Differentiation in phenology among and within natural populations of a South American Nothofagus revealed by a two-year evaluation in a common garden trial

Highlights

• Day of the year and growing degree days to bud burst varied among populations and years.

• Growing degree days and chilling hours have a role in the date of bud burst.

• A clear positive correlation was found for phenology of bud burst and altitude.

Abstract

Phenological traits are crucial for understanding adaptation to climate change due to their genetic control and association with abiotic factors. However, few data on phenology patterns are available for South American Nothofagus species, in particular for Nothofagus alpina, a key species of the temperate forests of Patagonia. Therefore, our aim was to analyze the variation among and within natural populations of N. alpina in two phenological traits (bud burst and foliar senescence), in growing season length and in relative growth height. We registered phenology in 65 open pollinated families of eight Argentinean natural populations installed in a common garden trial. Apical buds and foliar senescence were observed every three days in 6-year-old plants and again three years later in the same plants (N = 373). Day of the year until bud burst (DOY) and until the beginning (DOY10) and the end (DOY90) of foliar senescence were measured. Height was measured twice in a year in order to calculate the annual growth in both seasons. Growing degree days (GDD) and chilling hours (CH) until bud burst were also calculated, with two possible basal temperatures (5 °C and 7 °C) to evaluate their role in DOY. Significant differences among populations and years in DOY and growing season length were found using a linear mixed model (LMM), with the family factor explaining around 30% and 12% of the total variance respectively. The LMM for foliar senescence (DOY10 and DOY90) and the relative growth height (RGH) showed significant differences between years but not among populations. The family factor was significant for foliar senescence, although it only explained a small part of the total variance (DOY10: 4%; DOY90: 2%) and was not significant for relative growth height. A tight relationship between GDD and CH with DOY was found, and LMM showed significant differences among populations and years for both variables. The correlation between the altitude of natural populations and the mean DOY and GDD was high and positive. Our results reveal (i) the genetic control of bud burst and foliar senescence, and phenotypic plasticity of all analyzed traits, (ii) that GDD and CH are implicated in the DOY, and (iii) that altitude is probably conditioning thermal requirement of bud burst. This information suggests good perspectives to face the climate change scenario and highlight the importance of selecting appropriate populations and families for domestication and breeding of N. alpina at particular sites.

Introduction

Climate change imposes new environmental conditions to which tree species can show different biological responses (Parmesan, 2006), the most drastic being the local extinction of populations. When the “biological strategy” is the survival of the next generation, the species can persist either by adaptation (change of the genetic frequencies of the populations towards the most favorable genes for the new environment), or by migration to places where the new environment resembles that of the original places. Still, the current generation can persist in situ through phenotypic plasticity, which is the adequacy of the phenotypes to the new environment without any genetic change (Aitken et al., 2008).

Traits such as bud burst, frost resistance, growing season length and growth rate, may be variable among natural populations of a species and are key when selecting better-adapted genetic material (e.g. genotypes, populations) (Sotolongo Sospendra et al., 2010). These traits are considered as potentially adaptive, and knowledge about their variation is essential to promote species conservation, as well as for the development of domestication, breeding and restoration programs. Phenological traits are probably the most affected by climate change (Bertin, 2008). Early bud burst in temperate zones is advantageous to compete with other individuals of the same or different species for light and other resources. However, early leaf expansion also increases the risk of frost damage by late-spring freezing temperatures (Heide, 2003). On the other side, bud burst in late spring may avoid frost damage but with the cost of reducing the duration of the photosynthetically active foliage (Lechowicz, 1984, Leinonen and Hänninen, 2002). Likewise, late senescence can result in larger photosynthates storage, but can also increase the risks of incomplete nutrient remobilization due to early autumn frost affecting functional leaves (Keskitalo et al., 2005). Therefore, plant phenology is a trade-off between competence and avoidance of cold damage, and is essentially driven by air temperature and photoperiod (Menzel, 2002).

Active tissues in the buds are protected from frost damage by an endodormancy phase in winter, generally induced by photoperiod and released by chilling temperatures (Horvath et al., 2003, Campoy et al., 2012), followed by an ecodormancy phase released by heat temperatures (Lang, 1987). Within species, populations often have different chilling and heat requirements, mostly determined by adaptation to their home environment (Charrier et al., 2011, Polgar and Primack, 2011). The cessation of growth, formation of buds and subsequent senescence of leaves are adaptive responses to unfavorable conditions for photosynthesis and the occurrence of frost (Estrella et al., 2006). Therefore, trees adapted to colder climates concentrate their growth early in the season (Howe et al., 2003). Despite only a few studies have analyzed the effect of environmental factors on autumn phenology, it is generally assumed that under favorable conditions (i.e. no water or nutrition stresses) the two main environmental factors that regulate the occurrence of leaf senescence are photoperiod and temperature (Estrella et al., 2006, Koike, 1990).

Phenological traits also affect growth ability of plants (Howe et al., 2003), and have a great impact on functional processes (e.g. photosynthesis) (Tang et al., 2016). For deciduous trees, bud burst and leaf senescence determine the duration of the photosynthetically active foliage, defined as the lapse between both phenological events (Chmielewski and Rotzer, 2001, Kramer, 1995, Vitasse et al., 2009a), having a direct effect on the productivity of plants (Bennie et al., 2010, Tang et al., 2016). On the other hand, height growth (closely related to productivity) is the main morphological trait that characterizes plants ability for competition (Pinto et al., 2011).

Potential adaptive traits are usually studied through common garden trials, establishing in the same site (homogeneous environmental conditions) different genetic entities (typically populations and families), under the assumption that phenotypic differences will be mainly of genetic origin. On the other hand, the latitude, longitude and altitude of the natural populations are proxy for environmental conditions at their place of origin (e.g. temperature, humidity and solar radiation), and somehow reflect adaptive patterns of quantitative traits to local conditions (Alberto et al., 2013). While several studies have demonstrated the genetic determinism of phenological traits (e.g. Chmura and Rozkowski, 2002, in Fagus sylvatica; Premoli et al., 2007, in Nothofagus pumilio; Barbero, 2014, in Nothofagus obliqua; Torres-Ruiz et al., 2019, in Quercus petraea), less knowledge has been gained related to the phenotypic plasticity of such characters (e.g. de Villemereuil et al., 2018 in Arabis alpina; Gárate-Escamilla et al., 2019, in F. sylvatica). To test phenotypic plasticity, it is necessary to grow the same genetic pool in different environments (i.e. clonal, progeny or even provenance assays in different sites). Also, it could be achieved by measuring traits repeatedly in the same trial for different years (“temporal” plasticity; e.g. Mijnsbrugge and Janssens, 2019), especially if the years are climatically different and the traits are little influenced by the ontogenic states reached in those years.

The aim of this work was to study variation in phenological traits among natural populations of Nothofagus alpina (Poeppig & Endlicher) Oerst. (=Nothofagus nervosa (Phil.) Dim. et Mil.) growing in a common garden trial in climatically different years. Nothofagus alpina is a foundational and productive tree species of the temperate subantarctic forest from Chile and Argentina. Its very good quality wood (Tortorelli, 1956) and its breeding potential has led to develop genetic improvement programs both in Argentina and in Chile (Ipinza et al., 2000, Pastorino et al., 2016). Moreover, the species has been considered in Europe as a potential replacement of others of similar physiognomy but affected by climate change (e.g. Fagus sylvatica and Quercus spp.), although frost damage has been reported, probably caused by mistimed phenology (Mason et al., 2017). The species occurs mainly in Chile (Fig. 1), between 35° and 41° 30′ S, in both the Andes and Coastal Mountains, while in Argentina its range is almost restricted to the Lanín National Park (Sabatier et al., 2011). A Mediterranean-type climate characterizes its Argentinean range, with dry summers of great thermal amplitude and mean annual precipitation from 3000 mm close to the international boundary, up to 1200 mm near the ecotone with the Patagonian steppe, with a distance of only 50 km between both extremes (Veblen et al., 1996).

Specifically, the main objectives of this work were to analyze: (i) the variation in the date of bud burst, foliar senescence, growing season length and relative growth height (a) among natural populations and families within them and, (b) between two different growing seasons in a common garden trial. Additionally we also evaluated (ii) if growing degree days (GDD) and chilling hours (CH) have a role in bud burst date, and in case of having it, the variation of GDD and CH among populations / families and between years; (iii) the correlation between traits and geographical and environmental variables of the natural populations in order to discuss possible adaptation processes. Finally, an extra methodological objective was to evaluate the relationship between apical and lateral bud burst, to determine if lateral phenology could be a useful proxy of apical bud burst.