There is mounting evidence that tropical trees are becoming more and more affected by global climate change, with droughts (1) and increased temperatures becoming more prevalent (2), causing increased tree physiological stress and mortality in tropical forest (3, 4).  The global surface temperature increase by 2100 relative to 1850-1900 is predicted to between 0.3 and 4.8°C, depending on the emissions scenario, with half of the scenarios predicting a >2°C warming (5).  A >2°C warming will affect extensive areas of tropical forests, and there is great uncertainly whether tropical trees will be able to acclimate (6) and how the effects of warming large areas of tropical forest may feedback to affect the earth system (7).  When considering an individual tree, population of individual trees, or communities of tree species, the options are: adapt (i.e., migrate; 8) , acclimate, or go extinct(9).  There is evidence that tropical trees are migrating upslope along elevational profiles in areas where mountains exist (10), resulting in community thermophilization, albeit at low rates, 0.002-0.022 °C year-1 for adult trees(11).  Therefore, the question remains: how will the millions of trees inhabiting the extensive areas of lowland tropical forests (e.g. the Amazon) respond?

            The ability to rapidly and effectively acclimate to the projected increase in the temperature in the tropics has typically be thought of as unlikely(2).  Regardless, using the tenants of classical niche theory in relation to the realized thermal nice, I will entertain three theoretical possibilities of how tropical tree may adapt to climate change.  First, species migration is not needed if niche breath is large, highly plastic, or does not interact directly with the environmental stressor (temperature).  Second, I contend that the underappreciated molecular plasticity in gene expression (i.e. transcriptomic variability) allows species to interact and manipulate their functional niche space in response to an environmental stressor, permitting a large adaptability at evolutionary timescales.  Third, species may acclimate to an environmental stressor if it is spatially or temporally variable, allowing patch dynamics in susceptibility and recovery to drive stress-resistant selection. 

          The evidence that the temperature rise in tropical forest will happen abruptly enough to push tropical trees beyond their physiological limit is mixed, but mounting, and contains some uncertainty (1, 4, 12, 13).  The current ambient air temperature for most tropical evergreen forests lies below 28°C, in some cases well below that mark; areas where air temperatures exceed 28°C tend to support grassland and more-xeric deciduous woody plant communities, in part because these areas are drier(14).  This speaks to the cofounding effects of total precipitation and climate seasonality in understanding temperature responses; mainly, that although the optimal temperature for light saturating photosynthesis for seven tropical forest areas (four Asian and three Amazonian study plots) peaked around 28°C, there was substantial variability among them when accounting for vapor pressure deficit.  Nonetheless, modeled net ecosystem exchange for tropical forests declines above 28°C (14).  Recent work on leaf-level photosynthetic rates of tropical trees shows the optimal temperature for assimilation to be 29°C (15).  Thus, individual trees may be slightly more resistant to temperature increase, than understood when modeling the productivity whole forest s (12).  

         Some of the best, recent evidence suggests that tropical tree species are sensitive to warming, with relative growth and assimilation rates definitely declining above 30°C, and that later successional species with lower photosynthetic capacities being more vulnerable (16).  Slot and Winter (16) grew tree seedlings of three tropical tree species in controlled environments at 25, 30, and 35 °C, and all three species showed some degree of acclimation by increasing their thermal optimum of photosynthesis and reducing respiration rates.  But, increases in thermal optimums of photosynthesis lagged the environmental temperature increase, and rates of photosynthesis still declined significantly in the warmest treatment (with several seedlings dying completely).  Therefore, the thermal niche of tropical trees is somewhat plastic, and species can make physiological adjustments to increasing temperature, up to a point (probably around 30-32°C) (15, 16).  It is unlikely that tropical forest areas will warm fast enough to where such physiological adjustments will not permit the acclimatization by tropical trees; therefore, we can view tropical trees, and tropical forest albeit a bit less so, to have a wide thermal niche.  No doubt, there will likely be stress and mortality of individuals that exist at the edge of the thermal niche(10, 11), for example the warmest locations of the species distribution, or areas undergoing habitat change via increased drought.  Yet, tropical tree species do show physiological adaptability and the overall extinction risk related to thermal stress for tropical forests and most of the species they harbor should remain low.

