Hostname: page-component-74656f9884-8rzl6 Total loading time: 0 Render date: 2026-11-06T06:54:35.947Z Has data issue: false hasContentIssue false
  • English
  • Français

Changes in climate and vegetation with altitude on Mount Batilamu, Viti Levu, Fiji

Published online by Cambridge University Press:  10 September 2018

Jacynta Anderson ,
Gunnar Keppel Open the ORCID record for Gunnar Keppel [Opens in a new window] ,
Sophie-Min Thomson ,
Anthony Randell ,
Jone Raituva ,
Iliesa Koroi ,
Ramokasa Anisi ,
Tanya Charlson ,
Hans Juergen Boehmer  and
Sonia Kleindorfer
Jacynta Anderson
Affiliation:
Natural and Built Environments Research Centre, School of Natural and Built Environments, University of South Australia, GPO Box 2471, Adelaide, South Australia 5001, Australia
Gunnar Keppel*
Affiliation:
Natural and Built Environments Research Centre, School of Natural and Built Environments, University of South Australia, GPO Box 2471, Adelaide, South Australia 5001, Australia Future Industries Institute, University of South Australia, Mawson Lakes Campus, GPO Box 2471, Adelaide, South Australia 5001, Australia Biodiversity, Macroecology & Biogeography, Faculty of Forest Sciences, University of Goettingen, Büsgenweg 1, 37077 Göttingen, Germany
Sophie-Min Thomson
Affiliation:
Natural and Built Environments Research Centre, School of Natural and Built Environments, University of South Australia, GPO Box 2471, Adelaide, South Australia 5001, Australia
Anthony Randell
Affiliation:
Natural and Built Environments Research Centre, School of Natural and Built Environments, University of South Australia, GPO Box 2471, Adelaide, South Australia 5001, Australia
Jone Raituva
Affiliation:
South Pacific Regional Herbarium, Faculty of Science, Technology and Environment, University of the South Pacific, Suva, Fiji
Iliesa Koroi
Affiliation:
South Pacific Regional Herbarium, Faculty of Science, Technology and Environment, University of the South Pacific, Suva, Fiji
Ramokasa Anisi
Affiliation:
South Pacific Regional Herbarium, Faculty of Science, Technology and Environment, University of the South Pacific, Suva, Fiji
Tanya Charlson
Affiliation:
Natural and Built Environments Research Centre, School of Natural and Built Environments, University of South Australia, GPO Box 2471, Adelaide, South Australia 5001, Australia
Hans Juergen Boehmer
Affiliation:
School of Geography, Earth Science and Environment, Faculty of Science, Technology and Environment, University of the South Pacific, Suva, Fiji
Sonia Kleindorfer
Affiliation:
College of Science and Engineering, Flinders University, GPO Box 2100, Adelaide 5001, Australia
*
*Corresponding author. Email: gunnar.keppel@unisa.edu.au
Rights & Permissions [Opens in a new window]

Abstract:

To investigate changes in vegetation and climate with altitude, we established forest plots and recorded climatic data at 100-m intervals between 550–1100 m asl on the western slopes of Mount Batilamu, Mount Koroyanitu range, Viti Levu, Fiji. Trees with a dbh ≥10 cm were identified and measured in 21 10 × 10-m plots, starting at 750 m altitude. Temperature and relative humidity sensors were deployed in two habitats, leaf litter and 50 cm above the ground, and two vegetation types, grasslands and forest, at six altitudes over a 48-h period. Two significantly distinct forest types, lowland and montane, were present. Montane forest was found at higher elevations (>950 m asl) and had significantly higher stem density. Mean temperature decreased significantly with altitude and was strongly moderated by vegetation type (lower average and less variation in forest). While average relative humidity significantly increased with altitude, it was strongly moderated by both habitat and vegetation type (higher average and less variation in leaf litter and forest). The lapse rate varied with time of day (higher during the day) and vegetation type (higher in grasslands). Therefore, vegetation and microhabitats create unique microclimates, and this should be considered when investigating current or future climatic patterns along altitudinal gradients on forested mountains.

Keywords

altitudinal gradientendemismlowland rain forestmicroclimatemontane forestMount Koroyanitu National Heritage Parkrelative humiditySouth Pacifictemperaturetropical oceanic island

Information

Type
Research Article
Information
Journal of Tropical Ecology , Volume 34 , Issue 5 , September 2018 , pp. 316 - 325
Copyright
Copyright © Cambridge University Press 2018 

INTRODUCTION

As altitude increases, changes in temperature, rainfall, humidity, solar radiation and evapotranspiration occur (Strong et al. Reference STRONG, BOULTER, LAIDLAW, MAUNSELL, PUTLAND and KITCHING2011). Temperature decreases and humidity increases with altitude (Brehm et al. Reference BREHM, COLWELL and KLUGE2007, McVicar et al. Reference MCVICAR, VAN NIEL, LI, HUTCHINSON, MU and LIU2007, Shanks Reference SHANKS1954). The rate of temperature decline (the lapse rate) is usually about 0.5–0.6°C per 100 m (Osborne Reference OSBORNE2012). The lapse rate is influenced by season (McVicar et al. Reference MCVICAR, VAN NIEL, LI, HUTCHINSON, MU and LIU2007), time of the day (steeper during daytime; Pepin Reference PEPIN2001), and topography (e.g. effects of relative solar radiation, slope and proximity to streams; Lookingbill & Urban Reference LOOKINGBILL and URBAN2003). Different microhabitats may also modify temperature and humidity (Keppel et al. Reference KEPPEL, ANDERSON, WILLIAMS, KLEINDORFER and O'CONNELL2017a, Scheffers et al. Reference SCHEFFERS, EVANS, WILLIAMS and EDWARDS2014).

Altitudinal gradients in environmental parameters impact species distributions (Körner Reference KÖRNER2007, McCain & Grytnes Reference MCCAIN and GRYTNES2010, Strong et al. Reference STRONG, BOULTER, LAIDLAW, MAUNSELL, PUTLAND and KITCHING2011). As a result, the composition and structure of vegetation change along altitudinal gradients in response to climatic and environmental variables (McCain & Grytnes Reference MCCAIN and GRYTNES2010). These changes may be manifested in distinct vegetation zones at different altitudes, which have been variously classified (Ashton Reference ASHTON2003, Boehmer Reference BOEHMER, Brauch, Oswald Spring, Mesjasz, Grin, Kameri-Mbote, Chourou, Dunay and Birkmann2011, Richards Reference RICHARDS1996). An important transition between lowland and montane rain forest generally occurs between 800–1200 m asl (occasionally lower on islands), with montane forest being characterized by lower numbers of woody climbers, and a greater abundance of ground ferns, herbaceous angiosperms and epiphytes (Ashton Reference ASHTON2003, Boehmer Reference BOEHMER, Brauch, Oswald Spring, Mesjasz, Grin, Kameri-Mbote, Chourou, Dunay and Birkmann2011).

In addition to species composition, diversity and forest structure tend to change with altitude (Moser et al. Reference MOSER, RÖDERSTEIN, SOETHE, HERTEL, LEUSCHNER, Caldwell, Heldmaier, Jackson, Lange, Mooney, Schulze and Soomer2008, Richards Reference RICHARDS1996, Slik et al. Reference SLIK, AIBA, BREARLEY, CANNON, FORSHED, KITAYAMA, NAGAMASU, NILUS, PAYNE, PAOLI, POULSEN, RAES, SHEIL, SIDIYASA, SUZUKI and VAN VALKENBURG2010, Steinbauer et al. Reference STEINBAUER, FIELD, GRYTNES, TRIGAS, AH‐PENG, ATTORRE, BIRKS, BORGES, CARDOSO, CHOU, DE SANCTIS, DE SEQUEIRA, DUARTE, ELIAS, FERNÁNDEZ‐PALACIOS, GABRIEL, GEREAU, GILLESPIE, GREIMLER, HARTER, HUANG, IRL, JEANMONOD, JENTSCH, JUMP, KUEFFER, NOGUÉ, OTTO, PRICE, ROMEIRAS, STRASBERG, STUESSY, SVENNING, VETAAS and BEIERKUHNLEIN2016). Plant diversity usually decreases or displays a hump-shaped pattern with increasing altitude (Clark et al. Reference CLARK, HURTADO and SAATCHI2015, McCain & Grytnes Reference MCCAIN and GRYTNES2010, Rahbek Reference RAHBEK1995). Conversely, the proportion of endemic species increases with altitude, a trend believed to be caused by increased isolation promoting speciation (Steinbauer et al. Reference STEINBAUER, FIELD, GRYTNES, TRIGAS, AH‐PENG, ATTORRE, BIRKS, BORGES, CARDOSO, CHOU, DE SANCTIS, DE SEQUEIRA, DUARTE, ELIAS, FERNÁNDEZ‐PALACIOS, GABRIEL, GEREAU, GILLESPIE, GREIMLER, HARTER, HUANG, IRL, JEANMONOD, JENTSCH, JUMP, KUEFFER, NOGUÉ, OTTO, PRICE, ROMEIRAS, STRASBERG, STUESSY, SVENNING, VETAAS and BEIERKUHNLEIN2016, Trigas et al. Reference TRIGAS, PANITSA, TSIFTSIS and MOREAU2013). Stem density generally increases, canopy height decreases and basal area displays varying patterns with increasing altitude (Ibanez et al. Reference IBANEZ, MUNZINGER, DAGOSTINI, HEQUET, RIGAULT, JAFFRÉ and BIRNBAUM2014, Moser et al. Reference MOSER, RÖDERSTEIN, SOETHE, HERTEL, LEUSCHNER, Caldwell, Heldmaier, Jackson, Lange, Mooney, Schulze and Soomer2008, Slik et al. Reference SLIK, AIBA, BREARLEY, CANNON, FORSHED, KITAYAMA, NAGAMASU, NILUS, PAYNE, PAOLI, POULSEN, RAES, SHEIL, SIDIYASA, SUZUKI and VAN VALKENBURG2010). Similarly, lower tree height and higher stem densities at higher elevations are commonly observed in the Pacific (Ash Reference ASH1987, Ibanez et al. Reference IBANEZ, MUNZINGER, DAGOSTINI, HEQUET, RIGAULT, JAFFRÉ and BIRNBAUM2014, Kirkpatrick & Hassall Reference KIRKPATRICK and HASSALL1985).

Plant communities along mountain slopes are changing rapidly in response to anthropogenic climate change (Pauli et al. Reference PAULI, GOTTFRIED, DULLINGER, ABDALADZE, AKHALKATSI, ALONSO, COLDEA, DICK, ERSCHBAMER, CALZADO, GHOSN, HOLTEN, KANKA, KAZAKIS, KOLLÁR, LARSSON, MOISEEV, MOISEEV, MOLAU, MESA, NAGY, PELINO, PUŞCAŞ, ROSSI, STANISCI, SYVERHUSET, THEURILLAT, TOMASELLI, UNTERLUGGAUER, VILLAR, VITTOZ and GRABHERR2012, Penuelas & Boada Reference PENUELAS and BOADA2003). Increased global temperatures are causing plant species to move upslope as they track suitable environmental conditions (Gottfried et al. Reference GOTTFRIED, PAULI, FUTSCHIK, AKHALKATSI, BARANCOK, BENITO ALONSO, COLDEA, DICK, ERSCHBAMER, FERNANDEZ CALZADO, KAZAKIS, KRAJCI, LARSSON, MALLAUN, MICHELSEN, MOISEEV, MOISEEV, MOLAU, MERZOUKI, NAGY, NAKHUTSRISHVILI, PEDERSEN, PELINO, PUSCAS, ROSSI, STANISCI, THEURILLAT, TOMASELLI, VILLAR, VITTOZ, VOGIATZAKIS and GRABHERR2012, Parmesan Reference PARMESAN2006, Penuelas & Boada Reference PENUELAS and BOADA2003). Mountaintop species are thought to be particularly vulnerable to the effects of climate change due to their limited range (Boehmer Reference BOEHMER, Brauch, Oswald Spring, Mesjasz, Grin, Kameri-Mbote, Chourou, Dunay and Birkmann2011, Costion et al. Reference COSTION, SIMPSON, PERT, CARLSEN, JOHN and CRAYN2015, Keppel et al. Reference KEPPEL, ROBINSON, WARDELL-JOHNSON, YATES, VAN NIEL, BYRNE and SCHUT2017b, Parmesan Reference PARMESAN2006). However, we lack fine-scale climatic data on forest-covered tropical mountains, as most of our climatic data are derived from meteorological stations that are not shaded by vegetation (De Frenne & Verheyen Reference DE FRENNE and VERHEYEN2016). In the Pacific, basic ecological data, such as vegetation changes with altitude, are very limited (Keppel et al. Reference KEPPEL, TUIWAWA, NAIKATINI and ROUNDS2011).

