Increased night-time temperatures, through their influence on dark respiration, have been implicated as a reason behind decreasing growth rates in tropical trees in the face of contemporary climate change.
Seedlings of two neo-tropical tree species (Ficus insipida and Ochroma pyramidale) were grown in controlled-environment chambers at a constant daytime temperature (33°C) and a range of increasing night-time temperatures (22, 25, 28, 31°C) for between 39 d and 54 d. Temperature regimes were selected to represent a realistic baseline condition for lowland Panama, and a rise in night-time temperatures far in excess of those predicted for Central America in the coming decades. Experiments were complemented by an outdoor open-top chamber study in which night-time temperatures were elevated by 2.4°C above ambient.
Increasing night-time temperatures resulted in > 2-fold increase in biomass accumulation in growth-chamber studies despite an increase in leaf-level dark respiration. Similar trends were seen in open-top chambers, in which elevated night-time temperatures resulted in stimulation of growth.
These findings challenge simplistic considerations of photosynthesis-directed growth, highlighting the role of temperature-dependent night-time processes, including respiration and leaf development as drivers of plant performance in the tropics.
Contemporary global warming has been accompanied by the narrowing of the diel (24 h) temperature range by the asymmetric rise in night-time and daytime temperatures (Kukla & Karl, 1993; Easterling et al., 1997). The complex interaction of changing atmospheric chemistry, water-vapor feedback, and cloud behavior is difficult to predict (Dai et al., 1997; Lobell et al., 2007), yet it is likely that the emergence of novel temperature regimes in the tropics (Diffenbaugh & Scherer, 2011) will include the continued asymmetric rise of night-time temperatures (IPCC, 2007), with potentially profound implications upon plant growth in an already compromised ecosystem (Wright, 2010).
Several recent meta-analysis of plant responses to increasing temperature (Lin et al., 2010; Way & Oren, 2010) have highlighted how tropical species may be particularly vulnerable to increases in both daytime and night-time temperatures. In temperate and boreal systems productivity is often limited by seasonal minima, and an increase in temperature may be associated with the extension of the growing season or a shift towards a temperature optimum for growth. By contrast, humid tropical (lowland) systems with their relatively stable climatic envelope (Wright et al., 2009) are presumed to operate at, or close to their thermal optimum (Janzen, 1967; Ghalambor et al., 2006) and may possess only limited tolerance to increasing temperatures (Colwell et al., 2008; Laurance et al., 2011; Corlett, 2012).
Dark respiration (Rd) (i.e. non-photorespiratory mitochondrial respiration) provides cells with metabolic precursors, usable energy (i.e. ATP), and reducing equivalents (e.g. NADH) (Ryan, 1991; Amthor, 2000). As in all organisms, Rd of tropical trees is known to rise in response to short-term increases in temperature (Cavaleri et al., 2008) and, as such, the elevated respiratory-carbon loss at higher growth temperatures has been invoked in the explanation of empirical (Clark et al., 2003, 2010; Loescher et al., 2003; Larjavaara & Muller-Landau, 2012) and predicted (Cox et al., 2000; Cowling & Shin, 2006) decreases in productivity of tropical forests with changing climate. However, this interpretation precludes long-term acclimatization of Rd (Atkin & Tjoelker, 2003; Atkin et al., 2005) and is predicated upon carbon-source dynamics, the net balance between photosynthesis (A) and Rd determining plant growth. This framework does not address potential changes in carbon use efficiency (Hansen et al., 2009) or phenological development (Wolfe-Bellin et al., 2006), the highly dynamic nature of thermal responses seen in Rd (Atkin et al., 2005), or how plant metabolic rates may be sink rather than source mediated (Körner, 2003; Sala et al., 2012). Indeed, if we assume growth to be regulated within a simplistic source-centric framework increased Rd in response to temperature would need to reflect a substantive increase in maintenance costs (e.g. increased turn-over of cellular components, or energy requirements to maintain ion gradients; Penning de Vries, 1975) or an increase in wastage (e.g. futile cycling or decoupling of the mitochondrial electron transport chain from oxidative phosphorylation; Amthor, 2000).
