Research progress in monitoring and simulating stem radius growth: an overview

ZhangYong Wang , Bao Yang , Chun Qin , Feng Shi

1. Key Laboratory of Desert and Desertification, Cold and Arid Regions Environmental and Engineering Research Institute,Chinese Academy of Sciences, Lanzhou, Gansu 730000, China

2. Graduate School of the Chinese Academy of Sciences, Beijing 100049, China

Research progress in monitoring and simulating stem radius growth: an overview

ZhangYong Wang1,2, Bao Yang1*, Chun Qin1, Feng Shi1

1. Key Laboratory of Desert and Desertification, Cold and Arid Regions Environmental and Engineering Research Institute,Chinese Academy of Sciences, Lanzhou, Gansu 730000, China

2. Graduate School of the Chinese Academy of Sciences, Beijing 100049, China

Traditional dendroclimatology research is based on the statistical correlation between tree-ring chronology and modern observation data, but lacks consideration of the mechanisms of tree-ring growth. In order to reconstruct past climatic conditions more reliably, the response of stem radius growth dynamics should be studied in detail. At present, in international researches, monitoring and simulation methods are generally treated as two separate approaches. The monitoring method generally consists of dendrometer measurements, micro-coring, and pinning. Here, we present the research progress of stem radius growth dynamics studies derived from these methods, and provide prospects for future study.

model; dendrometer; micro-coring; pinning

1. Introduction

Dendroclimatology developed on the basis of tree physiology, namely, determining how certain limiting conditions in the forest environment can affect tree growth and the form of many plant structures. Thus, dendroclimatology has been considered a science that studies the relationships between climatic change and tree-ring width, density, and stable isotopes to reconstruct past climatic and environmental conditions. In the ongoing international PAGES (Past Global Changes) research project, tree-ring research is considered as one of the most important technological approaches to study past global change (Wu, 1990), and is very often applied in paleoclimate researches (Fritts, 1976; Liet al., 2000;Mann and Jones, 2003; Yuet al., 2008; Yuanet al., 2008;Zhanget al., 2008; Mannet al., 2009; Wanget al., 2009;Zhanget al., 2010).

In general, traditional dendroclimatology uses nearly linear equations to interpret the relationship between tree-ring growth and its main limiting factors. According to this principle, one common approach is to work through correlation and regression analyses, and then reconstruct past climatic conditions by using the transfer function which is established by the regression equations (Fritts, 1976;Hughes, 2002). However, the relationship between tree-ring growth and climate is not a simple linear correlation, but,rather, a complex nonlinear combination of the effects of various factors (Fritts, 1976; Vaganovet al., 2006).

At present there is no perfect method to separate the climatic signal from the biological signal in tree-ring formation that comes with increasing tree age (Cook and Kairiukstis,1990). The common detrending approaches may lose the low-frequency signal of climatic change. As temperatures increased in many regions during the last century, a phenomenon called "divergence" between tree growth and climate factors has been reported (Jacoby and D’Arrigo, 1995;Briffaet al., 1998; Oberhuber, 2004; Wilmkinget al., 2004;Wilmking and Juday, 2005; Carrer and Urbinati, 2006;D’Arrigoet al., 2008). This encompasses the divergence between tree-ring-based reconstruction and measured data,the divergence of tree growth and its main limiting climate factors, and the divergent growth responses of trees within study sites. All of these phenomena indicate that the statistically approximated linear relationship in traditional dendroclimatology research might not universally apply between tree growth and its limiting climate factors.

Thus, it is essential to determine the physiological response mechanism of tree-ring growth to climate change,such as the onset and cease of stem radial growth periods,the formation periods of earlywood and latewood, and the main limiting factors to tree-ring growth in different temporal and spatial scales. In this way, the disadvantages of traditional dendroclimatology statistical methods could be offset and the development of dendroclimatology could be accelerated to some extent.

As mentioned earlier, monitoring and simulation methods are currently treated as the two main approaches to study stem radial growth. Simulation methods use or establish models to study the response relationship between stem radial growth and climatic factors. The monitoring method mainly involves dendrometer measurements, micro-coring,and pinning. In this article, we discuss in detail the research progress on stem radial growth response to climatic change;and elucidate the current problems and development trends for future study.

2. The simulation method

Model simulation research is based on the response mechanism of tree growth to climatic factors. Several factors are considered in a model, and the variations of those factors are used to simulate the process of stem radial growth. A model enables the study of different limiting factors at different time scales, unlimited by the amplitude of calibration period, which can avoid problems such as the changed sensitivity of stem radial growth response to climate (Vaganovet al., 2006). Therefore, model simulation research is highly significant for both past climatic reconstruction and projected tree growth in the future climatic background (Yin,2009; Liet al., 2010).

From the 20th century to the present time, several physiological tree growth models gradually appeared. In 1968,Wilson and Howard (1968) built a model for the quantitative description of tree-ring formation. This model reproduced the structures formed in the tree stem during the growing season; it required at least 14 variables and the seasonal observational data had to be specifically defined or measured for any particular season. Furthermore, the model did not include external environmental parameters. Thus, this model has not been widely used. In 1975, Stevens (1975) modified the Wilson-Howard model, setting 22 parameters as basic inputs to simulate cambial activity and tree-ring formation.However, other than having more inputs, there seemed to be no essential differences between Stevens’s model and the Wilson-Howard model, as the former did not include the influence of environment on tree-ring formation. In 1990, a model called TRACH was developed (Vaganov, 1990).Daily meterological data were used to simulate the changes of cambial cell activities, and the links between environmental conditions and cambial cell growth were considered comprehensively. However, the model did not link soil moisture balance with a tree’s transpiration and photosynthesis. Then, Shashkin and Vaganov (1993), Vaganov(1996a) overcame these disadvantages with the new Vaganov-Shashkin model (V-S model). In 1999, Frittset al.(1999) united the TRACH and V-S models to create the TREE-RING model. Most of these models were designed to study softwoods. In 2010, Drewet al.(2010) built the CAMBIUM model, which was based on the CABALA model (Battagliaet al., 2004). This model is able to provide realistic estimates of short-term variation and temporal ranges in eucalypt fiber diameter and secondary wall development and wood density. Among such models, the V-S and TREE-RING models are the most popular and widely used,due to their simplicity and their proven success in simulating coniferous tree growth.