             Generally, research on the genetic variability in tropical trees has focused on the genomic diversity of trees in space, focusing primarily on their ecological interactions (17).  Analyzing the genomic diversity within communities is undoubtedly one of the major modern biological challenges, and it is becoming increasingly feasible to sequence and assemble transcriptomes (18) and genomes for tropical tree species (19).  Recent transcriptomic evidence using tropical tree communities showed that individual trees can regulate the expression of genes associated the pest and pathogen defense in relation exposure risk (20).  Trees that existed at higher densities with conspecifics had greater levels of gene expression associated with plant defense, showing that negative density dependence operates at both at physiological level via plant-pant interactions and invokes an underlying genetic response.  One can envision this operating with respect to thermal or drought stress, where individuals exposed to the stressor develop an increased ability to respond physiologically via genetic or transcriptomic memory.  In fact, transcriptomic expression by species subjected to experimental drought was a greater predictor of species co-occurrence in space than phylogenetic of functional trait relatedness in a Wisconsin forest (21).  Identifying how gene expression relates to the physiological adaptations of tropical trees subjected to thermal stress is a promising avenue in better understanding the biological processes underlying their responses to increasing temperatures. 

           Lastly, both the physiological (12, 14, 16) and genetic adaptations (20, 21) interact in time and space in a directionally stochastic fashion.  Increased temporal and spatial variability of rainfall and temperature, with the general increasing trend in temperature, are the two main predictions of climate change forecasts(5).  With respect to temperature, I predict persistence in realized niche thermal nice for tropical trees with increased temporal variability in species performance.  Stressful years will increase in frequency and magnitude, but so will years suitable to high performance, and reproduction of tropical trees.  For example high irradiance ENSO years will likely drive forest-wide reproductive bursts in fruit production (22).  If species can adapt physiologically and genetically, at least in the transcriptomic expression of genes associated with physiological adaptation to increased temperature, the logical pathway for evolutionary selection and diversification of species adapted to higher temperatures becomes reasonable.  The patch dynamics (sensu 23) of individual performance, or fitness in relation to thermal stress, and species interactions in space will then determine if tropical trees persist in the coming centuries.  Thus, it becomes increasingly important to maintain large areas of intact tropical forest in well-conserved condition.
​
            In conclusion, despite the doomsday rhetoric among some scientific camps and the undeniable facts of anthropogenic climate change, the thermal niche for tropical tree species is broad in most contexts, especially that of the typical lowland tropical forest community.  This allows species to shift within their realized thermal niche, at least with respect to projected temperature increase.  Furthermore, temperature increase. as a stressor. is spatially and temporally variable allowing ample time for physiological and genetic acclimation by species.  Such responses will interact over time and in space to determine the ability of tropical species to evolve in the face of increasing temperatures.  There is evidence that plant communities have responded to climate change in the past (24), yet uncertainty surrounds whether they will continue to do into the future. ♠