Previous studies on seedlings of tropical tree species have demonstrated that growth can be maintained at day/night temperatures in excess of those in the current home range of the species (Cunningham & Read, 2003) through, among other things, acclimatization of Rd in response to changes in mean growth temperature (A. W. Cheesman & K. Winter, unpublished). Yet, fundamental questions still exist as to how changes in night-time temperature and a reduction in diel temperature range may affect tropical forests. Although we recognize that the rise of night-time temperatures and reduction of temperature range in the tropics will have profound implications beyond the direct impact of temperature on plant physiological processes, including alteration of precipitation patterns (Kitoh & Arakawa, 2005) and nutrient cycling (Salinas et al., 2011), our aim was to elucidate the direct effects of night-time warming upon biomass accumulation and plant development. We examined growth in two neotropical pioneer tree species using both indoor controlled-environment growth-chambers, and outdoor open-top chambers. In linking results from highly-controlled growth chambers with those derived from outdoor studies under more natural conditions, including variation in diel temperature and natural sunlight, we demonstrate profound positive effects of elevated nocturnal temperatures on biomass accumulation.
Materials and Methods
We studied the effects of increasing night-time temperature on two fast-growing tropical pioneer tree species, Ficus insipida Wild. (Moraceae) and Ochroma pyramidale Cav. ex Lam.Urb. (Bombacaceae). Both species, found naturally throughout the lowland tropical Americas, are known to be shade intolerant only able to regenerate in natural clearings or anthropogenically disturbed landscapes (Phillips, 1990; Dalling et al., 1998). Seeds, collected from trees growing at 0–50 m above mean sea level within forests surrounding Panama City were set out in seed-tray flats and germinated in a ventilated screen-house (70% natural sunlight) at the Santa Cruz Experimental Field Facility, Smithsonian Tropical Research Institute, Gamboa, Panama. After emergence of cotyledons at least three seedlings were transplanted to experimental pots, and seedlings established before thinning and initiation of heat treatments.
Controlled–environment growth chambers
We employed four GC15 plant growth chambers (EGC, Chagrin Falls, OH, USA) housed at the Tupper Center of the Smithsonian Tropical Research Institute in Panama City. Seedlings were established in 1.75 l pots (TP-49 Treepot™; Stuewe & Sons, Tangent, OR, USA) containing a nonsterilized mix of native topsoil and 20% (by volume) vermiculite used to improve soil structure. Six individuals were harvested for determination of initial biomass (averaging 6 mg dry mass in F. insipida and 39 mg in O. pyramidale) while a further 10 seedlings of each species were assigned to each of the four chambers. All chambers were maintained with a 12 h light period (Philips TL841 fluorescent bulbs, 500 μmol photosynthetically active radiation (PAR) m−2 s−1 at pot height), and regulated air temperatures (Tair). Tair during the light period was set to 33°C and dark period temperatures randomly assigned as either 22, 25, 28 or 31°C. Transition of Tair between the light and dark set-points was completed within 10 min. The chamber treatment of 33°C/22°C represents a realistic mean daily temperature (27.5°C) and diel temperature range (11°C) for lowland Panama, while the 9°C rise in night-time temperatures exceeds even the most pessimistic scenarios for central America (IPCC, 2007). Plants were watered every day and rotated within each chamber every 3 d to standardize illumination. Final harvest and determination of biomass accumulation–allocation occurred after either 38 d in O. pyramidale or 54 d in F. insipida, with time-periods chosen to standardize for difference in absolute growth rates between species. Stem height was measured to apical meristem, and plants separated into leaf-lamina, stem + petioles, and roots. Leaf area was determined using a Li-3100 leaf area meter (Li-Cor, Lincoln, NE, USA) and roots were separated from soil by hand sorting, dry sieving and washing. All samples were dried to constant mass at 70°C, weighed, and stored under ambient laboratory conditions until further analysis. Biomass and leaf area data were used to calculate mean relative growth rates (RGR g g−1 d−1), leaf mass per area (LMA, g m−2), leaf area ratio (LAR, cm2 g−1) and shoot to root mass ratio (S : R g g−1).
In the days before final harvest, leaf-level gas exchange was measured on the youngest fully-developed leaf using a Li-6400 portable photosynthesis system (Li-Cor, Lincoln, NE, USA) under both night-time and daytime growth conditions. Cuvette temperature (Tcuv) of the Li-6400 was set to that of the specific growth chamber, and reference [CO2] regulated at c. 400 μl l−1. Dark respiration was determined during the last 5 h of the 12-h dark period. Net CO2 uptake rates (A500) were measured between the second and sixth hours of the light period using a red/blue LED light source (500 μmol m−2 s−1). For a subset of both F. insipida (five individuals per chamber) and O. pyramidale (four individuals per chamber) a more complete photosynthetic characterization was carried out; including light response curves (PAR = 1300, 1000, 500, 250, 100, 50 μmol m−2 s−1) and A–Ci curves (Ca = 400, 250, 50, 500, 750, 1000, 1250 μmol CO2 mol−1 air). During A–Ci curves Tcuv was 33°C and PAR saturating at 1000 μmol m−2 s−1.
