Tropical wet forest plants experience relatively stable temperatures throughout the year. However, tropical forests represent a mosaic of habitats characterized by different temperatures. Heat tolerances are expected to be adapted to temperatures specific to their habitats. Although the heat tolerance of species sharing similar environments is expected to be similar, it is also possible that heat tolerance is constrained by evolutionary history because closely related species usually display similar physiologies. When exotic species are introduced to novel communities, colonization may be facilitated by their previous adaptation to high temperatures and other physiological, genetic, and demographic traits, which may grant them some competitive advantage. Increasing temperatures may represent a strong environmental filter affecting community assembly, and higher heat tolerances could facilitate the persistence of exotic species in novel environments. 

Using a community of 32 native and 7 exotic Zingiberales species from different tropical habitats in Costa Rica, Central America, we aim to answer the following questions: a) does evolutionary history constrain heat tolerance? b) do plants in the same habitat display similar heat tolerances? c) do the heat tolerances of exotic species differ from those of native species?

We measured temperature-dependent changes in photosynthetic fluorescence to determine the temperature at which the first sign of damage to photosystem II is observed (T₁₅), and the temperature at which the fluorescence of photosystem II is reduced by 50% (T₅₀). Using a community phylogeny, we tested for phylogenetic signals in T₁₅ and T₅₀. In addition, we tested for differences in heat tolerance among Zingiberales from old growth, secondary forests, and open areas, as well as between native and exotic species.

Our results support a) a significant phylogenetic signal (Pagel’s λ) for both T₁₅ and T₅₀, b) communities from open areas displayed similar photosynthetic heat tolerance compared to species from old growth and secondary forests, c) exotic Zingiberales are marginally tolerant to high temperatures than native species, but only for T₁₅. Our results suggest that evolutionary history constraints heat responses of native and exotic Zingiberales in a warming world.

We conducted this research within the Caribbean lowlands of Costa Rica. 

Fluorescence: To determine the heat tolerance of each plant species, we collected two to five undamaged, fully expanded mature leaves per individual. Leaves were collected from 7:00 AM to 10:00 AM, from June 17 to July 23 of 2017. Four extra individuals of E. elatior and H. psittacorum were collected on July 23 ad 24 of 2018. Leaves collected from open areas were previously exposed to full sun, corresponding to their specific habitat. We cut six leaf disks per treatment of 1.9 cm in diameter, which were assigned to each temperature treatment at random. We used three leaf disks for the few species with small leaves. For all species in each habitat, we collected four to a maximum of six individuals per species, for a total of 198 individuals. The only exception was Costus pulverulentus for which only two individuals were sampled (see Table S2 for sample size by species).

To prevent anaerobiosis during the heat tolerance experiments, leaf disks were placed in Miracloth® (EMD Millipore Corp, Massachusetts, USA) (Krause et al., 2010). During heat treatments, the Miracloth® blocked some of the artificial light in the laboratory. Leaf disks were subsequently placed inside a waterproof Whirl-Pak® bag (Nasco, Wisconsin, USA) and immersed in one of several circulating hot water baths for 15 minutes (ANOVA Sous Vide Precision Cooker A2.2-120V-US 2014, San Francisco, California, USA). The water bath has a temperature accuracy of ± 0.1°C. The water bath has a temperature accuracy of ± 0.1°C. We constantly checked the temperature with an infrared thermometer (accuracy: ± 1.5°C). Water baths were programmed to reach the following temperatures: 23°C (representing our control), 38, 42, 44, 46, 48, 50, 52, 54, and 60 °C. We removed leaf disks from the water baths and placed them in Petri dishes lined with moist paper towels for 24 h. During heat treatments and when the leaves were placed in Petri dishes, leaves were kept under dim light. We dark-adapted each leaf disk for at least 20 minutes and then measured F_v/F_m (Model OS-30P, OptiSciences, New Hampshire, USA). At the beginning of the experiment, we selected six leaf disks per individual and measure F_v/F_m to confirm leaves had values high enough to be used in the experiment and to avoid stressed leaves.

Photosynthetic heat tolerance was measured using the maximum quantum yield (F_v/F_m) of the photosystem II (PSII). F_v/F_m was calculated as (F_m – F_o)/F_m, where F_m and F_o are the maximum and basal fluorescence yield, respectively. The maximum quantum yield is a well-established method for assessing heat tolerance that corresponds to damage to the photosynthetic apparatus and is not prone to error associated with changes in leaf optical properties during stress treatments such as heating (Baker, 2008; Baker & Rosenqvist, 2004; Maxwell & Johnson, 2000). We heated leaf samples to only one temperature each and quantified damage to PSII using F_v/F_m at each temperature. These values were used to calculate the heat tolerance for each of our study species.

Temperature data: We recorded temperatures in open areas, secondary and old growth forests. Temperatures were recorded every 15 minutes during the months of March and July 2022, for a total of 44 days. In each habitat, we located two iButton temperature data loggers (Maxim Integrated, California, USA) at a minimum distance of 400 m and 30 m above the ground. Please note that temperatures were recorded in Y 2022, but plant heat tolerance estimates were recorded during the same season in Y 2017. Extreme temperatures from the maximum daily temperature in the open area are the result of direct sun incidence on temperature sensors.

Forest Overstory Density: We measured canopy density for ten sites of each of the three habitats (open area, secondary forest, and old growth forest), at a minimum distance of 75 m. We used a convex Forest Densiometer (Forestry Suppliers, Inc, Florida, USA). For each of the ten site values, we averaged the values taken from readings facing North, South, East, and West.

Molecular Phylogeny: We assembled a molecular phylogeny of a subset of 27 species for which DNA sequences were readily available (Fig. S2). Relationships among species are based on two chloroplast markers (matK and trnL-trnF) and the nuclear ribosomal internal transcribed spacer (ITS). DNA sequences were mined from GenBank using PyPHLAWD (Smith & Brown, 2018) and aligned using MAFFT v7.407 (Katoh & Standley, 2013). We concatenated the alignments and inferred a tree using RAxML 8.2.11 (Stamatakis, 2014) with GTRCAT model for each gene partition and 100 bootstrap replicates.