Misión Biológica de Galicia, Spanish Research National Council (CSIC), Apartado 28, 36080

Transcription

1 Ex-situ conservation of maize germplasm from different latitudes Pedro Revilla*, Laura Campo, José Ignacio Ruiz de Galarreta, Domingo Rios, Ángel Álvarez, Jesús Moreno-González, Amando Ordás, Rosa Ana Malvar 0 Misión Biológica de Galicia, Spanish Research National Council (CSIC), Apartado, 00 Pontevedra, Spain. Centro de Investigaciones Agrarias de Maegondo, Aegondo, A Coruña, Spain NEIKER-Instituto Vasco de Investigación y Desarrollo Agrario, Arkaute, Vitoria, Spain Centro de Conservación de la Biodiversidad Agraria de Tenerife, Tacoronte, Santa Cruz de Tenerife, Spain Estación Experimental de Aula Dei (CSIC), Montañana, Saragossa, Spain. *Corresponding author: previlla@mg.csic.es, Phone 00, Fax

2 0 Astract Germplasm collections often include open-pollinated varieties from diverse latitudes and climatic areas that cannot e correctly conserved in a single environment. The ojective of this research was to investigate if open-pollinated maize varieties from diverse latitudinal and climatic conditions could e conserved in a single location. Twelve varieties from the humid Spain, the dry Spain, and the Canary Island were multiplied in three locations from the humid Spain, one from the dry Spain, and one from the Canary Islands. Hand pollinations were made in 00-plant plots following a randomized complete lock design with two replications in five locations during two years. Differences etween origins and etween varieties were significant for most traits, including those related to growth cycle (male and female flowering and grain moisture). The varieties from each latitudinal and climatic origin did not ehave as a consistent germplasm lock; contrarily, the location variety interaction was significant for most traits. The varieties from the humid Spain maintained their ranks of growth cycle across locations and differences etween varieties were more important in the dry Spain. The growth cycle of the Canarian varieties was completely different in the dry Spain. We can maintain the varieties from any Spanish environment and latitude at any location, ut most varieties could e modified y natural selection for adaptation. Key words: Zea mays; maize; germplasm; adaptation; latitude

3 0 0 Introduction Maize (Zea mays L.) variaility was generated during its expansion from tropical America towards higher latitudes and diverse climatic conditions. The main door way for maize introduction in the Old World was Spain eginning with the second Columus trip to America (Revilla et al. ). At that time, the Canary Islands were a usual station in the transoceanic trips. As a consequence, maize coming from the Cariean area could have een directly introduced in the Canary Islands. The varieties introduced in Spain were then transferred to the rest of the Old World, although further introductions have een made through other countries (Revilla et al. 00). Only a small part of the large maize variaility availale in the Cariean area was introduced in Spain and, susequently, in the Old World. That variaility was consideraly reduced due to limited sampling and natural selection for adaptation. Nevertheless, the mild sutropical conditions of the Canary Islands do not require strong selection for adaptation of tropical germplasm and could, therefore, hold a large proportion of the original variaility introduced from Central America. If this is true, the maize variaility availale in the Canary Islands could e unique and particularly interesting for roadening the reeding ase of temperate reeding programs. Several authors have encouraged the use of tropical and sutropical germplasm as a source of favorale alleles for roadening the genetic ase of reeding germplasm y releasing semi-exotic ase reeding populations (Glover et al. 00; Goodman 00). The introduction of exotic germplasm in temperate reeding programs is eing made at large y the Latin American Maize Project (LAMP) and the Germplasm Enhancement of Maize (GEM) project (Pollak 00). Hallauer and Carena (0) recently emphasized that temperate reeding programs would strongly enefit from the introduction of genetically road ased germplasm that can contriute useful and unique alleles not present in current U.S. genomes, which is the responsiility of pulic reeding programs. Since tropical and sutropical germplasm undergo strong selection for adaptation when multiplied in temperate environments, two main strategies have een undertaken for incorporating exotic germplasm into temperate areas: direct mass selection or autochthonous allochthonous

