Andrew C. Martin

Reducing the Powers of the Chief Minister. A Petition to Tynwald, July 2023

Wed, 19 Jul 2023 12:00:00 +0000

I handed a petition for redress to the Clark of Tynwald at the foot of Tynwald Hill on Tynwald Day 2023. The relevant Tynwald committee has now declared this petition ‘in order’, meaning it is available for any member of Tynwald to pick up for action / debate. The first oppertunity for this to happen is October’s Tynwald sitting, 2023.

The Committee’s report and my petition may be found in here.

Broadly, the petition calls for the stripping of powers from the Chief Minister to make the role closer to the principle of first among equals. Of relevance to very recent events since I handed in the petition, I called for the removal of the powers to hire and fire ministers from the Chief Minister. This week, Mr Thomas MHK has been sacked as Infrastructure Minister. Mr Cannan confirmed to Tynwald that the decision to sack Mr Thomas was not discussed by the Council of Ministers, but was a late Sunday night unilateral decision by Mr Cannan, Chief Minister (as confirmed by Mr Cannan in Tynwald, Tuesday 18th July).

The full petition is reproduced below.

To the Honourable Members of Tynwald Court

Movement of the Island’s Government from Boards of Tynwald and the Executive Council to the Council of Ministers (between 1981 and 1990) concentrated power into a small executive consisting of the Chief Minister and their Ministers. The provisions under which the transfer occurred have almost-certainly become a root cause of recently widespread governance failures of the Island and contributed to political malaise. Reform of the executive’s structure is urgently required to improve the state of governance of the Island. Concerns raised by Tynwald members on the introduction of the Council of Ministers during the 1980s – such as too greater concentration of power in the executive and development of a ‘one party state’ – have not been formally assessed against the evidence now accumulated over 30 years. It is now time for such an assessment.

Consensus-building is essential in a parliament where most members at any time are not party-affiliated. However, three key interlinked consequences of the creation of the Council of Ministers have weakened the tradition of consensus decision-making in Tynwald: (1) the powers and patronage of the Chief Minister; (2) the ability of Government to secure majorities in exchange for the perception of consensus decision- making; and (3) under-resourcing of Tynwald and its offices.

First, the position of Chief Minister carries patronage and power that is incompatible with our non-party politics. The cultural tradition of Tynwald and circumstances of the Island make the development of substantial party politics unlikely, as demonstrated by the fact that most members have been independent since the introduction of the Council of Ministers over 30 years ago. As it stands, the Chief Minister exerts four key distorting effects (to greater and lesser extent):

The Chief Minister’s ability to hire and fire enables their political views and ideas to carry more weight than other members of the Council of Ministers, despite them having no more public mandate than others.
The Chief Minister’s patronage does not ensure that the best Tynwald members are selected for Ministerial posts, but rather those most likely to support the Chief Minister.
The Chief Minister can mobilise the substantial resources associated with their office to support their policy positions, giving them undue power over both other Ministers, Departmental Members, and other members of Tynwald, who are not afforded as much support.
A no confidence vote in a Chief Minister’s administration is unlikely, as over three-quarters of the House of Keys are subject to the patronage of the Chief Minister because: (a) they hold Government roles that they may only retain with Government’s support; and (b) the opposition is largely unorganised with no real possibility of an ‘official opposition’ emerging due to lack of parties.

Solutions exist to minimise these concerns. Other consensus-based parliaments, such as Nunavut and the Northwest Territories of Canada, maintain that parliament selects Ministers rather than the head of the executive. Similarly, many coalition-based parliaments do not grant powers for the head of the executive to fire ministers outside of their party (e.g., Denmark). Tynwald had such a system to appoint members of the Executive Council and Chairs of Boards of Tynwald in the 1980s; this could be revitalised. Such an approach could work to widen the representation of views within the Council of Ministers.

Going further, making the Chief Minister first among equals – where the Council of Ministers collectively form the Government with the Chief Minister acting as Chair with limited specific powers – would maintain a figurehead and point of contact for external affairs, while addressing concerns that the Chief Minister is not popularly elected and has too considerable powers of patronage and institutional power when compared to their mandate as an independent individual in a single constituency. Such an appointment could be made automatically (e.g., based on seniority) and annually rotated, similarly to the President of the Swiss Confederation.

Second, the change from Boards – in which most Tynwald members were involved in Government policymaking – to the Council of Ministers has effectively reduced the number of members involved in initiating and considering policy choices. Departmental Members were created to address this shortcoming; however, their existence enables the Council of Ministers to secure block voting through collective responsibility – for some Departments creating automatic majorities in the House of Keys – in exchange for limited control of a relatively small policy area compared to older Board membership. Voting data from 2016-21 indicates that Department Members nearly always voted with their Department on Government- backed motions, negating the argument that the role of Departmental Members in block votes is insignificant. Previously, Board members voted as a collective on all matters before the Board, whereas Departmental Members are constrained to their policy area and under the control and patronage of the Minister.

