factors controlling the productivity of tropical andean

Biogeosciences 18 1525ndash1541 2021httpsdoiorg105194bg-18-1525-2021copy Author(s) 2021 This work is distributed underthe Creative Commons Attribution 40 License

Factors controlling the productivity of tropical Andean forestsclimate and soil are more important than tree diversityJuumlrgen Homeier12 and Christoph Leuschner12

1Plant Ecology and Ecosystems Research University of GoumlttingenUntere Karspuumlle 2 37073 Goumlttingen Germany2Centre for Biodiversity and Sustainable Land Use University of GoumlttingenUntere Karspuumlle 2 37073 Goumlttingen Germany

Correspondence Juumlrgen Homeier (jhomeiegwdgde)

Received 14 September 2020 ndash Discussion started 28 September 2020Revised 21 January 2021 ndash Accepted 22 January 2021 ndash Published 3 March 2021

Abstract Theory predicts positive effects of species rich-ness on the productivity of plant communities through com-plementary resource use and facilitative interactions betweenspecies Results from manipulative experiments with tropicaltree species indicate a positive diversityndashproductivity rela-tionship (DPR) but the existing evidence from natural forestsis scarce and contradictory We studied forest abovegroundproductivity in more than 80 humid tropical montane old-growth forests in two highly diverse Andean regions withlarge geological and topographic heterogeneity and relatedproductivity to tree diversity and climatic edaphic and standstructural factors with a likely influence on productivityMain determinants of wood production in the perhumid studyregions were elevation (as a proxy for temperature) soil nu-trient (N P and base cation) availability and forest structuralparameters (wood specific gravity aboveground biomass)Tree diversity had only a small positive influence on produc-tivity even though tree species numbers varied largely (6ndash27 species per 004 ha) We conclude that the productivity ofhighly diverse Neotropical montane forests is primarily con-trolled by thermal and edaphic factors and stand structuralproperties while tree diversity is of minor importance

1 Introduction

Research into the diversityndashproductivity relationship (DPR)has recently shifted its focus on forests because of their im-portance for the global carbon cycle and the high biodiver-sity especially of tropical forests (Mori et al 2017 Fei et

al 2018) Results from mixed species plantations with trop-ical trees indicate a positive DPR in the majority of cases(Sapijanskas et al 2014 Huang et al 2018 Schnabel et al2019) suggesting complementary resource use and facilita-tive interactions between different tree species (McIntire andFajardo 2014 Chen et al 2016) The results of these experi-ments may provide valuable insights into the mechanisms ofdiversity effects on productivity (Tuck et al 2016 Schnabelet al 2019) but they are difficult to extrapolate to tropicalold-growth forests due to the young age and cohort structureof these tree plantations Observational studies on the linkagebetween tree diversity and productivity in natural or near-natural tropical forests are usually burdened with additionalconfounding factors and thus are more difficult to interpretthan experiments but they have the advantage of consideringnatural systems The few existing observational studies onthe DPR in tropical forests have produced contradictory evi-dence On Mt Kinabalu Malaysia aboveground net primaryproduction (NPP) was higher in more diverse tropical mon-tane forest plots (Aiba et al 2005) but no effects of speciesrichness on productivity were found in logged tropical low-land forests in Guyana (van der Sande et al 2018) Chisholmet al (2013) analyzed the tree diversityndashproductivity relationin 25 permanent forest plots from temperate to tropical re-gions and found a positive diversity effect when plot sizewas small (004 ha) but the effect disappeared at larger spa-tial scales (025 and 1 ha) In a global analysis of forest pro-ductivity data Liang et al (2016) found that positive DPRspredominate but negative relations do also exist though theyare restricted to a few of the tested data sets

Published by Copernicus Publications on behalf of the European Geosciences Union

1526 J Homeier and C Leuschner Productivity of tropical Andean forests

The existing observational data from tropical forests sug-gest that (i) tree diversity may enhance productivity undercertain conditions and (ii) a positive DPR might be moreprominent in small plots To date it is not clear under whichabiotic conditions and stand structural settings a positiveDPR does emerge and how important tree diversity is com-pared to abiotic and other biotic determinants of productivity

The search for abiotic determinants of tropical-forest pro-ductivity has mainly concentrated on soil nutrient availabil-ity solar irradiance and precipitation revealing clear depen-dencies in studies along environmental gradients (Schuur2003 Girardin et al 2014 Malhi et al 2017) Many tropi-cal forests especially those in the lowlands seem to be P-limited (Aragatildeo et al 2009 Quesada et al 2012) whileN limitation may be more influential in tropical montaneforests (TMFs) (Tanner et al 1998 Moser et al 2008Homeier et al 2012 Fisher et al 2013) Fertilization ex-periments and studies on foliar nutrient levels across envi-ronmental gradients suggest that the NPP of tropical foreststends to increase with soil fertility notably N and P avail-ability but locally also with base cation supply (Kitayamaand Aiba 2002 Cleveland et al 2011 Malhi et al 2004Homeier et al 2010 Unger et al 2012 Fisher et al 2013Banin et al 2014 Hofhansel et al 2015) Bruijnzeel andVeneklaas (1998) assumed that the productivity of TMFsis primarily limited by cloudiness ie low solar radiationwhich is supported by a modeling study (Fyllas et al 2017)Precipitation should play a minor role in TMFs (Zimmer-mann et al 2010) Less clear is the role of temperature Theelevational temperature decrease can act on plants in multi-ple ways directly through a limitation of photosynthetic ac-tivity and of carbon water and nutrient cycling in the plantand it may influence stem leaf and root morphology Theimpairment of nutrient supply by low soil temperatures isan indirect effect Yet plants can adapt to lower tempera-tures thereby partly reducing these limitations (Lambers andOliveira 2019) The majority of studies along elevation tran-sects found an NPP decline with decreasing temperature intropical mountains (Kitayama and Aiba 2002 Girardin etal 2010 Cleveland et al 2011 Leuschner et al 2013van de Weg et al 2014) However an analysis of a globaldatabase failed to show a temperature dependence of NPP intropical forests (Luyssaert et al 2007) This fits to modelingresults suggesting that trait variation with elevational speciesturnover in TMFs may partly capture the effect of tempera-ture on tree metabolism and growth (Fyllas et al 2017)

Biotic controls related to stand structural properties no-tably stem density basal area and wood density and thefunctional composition of the community can also affectforest productivity Some studies found that forest NPP in-creased with standing aboveground biomass (AGB) (Keelingand Phillips 2007 Pan et al 2013 Lohbeck et al 2015) butthe relation does not seem to be very close since woody tissueresidence time ie the ratio between AGB and abovegroundproductivity varies considerably with soil fertility and cli-

mate in tropical moist forests (Malhi 2012 Quesada et al2012 Chisholm et al 2013 Malhi et al 2017) Variation inleaf area index and leaf functional traits such as LMA (leafmass per area) and foliar N and P content should be majordeterminants of net primary productivity (van de Weg et al2014 Wittich et al 2012 Finegan et al 2015) as they in-fluence the photosynthetic capacity of the canopy and radi-ation interception Indeed the modeling study of Fyllas etal (2017) indicated for tropical forests along an Andean el-evation gradient that the influence of LMA foliar N and Pcontent and wood density of the species was a more impor-tant determinant of NPP than temperature which is in linewith the findings of Luyssaert et al (2007) From the partlycontradicting results we conclude that a multi-factorial ap-proach is needed to understand the determinants of tropical-forest productivity Yet most of the existing studies have in-vestigated only one or two of the mentioned factors

Tropical montane forests cover all humid tropical moun-tains between sim 1000 and 3000 m asl with the widest dis-tribution in the Andes and Central American mountains andmore restricted occurrence in the mountains of the Pale-otropis They differ from tropical lowland forest in terms oflower canopy heights higher stem densities and lower AGBTMFs play important roles in the provision of water for hu-man populations and as stores of carbon and tropical moun-tains are known as hotspots of biodiversity (Ashton 2003Bruijnzeel et al 2010 Spracklen and Righelato 2014 Fa-hey et al 2015 Rahbeck et al 2019) TMFs occur alongsteep thermal and edaphic gradients making them attrac-tive for the study of abiotic controls of tropical-forest pro-ductivity and the role of tree diversity In this study in twoAndean elevation transects we combine biomass productiv-ity and tree diversity data with comprehensive soil chem-ical data to identify the most important abiotic and bioticdrivers of forest productivity in Neotropical montane forests(Fig 1) In permanent plots in old-growth forests we mea-sured coarse-wood production (WP) and fine-litter produc-tion (only one of the transects) and related these componentsof aboveground primary production to the availability of thefive macronutrients N P Ca K and Mg in the soil to el-evation as a proxy for temperature and to stand structuralproperties and tree diversity in the plots We chose a plot sizeof 004 ha (20 mtimes 20 m) in order to meet plot homogene-ity requirements in the rugged terrain Moreover assumeddiversity effects are more likely to emerge at small spatialscales (Chisholm et al 2013) This allowed us to investi-gate a larger number of plots across extended environmentalgradients (2000 m of elevation distance or sim 10 K variationin mean annual temperature large bedrock topographic andsoil variation) and to encounter considerable variation in treediversity (6 to 27 species per 004 ha) The analyses were re-stricted to old-growth forests with closed canopies in orderto exclude variation in stand structure and species composi-tion related to successional processes We employed struc-tural equation modeling for exploring mutual interrelation-