Works Cited
1. S. L. Lewis, P. M. Brando, O. L. Phillips, G. M. van der Heijden, D. Nepstad, The 2010 amazon drought. Science 331, 554-554 (2011).
2. T. M. Perez, J. T. Stroud, K. J. Feeley, Thermal trouble in the tropics. Science 351, 1392-1393 (2016).
3. N. McDowell et al., Drivers and mechanisms of tree mortality in moist tropical forests. New Phytologist,  (2018).
4. W. R. Anderegg et al., Tropical nighttime warming as a dominant driver of variability in the terrestrial carbon sink. Proceedings of the National Academy of Sciences 112, 15591-15596 (2015).
5. IPCC Climate change 2014 Synthesis Report-Summary for Policymakers (2014).
6. M. A. Cavaleri, S. C. Reed, W. K. Smith, T. E. Wood, Urgent need for warming experiments in tropical forests. Global Change Biology 21, 2111-2121 (2015).
7. V. K. Arora et al., Carbon–concentration and carbon–climate feedbacks in CMIP5 Earth system models. Journal of Climate 26, 5289-5314 (2013).
8. I.-C. Chen, J. K. Hill, R. Ohlemüller, D. B. Roy, C. D. Thomas, Rapid range shifts of species associated with high levels of climate warming. Science 333, 1024-1026 (2011).
9. K. J. Feeley, M. R. Silman, Extinction risks of Amazonian plant species. Proceedings of the National Academy of Sciences 106, 12382-12387 (2009).
10.  K. J. Feeley et al., Upslope migration of Andean trees. Journal of Biogeography 38, 783-791 (2011).
11. A. Duque, P. R. Stevenson, K. J. Feeley, Thermophilization of adult and juvenile tree communities in the northern tropical Andes. Proceedings of the National Academy of Sciences 112, 10744-10749 (2015).
12. M. E. Dusenge, D. A. Way, Warming puts the squeeze on photosynthesis–lessons from tropical trees. Journal of experimental botany 68, 2073 (2017).
13. C. Huntingford et al., Simulated resilience of tropical rainforests to CO 2-induced climate change. Nature Geoscience 6, 268 (2013).
14. Z.-H. Tan et al., Optimum air temperature for tropical forest photosynthesis: Mechanisms involved and implications for climate warming. Environmental Research Letters 12, 054022 (2017).
15. M. Slot, K. Winter, In situ temperature response of photosynthesis of 42 tree and liana species in the canopy of two Panamanian lowland tropical forests with contrasting rainfall regimes. New Phytologist 214, 1103-1117 (2017).
16. M. Slot, K. Winter, Photosynthetic acclimation to warming in tropical forest tree seedlings. Journal of experimental botany 68, 2275-2284 (2017).
17. C. O. Webb, Exploring the phylogenetic structure of ecological communities: an example for rain forest trees. The American Naturalist 156, 145-155 (2000).
18. S. R. Strickler, A. Bombarely, L. A. Mueller, Designing a transcriptome next‐generation sequencing project for a nonmodel plant species1. American Journal of Botany 99, 257-266 (2012).
19.  E. K. Lee et al., A functional phylogenomic view of the seed plants. PLoS Genetics 7, e1002411 (2011).
20. J. Zambrano et al., Neighbourhood defence gene similarity effects on tree performance: a community transcriptomic approach. Journal of Ecology 105, 616-626 (2017).
21.  N. G. Swenson et al., Tree co-occurrence and transcriptomic response to drought. Nature communications 8, 1996 (2017).
22.  M. Detto, S. J. Wright, O. Calderón, H. C. Muller-Landau, Resource acquisition and reproductive strategies of tropical forest in response to the El Niño–Southern Oscillation. Nature communications 9, 913 (2018).
23. R. Levins, Some demographic and genetic consequences of environmental heterogeneity for biological control. American Entomologist 15, 237-240 (1969).
24. M. B. Bush, M. R. Silman, D. H. Urrego, 48,000 years of climate and forest change in a biodiversity hot spot. Science 303, 827-829 (2004).

Density dependence, or a correlation of per capita population growth rate with population size, is a biological phenomenon that is ubiquitous in nature.  Density dependence may either be positive or negative, depending if ∆N(b-d), increases or decreases as population size (N) increases (where b and d are births and deaths, respectively).  Cases of positive density dependence (i.e., Alee effects) are rare in nature, however Negative Density Dependence (NDD; sometimes called a Malthusian-Darwinian process) is quite common.  Since the first basic identification that the human population is limited by food production (Malthus 1798), the study of density dependence has become increasingly nuanced.  The nuances of density dependence and it workings stem from the fact that it is impossible to identify process (i.e., the mechanism underlying density dependence) solely from pattern (i.e., a time series of population counts), and density dependence can be difficult to detect.  Long time-series data at the appropriate scale for population fluctuations of the study organism are needed.    

