References

Measurements of branch xylem density, D x , were made for 1466 trees representing 503 species, sampled from 80 sites across the Amazon basin. Measured values ranged from 240 kg m ?3 for a Brosimum parinarioides from Tapajos in West Para, Brazil to 1130 kg m ?3 for an Aiouea sp. from Caxiuana, Central Para, Brazil. Analysis of variance showed significant differences in average D x across the sample plots as well as significant differences between families, genera and species. A partitioning of the total variance in the dataset showed that geographic location and plot accounted for 33% of the variation with species identity accounting for an additional 27%; the remaining "residual" 40% of the variance accounted for by tree to tree (within species) variation. Variations in plot means, were, however, hardly accountable at all by differences in species composition. Rather, it would seem that variations of xylem density at plot level must be explained by the effects of soils and/or climate. This conclusion is supported by the observation that the xylem density of the more widely distributed species varied systematically from plot to plot. Thus, as well as having a genetic component branch xylem density is a plastic trait that, for any given species, varies according to where the tree is growing and in a predictable manner. Exceptions to this general rule may be some pioneers belonging to Pourouma and Miconia and some species within the genera Brosimum, Rinorea and Trichillia which seem to be more constrained in terms of this plasticity than most species sampled as part of this study.


Figures 1 Introduction
The Amazon Basin remains the home to the most diverse and largest contiguous tropical forest on the planet (Malhi and Grace, 2000;Laurance et al., 2004).Different ecological systems and vegetation formations exist within its boundary creating a mosaic of forests and vegetation types with a floristic complexity the basis of which is still not well understood (Phillips et al., 2003) How and why species are distributed and where trees can grow in this mosaic of tropical forest types is a central question for tropical ecology (Pitman et al., 2001;Phillips et al., 2003).On the one hand water availability has been considered as an important factor determining tree species distributions at global (Woodward, 1987) and regional scales (Borchert, 1998;Veenendaal and Swaine, 1998;Bongers et al., 1999).Tree physiology data also shows that it may be one key factor determining species distributions in tropical rain forest despite the substantial rainfall they receive (Meinzer et al., 1999;Engelbrecht et al., 2002Engelbrecht et al., , 2005Engelbrecht et al., , 2006Engelbrecht et al., , 2007;;Tyree et al., 2003;Baltzer et al., 2008).On the other hand, although water shortage (seasonality and rainfall) is not considered by some as an important selective pressure determining the functional composition of moist and wet tropical forest (Baker et al., 2004b;ter Steege and Hammond, 2004), it clearly does help to determine overall diversity and distribution patterns of canopy tree genera across the Amazon basin (ter Steege et al., 2006).Light as a factor explaining coexistence of trees has also been considered (Poorter and Arets, 2003) and climate more than geographic distance in lowland forest seems also to affect species composition (Pitman et al., 2001) and distributions (Bongers et al., 1999;Baltzer et al., 2008).Additionally, even if there clearly is some "distance-decay" i.e. the similarity between two sites declines as the distance between them increases, see (Soininen et al., 2007) in Amazonian forest communities (Condit et al., 2002;Tuomisto et al., 2003), consistent with the view of plant species distribution as a partly random controlled process (Hubbell, 2001), it is also clear that substrate-mediated processes play a significant deterministic role in controlling species distributions across Amazonian landscapes (Gentry, 1988;Tuomisto et al., 1995Tuomisto et al., , 2003;;Phillips et al., 2003), this also being the case for other tropical forests (Harms et al., 2001;Robert et al., 2007).These theories are by no means exclusive, and one of the main purposes of the work described in this special issue (Lloyd et al., 2008a 1 ) was to interface dedicated plant physiological trait data with species abundance, climate and soil type information to Figures In what is the first of a series of