Biogeosciences 18 1525ndash1541 2021 httpsdoiorg105194bg-18-1525-2021

J Homeier and C Leuschner Productivity of tropical Andean forests 1527

Figure 1 Conceptual model of plausible interaction pathways be-tween environment (elevation soil) and forest structure as driversof productivity Arrows drawn from the two boxes ldquostand proper-tiesrdquo and ldquosoil propertiesrdquo indicate that pathways from the respec-tive variables (forest stand properties wood specific gravity stemdensity leaf area index (LAI) soil properties first two principalcomponent analysis (PCA) axes) were included in the models

ships between the likely abiotic and biotic drivers of produc-tivity

Based on the assumption that precipitation is not a growth-limiting factor in these humid to perhumid study regions wehypothesized in accordance with the existing literature that(i) temperature is the principal productivity-determining fac-tor in the study regions (ii) soil nutrient availability is an-other ndash but secondary ndash influential abiotic factor acting inTMFs mainly through N availability and (iii) tree diversityis a less important driver of productivity than abiotic factors

2 Methods

21 Study transects

The study was conducted in permanent plots in old-growthmontane forests established along two elevation transects onthe eastern slope of the Ecuadorian Andes in the provincesNapo (Napo transect) Loja and Zamora-Chinchipe (Lojatransect) (Fig A1) The chosen plot area was 20 mtimes 20 m(400 m2) a size that has previously been used in surveysof tropical-forest diversity because the area is large enoughto give representative information on stand structure andspecies composition while the patch size in tropical mon-tane old-growth forests is small enough to guarantee suffi-cient stand structural and edaphic homogeneity on the plotlevel All plots were placed in forest patches without signsof recent disturbance and with sufficient distance to specialmicrohabitats such as ravines and ridges By selecting small-sized plots solely in old-growth portions of the forest weattempted to reduce the variation in structural and biologi-cal parameters resulting from the mosaic of different succes-sional stages typically existing in natural forests

The climate in both transect regions is perhumid through-out the year with a mean annual precipitation gt 2000 mm

at all elevations and lack of a distinct dry season (Bendixet al 2008 Salazar et al 2015) All study sites are char-acterized by aridity index values gt 074 after Trabucco andZomer (2009)

Soil properties vary with elevation and differ between thetwo transects The Loja transect is characterized by relativelynutrient-poor soils mostly on metamorphic schists and sand-stones with a slightly better nutrient supply at lower ele-vations and in valleys and less favorable supply of N andP on upper slopes and at higher elevations (Wolf et al2011 Werner and Homeier 2015) In contrast the Napotransect has more fertile soils which developed on volcanicdeposits or limestone In this transect the N mineraliza-tion rate decreases with elevation while plant-available Pincreases (Unger et al 2010) The availability of the fiveplant macronutrients N P Ca K and Mg was analyzed inthe soil of all study plots together with soil acidity in orderto cover the physiologically most meaningful soil chemicalproperties For characterizing N availability incubation ex-periments to determine the net release of NH+4 and NOminus3through mineralization and nitrification (Nmin) were con-ducted in the topsoil In addition the bulk soil CN ratiowas determined in the topsoil for characterizing the decom-posability of soil organic matter and thus gives an indepen-dent measure of potential N supply (Pastor et al 1984) Theincrease in organic-layer depth with elevation in both tran-sects accompanied by greater CN ratios and lower pH isrelated to higher soil-organic-matter content and indicates inour study areas a decrease in nutrient availability as a resultof reduced organic-matter turnover at lower temperatures Pavailability was estimated as resin-exchangeable P ie theexchange of phosphate ions by anion exchange resins whichmay give a minimum estimate of plant-available P (Pav) Theplant availability of Ca K and Mg (CaKMgex) and also of po-tentially toxic Al3+ was quantified by salt exchange (02 NBaCl2 solution) applying a standard protocol for the chem-ical analysis of forest soils (for analytical details see Ungeret al 2010 Wolf et al 2011) Soil pH was measured in soilsuspended in 1 M KCl Since most fine roots are concentratedin the topsoil (Soethe et al 2006) where the vegetation ef-fect on the soil is also largest the statistical analyses focus onthe chemistry of the 0ndash10 cm layer and its importance for thevegetation (see Supplement S2 for the soil chemical proper-ties of the plots)

22 Plot inventory and forest productivity

The Loja transect consists of 54 plots that are equally spreadover three chosen elevation levels (18 plots each at sim 1000sim 2000 and sim 3000 m asl) (see Supplement Sect S1 forthe stand properties of the plots in both transects) From theNapo transect we included data from 66 plots that were lo-cated in a comparable elevation range (960ndash3100 m asl) Inboth transects all trees with a diameter at breast height (dbh)ge 10 cm in the 400 m2 plot were recorded with their dbh and

httpsdoiorg105194bg-18-1525-2021 Biogeosciences 18 1525ndash1541 2021


identified according to species From unknown tree speciesspecimens were collected for later identification at Ecuado-rian herbaria (HUTPL LOJA QCA QCNE) or by interna-tional specialists for difficult groups

Data on AGB annual AGB increment (coarse-wood pro-duction WP) and WSG (wood specific gravity) of all treesge 10 cm dbh in the plots were available from earlier stud-ies (Loja transect Wolf et al 2011 Leuschner et al 2013Napo transect Unger et al 2012 Kessler et al 2014) AGBand WP of all trees ge 10 cm dbh in a plot were estimated asthe sum of the biomass of all tree individuals and the corre-sponding increment of these trees in a second stem diameterinventory after 1ndash5 years For quantifying AGB we appliedthe allometric equation of Chave et al (2005) for tropicalwet forests WP was calculated as the difference between thetwo AGB estimates on an annual basis For the Loja tran-sect we additionally calculated net aboveground productivity(NPPa) as the sum of WP and annual fine-litter productionthat was recorded in litter traps (data available from Wallis etal 2019)

Leaf area index (LAI) was estimated by synchronous mea-surement with two LAI 2000 plant canopy analyzers (LI-COR Inc Lincoln NE USA) that were operated in theremote mode during periods of overcast sky Simultaneousreadings were taken below the canopy at 2 m height abovethe ground and in nearby open areas (ldquoabove-canopyrdquo read-ing) to record incoming radiation Data for the Napo transectwere taken from Unger et al (2013) For the Loja transectthe LAI measurements were conducted in 2011 adopting themethods described in detail in Unger et al (2013)

23 Data analysis

Tree diversity was calculated according to the individual-based rarefaction method (Gotelli and Colwell 2001) as thenumber of species (stems with dbh ge 10 cm) expected in arandom sample of 14 trees in a 400 m2 plot (14 being thesmallest number of tree individuals recorded in our plots)Linear regression analyses were applied to identify signif-icant relationships between AGB WP tree diversity LAIstem density WSG and soil properties as dependent variablesand elevation (as a proxy for temperature) as an independentvariable

We used principal component analysis (PCA) to reducethe number of soil variables and to ensure that the subse-quent analyses were not affected by the problem of multi-collinearity We conducted two PCAs using the R packageldquopcaMethodsrdquo one with the merged data from both transectsand another one for the Loja transect The following 10 soilvariables were included in the PCAs organic-layer depthmineral soil pH (KCl) exchangeable K Mg Ca and Al con-tent resin-exchangeable P content CN ratio and topsoilN mineralization and nitrification rate The plot scores onthe first two resulting principal components (PC 1 and PC2) were included in the subsequent analyses

Structural equation modeling was used for identifying thedirect influence of abiotic factors (elevation soil principalcomponents) on forest stand properties and tree diversity andto assess the direct and indirect influences of these factorson productivity For each initial structural equation model(SEM) we included the two principal soil components (PC1and PC2) from the respective PCA and the factors elevationWSG and tree diversity to predict AGB and stand productiv-ity (WP or NPPa) WSG was selected from the stand proper-ties as it showed a stronger correlation to stand productivitythan the other measured variables (stem density and LAI)

Based on the existing knowledge about the dependency ofproductivity and AGB on environmental and stand proper-ties and assumed diversity effects on productivity we devel-oped an initial model of interaction paths between the envi-ronmental parameters (elevation soil PC 1 and PC 2) WSGand tree diversity and the target variables (AGB and WP orNPPa) (Fig 1) We then fitted two different SEMs one forthe complete set of 83 plots from both transects (where allinformation was available 54 Loja plots and 29 Napo plots)and another one for the Loja transect (54 plots) where NPPa(and not only WP) data are available

IBM SPSS AMOS 24 software (Arbuckle 2016) was usedto fit the normalized data to the hypothesized path model andto determine path coefficients using the maximum likelihoodmethod We assessed the goodness of model fit using the χ2

value the associated p value Akaike information criterion(AIC) RMSEA (root mean square error of approximation)and CFI (comparative fit index) Since we used SEM in anexplorative way the original model has been subject to mod-ification We iteratively removed insignificant paths to testwhether incorporation of those paths in the model signifi-cantly increased the χ2 value and CFI and reduced the AICand the RMSEA of the model For all dependent variableswe calculated R2 values that indicate the proportion of vari-ance explained by the model

3 Results

31 Patterns of tree diversity and productivity

The two transects differed considerably with respect to floris-tic composition The most important tree families (in orderof abundance) were Lauraceae Melastomataceae Cunoni-aceae Moraceae and Clusiaceae in the Loja transect andEuphorbiaceae Fabaceae Lauraceae Moraceae and Rubi-aceae in the Napo transect The number of stems per plotvaried between 14 and 61 and the number of tree speciesbetween 6 and 27 (Supplement Sect S1) Tree species rich-ness showed no significant decline with elevation (Fig 2a)AGB decreased on average by 50ndash90 Mg haminus1 per kilometerelevation in both transects (Fig 2b) Biomass was between200 Mg haminus1 (lower elevations) and 100 Mg haminus1 (upper ele-



vations) higher in the Napo transect on more fertile soil thanin the Loja transect

