22 APRIL 2011 VOL 332 SCIENCE www.sciencemag.org
ference of the Parties to the Convention on
Biological Diversity (COP-10) highlighted
ecological restoration as a signifi cant oppor-
tunity for achieving global conservation
goals ( 2). The restoration of nature, natural
assets, and biodiversity is now a global busi-
ness worth at least $1.6 trillion annually and
likely to grow substantially ( 3). Although
seed banks have emerged as a tool to pro-
tect wild plant species ( 4), off-site (ex situ)
conservation measures at seed banks must be
complementary to “on the ground” manage-
ment at the conservation site. For example,
whereas global targets are for restoration or
management of at least 15% of each ecologi-
cal region or vegetation type ( 1, 5), recogni-
tion of the mechanisms required to achieve
these goals is largely absent from policies.
How are large-scale (100- to 1000-km2)
plant reintroductions that recreate biodiverse
communities to be achieved? We argue that
use of native seeds underpins achievement of
these restoration targets and that seed banks
need to shift from being “stamp-collections”
of species to collections that can deliver res-
toration-ready seeds at the scale of a metric
ton and larger. We propose the concept of the
restoration seed bank as a facility that looks
beyond the core skills of collection and stor-
age of germplasm ( 6) to rigorous science-
based restoration-use of germplasm, seed
farming, training, and information dissemina-
tion. These functions must be linked to com-
munity- and industry-based restoration initia-
tives. Connecting science to the community is
particularly important ( 7), with opportunities
at a local scale to develop traditional foods and
medicines into the restoration palette through
traditional ecological knowledge ( 8, 9).
ith nearly two-thirds of the world’s
ecosystems degraded ( 1), the
October 2010 meeting of the Con-
Scale in Restoration
Seeds are the primary tool for reintroducing
plant species ( 10– 14). But effectively using
seeds of wild species in contemporary resto-
ration is facing a crisis of scale. The majority
of the world’s seed banks dedicated to wild
species have seed holdings that are barely suf-
fi cient to provide seed for but a few percent
of the areas in need. The Millennium Seed
Bank, for example, stores, as of March 2011,
1.8 billion seeds of 30,402 species ( 15). This
translates on average to around 60,000 seeds
per species, well short of the amount required
to restore at the landscape-scale.
It is common for restoration programs
to tackle thousands, or tens of thousands, of
hectares, often in poorly studied ecosystems.
The Gondwana Link project ( 16) in south-
western Australia is the world’s largest inte-
grated wild species restoration program in a
biodiversity hotspot. It aims to repatriate with
local indigenous species many thousands of
hectares of former farmland to create a bio-
logical corridor spanning 1000 km.
Ambitious restoration programs require
large volumes of seed, in the order of tens
to hundreds of metric tons. For example,
through the United Nations Compensation
Commission, Kuwait proposes to revegetate
720 km2 of land damaged during the 1990–
91 Gulf War ( 17). Applying even a modest
seeding rate of 2 to 4 kg of seed per hectare
will require 140 to 280 tons of native seed.
At present, there is no seed bank in Kuwait
to address this need, outside of the kilogram-
scale seed storage facility at the Kuwait
Institute for Scientifi c Research ( 18).
Current levels of seed wastage compound
the problem. Typical establishment rates in
biodiverse restoration result in seed losses of
>90% ( 19) through substandard storage, lack
of seed pretreatments for “on-demand” dor-
mancy release, and lack of precision in deliv-
ering seeds to sites at the appropriate time and
into a suitable soil environment for seedling
establishment ( 12, 14).
Technology Shifts, Infrastructure Capacity
A key issue is whether seed banks have the
infrastructure and scientifi c and technologi-
cal capacity to deploy seeds wisely for land-
scape-scale restoration. At present, the stor-
age of seeds for restoration is undertaken
largely by end users (e.g., seed suppliers,
mining companies, NGOs, and community-
based groups). These facilities often have
limited access to knowledge and training,
and as a result, most seed banks are missing
the technological resources and capacity to
deliver landscape-scale restoration. Resto-
ration seed banks must adapt the scientifi c
principles of germplasm storage developed
by seed banks conserving wild species (e.g.,
strict control over seed moisture content and
storage temperature) that have proven effec-
tive vehicles for the protection of biodiver-
sity ( 5, 20, 21). Unlike current technologi-
cally advanced seed banks conserving germ-
plasm, restoration seed banks need to be stor-
Restoration Seed Banks—
A Matter of Scale
David J. Merritt 1, 2 * and Kingsley W. Dixon 1 ,2
Seed banks must shift from being
“stamp-collections” of species to collections
that can provide tons of seeds and the
expertise to improve restoration efforts.