Recent voting records suggest that, apart from all but the most contentious policies, the Council of Minister’s core block vote of nine may be combined with Departmental Members’ support (usually 0 – 3 members) to ensure that the Council of Ministers’ collective position on a matter is very difficult – or even impossible – to challenge. Given a block vote of between nine and twelve votes, motions are highly likely to succeed in Keys with 23 members (i.e., minus the speaker), or in Keys votes in Tynwald Court with 24 members. Application of a dynamical mathematical model of opinion formation (based on a small-world network) to Tynwald’s structure suggests that the closeness of the block votes to a majority in the Keys ensures that the Council of Ministers’ position on motions in Tynwald can generally succeed in more than 90% of cases, as the Council of Minister’s initial view exerts substantial influence over other members’ opinions. This may occur despite potential disagreement between branches (succeeding on a combined vote) and despite the initial opinions of Keys members outside of the block vote. During the Quayle administration, the Department of Health had a block vote of 12 in the House of Keys (consisting of Mr Ashford as Minister, and Mrs Corlett, Mr Moorhouse, and Mrs Barber as Members), ensuring Government- backed health matters were highly unlikely to fail (unless there was breakdown of the block vote or Departmental Members’ support). Similarly, planning policy and implementation were split across the Cabinet Office and Department of Environment, Food and Agriculture; for planning policy matters, both departments generally voted together. The problem is compounded when one or more non-block MHKs are absent: block voting becomes more significant as the number of members required for a majority falls. Further, increasing use of secondary legislation by Government combined with block voting has arguably led to ill-thought-out legislation to enter force with unforeseen consequences, as exemplified by policies and debates stemming from the Action Plan to Improve the Planning System. Some of the policies that arose from the Plan are now recommended for reversal by the Built Heritage Select Committee only a few years later. Tynwald is not directly involved in the drafting of secondary legislation and has no powers of amendment when voting on regulations in Tynwald Court, leaving the totality of sometimes substantial documents only able to be voted down in their entirety rather than allowing small mistakes or issues to be caught and amended.

Removal of Departmental Membership would inhibit the ability of the executive to push through potentially minority positions by astute application of Department Members. This could be compensated for by strengthening Tynwald’s role in policy development, such as by creating Boards (or Committees) to intercept incoming secondary legislation and other policy from Government. Such Boards could contribute to policy before it comes into force through these new mechanisms rather than Departmental Membership, potentially enabling more considered and widely supported policy to progress.

Third, the current budget process is constraining the ability of Tynwald and its offices to operate effectively or at all. The Tynwald Commissioner for Administration requested a budget for this financial year of £50,000, recognising that she would be unable to fulfil her statutory duties effectively with the existing budget of £32,000; in other words, she would be unable to investigate legitimate complaints from the public about Government, leading to a possible loss of public confidence in the process. Only minimal budget uplift of £2,000 was granted rather than £18,000, thus – according to her recent report – leaving new complaints now not investigated. Such budgetary control is further concentrating power into the Government and away from Tynwald and the public. The role of Treasury and the Council of Ministers in setting Tynwald’s budgets must be addressed.

Wherefore your petitioner seeks that:

Tynwald appoints a Select Committee of five members (with representation from both the House of Keys and Legislative Council) to:

Consider, make recommendations and – if applicable – an implementation strategy to reduce the powers of the Chief Minister to:

a. Remove powers to appoint and fire Ministers and return this ability to Tynwald.
b. Re-establish a Selection Committee of Tynwald for the purpose of recommending appointments to the executive based on merit, broad representation, and experience.
c. Remove the present role of the Chief Minister and replace with a new role that is allocated to a Minister on the principle of first among equals. Specific consideration should be given to applying a method of automatic allocation that operates on a rotating basis.

Consider, make recommendations and – if applicable – an implementation strategy to strengthen Tynwald oversight of policy development by one or more of the following:

a. Introducing Tynwald policy board(s) to widen political involvement with proposed secondary legislation (and other policy) before consideration in Tynwald Court, with specific aims of improving quality and political consensus.
b. Ensuring the executive cannot gain automatic majorities in House of Keys votes while maximising Tynwald members’ involvement in policymaking by: i. Abolishing Departmental Membership; and ii. Replacing with an equal number of positions in boards proposed in (a).
c. Any other mechanism identified that may achieve the same aim.

Investigate if appropriate measures are in place to ensure Tynwald and its associated functions (e.g., Tynwald Commissioner for Administration) are adequately funded within the current budget process.

The Select Committee should bring recommendations for reform to Tynwald no later than one year after their appointment.

– end

Isle of Man Population Projections: an Uncertainty Estimate

Thu, 22 Aug 2019 10:00:00 +0000

The Island’s housing need relies on accurate and up-to-date population statistics and projections. Government have committed to using the latest demographic information for the Area Plan process in the Strategic Plan. However, Government’s latest housing need figure does not take into account new observational data that was released as part of the 2016 census. Government’s population projections are also deterministic and do not therefore quantify uncertainty arising from assumptions about population change and measurement error in the Isle of Man datasets. It is desirable to quantify uncertainty to be sure that the precautionary principle is respected when used in land-use planning terms. The United Nations have adopted a Bayesian approach that applies hierarchical models to estimate Total Fertility Rate (TFR), life expectancy, and future populations, all with model uncertainty (Raftery et al 2014). This approach uses the same underlying intercensal compartment approach that is used in the official Isle of Man Government projections, but properly accounting for uncertainty. Here, I apply the latest UN method using the latest publicly available information to create population projections and show the bounds of uncertainty.