Here we investigate the changes in climate and vegetation along an altitudinal gradient (500–1100 m asl) along the western slopes of Mount Batilamu in the Mount Koroyanitu Range, western Viti Levu, Fiji. Specifically, we predicted based on literature that with altitude, species composition would change; species richness, canopy height, tree volume and temperature would decrease; and endemism, stem density and relative humidity would increase.

STUDY SITE

Fiji originated through volcanic activity 30–40 million y ago (Neall & Trewick Reference NEALL and TREWICK2008). It consists of about 330 islands and is located in the seasonal tropics of the Pacific Ocean on the Australian and Pacific plate convergence boundary (Mueller-Dombois & Fosberg Reference MUELLER-DOMBOIS and FOSBERG1998). Fiji has a tropical climate with annual temperatures averaging 20–27°C, with a warmer wet season from November to April and a cooler dry season from May to October (Mataki et al. Reference MATAKI, KOSHY and LAL2006). Precipitation ranges from c. 1800 mm y−1 in the drier western part to about 3000 mm y−1 on the wetter eastern part of the archipelago, with ridges exposed to the south-east tradewinds receiving up to 10000 mm y−1 (Keppel & Tuiwawa Reference KEPPEL and TUIWAWA2007, Mataki et al. Reference MATAKI, KOSHY and LAL2006).

Fiji's vascular flora mostly originated from the Indo-Malesian and Australian regions (Keppel et al. Reference KEPPEL, LOWE and POSSINGHAM2009) and has c. 1315 native species (Heads Reference HEADS2006). The key natural vegetation types are lowland, lower montane and upper montane rain forests, mesic forest, tropical dry forest, mangroves, coastal forest and wetlands (Keppel & Tuiwawa Reference KEPPEL and TUIWAWA2007, Mueller-Dombois & Fosberg Reference MUELLER-DOMBOIS and FOSBERG1998). On the drier, western parts of the archipelago, grassland- and savanna-like talasiga (meaning ‘sunburnt’) vegetation has become dominant due to increased fire frequency since human settlement 3000–4000 BP (Ash Reference ASH1992, Mueller-Dombois & Fosberg Reference MUELLER-DOMBOIS and FOSBERG1998).

Mount Batilamu reaches about 1110 m asl in altitude and is located within the Koroyanitu National Heritage Park, which is a 25000-ha, community-managed protected area of conservation, ecotourism and experimental significance (Olson et al. Reference OLSON, FARLEY, PATRICK, WATLING, TUIWAWA, MASIBALAVU, LENOA, BOGIVA, QAUQAU, ATHERTON, CAGINITOBA, TOKOTA'A, PRASAD, NAISILISILI, RAIKABULA, MAILAUTOKA, MORLEY and ALLNUTT2010, Smith Reference SMITH1948, Thaman Reference THAMAN1996). The park includes Mount Koroyanitu (formerly Mount Evans), which, at 1195 m asl, is Viti Levu's third tallest peak (Smith Reference SMITH1948). Many of its over 700 plant species are culturally important and 11 species are endemic to the park, which includes lowland, lower and upper montane rain forests, talasiga and secondary vegetation (Thaman Reference THAMAN1996).

METHODS

Vegetation sampling and analyses

In July 2016, vegetation was sampled along a trail ascending the western slopes of Mount Batilamu (Figure 1), which was forested from ~700 m asl. Plots were established at ~750, 800, 900, 1000 and 1100 m asl, each measuring 10 × 10 m. Because the tree flora of Fijian rain forests can be diverse (Keppel et al. Reference KEPPEL, BUCKLEY and POSSINGHAM2010), we established five plots at each altitudinal band except at 750 and 800 m, where only three plots could be established due to very steep (>60° slopes) terrain. This resulted in a total of 21 plots, ranging in altitude from 751 to 1096 m asl (Figure 2). The four corners of each plot were permanently marked with 50-cm-long PVC pipes. Altitude and plot location were recorded with a Garmin GPSmap 62 GPS.

Figure 1.

Location of Fiji in the South-west Pacific (a), Mount Koroyanitu on Viti Levu Island in Fiji (b), and the study plots along the western slopes of Mount Batilamu in Mount Koroyanitu National Park (c).

Figure 2.

Non-metric MDS using the Jaccard similarity index, 21 plots along an altitudinal transect on Mount Batilamu, Viti Levu, Fiji. The plots are labelled with the altitude at which they are located.

In each plot, the following data were collected: the identity (name), dbh measured at 1.3 m above the ground, and height of each individual tree with a dbh ≥10 cm. For trees of unknown identity, herbarium-type specimens were collected, preserved in a 70% alcohol solution, pressed, transported to the South Pacific Regional Herbarium (SUVA) at the University of the South Pacific (USP), and dried at 50°C. Specimens were identified using the Flora Vitiensis Nova (Smith Reference SMITH1979–Reference SMITH1991) and herbarium collections at SUVA.

All analyses were undertaken in the R version 3.4.2 statistical computing environment (https://www.r-project.org/) using ‘ggplot’, ‘ggpubr’, ‘lme4’, ‘MASS’, ‘MuMin’ and ‘vegan’ packages. To investigate the species composition of the plots, non-metric multidimensional scaling (nMDS) was performed using the Jaccard similarity index to test for distinct forest types. A one-way Analysis of Similarity (ANOSIM; Clarke Reference CLARKE1993) using 95% (α = 0.05) confidence intervals was used to test for significant differences between forest types identified.