A classic example of how density dependence operates, can be understood from the data presented by Davidson and Andrewartha (1948) on Thrips imaginis and their host plants.  Davidson and Andrewartha sought to describe the interannual variability in the population growth of the thrips using an exponential (i.e. density independent) growth function.  Although, the fit of the exponential growth function described the data relatively well, Frederick Smith (1961) incorporated the inherit seasonality in the data with a piece-wise (i.e., multiple) regression with slopes that varied with winter degree days and amount of spring day.  He demonstrated that the population growth data for the thrips were highly seasonal (i.e., density dependent), with population growth distributed normally around November/December.  This can be seen visually from the data table represented in the original paper by Davidson & Andrewartha.  The thrips go through boom and bust cycles that are regulated by seasonality. 

Aside from thrips, NDD appears many times in the body of ecological literature.  In tropical forest ecology, the classic example is the famous “Janzen-Connell” hypothesis (now more of theory).  Around the boom of when population biology was in the renaissance of density dependence (i.e., the 1970s), Dan Janzen and Joe Connell were both walking around tropical forests, independently.  They observed that seedling densities were highest immediately below the tree crown.  This phenomenon can be attributed to dispersal limitation of seeds originating from the parent tree.  Trees have many dispersal mechanisms including wind dispersal (amenochory), animal dispersal (zoochory), and mechanical dispersal (autochory).  However, gravity dispersal is certainly the most common mode of dispersal by trees in that all seeds will fall to the ground in the absence of a seed disperser, demonstrating why most reproductive seeds do not fall from the parent tree. 

Janzen (1970) and Connell (1978) thought about density dependence and its implications for the maintenance of diversity, and came to the theoretical conclusion that that seedling survival increases with seed dispersal distance.  The mechanism by which increased mortality of seedlings closer to the parent tree occurs is understood to be pathogen and/or herbivore related (Comita et al. 2010).  One can envision how common species are more frequently consumed by generalist herbivores, which target seedlings directly beneath the parent tree.  Therefore, greater dispersal distances result in higher rates of seedling survival, and thereby maintain species diversity in highly-diverse tropical forests.  Furthermore, NDD in tropical forests can be a mechanism for population regulation for many species. 

​Recent work (Lebrija‐Trejos et al. 2016; Uriarte et al. 2018) has focused on how NDD is mediated via water availability in tropical seedlings, resulting in interacting effects of moisture and biotic interactions (likely belowground fungi or pathogens) on seedling distribution patterns and diversity.  At Barro Colorado Island in Panama, Conspecific NDD was stronger in years with higher soil moisture.  Also, trees with larger seed mass exhibited decreased levels of conspecific NDD, suggesting that seeds with greater seed mass have weaker NDD-regulation (i.e., a less negative slope of the density dependence as population size increases.  Similarly, at Luquillo, Puerto Rico, the positive effects of increased solar radiation on seedling survival and growth were more pronounced, favoring drought tolerant species (e.g., lianas) and light-demanding pioneers and palm seedlings.  In conclusion, NDD is a strong regulating force in population dynamics for many organisms and should be examined considering abiotic and biotic drivers of population change. ♠

Comita, L. S., H. C. Muller-Landau, S. Aguilar, and S. P. Hubbell. 2010. Asymmetric density dependence shapes species abundances in a tropical tree community. Science 329:330-332.
Connell, J. H. 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302-1310.
Davidson, J., and H. Andrewartha. 1948. Annual trends in a natural population of Thrips imaginis (Thysanoptera). The Journal of Animal Ecology:193-199.
Janzen, D. H. 1970. Herbivores and the number of tree species in tropical forests. American Naturalist:501-528.
​Lebrija‐Trejos, E., P. B. Reich, A. Hernández, and S. J. Wright. 2016. Species with greater seed mass are more tolerant of conspecific neighbours: a key driver of early survival and future abundances in a tropical forest. Ecology Letters 19:1071-1080
Malthus, R. T. 1798. An Essay on the Principle of Population. As it affects the future improvement of society, with remarks on the speculations of Mr. Godwin, M. Condorcet, and other writers. London.
Smith, F. E. 1961. Density dependence in the Australian thrips. Ecology 42:403-407.
Uriarte M., R. Muscarella, and J. K. Zimmerman. 2018. Environmental heterogeneity and biotic interactions  mediate climate impacts on tropical forest regeneration.  Global Change Biology 24: e692-e704.

