PS 77-72
Selection for isoprene-emitting trees in a warmed, artificial tropical forest: implications for climate change and community phylogenetic analyses

Friday, August 9, 2013
Exhibit Hall B, Minneapolis Convention Center
Tyeen C. Taylor, Ecology & Evolutionary Biology, University of Arizona, Tucson, AZ
Kolby Jardine, Earth Sciences Division Climate Sciences Department, US Department of Energy, Berkeley, CA
Marielle N. Smith, Ecology & Evolutionary Biology, University of Arizona, Tucson, AZ
Cyrille Violle, Centre d'Ecologie Fonctionnelle et Evolutive, CNRS, Montpellier, France
Brad Boyle, Ecology and Evolutionary Biology Department, University of Arizona, Tucson, AZ
Joost van Haren, Biosphere 2, University of Arizona, Tucson, AZ
Raphael Rosolem, Department of Civil Engineering, University of Bristol, Bristol, United Kingdom
Brian J. Enquist, Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ
Scott R. Saleska, Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ

The responses of tropical forests to 21st century climate change are uncertain, and their potential feedbacks to global climate contribute to uncertainty in climate predictions. Forest responses will largely be governed by abiotically driven shifts in community assembly mediated by plant traits. Here, we focus on an important plant trait that is not typically considered in community ecology studies: isoprene emission from leaves. Isoprene mitigates heat stress in leaves, contributes significantly to atmospheric aerosol formation, and is highest from tropical forests. We demonstrate positive selection for isoprene emitters in the warmed Biosphere 2 artificial tropical forest (B2-TF). We also discuss the unique effects of selection for this trait on a phylogenetic index of community assembly.

Nonrandom extinction patterns and enhanced thermal tolerance of the B2-TF have been previously demonstrated. We screened all extant trees in the B2-TF for isoprene emission. We applied a binary emission state (0=does not emit, 1=does emit) to extinct species by matching to species or taking genus averages from unbiased isoprene surveys from the literature. A test for nonrandom distribution of emission status on the phylogenetic tree and an analysis of method biases justified the use of genus averaging.


The proportion of extant trees that emit isoprene in the B2-TF after two decades of community change (1991 to 2010) was significantly higher than null expectation based on a model of random tree extinction (47% vs 30%, p = 0.01). This result demonstrates the potential for tropical forests to respond to climate warming with an increase in the proportion of isoprene emitting trees. Isoprene’s role in enhanced leaf performance at high temperatures is well demonstrated. As in the B2-TF, natural forests may gain enhanced thermal tolerance in conjunction with such community change. Increased per-tree contribution to atmospheric aerosols would also be expected.

When three inter-survey periods were analyzed independently, the period 1993-2000 showed the strongest and only significant selection for isoprene emitters. During the same period, the nearest taxon index of phylogenetic relatedness (NTI) significantly decreased (NTI = -3.1, p = 0.01), a trend usually interpreted as resulting from high competition between functionally similar species. However, it has been hypothesized (and simulated) that abiotic filtering on a trait with low phylogenetic conservatism (such as isoprene emission: Blomberg’s K = 0.37) can result in a negative NTI. Ours may be the first empirical results that are strongly consistent with that hypothesis.