For each plot the following variables were measured or calculated: species richness (number of species per stem, to account for differences in stem density), endemism (proportion of species endemic to Fiji), stem density (total number of stems per plot), height (average of the five tallest trees) and volume. Tree volume was calculated assuming a conical tree shape: volume = [(dbh ÷ 2)2 × π × tree height] ÷ 3. A conical shape was assumed because it provides more conservative volume estimates than assuming a cylindrical shape, and because there were no local or regional equations for estimating volume (Magnussen & Reed Reference MAGNUSSEN and REED2015).

In addition, the relative density (number of individuals of a species ÷ total number of individuals × 100), dominance (volume of a species ÷ total volume × 100) and frequency (number of plots a species occurs in ÷ sum of frequency for all species × 100) were calculated for each species in each of two forest types. The sum of the relative density, dominance and frequency constitutes the importance value of a species (Mueller-Dombois & Ellenberg Reference MUELLER-DOMBOIS and ELLENBERG2002), which was converted to relative importance (sum of all importance values ÷ 3).

Based on visual data inspection and the small sample sizes employed, we assumed that the data collected were not normally distributed. Therefore, non-parametric tests were used; Kruskal–Wallis tests to determine significance differences between variables at the various altitudes, a Wilcoxon test to determine which altitudes or forest types differed significantly, and a Spearman rs test to identify correlations between the various variables and altitude.

Microclimate sampling and analysis

Paired Maxim iButtons (DS1923) were placed at altitudinal intervals of about 100 m, recording instantaneous readings of temperature and relative humidity every 20 min. One sensor was suspended 50 cm above the ground on bamboo sticks on the inside of inverted, white plastic cups, which were covered with white duct tape on the upper half. The other sensor was positioned inside leaf litter by placing it inside a metal tea strainer, which was covered with white duct tape on the upper (and upward-facing) half and left uncovered in the lower half (Keppel et al. Reference KEPPEL, ANDERSON, WILLIAMS, KLEINDORFER and O'CONNELL2017a). Where feasible the paired iButtons were placed in both forested and grassland vegetation at each altitude in close proximity, resulting in four iButtons per altitude; one each in the air in grasslands, leaf litter in grasslands, air in forest, and leaf litter in forest. The iButtons were mostly placed close to the plots, but we also placed a quadruplet of sensors in grassland and a forested gully at ~550 m asl.

A 48-h data window (07h00 on 19 July–06h40 on 21 July, sunshine with little cloud cover was observed during this period) was used for analyses. For this time period we calculated the average, standard deviation, maximum and minimum values of temperature and humidity for each iButton, separately for day and night time. Day and night time periods were determined by using sunrise and sunset data from timeanddate.com for the nearby city of Lautoka. The hour prior to and after sunrise and sunset were removed in order to allow a settling period between the two phases.

For both temperature and humidity, we calculated the mean, standard deviation, maximum and minimum values, separately for day- and night-time. We then used these as response variables in generalized linear mixed-effect models, with altitude as the sole fixed effect and vegetation type (open grasslands versus forest), habitat type (air versus leaf litter) and time of the day (day versus night) as random effects. We extracted the slope of the fixed effect (i.e. the lapse rate) and the percentage variance explained by each random effect from this model. We determined the significance of the fixed effect by comparing the performance of the model with and without altitude using a chi-squared test. The explanatory power of selected models was assessed using a pseudo-r 2-value (Nakagawa & Schielzeth Reference NAKAGAWA and SCHIELZETH2013), which calculates the marginal (variation explained by fixed effects) and conditional r 2 (variation explained by fixed and random effects).

RESULTS

Vegetation

Overall, there were 292 stems belonging to 71 species (in 37 families) of trees with a dbh ≥10 cm in the 21 plots. Of these, 43 species (60.56%) were endemic to Fiji. Two seemingly distinct plot clusters were produced by non-metric multidimensional scaling (nMDS; Figure 2). One group (montane plots) consisted of 10 plots at higher (>950 m) altitudes. The other group (lowland plots) consisted of plots at lower (<885 m) altitudes. The plots at 888 and 795 m had intermediate species composition, including taxa otherwise associated with higher altitudes (e.g. Podocarpus neriifolius, Podocarpaceae). The plot at 751 m was dominated by Pterocymbium oceanicum (Sterculiaceae) and was the only plot with this species. The one-way analysis of similarity (ANOSIM) supported montane and lowland plots to be significantly different (P = 0.001, r = 0.622).

The lowland plots were dominated by Dendrocnide harveyi (Urticaceae), Dysoxylum aliquantulum (Meliaceae) and Bischofia javanica (Euphorbiaceae). Together, these species accounted for more than a third of all stems and more than 60% of the total basal area, while also having the highest importance values (Table 1). The montane plots were dominated by Agathis macrophyllla (Araucariaceae), which accounted for more than 10% of all stems and almost 50% of the total tree volume. Of the 17 most important species in this study (based on being among the 10 most important species in either the lowland or montane plots), only six species occurred in both forest types (Table 1).

Table 1.

Importance values of the 10 tree species with the highest importance value in either lowland or montane forest along an altitudinal transect on Mount Batilamu, Fiji. RDe = relative density (%), RDo = relative dominance (%), RF = relative frequency (%), IV = importance value (%).

Measures of species diversity (richness, endemism) and forest structure (density, height, volume) did not differ significantly among altitudes, except for volume (Kruskal–Wallis, P = 0.027). Tree volume at 1000 m asl was significantly smaller than at 750 m (P = 0.037), 900 m (P = 0.012) and 1100 m (P = 0.012). Only canopy height had a significant, linear correlation with altitude (Spearman rs, ρ = −0.9, P = 0.017). The measures of species diversity and forest structure generally displayed predicted differences between the two forest types, i.e. higher species richness and greater canopy height and tree volume in lowland forest, and higher endemism and stem density in montane forest. However, those differences were only significant for stem density (P = 0.0048).

