Metabarcoding and Biodiversity Soup

Many of the challenges of biodiversity conservation can be thought of as problems of management, and in management, it is a truism that you only get what you measure. A major problem therefore is the 'taxonomic impediment': standard biological surveys consume huge amounts of time, money and taxonomic expertise, and we are therefore impeded from addressing biodiversity loss as a normal management problem that can be dealt with wherever and whenever it arises. A partial solution has been to design indicators, but indicators are criticised for being unrepresentative of total biodiversity and likely to bias incentives if used as policy targets.  As one example, 19 species of farmland birds have been designated as the main biodiversity indicator on UK farmlands. However, an understandable response has been to ‘teach to the test’ via supplemental winter feeding of farmland birds.

However, recently, it has become feasible to combine DNA taxonomy and high-throughput DNA sequencing to create a rapid and taxonomically comprehensive method of biodiversity assessment, called metabarcoding. Metabarcoding uses PCR to mass-amplify a taxonomically informative gene from mass collections of organisms or from environmental DNA.

My lab is developing metabarcoding protocols (Yu et al. 2012 Meth Ecol Evol) and testing whether metabarcoding is a reliable way of measuring biodiversity (Ji et al. 2013 Ecol Letts). For instance, our collaborator Scott Pedley ran an ecological restoration experiment in which he applied a range of disturbance treatments to forest trackways, to make the habitat hospitable for a suite of heathland specialist arthropods. Scott used spiders, ants, and carabid beetles as his indicator taxa and found that the more severe disturbance treatments were more successful at converting trackways into suitable ecological corridors. We metabarcoded entire pitfall-trap samples and found the same result, but without needing to go through many months of laborious morphological identifications (Ji et al. 2013 Ecol Letts).

My lab is now using metabarcoding to measure the biodiversity responses of a wide range of policy interventions:  nature-friendly farming and forestry, systematic conservation planning, climate-change monitoring, and nature-reserve protection. We are also starting to use metabarcoding to study wood decomposition and salmon fisheries. Finally, we are working with the Saola Working Group and Tom Gilbert's lab at the University of Copenhagen to metabarcode leeches to search for the highly endangered rainforest saola antelope in Vietnam and Laos.



It is very natural to think of symbiosis as being like an employer-employee relationship.  Microeconomic game theory provides formal ways of modelling this relationship. For instance, screening theory shows how hosts solve the partner choice problem by designing appropriate entry costs, without needing to observe potential symbionts directly (Archetti et al. 2011 Am Nat).  Contract theory distinguishes two longstanding but weakly differentiated explanations of host response to cheaters, host sanctions and partner fidelity feedback, which can be understood using the game-theoretical concepts of commitment and subgame perfectness (Weyl et al. 2010 PNAS).

We have reviewed the interface between microeconomic theory and evolutionary biology (Archetti et al. 2011 Ecol Letts) . Did you know that the Tragedy of the Commons is only an edge case of the general public goods game? 

We also conduct a lot of empirical work.  In fig-wasp symbioses, fig wasps pollinate fig plants, but the wasps also lay eggs in seeds, so the plant must somehow limit wasp reproduction, even though we expect wasps to evolve faster than the plants.  We developed an optimal foraging model to explain how the mutualism is maintained (Yu et al. 2004 Proc Roy Soc Lond B) and tested these ideas in China and Australia (Dunn et al. 2008 PLoS Biology, J. Anim Ecology, Wang et al. 2009 PLoS ONE, Wang et al. 2013 Oikos). Among other things, we showed that interference competition prevents large numbers of wasp foundresses from laying eggs efficiently, which preserves fig seeds.

In leafcutter ant-bacteria symbioses, there are two competing ideas:  vertical transmission of a coevolved bacterial symbiont (Pseudonocardia) and environmental recruitment of actinobacteria. The vertical transmission idea must explain how it is that Pseudonocardia maintains its efficacy against coevolving fungal pathogens, and the environmental recruitment idea must explain how the ant is able to ‘screen-in’ mutualistic bacteria from the highly diverse soil environment.  Our work (Barke et al. 2011 BMC Biology and Comm & Int Biol, Seipke et al. 2011 PLoS ONE) has found evidence for both hypotheses and therefore supports the idea that attine ants use multidrug therapy to maintain their fungal cultivars. We have developed a model (Scheuring & Yu 2012 Ecol Letts) that explains how hosts (from insects to plants to corals to birds) are able to screen-in actinobacteria-dominated microbiomes, and we are now testing it in the lab.