4 0 0 crosses in intermediate locations. Direct mass selection was made in Iowa y using large samples and selecting the earliest plants (Hallauer and Carena 0) while autochthonous allochthonous crosses were made in Mexico (Whitehead et al. 00) with oth methods releasing new sources of diversity for temperate reeding programs. Presumaly, each strategy has some advantages and disadvantages; mass selection releases adapted exotic populations with large reduction of variaility while the crosses release more adapted varieties with closer genetic similitude to the temperate parents. A gradual adaptation from tropical to temperate environments going through sutropical and lower temperate latitudes would allow, presumaly, a softer natural selection for adaptation maintaining most of the exotic variaility. This is what has happened in maize introduction from Central America to Europe through the Canary Islands and southern Spain. Maize germplasm has een extensively collected in the Ierian Peninsula and the Balearic Islands since the 0s (Sánchez-Monge ; Ordás et al. ), and in the Canary Islands since the 0s; and we continue the collection of maize germplasm in the Canary Islands even today. Spain has three main climatic areas, namely the humid Peninsula, the dry Peninsula (including the Balearic Islands) and the sutropical Canary Islands, with several climatic zones in each main area. Maize reeding egun in Spain in in the Misión Biológica de Galicia (Pontevedra, northwestern Spain) and in this century other reeding programs were estalished in Madrid, Saragossa, Corunna, Vitoria and, finally, in Tenerife. Although most of those reeding programs have een almost closed nowadays, there are germplasm collections in these six locations, including the national collection in the Spanish Center for Phytogenetic Resources (CRF) in Madrid. Given the differences in latitude and climatic conditions among those locations, proaly the Spanish national collection cannot e maintained in a single location without losing variaility and changing the genetic structure of the open-pollinated varieties due to photoperiod sensitivity and natural selection for adaptation. The European Union and the USA germplasm collections have similar situations with larger latitudinal and climatic diversity. However, the possiility of preserving a

5 0 germplasm collection of different latitudes and climatic areas in a locality has not een investigated yet. Most Spanish varieties have een characterized and included in the reeding programs as ase germplasm. However, nowadays, the collection of Canarian maize includes over 00 openpollinated varieties and most of them have not een characterized. Our previous experience indicates that some of the Canarian open-pollinated varieties have an interesting potential for iomass production. Moreover, most Canarian varieties have not een included in reeding programs and no inred lines have een released from Canarian maize. Conversely, improved varieties from temperate areas have not een tested or included in reeding programs in the Canary Islands. Other authors have reported successful programs of adaptation of tropical or sutropical maize to temperate locations (Carena 0a; 0; 0c; Gouesnard et al. 00; Hallauer and Carena 0), although they did not intend the conservation of genetic resources ut the selection of exotic varieties for adaptation. Furthermore, no previous reports have assessed the adaptation of temperate germplasm to tropical or sutropical locations. The ojective of this research was to investigate if open-pollinated maize varieties from diverse latitudinal and climatic conditions could e conserved in a single location.

6 0 0 Materials and Methods We evaluated the aility to conserve germplasm from diverse latitudes and climatic areas ex-situ, i.e. to develop normally and to produce viale grain, of open-pollinated maize varieties (Tale ) in diverse climatic conditions and latitudes. Four varieties came from the humid Ierian Peninsula, four from the dry Ierian Peninsula (including the Balearic Islands) and four from the Canary Islands that are sutropical. The seed of each population had een multiplied in its original reeding station. Trials were carried out in five locations: Pontevedra (º 0 N, º W, 0 m aove the sea level), Vitoria ( N, W, m aove sea level), Corunna ( º 0 N, º, 00 m aove the sea level), on the European Atlantic coast, Saragossa ( º 0 N, 0 º 0 W, 0 m aove the sea level), and Tenerife ( ' '' N, ' '' W, m aove the sea level). The trials were laid out as a randomized complete locks with two replications in 0 and 0. One hundred grains were sown in each four-row experimental plot of 0 m at a planting density of 0,000 plants ha -. Rows were separated 0. m and plants within rows 0. m. In each experimental plot, we made as many hand-pollinations as possile y using each plant once either as male or as female. The traits measured were days to pollen and silking, plant and ear height (only in Pontevedra), plant and ear health (visual scale from = severely damaged to = healthy), pollinated ears, pollinated ears with grain, pollinated grain / m, grain moth damage (numer of grains damaged y Sitotroga cerealella / ear), grain moisture (g kg - ), 00-grain weight, and pericarp damage. Furthermore, we calculated anthesis-silking interval (ASI) as the difference etween days to pollen and days to silking, and pollination index (proportion of pollinated ears with grain). Analyses of variance were carried out using the procedure Mixed of SAS (SAS Institute Inc. 00) with years, locations, repetitions, and origin of varieties as class variales, locations and origin eing fixed effects. Susequent analyses of variance were made with years, locations, repetitions, and varieties as class variales, and locations and varieties as fixed effects. Comparisons of means