The AGB decrease was linked to a decline in WP be-tween 1000 and 3000 m by about 13ndash15 Mg haminus1 yrminus1 perkilometer elevation with a generally lower productivity (byabout 10ndash15 Mg haminus1 yrminus1) in the Loja transect comparedto the Napo transect (Fig 2c) LAI decreased with eleva-tion in both transects by about 1 m2 mminus2 per kilometer el-evation with slightly higher LAI values in the Loja transect(Fig 2d) WSG showed no systematic variation with eleva-tion (Fig 2e) The expected stem density increase with ele-vation was only observed in the Loja transect (Fig 2f)

32 Identification of principal soil factors

In the two data sets (both transects merged Loja transectonly) the first two PCA axes explained 63 and 61 re-spectively of the variation in soil factors across all plots Inthe merged data set (Figs 3 and B1 and Table B1) the firstaxis represented mainly the basic cations Mg Ca and K soilpH and organic-layer depth the second axis stood for N sup-ply and Pav In the PCA of the Loja transect (Fig B2 and Ta-ble B2) soil parameters were represented similarly but Pavhad a higher loading on the first axis In general the soilsof the Napo transect were characterized by higher fertility asindicated by on average markedly higher topsoil net N miner-alization rates and higher P availability in the mineral topsoil(see Supplement Sect S2)

33 Factors determining wood production in themerged data set

In the merged data set (83 plots Fig 4 Table 1) above-ground tree biomass (AGB) was positively influenced by soilfertility (standardized total effects of PC1 and PC2 of 030and 045 respectively) and tree diversity (023) The threemost important determinants of WP were elevation (nega-tive influence standardized direct effect of minus058) soil PC2(029) and WSG (minus025)

The direct negative effect of elevation on WP was promi-nent and this influence was enhanced through indirect re-lationships via negative effects of elevation on soil proper-ties and tree diversity (standardized indirect effect ofminus018)However soil factors were also important drivers of woodproduction either directly (in case of PC2 which was relatedto N mineralization 029) or indirectly through a soil N ef-fect (PC2) on AGB and diversity (together 010) and effectsof base cation availability and pH (PC1) on WSG and AGBTree diversity itself had only a relatively weak indirect posi-tive effect on WP through its influence on AGB (standardizedtotal effect of 005)

34 Factors determining NPPa in the Loja transect

The model for aboveground net primary production (ie WPplus fine-litter production) built only with the Loja transect

data identified elevation and WSG as the main direct de-terminants of NPPa (both effects negative standardized di-rect effects of minus067 and minus039 respectively Fig 5 Ta-ble 2) Soil properties were also influential (mainly PC1 withN availability) but only indirectly through a negative effecton WSG and a positive effect on diversity (standardized totaleffect of 030) Tree diversity had a significant positive butrelatively weak direct effect on NPPa (018) which was en-hanced by an indirect influence via AGB (standardized totaleffect of 027)

4 Discussion

In a global perspective forest productivity increases withtemperature in cooler regions and with precipitation indrier regions (Reich and Bolstad 2001 Schuur 2003)In the absence of temperature and precipitation measure-ments and soil moisture data from the plots we used ele-vation as a proxy for the thermal conditions We further as-sumed that soil moisture rarely constrains forest productiv-ity in both study regions as precipitation totals range wellabove 2000 mm yrminus1 throughout the Napo and Loja tran-sects (Bendix et al 2008 Salazar et al 2015) and soilmoisture is usually high in Andean TMFs (Zimmermann etal 2010) Our assumption is supported by the highly sig-nificant negative influence of elevation on WP suggestingthat the temperature decrease with elevation was physiolog-ically much more important than a possible water shortageeffect on productivity Both the SEMs and correlation anal-yses show that tropical-forest productivity largely decreases(by 13ndash15 Mg haminus1 yrminus1 kmminus1) with a concurrent decreasein mean annual temperature from sim 20 to sim 10 C in the twotransects confirming earlier observations (Aiba et al 2005Cleveland et al 2011) Reduced temperatures may nega-tively affect tree growth through a variety of direct and indi-rect influences on carbon gain carbon allocation and nutrientacquisition Photosynthesis measurements revealed commu-nity means of light-saturated net photosynthesis rates (Amax)that were 36 and 18 lower at 3000 m compared to thecommunity means at 2000 and 1000 m respectively (Wittichet al 2012) Reduced canopy carbon gain in combinationwith the elevational LAI decrease by 1 m2 mminus2 kmminus1 mightlargely explain the productivity decline from 1000 to 3000 min the Ecuadorian Andes (Leuschner et al 2013) In apparentcontradiction the tree-based carbon balance model of Fyllaset al (2017) suggests for the Peruvian Andes that the ap-parent temperature effect on photosynthesis and productivityin the 3300 m elevation gradient is manifested through leaftrait variation associated with the elevational species turnoverrather than through a temperature dependency of photosyn-thesis itself In fact light-saturated net photosynthesis (Amax)did not change between 250 and 3500 m asl in this transect(Fyllas et al 2017) These results contradict a pantropicalanalysis (ca 170 tree species 18 sites) showing that Amax



Figure 2 Variation in tree diversity (a) aboveground biomass (b) wood production (c) LAI (d) wood specific gravity (e) and stem density(f) along the two studied forest transects (red dots Loja transect blue dots Napo transect) The number of plots available for a givenparameter and transect varies as not all parameters were measured in all plots (a Loja 54 plots Napo 64 plots b 54 66 c 54 29 d 5360 e 54 66 f 54 66) Data presented in (b) (c) (e) and (f) were compiled from Unger et al (2012) Leuschner et al (2013) and Kessler etal (2014) data from the Napo transect in (d) are from Unger et al (2013)

decreases on average by 13 micromol mminus2 sminus1 per kilometer el-evation in tropical mountains (Wittich et al 2012) We as-sume that local differences in soil fertility and climatic fac-tors such as cloudiness or very high precipitation causingtemporal soil anoxia may have a large influence on eleva-tional patterns in foliar nutrient content and photosyntheticcapacity Moreover the Amax study of Wittich et al (2012)with 40 Andean tree species from 21 families points at a con-siderable influence of phylogeny on photosynthetic capacitySpecies turnover between elevations and also between dif-ferent locations might therefore explain some of the differ-ences in photosynthetic carbon gain and productivity inde-pendently of temperature differences

Low-temperature effects on nutrient supply and acqui-sition processes which enhance nutrient deficiency couldwell be additional factors causing productivity reductionsand increased belowground C allocation at higher elevations(Graefe et al 2008 Moser et al 2011 Leuschner et al2013) Indeed pronounced leaf trait changes were observedin the Loja transect from larger thinner and N-richer leavesat 1000 m to smaller thicker and N-poorer leaves at 3000 m(Homeier et al unpublished) suggesting that temperaturemay act on the forest carbon cycle partly through N and Pavailability as an environmental filter sorting tree speciesaccording to their leaf traits from more acquisitive to moreconservative at higher elevations



Figure 3 Variation in soil properties along the two forest transects (red dots Loja transect blue dots Napo transect) Shown are the (a) firstand (b) second principal components of the soil PCA (see Fig B1 and Table B1)

Figure 4 Final model for wood production (Loja transect and Napotransect 83 plots) Structural equation model (chi-square= 849 degrees of freedom p= 050 AIC= 604 RMSEAlt001CFI= 100) with standardized path coefficients The size of the ar-rows is proportional to the strength of the paths their significance isindicated by asterisks (lowast Plt005 lowastlowast Plt001 lowastlowastlowast Plt0001) Bluearrows indicate positive and red negative estimates AGB above-ground biomass WSG wood specific gravity PC1 PC2 first twoaxes of the soil PCA (Fig B1 and Table B1)

Temporal water limitation if it were relevant in the tworegions should be more prominent at lower elevations withlower rainfall which should lead to a positive and not a neg-ative elevation effect on productivity We cannot excludehowever that high soil moisture in combination with tempo-ral hypoxia in the rhizosphere of the high-elevation stands iscontributing to the productivity decline toward the timberline(Moser et al 2011)

Other climatic factors with a potential impact on produc-tivity such as solar radiation air humidity and leaf wetnesswere not considered in our analysis due to lack of suitable

Table 1 Standardized direct indirect and total effects of elevationvarious stand structural parameters and soil chemical parameterson coarse-wood production (WP) and aboveground biomass (AGB)according to the SEM analysis in Fig 4 Standardized path coeffi-cients are shown

Factors Direct Indirect Total

WP

Elevation minus0397 minus0183 minus0580PC1 0163 0163PC2 0285 0099 0385AGB 0222 0222WSG minus0245 minus0245Tree diversity 0051 0051

AGB

Elevation minus0302 minus0302PC1 0295 0295PC2 0447 0447Tree diversity 0230 0230

local data Solar radiation typically reaches a minimum atmontane or upper montane elevation due to cloud immersionwhich can reduce carbon gain (Malhi et al 2017) Includingthese factors in the analysis may lead to somewhat differentconclusions on the key determinants of forest productivity(Finegan et al 2015 Fyllas et al 2017 Malhi et al 2017)

Our study in two regions of the northern Andes is the firstworldwide to analyze the dependence of tropical-montane-forest production on the availability of all five quantitativelymost important plant nutrients (N P Ca K Mg) P availabil-ity has been reported to limit or co-limit forest productivityat many tropical lowland sites (Tanner et al 1998 Paoli andCurran 2007 Cleveland et al 2011 Wright et al 2011Dalling et al 2016b) whereas N seems to limit tropical-forest productivity more often at higher elevations (Tanner



