High risk, high gain? Trade-offs between growth and resistance to extreme events differ in northern red oak (Quercus rubra L.)

Abstract

Information about the resistance and adaptive potential of tree species and provenances is needed to select suitable planting material in times of rapidly changing climate conditions. In this study, we evaluate growth responses to climatic fluctuations and extreme events for 12 provenances of northern red oak (Quercus rubra L.) that were tested across three trial sites with distinct environmental conditions in Germany. Six provenances each were sourced from the natural distribution in North America and from introduced stands in Germany. We collected increment cores of 16 trees per provenance and site. Dendroecological methods were used to compare provenance performance and establish climate-growth relationships to identify the main growth limiting factors. To evaluate the provenance response to extreme drought and frost events, three site-specific drought years were selected according to the Standardized Precipitation Evapotranspiration Index (SPEI) and 2010 as a year with an extreme late frost event. Resistance indices for these years were calculated and assessed in relation to overall growth performance. We observed a high variation in growth and in the climate sensitivity between sites depending on the prevailing climatic conditions, as well as a high intra-specific variation. Overall, summer drought and low temperatures in the early growing season appear to constrain the growth of red oak. The resistance of provenances within sites and extreme years showed considerable rank changes and interaction effects. We did not find a trade-off between growth and resistance to late frost, namely, fast growing provenances had a high frost hardiness. Further, there was no evidence for a trade-off between growth and drought hardiness. Still, responses to drought or late frost differ between provenances, pointing to dissimilar adaptive strategies. Provenances from introduced (i.e. German) stands represent suitable seed sources, as they combine a higher growth and frost hardiness compared to their North American counterparts. Drought hardiness was slightly higher in the slow-growing provenances. The results provide a better understanding of the variable adaptive strategies between provenances and help to select suitable planting material for adaptive forest management.

Keywords: climate-growth relationships; dendroecology; drought hardiness; frost hardiness; introduced species; provenance trial; tree rings.

Figure 1

Distribution of provenances (circles) in North America (A) and in Germany (B) that were planted on the three sites of the provenance trial (triangles). The natural distribution of red oak (Little, 1971) is shown in grey. Provenances and sites are colored according to the summer heat:moisture index (SHM) at their origin to highlight the different moisture conditions during the growing season. Provenances names from the USA, Canada and Germany are colored black, red and blue, respectively; this color scheme is used consistently hereafter when referring to the specific origin of provenances.

Figure 2

Heatmap with bootstrapped correlation coefficients between provenance chronologies (y-axis) and monthly climate data (temperature, precipitation and SPEI) from previous-year June (jun) to current-year September (SEP). Sites are arranged according to the precipitation gradient from wet (Waechtersbach) to dry (Waldsieversdorf). Correlation coefficients range from negative (violet) to positive (green) values and asterisks indicate significant correlations.

Figure 3

Visual comparison between standardized climate variables and the basal area increment (BAI) [cm²] respectively (note different scales). As climate variables, the precipitation (P) and temperature (T) from May to September as well as SPEI in summer (Jun-Aug) indicate drought and temperature deviations during the growing season. Sites are ordered along the precipitation gradient from wet to dry.

Figure 4

Plasticity of the growth response in site-specific EYs separated between drought and frost events. Lines between EYs are drawn for visibility of interaction and indicate no time-series. Standard errors were omitted for a better visibility (presented in Supplementary Table 2 ).

Figure 5

Resistance (Rt) per provenance to late frost (2010) is plotted against the mean diameter at breast height (DBH) (A) and at each site (B). In the lower panels, mean Rt for site-specific drought years is plotted against the DBH (C) and at each site (D). Point shapes refer to the three sites and line types illustrate significant (solid) or non-significant (dashed) trends. Standard errors of the means were omitted for a better visibility (presented in Supplementary Table 2 ). Significant differences between the site means by Kruskal-Wallis test are indicated by *p < 0.05, **p < 0.01, ***p < 0.001.