Climate

Models consisting of altitude, vegetation type, habitat and time were good predictors of the temperature and humidity variables investigated, explaining 62–85% of the total variation (conditional r2). Altitude was a significant predictor of the average, maximum and minimum temperature, explaining about 15%, 8% and 30% of the variation in these variables, respectively (marginal r2). Time of day and vegetation type were other important predictors of these variables. Daytime temperature was generally lower in forests than in grasslands (black versus white shapes, Figure 3). The variation (standard deviation) in temperature was mostly driven by time of day (lower variation at night) and vegetation type (lower variation in forests).

Figure 3.

Variation of average (19–20 July 2016) daytime temperature (a) and relative humidity (b) along an altitudinal gradient on Mount Batilamu, Fiji.

The average lapse rate was c. −0.9°C per 100 m, being higher during the day (−1.1°C per 100 m) than at night (−0.6°C per 100 m). Comparing the different microhabitats, the lapse rate was higher in grasslands (−1.15°C per 100 m in the leaf litter and −0.88°C per 100 m in the open) than in the forest (−0.86°C per 100 m in the open air and −0.64°C per 100 m in leaf litter). From about 900 m (the break between the two forest types), the average temperature did not exceed 19°C and the average relative humidity did not drop below 75%.

Altitude was a significant predictor of the average and minimum humidity (though considerably less so than for temperature), explaining about 4% and 3% of the variation in these variables, respectively (marginal r2). Habitat type, time of the day and, to a lesser extent, vegetation type were other important predictors of these variables. Daytime humidity indeed tended to be higher in leaf litter (squares) than in the air (50 cm above the ground; triangles; Figure 3). The variation (standard deviation) in humidity was mostly driven by time of day (lower variation at night), with habitat type (lower variation in leaf litter) and vegetation type (lower variation in forests) also being important. Only habitat type was a good predictor (explaining almost 60% of the total variation in this variable) of maximum humidity, with leaf litter having consistently higher maximum humidity.

DISCUSSION

The two major forest types identified along the altitudinal transect on the western slopes of Mount Batilamu correspond to lowland and montane rain forest, with the transition occurring between 800–900 m asl and corresponding with the altitude suggested for the break between these forest types (Ashton Reference ASHTON2003). The montane forest appears to be the Agathis–Podocarpus type described by Mueller-Dombois & Fosberg (Reference MUELLER-DOMBOIS and FOSBERG1998; as upland rain forest), with the conifer Agathis macrophylla as the dominant element. The montane forest is lower montane forest sensu Hamilton et al. (Reference HAMILTON, JUVIK and SCATENA1995) and Ashton (Reference ASHTON2003), as it is different from cloud forest, lacking trees with stunted growth and twisted trunks. The lowland rain forest has common lowland rain-forest elements, such as Dysoxylum spp., Bischofia javanica, Hedycarya dorstenioides and Viticipremna vitilevuensis (Keppel et al. Reference KEPPEL, TUIWAWA, NAIKATINI and ROUNDS2011). Other altitudinal transects in Fiji have similarly found turnover of dominant species with altitude (Ash Reference ASH1987, Kirkpatrick & Hassall Reference KIRKPATRICK and HASSALL1985).

Although the variables related to rain-forest diversity and structure tended to differ between lowland and montane forest as expected based on literature, the differences were more subdued than expected – with only those in stem density being significant. The limited sample size and altitudinal range of this study may have limited our ability to detect differences. Furthermore, we used the number of species endemic to Fiji (as there were no species endemic to the Mount Koroyanitu Range in our plots) and this may not respond strongly to altitude, as the percentage of endemic tree species in Fijiś lowland rain forests exceeds 50% (Keppel et al. Reference KEPPEL, BUCKLEY and POSSINGHAM2010).

Decreasing temperature and increasing relative humidity with altitude are well documented (Grubb & Whitmore Reference GRUBB and WHITMORE1966, Pepin & Losleben Reference PEPIN and LOSLEBEN2002, Shanks Reference SHANKS1954). Our lapse rates are higher than the usual average of about 0.5–0.6°C per 100 m (Osborne Reference OSBORNE2012). However, the altitudinal range that we used to calculate lapse rates was comparatively small, which can affect results (Pepin & Losleben Reference PEPIN and LOSLEBEN2002). Furthermore, lapse rates are generally calculated as the average over an entire year and our data are based on a relatively short time period during which there was little cloud cover. Lapse rates are known to differ among seasons and with cloud cover (McVicar et al. Reference MCVICAR, VAN NIEL, LI, HUTCHINSON, MU and LIU2007, Pepin Reference PEPIN2001, Pepin & Losleben Reference PEPIN and LOSLEBEN2002) and we here show that they also differ among vegetation types and in different microhabitats.

Both canopy cover and microhabitats are known to moderate microclimate (Keppel et al. Reference KEPPEL, ANDERSON, WILLIAMS, KLEINDORFER and O'CONNELL2017a, Scheffers et al. Reference SCHEFFERS, EVANS, WILLIAMS and EDWARDS2014). In our study, the absence of canopy cover strongly increased temperature and decreased relative humidity. These microclimatic differences between forests and more open vegetation, such as grasslands, savannas and canopy gaps, are well documented (Grubb & Whitmore Reference GRUBB and WHITMORE1966, Holl Reference HOLL1999, Ibanez et al. Reference IBANEZ, HELY and GAUCHEREL2013, Turton & Sexton Reference TURTON and SEXTON1996). The leaf-litter microhabitat strongly moderated climatic variables, especially relative humidity. In addition, topographic variation can strongly affect microclimate measurements (Lookingbill & Urban Reference LOOKINGBILL and URBAN2003). While we minimized the effects of aspect, which can be significant, we could not include other important effects, such as solar radiation and distance to a stream (Lookingbill & Urban Reference LOOKINGBILL and URBAN2003), as fine-scale digital elevation models (DEMs) are not available for the area.

