Loading...
Publication
Linking climate, distribuon and growth in forest research: can dendrochronology improve species distribuon modelling?
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 8.59 MB | Adobe PDF |
Authors
Abstract(s)
Estudos em dendrocronologia são mais escassos em áreas com humidade relativa alta, baixa amplitude térmica e sazonalidade pouco pronunciada. Por isso, esta lacuna de conhecimento cria uma oportunidade para explorar os anéis de crescimento de espécies arbóreas da floresta Laurissilva e montana dos Açores. Esta trabalho teve como objetivo reduzir esta lacuna e compreender melhor o papel da dendrocronologia na modelação ecológica, sendo que para tal, foram recolhidas mais de 900 amostras de madeira de três espécies arbóreas nativas dos Açores: Juniperus brevifolia (Seub.) Antoine (Cupressaceae), Ilex azorica Gand. (Aquifoliaceae) e Laurus azorica (Seub.) Franco (Lauraceae). Dez diferentes locais foram escolhidos em duas ilhas, São Miguel e Terceira, e foram aplicados métodos padrão da dendrocronologia e dendroclimatologia, que incluem o alinhamento, detrending e exclusão de amostras ou porções de cronologias. Várias opções de modelação (por exemplo, Modelos Lineares Generalizados e Random Forest) foram usadas para determinar os principais fatores climáticos que afetam o crescimento radial das árvores. Todas as três espécies estudadas mostraram alguns limites de anéis parcialmente indistinto. No entanto, após a aplicação rigorosa de métodos dendrocronológicos, mostraram um potencial aceitável para investigação em dendrocronologia. A madeira de J. brevifolia apresentou uma distinção de cores entre o alburno (tons amarelados) e o cerne (tons avermelhados). No I. azorica, foi encontrada uma linha de vasos associada à fronteira dos anéis na maioria das amostras. Ambas as cronologias maiores e mais pequenas foram encontradas para o I. azorica, respetivamente nas Sete Cidades com 47 anos de comprimento e na Terra Brava com 128 anos. Diversas relações foram encontradas entre o crescimento das árvores e temperatura, principalmente um efeito positivo da temperatura da primavera (I. azorica e L. azorica) e efeito negativo das temperaturas de verão (todas as três espécies). A precipitação da primavera também afetou o crescimento radial do I. azorica. O clima do ano anterior foi também relevante no crescimento das árvores do I. azorica e L. azorica, indicando efeitos fisiológicos tardios. Nos Açores, as limitações nos estudos de dendrocronologia estão relacionadas com a baixa amplitude térmica entre estações do ano, com a sazonalidade pouco pronunciada, e com a elevada variação dos padrões dos anéis de crescimento. A variação encontrada entre as populações está relacionada com perturbações naturais (elevação, diferenças de solo e processos de sucessão e regeneração) ou antropogénicas (corte de árvores e plantação de espécies exóticas). Ambos, são fatores que influenciam os diferentes comprimentos das cronologias construídas e das pequenas variações das relações de clima-crescimento. Os resultados globais sugerem que, no futuro, ao usar ferramentas dendrocronológicas com medições refinadas de anéis de crescimento, pode ser possível calcular as taxas de crescimento e integrar esta informação como variáveis preditoras em modelos de distribuição de espécies, embora necessitem de validação com dados de campo ou de deteção remota. O sinal de temperatura encontrado nas relações de crescimento climático é alarmante, porque espera-se que as temperaturas de verão aumentem (alterações climáticas), o que pode levar à extinção local de espécies. Assim, os métodos utilizados nesta tese sugerem que, esta abordagem pode ser utilizada para compreender o impacto das alterações climáticas no crescimento das árvores. Também, esta abordagem complementada com outras medidas de crescimento radial das árvores e uso de dendrómetros, pode ser aplicada a outras espécies nativas dos Açores em baixa altitude, como Morella faya e Picconia azorica, espécies exóticas, como a Cryptomeria japonica, ou outras espécies já estudadas em climas temperados na Europa. Com a base de dados criada neste trabalho, uma análise global com diferentes métodos de detrending e de abordagens de modelação pode ser aplicada, de forma a dar pistas para uma metodologia abrangente adequada para a região.