Figure 5 Final model for NPPa (Loja transect 54 plots) Structuralequation model (chi-square= 99 10 degrees of freedom p= 045AIC= 599 RMSEAlt001 CFI= 099) with standardized pathcoefficients The size of the arrows is proportional to the strengthof the paths their significance is indicated by asterisks (lowast Plt005lowastlowast Plt001 lowastlowastlowast Plt0001) Blue arrows indicate positive and rednegative estimates AGB aboveground biomass WSG wood spe-cific gravity PC1 PC2 first two axes of the soil PCA (see Fig B2and Table B2)

Table 2 Standardized direct indirect and total effects of elevationvarious stand structural parameters and soil chemical parameters onNPPa and aboveground biomass (AGB) (Loja transect) accordingto the SEM analysis in Fig 5 Standardized path coefficients areshown


NPPa

Elevation minus0393 minus0272 minus0665PC1 0304 0304PC2 0042 0042AGB 0253 0253WSG minus0391 minus0391Tree diversity 0178 0088 0266

AGB

Elevation minus0447 minus060 minus0506PC1 0126 0126PC2Tree diversity 0346 0346

et al 1998 Benner et al 2010 Wolf et al 2011 Dalling etal 2016a) The prominent direct and indirect effects of PC1and PC2 on productivity suggest that all three nutrient com-ponents (N P and base cations) play roles in the limitation offorest productivity in the study regions

A high net N mineralization rate positively influenced WPin the merged data set but the indirect effect through AGBdominated The large difference in N mineralization rates be-

tween the two transects with high values in the predomi-nantly volcanic soils in the Napo region but low values in thenon-volcanic soils of the Loja region probably also explainsthe considerably higher AGB and consequently WP in theNapo transect as compared to the Loja transect (Fig 2) Thetrees in the less fertile Loja transect most likely allocate ahigher proportion of their photosynthates to root productionand to root symbionts (Vicca et al 2012 Doughty et al2018) which must reduce wood production The availabil-ity of base cations and soil acidity influenced productivityin both models only indirectly mainly through an effect onWSG (higher wood density on acidic base-poor soils)

The PCAs also showed that the three main nutrient cate-gories (P N base cations) varied more or less independentlyfrom each other across the plots Moreover their relative im-portance varied between the Loja and Napo data sets sug-gesting that soil nutrient limitation of forest productivity de-pends more on region bedrock type and soil age than on ele-vation The different paths from soil nutrients to productivityand the weak correlations among Pav Nmin and base cationsacross our data set imply that no single nutrient element canserve as a good indicator of overall nutrient availability orsoil fertility in the study regions This is also valid for pHwhich showed the tightest relation to CaKMgex (Loja tran-sect) or organic-layer depth (merged transects) but weaker orno association to Pav Nmin and soil CN ratio Soil fertil-ity assessments in Andean montane forests should considerN P and also basic cations when forest productivity shall berelated to soil fertility The widely held assumption that netprimary production in tropical lowland forests is primarilylimited by low P availability and TMF productivity predom-inantly by low N availability (Vitousek and Sanford 1986Tanner et al 1998 Fisher et al 2013) may thus requireadaptation

Various studies have identified WSG as a good predictorof the diameter increment of tropical trees and high WSGhas been linked to low aboveground productivity in tropi-cal lowland forests (Malhi et al 2004 Poorter et al 2008Finegan et al 2015) In the Ecuadorian Andes the nega-tive effect of average WSG on productivity was as strong asthe positive biomass (AGB) influence suggesting that nega-tive edaphic effects on productivity notably high soil acidityand low availability of basic cations are exerted in part indi-rectly through an increase in wood specific gravity (Unger etal 2012) A recent study in Southeast Asian tropical forestsshowed that WSG has a direct effect neither on biomassproduction (Kotowska et al 2021) nor on tree water con-sumption even though harder wood tends to be associatedwith lower growth rates (Muller-Landau 2004 Hoeber et al2014) Yet WSG is known to be associated with most struc-tural and functional wood properties (Chave et al 2009)and it may also be related to anatomical attributes such aspit membrane characteristics that influence sap flux densitywhich itself is positively related to productivity (Kotowskaet al 2021) In addition our data support a causal link from



higher N availability through lower WSG to higher produc-tivity which is suggested by N fertilization experiments inconifers (Jozsa and Brix 1989) and a negative correlation be-tween wood density and soil fertility in continent-wide woodproperty analyses (Chave et al 2009) N fertilization exper-iments in Andean TMFs support the positive effect of N onproductivity (Homeier et al 2012 Fisher et al 2013) butthey were not designed to prove a WSG-mediated mecha-nism

Unexpected is the insignificant effect of LAI on WP inboth transects visible in the exclusion of leaf area fromthe models This is difficult to reconcile with the fact thatcanopy carbon gain depends on the amount of interceptedlight and that the LAIndashproductivity relationship is usuallytight in forests (Reich 2012) A possible explanation couldbe that LAI measurements with the LAI2000 canopy ana-lyzer are known to lead to both underestimation where leafclumping is dominant and overestimation where stem den-sity is high This may have blurred LAI differences betweenstands differing in productivity

Tree diversity apparently has only a small influence onproductivity in the studied tropical montane forests eventhough tree species numbers varied largely across the plots(6ndash27 species per 004 ha) and plot size was small so thatthe effect should be more visible A positive diversity effectapparently was confined to the component fine-litter produc-tion (in the Loja transect) as the effect became more visiblewhen NPPa (which includes fine-litter production) was con-sidered and not wood production alone (Fig 4) A possibleexplanation of the weak diversity effect is that the speciesrichness in our plots may have been too high to show a pos-itive DPR as diversity effects on productivity generally aremore pronounced at lower species numbers (Paquette andMessier 2011 Tilman et al 1997) It would be interesting tosee how tree species richness translates into functional diver-sity in our two transects and if functional diversity is alreadysaturating at intermediate levels of tree diversity (as a resultof increasing functional redundancy) That would explain theweak diversity effect But functional trait data are only avail-able for a small fraction of the recorded tree species

Our results support our third hypothesis and fit the contro-versial picture that emerges from the existing observationalstudies on the DPR in tropical forests Chisholm et al (2013)systematically explored the role of diversity for productivityin natural tropical forests and found only a small absolute ef-fect when controlling for stem density effects in the 004 haplots a doubling of species richness corresponded in theirstudy to only a 5 increase in WP on average At a largerscale (025 1 ha) results were mixed and negative relation-ships became more common This points at a negligible roleof tree diversity for productivity in tropical-forest landscapesin line with our findings These results from tropical forestsmatch findings from more than 100 000 forest plots in theUnited States where the predominant DPR was concave andnegative in humid climates and primarily non-significant in

harsh climates (Fei et al 2018) Similar to the DPR a sig-nificant positive diversityndashaboveground biomass relationshipwas detectable in tropical lowland forests only in small plots(004 ha) while it disappeared at a larger scale (1 ha plots)(Sullivan et al 2017) Moreover this study showed that thediversity effect on AGB was small doubling species richnessat 004 ha increased biomass only by 69

An additional important result of our study is the high vari-ation in stand properties and WP across the plots of an ele-vation level (Fig 2) Environmental variation related to to-pography can be a strong driver of forest structure and treespecies composition and it influences productivity and nu-trient cycling through changes in traits as has been shownfor several tropical lowland forests (Fortunel et al 2018Jucker et al 2018) and also for the TMF of the Loja tran-sect (Werner and Homeier 2015 Pierick et al 2021)

Our analysis of multiple drivers of productivity furthershows that a diversity effect even if significant is only ofminor importance when other productivity-influencing fac-tors are also included in the analysis In our case diversityranked behind effects of elevation (temperature) soil nutrientavailability and wood specific gravity For understanding thecontrols of forest productivity it is more important to studythe influences of environmental factors and stand structuraland functional properties in more detail

5 Conclusions

From the analysis of more than 80 old-growth forest plotsin two highly diverse Andean regions with large geologicaland topographic heterogeneity we conclude that the maindeterminants of aboveground productivity in tropical mon-tane forests are elevation (primarily as a proxy for temper-ature) nutrient supply and stand structural properties givennon-limiting moisture conditions Nutrient availability seemsto affect productivity not only directly but also indirectlythrough stand properties while tree diversity has only asmall influence on productivity Tree species turnover andassociated changes in functional traits are far more impor-tant for productivity than changes in species diversity Func-tional biodiversity research has to proceed from the searchfor the significance of diversity effects which has domi-nated biodiversityndashecosystem functioning (BEF) research inthe past to an assessment of the importance of diversity ef-fects for productivity and other ecosystem processes in orderto better understand the functional role of diversity A deeperunderstanding of foliar and root trait variation with elevationand among the species of a community will help to achievea more mechanistic understanding of tropical-forest produc-tivity change along environmental gradients Our findings areof practical relevance as the analysis addresses spatial scalesrelevant for old-growth forest conservation and management



Appendix A Study sites and plots

Figure A1 Map showing the location of the two study sites in Ecuador



Appendix B PCAs of soil properties

Figure B1 Principal component analysis of the soil properties of 83 study plots (Loja transect and Napo transect) PC1 and PC2 explained374 and 261 of variance respectively

Table B1 Factor loadings of the variables in the PCA (Fig B1)

Variable PC1 PC2

Organic layer depth minus0319 minus0182pH KCl 0421 0208K 0349 minus0300Mg 0436 minus0258Ca 0438 minus0192Al minus0289 0146CN ratio minus0199 minus0181Pav 0111 minus0390Nmin S 0189 0507Nnitr S 0212 0517



Figure B2 Principal component analysis of the soil properties of 54 study plots of the Loja transect PC1 and PC2 explained 469 and146 of variance respectively