Only recently has the ability of microhabitats to retain microclimates that are more stable than external conditions been quantified and its potential importance under anthropogenic climate change been realized (Keppel et al. Reference KEPPEL, ANDERSON, WILLIAMS, KLEINDORFER and O'CONNELL2017a, Scheffers et al. Reference SCHEFFERS, EVANS, WILLIAMS and EDWARDS2014). As such microhabitats can maintain more consistent and favourable climates and, therefore, they are likely to be important in facilitating the persistence of biota as regional environmental conditions change (Lenoir et al. Reference LENOIR, HATTAB and PIERRE2017, Scheffers et al. Reference SCHEFFERS, EVANS, WILLIAMS and EDWARDS2014). Our study shows that vegetation and microhabitats interact to create a patchwork of microclimates (Keppel et al. Reference KEPPEL, ANDERSON, WILLIAMS, KLEINDORFER and O'CONNELL2017a), which is important for facilitating the persistence of species under forecast anthropogenic climate change, as species respond to this fine-scale interplay of abiotic and biotic factors (Keppel et al. Reference KEPPEL, ROBINSON, WARDELL-JOHNSON, YATES, VAN NIEL, BYRNE and SCHUT2017b, Oorebeek & Kleindorfer Reference OOREBEEK and KLEINDORFER2008).

Our findings therefore highlight the importance of including the moderating effects of vegetation and microhabitats on climatic conditions when forecasting future changes in climate and species distributions. This even holds true for the tropical South-West Pacific islands, which are experiencing milder climate change than higher latitudes (IPCC 2014). However, climatic data are currently mostly derived from meteorological stations in open areas without tall vegetation and we therefore cannot currently quantify the effects of canopy cover on microclimate (De Frenne & Verheyen Reference DE FRENNE and VERHEYEN2016).

ACKNOWLEDGEMENTS

We would like to thank Abaca Village, Vuda District, Ba Province, for kindly giving us permission for, and providing assistance, during the fieldwork. Geon C. Hanson and Leomar Longsworth assisted with data collection. This research was undertaken as part of a joint undergraduate student project between Flinders University, the University of South Australia and the University of the South Pacific, and funded through the Department of Foreign Affairs and Trade of the Australian Government as part of the New Colombo Plan initiative. During part of the analyses and writing GK was supported by an Alexander von Humboldt fellowship. A study tour travel grant by the University of South Australia assisted JA, SMT, AR and TC.

Footnotes

§

Authors contributed equally to this paper.

