Fire shapes plant traits: Introduction & Evidence from the Mediterranean Basin

It is now well accepted that wildfires are fundamental process in many ecosystems, including mediterranean ones. It is also becoming clear that fires has been for long in the Earth, almost since the origin of land plants. However, the assumption that some plant traits are adaptations to fire has remained controversial. In fact, demonstrating that fire, as any other process, has been a selective force for in the evolution of plants is not an easy task, and distinguishing between adaptations and exaptations is a high bar. In this session we aim to compile evidence that many plant traits are indeed fire adaptations as they have been shaped by specific fire regimes through their evolutionary history. Specifically in this talk we review recent experimental studies in the Mediterranean Basin providing evidence that natural selection is shaping fire related traits within different model species. Fire enhance flammability in Ulex parviflorus (Fabaceae); and different fire regimes selects for different levels of serotiny and bark thickness within Pinus species (e.g., P. pinaster, P. halepensis). In addition, we found evidence that heat-stimulated germination of different species cannot be explained by summer temperatures in the bare soil. All these experiments contribute to demonstrate that fire is a selective agent shaping plant traits.

Fire and the angiosperm revolutions

Climate is generally considered the main factor controlling plant form and the distribution of vegetation (locally modified by soils). This assumption was central to seminal studies of convergent evolution in shrublands growing in Mediterranean-type climates. However mediterranean shrublands are not necessarily at equilibrium with climate or soils. Instead their stature and other attributes may be shared because of convergent fire regimes. All Mediterranean climate regions also support non-shrubby vegetation, including forests, with quite different plant attributes and different fire regimes. The question, then, is whether the fire regimes are primarily determined by extrinsic factors, especially climate, or by the intrinsic properties of the plants themselves. If intrinsic factors are important, then it would be possible to have very different ecosystems, such as shrublands vs. forests, sharing the same extrinsic environmental settings but diverging in fire regime. Changing proportions of alternative states through space and time would depend on factors, intrinsic and extrinsic, influencing the way in which vegetation creates and responds to fire regimes. I explore these questions in the context of angiosperm evolution starting from the spread of early flowering plants in the Cretaceous, the emergence of broad-leaved forests in the early Cenozoic and the rise of grassland and shrubland biomes in the later Cenozoic. Both palaeontological and phylogenetic studies indicate that vegetation fires are ancient and that pyrophytic (flammable shrublands) and pyrophobic (closed forest) formations have existed for many millions of years. The relative areas occupied have fluctuated through time because of changes in both extrinsic (climate, atmosphere) and intrinsic (biological) factors.

Role of fire in evolution of the great Gondwanan family Proteaceae

In addressing the issue of how to demonstrate that a particular fire-related trait is actually an adaptation to fire, we believe that two criteria need to be satisfied: 1. the trait gives the species a greater fitness advantage in the presence of fire than any other likely agent of natural selection in its particular habitat; 2. the origin and proliferation of that trait through evolutionary time coincides with the occurrence of fire as the agent of natural selection. Neither is easy to demonstrate in practice. Sufficient is known about the ecology of the great Gondwanan family, Proteaceae, to reveal the relative fitness benefits of certain traits in the presence of recurrent fire, though few comparative experiments have been conducted. Recently, DNA-based phylogenies have been converted to time-based chronograms using the excellent fossil record for Proteaceae to give good estimates of the molecular clock. This enabled us to superimpose fire-related traits and fire-proneness of species in the genus Banksia (He, Lamont, Downes 2011) and all genera in Proteaceae (Sauquet et al. 2009; Lamont, He, unpublished) on the chronograms to gain insights into their origin and evolution using novel Bayesian MCMC procedures. We show that Banksia has been fire-prone since its origin 61 Mya and that serotiny (fire-dependent seed release, drought is ineffective) and dead floret retention (fire-enhancing for seed release) probably arose concurrently. Clonality (fire-avoiding rhizomes) and dead leaf retention (fire-enhancing for seed release) are derived traits and arose < 25 Mya. For the entire family, five episodes of shift from rainforest to fire-prone environments with massive speciation began 85 Mya. Serotiny can be traced back to 74 Mya, passive soil-storage to 71 Mya, and ant-dispersed soil-storage to 44.5 Mya, all showing negligible germination in the absence of fire and their proliferation highly correlated with the historical occurrence of fire.

Anthropogenic Fires Shape Seed Traits in the Chilean Matorral

Fire is a major disturbance affecting ecosystems worldwide. Phylogenetic studies have shown that seed persistence (fire-resistance) has evolved with fire in Mediterranean-type ecosystems. However, the existence of specific seed traits resulting from natural selection mediated by fire remains as a key question in plant evolution. We evaluated the role of fire in the evolution of seed traits from a microevolutionary perspective, using as study system a native annual plant (Helenium aromaticum) from the Chilean matorral, where fire is a novel disturbance. The interpopulation variation of seed pubescence, seed shape and pericarp thickness was strongly associated with fire frequency, and not to site productivity. Within a given H. aromaticum population, fire produced directional selection on seed pubescence and stabilizing selection on seed shape. These were shown to be heritable traits. Our findings provide insights into the understanding of seed trait evolution in Mediterranean ecosystems and demonstrate that human-made fires can be driving evolutionary changes in plant species from ecosystems where fires do not occur naturally. Funded by FONDECYT-3090018.