Open top chambers
Six open-top chambers were installed at the Santa Cruz Experimental Field Facility (Fig. 1; see the Supporting Information, Fig. S1). Chambers consisted of a hexagonal steel frame (volume of 2.55 m3) with lower chamber walls of reflective insulation and an upper portion (87 cm) of clear film gauge-50 Tefzel ETFE film (CS Hyde Company Inc., Lake Villa, IL, USA). Ventilation was provided by six independent fans (TD-150; Soler &Palau, Jacksonville, FL, USA) drawing ambient air from a shaded position and delivering to a polyethylene distribution plenum in the lower portion of each chamber. Airflow in three of the chambers was heated from 18:00 h to 06:00 h by 2 × 400 W inline electrical resistance heaters (cat. no 20055K112; McMaster-Carr, Los Angeles, CA, USA) per chamber. The use of inline heating ensured that heated air was well mixed before delivery to the chamber. Temperatures within each chamber were monitored via T-type (copper-constantan) thermocouples within standard radiation shielding linked to a CR1000 data logger (Campbell Scientific, Logan, UT, USA). Tair during the hours of daylight showed little difference between treatments (Fig. 1) with temperatures at midday averaging c. 3°C above those recorded at a nearby (< 10 m) meteorological station (Fig. S1). The PAR at the site averaged a daily maximum of 1200 μmol m−2 s−1 over the course of the experiment, with some cloudless days recording levels up to 2000 μmol m−2 s−1. Air vapor pressure deficit, calculated outside the chambers, averaged a daily high of 1.5 kPa but did climb to 3 kPa on certain days (Fig. S1). During the night the use of in-line heating provided a robust method of tracking ambient temperature +2.4°C over the course of the experiment (Fig 1).
Five seedlings of O. pyramidale were established in 30 19-l plastic buckets containing the same fertile soil mix as in previous experiments. Seedlings were transplanted to buckets and grown for c. 3 wk under a transparent rain-out shelter before seedlings were thinned to a single individual per pot (mean plant total dry mass = 1.2 g) and randomly assigned to a chamber. After a further 10 d of nonheated equilibration plants were again randomized between the chambers and the warming treatment in three of the chambers was initialized. Warming treatment continued for 33 d before determination of leaf-level gas flux measurements and biomass accumulation/allocation as in previous experiments, with additional stem cross-section area determined at the root-shoot interface using image-J (www.rsb.info.nih.gov/ij) analysis of scaled photographs. Pots received ambient rainfall, with additional daily watering, as required, to maintain soils at close to field capacity.
Given practical limitations on the number of environmental chambers available, and to avoid potentially confounding issues of pseudoreplication, data were analysed at the level of the chamber within species. Biomass accumulation/allocation was analysed using the chamber mean of 10 replicate plants, with simple linear regression employed to determine potential relationships between night-time temperature and response parameter. Further analysis of derived variables LMA and LAR were also carried out using linear regression of chamber means. Leaf-level gas fluxes A500 (rate of CO2 uptake at 500 μmol PAR m−2 s−1) and Rd of plants measured at growth conditions were initially analysed by linear regression against night-time temperature. The apparent Q10 of Rd was determined by linearization of the exponential relationship between mean leaf temperature and Rd across treatments, while light response and A–Ci curve fitting was carried out using tools developed by Kevin P. Tu (www.landflux.org) following Thornley & Johnson (1990) and Ethier & Livingston (2004), respectively. Biomass allocation (total biomass, shoot height, shoot : root ratio, leaf area, LAR), morphological characteristics (LMA, stem cross-section), and physiological performance Asat (net CO2 uptake at saturating light intensity – 1500 μmol PAR m−2 s−1), leaf-level Rd of plants from the open-top chamber experiment were first averaged within chambers (n =5) and then compared between treatments (n =3) using simple t-test. All statistical tests were carried out using R (R Development Core Team, 2012) with reported values representing mean ± 1 SD unless otherwise stated.