papers, we here analyse in detail geographic and taxonomic patterns of tree branch wood density across Amazonia.Although wood density (D w ) is most often measured on tree boles sampled after destructive harvesting, often in association with commercial operations (Baker et al., 2004b;Chave et al., 2006), here we have used an alternative approach -the sampling of the density of the functional xylem or sapwood of small (ca.1.5 cm diameter) branches.This measurement, referred here as "xylem density", D x , is a plant trait usually considered in physiological processes studies; for example when looking at wood properties in relation to transpiration, stem water storage capacity, cavitation resistance, mechanical support, photosynthesis and/or growth (Niklas, 1997;Enquist et al., 1999;Stratton et al., 2000;Hacke et al., 2001;Meinzer, 2003;Bucci et al., 2004;Santiago et al., 2004;Jacobsen et al., 2007aJacobsen et al., , 2007b;;Pratt et al., 2007;Scholz et al., 2007).We rationalised that these are all processes that should be expected to influence where a given tree species can survive and thus should ultimately contribute to the shaping of species distributions across landscapes.
For any given species D x and D w should be related as both reflect an individual species' water transport strategy and the invariable trade-off against mechanical strength (Taneda and Tateno, 2004;Wagner et al., 1998;Asner and Goldstein, 1997).But D w may well be affected by factors in addition to those modulating D x .For example it may also reflect differences in the storage of resins and/or increased of secondary compounds of bole heartwood with time, different biomechanical requirements for maintaining vertical position and support of the whole tree and individual branches, also generally reflecting the longer term growth history of the tree.It means that, D w may be considered to effectively integrate plant ontogeny, edaphic and climatic effects over a plant's lifetime.
It has long known that D w is a genetically conserved trait, and this characteristic has been used extensively in tree breeding (Yang et al., 2001) also leading to studies of its phylogenetic inheritance (Enquist et al., 1999;Chave et al., 2006 also long been known that for a given tree species, marked variations may be observed attributable to differences in site quality and/or forest type.There have already been several reports of variations with site/growth conditions of order 10% for plantation tropical tree species (Roque, 2004;Erskine et al., 2005).Similarly, site specific differences have been noted when comparing the same species growing in different forests or site conditions (Gonzalez and Fisher, 1998;Woodcock et al., 2000;Muller-Landau, 2004;Parolin and Ferreira, 2004;Roque, 2004;Nogueira et al., 2005;Nogueira et al., 2007).Lower D w is often associated with early and secondary succession vegetation (Saldarriaga 1989;Woodcock et al., 2000) and higher growth rates (Wiemann and Williamson, 1989;Erskine et al., 2005;Nogueira et al., 2005;King et al., 2005King et al., , 2006)).
We thus hypothesised that D x is a "plastic" trait that reflects both phylogenetic heritage of trees and overall site conditions.Therefore variations in D x at the stand level across Amazonia should not only reflect differences in species composition, but differences in soil and climate conditions.If a given species can grow in different sites, D x of that species may converge to an "average" value of that of co-occurring species within each site.If this is true, species with enough phenotypic plasticity can occupy more and larger areas and become more common than species with small plasticity which will occupy restricted areas and become "rare" species at a large-scale but common in just one site.

Study sites
Eighty forest plots from across the Amazon basin were sampled normally, at the end of the rainy season of the region between January 2001 and December 2005.Additionally, two plots were sampled in Paracou, French Guiana in September 2007 (see details below).These plots form part of the RAINFOR project (www.rainfor.org;Malhi et al., 2002) and span local, regional and Basin-wide environmental gradients.of the plots have been described in detail elsewhere (Baker et al., 2004a, b;Malhi et al., 2002;Phillips et al., 2004;Vinceti, 2003) and Appendix A lists all the plots visited, including those not previously described and in some cases with updated information.

