Adaptation in Plant Breeding (1997) pp. 81-84
edited by P.M.A Tigerstedt

Adaptive properties of Picea abies progenies are influenced by environmental signals during sexual reproduction
ěystein Johnsen & Tore Skr°ppa

Division of Silviculture, Norwegian Forest Research Institute, Høgskoleveien 12, 1432 Ås, Norway.


Several independent tests have shown that climate and weather during sexual reproduction influence the adaptive properties of the Picea abies progenies. This phenomenon is expressed in seed orchards established by moving parent trees, propagated as grafts, from north to south, from high to low elevation, or from outdoor to indoor greenhouse conditions. The progenies exhibit delayed flushing in the spring, later growth cessation of leader shoots in the summer, delayed bud-set, higher frequency of lammas shoots and delayed development of frost hardiness in the autumn compared to progenies reproduced in the colder native environment. The altered performance is persistent. We have found no effect on progenies of photoperiod and temperature treatments given to the males daring meiosis and pollen production. However, when crosses were made in early spring (March), inside a heated greenhouse (short day, high temperature), the progenies are less frost hardy daring cold acclimation than progenies from identical crosses performed in late spring (May; long day high temperature) in the greenhouse. The most hardy offspring were from crosses performed under outdoor conditions in May (long day, low temperature). These results indicate that some stages in reproduction, such as female meiosis, pollen tube growth, syngamy and early embryo development, are sensitive to temperature and/or photoperiod which then alter the phenotypic performance of the offspring. The most likely explanation is the existence of a regulatory mechanism affecting the expression of genes controlling adaptive traits. If this is true, it must have implications for the genetic interpretation of provenance differences in Norway spruce.


Norway spruce (Picea abies (L.) Karst) is the most widely planted conifer tree species in Europe. Seeds for practical plantations are collected in forest stands or in seed orchards after mating between clones that have been selected for superior phenotypic performance. Within the natural range of distribution a great number of climatic races (provenances) have developed (Schmidt-Vogt, 1978). They show clinal variation along latitudinal and altitudinal gradients for traits that describe the annual growth cycle, in particular growth start and cessation and the development of autumn frost hardiness. This variation is thought to reflect climatic adaptation to photoperiod and temperature (Ekberg et al., 1979).

One of the constraints of seed production and breeding in Norway spruce is scarce flowering and poor seed ripening at northern latitudes and high altitudes. An important strategy has thus been to locate seed orchards and clone banks in warm regions. A more frequent, abundant flowering and better seed ripening under favorable climatic conditions are the main reasons for this practice (Werner, 1975; Schmidtling, 1983, 1984). In some cases the vegetatively propagated clones (grafts or cuttings) have been moved long distances from north to south or from high to low elevation, or indoor containerized seed orchards have been tried out as a supplement to outdoor seed orchards (Ross et al., 1986; Philipson, 1990; Johnsen et al., 1994a, 1994b).

The idea to produce seeds in a warm non-native environment, either by clonal transfer to an outdoor seed orchard or by use of heated greenhouses, is based on the assumption that the performance of offspring is not affected by changed environmental conditions. However, several independent studies have shown that the climate and weather during sexual reproduction influence adaptive properties of Picea abies progenies. In this paper we give an overview of the main results from these studies. The change in phenology of the reproductive process is discussed in relation to the environmental signals given by temperature and photoperiod. Possible mechanisms are briefly discussed in relation to possible sensitive stages daring reproduction

Sexual reproduction in warm non-native environments

Provenance and half-sib family comparisons

In Norway, seeds for the central and northern part of the country (lat. 63-67° N) are produced in an orchard located in the southern part at lat. 58° N, and seeds for high altitudes are produced in an orchard near the sea level. The parental clones in these orchards are phenotypically selected trees from native stands in the same area where the seedlings from the orchard seeds are to he planted. Both orchards are well isolated from outside pollen sources, and were expected to give rise to seedlings well adapted to the native areas.

When seedlings from the southern seed orchard were grown in nurseries in the northern environment, it was observed that the orchard progenies exhibited a phenology different from seedlings of autochthonous northern forest stands. With these casual observations in mind, Bjørnstad (1981) studied plants from open pollinated seed after cone collection on the selected plus trees in the native stands, and compared them with their half-sibs after controlled pollination (using pollen mix from northern males) in the southern orchard. He found that the seed orchard progenies produced terminal buds l-4 weeks later in the autumn than their northern half-sibs. Further studies showed that the seed orchard progenies were different from their northern half-sibs in many traits related to phenology. They flushed later in the spring, terminated leader growth later in the summer, they had higher frequencies of lammas shoots, lignified the annual ring later in autumn, and were more damaged in artificial freezing tests when plants developed frost hardiness an the autumn (Johnsen, 1989a, 1989b). The difference between pairs of half-sibs persisted even after a clonal propagation with rooted cuttings, and the seed orchard progenies were 15% taller than northern half-sibs after 6 years from seeds (Johnsen, 1989b). Phenological observations in progeny tests with families from the southern seed orchard show that they perform like a more southern provenance even after 17 years from seeds (Edvardsen, 1995). The hypothesis that the observed differences were an effect of phenotypic selection had to be rejected as progenies of the plus trees did not perform differently from progenies of non-selected trees from the same stands (Johnsen & Østreng, 1994).