Variable PC1 PC2

Organic-layer depth minus0270 minus0182pH KCl 0392 minus0020K 0343 minus0419Mg 0431 minus0173Ca 0413 minus0125Al minus0278 0280CN ratio minus0144 minus0214Pav 0218 minus0104Nmin S 0248 0621Nnitr S 0305 0475



Data availability All data sets used are included in the Supple-ment

Supplement Supplement S1 contains a table with the stand param-eters of the study plots Supplement S2 contains a table with the soilparameters of the study plots The supplement related to this arti-cle is available online at httpsdoiorg105194bg-18-1525-2021-supplement

Author contributions JH and CL designed the study JH performedthe research and analyzed the data and JH and CL wrote themanuscript

Competing interests The authors declare that they have no conflictof interest

Acknowledgements We thank the universities in Loja (UTPLUNL) and in Quito (PUCE) for continuous support during our fieldstudies We thank the Ministerio de Ambiente del Ecuador for grant-ing the research permits We further acknowledge the assistanceof Nixon Cumbicus Miguel-Angel Chinchero Jaime Pentildea Ro-man Link Malte Unger and Katrin Mikolajewski during fieldworkand lab work We thank the two anonymous reviewers for their con-structive criticism of an earlier draft of the paper

Financial support This research has been supported by theDeutsche Forschungsgemeinschaft (grant nos Ho32962Ho32964 and Le76210) and the Bundesministerium fuumlr Bil-dung und Forschung (grant no Pro Benefit)

This open-access publication was fundedby the University of Goumlttingen

Review statement This paper was edited by Sara Vicca and re-viewed by two anonymous referees

References

Aragatildeo L E O C Malhi Y Metcalfe D B Silva-Espejo J EJimeacutenez E Navarrete D Almeida S Costa A C L Sali-nas N Phillips O L Anderson L O Alvarez E BakerT R Goncalvez P H Huamaacuten-Ovalle J Mamani-SoloacuterzanoM Meir P Monteagudo A Patintildeo S Pentildeuela M C Pri-eto A Quesada C A Rozas-Daacutevila A Rudas A Silva Jr JA and Vaacutesquez R Above- and below-ground net primary pro-ductivity across ten Amazonian forests on contrasting soils Bio-geosciences 6 2759ndash2778 httpsdoiorg105194bg-6-2759-2009 2009

Arbuckle J S IBM SPSS Amos 24 Userrsquos Guide IBM ChicagoUSA 2016

Aiba S Takyu M and Kitayama K Dynamics productivityand species richness of tropical rainforests along elevational andedaphic gradients on Mount Kinabalu Borneo Ecol Res 20279ndash286 2005

Ashton P S Floristic zonation of tree communities on wettropical mountains revisited Perspect Plant Ecol 6 87ndash104httpsdoiorg1010781433-8319-00044 2003

Banin L Lewis S L Lopez-Gonzalez G Baker T R Que-sada C A Chao K J Burslem D F R P Nilus R SalimK A Keeling H C Tan S Davies S J Mendoza A MVaacutesquez R Lloyd J Neill D A Pitman N and PhillipsO L Tropical forest wood production a cross-continental com-parison J Ecol 104 1025ndash1037 httpsdoiorg1011111365-274512263 2014

Bendix J Rollenbeck R Fabian P Emck P Richter M andBeck E Climate variability in Gradients in a tropical mountainecosystem of Ecuador edited by Beck E Kottke I Bendix JMakeschin F and Mosandl R Springer Berlin and HeidelbergGermany 281ndash290 2008

Benner J Vitousek P M and Ostertag R Nutrient cycling andnutrient limitation in tropical montane cloud forests in Tropi-cal montane cloud forests science for conservation and manage-ment edited by Bruijnzeel L A Scatena F N and HamiltonL Cambidge University Press Cambidge UK 90ndash100 2010

Bruijnzeel L A and Veneklaas E J Climatic conditions and trop-ical montane forest productivity the fog has not lifted yet Ecol-ogy 79 3ndash9 1998

Bruijnzeel L A Scatena F N and Hamilton L S Tropical mon-tane cloud forests science for conservation and managementCambridge University Press Cambridge UK 2010

Chave J Andalo C Brown S Cairns M A Chambers J QEamus D Foumllster H Fromard F Higuchi N Kira T Les-cure J P Nelson B W Ogawa H Puig H Rieacutera B andYamakura T Tree allometry and improved estimation of carbonstocks and balance in tropical forests Oecologia 145 87ndash99httpsdoiorg101007s00442-005-0100-x 2005

Chave J Coomes D Jansen S Lewis S L Swenson N Gand Zanne A E Towards a worldwide wood economics spec-trum Ecol Lett 12 351ndash366 httpsdoiorg101111j1461-0248200901285x 2009

Chen Y Wright S J Muller-Landau H C Hubbell S P WangY and Yu S Positive effects of neighborhood complementarityon tree growth in a Neotropical forest Ecology 97 776ndash785httpsdoiorg10189015-06251 2016

Chisholm R A Muller-Landau H C Abdul Rahman K BebberD P Bin Y Bohlman S A Bourg N A Brinks J Bunyave-jchewin S Butt N Cao H Cao M Caacuterdenas D ChangL-W Chiang J-M Chuyong G Condit R Dattaraja H SDavies S Duque A Fletcher C Gunatilleke N GunatillekeS Hao Z Harrison R D Howe R Hsieh C-F Hubbell SP Itoh A Kenfack D Kiratiprayoon S Larson A J LianJ Lin D Liu H Lutz J A Ma K Malhi Y McMahonS McShea W Meegaskumbura M Mohd Razman S More-croft M D Nytch C J Oliveira A Parker G G Pulla SPunchi-Manage R Romero-Saltos H Sang W Schurman JSu S-H Sukumar R Sun I-F Suresh H S Tan S ThomasD Thomas S Thompson J Valencia R Wolf A Yap S YeW Yuan Z and Zimmerman J K Scale-dependent relation-ships between tree species richness and ecosystem function in



forests J Ecol 101 1214ndash1224 httpsdoiorg1011111365-274512132 2013

Cleveland C C Townsend A R Taylor P Alvarez-Clare SBustamante M M Chuyong G Dobrowski S Z Grierson PHarms K E Houlton B Z Marklein A Parton W PorderS Reed S C Sierra C A Silver W L Tanner E V J andWieder W R Relationships among net primary productivitynutrients and climate in tropical rain forest a pan-tropical anal-ysis Ecol Lett 14 939ndash947 httpsdoiorg101111j1461-0248201101658x 2011

Dalling J W Heineman K Lopez O R Wright S J andTurner B L Nutrient availability in tropical rain forests theparadigm of phosphorus limitation in Tropical Tree Physiologyedited by Goldstein G and Santiago L S Springer ChamUK 261ndash273 2016a

Dalling J W Heineman K Gonzaacutelez G and Ostertag R Ge-ographic environmental and biotic sources of variation in thenutrient relations of tropical montane forests J Trop Ecol 32368ndash383 httpsdoiorg101017S0266467415000619 2016b

Doughty C E Goldsmith G R Raab N Girardin C A JFarfan-Amezquita F Huaraca-Huasco W Silva-Espejo J EAraujo-Murakami A da Costa A C L Rocha W GalbraithD Meir P Metcalfe D B and Malhi Y What controls varia-tion in carbon use efficiency among Amazonian tropical forestsBiotropica 50 16ndash25 httpsdoiorg101111btp12504 2018

Fahey T J Sherman R E and Tanner E Tropical mon-tane cloud forest environmental drivers of vegetation struc-ture and ecosystem function J Trop Ecol 32 355ndash367httpsdoiorg101017S0266467415000176 2015

Fei S Jo I Guo Q Wardle D A Fang J Chen A Os-walt C M and Brockerhoff E G Impacts of climate onthe biodiversity-productivity relationship in natural forests NatCommun 9 1ndash7 httpsdoiorg101038s41467-018-07880-w2018

Finegan B Pentildea-Claros M de Oliveira A Ascarrunz N Bret-Harte M S Carrentildeo-Rocabado G Casanoves F Diacuteaz SEguiguren Velepucha P Fernandez F Licona J C LorenzoL Salgado Negret B Vaz M and Poorter L Does functionaltrait diversity predict above-ground biomass and productivity oftropical forests Testing three alternative hypotheses J Ecol103 191ndash201 httpsdoiorg1011111365-274512346 2015

Fisher J B Malhi Y Torres I C Metcalfe D B van de WegM J Meir P Silva-Espejo J E and Huaraca Huasco W Nu-trient limitation in rainforests and cloud forests along a 3000-melevation gradient in the Peruvian Andes Oecologia 172 889ndash902 httpsdoiorg101007s00442-012-2522-6 2013

Fortunel C Lasky J R Uriarte M Valencia R Wright S JGarwood N C and Kraft N J B Topography and neighbor-hood crowding can interact to shape species growth and distri-bution in a diverse Amazonian forest Ecology 99 2272ndash2283httpsdoiorg101002ecy2441 2018

Fyllas N M Bentley L P Shenkin A Asner G P AtkinO K Diacuteaz S Enquist B J Farfan-Rios W Gloor EGuerrieri R Huasco W H Ishida Y Martin R E MeirP Phillips O Salinas N Silman M Weerasinghe L KZaragoza-Castells J and Malhi Y Solar radiation and func-tional traits explain the decline of forest primary productivityalong a tropical elevation gradient Ecol Lett 20 730ndash740httpsdoiorg101111ele12771 2017

Girardin C A J Malhi Y Aragatildeo L E O C Mamani MHuaraca Huasco W Durand L Feeley K J Rapp J Silva-Espejo J E Silman M Salinas N and Whittaker R J Netprimary productivity allocation and cycling of carbonalong atropical forest elevational transect in the Peruvian An-des GlobChange Biol 16 3176ndash3192 httpsdoiorg101111j1365-2486201002235x 2010