2.1. V-S and TREE-RING models

The V-S model (Vaganovet al., 2006) was built based on the TRACH model. Several climatic factors were used to simulate tree-ring formation in the V-S model, and this model simplified the complex process of tree-ring formation by assuming that the influence of environmental conditions on tree-ring formation could be reflected in cambial activity.The "main limiting factor principle" was used to simulate the growth rate of cambial cells, that is, the growth rate at a certain time of a season cannot be higher than that allowed by the most limiting factor. The V-S model makes use of a limited number of equations relating daily temperature, precipitation, and sunlight to the kinetics of secondary xylem development in order to model tree-ring growth and the internal characteristics. The climatic variabilities, including temperature, soil moisture balance, and solar irradiance, are used to simulate the rate and timing of growth and division of cells in the cambium. In this way, the kinetics of xylem formation is explicitly modeled as a function of climate variability modified by parameterized environmental and cambial processes. Compared with the TREE-RING model,which explicitly treats photosynthesis, respiration, and the partitioning of assimilates, its applicability is limited (Vaganovet al., 2006; Yin, 2009).

The TREE-RING model was first presented by Frittset al.(1999), and was derived from the V-S model and the TRACH model. At present, it is commended as being the most comprehensive model simulating growth processes of needle species (Downeset al., 2009). The TREE-RING model improved and expanded the calculation of cambial zone cell growth rate from part of the V-S model. Four main modules are included in the TREE-RING model: the micro-climatic factors module, the soil moisture balance module, the photosynthesis and respiration module, and the growth module. The growth module is the heart of the whole model; it simulates the growth of root and leafage, and the processes of cambial cell division, enlargement, and cell wall thickening. The daily maximum and minimum air temperature, precipitation, and illumination conditions are used as the main inputs for the model. Several factors, such as the gradient of the research site, the forest crown density,the depth of the main roots, the soil layer thickness, and the maximum soil water content are considered by carefully setting the parameters. The simulation of photosynthetic rate is effectively linked with stomatal activities of leafage, light intensity, temperature, and carbon dioxide concentration,which makes the physical process more intact (Yin,2009).

2.2. Applications using the models

Physiological tree growth models have been widely applied abroad, especially in North America and Russia (Fritts,1990; Vaganov, 1996b; Arbatskaya and Vaganov, 1997; Vaganovet al., 1999). Evanset al.(2006) investigated the interpretation of tree-ring data using the V-S model in North America and Russia from 198 data sets, and found the V-S model to be as capable of accurately simulating intraseasonal to interdecadal climate variability as the classical dendrochronological statistical modeling techniques. Also, the model showed a good simulation of actual tree-ring width chronologies from a variety of environments using a single fixed set of model parameters.

Vaganovet al.(2006) used the V-S model to simulate seasonal growth and tree-ring formation dynamics under different climatic conditions, from subarctic to monsoon, of different species such asPinus rigida,P. sylvestris,P. densiflora, andP. ponderosa; different ecosystems such as taiga zones, steppe zones, and semi-arid forests were compared to ascertain the correct determination of model parameters,demonstrating that this model has considerable potential for the prediction of tree-ring growth response under different climate situations. The TREE-RING model performed well withP. ponderosaandP. sylvestrisin Arizona and western Siberia, respectively.

Tree-ring formation and structure can be affected by environmental pollution; therefore, some scholars have used model simulation to study the response dynamics of tree-ring growth to pollution events. Simachevet al.(1992),Ivshin and Shiyatov (1995), and Kharuket al.(1996) conducted simulation studies of tree-ring growth in the industrial emissions area in Norilsk, Russia.

In China, Yin and Wu (1995) used the TRACH model to simulateP. armandiiin the Qinling Mountains, and found that tree-ring formation was limited by the current growing season and previous annual precipitation, and that temperature only has effects during the beginning of the growing season. Shiet al.(2005, 2006, 2008) used the V-S model in the Helan Mountains between the Yinchuan Plain and the Inner Mongolia Prairie to simulate the growth-climate relationship ofP. tabulaeformis, and found that tree-ring formation was mainly limited by temperature from April to mid-May, and was also limited by precipitation from late May to late August. Liet al.(2010) used the TREE-RING model found that the growth ofPicea crassifoliatree-ring formation in the Halihatu region started on May 9 and stopped on October 28, and the soil moisture condition in the early period of the growing season was the main climate factor limiting tree growth. Yinet al.(2009) used the TREE-RING model to simulatePinus koraiensisin Xiaoxing’anling, and showed that the model could simulate tree growth accurately. The tree-ring width is mainly controlled by October temperatures of the year prior to growth and April temperatures of the growth year, and cambium growth begins in late April and ends in early October.

In China, simulation researches have been concentrated in the northwestern regions and have played important roles in illustrating tree-ring formation dynamics and the meaning of dendroclimatological proxy data. However, compared with international researches, the limiting aspects in China are that the simulation contents are not abundant enough and there are no long-term measurements to support the model parameters set.