Similar effects have been observed on progenies after crosses performed at low elevation sites between Norway spruce clones originating from high altitudes (Johnsen, 1988). Seedlings from controlled crosses had an extended shoot elongating period at ages nine and ten years compared with trees from high altitude provenances and had a lower level of autumn frost hardiness than comparable seedlings from natural stands (Skrøppa, 1994).

Results from both field trials and practical plantings indicate that the observed influences of the crossing environment may have practical consequences under extreme climatic situations (Johnsen et al., 1989; Skrøppa, in preparation). The effects may either be positive or negative for survival and vigor of the plantation, depending on how the climatic extremes are related to the annual growth rhythm of the material. The situation seems to be similar to that of a provenance transfer, it will be advantageous under certain conditions (e.g. avoidance of late sprang frost damage), but the opposite under other conditions (e.g. tolerance of early autumn frost).

Full-sib comparisons

In order to give conclusive evidence that progenies are influenced by climate and weather during sexual reproduction, more precise comparisons between progenies from the same parents produced in contrasting environments were needed. The same controlled crosses were thus performed to create identical full-sib families under different climatic conditions, both at different latitudes in Finland (Skrøppa et al., 1994), and inside a heated greenhouse vs. outside in a nearby seed orchard (Johnsen et al., 1995). The progenies that were sexually reproduced under warm conditions (southern latitudes or inside the greenhouse) were generally less frost hardy in the autumn than their full-sibs from the cold conditions. From these studies we could conclude that Norway spruce progenies were influenced both by the genetic constitution of their parents and by environment during sexual reproduction. This has also been found in Pinus sylvestris, although the effects of the parental environment seem to be less pronounced in this species (Lingren & Wang, 1986; Dormling & Johnsen, 1992; Andersson, 1994; Lindgren & Wei, 1994).

Interaction between photoperiod and temperature and timing of sexual reproduction

A common factor involved in seed production under warm non-native conditions is that the increased heat sum in spring induces an earlier start of the reproductive process at a shorter photoperiod compared with the cooler native environments far north or at high elevation. Especially a north to south transfer accentuates the differences in photoperiod, but even a strict high to low altitude transfer on the same latitude can result in as much as 2h difference in photoperiod when pollination starts in the first half of May at low elevation compared to four weeks later at high elevations. In addition, a higher temperature sum is normally attained during seed maturation late in summer and early autumn. If this interaction between temperature and photoperiod actually starts the triggering signal which then influence the progeny, a warm seed year with an early flowering could give less hardy progenies than a cool seed year with a late flowering even when seeds are produced from the same seed orchards. One study has indeed shown that seedlings from bulked seed lots of two seed orchards, from a cool (1987) versus a warm (1989) seed year, responded differently to photoperiodic treatments. Seedlings of the cool seed year formed terminal buds at shorter nights in growth chambers and formed buds earlier under natural days in nurseries than did seedlings from the warm seed year (Kohmann & Johnsen, 1994). In this connection it is interesting to note that in the study comparing full-sib pairs from the heated greenhouse and the outdoor seed orchard (Johnsen et al., 1995), the potted grafts were located inside the greenhouse for only three weeks, starting some few days before pollination until the female flowers were no longer receptive. However, due to this treatment, the reproductive process started three weeks earlier in the greenhouse that in the outdoor seed orchard. We thus suspect that an interaction between photoperiod and temperature is the major environmental trigger starting a signal which is then transduced to exert its influence on the progenies by a yet unknown mechanism (see Skrøppa & Johnsen, 1994 for a comprehensive discussion).