7 were made y using the LSD at P=0.0. Spearman correlations were calculated etween each trait at different locations with the procedure CORR of SAS. Finally, principal component analysis was carry out with the values of each population in each environment in order to figure out the ordination of populations ased on environments of origin and evaluation.

8 Results The comined analyses of variance for origins (humid and dry Ierian Peninsula, and the Canary Islands) showed significant differences among origins for pollen shedding, silking, grain /m, grain moisture, and 00 kernel weight. The differences were significant among locations for all traits except ASI, pollination index, plant and ear height, pericarp damage, germination, and grain moth. The location origin interaction was significant for ASI, grain /m, and ear health. The Canarian 0 varieties had longer time to oth male and female flowering and a higher grain moisture and kernel weight (Tale ). Accordingly, the northern varieties from the humid Spain had the shortest time to male and female flowering, and the lower grain moisture and grain production. On the other hand, differences etween germplasm origins were not significant for most traits, including numer of pollinated ears or any trait related to adaptation or viaility of the genotypes when grown in other environments. When the comined analyses of variance were made for the varieties instead of origins, differences among varieties were significant for all traits except pollination index, plant and ear height (ecause plant and ear height were measured only in one location), germination, ear health, and grain moth damage (data not shown). The location variety interaction was significant for all 0 traits except pollination index, plant and ear height, pericarp damage, germination, and grain moth damage. Differences among varieties were significant for all locations for male and female flowering, grain moisture and grain weight, and for some locations for other traits. Some of the varieties from the humid Spain had always the least numers of days to flowering, particularly Viana was the earliest variety in all locations, and Donostia was one of the earliest and was not significantly different from Viana in Vitoria and Tenerife, suggesting that these varieties from the humid Spain were not sensitive to the differences in photoperiod among locations (Tale ). The other varieties from the humid Spain (Mondoñedo and Tuy) had intermediate growth cycles, although Tuy was not significantly different from Viana for male flowering in Vitoria and Tenerife and for female

9 0 0 flowering in Vitoria. Mondoñedo was the variety with the most days to pollen shedding in all locations and Viana the fewest, ut the differences among the four varieties were all significant only in Saragossa. The Canarian varieties Orotava and Tacoronte had the longest times to flowering; Orotava had the largest numer of days to flowering in all locations except Saragossa; Tacoronte had the largest numer of days to flowering in three locations and was one of the latest varieties in Pontevedra and Tenerife (Tale ). The other Canarian varieties, Arucas and Oliva, had intermediate growth cycles, although Oliva was not significantly different from the variety with latest male and female flowering in Vitoria and Tenerife. Interestingly enough, the rank of Canarian varieties ased on male flowering was not consistent across locations; in Tenerife, the rank from latest to earliest was Orotava, Oliva, Tacoronte and Arucas with no significant differences etween the first and the second, the second and the third, and the third and the fourth variety. Similar ranks were oserved in the three locations of the Humid Spain, although the significant differences among populations were slightly different in these locations. However, the rank was completely different in Saragossa, where the latest variety was Tacoronte and the earliest Orotava. Another growth period-related trait is grain moisture at harvest that is normally higher for varieties with long times to flowering and vice versa (Tale ). Some of the varieties that had one of the highest grain moistures in some of the locations were not among those with longest flowering time, e.g. Castellote, Oliva and Vejer in Maegondo, Arucas in Saragossa, and Vejer in Tenerife. Conversely, Oliva, one of the varieties with largest growth cycle in Vitoria and Saragossa, was not among those with highest grain moisture. Flowering time and grain moisture completely agreed only for Viana which was the earliest in all locations from oth flowering and moisture perspectives, and for Tacorente and Orotava, which were the latest from oth perspectives in the same locations. For the other varieties, there were some inconsistencies in some locations. Other traits are not directly related to photoperiod or climatic adaptation, ut with the aility to produce seed in different environments. Among these, differences were significant among