3. The dendrometer measurement method

Dendrometers are instruments used to measure the diameters of the stems of trees. High-resolution dendrometer data were already being used in the 19th century, in research of looking at stem growth responses in multiple tree species under various weather conditions (Drew and Downes, 2009).High-resolution dendrometers are classified as two types:the point-type and the band-type (Breitsprecher and Hughes,1975). Point dendrometers usually measure stem radius and diameter changes, and are effective for measuring the variations at a certain point or direction of a stem. Thus, they are used to study such important events of stem radius growth as the onset, duration, and cessation of growing seasons (Drew and Downes, 2009). Band dendrometers measure stem circumference variations and are effective for studying the mean level of stem radius growth.

3.1. The usage of dendrometer data

Dendrometers measure stem radius variations caused by diurnal rhythms of water storage depletion and replenishment, and seasonal tree growth. In all the approaches described here to determine stem radius growth, the first step consists of removing the daily shrinking and swelling fluctuations caused by stem water status. According to how the stem variation is determined, the three main ways to do this are the stem cycle approach, the daily mean approach, and the daily maximum approach (Deslaurierset al., 2007).

Downeset al.(1999), based on the patterns of stem shrinking and swelling, divided the daily stem radius variations into three distinct phases (Figure 1): shrinkage, recov-ery, and increment. The period from a previous local maximum to a local minimum can be defined as shrinkage (S).Variation from a local minimum until the magnitude of the previous maximum can be defined as recovery (R). Variation from a previous maximum to a new maximum size can be defined as increment (I); when the previous cycle maximum was not reached, the stem radius decrease can be calculated but no increment phase is defined. The stem circadian cycle often lasts approximately for 24 h, but heavy rain in summer or freeze-thaw events in the rest of the year may cause long cycles of more than 24 hours (Zweifel and Häsler, 2000). Deslaurierset al.(2003b) and Giovannelliet al.(2007) used similar defined phases, but combined the recovery and increment phases into an expansion phase(Figure 1). Herzoget al.(1995) defined the stem circadian cycle as having five phases. Although those phase partitions were essentially similar, they allowed the amplitude of stem radial variations as well as their duration to be calculated.

Figure 1 Measurements of stem radius variations of Qilian juniper in the Qilian Mountains during May 14-15, 2010,showing phases of increment (I), shrinkage (S), recovery (R), as well as expansion (E).

The daily mean and daily maximum approaches consist of extracting one value (average and maximum, respectively)from the daily time series. Then the time series can be quickly reduced from raw data per day to just one parameter.Stem radius change information can thus be extracted by the computation in different time scales (Tardifet al., 2001;Bouriaudet al., 2005). Several scholars have used the difference between two consecutive daily maxima as the net rate of daily radius change (dR) in the growing season.When active growth that the positive dR indicates the formation of new cells in the xylem, however, stem diameters can also decrease and dR can become negative (in the case of cambial dormancy and net water loss of the trunk)(Bräuninget al., 2009).

3.2. Applications using dendrometer measurements

In the past decade, several scholars used high-resolution dendrometers to extract stem radius growth information in different regions, from northern frigid zones to tropics, and abundant results have been achieved.

Drewet al.(2009) used point dendrometers to measure two commercial hybridEucalyptusclones over 3.5 years in South Africa, and found that GU (E. grandis×urophylla) is more susceptible to drought stress than GC (E. grandis×camaldulensis). Downeset al.(1999) took dendrometer measurements in irrigatedE. nitensandE. globulustrees in southeastern Tasmania to identify the optimal means of relating weather variation to radial increments when water is not limiting growth. They found that the correlation between temperature and stem growth varied from positive in spring to zero or negative in summer.

Bouriaudet al.(2005) studied intra-annual radial growth variations with stem height change of Norway spruce trees by point dendrometers over four years in northeastern France, and found that radial growth started in early April and ended in August or September. The sensitivity of radial growth to climate decreased with stem height, and a dry period in June 1996 induced false-ring formation. Deslaurierset al.(2007) used automatic band dendrometers to monitorPicea abiesandLarix deciduaat different altitudes in the eastern Italian Alps; different approaches, such as the stem cycle, daily mean, and daily maximum approaches,were compared, and they found that these approaches produced similar results of stem radius variation but there were differences in the amplitude of the stem variation. The stem cycle approach has the advantage of extracting continuous stem contraction or expansion, and is recommended when a high frequency of long circadian cycles (more than 28 hours)occurs. The main growth period from May to July corresponded mainly with earlywood cell formation. Carreret al.(1998) used band dendrometers to monitor the intra-annual radial growth dynamics ofPicea abies,Pinus cembra, andLarix deciduain timberline mixed forests in the eastern Italian Alps, and found that the radius growth rate of the three species in June and July was greater when air temperatures were higher.

Tardifet al.(2001) used band dendrometers to measure seven species trees in northwestern Québec. Their results indicated that air temperature was positively related to stem swelling during the late winter to early spring period. Starting in early June, all species registered a sustained increase in radial increments, possibly associated with active cell division; then radial expansion was negatively related to air temperature and positively to rainfall. Deslaurierset al.(2003b) used point dendrometers to study daily stem radius growth ofAbies balsameain the boreal forests of Québec,and found that total rainfall in phases 2, 3, and 4 was correlated positively with SRI (stem radius increment); a high vapor pressure deficit in Phase 2 decreased the possibility of cell radial expansion; and the night temperature was more important than day temperature in controlling radial growth. McLaughlinet al.(2007) took manual measurements of 86 trees and automatic dendrometer measurements of six trees in eastern Tennessee (USA) over two to three years, and found that ambient O3caused a periodic slowdown in seasonal growth patterns that was attributable, in part, to amplification of diurnal patterns of water loss in tree stems. Biondi and Hartsough (2010) used point dendrometers to studyPinus hartwegiiat the tropical treeline of North Amercia; their results showed that maximum radial growth rates occur in late spring (May), as soil temperature increases and incoming short-wave radiation reaches its highest values.