Controlled-environment growth chambers
Both F. insipida and O. pyramidale (0.95) showed a substantial, and significant, increase in total biomass with increasing night-time temperatures (Fig. 2), from a mean of 0.85 ± 0.26 to 2.24 ± 0.39 g in F. insipida (F1,2 = 72.21, R2 = 0.96, P <0.05) and from 1.21 ± 0.18 to 2.65 ± 0.45 g in O. pyramidale (F1,2 = 56.03, R2 = P <0.05). The rate of increase in total biomass with increasing night-time temperature was similar in both species (equal to 0.16 g °C−1). This increase in biomass represented an increase in mean RGR from 0.091 ± 0.005 to 0.109 ± 0.003 g g−1 d−1 in F. insipida and from 0.090 ± 0.004 to 0.110 ± 0.005 g g−1 d−1 in O. pyramidale (Table S1).
Associated with the substantial change in total biomass and relative growth rate, there was pronounced change in biomass allocation (Fig. 2a,b, Table S1). In both F. insipida and O. pyramidale the proportion of seedlings found as shoot biomass (i.e. leaf + stem + petiole) increased, from 72.8 ± 2.6 to 75.4 ± 1.8%, and from 77.8 ± 2.2 to 82.3 ± 2.1% of total biomass respectively, although only in O. pyramidale was this shown to be a significant increase in chamber means (F1,2 = 29.78, R2 = 0.91, P <0.05). The leaf area of the latest fully developed leaf targeted for gas flux measurements more than doubled in F. insipida and O. pyramidale, from 23 ± 6 to 51 ± 8 cm2 in F. insipida, and from 54 ± 11 to 131 ± 28 cm2 in O. pyramidale as night-time temperature rose from 22 to 31°C (Fig. 3a). Associated with this increase in individual leaf area was either a marginally significant (F1,2 = 16.2, R2 = 0.84, P =0.056) decrease in LMA of F. insipida, from 48 ± 3.5 to 41 ± 3.1 g m−2, or a significant (F1,2 = 21.5, R2 = 0.87, P <0.05) reduction in LMA of O. pyramidale, from 46 ± 4.1 to 39 ± 3.7 g m−2. Leaf area ratio showed a significant increase in both F. insipida (F1,2 = 106, R2 = 0.97, P <0.01) and O. pyramidale (F1,2 = 61, R2 = 0.95, P <0.05) with increasing nighttime temperatures, from 123 ± 11 to 152 ± 11 cm2 g−1 in F. insipida, and from 152 ± 13 to 183 ± 17 cm2 g−1 in O. pyramidale, respectively (Fig. 3c).
Analysis of leaf-level gas flux showed no monotonic change in A500 at 33°C across treatments (Fig. 4), with rates averaging 12.4 ± 0.6 and 15.1 ± 0.6 μmol m−2 s−1 in all treatments of F. insipida and O. pyramidale, respectively. In F. insipida Rd measured at night-time growth temperatures increased significantly (F1,2 = 49.1, R2 = 0.94, β = 0.092, P <0.05) from 1.02 ± 0.10 to 1.82 ± 0.21 μmol m−2 s−1 across the four treatments. By contrast, the increase in Rd in O. pyramidale from a chamber mean of 1.17 ± 0.15 to 1.46 ± 0.13 μmol m−2 s−1 was found to be nonsignificant (F1,2 = 4.1, R2 = 0.5, β = 0.027, P =0.182). The ‘apparent Q10’ of Rd determined across plants developed under, and measured at increasing night-time temperatures was found to be 1.95 in F. insipida, but only 1.22 in O. pyramidale. There was no significant difference in the light response curves of photosynthetic CO2 uptake in either species across the four temperature treatments, with Asat rates averaging 17.2 ± 0.9 and 20.0 ± 0.8 μmol m−2 s−1, and light compensation point averaging 28 ± 3.3 and 26 ± 2.5 μmol PAR m−2 s−1 in F. insipida and O. pyramidale, respectively (Table S2). Analysis of A–Ci curves revealed no significant difference in either Jmax or Vcmax, averaging 154 ± 5.7 μmol m−2 s−1 and 93 ± 6.4 μmol CO2 m−2s−1 or 158 ± 2.7 μmol m−2 s−1 and 96 ± 2.8 μmol CO2 m−2s−1 across all treatments in F. insipida and O. pyramidale, respectively (Table S2).