In earlier work, my colleagues and I used an ant-plant symbiosis to develop and test the first successful spatial ecology model of species coexistence (Yu & Pierce 1998 Proc Roy B, Yu et al. 2001 Ecology, 2004 J Anim Ecol, Yu & Wilson 2001 Am Nat, Szilagyi et al. 2009 Ecol Letts).


Conservation in Low-Governance Environments

Conservation biology has made three big advances:

It has documented trade-offs between the welfare of humans and nature.  For example, modern mechanised agriculture reduces food availability for farmland birds. 

It has designed interventions to reduce the severity of those trade-offs in high-governance settings.  For example, controlled burning partially replaces natural but suppressed fire cycles on managed grasslands and forests.  These interventions, which can be costly and complex, must be designed well and be carried out reliably, an assumption that applies mainly to rich countries.

It is showing in a growing number of cases that trade-offs between human and natural welfare can be less severe than they first appear, and might not even exist.  This is because intact ecosystems provide valuable services to humans, such as the provision of natural products and the regulation of waterflows.


The fourth advance, far from fully achieved, is to design conservation interventions that work in low-governance settings, especially in economically poor but biodiversity-rich countries.  For example, in tropical countries, we mainly rely on setting aside natural protected areas:  parks.  But what should we do about the presence of people inside those parks?  How do we guarantee the continued existence of protected areas in the presence of powerful actors that want to exploit or convert the land in parks?  How do we protect natural ecosystems outside parks?

To achieve this fourth advance, we must design interventions that take advantage of governance structures that already exist, including those not derived from the state. 

We have worked on two applications of the fourth problem:  (1) indigenous hunting in Manu National Park, Peru, and (2) ecotourism in Tambopata, Peru.

(1) We showed that decades of bushmeat hunting by the Matsigenka indigenous population living in Manu Park have not harmed the viability of vulnerable species, such as large monkeys.  This is because Matsigenka are central-place foragers, and so hunting is centered around Matsigenka settlements. The result is that most of Manu Park contains monkey source populations, slowly feeding the hunting zone via monkey dispersal. Instead of trying to measure and limit hunting effort, which is impossible in this low-governance environment, it is much easier to manage the number of settlements. Fewer, larger settlements will have less total hunting impact on Manu Park, especially if Matsigenka eventually switch to shotguns.

But how to limit settlement number?  We propose a grand trade:  increased investment in schools, potable water, and healthcare in the existing settlements. Such centripetal attractions will naturally reduce the incentive to build new settlements in the short term, and in the long term, increased education will increase the ability of the Matsigenka population to defend all of Manu Park via political means. We call this the Occupy Amazon strategy (Yu et al. 2013, Reporte Manu).   

The background to this was years of ecological and anthropological fieldwork, during which we studied ecotourism (Ohl-Schacherer et al. 2008 Environmental Conservation), slash-and-burn agriculture (Ohl et al. 2007 Environment, Sustainability, and Development), and hunting (Ohl-Schacherer et al. 2007 Conservation Biology cover article) by the Matsigenka.  This work allowed us to develop a modeling technique to map, quantify, and predict the impact of hunting on a landscape, which was used to quantify the conflict between livelihoods and conservation (Levi et al. 2009, J. Applied Ecology, 2011 Ecological Applications).

2) The Interoceanica Highway, which connects Brazil to the Pacific Ocean via Peru, and the gold mining industry are fomenting deforestation and environmental pollution across a large swathe of Peru’s southern rainforests.  The most vulnerable protected area is the Bahuaja-Sonene National Park, which lies within a few kilometres of the highway.  However, a multi-million dollar ecotourism industry is also established at this frontier zone.  My student, Chris Kirkby, showed that a land tenure scheme introduced by the Peruvian government has allowed the most profitable tour operators to set aside strategic land parcels to protect forest cover around their lodges, thereby building a well-financed buffer zone between the road and the park (Kirkby et al. 2010 PLoS ONE, 2011 Env Cons).  Ecotourism has become a privately financed source of environmental governance in this part of Peru.