Volland-Voigtet al.(2009) used point dendrometers to measureTabebuia chrysanthaandTerminalia valverdeaein two different ecological forest types in South Ecuador. Their results demonstrated that even in very humid tropical mountain climates, cambial activity reacted very sensitively to moisture availability; a few consecutive dry days were enough to cause a drastic decrease in cambial activity.Worbes (1999) used band dendrometers to study different tree species’ growth patterns in the Caparo Forest Reserve in Venezuela, and found that annual rings were formed by many species, and the distinctiveness of growth zones was usually greater in deciduous species than in evergreen species.

Similar researches conducted in China began later than in other countries, and the research sites were limited to arid and semi-arid regions. Guanet al.(2007) and Xionget al.(2007) used band dendrometers to monitor stem radius growth ofLarix principis-rupprechtiiin the Liupan Mountains. They found that June to July was the rapid stem-growing season, and stem growth may stop in late August. The pattern of cumulative stem diameter growth could be satisfactorily fit by the power function, and temperature and moisture were the main limiting factors to stem radius growth. Liet al.(2007) studiedPinus tabulaeformis,Populus davidiana,Quercus liaotungensis, andElaegnus umbellataby point and band dendrometers in a loess hilly region, and found that the seasonal growth of stems was a gradual increase accompanying diurnal shrinkage and swelling.Polulus davidianahad the longest growth period.When stem radius growth was associated with climate, the factors affecting stem radius growth varied across tree species. Yanget al.(2009) monitoredLarix principis-rupprechtii’s stem radius growth by point dendrometers in the Luya Mountains, and found that positive values of stem radius growth appeared in summer, while negative values appeared in autumn; the stem radius increment had a significant positive correlation with soil moisture and a negative correlation with air temperature in the two measurement periods.

Generally speaking, compared with the international literatures, the dendrometer measurement studies in China have been characterized by short data series which did not span even a length of a whole year, and some results which did not apply to tree-ring/climate dynamics interpretation.For dendrometer-measured stem radius variations contained long-term irreversible stem growth and short-term reversible shrinking and swelling change due to stem water status change, the stem radius growth information could not be precisely extracted from the raw data series. Thus, dendrometers have been criticized when used to measure short-term growth rates (Zweifel and Häsler, 2000, 2001;Zweifelet al., 2000; Mäkinenet al., 2003; Mäkinenet al.,2008). However, the usefulness of this instrument, with which stem radius variations can be continuously and easily measured with high resolution and over a long study period,makes it a unique method when compared with other methods. It should be noted that, when using a dendrometer to study stem radius growth, appropriate approaches to compute data series should be selected with caution, and other research methods could be utilized for comparison purposes.

4. The micro-coring and pinning methods

Micro-coring and pinning are two monitoring methods that focus on cellular analysis of cambium activity and wood formation. Thus, the two methods have many similarities in some operating steps (Seoet al., 2007).

4.1. Micro-coring

The micro-coring method is based on the different patterns evident during the radial enlargement phase, the wall thickening phase, and the mature cell phase of cell formation in growth rings during the growing season. By repeated sampling and cellular analyses, the cell number and growth rate in different phases, and the onset and cessation of the growing season, can be determined. For example, Riding and Little (1984) reported that the dormant cambium before the beginning and at the end of the growing season was easily identified by two to four radially flattened fusiform cells.

A good description of the micro-coring method is given in Deslaurierset al.(2003a). A trephor (Rossiet al., 2006a)is usually used for the extraction of small cores weekly or biweekly in the growing seasons. The cores were 1-2 mm in diameter and 15-20 mm long, containing four to six rings.Samples were taken following a spiral trajectory up to the stem, from 30 cm below the breast height to 30 cm above.Wood cores were always taken at least 10 cm apart from each other in adult trees and 5 cm apart in young trees to avoid getting resin ducts on adjacent cores. In the laboratory,the small cores were fixed in paraffin, and then transverse sections were prepared and immersed in solution (Histo-ClearTMand alcohol) to remove the paraffin. The sections were then dehydrated and put into water before being stained with a water solution of cresyl fast violet and kept in water to assess the ring development. Sections were stained with safranin (Antonova and Shebeko, 1981) and permanently fixed to assess cell measurement with a computer program. A Polaroid camera fixed on an optical microscope and connected to a computer was used for numerical image analysis. The measured parameters could be lumen area,single cell wall thickness, lumen diameter, and total cell width. Depending on different study aims, the number of cells was counted on radial files on the rings and subsequently used for a cell number standardization for each sample.

4.2. Pinning

The concept of pinning is based on the simple fact that the cambium of trees is highly responsive to external impacts (Larson, 1994). In general, in response to the action of inserting a pin through the bark and cambium into the xylem of a tree, the cambium will, first, just go through the pinning canal but will stop forming new cells, and second, apart from the pinning canal it will develop modified cells. Both of these responses allow the amount of wood formation to be monitored within a certain period of time. After the growing season, the pinning canal with some tissue around is removed from the standing tree using a chisel and is then processed in the laboratory.

4.3. Applications using the micro-coring and pinning methods

Rossiet al.(2006b) used the micro-coring method to study three adult tree species,Larix decidua,Pinus cembra,andPicea abies, at the timberline in the eastern Italian Alps.They found that all of the species showed the same trend of xylem formation, a term from 100 to 130 days was required to complete cell differentiation, and tree-ring formation ended in September. Seoet al.(2007) studiedPinus sylvestrisandPicea abiesin northern and southern Finland, and concluded that the pinning method was useful as a "timer" to split up the coherent process of wood formation during the growing season. Van der Werfet al.(2007) used micro-coring and pinning methods by biweekly sampling to study the intra-annual tree-ring formation ofFagus sylvaticaandQuercus roburin the eastern part of the Netherlands;their results indicated that oak and beech reacted differently to the summer drought in 2003.