References

LITERATURE CITED

ASH, J. 1987. Stunted cloud-forest in Taveuni, Fiji. Pacific Science 41:191–199.Google Scholar
ASH, J. 1992. Vegetation ecology of Fiji: past, present, and future perspectives. Pacific Science 46:111–127.Google Scholar
ASHTON, P. S. 2003. Floristic zonation of tree communities on wet tropical mountains revisited. Perspectives in Plant Ecology, Evolution and Systematics 6:87–104.Google Scholar
BOEHMER, H. J. 2011. Vulnerability of tropical montane rain forest ecosystems due to climate change. Pp. 789–802 in Brauch, H. G., Oswald Spring, Ú., Mesjasz, C., Grin, J., Kameri-Mbote, P., Chourou, B., Dunay, P. & Birkmann, J. (eds). Coping with global environmental change, disasters and security – threats, challenges, vulnerabilities and risks. Springer, Berlin.Google Scholar
BREHM, G., COLWELL, R. K. & KLUGE, J. 2007. The role of environment and mid-domain effect on moth species richness along a tropical elevational gradient. Global Ecology and Biogeography 16: 205–219.Google Scholar
CLARK, D. B., HURTADO, J. & SAATCHI, S. S. 2015. Tropical rain forest structure, tree growth and dynamics along a 2700-m elevational transect in Costa Rica. PLoS ONE 10:e0122905.Google Scholar
CLARKE, K. R. 1993. Non-parametric multivariate analyses of changes in community structure. Journal of Ecology 18:117–143.Google Scholar
COSTION, C. M., SIMPSON, L., PERT, P. L., CARLSEN, M. M., JOHN, K. W. & CRAYN, D. 2015. Will tropical mountaintop plant species survive climate change? Identifying key knowledge gaps using species distribution modelling in Australia. Biological Conservation 191:322–330.Google Scholar
DE FRENNE, P. & VERHEYEN, K. 2016. Weather stations lack forest data. Science 351:234.Google Scholar
GOTTFRIED, M., PAULI, H., FUTSCHIK, A., AKHALKATSI, M., BARANCOK, P., BENITO ALONSO, J. L., COLDEA, G., DICK, J., ERSCHBAMER, B., FERNANDEZ CALZADO, M. R., KAZAKIS, G., KRAJCI, J., LARSSON, P., MALLAUN, M., MICHELSEN, O., MOISEEV, D., MOISEEV, P., MOLAU, U., MERZOUKI, A., NAGY, L., NAKHUTSRISHVILI, G., PEDERSEN, B., PELINO, G., PUSCAS, M., ROSSI, G., STANISCI, A., THEURILLAT, J.-P., TOMASELLI, M., VILLAR, L., VITTOZ, P., VOGIATZAKIS, I. & GRABHERR, G. 2012. Continent-wide response of mountain vegetation to climate change. Nature Climate Change 2:111–115.Google Scholar
GRUBB, P. J. & WHITMORE, T. C. 1966. A comparison of montane and lowland rain forest in Ecuador: II. The climate and its effects on the distribution and physiognomy of the Forests. Journal of Ecology 54:303–333.Google Scholar
HEADS, M. 2006. Seed plants of Fiji: an ecological analysis. Biological Journal of the Linnean Society 89:407–431.Google Scholar
HAMILTON, L. S., JUVIK, J. O. & SCATENA, F. N. (eds). 1995. Tropical montane cloud forests. Springer, Berlin.Google Scholar
HOLL, K. D. 1999. Factors limiting tropical rain forest regeneration in abandoned pasture: seed rain, seed germination, microclimate, and soil. Biotropica 31:229–242.Google Scholar
IBANEZ, T., HELY, C. & GAUCHEREL, C. 2013. Sharp transitions in microclimatic conditions between savanna and forest in New Caledonia: insights into the vulnerability of forest edges to fire. Austral Ecology 38:680–687.Google Scholar
IBANEZ, T., MUNZINGER, J., DAGOSTINI, G., HEQUET, V., RIGAULT, F., JAFFRÉ, T. & BIRNBAUM, P. 2014. Structural and floristic diversity of mixed tropical rain forest in New Caledonia: new data from the New Caledonian Plant Inventory and Permanent Plot Network (NC‐PIPPN). Applied Vegetation Science 17:386–397.Google Scholar
IPCC. 2014. Climate Change 2014: Synthesis Report. Fifth Assessment Report of the Intergovernmental Panel on Climate Change. IPCC, Geneva.Google Scholar
KEPPEL, G. & TUIWAWA, M. V. 2007. Dry zone forests of Fiji: species composition, life history traits, and conservation. New Zealand Journal of Botany 45:545–563.Google Scholar
KEPPEL, G., LOWE, A. J. & POSSINGHAM, H. P. 2009. Changing perspectives on the biogeography of the tropical South Pacific: influences of dispersal, vicariance and extinction. Journal of Biogeography 36:1035–1054.Google Scholar
KEPPEL, G., BUCKLEY, Y. M. & POSSINGHAM, H. P. 2010. Drivers of lowland rain forest community assembly, species diversity and forest structure on islands in the tropical South Pacific. Journal of Ecology 98:87–95.Google Scholar
KEPPEL, G., TUIWAWA, M. V., NAIKATINI, A. & ROUNDS, I. A. 2011. Microhabitat specialization of tropical rain-forest canopy trees in the Sovi Basin, Viti Levu, Fiji Islands. Journal of Tropical Ecology 27:491–501.Google Scholar
KEPPEL, G., ANDERSON, S., WILLIAMS, C., KLEINDORFER, S. & O'CONNELL, C. 2017a. Microhabitats and canopy cover moderate high summer temperatures in a fragmented Mediterranean landscape. PLoS ONE 12:e0183106.Google Scholar
KEPPEL, G., ROBINSON, T. P., WARDELL-JOHNSON, G. W., YATES, C. J., VAN NIEL, K. P., BYRNE, M. & SCHUT, A. G. T. 2017b. A low-altitude mountain range as an important refugium for two narrow endemics in the Southwest Australian Floristic Region biodiversity hotspot. Annals of Botany 119:289–300.Google Scholar
KIRKPATRICK, J. B. & HASSALL, D. C. 1985. The vegetation and flora along an altitudinal transect through tropical forest at Mount Korobaba, Fiji. New Zealand Journal of Botany 23: 33–46.Google Scholar
KÖRNER, C. 2007. The use of ‘altitude’ in ecological research. Trends in Ecology and Evolution 22:569.Google Scholar
LENOIR, J., HATTAB, T. & PIERRE, G. 2017. Climatic microrefugia under anthropogenic climate change: implications for species redistribution. Ecography 40:253–266.Google Scholar
LOOKINGBILL, T. R. & URBAN, D. L. 2003. Spatial estimation of air temperature differences for landscape-scale studies in montane environments. Agricultural and Forest Meteorology 114: 141–151.Google Scholar
MAGNUSSEN, S. & REED, D. 2015. Modelling for estimation and monitoring. Pp. 111–136 in Knowledge reference for national forest assessments. Food and Agriculture Organisation of the United Nations, Rome.Google Scholar
MATAKI, M., KOSHY, K. C. & LAL, M. 2006. Baseline climatology of Viti Levu (Fiji) and current climatic trends. Pacific Science 60: 49–68.Google Scholar
MCCAIN, C. M. & GRYTNES, J.-A. 2010. Elevational gradients in species richness. Pp. 1–10 in Encyclopedia of life sciences. John Wiley & Sons, Chichester.Google Scholar
MCVICAR, T. R., VAN NIEL, T. G., LI, L., HUTCHINSON, M. F., MU, X. & LIU, Z. 2007. Spatially distributing monthly reference evapotranspiration and pan evaporation considering topographic influences. Journal of Hydrology 338:196–220.Google Scholar