Sampling of plant material
Normally, around 20 trees greater than 10 cm dbh (diameter at breast height i.e. at 1.3 m from the base of the tree) were chosen in each plot.On some occasions, for instance when plots were unusually heterogeneous, as a consequence of topographic variations and/or shape (i.e.1000×10 m) more trees were sampled (for example Bogi plots in Ecuador).For two of the Caxiuana plots (Central Par á, Brazil) we had sampled in two consecutive years (2002 and 2003) and since there was no statistical difference in D x for the two years, we merged the data in just one set.When a plot was clearly composed of different defined landscapes, and each landscape was considered as an individual plot with often less than 20 trees sampled within it (e.g.Jacaranda Plots, Km 34 Manaus, Brazil).Trees were not chosen completely at random but taking into account two factors.
First, often there was a selection of three to six contrasting areas (slopes, valleys, gaps, flood, etc) where these were present, a professional tree climber then choose a "climbable tree" within the generally identified areas.Naturally, this "climbable tree" varied from climber to climber according to the technique employed and overall climbing skills.Nevertheless a general rule was that the "climbable tree" was >10 cm dbh with at least some neighbouring trees reachable by a clipper pole.From the climbed tree, the climber sampled one branch from each of the upper, (exposed to the sun), middle (semi light) and lower crown (shade), and one branch from the upper crown of the (usually three to five) surrounding trees.
The only difference for this sampling strategy was for two of the Guyaflux plots at Paracou, French Guiana where only lower branches from subcanopy trees were sampled.An analysis of over 200 trees comparing D x of branches from the upper, middle and lower crown of the trees using the techniques described in Lloyd et al. (2008b) 2 has, however, shown no statistically significant effect of height on D x (data not shown).
The subcanopy trees sampled at Paracou represented the distribution of the main families present in those plots with D x showing the same distribution of frequencies found for D w of 309 trees from Paracou (J.Beauchen, personal communication).In any case, because these measurements were made after the main analysis of this paper we have simply included these values to help illustrate regional patterns (Fig. 1).

Species identification
Details of the species identification from the permanent plots are described elsewhere (Baker et al., 2004b)

Xylem density determinations
A pair of stem segments of 0.05 to 0.1 m long and 0.01 to 0.02 m diameter were cut from each branch after harvesting and immediately placed in plastic bags to avoid desiccation and returned to the laboratory or field station.Normally within 12 h of sampling (but sometimes as long as 36 h later) the outer bark and phloem were removed from one of the two sample stems and its fresh volume calculated from its stem length and the average diameter of the two equiaxial diameters at each end.When the pith was wider than 2 mm diameter the stem was cut into a small segment (0.02 to 0.05 m long) and the pith removed with a small screw driver or scalpel.When the pith was thinner than 2 mm it was not removed from the stem (as it was thus assumed to be of negligible mass) but its volume subtracted from the volume of the stem without bark.Pith volume was calculated by measuring the average diameter (two measurements of diameter at each end of the stem) and stem length.All stems were then dried at 70-90 • C for three to four days (to constant mass) and weighed.Xylem density, D x , was then determined as the dry mass divided by the green volume of the sample.

Statistical analysis
Basic statistics shown in Figs.
(1), (2), and (3) were performed with Minitab 14 (Minitab Inc.).All Standard Major Axis Regression analysis (Fig. 5) were done using SMATR (Warton et al., 2006).Variance partitioning for Fig. 4 was accomplished by applying Residual Error Maximum Likelihood (REML) analysis (Gilmour et al., 1995) employing GENSTAT Discovery Edition.Mixed-effect modelling (Fig. 6) was carry out with the "lmne" package (Bates et al., 2007) and rank-based linear regression (Fig. 7) accomplished as in Terpstra and McKean (2005), both using the "R" statistical computing package (R Development Core Team, 2007).For the latter analysis, we applied the "high-breakpoint" option (Chang et al., 1999) to account for the possibility of "contaminated" data having been included in any of the D w values assimilated from a wide range of sources into the RAINFOR "wood density" database.
In order to determine the extent to which D x changes in a given species within the same plot and between plots, IPP, index of phenotypic plasticity (Valladares et al., 2000), was calculated as the absolute difference between the maximum value and the minimum value divided by the maximum value.