Note: I’ll update this post shortly with the full input data and original data sources in due course.*

Method

I collated historic datasets on Isle of Man demographics and integrated these into the United Nation’s global hierarchical population model. I conducted analysis using the UN-authored R package bayesPop (Ševčíková and Raferty 2016), which allows the implementation of thr UN modelling approach with additional data sources. The following inputs were required for the model: age- and gender-specific population counts, death rates, and fertility rates as a percentage of the total fertility rate; and projections of future sex ratio and age/gender-specific migration. I compiled a number of datasets from Government sources. Age- and sex-specific population structure for each census period was determined from census reports and academic publications (Kelly 1999). Due to lack of publicly available information in the Isle of Man, I used UK average age- and sex-specific mortality rates within the model.

Total Fertility Rate (TFR) was estimated with uncertainty by fitting a global hierarchical model to the United Nations global population projections dataset that was extended with Isle of Man TFR data obtained from Government reports (Alkema et al 2011, Fosdick and Raftery 2014). Similarly, uncertainty-based projections of male and female life expectancy were estimated from 1980-present Isle of Man life expectancy data (Chunn et al 2013). The predicted life expectancy and TFR were provided as inputs to the probabilistic population model. I used the same future projections of migration as the Government in their projections: zero, 500 and 1000 net migration based on the 2006 – 2016 age- and sex-related migration trend. Projections were calculated to 2100 but are truncated to 2040 in this post.

Result: Population Projections with Uncertainty

The probabilistic population projections were broadly similar to the Government’s population projections (see below graphs). From the 2016 population total of 83,314, my median prediction to 2038 under zero migration predicts less of a decline in population than the official Government projection to 2036.

Isle of Man projected population using deterministic future migration scenarios that are the same as the Isle of Man Government’s projections (Top = zero migration; middle = net 500 migration, bottom = net 1000 migration). The 95% Prediction Intervals (PI) represent the region in which we can be 95% confident of the projection given previous observations.

I also created another projection using a crude migration scenario where the age- and sex-related migration trend into the future is the same as the 40-year average (from 1976 - 2016). This probably isn’t realistic given the Island’s current economic conditions:

Isle of Man population projection based on the age- and sex-specific 1976-2016 migration trend.

The mean net migration rate to the Island from 1976 to 2016 was 480 people per year. This is similar to the Government’s current target of 500 per annum. However, the age-distribution is very different between the 2006-2016 and 1976-2016 means.

Implications for land-use planning and the ‘Area Plan for the East’

The 95% prediction intervals indicate uncertainty for population change for different migration scenarios. I suggest that the Area Plan should be refined to account for a median population prediction until 2026, with strategic reserves only allocated for the difference between the median projection and the upper 95% prediction interval. The complexity, however, lies in the assumption about the rate of future migration.

In the next post, I’ll be assessing the most appropriate future migration trend in light of demographic versus pension problems.

References

L. Alkema, A. E. Raftery, P. Gerland, S. J. Clark, F. Pelletier, Buettner, T., Heilig, G.K. (2011). Probabilistic Projections of the Total Fertility Rate for All Countries. Demography, Vol. 48, 815-839.

Fosdick, B., Raftery, A.E. (2014). Regional Probabilistic Fertility Forecasting by Modeling Between-Country Correlations. Demographic Research, Vol. 30, 1011-1034.

J. L. Chunn, A. E. Raftery, P. Gerland, H. Sevcikova (2013): Bayesian Probabilistic Projections of Life Expectancy for All Countries. Demography 50(3):777-801. doi:10.1007/s13524-012-0193-x

Kelly (1999), The demographic implications of economic growth in the Isle of Man. Population and the Environment.

Raftery, A.E., Alkema, L. and Gerland, P. (2014). Bayesian Population Projections for the United Nations. Statistical Science, Vol. 29, 58-68.

Tiny grains, big data: the Global Pollen Project (crossposted from Methods.blog)

Thu, 18 May 2017 10:00:00 +0000

The Global Pollen Project is a new, online, freely available tool developed to help people identify and disseminate palynological resources. This article was originally published for Plant Conservation Day, on Method.Blog.

The Global Pollen Project is a new, online, freely available tool developed to help people identify and disseminate palynological resources. Palynology – the study of pollen grains and other spores – is used across many fields of study modern and fossil vegetation dynamics, forensic sciences, pollination, beekeeping, and much more. This platform helps to facilitate cross/multi-disciplinary integration and discussion, outsourcing identifications, expertise and the sharing of knowledge.

Pollen’s Role in Plant Conservation

Successful conservation of rare, threatened, and valuable plants is dependent on an understanding of the threats that they face. Also, conservationists must prioritise species and populations based on their value to humans, which may be cultural, economic, medicinal, etc. The study of fossil pollen (palaeoecology), deposited through time in sediments from lakes and bogs, can help inform the debate over which species to prioritise: which are native, and when did they arrive? How did humans impact species richness? By establishing such biodiversity baselines, policymakers can make more informed value judgements over which habitats and species to conserve, especially where conservation efforts are weighted in favour of native and/or endemic flora.