10 0 0 varieties in all locations for 00-kernel weight, which was highest for the latest varieties such as Tacoronte, except in Tenerife, and Oliva and Orotava, except in Vitoria and Saragossa (Tale ). The smallest grains were produced y the earliest variety Viana, while the other early varieties had medium-size grains. Finally, plant and ear health did not provide significant information except in Saragossa, where Rastrojero had the healthiest plants, and the healthiest ears were produced y two varieties from dry Spain and two Canarian varieties. The similitude etween the five locations could e shown y the correlations etween these locations ased on the morphological variaility of the diverse varieties. Most correlations etween locations were positive although not all were significant (Tale ). The highest similitude was found etween the northernmost locations Maegondo and Vitoria for pollen shedding and silking and these locations had positive correlations for other traits. Maegondo had also some of the highest correlations with Saragossa for grain/m, and with Pontevedra for kernel weight. Finally, Pontevedra and Tenerife had some of the highest correlations, particularly for ASI and grain /m. The main dissimilitude was etween Saragossa and Pontevedra for male and female flowering, and Saragossa and Tenerife for ASI, as the lowest correlations for many traits were not significant. Therefore, none of the locations would guarantee the conservation of maize varieties from the other regions without natural selection, particularly for time to flowering. Principal component analysis allows a general view of the performance of populations ased on environment of origin and evaluation (Figure ). The first two principal components (PC) explain almost 0% of the variaility. The traits with largest contriution to PC were flowering times, and for PC the main contriutions were made y yield; therefore, populations flowered earlier in Tenerife and Saragossa, proaly due to heat units, and yielded more in Saragossa. Canarian populations were more adapted to Saragossa than to Pontevedra. 0

11 Discussion Population origins affected flowering and grain production and quality, and differences were significant among locations for most traits, with significant location origin interaction for few 0 0 traits. As expected, Canarian varieties had longer growth cycle while the shortest cycle was detected for the varieties from northern Spain due to the reported relationship etween latitude and growth cycle (Hallauer and Carena 0). The growth cycle of maize was larger as the latitude of the origin decreases ut we do not expect a general failure when trying to multiply maize from different latitudes of Spain (etween and North) in different climatic areas (dry and humid). The adaptation of late varieties from lower latitudes might require selection for earliness, which is a common method used y several reeders, although involves changes in genetic structure and loss of variaility (Samayoa et al. 0; Horner 0; Hallauer ). Alternatively, other authors make a screening of the exotic varieties for earliness that implies a reduction of the availale variaility (Moreno-González et al. ). The four varieties coming from each latitudinal and climatic origin did not ehave as a consistent germplasm lock ut there were different degrees of adaptation within each group. Accordingly, other authors have shown that genotypes from tropical origin have large variaility for photoperiod sensitivity when grown in temperate locations (Gouesnard et al. 00). Our results suggest that adaptation of diverse varieties should e investigated separately for each location. As Gouesnard et al. (00) reported, adaptation of exotic germplasm to temperate environments depended not only on the latitude ut also on the altitude of the environment. Therefore, we carried out individual analyses of variance for each location. The varieties from the dry Spain had always intermediate flowering cycles and were always significantly different from the earliest and the latest varieties in all locations. The ranks of the four varieties from the dry Spain were slightly different depending of the location trial, ut the differences etween varieties were not significant. The differences etween varieties due to latitude or climatic conditions were not significant proaly ecause these varieties had medium cycles and

