Bioclimate and reproductive potential at the cold limit of European deciduous tree species

This thesis assessed factors that contribute to the explantion of the upper elevational and latitudinal limits of European broad-leaved tree species by using a set of ecological analysis, including in situ temperature measurements, a transplant experiment and a survey of the reproductive cycle and demography at the species limits.
The following key questions were answered in five chapters:
Do the elevational thermal limits of broad-leaved tree species correspond to their latitudinal thermal limits?
Can we use temperature data from weather stations to predict temperature at broad-leaved tree species limits?
Which facet of temperature at which phenological stage explain the most the elevational and latitudinal cold limits of broad-leaved tree species?
Does seed quality restrict the reproduction of species at high elevation?
Do the elevational limits of juvenile trees match with the respective adult tree limits or do the results offer an indication for recent species upward shift?
Studied species & study region
Ten broad-leaved tree species were selected for this thesis: Acer pseudoplatanus L., Fagus sylvatica L., Fraxinus excelsior L., Laburnum alpinum Mill.) Bercht. & J. Presl, Prunus avium L., Quercus petraea (Matt.) Liebl.,Sorbus aria L., Tilia platyphyllos Scop., Sorbus auccuparia L. and Ilex aquifolium L. The choice of species was driven by the aim of drawing conclusions beyond species level and beyond the regional focus. Thus, the selected species have a wide European distributional range and a wide elevational range in common. At the same time they cover a broad spectrum of plant families, namely Rosaceae, Fagaceae, Aceraceae, Oleaceae, Tiliaceae, Aquifoliaceae and Fabaceae.
The studies were carried out in three distinct regions: In the Western Swiss Alps near Martigny (46°6'N, 7°4'E), in the Eastern Swiss Alps near Chur (46°51'N, 9°32'E) and in the South of Sweden (from 57°42' to 59°39'N, ca. 12°E). Both regions in Switzerland were selected because most broad-leaved tree species studied here, find their elevational limit in these regions and reach their highest positions in Switzerland. Furthermore, the large and steep elevational gradients found in these regions facilitated surveys, specimen collections and common garden experiments. In southern Sweden we studied species at their latitudinal limits, allowing a large-scale comparison.
Comparisons of the elevational and latitudinal upper limits of broad-leaved tree species
Christophe F. Randin, Jens Paulsen, Yann Vitasse, Chris Kollas, Thomas Wohlgemuth, Niklaus E. Zimmermann and Christian Körner
We compared temperatures at the upper elevational and latitudinal limits of 18 deciduous tree species respectively in Switzerland and North of Europe, and hypothesized that species exhibit the same position (relative to the treeline) along the elevational and latitudinal gradients, which can be expected if species have reached their thermal cold limit at both high latitude and high elevation. We developed a method to identify a least-biased estimate of the elevational and latitudinal cold temperature limits and for comparing relative rank positions of species. We applied an algorithm to calculate the elevation of the potential treeline for each point in the gridded landscape of Europe and Switzerland. For each occurrence of each species, elevation was extracted from digital elevation models. The vertical distance between the elevation of the potential regional climatic treeline and the uppermost tree positions of a species was calculated and used for comparisons between elevation and latitude. We found that each species’ elevational distance to the treeline was strongly correlated to their latitudinal distance to the treeline (R2 = 0.65; P-value < 0.001).
A first group of nine species showed very similar thermal distances to the potential treelines along both, the elevational gradient and the latitudinal gradient. A second group of seven species occupied climatic niche closer to the thermal treeline at their latitudinal range edge. Only two species Sorbus aria and Quercus pubescens
appeared to have not filled their thermal niche latitudinally, because their elevational limits in the Alps were found to be at significantly cooler growing season temperatures than their latitudinal distribution limits. Our study thus, provides support for the common concept of the species range–environment equilibrium for most investigated species. Notably, we did not find stronger deviations in filling thermal niches at the latitudinal limits as compared to the elevational limits, although the former requires a species to expand over a greater geographical distance.