A selection of pollen grains from the Global Pollen Project

Fossil pollen also contributes greatly to our knowledge of spatial-temporal dynamics of plant populations through time. Using this information, it is possible to assess the resilience of species to environmental change: how have changes in past climate and environment influenced the presence of certain species? Where did they persist and where did they die out? Such research can assist efforts to understand how resilient plant populations are to future climate and environmental changes, showing us which areas will be suitable now and in the future.

The Process of Pollen Identification

A major limitation when using pollen to inform biodiversity conservation is taxonomic identification. The identification of a pollen grain to a plant species is traditionally carried out via morphological analysis under a light microscope. Each grain is compared morphologically to reference material in the form of reference collections. These collections are often only accessible in the form of glass slides (like in the image above) or in printed books, which are bound to a specific physical location, limiting their access and use.

The pollen identification and counting process is labour intensive. Identification is conducted based on the morphology of pollen traits, such as their size, shape, pores and furrows, surface patterning, and surface sculpture. This process, by its nature, is hard to replicate, with identifications requiring the build-up of significant identification skill, especially for tropical environments with high species richness. The ease of identification also depends on the location and preservation of grains. For example, pollen from dry, tropical environments tends to be more degraded, which makes it harder to give a species identification. So in the tropics, there’s a significant overlap between rare and threatened biodiversity, and difficulty in pollen identification.

The Global Pollen Project: Crowdsourced Identifications and Global Reference Material

To help make pollen identification quicker and easier, we developed the Global Pollen Project (GPP) – an open, peer-reviewed database of global pollen morphology, where content and expertise is crowdsourced from across the world. Our approach to developing this tool was open: open code, open data, open access. It connects to external services, including the Global Biodiversity Information Facility and Neotoma Palaeoecology Database, to provide botanical descriptions and occurrence data for each taxon, alongside pollen images and metadata. There are currently two major avenues through which data is fed into the GPP: digitisation of reference material and the submission of pollen grain images of unknown taxonomic identity.

Digitising Pollen Data

The digitisation of existing reference material is at the heart of the GPP, providing the bulk of the digitised information. Most submitted reference material is modern and sampled from individual plants. This means that it has greater taxonomic confidence than morphological identifications. The digitisation tools are also open to any researcher – from large institutions to individual PhD students – to archive their pollen reference material, and stop it lapsing into ‘dark data’, or becoming degraded. We emphasise the importance of focusable images (or z-stacked images), which provide essential depth-based perspectives on the 3D structure of grains (although we do accept static images too). You can see some examples here, here, and here.

Crowdsourced Identification

A key aim of the GPP is to help improve identification skills through sharing and learning. Users can submit images using a fixed microscope camera, or even a smartphone, providing additional spatial-temporal metadata, which then enter the pool of unidentified material. Other users can then submit morphological identifications to family, genus, or species rank. The GPP weighs up different types of taxonomic identification using in-built algorithms and provides a ‘confirmed’ identification to the material when conditions are met (we currently require at least three identifications with an agreement of 70%). Such ‘confirmed’ grains are then added to the global reference collection with its metadata. In this way, crowdsourcing identifications provides benefits to the submitter and the wider community. To incentivise identifications, each unknown grain has an available score, which an institution or individual can gain through helping to identify the grain in question, providing friendly competition between labs (there’s nothing wrong with a bit of healthy gamification).

Education

The GPP has been incorporated into undergraduate teaching in many institutions since its launch, including the University of Southampton and UCLA. Although offline ‘physical’ reference sets are still useful, this tool can provide access to training in places where large reference collections are not accessible. It also preserves material for the long-term, ensuring that it will be available to the next generation of students.

Bridging the Gap – Palaeoecology to Biodiversity Conservation

Palaeoecology can provide rich insights into environmental history that are highly relevant for plant conservation concerns. With the GPP, we have attempted to bridge some of the divides that limit the direct applicability of pollen research to biodiversity conservation.

Translating Botanical Nomenclature

Pollen research has traditionally used morphological types are the basic taxonomic unit, which do not necessarily translate to botanical species or genera. The GPP was constructed with a dynamic taxonomic backbone. This currently uses the latest version of The Plant List, ensuring that every identity within the system matches a confirmed taxonomic name. The Plant List is the only complete working list of all known plant species detailing all current accepted names and synonyms in floristic taxonomy. Pollen morphotypes are sometimes used in taxonomic nomenclature, particularly in fossil pollen identifications. This is useful and important information, but we think that it’s best preserved as tagged metadata. It’s important to keep this practice separate from attributing pollen grains to their associated plant.

The GPP automatically keeps the botanical identity of every pollen grain up-to-date, by assessing the splitting, clumping, and renaming of taxa through time. The backbone is essential in ensuring that identifications are valid and comparable between pollen grains or slides collected at different times. We also leave the ‘taxonomic trail’ to follow from the original identification, so that the original slides can be traced from digitised slides, even after taxonomic changes. This also lets us directly display our pollen morphology data with modern distributions and data from the Global Biodiversity Information Facility.