In case you were wondering, CTFS stands for the Center for Tropical forest Science, also know as Forest-GEO: www.forestgeo.si.edu/; and is a global network of forest dynamics plots, aimed at documenting the diversity and dynamics of the world's forests. 

My favorite ecosystem is the subtropical montane Tabonuco forest of the Luquillo Mountains in northeastern Puerto Rico.  The system is described as “aseasonal” and everwet, receiving >3500 mm/year rainfall with temperatures ranging between 20 and 23°C.  The complete food web is documented in Reagan and Waide (1996).  When published this book revolutionized the understanding of food webs, as it was the first book to fully document a food web of tropical rain forest.  Forest species richness is greater than temperate forests, but less than tropical forests at similar latitudes due to an island effect, and measures around 200 species.  Bird and ant richness is on par with other tropical forests, but there are fewer beetles than continental tropical forests.  Faunal richness is low, lacking large-bodied herbivores, however frogs and lizards are present at very high densities.  In fact, the ecosystem has the highest densities of Etherodactylus coqui frogs anywhere in the world (3200 adults and 17000 juveniles/ha). 

First and second-order headwater streams dissect the forest and harbor high densities of freshwater shrimp, land-dwelling crabs and small-bodied fish (e.g., gobies).  Given the community and abundances of organisms, traditional food web theory would predict that the food web of such an ecosystem should be relatively simple, with 3 to 4 trophic levels.  However, researchers found that on average the food web was more complex, with on average 8.5 links per food chain, with several reciprocal loops.  They also found the number of links in the food web was variable across different habitats in the ecosystem, and that within habitats the food web is partitioned into diurnal and nocturnal components with different predator and prey interactions.  Notably, without large herbivores, food chains were longer (not shorter) due to the higher trophic efficiency of cold-blooded organisms (i.e., frogs, lizards) relative to larger warmer-bodied organisms with higher metabolic rates.  Most of the organisms are omnivorous making species interactions strong, but complex and variable. ​

Understanding carbon (C) dynamics in ecosystems is important for understanding ecosystem energy balance over time, the biological processes within an ecosystem, and projecting ecosystem responses to global climate change.  C molecules are a principle element in most biological compounds used by organisms.  At the ecosystem scale, the net C balance is defined as the difference between the amount of C entering versus leaving the system, or the change in C storage over time; and because ecosystems are complex the net ecosystem C balance must quantify the C flux of several important C pools, gaseous, inorganic and organic.  The gaseous C pool is mainly CO2, but also contains methane and other volatile organic C gas derivatives [e.g., isoprenoids].  The inorganic C pool is comprised of elemental forms of C [e.g., carbonates], which are usually found in the soil, that are freely available to plants and microbes.  The organic forms of C are the biologically-bound forms of C in the ecosystem and are found in alive or dead biomass.  The amount of C in each of the C pools depends on the energy balance of the ecosystem and rate of biological processes that can transfer the C from one pool to another.  Photosynthesis, or the anabolic, autotrophic production of organic C compounds using CO2 and sunlight, is the primary process that moves C from its gaseous forms into organic forms.  Decomposition is the breakdown of organic material transforming its C into simpler, inorganic elemental forms. 
            The amount total biomass in an ecosystem is a function of the balance between photosynthesis and decomposition, minus respiration (the amount of C used in metabolic processes), and the residence time of C in the organic vs. inorganic pools (Figure 1, Odum 1969).  As plants sequester C from is gaseous form, some of the C may be transforming it into sugars, carbohydrates which in turn may be returned to gaseous C pool via respiration, but some of the C may be stored or converted to more complex organic forms within the plant (e.g. tissues), thereby fixed into the inorganic C pool.  C from the organic pool must be converted to an inorganic form before eventually returning to the gaseous C reservoir.  Usually, this happens through the decomposition of necromass (or dead biomass), where microbes and other decomposers catabolically break down biological tissues containing organic forms of C, converting them to simpler inorganic forms.  These balance of these processes is termed the net ecosystem C balance, and is limited by the rate of the biological process at the organismal level, and the abundance of that organism in the ecosystem.  The majority of terrestrial ecosystems are C sinks, meaning that the ability of plants to sequester C via photosynthesis largely outweighs ecosystem C release through respiration, decomposition and lateral C transfer (leaching etc.), resulting in the accumulation of biomass over time.  As terrestrial ecosystems mature, the accumulation of biomass asymptotes as the net positive difference between productivity and respiration peaks and then decreases.
            Plants have a variety of strategies for accumulating C via photosynthesis, and rates of C uptake by plants (i.e., assimilation rates) vary and are determined by their physiological traits (stomatal density, leaf area, leaf turnover rate, photosynthesis type etc.).  This leads to differences in residence times for C in the biological component of the C cycle, and differences in the chemical makeup of plant biomass.  At the ecosystem-scale, this residence time is a function of the transfer of C from primary producers to higher trophic levels, which mainly happens through herbivory, or the consumption of some form of plant biomass.  Plant material quality is often characterized by its nutrient content or the C to N ratio, with material that has more N content being considered higher-quality.  Higher quality plant materials are preferentially selected by both herbivores and decomposers, influencing C and nutrient cycling rates. ♠