Geographic variation
Arithmetic mean D x for the 80 plots are shown in Fig. 1, which also shows our separation into 13 specified geographical regions and used for subsequent analysis.
From Fig. 1 a gradient of increasing D x from south and north towards the Amazon River is apparent with high D x being concentrated along the river itself.Plots located close to the Andes tend to have the lowest D x .For example, all plots in Ecuador, some in North Per ú, South Per ú and Bolivia had a relatively low D x compared to plots in Colombia and North Per ú which were all at a lower altitude and closer to the Amazon River (see Appendix A).Similar patterns of low D w in forests close to the Andes have previously been reported (Chave et al., 2006;Baker et al., 2004b;ter

2006
). Taking the basin as a whole (no division into regions), statistically significant differences existed between plot means (P <0.001) ranging from 800±50 kg m −3 (± standard deviation) at the dry experiment plot at Caixuana (Projecto Secaflor), CAX-04, with the nearby control plot CAX-03 being the second highest at 780±120 kg m −3 .These are both terra firme forests located in Central Par á, Brazil (CP-Brazil region).
The lowest plot means were for TAM-03 a swamp forests in Tambopata, South Per ú (S-Per ú & AC-Brazil region) and JAS-05 a forest growing on recently deposited river sediments (fluvisol) in Jatun Sacha in the Ecuador region.Both these plots had a mean D x of 470 kg m −3 .Data for all 80 plots are summarised in Appendix B.
Figure 2 gives means (± standard deviations) for all plots, grouped according to region, with regions being presented sequentially from top to bottom according to the overall mean D x for the trees sampled within them.This shows that, although considerable plot-to-plot variation existed within regions (e.g.N. Per ú and Colombia) large statistical differences between regions also existed (P <0.001).Of these, the highest overall value was for Central Par á in Brazil (760±130 kg m −3 , N=144) which had significantly higher D x (Tukey Test) than the rest of the regions while Ecuador had the lowest overall values (540±92 kg m −3 ).Nevertheless, this region did not differ significantly from Acre, Mato Grosso (Brazil), Colombia and Bolivia.Within some regions: PC-Brazil, PE-Brazil, N-Per ú, PW-Brazil, Colombia, S-Per ú, MT-Brazil, and Ecuador mean D x of plots varied considerably (Appendix B), while in some regions Bolivia, AC-Brazil, NE-Venezuela, SW-Venezuela, plots were not significantly different from each other.

Taxonomic variation
In a similar manner to the Region/Plot analysis above, variation in D x at the family and genera level is summarised in Fig. 3. Overall there were significant differences between the families sampled (F =8.08 DF =57P <0.001).Families with D x higher than the basin mean were Olacaceae, Celastraceae, Chrysobalanaceae, Humiriaceae, Ochnaceae, Linaceae, Scrophulariaceae, Myrtaceae, and Lecythidaceae (Fig. 3).The remaining families all contained genera characterised by both high and low D x and include some of the most abundant families across the basin: Fabaceae, Rubiaceae, Lauraceae, Sapotaceae, Apocynaceae (Fig. 3).There were also significant differences between genera (F =3.78 DF =249 P <0.0001) with the highest density genera being Aiouea, Callichlamys, Pithecellobium, Vatairea, Stachyarrhena, Dipteryx, Machaerium.

