypic variance. The trial setup practices vary between European countries, as they are prescribed by the respective national regulations. One of the features specified by the regulations is trial design, and there are reasons to believe that the original trial design in most countries was a complete block design, e.g. the randomized complete block design (RCBD). Trials with the most economically important crops often include a large number of candidate genotypes (sometimes even more than a hundred). Consequently, when such a large trial is carried out using RCBD, large block sizes lead to their heterogeneity, which introduces the bias and directly reduces the trial precision. In order to investigate the possibility of improving the efficiency of variety trials, a reanalysis of the RCBD experiment was performed using two different methods: postblocking and spatial analysis. Choosing maize hybrids is one of the most important decisions a maize grower makes each year. Grain yield potential along with relative maturity are the decisive factors that need to be addressed. Hybrid relative maturity, i.e. plant cycle duration is becoming more important in the context of climate change to maximize yield, whereby farmers should continuously adapt maize cycle duration and planting dates to the diversity of environmental conditions. In this study were to determine environmental effects on grain yield across three maturity groups in Croatian VCU maize trials over the last two decades, and to evaluate the use of SDD as a climatic covariate to determine the impact of climate change on grain yield in maize.
Materials and methods: Yield data from official Croatian variety trials assessing VCU of seven crops were used in this study for the period 2001 – 2010. All trials were established in randomized complete block designs with four replications and were machine-planted. Plot sizes were 10 m2 in winter wheat, winter barley and winter oil seed rape; 11,2 m2 in maize and sunflower; and 8 m2 in sugar beet. Grain yield in winter wheat, winter barley and maize were calculated on the basis of 14 % moisture, whereas grain yield in winter oil seed rape and sunflower on the basis of 9 % moisture. The data sets of seven crops were subdivided into four groups: cereals, maize, oil seed crops and sugar beet (Table 1). The number of genotypes included all entries in all trials: controls, the genotypes entering the first trial year, subsequent withdrawn genotypes by breeders, as well the genotypes (varieties) finally released. The trials of all crops were about equally distributed across the individual crop’s typical growing region of the continental, northern part of Croatia including 5-7 locations and making total of 37-49 trials per crop in the 10-year period. The data sets were both non-orthogonal and unbalanced due to ever-changing genotype sets, caused by a number of genotypes leaving or entering the trials from one year to another. Furthermore, even the set of sites was subject to some minor changes. Pre-processing of the data was done by analysing separately each of the 294 included trials to check for recording errors and outliers in order to avoid biased results. The G, L, Y, G × L, G × Y, L × Y, G × L × Y and the residual variance components were estimated using the REML method implemented in lmer function from the R package lme4 (Bates et al. 2012). Since the size of estimated variance components is related to the mean performance of individual crops, the effects were also presented as coefficients of variation, i.e. square roots of the variance component expressed as a percentage of the mean yield in order to enable comparability between crops, and with results of other authors. Data from official Croatian variety trials for seven crops conducted from 2001 to 2010 were reanalyzed to predict the efficiency that could be achieved by using incomplete rather than complete block designs. The data analysis started with fitting the baseline RCBD model to each subset. Block size for postblocking was selected following P&H’s “rule of thumb” to be a rounded value of square root of number of genotypes v. First version of postblocking was carried out using the original method of Patterson and Hunter (1983), by iteratively shifting superimposed blocks of the selected size k, and calculating EMS in each iteration by fitting the IBD model. Final estimate of the EMS was calculated as the mean of all iteration EMSs weighted by their degrees of freedom. An alternative approach to postblocking was developed following the principles for generating the alpha design for unequal block sizes (Patterson and Williams 1976). Spatial analysis model was constructed by imposing the autoregressive structure to variance-covariance matrix of the errors from model, assuming the correlation between the columns. Relative efficiency of the spatial model (RESP) was scored in the same way as for the postblocking: where SEDSP is the average standard error of the difference for pairwise genotype comparisons based on spatial model. Significance of added structure, either incomplete blocks or spatial, was tested using the likelihood ratio test (LRT). For all crops except winter oilseed rape (trials included both hybrids and pure lines), multiple comparisons with Bonferroni adjustment at p = 0.05 level were carried out to test the differences between the candidate genotypes and controls. Due to nonuniqueness of postblocking solutions, results of RCBD analysis were compared only with the results of spatial analysis, by counting the numbers of matches and mismatches in different categories (equal to, better or worse than a control). All statistical analyses were conducted within R environment (R Core Team, 2020). As the postblocking analysis had to be carried out in many iterations for each subset (for all possible layouts), an R function was created to automate the process. It utilizes the functions from specialized packages “combinat” (Chasalow, 2012), “lme4” (Bates et al., 2015) and “multcomp” (Hothorn et al., 2008). Spatial analyses required the use of commercial package “asreml” (Butler, 2020) along with the freely available companion package “asremlPlus” (Brien, 2020). The impact of heat stress on maise grain yield (FAO 300, FAO 400 and FAO 500) was estimated using stress degree days (SDD) concept. The SDD index was chosen as an environmental covariate in this study due to highest correlations with grain yield. A simple linear regression model was used for fitting the data for grain yield on SDD values across the five locations.
Results and conclusions: In the groups of cereals and maize, the greatest variance components were of the interaction location × year (L × Y), whereas in oil seed crops and sugar beet, the greatest variance components were due to year main effect (Y). On the other hand, genotype × location interaction (G × L) was the smallest among the estimated variance components in cereals, maize and winter oil seed rape, while the location main effect (L) was negligible in sunflower and sugar beet. The coefficients of variation of components were the greatest for the environmental coefficients of L × Y, Y and L as well as of the residual. Variability of Y was dominant in oilseed crops and sugar beet. The coefficient for the residual was larger than the values of G coefficients in cereals, oilseed crops and sugar beet. Genotypic variation did not differ considerably among the four crop groups, while the smallest coefficients were observed for the G × L interaction. The current results are comparable with the relative sizes of the variance components obtained in studies from four-to sixfold larger countries, indicating that the environments within Croatia, if sufficiently widely sampled, can provide as extreme cultivar responses as a geographically more dispersed set of VCU trials.
The use of postblocking revealed that significant efficiency gains were expected in at least 20 % of the cereal crop trials and less than 10 % of the non-cereal crop trials. On average, spatial analysis resulted in higher efficiency gains, obtained by capturing different patterns of field trends. The ability of incomplete block designs to increase trial efficiency increases with trial and plot size, so the subsetting strategy did have a mitigating effect on the lower efficiency of complete block designs.
Grain yield varied substantially over locations with no notable positive trend over the years. Random effects of location-year interaction showed no different patterns between maturity groups. Stress degree days (SDD) showed mostly significant coefficients of regression on location effects, except in two locations. Apparently, the effects of management might play more critical role in maize phenology and yield formation compared to climate change, at least in suboptimum growing conditions often found in Southeast Europe. To facilitate robust analysis of the crop improvement, the traditional forked approach dealing with G×E by breeders and E × M by agronomists should be integrated to G × E × M framework, to assess the full gradient of combinations forming the adaptation landscape. |