• Genetic provenance
• Purity assessment
• Viability assessment
• Pre-storage conditioning
• Germination testing
• Short-term storage
• Long-term storage
• Germination testing
• Viability monitoring
• Dormancy release
• Germination enhancement
• Seed delivery
PHOTO CREDIT: SASHA RADOSAVLJEVIC/ISTOCKPHOTO.COM
*Author for correspondence. E-mail: david.merritt@bgpa.
1Kings Park and Botanic Garden, West Perth, Western Aus-
tralia 6005, Australia. 2School of Plant Biology, Faculty of
Natural and Agricultural Sciences, The University of Western
Australia, Crawley, Western Australia 6009, Australia.
Integrated seed curation and research functions of a restoration seedbank.
Published by AAAS
on April 29, 2011
www.sciencemag.org SCIENCE VOL 332 22 APRIL 2011
ing tens to hundreds of tons of seeds. Some
seed banks on this scale are emerging, such
as the Seed Warehouse of the Utah Division
of Wildlife Resources, with a storage capac-
ity of 340 tons ( 22).
For policy-makers, the science to deliver
effective seed for restoration does not rely
on an expectation of open-ended funding,
as solutions found for one plant species may
apply across many other species. For example,
the discovery of smoke-stimulated germina-
tion in 1990 ( 23) revolutionized propagation
for thousands of wild plant species ( 24).
Integrating Science with Practice
Although seeds are recognized as a viable
means for accelerating plant establishment,
rarely is the source, availability, or effective
management and use of the seeds considered
( 10, 13). Commonly encountered shortfalls
in seed knowledge and handling practices
that hamper restoration outcomes include a
lack of research data on the phenology of seed
development and maturation for most wild
species and the spatial and temporal varia-
tion for these factors ( 10, 25) that can lead
to inappropriate timing of seed collection; the
failure to document or understand the qual-
ity and viability of collected seeds at input
into the seed bank resulting in no knowledge
of the potential of the seed resource to pro-
duce plants ( 25); an inability to break seed
dormancy for many plant families prevent-
ing germination at the time seeds are sown
( 26); the use of poor storage procedures such
as seed stored in uncontrolled environments
where humidity and temperature fl uctuations
occur ( 26); and low seedling establishment
rates from broadcast seed (<10% of delivered
seeds establish) for biodiverse ecosystems
such as in Mediterranean environments ( 19,
25). Seed farming of wild species is a critical
area in need of broader recognition and rapid
development to address shortfalls in seed
supply to restoration seed banks.
Achieving effective landscape-scale resto-
ration will need seed banks to scale-up capac-
ity in a number of key areas: seed technology
for effective seed use and site delivery, large-
scale native seed farming enterprises to gen-
erate the seed needed to reduce the impact of
seed collection on wild sources, and genetic
analysis tools to ensure provenance issues are
addressed in seed farming. Critically, it is the
integration of these research areas that is nec-
essary to improve wild seed use in restoration
(see the chart).
Examples of research-driven improve-
ments in seed-use effi ciency for restoration
are not available. However, as an example,
land areas disturbed by mining in the biodi-
verse semiarid region of the Pilbara in West-
ern Australia exceed 20,000 ha. Current seed-
ing rates for restoration in this region are 5
to 7kg/ha ( 27), and an average seed price
is $749 ± 65 per kg, based on the available
seed prices from commercial suppliers for
88 dominant Pilbara plant species ( 28) [sup-
porting online material (SOM)]. If restora-
tion research reduced the number of seeds
required to achieve plant establishment tar-
gets by 30%, this research effort would repre-
sent a reduction in seed use of ~30 to 42 tons
and a saving of $20 to $34 million in seed
purchase costs (SOM).
A Role for Botanic Gardens
Botanic gardens have the potential to contrib-
ute to global restoration outcomes through
their infrastructure, the knowledge gener-
ated by their scientifi cally curated plant col-
lections, and their seed bank technological
capacity ( 29). With 2700 botanic gardens in
100 countries, including representation in
all of the world’s biodiversity hotspots ( 21),
botanic gardens have the geographical reach
and networking capability to retool for deliv-
ery of a global restoration capability ( 20,
30). Whereas botanic gardens promote their
seed bank collections as a viable restora-
tion resource ( 20), most fall well short of the
capacity for delivering large-scale restoration.
To move botanic garden seed banks from
single-species conservation to landscape-
scale restoration will require major refocus-
ing on collection strategies, seed utilization,
and funding. For example, Kew’s Millennium
Seed Bank Project achieved its seed collec-
tion target of 10% of the world’s fl ora in 2009
( 20, 30). But the effort is now refocusing to
expand its mandate from germplasm conser-
vation to restoration and is planning a global
seed-based initiative through the Breathing
Planet Program for botanic gardens and local
communities to begin developing and deliv-
ering in-country restoration capacity through
worldwide partnerships ( 31).