In situ temperatures at the cold limits of broad-leaved tree species and their deviation from records of weather stations
Chris Kollas, Christophe F. Randin, Yann Vitasse, Christian Körner
Most plant physiological processes act on micro-geographic scales within meters or less and on temporal scales of minutes or less. Yet, most studies relating species distribution to climate used typical resolutions of kilometers and months at best. Commonly available climate records from weather stations or freely available coarse-resolution geographic climatic layers thus, do not reflect the actual climatic conditions experienced by trees. In this study I selected sites where eight temperate deciduous tree species are growing at their cold upper elevational and latitudinal limits in the Swiss Alps (from 1165 m a.s.l to 1804 m a.s.l.) and in Sweden (from 58°18' N to 59°27' N). At each site, temperature was recorded for 1-2 years in different positions in the forest: at understorey height (50 cm), 2-m above ground, in the top of tree canopies and at 10 cm depth in the soil. I compared these biologically meaningful temperatures with the closest weather station data after correction for elevation. The data evidence that in mountain terrain, scaling from weather station data to on-site forest conditions requires month-specific lapse rates of temperatures, separated for means and extremes (e.g. minima). Besides best elevation-correction procedures, monthly absolute minimum temperatures of weather station data remained 1.4 +- 0.2 K (mean +- se, 12 sites) cooler than in situ conditions during the growing season (2.0 +- 0.2 K cooler during the dormant season). At the time when 2-m air temperature reached its absolute monthly minimum, temperature in the top of the tree canopy was found 0.4 +- 0.1 K colder (mean +- se, 12 sites) during the growing season and 0.9 +- 0.1 K during the dormant season.
The results add to the evidence that spatially and/or temporally coarse-resolution climatic data are not matching actual life conditions of plants. The systematic deviations of low temperature extremes from those predicted from weather stations close the gap between geographical range limits of species, their physiological limits (e.g. freezing resistance) and meteorological information. Thus, (1) the “thermal niche” concept of species range limits needs to account for such deviations of life conditions from meteorological data, should the niche boundaries have a functional meaning rooted in plant biology. And (2), improved geographic climatic layers providing resolution refinement in space and/or time may turn out to be meaningless, unless they are corrected for microclimatic differences. These proxy data for the actual bioclimate may be significantly improved by accounting for the systematic deviations reported here.
Critical temperatures determining the elevational and latitudinal range limits of European broad-leaved trees
Chris Kollas, Christophe F. Randin, Christian Körner
In this study I aimed to test three temperature-related hypotheses of European broad-leaved tree species limits in altitude and latitude. I hypothesized that: (1) annual absolute minimum temperature at the species limit does not exceed the maximum freezing tolerance of buds (2) Late spring frost events during bud-break are regularly exceeding the freezing tolerance of unfolding leaves and flowers at the species boundary (3) Minimum requirement of warm temperature during the growing season can not be met above the current range limit (incomplete tissue and seed maturation).
The study was conducted on European broad-leaved tree species elevational limits in the Swiss Alps (1165-2160 m a.s.l.) and the respective latitudinal limits in South Sweden (57°N – 59°N).
I used miniature data loggers to assess {in situ temperatures at seven broad-leaved tree species limits in the study areas. This temporal high-resolution data was then correlated with nearest weather stations to reconstruct 50 years of daily climate data and to predict weekly absolute minimum temperatures within 100 years at the species elevational and latitudinal limits.
I found that at their elevational limit most studied tree species experienced mid-winter lowest temperatures considerably warmer than their known winter frost resistance (work in progress in TREELIM work package 3, A. Lenz et al.). In contrast, the lowest temperatures that species experienced at their latitudinal limit were considerably lower than at their elevational limits in the Alps, and far below their expected winter frost resistance (LT50 gained in the Alps and Germany). Winter minima, thus, do not seem to be decisive in the Alps.
Along both gradients (elevation & latitude), I found similar growing season lengths and deleterious late spring frost events to occur during and shortly after the period of bud-break at 41 to 100 years return periods.
Thus, minimum temperatures during bud-break are most likely causing damage to broad-leaved trees and thus, along with short growing seasons, exert a significant influence on species' low temperature elevational and latitudinal limits.
Hence, process-based growth models – that depend to great extent on empirically derived thresholds for species growth and survival - will be greatly improved by incorporating these constraints during sensitive phenological stages in spring.
Further, the study showed that coupling microclimatic measurements (by data-loggers) with regional climate measurements (by weather stations) helps to overcome specific short-comings of each method (i.e. short period of employment for data loggers and inflexibility for weather stations). The combination of both devices, thus provides an excellent tool to scale up temperature measurements.
Reproduction and Recruitment limitation at the upper elevational limits of tree species in Switzerland
Chris Kollas, Yann Vitasse, Christophe F. Randin, Günter Hoch and Christian Körner (on reproduction, Chapter 5) and
Yann Vitasse, Günter Hoch, Christophe F. Randin, Armando Lenz, Chris Kollas and Christian Körner (on recruitment, Chapter 6)
The low-temperature range limit of tree species might be determined by their ability to produce and disperse viable seeds or by failure during the establishment of recruits. Biological processes such as flowering, pollen transfer, pollen tube growth, fertilization, embryogenesis, seed maturation, germination and seedlings growth into samplings are known to be affected by cold temperatures. The aim of the studies was to (1) assess the quality of seeds and (2) to assess current tree recruitment near the cold- temperature limit of 10 major European tree species in the Swiss Alps.