Sensitive stages during reproduction

To be able to understand the process we need to know the stages in the reproductive process which are sensitive to the environmental trigger. Possible selective events occurring during the reproduction could be preferential segregation or meiotic drive during formation of pollen and megaspores, megaspore degeneration, competition among 3-7 pollen grains during tube growth, or postmeiotic embryo competition among the 2-4 genetically different embryos in each developing seed (Owens & Blake, 1985). We have tested whether photoperiod and temperature treatments (long and short days/high and low temperature) during male meiosis and pollen maturation (before pollination) could give any differences in progeny performance, but in two independent experiments we found no evidence for altered autumn hardiness related to these pollen treatments (Johnsen et al., 1996 in press). Thus, the process with greatest potential for genetic selection, due to the enormous amount of pollen grains produced, does not seem to be involved as a sensitive stage. In a third experiment we made crosses inside a heated greenhouse in early spring (March) when days were short (14h), repeated the crosses inside the greenhouse in May when the days were longer (18h), and in addition made the identical crosses outside the greenhouse in May. No treatments were given to the male parents in this experiment. The progenies reproduced in early spring were less frost hardy than their full-sibs reproduced outdoors in May (Johnsen et al., 1996 in press). This result indicates that somewhere during the time interval from female meiosis, pollen tube growth, syngamy, embryo competition and to early embryogenesis, a signal is transduced, triggered by photoperiod and/or temperature, which then alter the phenotypic performance of the progenies.

A more precise timing of photoperiodic and temperature treatments to the various reproductive stages in the female cones is needed in future experiments to reveal the stages which are sensitive to environmental influence, and to find out whether the phenomenon is triggered only by temperature or photoperiod or an interaction between the two climatic factors. Sarvas (1968) has related pre- and postzygotic stages in Norway spruce to accumulated heat in day degrees above 50°C. This summer (1995) we timed heat treatments according to these data and we intend to test progenies from these treatments. In addition, ovules have been dissected and fixed throughout the experimental period for light microscopy examinations in order to determine phenological development in relation to the accumulated heat, and to detect potential effects of the heat treatments on anatomical and histological aspects in the ovules. This information will be used in future experiments for more precise timing of treatments to significant events in the reproductive process. If we find the developmental stages in question, a search for mechanisms can hopefully be initiated.

Possible mechanisms

From the data referred to above, it can be concluded that the environmental signal is transduced during sexual reproduction in the female flowers. However, the nature of this particular signal transduction (see Bowler & Chua, 1994) and how it exerts its influence at the molecular level is unknown. Our data indicate that the effect of the parental environments is persistent in the progeny. Thus, the mechanism should to our opinion operate either on a genetic (change in gene frequency caused by selection, or rapid genomic changes) or an epigenetic (gene regulation altered by imprinting) level (see Skrøppa & Johnsen, 1994). The reproductive process offers several possibilities for directed selection. However, the potential effect of gametophytic and sporophytic selection is rather limited due to the low number of pollen grains in each pollen chamber and embryos in a developing seed in spruce (Owens & Blake, 1985). Thus, unless the selection is going in the same direction in a three stage sequence (megaspore degeneration, pollen grain and embryo competition) it can hardly account for the observed effects reported.

Virtually no information is available about rapid genomic changes or epigenetic effects in spruce. Genomic imprinting is, however, being increasingly accepted as a fundamental and widespread process that determines, in ways not predicted by the laws of Mendelian inheritance, whether a particular gene will be expressed or not (Matzke & Matzke, 1993). Interestingly, Meyer et al. (1992) found that environmental factors influenced 35S promoter methylation of a maize A1-gene construct in transgenic petunia and its colored phenotype. While blossoms on field-grown plants flowering early in the season were predominantly red, later flowers on the same plants showed weaker coloration. The reduction of the A1-specific phenotype correlated with methylation of the 35S promoter. Moreover, they found that the stability of pigmentation correlated with the time of seed production. The A1-gene construct was rather insensitive to DNA methylation in progeny from flowers of young parental plants produced early in the season, but became susceptible to methylation within progeny from subsequent later crosses. Similar observations have been made in nontransgenic Zea mays where methylation was more pronounced in upper ears and tassels (Federoff & Banks, 1988). So far, we can only speculate that such gene regulation, caused by activation or deactivation of certain genes by environmental conditions during reproduction, regulate the phenotypic expression of adaptive traits in Norway spruce.

Genetic interpretation of provenance variation

Traits related to phenology of Norway spruce provenances are clinally correlated to latitude and altitude. As much as 60-93% of the variation have been explained by these parameters for traits such as bud set in the autumn, cessation of leader growth, duration of the growth period and development of autumn frost hardiness (Dormling, 1979; Skrøppa & Magnussen, 1993; DaehIen et al., 1995). Our results indicate, regardless of type mechanism involved, that phenotypic provenance variation is not only caused by natural selection among different genotypes within populations, but also directed by environmental signals received by the female parents during sexual reproduction. Both processes seem to change phenology traits of the progeny in the same direction, and may explain why Norway spruce provenances so strongly show clinal variation along latitudinal and altitudinal gradients for traits that describe the annual growth cycle.


This research have been supported by The Norwegian Ministry of Agriculture and the Nordic Forest Research Co-operation Committee.