Deslaurierset al.(2003a) used the micro-coring method to study the timing of ring formation and the development patterns of earlywood and latewood ofAbies balsameain the Québec boreal forest. Bräuninget al.(2009) combined the dendrometer measurement and micro-coring methods,and monitored the seasonal stem growth dynamics of fourCedrela montanaspecimens in the humid mountain rainforests of southern Ecuador from 2006 to 2009. They found that high-resolution dendrometer data indicated a regular seasonal growth rhythm with cambial activity during January and April, and the cambial activity is limited by available moisture even in a very humid mountain climate. They concluded that ring width might not be the optimal wood parameter to detect the influence of climate on the trees and to reconstruct former environmental conditions.

Similar studies conducted in China are relatively rare.According to our collected literatures, only Lianget al.(2009) used the micro-coring method to study the seasonal dynamics of cambial activity ofPinus tabulaeformisin the eastern Ortindag Sand Land in North China. They found that cell division in the cambial zone started within the third week of May, in June and July the rate of xylem cell production was the highest, and around mid-August cell division ended. When associated with climate, the cell-division period appears to coincide with the time of the highest monsoon precipitation and the time of ＞0 °C daily minimum temperature.

Compared with dendrometer measurements, the micro-coring and pinning methods, which can observe the process of cellular activity, are considered the most reliable techniques to monitor wood formation during a growing season. However, performing micro-coring and pinning in the forest and laboratory are time-consuming and the microscopic evaluation requires experience in wood anatomy.Also, due to the gradient of a slope and the competition between individual tree specimens, cambial activity around the stem circumference may differ in short tangential distances,obscuring measurements of weekly or biweekly increments between adjacent samples. Given the advantages and disadvantages of the dendrometer measurement method and the micro-coring and pinning methods, the accuracy of all the methods could be improved if they are used in combination.

5. Conclusions and prospects

In the past five to six decades, researches on stem radius growth using simulation and monitoring methods have produced abundant results. The timing of some important events of stem radius growth in different climate zones has been elucidated, such as the onset, duration, and cessation of the growing season; certain responses of tree-ring formation to special environmental events, such as extreme drought,outbreak of insect pests, or environmental pollution, have been distinguished; and tree-ring/climate dynamics in some regions have been explained. All of these results have improved our understanding of tree-ring/climate relationships,have laid a good foundation for similar researches in other areas, and have provided many new ideas for reconstructing past climatic changes reliably.

However, there are still some problems to be considered in future study:

(1) The combination of simulation and monitoring methods should be further investigated. At any given site,setting the model parameters needs long-term instrument monitoring studies. In the existing researches, the model parameters were usually derived from existing results of other climatic regions and tree species; this may increase uncertainties in simulation results. The monitoring research of stem radius growth could provide deeper insights into continuous changes of cambial activities in different time scales. Combining simulation and monitoring studies will provide more accurate parameters for a model, leading to a much better understanding of the tree-ring/climate response mechanism.

(2) The combination of different monitoring methods should be considered, and monitoring data series need to be extended. Monitoring techniques such as dendrometer measurements, micro-coring, and pinning, all have their strengths and weaknesses. If they are used in combination,the accuracy of all the methods could be improved. Continuous monitoring data series should be extended to more than two years, which may reveal the true rules of stem radius growth.

(3) Different climatic zones and different tree species should be included in future studies. In China, the stem radius growth research regions have been limited to northwestern China, and the trees have been limited to the traditional species of dendroclimatological studies. Simulation and monitoring studies of tree-ring/climate response dynamics should be extended to wider regions and more tree species. This could provide more useful information for the past climatic reconstruction in large temporal and spatial scales. Furthermore,the comparisons of tree-ring/climate response dynamics at different altitudes should be considered.

The study was jointly funded by the CAS Strategic Priority Research Program (Grant No. XDA05080801), the National Science Foundation of China (Grant No. 41071130), and the National Basic Research Program of China (973 Program)(No. 2010CB950104). The authors thank YangYang Li,ErYuan Liang, and Wei Guan for their patient discussions during the process of writing this article.

Antonova G, Shebeko V, 1981. Applying cresyl violet in studying wood formation. Khimiya Drevesiny, 4: 102-105.

Arbatskaya M, Vaganov E, 1997. Long-term variation in fire frequency and radial increment in pine from the middle taiga subzone of central Siberia.Russian Journal of Ecology, 28(5): 291-297.

Battaglia M, Sands P, White D, Mummery D, 2004. CABALA: A linked carbon, water and nitrogen model of forest growth for silvicultural decision support. Forest Ecol. Manag., 193(1-2): 251-282. DOI:10.1016/j.foreco.2004.01.033.

Biondi F, Hartsough P, 2010. Using automated point dendrometers to analyze tropical treeline stem growth at Nevado de Colima, Mexico.Sensors, 10(6): 5827-5844. DOI: 10.3390/S100605827.

Bouriaud O, Leban JM, Bert D, Deleuze C, 2005. Intra-annual variations in climate influence growth and wood density of Norway spruce. Tree Physiol., 25(6): 651-660.

Bräuning A, Volland-Voigt F, Burchardt I, Ganzhi O, Nauss T, Peters T, 2009.Climatic control of radial growth of cedrela montana in a humid mountain rainforest in southern Ecuador. Erdkunde, 63(4): 337-345.DOI: 10.3112/erdkunde.2009.04.04.