Tiny Grains, Big Data

The GPP aims to provide the first global key of pollen morphology. Although there are still significant taxonomic gaps, we’re working hard to fill them. There are currently 2,877 images of pollen grains in the online reference collection. These cover 167 families, 767 genera and 1,565 species from Britain & Ireland, Europe, North America, Mexico, Mongolia, Madagascar, and the Canary Islands. This expanding collection could be developed to create a more accessible, interactive pollen key, aiding efficiency in pollen analysis, and allowing the pace of data generation to quicken. There is also the potential to ask new questions of the data through the connection of morphology to botany, especially when combined with image-based machine learning.

With the GPP we aimed to bridge the gaps between plant and environmental scientists, ecologists, and palaeoecologists, hoping to stimulate cross-disciplinary research. We hope that better keys will enable more successful differentiation between genera and species, some of which may be of conservation concern while others not, to enable the creation of more accurate biodiversity baselines. Ultimately, this should help policy-makers to make informed judgements about plant conservation.

Using pollen to map plant species occurrence through time: a D3.js example

Sat, 30 Jan 2016 10:36:17 +0000

For the Global Pollen Project, I wanted to create an interactive map that integrates modern observations of plant species with evidence of their distribution in the past.

Past Occurrences of Betula

Colour scale from blue to yellow indicates the ‘last seen date’ at a location, with yellow the most recent.

Modern occurrence data is relatively easy to access: the Global Biodiversity Information Facility (GBIF) provides an open API to developers to query occurrences of species. The GPP automatically assigns a GBIF taxon ID to each family, genus, and species.

Obtaining past occurrence data using paleoecological / long-term ecological data was a bit more tricky. Neotoma is a great database for paleoecological data, including pollen, diatoms and ostracods. Neotoma also provides an open API, and calibrates the age-depth models of sedimentary data, meaning that the dates given to the presence of taxa are more reliable. Using pollen data from Neotoma, it is possible to plot distributions through time.

There are some issues when comparing the modern observations from GBIF to pollen presence in Neotoma data.

First, Neotoma does not return recursive results up the taxonomic heirarchy. For example, if a pollen grain is identified as a Betula, this grain will not be returned when searching the API for Betuleacae. GBIF, on the other hand, has a solid taxonomic backbone and accounts for heirarchical relations.

Second, the presence of pollen is not just reliant on the presence of a taxon, but also on the preservation characteristics of the grain, the environment in which it was deposited, and the ability to distinguish this grain from others, at family, genus and species resolutions. There are also radically different spatial patterns of sampling effort between modern observations and sediment core data.

Third, the palaeoecological data is sorted by ‘pollen types’, rather than linked to botanical species designations. The interpretation in these maps therefore contains certain assumptions about the relations between pollen morphology and species.

The example on this page just plots the Neotoma pollen data. Check out the code at my GitHub repo.

Visualising climate model ensemble outputs in D3.js

Sat, 21 Nov 2015 10:36:17 +0000

Recently I was tasked with prsenting data for future climates to a diverse user base (public, professional, and scientific). It took a lot of time to understand how best to summarise complex outputs of global climate models, and how to display this information in an easy to understand format to end users of our software.

I found some great visualisations of climate ensemble statistics at the Climate Service Centre Germany, and thought these would look good in our climate data interface to highlight the differences between emissions scenarios.

Isle of Man: Projected Mean July Temperatures

On the left panel of this figure, the light grey area represents the likely range of mean July air temperature (10m above ground). The darker grey area represents the very likely range of future temperatures. The coloured lines indicate the ensemble medians for the various emissions scenarios (RCP2.6, 4.5, 6.0 and 8.5).

The right panel is designed to highlight how the ensemble medians differ through time. In the near future, the predicted values are very similar, but as time progresses the high emissions scenario develops a larger range and median temperature.

While I was working at Microsoft, I developed an Azure-powered system that collected the latest climate model outputs, created summary statistics and presented these in FetchClimate. This data was pulled from the datasets that I produced.

Check out the source code for this graph at my GitHub repo.

An Infinite Variety of Names

Thu, 04 Dec 2014 10:36:17 +0000

Despite a rich scientific heritage of environmental exploration and classification, we are still not much closer to understanding how many unique species populate our planet.

Estimates vary widely from as little as 5 million up to 100 million, a staggering degree of difference. To reach these numbers, the majority of studies extrapolate from historic rates of species description, but how many unique species have already been described? This question is more difficult to answer than it initially appears, as our taxonomies are littered with many names for a single species. This makes interpreting both species richness and the attached occurrence data more complicated.

What’s in a name?

Taxonomic nomenclature is based on the Linnaean system, comprising of a genus, species and the author who described the specimen. The name must be linked to a type specimen to be valid. When the species is formally described twice, either through coincidence or a misinterpretation of population-level variation, it can end up with two different names. As the names chosen are often subjective and based on a variety of social and environmental cues, the resultant names may be very different. In more recent decades, reinterpretation of existing specimens and molecular investigations have led to a proliferation of name changes. Species are now often moved to a different genus or split into multiple species, creating more synonyms.