Total biomass accumulation showed a significant (t(4) = 3.09, P <0.05) increase under elevated night-time temperatures, averaging 28.9 ± 0.9 and 32.7 ± 1.9 g in control and elevated nighttime chambers respectively (Fig. 5). This increase in total biomass was attributed to a significant increase in both leaf (t(4) = 4.44, P <0.05) and stem + petiole (t(4) = 2.95, P <0.05) biomass of plants grown under elevated night-time temperature, with root mass showing no substantial change (Fig. 5). This resulted in a significant (t(4) = 2.88,P <0.05) difference in the shoot to root ratio between treatments, averaging 2.8 ± 0.08 and 3.3 ± 0.28 in control and elevated nighttime temperature chambers, respectively (Table 1). Associated with increased leaf mass under elevated temperatures was a significant (t(4) = 3.16, P <0.05) increase in total leaf area (0.41 ± 0.03 to 0.47 ± 0.02 m2), but unlike the growth chamber study there was no significant decrease in LMA or increase in LAR (Table 1). Leaf-level gas flux rates in leaves developed under chamber conditions showed no difference between treatments when measured at a common temperature; Asat averaged 27.4 ± 1.8 μmol m−2 s−1 at a temperature of 33°C, while Rd(25) averaged 0.76 ± 0.04 μmol m−2 s−1 across all six chambers. Similarly, both daytime and night-time transpiration rates showed no significant difference between treatments, averaging 4.7 ± 0.2 and 0.36 ± 0.4 mmol H2O m−2 s−1, respectively, across all chambers (Table 1).
Table 1. Growth and morphological characteristics of Ochroma pyramidale grown in open-top chambers under control or elevated night-time (+2.4 °C) temperatures
Values represent arithmetic mean of chamber means (n =3) ± 1 SD. ns, not significant.
Total biomass (g)
28.9 ± 0.9
32.7 ± 1.9
Shoot to root ratio
2.8 ± 0.1
3.3 ± 0.3
Leaf area (m2)
0.41 ± 0.03
0.47 ± 0.02
Stem height (cm)
39.5 ± 1.9
43.5 ± 2.4
Leaf area ratio (cm2 g−1)
141 ± 6.0
145 ± 3.1
Stem cross section × 104 (m2)
1.56 ± 0.15
1.75 ± 0.12
Leaf mass per area (g m−2)
39.1 ± 1.7
39.3 ± 0.9
Asat (µmol CO2 m−2s−1)
27.5 ± 1.09
27.3 ± 2.61
Transpiration – light (mmol H2O m−2 s−1)
4.7 ± 0.15
4.8 ± 0.25
Transpiration – dark (mmol H2O m−2 s−1)
0.38 ± 0.22
0.35 ± 0.15
Rd(25) µmol CO2 m−2 s−1
0.77 ± 0.09
0.76 ± 0.04
Contrary to the notion of adverse effects of increasing night-time temperatures on tropical tree performance (Clark et al., 2003, 2010), our results demonstrate that under well-watered conditions elevated night-time temperature promotes growth in seedlings of two neotropical pioneer tree species. In the experiments that employed controlled-environment chambers growth was positively affected in both species, even at 31°C, far in excess of night-time temperatures currently seen in lowland Panama (Fig. 2). Results from naturally lit open-top chambers were consistent with this finding, showing an increase in biomass of O. pyramidale in response to moderate night-time warming (Fig. 5).
In the indoor study, a positive growth response was paralleled by pronounced changes in leaf morphology (Fig. 3), with an expansion of leaf area driving both a decrease in LMA and an increase in LAR. Leaf morphology may be particularly sensitive to elevated night-time temperatures given that the biophysical basis of expansive growth (Pantin et al., 2012) often causes leaf expansion to peak during the night (Schurr et al., 2000; Pantin et al., 2011). In addition, and although leaf physiognomy (Royer et al., 2009) and LMA (Cordell et al., 1998; Poorter et al., 2009) are known to vary in response to climatic conditions, differences in leaf morphology may represent an indirect effect mediated through enhanced phenological development, that is, the speeding up of plant development, under higher night-time temperatures. In the open-top chamber study, with substantially greater final biomass than in the growth-chamber study, there was no difference in LMA observed as a result of night-time warming: LMA values (c. 39 g m−2) were similar to those seen at high night-time temperatures in the growth chamber study (Fig. 3b).
Leaf-level Rd measured as part of the growth-chamber study showed a general increase with temperature, but to a level generally below that predicted by instantaneous Q10 measurements (Cavaleri et al., 2008). Although the apparent Q10 of F. insipida was 2, it was only 1.2 in O. pyramidale, suggesting significant acclimatization of Rd. Indeed, both species have previously been shown to demonstrate pronounced thermal acclimatization of Rd in response to mean growth temperature (A. W. Cheesman & K. Winter, unpublished) and our results highlight the need to consider long-term acclimatization of Rd in tropical trees (Atkin & Tjoelker, 2003).