Breitsprecher A, Hughes W, 1975. A recording dendrometer for humid environments. Biotropica, 7(2): 90-99.

Briffa KR, Schweingruber FH, Jones PD, Osborn TJ, Shiyatov SG, Vaganov EA, 1998. Reduced sensitivity of recent tree-growth to temperature at high northern latitudes. Nature, 391(6668): 678-682.

Carrer M, Anfodillo T, Urbinati C, Carraro V, 1998. High-altitude forest sensitivity to global warming: Results from long-term and short-term analyses in the eastern Italian Alps. The Impacts of Climate Variability on Forests, 74: 171-189.

Carrer M, Urbinati C, 2006. Long-term change in the sensitivity of tree-ring growth to climate forcing inLarix decidua. New Phytol., 170(4):861-872. DOI: 10.1111/j.1469-8137.2006.01703.x.

Cook ER, Kairiukstis L, 1990. Methods of Dendrochronology: Applications in the Environmental Sciences. Kluwer, Dordrecht.

D’Arrigo R, Wilson R, Liepert B, Cherubini P, 2008. On the "divergence problem" in northern forests: A review of the tree-ring evidence and possible causes. Global and Planetary Change, 60(3-4): 289-305. DOI:10.1016/j.gloplacha.2007.03.004.

Deslauriers A, Morin H, Begin Y, 2003a. Cellular phenology of annual ring formation ofAbies balsameain the Québec boreal forest (Canada).Canadian Journal of Forest Research, 33(2): 190-200.

Deslauriers A, Morin H, Urbinati C, Carrer M, 2003b. Daily weather response of balsam fir (Abies balsamea(L.) Mill.) stem radius increment from dendrometer analysis in the boreal forests of Québec(Canada). Trees--Structure and Function, 17(6): 477-484. DOI:10.1007/s00468-003-0260-4.

Deslauriers A, Rossi S, Anfodillo T, 2007. Dendrometer and intra-annual tree growth: What kind of information can be inferred? Dendrochronologia,25(2): 113-124. DOI: 10.1016/j.dendro.2007.05.003.

Downes G, Beadle C, Worledge D, 1999. Daily stem growth patterns in irrigatedEucalyptus globulusandE.nitensin relation to climate.Trees--Structure and Function, 14(2): 102-111.

Downes GM, Drew D, Battaglia M, Schulze D, 2009. Measuring and modelling stem growth and wood formation: An overview.Dendrochronologia, 27(2): 147-157. DOI: 10.1016/j.dendro.2009.06.006.

Drew DM, Downes GM, 2009. The use of precision dendrometers in research on daily stem size and wood property variation: A review.Dendrochronologia, 27(2): 159-172. DOI: 10.1016/j.dendro.2009.06.008.

Drew DM, Downes GM, Grzeskowiak V, Naidoo T, 2009. Differences in daily stem size variation and growth in two hybrid eucalypt clones.Trees--Structure and Function, 23(3): 585-595. DOI:10.1007/s00468-008-0303-y.

Drew DM, Downes GM, Battaglia M, 2010. CAMBIUM, a process-based model of daily xylem development inEucalyptus. Journal of Theoretical Biology, 264(2): 395-406.

Evans M, Reichert B, Kaplan A, Anchukaitis K, Vaganov E, Hughes M,Cane M, 2006. A forward modeling approach to paleoclimatic interpretation of tree-ring data. Journal of Geophysical Research,111(G3): G03008. DOI: 10.1029/2006JG000166.

Fritts HC, 1976. Tree Rings and Climate. Academic Press, London.

Fritts HC, 1990. Modeling tree-ring and environmental relationships for dendrochronological analysis. In: Dixon R, Meldahl R, Ruark G, Warren W (eds.). Process Modeling of Forest Growth Response to Environmental Stress.Timber Press, Portland, Oregon, pp. 360-382.

Fritts H, Vaganov E, Scviderskaya I, Shashkin A, 1991. Climate variation and tree-ring structure in conifers: Empirical and mechanistic models of tree-ring width, number of cells, cell size, cell-wall thickness and wood density. Clim. Res., 1(2): 97-116.

Fritts HC, Shashkin A, Downes GM, 1999. A simulation model of conifer ring growth and cell structure. In: Tree-Ring Analysis: Biological,Methodological and Environmental Aspects. CABI Publishing,Wallingford, UK, pp. 3-32.

Giovannelli A, Deslauriers A, Fragnelli G, Scaletti L, Castro G, Rossi S,Crivellaro A, 2007. Evaluation of drought response of two poplar clones(PopulusxcanadensisMönch "I-214" andP. deltoidesMarsh. "Dvina")through high resolution analysis of stem growth. J. Exp. Bot., 58(10):2673-2683. DOI: 10.1093/Jxb/Erm117.

Guan W, Xiong W, Wang Y, Yu P, He C, Du A, Liu H, 2007. Stem diameter growth ofLarix principis-rupprechtiiand its response to meteorogical factors in the north of Liupan Mountain. Scientia Silvae Sinicae, 43(9):1-6.

Herzog K, Häsler R, Thum R, 1995. Diurnal changes in the radius of a subalpine Norway spruce stem: Their relation to the sap flow and their use to estimate transpiration. Trees--Structure and Function, 10(2):94-101.

Hughes MK, 2002. Dendrochronology in climatology—The state of the art.Dendrochronologia, 20(1-2): 95-116. DOI: 10.1078/1125-7865-00011.

Ivshin A, Shiyatov S, 1995. The assessment of subtundra forests degradation by dendrochronological methods in the Norilsk industrial area.Dendrochronologia, 13: 113-126.

