Climate warming and disturbance influences on Nothofagus treelings in North Westland, New Zealand

Abstract

Globally, there is a great deal of concern over the impact of climate warming and other associated changes in climate. Treeline forests are generally thought to be highly sensitive to changes in temperature, as they occur at the temperature limits for tree growth. Consequently, in the Northern Hemisphere considerable research effort has focused on assessing whether treelines have responded to warming over the past century. However, few studies have considered the potential role of disturbance in modifying or even preventing a response by treelines to increased temperatures. This is despite the importance of disturbance in influencing both forest structure and growth. Therefore, this research has investigated the relative roles of disturbance and recent climate warming in driving the population dynamics and growth trends of Nothofagus menziesii (silver beech) dominated treelines in the Rahu Saddle area, north Westland, New Zealand.

This study combines forest stand history reconstruction and dendrochronological techniques to assess the relative roles of climate warming and disturbance at treeline. Seven 0.05 ha plots were established at three treeline sites in the Rahu Saddle area. Within and above each plot the species, size and spatial position of each individual stem (which can establish from seed or through vegetative reproduction) was recorded, with cores extracted from each stem ≥10 cm diameter for age determination (to reconstruct stand history) and chronology development (using silver beech stems only, to assess the relationship between growth and climate). While this study focused on the dynamics and growth trends at treeline, comparison of growth and growth-climate relationships between treeline and below may aid interpretation of the relative sensitivity of treeline growth to temperature. Therefore, further sampling from below treeline at each site was carried out. In total, 12 silver beech chronologies, 7 from the treeline plots and 5 from below treeline, were developed (these can be downloaded from the International Tree-Ring Data Bank; http://www.ngdc.noaa.gov/paleo/ftp-treering.html).

Based on the reconstruction of temporal and spatial patterns of recruitment, and the occurrence of abrupt growth declines over the last 300 years, it is inferred that disturbance is the primary control on the regeneration of Nothofagus treeline forests. Recruitment is highly episodic, and tends to occur in small, scattered patches. While specific windstorm events appear to be the cause of abrupt growth suppressions, periods of increased recruitment tend to occur over broad periods (c. 50 years) making it inappropriate to link recruitment to specific events. Rather, recruitment at treeline may be related to large-scale changes in circulation which may increase storminess as well as make treeline forests more susceptible to storms from unusual directions. The lack of recent recruitment within treeline forests or above treeline indicates that climate warming since c. 1950 in New Zealand has had little effect on Nothofagus treeline recruitment.

Dendrochronological analysis of the relationship between silver beech radial growth and climate (using response function analysis) found that year-to-year ring-width variation is related to temperature. However, the relationship between growth and temperature was more complicated than expected. Growth was positively related to current mid to late summer (January-February) temperatures, negatively related to previous summer temperatures, and possibly positively related to previous winter temperatures. Growth characteristics and growth-temperature relationships were compared at and below treeline to assess whether growth at treeline was more sensitive to temperature warming. The growth-temperature relationships were comparable, suggesting temperature is limiting in a similar way at and below treeline. Nevertheless, there were significant differences in the amount of common growth variation, with trees in the below treeline chronologies showing greater similarity in growth trends. It is suggested that this is related to the disturbance regime at treeline, where the stand-wide but small and patchy canopy-openings following windstorms may reduce the level of common growth variation at treeline.

Given that summer temperatures have increased since c. 1950 and summer temperature is the main limit on treeline growth, growth rates should have increased in recent decades. Although this appears to be true for three of the chronologies, in the remaining chronologies growth has actually declined since c. 1960, indicating spatial variation in response to climate warming. The growth decline strongly suggests that the growth-temperature relationship has changed in recent decades and that growth no longer responds to temperature trends. This visual observation was confirmed by quantifying the synchronicity and divergence between growth and temperature trends over the last 150 years, which indicated that radial growth has generally lost sensitivity to temperature forcing over the last 30 years.

As disturbance is a major influence on the dynamics of Nothofagus treeline forests, it seemed possible that disturbance and/or changes in competition related to stand dynamics could be the cause of the recent loss of growth sensitivity to temperature. This was investigated using split and stepped response functions to determine whether the growth-temperature relationship had changed through time. Early in the 20th century specific disturbance events (already identified as causing severe growth suppressions) appear to affect the growth-temperature relationship, although usually for only 5-10 years. The more recent period of loss of sensitivity to temperature appears to be related to an increase in competition as stems initiated during a late 19th/early 20th century recruitment period grow larger and move into the canopy. This competition effect was evident as a strong negative relationship between neighbour density and average growth rates between 1960 and 1989. However, the effect of competition on growth is highly spatially variable. Those stems surrounded by many neighbours show a growth decline and the competition for resources in these dense patches appears to prevent or maybe delay a response to warmer temperatures. In contrast, stems surrounded by fewer neighbours are unaffected by competition and show the expected increase in growth in recent decades. This spatial variation in competition is a consequence of the spatial patterns of recruitment at treeline. Therefore, a growth response to climate warming appears to be strongly disturbance-mediated.

Overall, this study suggests that the main driver of Nothofagus treeline dynamics and growth trends over the last 100 years has been disturbance, not climate warming.