These taxonomic synonyms are not synonyms in the literary sense. Rather, there can only be one ‘correct’ biological synonym for a species at any time. For example, the Western Hoolock Gibbon, Hoolock hoolock, has also been known in the genera Simia, Hylobates and Bunopithecus, but these names are now antiquated. When a species is described in parallel or with different regional names, who decides the ‘correct’ name? This depends on the taxonomic group. For example, birds are well studied and there are governing bodies to make these decisions. However, for plants there has historically been no central authority. This has led to highly variable names on a continental, country-specific and even regional level. Some herbaria and gardens have recently attempted to collate lists of known plant species and their synonyms, a famous example being Kew’s ‘The Plant List’. This database is not freely downloadable, however, making it difficult to access. Regardless, synonyms provide a layer of complication for managing and interpreting biodiversity data.

Identifying the Identifiers

For the conservation of biodiversity, synonyms are an inconvenience. An ideal species identifier would not vary through space and time, whereas current taxonomic names can change rapidly. Without a coordinated central taxonomy, remnant synonyms can float around inflating species counts, creating problems for biogeographical studies. For example, one interested in the Western Hoolock Gibbon would have to know of the four synonyms, search for occurrences of these, and aggregate the data. As these untidy elements litter our taxonomic information, more time and effort is wasted sorting our data.

Bioinformatics — the application of technology to biodiversity data — offers a bright future to overcome this problem. For example, the Global Biodiversity Information Facility (GBIF) collates species globally from 40 taxonomic databases and occurrence data from a range of contributors, and has recently created programatic public access to its species database. Cross-validation between different records is used to classify names into accepted, synonym and questionable. Thus, occurrence data can be grouped for a species despite being recorded against synonym taxa. However, this approach is not without its faults. Last September, a GitHub ‘paper’ criticised the level of synonym detection in GBIF. The author used a simple method of detecting when a species, author and year combination existed more than once. Although this may include false positives, it flagged taxa that may be problematic. For example, it identified the frog family Rhacophoridae, which on further examination had competing classifications from two different databases both appearing in GBIF.

Online data aggregation efforts have shown great potential to decrease uncertainty relating to the species synonym problem. I believe that open, accessible and aggregated data is key to mobilising experts, both from institutions and the armchair, to create a taxonomy with minimal species duplication. GBIF, working with other initiatives, still has a long way to go though. The Plant List, for example, has 22.8% unresolved plant names. Using cross-tabulation, we could make better informed estimates of what we have and better models of where it is, greatly enhancing our knowledge for biodiversity conservation.

Barcoding our Biodiversity

Thu, 04 Dec 2014 10:36:17 +0000

The huge variety of traits in fauna and flora became an obsession for seventeenth century colonial travellers, who catalogued and classified their new encounters, sending samples home to form large collections, such as those in the Oxford Museum of Natural History and Kew Gardens.

To delineate their differences, ‘species’ were determined based on morphological traits, following the naming conventions of Linnaeus. Darwin famously wrote of species:

“No one definition has as yet satisfied all naturalists, yet every naturalist knows vaguely what he means when he speaks of a species.” (Darwin 1859)

Surprisingly, little has changed in this regard in the 155 years that have passed. The most universal species definition is biological, stating that species can breed to produce viable offspring. However, the biological definition fails to account for species that interbreed and produce hybrids, such as oak trees, and is not useful for populations that become geographically separated and as a result cannot interbreed. Similarly, the morphological definition struggles with cryptic species, where their traits are very similar but their genetic makeup distinct. As conservationists, how can we best identify ‘species’ to maximise the protection that can be provided to biodiversity?

This is where the mystery mushrooms come in. A recent paper – picked up by popular media – found that a package of dried porcini mushrooms purchased in a London grocers contained three species new to science. The researchers from Kew identified the species using DNA barcoding, a relatively recent concept that uses a standard snipped of DNA to identify species. Although this technique is not intended to determine the evolutionary trees that connect species, it has been used to identify cryptic species and quickly identify morphologically difficult individuals. As an example, researchers in the USA identified 15 new North American bird species using barcoding. This is especially surprising, as scientists and birdwatchers have heavily studied these birds. However, DNA barcoding cannot replace traditional taxonomic methods as commonly suggested in the media, as although it can show that a species is not in the current taxonomy, it cannot fully characterise the new species. More thorough genetic analysis, which can reconstruct evolutionary lineages and formally analyse the genetic traits of species, can be used for species where morphological approaches are difficult. These methods are intensive and expensive, so cannot be applied with any great haste.

DNA barcoding, although not a fully featured taxonomic tool, is on track to revolutionise biodiversity conservation. In both science and practice, conservation can benefit from the rapid ability to identify morphologically difficult taxa, such as our mushrooms, to a high accuracy without time-consuming analysis by taxonomists. This does not make taxonomists redundant, but the contrary. As the use of barcoding identifies more unknown species, expertise will be required to classify them correctly into the species hierarchy. Not only will occurrence identification improve, but also data sharing. As DNA barcodes are small information snippets, they are perfect for ‘big data’ uses, connecting all known information of a species together with a central identifier. It is also useful for ‘crime scene conservation’, such as identifying illegal species trading. This is because it can be completed on ‘incomplete’ specimens, such as shark fins. Barcoding is often seen as a method of classification, but its potential to help biodiversity science is more profound as a common identification tool of the future.