Tradeoffs between seedling recruitment and sprouter persistence shape the evolution of chaparral shrub life history types

Shrubs that inhabit Mediterranean-type climate regions have different life history types. One of these types is facultative sprouters (FS) that resprout after fire and recruit seedlings from a fire-cued dormant seed bank. Another type is obligate sprouters (OS) that resprout after fire and only recruit seedlings in mature shady sub-canopy sites in between fires. Unpublished and published data for southern California chaparral shrubs showed that FS have lower resprouting success after fire (48% resprout survival) than OS species (96% survival). This raises the question as to why FS species do not resprout as successfully as OS species? An FS individual that did resprout as successfully as the OS would have a clear fitness advantage. One answer to this question could be a tradeoff between seedling recruitment and resprout success. This tradeoff could work by reducing seedling recruitment if the numbers of seeds produced was traded off against resources allocated to storage for sprouting; however, this model is not well supported among chaparral species. Alternatively, a trade off could be at work at the level of the traits that facilitate seedling recruitment post-fire that come at the cost of those that promote sprouting and persistence. In a common garden study of nine species, we found that FS seedlings grew and assimilated carbon more rapidly, had lower root to shoot ratios, and lower shade tolerance than OS seedlings. This suite of traits of FS seedlings would facilitate recruitment in a resource rich post-fire environment in competition with sprouting species and herbaceous annuals and would not be conducive to storage and thus persistence. By contrast, slower growing OS species that recruit in shady microsites in between fires display a suite of traits conducive to allocation of resources to storage that would facilitate resprouting at an early age.

Evolution of Postfire Seedling Recruitment

The vast majority of woody species in fire-prone environments persist in the face of recurring fire by vegetative resprouting. In Mediterranean-type climate shrublands subjected to a predictable crown fire regime there also has been selection to capitalize on burned sites for reproduction. In these species seedling recruitment is delayed to a single postfire pulse. This talk will focus on the factors selecting for postfire recruitment and the trade-offs involved in different soil vs canopy seed storage.

Dispersal and seed bank shifts in the Arbutoideae (Ericaceae): the potential role of fire.

The Arbutoideae is comprised of six genera principally restricted to regions with Mediterranean-type climates. Based on a molecular phylogeny, seed dispersal and germination characteristics have shifted from bird dispersal with transient seed banks to mammal dispersal with fire-stimulated, persistent seed banks. Fire regimes vary among sites in frequency and type. In this talk, I will explore the extreme differences that exist between dispersal types found in Arbutus and Arctostaphylos in North America. Recent work indicates that seedlings of species in both genera are relatively shade intolerant and sensitive to damping off by fungi. Arbutus menziesii is a relatively tall tree that is a major component of mixed evergreen forests along the central Pacific Coast region. This ecosystem normally experiences high frequency surface fires. Arbutus seeds recruit into gaps of post-fire forest habitats by bird dispersal from adjacent intact forests. Arctostaphylos has a high diversity of species in chaparral habitats that principally experience canopy wildfire at lower frequencies. In contrast to Arbutus menziesii, Arctostaphylos species produce persistent soil seed banks that germinate in place only after a fire. Evidence indicates that the principal dispersal agents of Arctostaphylos are scatter-hoarding rodents that provide a significant process in seed bank formation. The caching activities of rodents turn out to be an inadvertent fire adaptation in the context of the lower frequency, high intensity wildfires that characterize chaparral habitats.

Fire-stimulated flowering in mediterranean regions – a review

There is no better demonstration of a fire-dependent process than fire-stimulated flowering (fsf). We are doing a world survey of species with fsf and have comprehensive data on 550 species so far, although the data for mediterranean Europe and California are meagre and fsf is negligible in Chile. Here, we synthesise our data on 386 species for Australia (plus New Zealand) and South Africa (plus Madagascar). It is best expressed in the mediterranean regions of these (sub)continents (71.2%). Overall , most are herbaceous monocots (72.3%) but ‘woody’ monocots and gymnosperms as well as herbaceous/woody dicots are also represented. 16 orders are spread throughout the seed plant phylogeny from cycads to the most advanced Asterids (only the Magnoliids are missing), with at least six independent origins at the Class level, suggesting a long association with fire. Of the 35 families represented, the most speciose are terrestrial orchids (at 45.1%), followed by Xanthorrhoeaceae, Iridaceae, Asteraceae, Proteaceae, Haemodoraceae, Faboideae, Amaryllidaceae, Droseraceae, Poaceae, Dasypogonaceae, Zamiaceae and Myrtaceae (at 1.6%). Surprises include three parasites, a podocarp, 20 aloes, two trigger plants (Stylidium) and several trees. Flowering usually peaks 5?12 months after summer-autumn fire in mediterranean regions, but can be as short as a few days to as long as 2?3 years. 40% show obligate fsf (some species become dormant in the absence of fire) while the rest are facultative. We have identified ten potential fitness benefits of fsf, centred around pre-empting resources when they are most available. Work on the South African orchid, Disa (Bytebier et al. 2011) traces fsf back to 18 Mya showing that fire was an effective evolutionary force then. Among the six proteas with fsf, their ages range 1.2?10.8 My (Valente et al. 2010), indicating that fsf is a highly advanced condition in Protea. Much remains to be known about fsf.