by Patrick van Zwanenberg and Anabel Marin (Conicet / Cenit / UNSAM)
It is sometimes said that plant breeders breed their aspirations about how agricultural production systems ought to function directly into their seed varieties. Over the last three decades there has been a collapse in the diversity of seed breeders (and with it the range of different ideas about agricultural systems that get inscribed into seed varieties), as a handful of giant chemical/seed biotech firms have come to dominate both seed R&D and commercial seed markets. This rapid, unprecedented shift to an oligopolistic industrial seed sector has been enabled in large part by the worldwide imposition and diffusion of strict intellectual property rights, especially patents, over biological material.
Against that backdrop, STEPS America Latina is co-designing an open source license for seeds, with the purpose of defending and recovering a traditional culture of openness, sharing and and collaboration in seed breeding. Our work, conducted as part of the Pathways Network research project on transformations to sustainability, aspires to open up space for people interested in increasing rather than diminishing the diversity and availability of plant genetic resources, and who want to develop plant varieties adapted to many different agricultural settings, and for diverse agricultural practices and production systems. Such diversity is badly needed as a font of ideas for building more sustainable pathways of agricultural change.
The notion of ‘open source’ innovation originated in the field of computer software in the 1980s, as a response to decisions by software firms to restrict access to software code that programmers had traditionally shared freely amongst themselves. Copyright law was used to create a license for releasing software programmes that specified that the source code used in the programme was freely available to anyone to study, change and distribute for any purpose, but only under the condition that all subsequent changes were distributed under the same terms. The latter part of that sentence is critical. Software innovation is cumulative, and the license means that as a piece of open-source software is improved upon, over time, no one can exclusively appropriate new versions, or new components they have designed, or restrict others from using and further improving upon them. Software released under such ‘copy-left’ licenses therefore becomes part of a ‘protected commons’, available to all who agree to the rules not to exclusively appropriate.
Seed innovation, just like software, is a cumulative process, but in two senses. First, when a plant breeder crosses two plant varieties to create a new one, her novel variety comprises an incremental change to germplasm that has itself been repeatedly altered over decades of previous plant breeding and thousands of years of plant selection. Second, germplasm is both an output of the innovation process and an input, and so plant breeding depends critically on unhindered physical access to diverse sources of interesting germplasm. The intellectual property rules that have applied to seeds over the last half a century, known as plant variety rights, recognised that breeding was not possible without physical access to previous innovations. Those rules allowed plant breeders and farmers to freely use protected seeds as parental material for creating and commercialising new varieties. Those so called ´breeders rights´ – a recognition of both the need for germplasm as an input to further breeding, but also that seed is the cumulative result of many previous breeders and farmers – are increasingly being restricted, especially by the ability to take out patents on the gene sequences inserted into genetically engineered crop varieties (and in some countries on whole plant varieties). Patented material cannot be used by anyone else for commercial ends without a license from the patent holder, and plant breeders cannot use seed varieties that contain patented genes as a basis for breeding and commercializing new varieties because patented gene sequences cannot be practically separated from the rest of the so called ‘background’ germplasm. It is as if the entire germplasm – containing the efforts of countless historical breeders and farmers – has been captured by the patent holder of just one or two inserted genes and locked away.
Since patent protection became available for engineered gene sequences a small handful of global chemical/seed firms have bought up hundreds of smaller seed breeding firms, and now control most of the world’s commercial breeding and more than 60% of global seed sales. Those firms tend to focus on large commercial seed markets, and on resolving production problems that are commercially significant and/or that can be tied to use of their chemical pesticide products. Minor crops, marginal agro-ecological environments, and production constraints that are not of significant commercial importance are neglected, resulting in diminishing crop diversity, and a much narrower range of agricultural systems that the seed sector is able to support.
In response to these developments, groups of people in a number of countries are experimenting with open source ideas for seed innovation. The first to do so was a group of plant breeders and social scientists based at the University of Wisconsin in the USA, a country where patents are widely used to protect both gene sequences and entire plant varieties. In 2012, the group launched an open source pledge to accompany the sale or transfer of new seed varieties, based on copy-left principles. The pledge states “You have the freedom to use these [Open Source Seed Initiative]-Pledged seeds in any way you choose. In return, you pledge not to restrict others’ use of these seeds or their derivatives by patents or other means, and to include this Pledge with any transfer of these seeds or their derivatives.” Some 450 seed varieties supplied by 50 small breeding firms have been released with that pledge.
The pledge effectively creates an alternative pool of germplasm to that currently held under proprietary control through the use of patents, and which can be freely accessed and exchanged and used for further breeding, or anything else, now and into the future. The giant chemical/seed biotech companies, like anyone else, could of course access and use germplasm from that pool as a parent from which to create new varieties but they would then be obliged (assuming the pledge could be enforced) to release their new varieties under the same open source conditions. This is a requirement that is incompatible with those firms’ current business models, which rely on the use of patents.