Girardin C A J Malhi Y Feeley K J Rapp J M Sil-man M R Meir P Huaraca Huarasco W Salinas N Ma-mani M Silva-Espejo J E Garcia Cabrera K Farfan RiosW Metcalfe D B Doughty C E and Aragatildeo L Season-ality of above-ground net primary productivity along an An-dean altitudinal transect in Peru J Trop Ecol 30 503ndash519httpsdoiorg101017S0266467414000443 2014

Gotelli N J and Colwell R K Quantifying biodiversity proce-dures and pitfalls in the measurement and comparison of speciesrichness Ecol Lett 4 379ndash391 httpsdoiorg101046j1461-0248200100230x 2001

Graefe S Hertel D and Leuschner C Estimating FineRoot Turnover in Tropical Forests along an ElevationalTransect using Minirhizotrons Biotropica 40 536ndash542httpsdoiorg101111j1744-7429200800419x 2008

Hoeber S Leuschner C Koumlhler L Arias-Aguilar D andSchuldt B The importance of hydraulic conductivity and wooddensity to growth performance in eight tree species from atropical semi-dry climate Forest Ecol Manag 300 126ndash136httpsdoiorg101016jforeco201406039 2014

Hofhansl F Schnecker J Singer G and Wanek WNew insights into mechanisms driving carbon alloca-tion in tropical forests New Phytol 205 137ndash146httpsdoiorg101111nph13007 2015

Homeier J Breckle S-W Guumlnter S Rollenbeck R T andLeuschner C Tree Diversity Forest Structure and Productiv-ity along Altitudinal and Topographical Gradients in a Species-Rich Ecuadorian Montane Rain Forest Biotropica 42 140ndash148httpsdoiorg101111j1744-7429200900547x 2010

Homeier J Hertel D Camenzind T Cumbicus N MaraunM Martinson G O Poma L N Rillig M C SandmannD Scheu S Veldkamp E Wilcke W Wullaert H andLeuschner C Tropical Andean forests are highly susceptibleto nutrient inputs ndash Rapid effects of experimental N and P ad-dition to an Ecuadorian montane forest PLOS ONE 7 e47128httpsdoiorg101371journalpone0047128 2012

Huang Y Chen Y Castro-Izaguirre N Baruffol M Brezzi MLang A Li Y Haumlrdtle W von Oheimb G Yang X LiuX Pei K Both S Yang B Eichenberg D Assmann TBauhus J Behrens T Buscot F Chen X-Y Chesters DDing B-Y Durka W Erfmeier A Fang J Fischer M GuoL-D Guo D Gutknecht J L M He J-S He C-L Hec-tor A Houmlnig L Hu R-Y Klein A-M Kuumlhn P LiangY Li S Michalski S Scherer-Lorenzen M Schmidt KScholten T Schuldt A Shi X Tan M-Z Tang Z Tro-gisch S Wang Z Welk E Wirth C Wubet T Xiang WYu M Yu X-D Zhang J Zhang S Zhang N Zhou H-Z Zhu C-D Zhu L Bruelheide H Ma K Niklaus P Aand Schmid B Impacts of species richness on productivity ina large-scale subtropical forest experiment Science 362 80ndash83httpsdoiorg101126scienceaat6405 2018



Jozsa L A and Brix H The effects of fertilization and thinningon wood quality of a 24-year-old Douglas-fir stand Can J ForestRes 19 1137ndash1145 httpsdoiorg101139x89-172 1989

Jucker T Bongalov B Burslem D F R P Nilus R DalponteM Lewis S L Phillips O L Qie L and Coomes DA Topography shapes the structure composition and func-tion of tropical forest landscapes Ecol Lett 21 989ndash1000httpsdoiorg101111ele12964 2018

Keeling H C and Phillips O L The global relationship be-tween forest productivity and biomass Global Ecol Biogeogr16 618ndash631 httpsdoiorg101111j1466-8238200700314x2007

Kessler M Salazar L Homeier J and Kluge J Speciesrichness ndash productivity relationships of tropical terrestrialferns at regional and local scales J Ecol 102 1623ndash1633httpsdoiorg1011111365-274512299 2014

Kitayama K and Aiba S-I Ecosystem structure and pro-ductivity of tropical rain forests along altitudinal gradientswith contrasting soil phosphorus pools on Mount KinabaluBorneo J Ecol 90 37ndash51 httpsdoiorg101046j0022-0477200100634x 2002

Kotowska M Link R Roumlll A Hertel D Houmllscher D WaiteP-A Moser G Tjoa A Leuschner C and Schuldt B Ef-fects of wood hydraulic properties on water use and productivityof tropical rainforest trees Front For Glob Change 3 598759httpsdoiorg103389ffgc2020598759 2021

Lambers H and Oliveira R S Plant Physiological Ecology 3rdedition Springer Nature Cham UK 2019

Leuschner C Zach A Moser G Homeier J Graefe SHertel D Wittich B Soethe N Iost S Roumlderstein MHorna V and Wolf K The carbon balance of tropical moun-tain forests along an altitudinal transect in Ecosystem ser-vices biodiversity and environmental change in a tropical moun-tain ecosystem of South Ecuador edited by Bendix J BeckE Braumluning A Makeschin F Mosandl R Scheu S andWilcke W Springer Berlin and Heidelberg Germany 117ndash139httpsdoiorg101007978-3-642-38137-9_10 2013

Liang J Crowther T W Picard N Wiser S Zhou M Al-berti G Schulze E-D Mcguire A D Bozzato F Pret-zsch H Paquette A Heacuterault B Scherer-Lorenzen M Bar-rett C B Glick H B Hengeveld G M Nabuurs G-JPfautsch S Viana H Vibrans A C Ammer C Schall PVerbyla D Tchebakova N Fischer M Watson J V ChenH Y H Lei X Schelhaas M-J Lu H Gianelle D Par-fenova E I Salas C Lee E Lee B Kim H S Bruel-heide H Coomes D A Piotto D Sunderland T SchmidB Gourlet-Fleury S Sonkeacute B Tavani R Zhu J BrandlS Baraloto C Frizzera L Ba R Oleksyn J Peri P LGonmadje C Marthy W Brien T O Martin E H Mar-shall A R Rovero F Bitariho R Niklaus P A Alvarez-Loayza P Chamuya N Valencia R Mortier F Wortel VEngone-Obiang N L Ferreira L V Odeke D E VasquezR M Lewis S L and Reich P B Positive biodiversity-productivity relationship predominant in global forests Science354 aaf8957 httpsdoiorg101126scienceaaf8957 2016

Lohbeck M Poorter L Martiacutenez-Ramos M and Bongers FBiomass is the main driver of changes in ecosystem processrates during tropical forest succession Ecology 96 1242ndash1252httpsdoiorg10189014-04721 2015

Luyssaert S Inglima I Jung M Richardson A D Reich-stein M Papale D Piao S L Schulze E-D Wingate LMatteuci G Aragao L Aubinet M Beer C BernhoferC Black K G Bonal D Bonnefond J-M Chambers JCiais P Cook B Davis K J Dolman A J Gielen BGoulden M Grace J Granier A Grelle A Griffis T Gruumln-wald T Guidolotti G Hanson P J Harding R HollingerD Y Hutyra L R Kolari P Kruijt B Kuzsch W Lager-gren F Laurila T Law B E Le Maire G Lindroth ALoustau D Malhi Y Mateus J Migliavacca M MissonL Montagnani L Moncrieff J Moors E Munger J WNikinmaa E Ollinger S V Pita G Rebmann C Roup-sard O Saigusa N Sanz M J Seufert G Sierra C SmithM-L Tang J Valentini R Vesala T and Janssens I ACO2 balance of boreal temperate and tropical forests derivedfrom a global database Glob Change Biol 13 2509ndash2537httpsdoiorg101111j1365-2486200701439x 2007

Malhi Y The productivity metabolism and carbon cy-cle of tropical forest vegetation J Ecol 100 65ndash75httpsdoiorg101111j1365-2745201101916x 2012

Malhi Y Baker T R Phillips O L Almeida S Alvarez EArroyo L Chave J Czimczik C I Fiore A D Higuchi NKilleen T J Laurance S G Laurance W F Lewis S LMontoya L M M Monteagudo A Neill D A Vargas P NPatintildeo S Pitman N C Quesada C A Salomatildeo R SilvaJ N M Lezama A T Martiacutenez R V Terborgh J VincetiB and Lloyd J The above-ground coarse wood productivity of104 Neotropical forest plots Glob Change Biol 10 563ndash591httpsdoiorg101111j1529-8817200300778x 2004

Malhi Y Girardin C A J Goldsmith G R Doughty C E Sali-nas N Metcalfe D B Huaraca Huasco W Silva-Espejo J Edel Aguilla-Pasquell J Farfaacuten Ameacutezquita F Aragatildeo L E OC Guerrieri R Ishida F Y Bahar N H A Farfan-Rios WPhillips O L Meir P and Silman M The variation of pro-ductivity and its allocation along a tropical elevation gradient awhole carbon budget perspective New Phytol 214 1019ndash1032httpsdoiorg101111nph14189 2017

McIntire E J B and Fajardo A Facilitation as a ubiq-uitous driver of biodiversity New Phytol 201 403ndash416httpsdoiorg101111nph12478 2014

Mori A S Lertzman K P and Gustafsson L Biodiver-sity and ecosystem services in forest ecosystems a researchagenda for applied forest ecology J Appl Ecol 54 12ndash27httpsdoiorg1011111365-266412669 2017