Partialling out geographical and taxonomic differences
In order to apportion the total variance in the dataset observed (Searle et al., 2006) into geographical and taxonomic components, we used Residual Maximum Likelihood (REML) Analysis to fit a model according to where µ represents the overall mean of the dataset (620 kg m −3 ); effects of location are incorporated in the term r/p, which denotes that within each region (r) are nested more than one plot (p); genetic effects are represented by the term f /g/s, which denotes that within each family (f ) are nested various genera (g), within which are nested several species (s); and (ε) represents the residual variance.All effects were taken as random variables, as we had sampled only a limited subset of plots within distinct but not comprehensive regions; also sampling a more or less random (and incomplete) selection of Amazon families, genera and species.
Results of this analysis are shown in Fig. 4. Taken together, the geographical parameters (region and plot) accounted for 33% of the total variation with 26% of this being attributable to between region variations (this effectively representing the average variation between plots in any one region).Taxonomic variability in D x , was inferred as 2015 Introduction

Conclusions References
Tables Figures

Back Close
Full 27% of the total variability in the dataset, with species and family accounting for about 10% and genus per se accounting only for 6%.Overall, the proportion of the variance in the dataset that remained unexplained was 40%, this also being the "residual variance" reflecting within tree variation (only one branch per tree was sampled), treeto-tree variation and also may reflect measurement errors.Note that in this analysis we only included species that were observed to occur in more than one site and included families and genera for which more than one species had been sampled.The analysis presented here for D x differs from others (Chave et al., 2006;Baker et al., 2004b) in that we have not taken overall means for each species; but rather included intra-specific variation and the possibility of systematic plot-to-plot variations in our interpretation.
Figure 4 suggests that geographic location is as important, if not more important than taxonomic identity in determining the value of D x observed for any given tree but with considerable variation accountable for by neither.This point is demonstrated further in Fig. 5, where we have taken the more widely abundant families, genera and species in our data set and plotted the values observed in all plots were they were sampled as a function of the average tree density of all other trees sampled in the same plot.Detailed results from this analysis (which also included a consideration of likely errors in both mean plot and mean family/genus/species values for that plot) are also given in Appendix C.
Figure 5A shows that nearly all families examined converged to a statistically significant trend for an increase in the average D x observed with a slope close to 1:1 as the values observed by other families at that site also increased.Likewise the same trend was observed when different genera were examined (Fig. 5B) and, somewhat surprisingly, also for species (Fig. 5C).Notable exceptions did however exist for the families Urticaceae C). Figure 5C shows the same trend for the species examined.Nearly half of the species examined showed significant relationships.Non significant relationships might be related to the "pioneer" character of the examined species i.e.Pourouma minor, Pourouma guianensis and perhaps to the reduced number of plots where they were collected and use for this analysis (Appendix C).

Geographical and taxonomic contributions to stand level differences
In order to evaluate the extent to which overall plot-to-plot variations might be accountable in terms of geographical versus taxonomic effects, we utilised estimates of the individual plot and species effects from Eq. ( 1) and compared them to direct stand level calculations.This was achieved by first estimating the average value for each species within each plot and then obtaining a weighted average value for D x for that plot according to the observed abundance of each species within it, denoted here as D x .
A similar calculation was done for the REML "species effects" which are plotted along with REML fixed plot effects (the r/p term from Eq. ( 1) as a function of D x in Fig. 6.This analysis shows that by far the majority of the variation in D x was accountable in terms of plot-to-plot differences, with the plot effects increasing almost linearly with D x with a slope close to 1.0.By contrast the species (i.e.f /g/s) effects were more or less constant (and close to zero) for D x > ca.550 kg m −3 , although declining slightly thereafter.We treated our plot term as a fixed effect for the analysis in Fig. 6 (as opposed to a random effect in Fig. 4), as this permitted us to allow for different plots to have different intrinsic variances which were broadly consistent with difference of plot topographic and soil heterogeneity, also removing a slight bias in residuals which was present when treating the r/p term as random.

Phenotypic plasticity
In order to determine inter and intra-specific variation we compared the IPP of the same species collected several times within one plot and in many plots.The IPP of one species collected in more than two plots (mean=0.29±0.12,N=86) was significantly higher (mean=0.14±0.10,N=86) than the variation of the same species collected more than twice within one plot (DF =1, F =77.94, P <0001).