To learn more, check out the International Business Times, Barcode of Life project and Rockefeller.

The Ancient Lake that Fertilises the Amazon

Tue, 11 Nov 2014 10:36:17 +0000

The Amazon Rainforest is one of Earth’s most biodiverse locations, and recent evidence suggests that it is tightly coupled to one of the most barren. Using satellite observations, Saharan dust storms have been tracked across the Atlantic to the Caribbean and South America. A unique combination of geological and climatic drivers has created these conditions, which are hypothesised to transport essential nutrients to the Amazon.

An Algal Fertiliser

The Sahara Desert has not always been barren, but was wet at various times between approximately 4,200 and 12,000 years ago. Fossil pollen provides evidence of diverse tropical plant communities, while fossil lakes, rivers and fish show a network of ancient watercourses and lakes. During these pluvial phases, the filling aquifers enabled the formation of Lake Megachad, a lake with a surface area of around 400,000 square kilometres at its peak. As the basin is relatively flat, the lake was able to expand and detract rapidly, and now Lake Chad is only 1,350 kilometres squared. The lake was teaming with algal life. Diatoms are one of the most common groups of phytoplankton, and were present in large quantities in Lake Megachad. When it dried up, it left behind thick deposits of freshwater diatom shells, which hardened to form diatomite. This substance is rich in phosphorous, an essential plant macronutrient, as well as iron. The northern section of Lake Megachad is often termed the Bodélé depression, and is responsible for 95% of the Saharan dust exported to the Atlantic. The geological layout of the region creates a wind tunnel into the depression between the Tibesti and Ennedi mountain ranges, whipping up huge dust storms of light, fertile diatomite. The dust is so fine that it can be carried out across the Canary Islands and beyond, over the Caribbean and even as far as the Amazon.

The export of nutrient-rich dust has implications for the productivity of the Amazon Rainforest, as well as complex cloud-seeding properties for weather systems. In the Amazon, intense competition between plants causes nutrients to be cycled from soil matter quickly, leaving a relatively nutrient- poor soil. The Amazon is also home to some of Earth’s oldest soils, as the tropics have not been glaciated like mid- and high-latitudes. This creates a phosphorous-limited ecosystem, as this macronutrient is obtained from breaking down underlying rock. The Saharan dust is suspected to have ‘fertilised’ the Amazon, providing a steady covering of phosphorous and iron-rich dust, supporting biomass and biodiversity.

Dust: Past, Present and Future

There are many unanswered questions regarding the transport of dust to the Amazon, including the nature of fertilisation and the resilience of any beneficial effects to climate change. For climate scientists, how will rainfall vary during the 21st century, and how will this affect dust transport? Since the 1970s, there was concern that large-scale desertification was taking place in Africa. However, recent research suggests that there has been a regionally variable pattern of greening and desertification, while climate models cannot agree on a future trajectory. The nutrient flow also requires full characterisation, to determine the relative importance of this nutrient flow to the Amazon’s health. It has been suggested that, if Lake Chad was to expand towards its past extent, the export of dust from the Bodélé depression could be hindered, and that this would cause lesser productivity and ‘dieback’ of the Amazon. However, the Amazon did not ‘die back’ during the Holocene Saharan pluvial phases. To fully understand this system, collaboration between biodiversity scientists, geographers and climate scientists will be required. For more information on this topic, have a look at my sources: African dust keeps Amazon blooming, Shorelines in the Sahara and the Oxford Drylands Group.

The Dangers With Economically ‘Valuing’ Nature’s Benefits

Wed, 05 Nov 2014 10:36:17 +0000

Ecosystem services - the benefits of nature to humanity - are increasingly an essential concept in policy that promotes the sustainable use of natural resources.

A range of complex ecosystem processes lead to the production of so-called provisioning (direct products), regulating (indirect benefits) and cultural (intangible) ecosystem services. Through an understanding of the processes that lead to their production, these services can be quantified and compared within spatial land-use planning frameworks, such as Co$tingNature and ARIES. This approach has been extended in many circumstances to bring ecosystem services into economic markets and assigning monetary values, with well-known examples including REDD+ and EU carbon credits. The commodification of nature, however, is not without controversy, and a range of issues have been identified. The October 2014 issue of Science contains a perspective by Bill Adams, who suggests that valuing nature in this way is not always beneficial for biodiversity conservation, and should only be accepted when it has benefits for both the environmental and society. Here I add my own thoughts to this work, outlining briefly what I believe to be some of the challenges of assigning values to ecosystem services.