Moser G Roumlderstein M Soethe N Hertel D and LeuschnerC Altitudinal changes in stand structure and biomass allocationof tropical mountain forests in relation to microclimate and soilchemistry in Gradients in a Tropical Mountain Ecosystem ofEcuador edited by Beck E Bendix J Kottke I MakeschinF and Mosandl R Springer Berlin and Heidelberg Germany229ndash242 httpsdoiorg101007978-3-540-73526-7_22 2008

Moser G Leuschner C Hertel D Graefe S Soethe Nand Iost S Elevation effects on the carbon budget of trop-ical mountain forests (S Ecuador) the role of the below-ground compartment Glob Change Biol 17 2211ndash2226httpsdoiorg101111j1365-2486201002367x 2011

Muller-Landau H C Interspecific and inter-site variation inwood specific gravity of tropical trees Biotropica 36 20ndash32httpsdoiorg101111j1744-74292004tb00292x 2004



Pan Y Birdsey R A Phillips O L and JacksonR B The structure distribution and biomass of theworldrsquos forests Annu Rev Ecol Evol S 44 593ndash622httpsdoiorg101146annurev-ecolsys-110512-135914 2013

Paoli G D and Curran L M Soil nutrients limit finelitter production and tree growth in mature lowland for-est of southwestern Borneo Ecosystems 10 503ndash518httpsdoiorg101007s10021-007-9042-y 2007

Paquette A and Messier C The effect of biodiversity ontree productivity from temperate to boreal forests GlobalEcol Biogeogr 20 170ndash180 httpsdoiorg101111j1466-8238201000592x 2011

Pastor J Aber J D McClaugherty C A and Melillo J MAboveground Production and N and P Cycling Along a Nitro-gen Mineralization Gradient on Blackhawk Island WisconsinEcology 65 256ndash268 httpsdoiorg1023071939478 1984

Pierick K Leuschner C and Homeier J Topography as a factordriving small-scale variation in tree fine root traits and root func-tional diversity in a species-rich tropical montane forest NewPhytol httpsdoiorg101111nph17136 2021

Poorter L Wright S J Paz H Ackerly D D Condit R Ibarra-Manriacutequez G Harms K E Licona J C Martiacutenez-RamosM Mazer S J Muller-Landau H C Pentildea-Claros M WebbC O and Wright I J Are functional traits good predictors ofdemographic rates Evidence from five neotropical forests Ecol-ogy 89 1908ndash1920 httpsdoiorg10189007-02071 2008

Quesada C A Phillips O L Schwarz M Czimczik C I BakerT R Patintildeo S Fyllas N M Hodnett M G Herrera RAlmeida S Alvarez Daacutevila E Arneth A Arroyo L ChaoK J Dezzeo N Erwin T di Fiore A Higuchi N Hono-rio Coronado E Jimenez E M Killeen T Lezama A TLloyd G Loacutepez-Gonzaacutelez G Luizatildeo F J Malhi Y Mon-teagudo A Neill D A Nuacutentildeez Vargas P Paiva R PeacockJ Pentildeuela M C Pentildea Cruz A Pitman N Priante Filho NPrieto A Ramiacuterez H Rudas A Salomatildeo R Santos A JB Schmerler J Silva N Silveira M Vaacutesquez R VieiraI Terborgh J and Lloyd J Basin-wide variations in Amazonforest structure and function are mediated by both soils and cli-mate Biogeosciences 9 2203ndash2246 httpsdoiorg105194bg-9-2203-2012 2012

Rahbek C Borregaard M K Colwell R K Dalsgaard BHolt B G Morueta-Holme N Nogues-Bravo D WhittakerR J and Fjeldsaring J Humboldtrsquos enigma What causes globalpatterns of mountain biodiversity Science 365 1108ndash1113httpsdoiorg101126scienceaax0149 2019

Reich P B Key canopy traits drive forest pro-ductivity Proc R Soc Ser B 279 2128ndash2134httpsdoiorg101098rspb20112270 2012

Reich P B and Bolstad P Productivity of evergreen and decid-uous temperate forests in Terrestrial global productivity editedby Roy J Saugier B and Mooney H A Academic Press SanDiego USA 245ndash283 2001

Salazar L Homeier J Kessler M Abrahamczyk S Lehnert MKroumlmer T and Kluge J Diversity patterns of ferns along ele-vational gradients in Andean tropical forests Plant Ecol Divers8 13ndash24 httpsdoiorg101080175508742013843036 2015

Sapijanskas J Paquette A Potvin C Kunert N and LoreauM Tropical tree diversity enhances light capture through crown

plasticity and spatial and temporal niche differences Ecology95 2479ndash2492 httpsdoiorg10189013-13661 2014

Schnabel F Schwarz J A Danescu A Fichtner A Nock C ABauhus J and Potvin C Drivers of productivity and its tempo-ral stability in a tropical tree diversity experiment Glob ChangeBiol 25 4257ndash4272 httpsdoiorg101111gcb14792 2019

Schuur E A Productivity and global climate revisitedthe sensitivity of tropical forest growth to precipitationEcology 84 1165ndash1170 httpsdoiorg1018900012-9658(2003)084[1165PAGCRT]20CO2 2003

Soethe N Lehmann J and Engels C The vertical pat-tern of rooting and nutrient uptake at different altitudes of asouth Ecuadorian montane forest Plant Soil 286 287ndash299httpsdoiorg101007s11104-006-9044-0 2006

Spracklen D V and Righelato R Tropical montane forests area larger than expected global carbon store Biogeosciences 112741ndash2754 httpsdoiorg105194bg-11-2741-2014 2014

Sullivan M J P Talbot J Lewis S L Phillips O L Qie LBegne S K Chave J Cuni-Sanchez A Hubau W Lopez-Gonzalez G Miles L Monteagudo-Mendoza A Sonkeacute BSunderland T ter Steege H White L J T Affum-BaffoeK Aiba S de Almeida E C de Oliveira E A Alvarez-Loayza P Daacutevila E Aacute Andrade A Aragatildeo L E O CAshton P C G A A Baker T R Balinga M Banin LF Baraloto C Bastin J-F Berry N Bogaert J Bonal DBongers F Brienen R Camargo J L C Ceroacuten C MoscosoV C Chezeaux E Clark C J Pacheco Aacute C Comiskey JA Valverde F C Coronado E N H Dargie G Davies SJ De Canniere C K M N D Doucet J-L Erwin T LEspejo J S Ewango C E N Fauset S Feldpausch T RHerrera R Gilpin M Gloor E Hall J S Harris D J HartT B Kartawinata K Kho L K Kitayama K Laurance SG W Laurance W F Leal M E Lovejoy T Lovett J CLukasu F M Makana J-R Malhi Y Maracahipes L Ma-rimon B S Junior B H M Marshall A R Morandi P SMukendi J T Mukinzi J Nilus R Vargas P N CamachoN C P Pardo G Pentildea-Claros M Peacutetronelli P PickavanceG C Poulsen A D Poulsen J R Primack R B PriyadiH Quesada C A Reitsma J Reacutejou-Meacutechain M RestrepoZ Rutishauser E Salim K A Salomatildeo R P Samsoedin ISheil D Sierra R Silveira M Slik J W F Steel L Tae-doumg H Tan S Terborgh J W Thomas S C Toledo MUmunay P Valenzuela Gamarra L Vieira I C G Vos V AWang O Willcock S and Zemagho L Diversity and carbonstorage across the tropical forest biome Sci Rep-UK 7 39102httpsdoiorg101038srep39102 2017

Tanner E V J Vitousek P M and Cuevas E Experimental in-vestigation of nutrient limitation of forest growth on wet tropi-cal mountains Ecology 79 10ndash23 httpsdoiorg1018900012-9658(1998)079[0010EIONLO]20CO2 1998

Tilman D Lehman C L and Thomson K T Plantdiversity and ecosystem productivity theoretical con-siderations P Natl Acad Sci USA 94 1857ndash1861httpsdoiorg101073pnas9451857 1997

Trabucco A and Zomer R J Global Aridity Index (Global-Aridity) and Global Potential Evapo-Transpiration (Global-PET)Geospatial Database CGIAR Consortium for Spatial Informa-tion CGIAR-CSI GeoPortal available at httpwwwcsicgiarorg (last access 31 August 2018) 2009



Tuck S L OrsquoBrien M J O Philipson C D Saner P TanadiniM Dzulkifli D Godfray H C J Godoong E Nilus R OngR C Schmid B Sinun W Snaddon J L Snoep M TangkiH Tay J Ulok P Wai Y S Weilenmann M Reynolds Gand Hector A The value of biodiversity for the functioning oftropical forests insurance effects during the first decade of theSabah biodiversity experiment Proc R Soc B 283 20161451httpsdoiorg101098rspb20161451 2016

Unger M Leuschner C and Homeier J Variability of indices ofmacronutrient availability in soils at different spatial scales alongan elevation transect in tropical moist forests (NE Ecuador) PlantSoil 336 443ndash458 httpsdoiorg101007s11104-010-0494-z2010

Unger M Homeier J and Leuschner C Effects of soil chem-istry on tropical forest biomass and productivity at differentelevations in the equatorial Andes Oecologia 170 263ndash274httpsdoiorg101007s00442-012-2295-y 2012

Unger M Homeier J and Leuschner C Relationships amongleaf area index below-canopy light availability and tree diversityalong a transect from tropical lowland to montane forests in NEEcuador Trop Ecol 54 33ndash45 2013

van der Sande M T Arets E J Pentildea-Claros M Hoosbeek MR Caacuteceres-Siani Y van der Hout P and Poorter L Soil fer-tility and species traits but not diversity drive productivity andbiomass stocks in a Guyanese tropical rainforest Funct Ecol32 461ndash474 httpsdoiorg1011111365-243512968 2018

van de Weg M J Meier P Williams M Girardin C MalhiY Silva-Espejo J and Grace J Gross primary productivity ofa high elevation tropical montane cloud forest Ecosystems 17751ndash764 httpsdoiorg101007s10021-014-9758-4 2014