12 0 0 also ecause the intermediate adaptation area of these varieties. Furthermore, this intermediate ehavior could e explained y the presumaly genetic origin of this germplasm from tropical latitudes (Revilla et al. ). Varieties from humid and dry Spain represent also the variaility availale in Western Europe, as maize introduced in the humid Spain is similar to that introduced in the European Atlantic coast, while the varieties of dry Spain are similar to those of the Mediterranean Europe (Revilla et al. 00). Furthermore, the heterotic pattern humid Spain dry Spain (Ordás ) is equivalent to the heterotic pattern North (humid) Europe South (dry) Europe (Revilla et al. 00a). The variaility availale among European maize populations is quite large and involves adaptation to cold or warm conditions and growth cycle (Revilla et al. 00). According to these authors, European climatic regions and latitudes are not large enough as to cause prolems for maize production in diverse European environments. However, previous studies did not include germplasm from the Canary Islands. The aility of Canarian varieties for producing seed agrees with previous finding of diversity for photoperiod adaptation among tropical genotypes, for example Gouesnard et al. (00) found that some of the Cariean varieties exhiited a good grain production although sensitive to photoperiod in temperate areas, supporting the hypothesis of Cariean introductions in southern Spain. The inconsistent ehavior on growth cycle of these varieties could e due to photoperiod sensitivity or climatic adaptation. Conversely, other locations ehave in an exceptional way concerning the flowering time of some of the varieties, particularly the Canarian location Tenerife. These results are consistent with previous reports showing that within each group of origin there was variaility for adaptation (Hallauer and Carena 0). Our results indicate that sutropical varieties can e multiplied in temperate areas and vice versa. Carena (0a,, c) also found that tropical American maize could e adapted to temperate

13 0 0 areas, although this author ased conclusions on reeding programs for adaptation and did not check adaptation of temperate maize in tropical locations. Therefore, it is possile to multiply the genetic resources from any Spanish source in almost every Spanish area, ut the morphological characters of different varieties can e inconsistent when multiplied in areas different to the original location. Such inconsistencies should e expected for sutropical germplasm as well as for northern varieties, while the genotypes coming from central areas are more likely to maintain their characteristics throughout the country. These results are consistent with previous pulications (Hallauer and Carena 0). When multiplying allochthonous varieties, we should expect changes in some adaptation-related traits, such as flowering time or grain moisture (Peña-Asín et al. 0). Steinhoff et al. (0) concluded that it is possile to use some genotypes from southern Europe for reeding in the USA, ut the election of genotypes cannot rely on data determined in Southern Europe ecause gene genetic ackground interactions and epistasis effects affect the morphological variaility of genotypes. Therefore, there are effects of environmental adaptation on oth the phenotypic and the genetic structure when genotypes are used in environments different to their original locations (Revilla et al. 00). We can maintain the varieties from any Spanish environment and latitude at any location, ut most varieties could e modified y natural selection for adaptation; nevertheless, some locations with intermediate conditions and central latitudes are more appropriate to maintain the exotic varieties without strong natural selection for adaptation. The main constrains for conservation of these populations in a single location are that they come from a relatively wide latitudinal range (from to N) and were adapted to diverse climatic conditions ased on temperature and humidity. These results indicate that the limiting factors for ex-situ conservation of exotic populations, within a moderate range of diversity, are more related to the climatic conditions of the environment of conservation than to the latitude of origin of the populations.

14 Acknowledgements Research was supported y the Spanish Plan for Research and Development (project code RF C0 and AGL0--R).

15 0 0 References Carena MJ (0a) Challenges and opportunities for developing maize cultivars in the pulic sector. Euphytica :. Carena MJ (0) Development of cold and drought tolerant short-season maize germplasm for fuel and feed utilization. Crop Breed Appl Biotechnol :. Carena MJ (0c) Developing the next generation of diverse and healthier maize cultivars tolerant to climate changes. Euphytica 0:. Glover MA, Willmot DB, Darrah LL, Hiard BE, Zhu X (00) Diallel analyses of agronomic traits using Chinese and U.S. maize germplasm. Crop Sci :0 0. Goodman MM (00) Broadening the US maize germplasm ase. Maydica 0: 0. Gouesnard B, Reourg C, Welcker C, Charcosset A (00) Analysis of photoperiod sensitivity within a collection of tropical maize populations. Gen Res Crop Evol :. Hallauer AR () Registration of BS maize germplasm. Crop Sci :. Hallauer AR, Carena MJ (0) Adaptation of tropical maize germplasm to temperate environments. Euphytica :. Horner ES (0) Registration of Maize Germplasm FSA(S), FSA(T), FSB(S), and FSB(T). Crop Sci 0:. Moreno-Gonzalez J, Ramos-Gourcy F, Losada E () Breeding potential of European flint and earliness-selected U.S. Corn Belt Dent maize populations. Crop Sci :. Ordás A () Heterosis in crosses etween American and Spanish populations of maize. Crop Sci :. Ordás A, Malvar RA, De Ron AM () Relationships among American and Spanish populations of maize. Euphytica :. Peña-Asín J, Álvarez A, Ordás A, Ordás B (0) Evaluation of three cycles of full-si reciprocal recurrent selection in two maize populations from the Northeast of Spain. Euphytica :0-0.