In Argentina, we are exploring, with plant breeders and other stakeholders, how open source principles could be used to create a protected commons in germplasm, and how it might function. We are developing a material transfer agreement, a form of contract often used to exchange biological material, or a clause to add to an existing material transfer agreement, that reflects open source principles. We will call that kind of contract or clause ´Bioleft´, and we are developing a web-based platform for recording transfers of Bioleft-protected material.
Creating a workable and working open source contract or clause is challenging, for many reasons. One is that most plant breeders in Argentina are currently able to access and exchange germplasm without too much difficulty, and so few see an immediate need for such a license, mostly those who have previously encountered patent-based restrictions on their work. However, where we have had a chance to discuss open source innovation ideas with plant breeders they have generally been overwhelmingly supportive of an open source transfer agreement so as to ensure that new varieties remain freely available for further breeding, if or when restrictions were to emerge in the future.
Another challenge concerns the scope of an open source contract. We envisage, several variants of the Bioleft contract, though all would contain the same core clause that will stipulate that when seed material is released or transferred, the recipient cannot place any restrictions on access to and use of that material for research and for the commercialization of new derived varieties.
One of the variants would stipulate that when seed material is transferred, the recipient cannot place restrictions on any use of that material. This is the approach that has been taken by groups in other countries that are experimenting with open source seed licenses or pledges. It would mean, for example, not only that access to germplasm for breeding was unrestricted, but also that farmers would be free to save, replant, and exchange germplasm released under such a contract. The practice of saving, replanting and sharing open pollinated seed is a long-standing one amongst farmers, but it too is outlawed by the extension of patents to seed material (and often limited by plant variety rights). Patents enable seed companies to insist, in effect, that farmers rent their seed for a single growing season. This issue, more so perhaps that the restrictive effect of patents on plant breeding has provided much of the focus for international and national campaigns against the extension of patents to biological material. So why are we not advocating only this variant of a Bioleft contract?
The main reason is that no restrictions on any use would also mean that anyone could multiply and sell a seed variety released under those terms. Commercial seed firms would probably be unwilling to release varieties under those terms because competitors or anyone else could immediately multiply and market the same variety. Furthermore, some public sector breeders have told us that whilst their institutions would be likely to support use of a Bioleft contract that ensures that germplasm is always freely available to other breeders for research and the commercialization of derivative varieties, those institutions would want to retain the option to exclusively license their germplasm to commercial seed firms, in order that those institutions can continue to earn royalties, which are used to maintain their breeding activities.
We therefore plan to develop other variants of a Bioleft license, for example, that stipulate that while a recipient cannot restrict access and use of the material for breeding and commercialization of new derived varieties, or for non-commercial uses (i.e. saving, replanting and exchange), restrictions on multiplication are possible. That kind of license could, in effect, allow a seed breeder to maintain a monopoly on commercial sales of their variety, as domestic seed intellectual property legislation permits, or allow, say, a public sector breeder to transfer material to a seed firm with exclusive rights to multiplication and sale.
With such variants, we do not see how it would be possible to avoid different version of a Bioleft contract being used for the same seed material, depending on who is making the transfer.
Bioleft contacts that allow some restrictions to be placed on seed material might be seen by some commentators as inconsistent with the philosophy and ambitions underpinning the open-source movement. The vision is of a distributed, decentralized system of peer production of seed material, akin to the experience of free/libre open source software production, in which small farmers develop improved varieties collectively (as they have done for centuries) but this time in conjunction with supportive public sector breeders. This would require a means to prevent any exclusive appropriation, and would therefore probably be of little interest to commercial seed firms. Our conversations, thus far, with breeders, indicate that support for an entirely unrestricted use contract or clause exists amongst public sector breeders developing varieties for and with small-scale family farmers, but less so elsewhere.
Our view, for the time being, is that it is desirable to enrol as many different actors as possible into an open source seed innovation model. A contract that is attractive to public sector breeding institutions that engage with different kinds of agricultural systems, and that is also attractive to some domestic seed firms, might mean that an alliance of actors is created that effectively rejects the patenting of seed material – and that produces a growing pool of open source germplasm that cannot be patented.
The ambition, and one that is close to the spirit of the research programme that is funding us, is to co-design a workable open source model in conjunction with as many users as possible, and to ensure that such a model is put into practice; that is, to work with breeders to release new plant material under a Bioleft contract. Many challenges loom, but the learning is in the doing. The practice of co-design is central. It not only enables collective learning about how different actors view problems, priorities and possible solutions, but also occasionally throws up unforeseen opportunities (unforeseen, that is, to the research team). For example, our plans for a digital platform to record transfers of Bioleft-protected material were seen by some public sector breeders working with small scale family farming as also a means of facilitating participatory breeding initiatives (i.e. seed improvement that involves close collaboration between breeders and farmers). This too is an opportunity that we plan to support.