Jacoby GC, D’Arrigo RD, 1995. Tree ring width and density evidence of climatic and potential forest change in Alaska. Global Biogeochemical Cycles, 9(2): 227-234.

Kharuk V, Winterberger K, Tsibulskii G, Yakhimovich A, Moroz S, 1996.Technogenic disturbance of pretundra forests in Norilsk valley. Russian Journal of Ecology, 27(6): 406-410.

Larson PR, 1994. The Vascular Cambium: Development and Structure.Springer Series in Wood Science. Springer-Verlag, Berlin.

Li J, Yuan Y, You X, 2000. The Research and Application of Dendrohydrology. Science Press, Beijing.

Li Y, Shao X, Guo J, 2010. Primary discussion of the climate mechanism response ofPicea crassifoliaon east edge of Qaidem basin. Plateau Meteorology, 29(2): 349-358.

Li Y, Shi H, Zhang A, Tan H, 2007. Daily changes in radial growth of several tree stems and their response to environmental factors in loess hilly region. Journal of Soil and Water Conservation, 21(6): 170-196.

Liang E, Eckstein D, Shao X, 2009. Seasonal cambial activity of relict Chinese pine at the northern limit of its natural distribution in north China—Exploratory results. IAWA Journal, 30(4): 371-378.

Mäkinen H, Nöjd P, Saranpää P, 2003. Seasonal changes in stem radius and production of new tracheids in Norway spruce. Tree Physiol., 23(14):959-968.

Mäkinen H, Seo JW, Nöjd P, Schmitt U, Jalkanen R, 2008. Seasonal dynamics of wood formation: A comparison between pinning,microcoring and dendrometer measurements. Eur. J. Forest Res., 127(3):235-245. DOI: 10.1007/s10342-007-0199-x.

Mann ME, Jones PD, 2003. Global surface temperatures over the past two millennia. Geophysical Research Letters, 30(15): 15-18. DOI:10.1029/2003GL017814.

Mann ME, Zhang ZH, Rutherford S, Bradley RS, Hughes MK, Shindell D,Ammann C, Faluvegi G, Ni F, 2009. Global signatures and dynamical origins of the Little Ice Age and Medieval Climate Anomaly. Science,326: 1256-1260. DOI: 10.1126/science.1177303.

McLaughlin SB, Nosal M, Wullschleger SD, Sun G, 2007. Interactive effects of ozone and climate on tree growth and water use in a southern Appalachian forest in the USA. New Phytol., 174(1): 109-124. DOI:10.1111/j.1469-8137.2007.02018.x.

Oberhuber W, 2004. Influence of climate on radial growth ofPinus cembrawithin the alpine timberline ecotone. Tree Physiol., 24(3): 291-301. DOI:10.1093/treephys/24.3.291.

Riding RT, Little CHA, 1984. Anatomy and histochemistry ofAbies balsameacambial zone cells during the onset and breaking of dormancy.Canadian Journal of Botany, 62(12): 2570-2579. DOI: 10.1139/b84-349.

Rossi S, Anfodillo T, Menardi R, 2006a. Trephor: A new tool for sampling microcores from tree stems. IAWA Journal, 27(1): 89-97.

Rossi S, Deslauriers A, Anfodillo T, 2006b. Assessment of cambial activity and xylogenesis by microsampling tree species: An example at the alpine timberline. IAWA Journal, 27(4): 383-394.

Seo JW, Eckstein D, Schmitt U, 2007. The pinning method: From pinning to data preparation. Dendrochronologia, 25(2): 79-86. DOI:10.1016/j.dendro.2007.04.001.

Shashkin A, Vaganov E, 1993. simulation model of climatically determined variability of conifer’s annual increment (on the example of common pine in the steppe zone). Russian Journal of Ecology, 24(5):275-280.

Shi J, Liu Y, Vaganov E, Cai Q, Shishov V, 2005. A primary discussion on the climatic response ofPinus tabulaeformisin the Helan Mountains.Quaternary Sciences, 25(2): 245-251.

Shi J, Liu Y, Cai Q, Yi L, 2006. A case study of physiological characteristics of statistical correlation betweenPinus tabulaeformistree-ring widths and climatic factors. Acta Ecologica Sinica, 26(3): 697-705.

Shi J, Liu Y, Vaganov EA, Li J, Cai Q, 2008. Statistical and process-based modeling analyses of tree growth response to climate in semi-arid area of north central China: A case study ofPinus tabulaeformis. Journal of Geophysical Research, 113: G01026. DOI: 10.1029/2007JG000547.

Simachev I, Vaganov E, Vysotskaya L, 1992. Dendroclimatic analysis of larch growth under pollution impact of Norilsk metallurgical plant. Geogr.Prirod Resursi., 4: 65-73.

Stevens DW, 1975. A computer program for simulating cambial activity and ring growth. Tree-Ring Bulletin, 35: 49-56.

Tardif J, Flannigan M, Bergeron Y, 2001. An analysis of the daily radial activity of 7 boreal tree species, northwestern Quebec. Environ. Monit.Assess, 67(1-2): 141-160.

Vaganov E, 1990. The tracheidogram method in tree-ring analysis and its application. In: Methods of Dendrochronology: Applications in the Environmental Sciences. Kluwer Academic Publishers, The Netherlands,pp. 63-75.

Vaganov E, Sviderskaya I, Kondrateva E, 1990. Weather conditions and structure of the annual rings of trees: Simulation model of the tracheidogram. Soviet Forest Sciences, (2): 32-40.

Vaganov E, 1996a. Analysis of a seasonal growth patterns of trees and modelling in dendrochronology. In: Dean JT, Swetnam T, Meko D (eds.).Tree-Rings, Climate and Humanity. Arizona, Radiocarbon, pp. 73-87.