16 0 0 Pollak LM (00) The history and success of the pulic-private project on Germplasm Enhancement of Maize (GEM). Adv Agron :. Revilla P, Auín MC, Malvar RA, Soengas P, Ordás B, Ordás A (00) Genetic variation etween Spanish and American versions of sweet corn inred lines. Plant Breed :. Revilla P, Boyat A, Álvarez A, Gouesnard B, Soengas P, Ordás A, Malvar RA (00a) Heterotic patterns among French and Spanish maize populations. Maydica :. Revilla P, Boyat A, Álvarez A, Gouesnard B, Ordás B, Rodríguez VM, Ordás A, Malvar RA (00) Contriution of autochthonous maize populations for adaptation to European conditions. Euphytica :. Revilla P, Soengas P, Cartea ME, Malvar RA, Ordás A (00) Isozyme variaility among European maize populations and the introduction of maize in Europe. Maydica :. Revilla P, Soengas P, Malvar RA, Cartea ME, Ordás A () Isozyme variation and Historical relationships among the maize races of Spain. Maydica :. Samayoa LF, Butrón A, Revilla P, Álvarez A, Malvar RA (0) Five cycles of mass selection for earliness and ear appearance under corn orer infestation in the maize synthetic BS. Crop Sci :. Sánchez-Monge E () Razas de maíz en España. Pulicaciones del Ministerio de Agricultura. Madrid. SAS Institute. 00. SAS Version.. The SAS Institute, Cary, NC. Steinhoff J, Liu W, Maurer HP, Würschum T, Longin CFH, Ranc N, Reif JC (0) Exploitation of Elite Maize (Zea mays L.) Germplasm across Maturity Zones. Crop Sci :. Whitehead FC, Caton HG, Hallauer AR, Vasal S, Cordova H (00) Incorporation of elite sutropical and tropical maize germplasm into elite temperate germplasm. Maydica :.

17 Tale. Open-pollinated maize varieties from diverse climatic conditions and latitude evaluated in five locations for two years in Spain. Bank code Variety name Spanish area ESP0000 Tuy Humid Ierian Peninsula ESP000 Viana Humid Ierian Peninsula ESP0 Mondoñedo Humid Ierian Peninsula ESP000 Vejer Dry Ierian Peninsula a ESP000 Castellote Dry Ierian Peninsula ESP0000 Hemrilla/Queixalet Dry Ierian Peninsula ESP000 Rastrojero Dry Ierian Peninsula ESP00 Donostia Humid Ierian Peninsula ESP0 La Oliva Canary Islands ESP0 Arucas Canary Islands ESP0 Tacoronte Canary Islands ESP0 La Orotava Canary Islands a Includes the Balearic Islands.

18 Tale. Comparisons of means among three groups of open-pollinated varieties of maize from Humid and Dry Ierian Peninsula and the Canary Islands evaluated in five locations during two years Groups of origin Male Female Anthesis Numer Grain Grain 00 kernel Pericarp Germi- Plant Ear for the maize flowering flowering silking of ears weight / moisture weight (g) damage nation health health varieties (days) (days) interval / m m (g) (%) (%) (%) (-) a (-) a Canary Islands Dry Ier Pen Humid Ier Pen LSD (0.0) a Following a scale from = completely damage e diseases to = healthy

19 Tale. Comparisons of means among open-pollinated varieties of maize from humid and dry Ierian Peninsula and the Canary Islands evaluated in five locations during two years. Means are reported for those traits with significant differences among varieties for at least one location Male flowering (days) Female flowering (days) Maize varieties Maegondo Pontevedra Vitoria Saragossa Tenerife Maegondo Pontevedra Vitoria Saragossa Tenerife Arucas. c. cde 0. c.. cde c. c. c.. cd Castellote. c. c 0 c. c cd. c c 0. c. c Donostia e. g. e. h. fg. e. e. de. f. ef Hemrilla/Queixalet. c def c. c. cde 0. c c c 0.c. cd Mondoñedo d ef cd 0. f. def d 0. cd cd e. cde Oliva... a. a.. 0. a 0.. a Orotava 0. a. a. a c a 0. a 0. a. a 0. c. a Rastrojero. c. cd c. e. cd. c 0. c. d. c Tacoronte. a.. a. a. c. a. a. a Tuy e. fg. de. g. efg e. d. de. f. de Vejer. cd cd. c d cde. cd.. c. c. cd Viana. f. h. e i. g. f. f e. g. f