ABSTRACT: Dendrochronological studies are scarcer in areas with high relative humidity, low thermal amplitude and without a very pronounced climatic seasonality, such as the Azores Islands. Hence, this knowledge gap extends to the native and endemic Azorean tree species, for which almost no tree ring research was available at the beginning of the present work, despite the available knowledge about their ecological amplitudes, occurrences, and potential distribution. This research aimed to start fulfilling that gap and to clarify if dendrochronological research could support ecological modelling and, consequently, a better understanding of forest dynamics in the Azores. To accomplish this, more than 900 wood cores were collected from three Azorean endemic woody species: Juniperus brevifolia (Seub.) Antoine (Cupressaceae), Ilex azorica Gand. (Aquifoliaceae) and Laurus azorica (Seub.) Franco (Lauraceae). Ten different stands were sampled in São Miguel and Terceira islands, applying standard dendrochronological and dendroclimatic methods, including crossdating, detrending, and the exclusion of samples or chronology portions, due to tree ring anomalies or to low sampling depth. Likewise, several modelling options were used, such as Pearson correlation, Generalized Linear Models and Random Forest, to determine the most relevant climatic drivers of tree growth. Although the three species showed some partially indistinct ring boundaries and wedging rings, globally, the samples we considered as adequate for dendrochronological research. A clear separation between a yellowish sapwood and the reddish-brown heartwood was found in J. brevifolia. For I. azorica, a layer of vessels associated to the ring boundary was found in most growth rings. Both the shortest and longest chronologies were found for I. azorica, respectively in Sete Cidades with a length of 47 years and in Terra Brava with a length of 128 years. Diverse relations between tree growth and temperatures were found, mainly a positive effect of spring temperature (I. azorica and L. azorica) and a negative effect of temperatures (all three species). Precipitation also affected tree radial growth, especially the positive effect of spring precipitation (I. azorica). Additionally, previous year climate was also relevant on tree growth in I. azorica and L. azorica, indicating lagged physiological effects. In the Azores, limitations in dendrochronological studies are related to the low thermal amplitude, to the absence of a pronounced climatic seasonality, and to the high within stand variation in tree ring patterns. Moreover, the high variability found between stands is likely related to natural (e.g., elevation, terrain and soil differences, succession and regeneration processes) or anthropogenic disturbance (e.g, tree cutting, plantation and spread of exotic species), thereby contributing for a considerable variation in chronology length and to slightly different climate-growth relationships at different sites. Nevertheless, the global results suggested that, in the future, by using this dendrochronological approach with refined tree-ring measurements, it would be possible to calculate radial growth rates and to integrate this information in species distribution models, to be validated using field and remote sensing data. The temperature signal found in climate-growth relationships raises concern, because summer temperatures are expected to increase, which could lead to local extinctions. Our results also suggest that this approach could be used to better understand the impact of climate change on tree growth. Also, this approach, complemented with other radial growth proxies and the use of dendrometers, could be extended to other native species with lower elevation distribution ranges, such as Morella faya and Picconia azorica, or to exotic species, such as Cryptomeria japonica and tree species already studied in temperate Europe. Additionally, a global analysis of the entire dataset, exploring different detrending and modeling methods could also be undertaken, providing hints for a comprehensive methodology suitable for the region.
ABSTRACT: Dendrochronological studies are scarcer in areas with high relative humidity, low thermal amplitude and without a very pronounced climatic seasonality, such as the Azores Islands. Hence, this knowledge gap extends to the native and endemic Azorean tree species, for which almost no tree ring research was available at the beginning of the present work, despite the available knowledge about their ecological amplitudes, occurrences, and potential distribution. This research aimed to start fulfilling that gap and to clarify if dendrochronological research could support ecological modelling and, consequently, a better understanding of forest dynamics in the Azores. To accomplish this, more than 900 wood cores were collected from three Azorean endemic woody species: Juniperus brevifolia (Seub.) Antoine (Cupressaceae), Ilex azorica Gand. (Aquifoliaceae) and Laurus azorica (Seub.) Franco (Lauraceae). Ten different stands were sampled in São Miguel and Terceira islands, applying standard dendrochronological and dendroclimatic methods, including crossdating, detrending, and the exclusion of samples or chronology portions, due to tree ring anomalies or to low sampling depth. Likewise, several modelling options were used, such as Pearson correlation, Generalized Linear Models and Random Forest, to determine the most relevant climatic drivers of tree growth. Although the three species showed some partially indistinct ring boundaries and wedging rings, globally, the samples we considered as adequate for dendrochronological research. A clear separation between a yellowish sapwood and the reddish-brown heartwood was found in J. brevifolia. For I. azorica, a layer of vessels associated to the ring boundary was found in most growth rings. Both the shortest and longest chronologies were found for I. azorica, respectively in Sete Cidades with a length of 47 years and in Terra Brava with a length of 128 years. Diverse relations between tree growth and temperatures were found, mainly a positive effect of spring temperature (I. azorica and L. azorica) and a negative effect of temperatures (all three species). Precipitation also affected tree radial growth, especially the positive effect of spring precipitation (I. azorica). Additionally, previous year climate was also relevant on tree growth in I. azorica and L. azorica, indicating lagged physiological effects. In the Azores, limitations in dendrochronological studies are related to the low thermal amplitude, to the absence of a pronounced climatic seasonality, and to the high within stand variation in tree ring patterns. Moreover, the high variability found between stands is likely related to natural (e.g., elevation, terrain and soil differences, succession and regeneration processes) or anthropogenic disturbance (e.g, tree cutting, plantation and spread of exotic species), thereby contributing for a considerable variation in chronology length and to slightly different climate-growth relationships at different sites. Nevertheless, the global results suggested that, in the future, by using this dendrochronological approach with refined tree-ring measurements, it would be possible to calculate radial growth rates and to integrate this information in species distribution models, to be validated using field and remote sensing data. The temperature signal found in climate-growth relationships raises concern, because summer temperatures are expected to increase, which could lead to local extinctions. Our results also suggest that this approach could be used to better understand the impact of climate change on tree growth. Also, this approach, complemented with other radial growth proxies and the use of dendrometers, could be extended to other native species with lower elevation distribution ranges, such as Morella faya and Picconia azorica, or to exotic species, such as Cryptomeria japonica and tree species already studied in temperate Europe. Additionally, a global analysis of the entire dataset, exploring different detrending and modeling methods could also be undertaken, providing hints for a comprehensive methodology suitable for the region.
Description
Tese de Doutoramento, Ciências Biológicas, 26 de julho de 2023, Universidade dos Açores.
Keywords
Dendrocronologia Laurissilva Forest Azores Ecological Modelling
Citation
Pavão, Diogo Cláudio. (2023). "Linking climate, distribuon and growth in forest research: can dendrochronology improve species distribuon modelling?". 229 p. (Tese de Doutoramento em Ciências Biológicas). Ponta Delgada: Universidade dos Açores, 2023. Disponível em http://hdl.handle.net/10400.3/6782