Discussion
Our results show that there is a large variation of branch xylem density across Amazonia and that considerable plasticity occurs for many species growing in different forests.This suggests that branch xylem density may not be a simple genetically inherited trait that is predictable on the basis of the knowledge of plant taxonomy alone, and that across basin patterns may not be only explained by patterns of species composition and abundance as has been previously suggested for bole wood density (Enquist et al., 1999;Baker et al., 2004b;Chave et al., 2006).But, knowledge of the site conditions is also necessary.For example, irrespective of the genotypic level examined (Fig. 5C), D x observed varied by as much as 400 kg m −3 across sites.Moreover, this variation was systematic with different trees sampled within a given family/genus/species tending to have higher values of D x along with other trees in the same plot (and vice versa).
Additional evidence for widespread plasticity comes from the REML variance partitioning of Fig. 4 in which the combined effects of Region/Plot are shown to have contributed to about the same proportion of the overall variation in the data set as did Family/Genus/Species.This result differs significantly from those for D w (Baker et al., 2004b).
Nevertheless, when our species level means for D x are examined as a function of species mean D w using an expanded database from that presented in (Baker et al., 2004b) then there is reasonably good relationship (Fig. 7a).And the average D x for this study, 620 kg m −3 is also very similar to previous values reported for D w for Amazonia.For example (Brown and Lugo, 1984) estimated 620 kg m −3 as the average wood density of tropical America, (Chave et al., 2006)  overall species-level mean for Amazonia.
As reflected in Fig. 7b, our results do, however, differ from the above authors in that much of the variation in D w is attributed to genotype while in our case, variations in D x does include site and genetic variations.This difference is understandable, as there is a strong tendency of many species, genera and even families to be geographically confined to certain areas of the Basin (ter Steege et al., 2006).Thus, if there is equivalence between D x and D w what appears to be a genetic effect for the latter, may in fact be mostly a geographic (site and regional) effect attribute to variations in soils, climate and/or geological origin.In that respect it is only by studying replicated species growing across a wide range of environments that we have been able to show the strong environmental influence on D x (and by implication D w ).In essence the REML species effect in Fig. 7b represents the inferred value that each species would have were it to be found on some sort of "overall average site".
Also worth noting is that, in contrast to the general trend, pioneer species (Whitmore 1989) -either short and long-lived such as within the genera Pourouma, (Urticaceae) and Miconia (Melastomataceae) and some species belonging Brossimum (Moraceae), Trichillia (Meliaceae) and Myrcia (Myrtaceae) among others, often associated to secondary vegetation and/or late stages of forest succession (Banana and Tweheyo, 2001;Pe ña-Claros, 2003;Vieira et al., 2003;Poorter et al., 2006;Viera and Proctor, 2007;Selaya and Anten, 2008)., showed little tendency to exhibit variation in D x across the sites where they were found.This brings the question whether species showing little phenotypic plasticity and intermediate D x values even when present in sites with relative low or high xylem density are more restricted to specific edaphic and microclimatic conditions that sustain colonisation and fast growth i.e. gaps with enough water supply from the soil, light, and nutrients.Also, it is worth noticing that species such as But why should such plasticity in D x generally occurs?Most likely it is related to variation in resource availability and/or different site dependent soil physical characteristics and hydrological constraints to plant function.There is already abundant evidence of this from across a range of species, environments and controlled experiments (Sabatier et al., 1997;Laurance et al., 1999;Hacke et al., 2000;Vormisto et al., 2000;Pelissier et al., 2002;Baraloto and Goldberg, 2004;Kursar et al., 2005;Baraloto et al., 2007;Madelaine et al., 2007).Further evidence of the influence of site conditions on D x of trees comes from our own data.For example, in a Mangrove forest in East Par á, Brazil (EP-Brazil, BRA-01, Appendix A).There, only two species were sampled (10 individuals per species) Avicennia germinans and Rhizophora mangle which mean D x were 722±87 kg m −3 and 723±99 kg m −3 respectively.The two species are not phylogenetically related since they belong to two different families (Scrophulariaceae and Rhizophoraceae) and two different orders (Lamiales and Malpighiales).Nevertheless they converge to an almost identical D x .An analysis of the most variable species in our data set Protium paniculatum (IPP=0.66),collected in two different plots in Caxi- influence of soil texture and porosity on the hydraulic properties of Pinus taeda.
In conjunction with concurrent measurements of associated foliar physiological characteristics undertaken on leaves from the branches with which D x was determined in this study (Lloyd et al., 2008c 5 ) as well as associated soil physical and chemical characteristics (Quesada et al., 2008  6 ), further papers in this series consider the underlying factors contributing to the strong site effects in Fig. 6  West,.** not a permanent plot, samples were taken from trees around the Eddy covariance tower.Additional information of plots may be found in: (Malhi et al., 2002(Malhi et al., , 2003;;Baker et al., 2004;Vinceti, 2003).