20 LSD (0.0) Anthesis silking interval (days) Numer of ears / m Maize varieties Maegondo Pontevedra Vitoria Saragossa Tenerife Maegondo Pontevedra Vitoria Saragossa Tenerife Arucas a c. 0. Castellote a c. 0. Donostia...00 a c. 0. Hemrilla/Queixalet a c. 0. Mondoñedo c c. 0. Oliva a c. 0. Orotava a d. 0. Rastrojero a c. 0. Tacoronte.0.. a d. 0. Tuy c a. 0. Vejer a a. 0. Viana c. 0. LSD (0.0)

21 Grain moisture (%) 00 kernel weight (g) Euphytica : (0) Maize varieties Maegondo Pontevedra Vitoria Saragossa Tenerife Maegondo Pontevedra Vitoria Saragossa Tenerife Arucas.0 cd. c. cd.0 a. c. cd. c.0 ef. a 0. d Castellote 0.0 ac. c.0 cd. c. c. de. c. d.. cd Donostia. f.0 f. e. d. c. ef. cd. cd.. cd Hemrilla/Queixalet. cd. e. -e. d. c. h. e. g. d 0. e Mondoñedo. de 0. cd. cde. d. c 0. c. c.0 a.. cd Oliva 0. a.. c. 0.. a.0 a. de.. a Orotava. a. a. a.0. a. a. a. cd.0 c. a Rastrojero. cd 0. cd. -e. d. c 0. f. d. de. c. cd Tacoronte 0. a.. a. a. c. a. a. a. a. c Tuy. ef. de. de. d. c. f.0 d. ef. c. cd Vejer.0 a.0.0 -e.. a. cd. a 0.0 c. a. a Viana. f. f. cde. d. c. g. d. fg.0 c. d LSD (0.0) Maize varieties Plant health ( ) a Ear health ( ) a

22 Arucas.0. c cde. Castellote a. Donostia cde. Hemrilla/Queixalet.0. c.... a. Mondoñedo.0. c...0. de. Oliva.. c a. Orotava.. c...0. a.0 Rastrojero.0. a...0. c. Tacoronte..0 c cd. Tuy.0.0 c Vejer.. d Viana.0. cd e. LSD (0.0). 0. a Following a scale from = completely damage e diseases to = healthy Damage was not recorded ecause it was negligile

23 Tale. Spearman correlations etween several traits of maize measured on varieties grown at different locations Pollen shedding (aove the diagonal) and silking (elow the diagonal) Maegondo Pontevedra Tenerife Vitoria Saragossa Maegondo 0.** 0.** 0.** 0.** Pontevedra 0.** 0.** 0.** 0.** Tenerife 0.** 0.** 0.** 0.** Vitoria 0.** 0.** 0.** 0.** Saragossa 0.** 0.** 0.** 0.** ASI (aove the diagonal) and grain moisture (elow the diagonal) Maegondo Pontevedra 0.0** Tenerife 0.0** 0.** Vitoria 0.** 0.** 0.* 0. Saragossa 0.** 0.** 0.** 0.** Ears /m (aove the diagonal) and grain /m (elow the diagonal) Maegondo Pontevedra 0.0* Tenerife 0.* Vitoria Saragossa 0.* 0.* kernel weight (aove the diagonal) Maegondo 0.** 0.0** 0.* 0.** Pontevedra 0.** 0. 0.** Tenerife 0.* 0.

24 Vitoria 0.* *, ** significant at P = 0.0 and 0.0, respectively

25 Figure. Principal component analysis using eleven standardized variales of populations from different Spanish areas evaluated in five environments a. a Letters represent environments (M=Maegondo, P=Pontevedra, T=Tenerife, V=Vitoria and Z=Zaragoza) and colors represent populations (green= from Humid Ierian Peninsula, red= from Canary Island, yellow= from Dry Ierian Peninsula)