MOSER, G., RÖDERSTEIN, M., SOETHE, N., HERTEL, D. & LEUSCHNER, C. 2008. Altitudinal changes in stand structure and biomass allocation of tropical mountain forests in relation to microclimate and soil chemistry. Pp. 229–242 in Caldwell, M. M., Heldmaier, G., Jackson, R. B., Lange, O. L., Mooney, H. A., Schulze, E. D. & Soomer, U. (eds). Gradients in a tropical mountain ecosystem of Ecuador. Springer, Berlin.Google Scholar
MUELLER-DOMBOIS, D. & ELLENBERG, H. 2002. Aims and methods of vegetation ecology. Blackburn Press, Caldwell. 547 pp.Google Scholar
MUELLER-DOMBOIS, D. & FOSBERG, F. R. 1998. Vegetation of the Tropical Pacific Islands. Springer, New York. 733 pp.Google Scholar
NAKAGAWA, S. & SCHIELZETH, H. 2013. A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods in Ecology and Evolution 4:133–142.Google Scholar
NEALL, V. E. & TREWICK, S. A. 2008. The age and origin of the Pacific islands: a geological overview. Philosophical Transactions of the Royal Society of London, Series B, Biological Sciences 363: 3293–3308.Google Scholar
OOREBEEK, M. & KLEINDORFER, S. 2008. Understorey predicts the spatial distribution of Ixodes hirsti in South Australia. Australian Journal of Zoology 56:123–127.Google Scholar
OLSON, D., FARLEY, L., PATRICK, A., WATLING, D., TUIWAWA, M., MASIBALAVU, V., LENOA, L., BOGIVA, A., QAUQAU, I., ATHERTON, J., CAGINITOBA, A., TOKOTA'A, M., PRASAD, S., NAISILISILI, W., RAIKABULA, A., MAILAUTOKA, K., MORLEY, C. & ALLNUTT, T. 2010. Priority forests for conservation in Fiji: landscapes, hotspots and ecological processes. Oryx 44:57–70.Google Scholar
OSBORNE, P. L. 2012. Tropical ecosystems and ecological concepts. (Second edition). Cambridge University Press, Cambridge.Google Scholar
PARMESAN, C. 2006. Ecological and evolutionary responses to recent climate change. Annual Review of Ecology, Evolution, and Systematics 37:637–669.Google Scholar
PAULI, H., GOTTFRIED, M., DULLINGER, S., ABDALADZE, O., AKHALKATSI, M., ALONSO, J. L. B., COLDEA, G., DICK, J., ERSCHBAMER, B., CALZADO, R. F., GHOSN, D., HOLTEN, J. I., KANKA, R., KAZAKIS, G., KOLLÁR, J., LARSSON, P., MOISEEV, P., MOISEEV, D., MOLAU, U., MESA, J. M., NAGY, L., PELINO, G., PUŞCAŞ, M., ROSSI, G., STANISCI, A., SYVERHUSET, A. O., THEURILLAT, J.-P., TOMASELLI, M., UNTERLUGGAUER, P., VILLAR, L., VITTOZ, P. & GRABHERR, G. 2012. Recent plant diversity changes on Europe's mountain summits. Science 336: 353–355.Google Scholar
PENUELAS, J. & BOADA, M. 2003. A global change-induced biome shift in the Montseny Mountains (NE Spain). Global Change Biology 9:131–131.Google Scholar
PEPIN, N. 2001. Lapse rate changes in northern England. Theoretical and Applied Climatology 68:1–16.Google Scholar
PEPIN, N. & LOSLEBEN, M. 2002. Climate change in the Colorado Rocky Mountains: free air versus surface temperature trends. International Journal of Climatology 22:311–329.Google Scholar
RAHBEK, C. 1995. The elevational gradient of species richness: a uniform pattern? Ecography 18:200–205.Google Scholar
RICHARDS, P. W. 1996. The tropical rain forest. An ecological study. Cambridge University Press, Cambridge.Google Scholar
SCHEFFERS, B. R., EVANS, T. A., WILLIAMS, S. E. & EDWARDS, D. P. 2014. Microhabitats in the tropics buffer temperature in a globally coherent manner. Biology Letters 10:20140819.Google Scholar
SHANKS, R. E. 1954. Climates of the Great Smoky Mountains. Ecology 35:354–361.Google Scholar
SLIK, W. F., AIBA, S., BREARLEY, F. Q., CANNON, C. H., FORSHED, O., KITAYAMA, K., NAGAMASU, H., NILUS, R., PAYNE, J., PAOLI, G., POULSEN, A. D., RAES, N., SHEIL, D., SIDIYASA, K., SUZUKI, E. & VAN VALKENBURG, J. L. C. H. 2010. Environmental correlates of tree biomass, basal area, wood specific gravity and stem density gradients in Borneo's tropical forests. Global Ecology and Biogeography 19:50–60.Google Scholar
SMITH, A. C. 1948. An ascent of Koroyanitu. Proceedings of the National Academy of Sciences USA 34:579–585.Google Scholar
SMITH, A.C. 1979–1991. Flora Vitiensis nova: a new flora of Fiji (spermatophytes only). Volumes 1–5. Pacific Tropical Botanical Garden, Lawai, Hawai'i.Google Scholar
STEINBAUER, M. J., FIELD, R., GRYTNES, J. A., TRIGAS, P., AH‐PENG, C., ATTORRE, F., BIRKS, H. J. B., BORGES, P. A. V., CARDOSO, P., CHOU, C. H., DE SANCTIS, M., DE SEQUEIRA, M. M., DUARTE, M. C., ELIAS, R. B., FERNÁNDEZ‐PALACIOS, J. M., GABRIEL, R., GEREAU, R. E., GILLESPIE, R. G., GREIMLER, J., HARTER, D. E. V., HUANG, T. J., IRL, S. D. H., JEANMONOD, D., JENTSCH, A., JUMP, A. S., KUEFFER, C., NOGUÉ, S., OTTO, R., PRICE, J., ROMEIRAS, M. M., STRASBERG, D., STUESSY, T., SVENNING, J. C., VETAAS, O. R. & BEIERKUHNLEIN, C. 2016. Topography‐driven isolation, speciation and a global increase of endemism with elevation. Global Ecology and Biogeography 25:1097–1107.Google Scholar
STRONG, C. L., BOULTER, S. L., LAIDLAW, M. J., MAUNSELL, S. C., PUTLAND, D. & KITCHING, R. L. 2011. The physical environment of an altitudinal gradient in the rainforest of Lamington National Park, southeast Queensland. Memoirs of the Queensland Museum 55:251–270.Google Scholar
THAMAN, R. 1996. The biodiversity of Koroyanitu National Park. Domodomo 10:28–51.Google Scholar
TRIGAS, P., PANITSA, M., TSIFTSIS, S. & MOREAU, C. S. 2013. Elevational gradient of vascular plant species richness and endemism in Crete – the effect of post-isolation mountain uplift on a continental island system. PLoS ONE 8:e59425.Google Scholar
TURTON, S. M. & SEXTON, G. J. 1996. Environmental gradients across four rainforest-open forest boundaries in northeastern Queensland. Austral Ecology 21:245–254.Google Scholar
Figure 0

Figure 1. Location of Fiji in the South-west Pacific (a), Mount Koroyanitu on Viti Levu Island in Fiji (b), and the study plots along the western slopes of Mount Batilamu in Mount Koroyanitu National Park (c).

Figure 1

Figure 2. Non-metric MDS using the Jaccard similarity index, 21 plots along an altitudinal transect on Mount Batilamu, Viti Levu, Fiji. The plots are labelled with the altitude at which they are located.

Figure 2

Table 1. Importance values of the 10 tree species with the highest importance value in either lowland or montane forest along an altitudinal transect on Mount Batilamu, Fiji. RDe = relative density (%), RDo = relative dominance (%), RF = relative frequency (%), IV = importance value (%).

Figure 3

Figure 3. Variation of average (19–20 July 2016) daytime temperature (a) and relative humidity (b) along an altitudinal gradient on Mount Batilamu, Fiji.