Appendix B
Variation of D x within regions.In the first column the number below the name of the country is the mean followed by the standard deviation in parenthesis.DF=degrees of freedom; F =statistical values, P =probability, N=number of samples, SE=standard error of mean.* means "significantly different".
underlying factors controlling species distribution, composition and stand dynamics at the Basin-wide scale.
(Fig. 4A, panel M) the family containing the genus Pourouma which almost entirely consist of pioneer species (Withmore 1989) and Myrtaceae (Fig. 4A, panel U) which in our data set includes the genera Eugenia and Myrcia.Likewise at the genus level Pourouma and Brosimum also seemed to vary less than all the genera examined.The relationships for Miconia and Trichilia were also not highly significant (Appendix Introduction reported 650 kg m −3 for Central and South America together and(Baker et al., 2004b) estimated 620 kg m −3 as the Introduction Pourouma minor and P. guianensis, which are generally considered low density-fastgrowing species; were by far the species with the lowest densities in our study; xylem density varied from 410 to 690 kg m −3 , being comparable to any slow-growth old forest species in our data set.
uana (Par á, Brazil): CAX-02 (D x =280 kg m −3 ) and CAX-04 (D x =800 kg m −3 ), and Perebea tessmannii (IPP=0.58)two samples collected in the same plot but 900 m apart in BOG-01 with 320 and 760 kg m −3 , in both cases low pH, high proportion of sand and low minerals concentrations were be associated with the trees exhibiting high D x(Que- sada et al., 2008  3 ; Pati ño et al., 2008a  4 ).Hacke et al. (2000) have shown remarkable (Pati ño et al., 2008a  4 ), along with the extent to which these site dependent variations in xylem density reflect a co-ordinated physiological response of different Amazonian tree species to variations in their growth environment (Pati ño and Lloyd, 2008b 7 ).

Fig. 2 .Fig. 3 .
Fig. 2. D x distribution across the Amazon basin.Plots and Regions are indicated in the right and left axes respectively.Horizontal lines represent the standard deviation.Vertical straight lines represent the Tukey test in which regions joined by the line are not significantly different while regions not included within a line are significant different.

Fig. 4 .Fig. 6 .
Fig. 4. Apportion of the total variance of D x in the data set.
).Nevertheless, it has Figures Many Figures and in this work we have used the new classification given by the Angiosperm Phylogeny Group II (APG 2003, http://www.mobot.org/MOBOT/Research/APweb/), in which Bombacaceae, Sterculiaceae, and Tiliaceae are all included in the Malvaceae; Papillionaceae, Caesalpinaceae, and Mimosaceae are included in the Fabaceae; Cecropiaceae in the Urticaceae; and Flacoutiaceae in the Salicaceae.

Table B1 .
D x distribution across the Amazon basin.Plots and Regions are indicated in the right and left axes respectively.Horizontal lines represent the standard deviation.Vertical straight lines represent the Tukey test in which regions joined by the line are not significantly different while regions not included within a line are significant different.