Biodiversity vs Ecosystem Services

How does an ecosystem produce services? The combination of species and their interactions with the abiotic environment gives rise to services, but the role of biodiversity in the processes governing service production is poorly understood. There is also no guarantee that rare or threatened species are responsible for ecosystem service production. For example, a monoculture of plantation trees can provide comparable soil stability to a mixed forest. In fact, often the common species provide the majority of ecosystem functioning, such as oak and horse chestnut trees in an English forest. As a consequence, biodiversity is sometimes listed as a service in and of itself, but this ignores the role of biodiversity in the production of services. Through focusing on particular services within policy, as a consequence changes in biodiversity may (a) not conform to the goals of species-based conservation and (b) modify production of other connected services.

Back To The Future

By giving nature a value, we bring it into the human domain. Nature no longer exists for its own sake, but because as humans we find it beneficial. There is also a question regarding potential value, as new discoveries or scientific evidence may identify previously unseen benefits from ecosystems. Current valuation methods cannot account for these precautionary elements. For example, frameworks that attempt to value intangible cultural services, such as our enjoyment of nature, cannot ascertain future values from current consumers. Such imperfect methods of valuation may be misleading within a policy framework where services must be prioritised.

The ecosystem services framework should not be used alone, but complementary to spatial conservation prioritisation. Without a dual consideration, we could see the irreversible loss of unique biodiversity that may hold value to future cultures. There is also no guarantee that attaching an economic value to essential services will protect them from loss. For example, the Stern report has failed to generate significant action despite the clear economic benefits of early climate mitigation. The ecosystem services approach should be applied as an instrument from the conservation toolbox, but is not a panacea.

Britain's Hidden Nitrogen Addiction

Mon, 29 Sep 2014 10:36:17 +0000

The British Isles are home to a group of farming nations. In the United Kingdom, approximately 77% of the total land area is used for farming, providing a substantial contribution to regional and global food security. Since the mid 19th century, increasing intensification of our agricultural system has led to a hugely important increase in food production, but also a dependency on inputs that is spiralling out of control.

Nitrogen is an essential nutrient for plant growth, and is required in large amounts to sustain this production level. In the UK, approximately 1.375 million tonnes of synthetic nitrogen fertiliser is applied annually, with around 0.5 million tonnes imported from overseas1. It not only travels long distances, but is very energy-intensive to produce (through the Haber-Bosch process), making nitrogen fertiliser the dominant contributor to energy use in British agriculture. These inputs must be reduced because of this energy imbalance, but also as they are environmentally damaging for biodiversity. The relative inefficiency of conventional agricultural systems in retaining nitrogen inputs means that high inputs cause high environmental leakage. In this reactive form, nitrogen causes large-scale changes in the natural nitrogen balance of ecosystems, removing nutrient limitations and changing natural communities. Predominantly though agriculture, we have hijacked natural nitrogen cycling, bringing it firmly into the ‘anthroposphere’.

Organics and the Anthroposphere

Current environmental and agricultural policymakers have hit a conundrum. They strive for ‘sustainable intensification’, as growing populations are increasing food demand while the environment requires a drastic reduction of synthetic nitrogen inputs. Cutting edge research at Oxford is currently looking to engineer crop plants to associate with nitrogen-fixing bacteria and express nitrogenenase, which would enable them to ‘fix nitrogen’. However, Oldroyd et al (2014) suggest that this complex engineering conundrum could be a decade or more away2.

In the absence of technological advancements, how can we achieve sustainable intensification? Organic agriculture has embraced many input-reducing techniques under the guise of environmental sustainability and concerns over health. These include increasing soil-organic-matter through low tillage, intercropping with nitrogen-fixing legumes and more efficient use of green manure (including recycling). However, organic farming methods have been demonstrated to reduce yields by 5-40% over conventional farming in a European context3. A recent meta-analysis of European organics research found that organic farming requires 84% more land because of this yield drop and the necessity to intercrop with legumes4. Although identified as giving valuable contributions to biodiversity conservation and reducing our synthetic nitrogen reliance, the decreasing yield is a threat to achieving sustainable intensification. A better solution has been suggested as taking some of the best nitrogen-conserving approaches in the organic method and applying them to conventional systems, so that highly productive non-organic species and effective pesticides are not shunned to realise higher yields than a completely organic system.

A Question of Scale

Organic methods contribute to solving the ‘sustainability’ aspect of sustainable intensification, but at the cost of yield. On a British scale, the primary concern is environmental, whereas on a global scale the concern is food security. The solution, I believe, lies with the spatial scale in which the problem is defined. Badgley et al (2007) have suggested that if all land in the world currently used for agriculture was converted to an ‘idealised’ organic system with little synthetic nitrogen input, more than enough food to feed the current population could be produced5. Perhaps then, British and European farming could incorporate nitrogen- and biodiversity-conserving traits from the organic system, resulting in lower yields, whereas regions blanketed in low-productivity subsistence agriculture could be encouraged towards higher-yielding and more intensive organic methods. These policy-driven changes would help to limit the damage of unsustainable conventional agricultural expansion. For the British Isles, the pastoral ideal envisaged by conservation organisations would be more attainable, while local food security could be increased for developing nations. These ideas are of course a simplification of a complex problem, as the nutrient requirements and productivity of agriculture are subject to high local heterogeneity, so the most appropriate methods from the ‘organic toolkit’ would need to be implemented through policy at the regional and local scales.