Vaganov E, 1996b. Recording of warming in current century by tracheids of the annual tree rings. Doklady Akademii Nauk, 351(2): 281-283.

Vaganov E, Hughes M, Kirdyanov A, Schweingruber F, Silkin P, 1999.Influence of snowfall and melt timing on tree growth in subarctic Eurasia.Nature, 400(6740): 149-151.

Vaganov EA, Hughes MK, Shashkin AV, 2006. Growth Dynamics of Conifer Tree Rings: Images of Past and Future Environments.Springer-Verlag, Berlin.

van der Werf GW, Sass-Klaassen UGW, Mohren GMJ, 2007. The impact of the 2003 summer drought on the intra-annual growth pattern of beech(Fagus sylvaticaL.) and oak (Quercus roburL.) on a dry site in the Netherlands. Dendrochronologia, 25(2): 103-112. DOI:10.1016/j.dendro.2007.03.004.

Volland-Voigt F, Bräuning A, Ganzhi O, 2009. High-resolution dendrometer measurements in a tropical mountain rainforest and a dry forest in south Ecuador. Tree Rings in Archaeology, Climatology and Ecology, 7:85-88.

Wang Y, Ma Y, Zhang Y, Lu R, Sang Y, Meng H, 2009. Response of tree-ring width ofPinus tabulaeformisto climate factors in Luoshan Mountains of Ningxia. Journal of Desert Research, 29(5): 971-976.

Wilmking M, Juday GP, Barber VA, Zald HSJ, 2004. Recent climate warming forces contrasting growth responses of white spruce at treeline in Alaska through temperature thresholds. Global Change Biol., 10(10):1724-1736. DOI: 10.1111/j.1365-2486.2004.00826.x.

Wilmking M, Juday GP, 2005. Longitudinal variation of radial growth at Alaska’s northern treeline—Recent changes and possible scenarios for the 21st century. Global and Planetary Change, 47(2-4): 282-300. DOI:10.1016/j.gloplacha.2004.10.017.

Wilson BF, Howard RA, 1968. A computer model for cambial activity.Forest Science, 14(1): 77-90.

Worbes M, 1999. Annual growth rings, rainfall-dependent growth and long-term growth patterns of tropical trees from the Caparo Forest Reserve in Venezuela. J. Ecol., 87(3): 391-403.

Wu X, 1990. Tree Rings and Climate Change. China Meteorological Press,Beijing.

Xiong W, Wang Y, Yu P, Liu H, Shi Z, Guan W, 2007. The growth in stem diameter ofLarix principis-rupprechtiiand its response to meteorological factors in the south of Liupan Mountain. Acta Ecologica Sinica, 27(2): 432-441.

Yang Y, Jiang Y, Zhang W, Wang G, 2009. Stem radial growth features ofLarix principis-rupprechtiiresponding to environmental factors at the treeline of Mt. Luya, Shanxi, China. Chin. J. Appl. Environ. Biol., 15(2):169-174.

Yin H, 2009. The Study on the Climatic Response of the Growth of Pinus koraiensis in XiaoXing’AnLing Based on the TREE-RING Eco-Physiological Model. Nanjing University of Information Science and Technology, Nanjing.

Yin H, Liu H, Guo P, Linderhollm H, 2009. An analysis on the climatic response mechanism of the growth ofPinus koraiensisin the lower mountains of Xiaoxing’anling. Acta Ecologica Sinica, 29(12):6333-6341.

Yin X, Wu X, 1995. Modelling analysis of Huashan pine growth response to climate. Quarterly Journal of Applied Meteorology, 6(3): 257-264.

Yu S, Yuan Y, Wei W, Shang H, Zhang T, Tong Z, Du M, 2008. Response of tree-ring to climate change and temperature reconstruction in west of Tianshan Mountains north slope. Journal of Desert Research, 28(5):827-832.

Yuan Y, Wei W, Esper J, Yu S, Zhang R, 2008. Influence of sampling site and detrending method on correlations and climate signals of spruce tree-ring width chronologies at upper tree line in western Tianshan Mountains of Xinjiang. Journal of Desert Research, 28(5): 809-814.

Zhang R, Yuan Y, Wei W, Yu S, Chen F, 2008. Mean minimum temperature series of April to May reconstructed by tree-ring gray chronology in northern slope of Wusun Mountains, North Xinjiang. Journal of Desert Research, 28(5): 848-854.

Zhang R, Wei W, Yuan Y, Yang Q, Yu S, Zhang T, Chang H, Chen F, Fan Z,2010. Sand-dust days constructed by tree-ring in Aksu River basin during historical period. Journal of Desert Research, 30(5): 1040-1046.

Zweifel R, Häsler R, 2000. Frost-induced reversible shrinkage of bark of mature subalpine conifers. Agr. Forest Meteorol., 102(4): 213-222.

Zweifel R, Häsler R, 2001. Dynamics of water storage in mature subalpinePicea abies: Temporal and spatial patterns of change in stem radius. Tree Physiol., 21(9): 561-569.

Zweifel R, Item H, Häsler R, 2000. Stem radius changes and their relation to stored water in stems of young Norway spruce trees. Trees-Structure and Function, 15(1): 50-57. DOI: 10.1007/s004680000072.

10.3724/SP.J.1226.2012.00175

*Correspondence to: Prof. Bao Yang, Key Laboratory of Desert and Desertification, Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences. No. 320, West Donggang Road, Lanzhou, Gansu 730000,China. Tel: +86-931-4967538; Fax: +86-931-4967488; Email: yangbao@lzb.ac.cn

September 20, 2011 Accepted: December 23, 2011