Globally, communities need knowledge
on using seeds and restoring landscapes. In
the last session of COP-10, a village cattle-
man from Sudan “talked about how his fi elds
are drying up, his cattle are dying, his family
is suffering, and he is witnessing a continu-
ing and accelerating loss of plants and ani-
mals that provide subsistence to his commu-
nity” ( 32). He pleaded for answers, technical
guidance, and support that he could take back
to his community so he and his family could
cope with the impending changes to his liveli-
hood ( 32). Delivering effective and timely res-
toration and technology sets that build healthy
environments and sustainable livelihoods is
therefore a global issue for wild seed banks—
and none too soon if we are to stem the tide
of extinction and environmental degradation.
References and Notes
1. C. Nelleman, E. Corcoran, Eds., Dead Planet, Living
Planet—Biodiversity and Ecosystem Restoration for
Sustainable Development: A Rapid Response Assessment.
(United Nations Environment Programme, GRID-Arendal,
2. K. Bowers, http://blog.biohabitats.com/2010/10/cop10-
3. S. Cunningham, ReWealth! (McGraw-Hill, New York, 2008).
4. D.-Z. Li, H. W. Pritchard, Trends Plant Sci. 14, 614
5. Secretariat of the Convention on Biological Diversity,
Report of the Tenth Meeting of the Conference of the
Parties to the Convention on Biological Diversity (CBD
Publication UNEP/CBD/COP/10/27, 2010); www.cbd.int/
6. R. D. Smith, G. Hawtin, in Seed Conservation: Turning
Science into Practice, R. D. Smith, J. B. Dickie, S. H.
Linington, H. W. Pritchard, R. J. Probert, Eds. (Royal
Botanic Gardens, Kew, UK, 2003), chap. 56.
7. D. Lamb, P. D. Erskine, J. A. Parrotta, Science 310, 1628
8. E. Higgs, Restor. Ecol. 13, 159 (2005).
9. P. Wehi, Ecol. Appl. 19, 267 (2009).
10. L. M. Broadhurst et al., Evol. Appl. 1, 587 (2008).
11. J. M. Koch, Restor. Ecol. 15, S26 (2007).
12. D. P. Rokich, K. W. Dixon, Aust. J. Bot. 55, 375 (2007).
13. K. N. Suding, K. L. Goss, G. R. Houseman, Trends Ecol.
Evol. 19, 46 (2004).
14. S. M. Wijdeven, M. E. Kuzee, Restor. Ecol. 8, 414 (2000).
15. Royal Botanic Gardens, Kew, Millennium Seed Bank
16. Gondwana Link, www.gondwanalink.org.
17. S. A. S. Omar, N. R. Bhat, A. Asem, Hdb. Env. Chem. 3,
18. Kuwait Institute for Scientifi c Research, www.kisr.edu.kw/.
19. M. I. Williams, G. E. Schuman, A. L. Hild, L. E. Vicklund,
Restor. Ecol. 10, 385 (2002).
20. P. Smith, J. Dickie, S. Linington, R. Probert, M. Way, Seed
Sci. Res. 21, 1 (2011).
21. Botanic Gardens Conservation International, www.bgci.org.
22. Utah Division of Wildlife Resources, Great Basin Research
Center, Seed Warehouse, http://wildlife.utah.gov/gbrc/
23. J. H. de Lange, C. Boucher, S. Afr. J. Bot. 59, 145 (1990).
24. K. W. Dixon et al., Acta Hortic. 813, 155 (2009) (ISHS).
25. W. Mortlock, Ecol. Manage. Restor. 1, 93 (2000).
26. D. J. Merritt, S. R. Turner, L. E. Commander, K. W. Dixon,
in Proceedings of the Fifth Australian Workshop on
Native Seed Biology, S. W. Adkins, P. J. Ainsley, S. M.
Bellairs, D. J. Coates, L. C. Bell, Eds. (Australian Center
for Mining Experimental Research, Brisbane, Australia,
2005), pp. 69–76.
27. BHP Billiton Iron Ore, Jimblebar-Wheelarra Hill Mine
Progressive Rehabilitation Management Plan (BHP Billiton
Iron Ore, Perth, Australia, 2006); www.bhpbilliton.com/
28. Kimseed International, www.kimseed.com.au.
29. K. A. Hardwick, Conserv. Biol. 25, 265 (2011).
30. P. R. Crane, S. D. Hopper, P. H. Raven, D. W. Stevenson,
Trends Plant Sci. 14, 575 (2009).
31. Royal Botanic Gardens, Kew, Breathing Planet Programme;
32. K. Bowers, Rhizome, http://blog.biohabitats.
Supporting Online Material
Published by AAAS
on April 29, 2011