​Above: an excerpt from Edward De Bono's book: 'Practical Thinking'  (1971), where he argues that the principle objective of a good scientist is to be wrong.  That is to say, that scientists try to prove that their ideas hold water by constantly trying to disprove them; or in essence asking a "yes" question and concluding an answer of "no".  In Practical Thinking, De Bono advocates for a third answer to the usual yes/no question.  A usual "yes" or "no" can be circumvented with a Bonian "po". 
​
A "po" in short means possibly.  It can be the launching point for scientific inquiry.  

Thinking pioneer, psychologist, physician, and all-around intellectual, Edward De Bono pioneered a new school of thought that incorporates six thinking caps. ​Of his six caps science uses the discernment cap, defined as: 

• logic applied to identifying reasons to be cautious and conservative. Practical and realistic thinking primarily relying on logical reasoning.

Science, as a discipline, relies on using the scientific method.  
Observation -> Hypothesis -> Experimental design -> Experiment -> Analysis of results -> Careful thought and interpretation of results -> Generation of scientific knowledge and understanding (not guaranteed).  Implicit in a good experiment, is the isolation of variables of interest. 

Ecology is a science were the isolation of experimental variables is unusually difficult. Natural experiments tend to include many confounding factors.  For example, when trying to quantify the successional responses of 0.1 Ha forest patches to a human-induced experimental hurricane ("The Canopy Trimming Experiment") in Puerto Rico, forest recovery to the experimental trimming was confounded by differences in community composition among replicates.  (Zimmerman et al. (2014), Forest Ecology and Management 322 64-74).  

In a classic note that has now been blown-up to epic proportions, Dan Janzen published arguably the shortest academic note ever.  Two measly words.  Yes? No. The dichotomy of these two simple words embodies the essence of scientific process and ensuing debate.  As illustrated by De Bono, a good scientists ask questions, with the sole motivation of proving them wrong. 

What if we don't know?  We can simply say "po".  And the voice of reason speaks; Leave it to the British to pick it apart; Thank you C.H. Sterling!  A Bonian "po" allows for the scientific method to decide, as if to say, "well, possibly", "I simply do not know", or "Why don't we find out".  

Modern day scientists have immense pressure to deliver concrete, yes/no answers to weighty questions that oftentimes impact policy, decisions and livelihoods.  Usually, the answers is more nuanced than a simple, yes/no, therefore I advocate for the revival of the "it depends" or Bonian "po". ♠