We saw no increase in photosynthetic carbon gain per unit leaf area as a result of altered sink strength (indicated by higher Rd) at higher night-time temperatures (Turnbull et al., 2004). Yet, there was clearly a pronounced increase in biomass accumulation. Increasing LAR and decreasing LMA could both account for an increase in C gain per unit biomass. Similarly, changes in Rd of non-photosynthetic tissues (Saveyn et al., 2008), or a change in carbon-use efficiency (Hansen et al., 2009), could contribute to increased growth under higher night-time temperatures.
The results of the open-top chamber study were consistent with those seen in controlled-environment chambers with the one species tested, O. pyramidale, showing a significant difference in total biomass after only 1 month of elevated (+2.4°C) night-time temperatures. We had expected to find a difference in leaf-level Rd when measured at a common temperature (25°C) as a result of Type II acclimatization (Atkin & Tjoelker, 2003), but instead saw no significant difference in either Rd(25) or Asat(33). This finding does not preclude changes in total night-time Rd and suggests a need for a characterization of diel gas fluxes as part of further acclimation studies.
In the agronomic literature the influence of elevated night-time temperature on phenological development has long been recognized (Grace, 1988). Despite increased respiration rates, elevated night-time temperatures have been shown to result in no change, or indeed an increase, of total biomass accumulation in crop plants such as rice (Ziska & Manalo, 1996; Cheng et al., 2009), sorghum and sunflower (Manunta & Kirkham, 1996) as well as tobacco (Camus & Went, 1952) and cotton (Königer & Winter, 1993). However, many researchers have considered this finding as secondary to the noted catastrophic loss of yield resulting from the temperature sensitivity of flower development (Nielsen & Hall, 1985) and seed maturation (Seddigh et al., 1989; Peng et al., 2004; Prasad et al., 2006). Studies in more natural systems have also seen impacts on plant phenological development. For example, the large herbaceous perennial, Phytolacca americana (Phytolaccaceae) demonstrated no difference in biomass accumulation but achieved flower set faster as a result of night-time temperatures increasing from 20°C to 24°C (Wolfe-Bellin et al., 2006). More recently, studies in temperate North America have demonstrated that elevated night-time temperatures in urban environments increase growth rates in seedlings of red oak, Quercus rubra (Searle et al., 2012).
There is a growing consensus on the need to re-examine the dynamic interplay between respiration and plant growth in a changing environment (Atkin et al., 2010). The general paradigm of plant respiration, as codified in Amthor's (2000) reassessment of the literature following McCree (1974), distances itself from previous models, which distinguished between growth and maintenance respiration. Instead, it considers individual processes requiring support from respiration (e.g. maintenance of ion gradients, nitrate reduction–assimilation, cell division, phloem loading, etc.) and how these processes vary in response to both ontogenic and environmental change. The general paradigm model moves away from considering Rd merely leading to the loss of carbon that would otherwise be used for growth. It provides a means of understanding the dynamic nature of Rd in response to environmental factors (Atkin et al., 2007; Slot et al., 2008) and highlights the need to gain a better understanding of the temperature response of component processes (Ryan, 2010). Our results, which demonstrate increased growth at elevated nighttime temperatures, emphasize the role of temperature-dependent night-time processes as drivers of plant performance in the tropics while affirming the productive role of plant respiration.
We do not currently know the degree to which our results, obtained with seedlings of two pioneer species, can be extended to mature trees; physiological processes uniquely associated with adult trees (i.e. secondary woody thickening, extensive hydraulic structures, and reproductive effort) may be particularly vulnerable to increasing night-time temperatures. Nonetheless, it is apparent that elevated night-time temperatures can markedly stimulate growth of tropical tree species at the critical seedling stage. Further work is required, especially on later life stages, to better assess the implications of our findings as they pertain to the structure and functioning of tropical forests in the face of ongoing climate change.
This work was supported by the Smithsonian Tropical Research Institute through the Center for Tropical Forest Studies (CTFS) Smithsonian Institute Global Earth Observatory (SIGEO). We thank Jorge Aranda for logistical support at the Santa Cruz facility; and Timothy Davidson, Kelly Anderson, Helene C. Muller-Landau and three anonymous reviewers for constructive comments on previous drafts of this manuscript.