Novel tools for an old lineage

Abstract

With a ca. 300 million year-old evolutionary history, cycads are often perceived as ‘living fossils’, relicts of their previously widespread dominance. Patterns of genetic variation for a member of the most basal cycad genus, Cycas micronesica, support the notion that cycads are a dynamic group with ongoing diversification. Herein we hypothesize that cycad’s hefty genomes enable rapid adaptive change and facilitate specific beneficial interactions with varying assemblages of symbionts. Characterizing population-level genomic patterns of cycads and their symbionts, pollinators in particular, will enlighten our understanding of these mechanisms and of adaptive variation that underlies cycad evolution. In the light of rapid climate and landscape change, cycads are a beacon for understanding the ecological processes that ultimately enable species long-term survival.

This article refers to:

Introduction

Of at least Tertiary origin, the Cycas genus currently has ca. 100 species, mainly Australian and Indo-Chinese extending as far as East Africa.1 High intra-genomic polymorphism, complex morpho-taxonomy, unresolved species phylogenies and signs of recent dispersal suggest the Cycas lineage is ongoing diversification.2–6 Our results from Cycas micronesica in Guam7 and sister species (Cibrián-Jaramillo A, et al. unpublished) support this notion. In Cibrián-Jaramillo et al.7 we showed that allelic richness does not suggest a “genetic or evolutionary relict,”8 and that genetic relationships among populations, even within a small island, are complex. Here, we suggest that all cycads harbor and maintain vast genomic information within their ∼30G genomes and life longevity (with average minimum age of 500 years in some species9). This store-house likely offsets genetic erosion and provides a genetic platform for cycads to continuously develop de novo interactions with pollinators and soil microbionts, enabling cycads to adapt to ever-changing habitat and fauna despite minimal morphological changes. Their genomes also maintain genes associated to traits only found in angiosperms, constituting the link between the rise of the seed plants and angiosperm diversification.10 Cycad genomes are repositories of changes in a ∼300 my evolutionary history,10,11 making them an exciting group for exploring long-term genomic patterns in an ecological timeframe.

Next-generation (NextGen) sequencing techniques (sequencing entire genomes),12,13 are an unprecedented tool to enable characterization of a uniquely old but dynamic lineage such as cycads. Our recent study provides information on cycad microevolution based on population-level patterns of neutral polymorphisms.7 Genomewide markers identified with NextGen provide the opportunity to test the role of natural selection on polymorphisms associated with adaptive variation. In this way genomewide variation in populations or population genomics, can reveal the intersections between genomes and biological processes that have enabled the survival of plant lineages over time. We communicate here that the most interesting areas in cycad population genomics include adaptation to local environments, seed and pollen-mediated gene flow, and multiple-species interactions.

Adaptation to Local Environments

We showed that genetic patterns among north and south of Guam are associated to soil type, seed size differences and secondary-compounds.7,14 Invoking population genomic approaches would help identify the extent of cycad phenotypic plasticity from a genetic basis to adaptation across microenvironments.15–17 Of the emerging strategies,13,18 measuring genetic diversity and differentiation of SNPs, EST- or 454-microsatellites or RAD-TAGs7,13,19–21 is feasible for cycads.7,22,23 Patterns of polymorphism can be measured on outlier candidate genes related to phenotypic traits13,15,16 among the genetic populations we previously identified.7 Additional candidates can be identified using an ortholog-based approach to identify functionally important genes,24 for example genes that confer tolerance to iron-poor soils.7 Alleles associated to soil type can be identified with a GIS-based Fst-He association.25 Cycads in Guam's resource-limited sandy littoral habitats for instance, may segregate from those in the resource-abundant upland habitats.7 Studying these patterns will help understand the balance between constitutive and plastic traits in this ancient genus, as well as set the stage for downstream validation experiments.16