Vicca S Luyssaert S Pentildeuelas J Campioli M Chapin III FS Ciais P Heinemeyer A Houmlgberg P Kutsch W L Law BE Malhi Y Papale D Piao S L Reichstein M Schulze ED and Janssens I A Fertile forests produce biomass more effi-ciently Ecol Lett 15 520ndash526 httpsdoiorg101111j1461-0248201201775x 2012

Vitousek P M and Sanford R L Nutrient cycling in moist tropi-cal forest Annu Rev Ecol Syst 17 137ndash167 1986

Wallis C I Homeier J Pentildea J Brandl R Farwig Nand Bendix J Modeling tropical montane forest biomassproductivity and canopy traits with multispectral re-mote sensing data Remote Sens Environ 225 77ndash92httpsdoiorg101016jrse201902021 2019

Werner F A and Homeier J Is tropical montane forest hetero-geneity promoted by a resource-driven feedback cycle Evi-dence from nutrient relations herbivory and litter decomposi-tion along a topographical gradient Funct Ecol 29 430ndash440httpsdoiorg1011111365-243512351 2015

Wittich B Horna V Homeier J and Leuschner C Altitudi-nal change in the photosynthetic capacity of tropical trees a casestudy from Ecuador and a pantropical literature analysis Ecosys-tems 15 958ndash973 httpsdoiorg101007s10021-012-9556-92012

Wolf K Veldkamp E Homeier J and Martinson G O Ni-trogen availability links forest productivity soil nitrous ox-ide and nitric oxide fluxes of a tropical montane forestin southern Ecuador Global Biogeochem Cy 25 GB4009httpsdoiorg1010292010GB003876 2011

Wright S J Yavitt J B Wurzburger N Turner B L TannerE V J Sayer E J Santiago L S Kaspari M Hedin L OHarms K E Garcia M N and Corre M DPotassium phos-phorus or nitrogen limit root allocation tree growth or litterproduction in a lowland tropical forest Ecology 92 1616ndash1625httpsdoiorg10189010-15581 2011

Zimmermann M Meir P Bird M I Malhi Y and CcahuanaA J Q Temporal variation and climate dependence of soil res-piration and its components along a 3000 m altitudinal trop-ical forest gradient Global Biogeochem Cy 24 GB4012httpsdoiorg1010292010GB003787 2010


Abstract

Introduction

Methods

Study transects

Plot inventory and forest productivity

Data analysis

Results

Patterns of tree diversity and productivity

Identification of principal soil factors

Factors determining wood production in the merged data set

Factors determining NPPa in the Loja transect

Discussion

Conclusions



Data availability

Supplement

Author contributions

Competing interests

Acknowledgements

Financial support

Review statement

References


The existing observational data from tropical forests sug-gest that (i) tree diversity may enhance productivity undercertain conditions and (ii) a positive DPR might be moreprominent in small plots To date it is not clear under whichabiotic conditions and stand structural settings a positiveDPR does emerge and how important tree diversity is com-pared to abiotic and other biotic determinants of productivity

The search for abiotic determinants of tropical-forest pro-ductivity has mainly concentrated on soil nutrient availabil-ity solar irradiance and precipitation revealing clear depen-dencies in studies along environmental gradients (Schuur2003 Girardin et al 2014 Malhi et al 2017) Many tropi-cal forests especially those in the lowlands seem to be P-limited (Aragatildeo et al 2009 Quesada et al 2012) whileN limitation may be more influential in tropical montaneforests (TMFs) (Tanner et al 1998 Moser et al 2008Homeier et al 2012 Fisher et al 2013) Fertilization ex-periments and studies on foliar nutrient levels across envi-ronmental gradients suggest that the NPP of tropical foreststends to increase with soil fertility notably N and P avail-ability but locally also with base cation supply (Kitayamaand Aiba 2002 Cleveland et al 2011 Malhi et al 2004Homeier et al 2010 Unger et al 2012 Fisher et al 2013Banin et al 2014 Hofhansel et al 2015) Bruijnzeel andVeneklaas (1998) assumed that the productivity of TMFsis primarily limited by cloudiness ie low solar radiationwhich is supported by a modeling study (Fyllas et al 2017)Precipitation should play a minor role in TMFs (Zimmer-mann et al 2010) Less clear is the role of temperature Theelevational temperature decrease can act on plants in multi-ple ways directly through a limitation of photosynthetic ac-tivity and of carbon water and nutrient cycling in the plantand it may influence stem leaf and root morphology Theimpairment of nutrient supply by low soil temperatures isan indirect effect Yet plants can adapt to lower tempera-tures thereby partly reducing these limitations (Lambers andOliveira 2019) The majority of studies along elevation tran-sects found an NPP decline with decreasing temperature intropical mountains (Kitayama and Aiba 2002 Girardin etal 2010 Cleveland et al 2011 Leuschner et al 2013van de Weg et al 2014) However an analysis of a globaldatabase failed to show a temperature dependence of NPP intropical forests (Luyssaert et al 2007) This fits to modelingresults suggesting that trait variation with elevational speciesturnover in TMFs may partly capture the effect of tempera-ture on tree metabolism and growth (Fyllas et al 2017)

Biotic controls related to stand structural properties no-tably stem density basal area and wood density and thefunctional composition of the community can also affectforest productivity Some studies found that forest NPP in-creased with standing aboveground biomass (AGB) (Keelingand Phillips 2007 Pan et al 2013 Lohbeck et al 2015) butthe relation does not seem to be very close since woody tissueresidence time ie the ratio between AGB and abovegroundproductivity varies considerably with soil fertility and cli-

mate in tropical moist forests (Malhi 2012 Quesada et al2012 Chisholm et al 2013 Malhi et al 2017) Variation inleaf area index and leaf functional traits such as LMA (leafmass per area) and foliar N and P content should be majordeterminants of net primary productivity (van de Weg et al2014 Wittich et al 2012 Finegan et al 2015) as they in-fluence the photosynthetic capacity of the canopy and radi-ation interception Indeed the modeling study of Fyllas etal (2017) indicated for tropical forests along an Andean el-evation gradient that the influence of LMA foliar N and Pcontent and wood density of the species was a more impor-tant determinant of NPP than temperature which is in linewith the findings of Luyssaert et al (2007) From the partlycontradicting results we conclude that a multi-factorial ap-proach is needed to understand the determinants of tropical-forest productivity Yet most of the existing studies have in-vestigated only one or two of the mentioned factors

Tropical montane forests cover all humid tropical moun-tains between sim 1000 and 3000 m asl with the widest dis-tribution in the Andes and Central American mountains andmore restricted occurrence in the mountains of the Pale-otropis They differ from tropical lowland forest in terms oflower canopy heights higher stem densities and lower AGBTMFs play important roles in the provision of water for hu-man populations and as stores of carbon and tropical moun-tains are known as hotspots of biodiversity (Ashton 2003Bruijnzeel et al 2010 Spracklen and Righelato 2014 Fa-hey et al 2015 Rahbeck et al 2019) TMFs occur alongsteep thermal and edaphic gradients making them attrac-tive for the study of abiotic controls of tropical-forest pro-ductivity and the role of tree diversity In this study in twoAndean elevation transects we combine biomass productiv-ity and tree diversity data with comprehensive soil chem-ical data to identify the most important abiotic and bioticdrivers of forest productivity in Neotropical montane forests(Fig 1) In permanent plots in old-growth forests we mea-sured coarse-wood production (WP) and fine-litter produc-tion (only one of the transects) and related these componentsof aboveground primary production to the availability of thefive macronutrients N P Ca K and Mg in the soil to el-evation as a proxy for temperature and to stand structuralproperties and tree diversity in the plots We chose a plot sizeof 004 ha (20 mtimes 20 m) in order to meet plot homogene-ity requirements in the rugged terrain Moreover assumeddiversity effects are more likely to emerge at small spatialscales (Chisholm et al 2013) This allowed us to investi-gate a larger number of plots across extended environmentalgradients (2000 m of elevation distance or sim 10 K variationin mean annual temperature large bedrock topographic andsoil variation) and to encounter considerable variation in treediversity (6 to 27 species per 004 ha) The analyses were re-stricted to old-growth forests with closed canopies in orderto exclude variation in stand structure and species composi-tion related to successional processes We employed struc-tural equation modeling for exploring mutual interrelation-






2 Methods

21 Study transects












23 Data analysis







3 Results















4 Discussion















WP


AGB









NPPa


AGB
















5 Conclusions











Variable PC1 PC2






Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References





2 Methods

21 Study transects












23 Data analysis







3 Results















4 Discussion















WP


AGB









NPPa


AGB
















5 Conclusions











Variable PC1 PC2






Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References





23 Data analysis







3 Results















4 Discussion















WP


AGB









NPPa


AGB
















5 Conclusions











Variable PC1 PC2






Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References












4 Discussion















WP


AGB









NPPa


AGB
















5 Conclusions











Variable PC1 PC2






Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References













WP


AGB









NPPa


AGB
















5 Conclusions











Variable PC1 PC2






Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References





NPPa


AGB
















5 Conclusions











Variable PC1 PC2






Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References









5 Conclusions











Variable PC1 PC2






Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References









Variable PC1 PC2






Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References




Variable PC1 PC2












References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References










References

































































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References



















































































Abstract

Introduction

Methods

Study transects


Data analysis

Results





Discussion

Conclusions



Data availability

Supplement


Competing interests

Acknowledgements

Financial support

Review statement

References

factors controlling the productivity of tropical andean

Documents