We first collected seeds for each species (five seed families per location and species) from stands close to optimal growth conditions (low elevation) and from marginal stands (highest elevation), replicated in two regions in the Swiss Alps in 2009. Further, we measured seed weight, seed size, storage tissue quality, seed viability and germination success.
Second, for each species, occurrences were recorded along six elevational transects (in the western and eastern Swiss Alps) according to three size classes from seedlings to adult trees in 25-m elevation steps above and below their regional upper elevational limit, in 2010. Two methods were used to compare upper elevational limits between seedlings, saplings and adults within species. First, we focused on the uppermost occurrence observed in each life stage for a given species within each studied region; and second, we predicted their upper distribution range using polynomial models fitted to presence/absence data.
We found that all species examined produced a lot of viable seeds at their current high-elevation range limit during a summer ranked ‘normal’ by long-term temperature records. Low- and high-elevation seed sources showed hardly any trait differences. The concentration of non-structural carbohydrates in seeds tended to be higher at high elevation. Additionally, in one species, Sorbus aucuparia, all measured traits showed significantly higher seed quality in high-elevation seed sources.
In addition, the observed and predicted limits of tree species occurrences (each life stage) showed that all species were represented by young individuals in the vicinity of the limit of adult trees. Moreover, tree recruitment (both seedlings and saplings) was detected and predicted significantly beyond adult tree limits in most of the species. Across regions, seedlings and saplings were on average found at elevations 73 m higher than adult trees.
Thus, under current conditions, neither seed quality nor seedling establishment constitutes a serious limitation of recruitment at the current upper elevational limits of major European trees in the Swiss Alps. The recruits found beyond the adult limits demonstrate the potential for an upward migration of trees in the Alps in response to ongoing climate warming.
Elevational adaptation and plasticity in seedling phenology of temperate deciduous tree species
Yann Vitasse, Günter Hoch, Christophe F. Randin, Armando Lenz, Chris Kollas, J. F. Scheepens, Christian Körner
Phenological events, such as the initiation and the end of seasonal growth, are thought to be under strong evolutionary control because of their influence on tree fitness. Although numerous studies highlighted genetic differentiation in phenology among populations from contrasting climates, it remains unclear whether local adaptation could restrict phenological plasticity in response to current warming. Seedling populations of seven deciduous tree species from high and low elevations in the Swiss Alps were investigated in eight common gardens located along two elevational gradients from 400 to 1700 m. We addressed the following questions: Are there genetic differentiations in phenology between populations from low and high elevations? Are populations from the upper elevational limit of a species’ distribution able to respond to increasing temperature to the same extent as low-elevation populations? Genetic variation of leaf unfolding date between seedlings from low and high populations was detected in six out of seven tree species. Except for beech, populations from high elevations tended to flush later than populations from low elevations, emphasizing that phenology is likely to be under evolutionary pressure. Furthermore, seedlings from high elevation exhibited lower phenological plasticity to temperature than low-elevation provenances. This difference in phenological plasticity may reflect the opposing selective forces involved (i.e. a trade-off between maximising growing season length and avoiding frost damages). Nevertheless, environmental effects were much stronger than genetic effects, suggesting a high phenological plasticity to enable tree populations to track ongoing climate change, which includes the risk of tracking unusually warm springs followed by frost.
Conclusions and future perspective
Species' constraints at their elevational and latitudinal limits
The possible factors that could constrain the cold, high elevation and high latitude limits of European broad-leaved tree species were assessed. Seed quality and demography indicate that reproduction and recruitment are not constraining the studied broad-leaved tree species upper elevational limits. The data show that minimum temperatures in winter are not posing any risk at these species at high elevation. In contrast, late spring frost along with short growing seasons are more likely decisive for the natural elevational and latitudinal limits of these broad-leaved tree species.
Linking eco-physiology to biogeography
The evidences that suggest that broad-leaved tree species upper limits are most likely constrained by two main temperature-based candidates (growing season length and late spring frost) underline the necessity to go further towards new assessments for testing these assumptions. An assessment via geographical information system could first provide the spatial extent where freezing temperatures (-3 +- 1°C) occurred during bud-break of dominant species such as e.g. Fagus sylvatica. Second, the same approach could be used to calculate the range deliminated by a certain minimum growing season length (e.g. 159 days for F. sylvatica, Chapter 4). These two mapping data sets could then be compared with the maps of the actual distribution of F. sylvatica . Emergent results could be: (1) the most likely temperature based constraint of species limits and (2) regions within the distribution of the species where other abiotic factors such as land-use history may override the temperature constraints.