Gene Flow

Regional gene flow of four Cycas species across the South Pacific suggests ongoing differentiation and uncovers natural hybrids in this genus (Cibrián-Jaramillo A, et al. unpublished). Population genomics can identify hybrids and detect early phases of species divergence informing the biogeographic history of oceanic cycads. Other species with similar genetic patterns on islands or mainland (Macrozamia riedlei,26 Cycas debaoensis,27 Zamia spp.,28 Dioon spp.29) are also worth exploring in this context. Understanding seed versus pollen-mediated gene flow is especially interesting and critical to comprehend plant reproductive biology and ecology. Decades of population genetic research have scarcely disentangled their distinct spatial and temporal contribution in plant genetic variation.30,31 Cycads are ideal for studying this because embryo (F1) tissue can be unambiguously separated and genotyped from the haploid megagametophyte and maternal integument tissues (Fig. 1, top). Local spatial genetic structure (SGS),32,33 using spatial autocorrelation and parentage analysis can tease apart seed vs. pollen-mediated gene flow34 and their association to seed colonization rates, population size and local habitat requirements (SNPs, EST-microsatellites). Genomic patterns associated to seed morphology and size35 or seed's flotation layer in Cycas species would provide a phylogenetic context to adaptive changes in dispersal, which likely underlies their recent diversification in the Pacific.

Multi-Species Interactions

Cycads are also interesting because instead of favoring increased reproduction through selfing or mixed-mating strategies as most plants do, they can elicit increases in plant fitness through multi-layered interactions with pollinators. This is a previously overlooked mechanism that likely plays a major role in cycad population genetic structure. Most genetic plant studies ignore the influence of pollination on reproductive effort over time via tradeoffs between plant and pollinator(s) as a dynamic system. In the obligate outcrosser C. micronesica, male cone herbivory by larvae of the putative pollinator moths (Anatrachyntis spp.) accelerates the plant's subsequent reproductive event, thus timing of cone production in one cycle directly depends on cone predation of the previous cycle.14 Because pollination services covary with plant population size or density36,37 this dispersal-herbivory dynamic increases temporal and spatial variation in pollen availability that can favor reproductive assurance. This interplay between cycads and their pollinators may help explain genetic recruitment patterns among cycad populations that we observed in Guam, in which populations tend to self-seed but also exchange migrants with other populations.7 Measuring pollen-mediated gene flow would help tease out if observed complex relationships of cycads and insects are specific enough to affect reproductive strategies. It would be interesting to test if the increase in seed set is associated to pollen-bearing Anatrachyntis moths in receptive megastrobili14 or to the presence of wind-blown pollen.

The ecological genomics of pollinator-plant interactions is essentially unexplored. There are examples of species-specific pollinator associations in cycads38,39 but their recent island colonization and recurrent bottlenecks4 imply that Cycas species face novel pollinator associations continuously. In Guam, pollen records suggest that major habitat changes have driven C. micronesica bottlenecks at least four times in the past 9,000 years40 with potentially novel cycad-pollinator interactions forming. Cycas micronesica's current Guam putative pollinators are possibly of recent establishment.41 How often do cycads shift pollinator complexes? Phylogenomic correlations of functionally important genes involved in pollination (reviewed in ref. 42) or processes which may have coevolved to attract pollinators from herbivory-defense genetic pathways (e.g., thermogenesis)1,14,39,43–46 (Specht and Terry I, unpublished data) would help answer this question. Expression of genes related to cone production and cone timing can also be measured under different environmental triggers such as fire and/or predation by insects.

NextGen applied to other multi-layered relationships will provide further insights on the diversification of symbionts and hosts. For example, the tripartite symbiotic cycad root system with mycorhizae and cyanobacteria44 may be studied with speed and depth that is unprecedented.

Conclusions

Next-generation tools will enable the integration of genomic patterns to understand the ecological circumstances that make population-level patterns adaptive. By understanding population genomic patterns in the context of multiple-species interactions, we will be able to better manage and conserve vulnerable taxa. This is particularly relevant for cycads, which are arguably the most threatened group of plants.47 Ultimately, we will gain information regarding processes that have enabled plants to adapt to ever-changing landscapes.

Figures and Tables

Figure 1 Phenotype of Cycas micronesica seed (top). External maternal integument tissue (I), haploid gametophyte tissue (G) and F1 embryo (E) are easily distinguished within these large seeds. Typical tree in habitat (bottom).

Acknowledgments

Thank you to Irene Terry for helpful comments on this manuscript.

Addendum to: