WCMC handbooks on biodiversity information management

WCMC Handbooks on Biodiversity Information Management

A DARWIN
INITIATIVE PROJECT

WORLD CONSERVATION
MONITORING CENTRE

Te

Aiiw~

COMMONWEALTH
SECRETARIAT

Data
Custodianship
and Access

Volume 5

Digitized by the Internet Archive
in 2010 with funding from
UNEP-WCMC, Cambridge

http://www.archive.org/details/wcemchandbooksonbO5reyn

WCMC Handbooks on Biodiversity Information Management

Volume 5
Data Custodianship and Access

World Conservation Monitoring Centre

WORLD CONSERVATION
MONITORING CENTRE

Series Editor J.H. Reynolds

Commonwealth Secretariat 1998

The World Conservation Monitoring Centre, based in Cambridge, UK, is a joint venture between three partners in
the World Conservation Strategy and its successor Caring for the Earth: \UCN — The World Conservation Union,
UNEP — United Nations Environment Programme, and WWF — World Wide Fund for Nature. The Centre provides
information services on the conservation and sustainable use of species and ecosystems and supports others in the
development of their own information systems.

The United Kingdom’s Darwin Initiative for the Survival of Species, launched at the 1992 Earth Summit in Rio de
Janeiro, aims to support the Convention on Biological Diversity by drawing on Britain’s scientific, educational and
commercial strengths to assist in the conservation and sustainable use of the world’s biodiversity and natural habitats.
Key tenets of the Darwin Initiative include collaboration and cooperation with local people, capacity building,
distinctiveness and complementarity of project initiatives, poverty alleviation, and long-term sustainability. Through
training, awareness raising, and research on undervalued areas of biodiversity, Darwin support is particularly aimed
at strengthening links between Britain and those countries rich in biodiversity but poor in financial resources.

Under the auspices of its Environmental Training for Sustainable Development initiative, the Management and
Training Services Division of the Commonwealth Secretariat supports short- and long-term training, internships
and institution development for environmental policy makers, environmental ‘operatives’, and environmental
information professionals in the Commonwealth, in various areas of the environment including biodiversity and
gender. Funding support for training, institution development and publications under the aegis of the Management
and Training Services Division is provided by the Fund for Technical Co-operation (CFTC).

‘WORLD CONSERVATION

MONITORING CENTRE

Published by Commonwealth Secretariat

ISBN 0-85092-548-7

Copyright © 1998 World Conservation Monitoring Centre
Reproduction of this publication for educational or other non-commercial purposes is
authorised without prior permission from the copyright holders, provided the source is
acknowledged.
Reproduction for resale or other commercial purpose is prohibited without the prior written
permission of the copyright holders.
The views expressed in this book do not necessarily reflect those of WCMC or its
collaborators.
The designations of geographical entities in this report and the presentation of the material do
not imply the expression of any opinion whatsoever on the part of WCMC, the
Commonwealth Secretariat, the Darwin Initiative for the Survival of Species, or other
participating organisations concerning the legal status of any country, territory, or area, or of
its authorities, or concerning the delimitation of its frontiers or boundaries.

Citation World Conservation Monitoring Centre. 1998. WCMC Handbooks on Biodiversity Infor-
mation Management. Volume 5: Data Custodianship and Access. Reynolds, J.H. (Series
Editor). Commonwealth Secretariat, London. ix + 24pp.

Typeset by Bookcraft Ltd, Stroud, Gloucestershire, England

Cover design Michael Edwards

Photography by J.S. Donaldson (Wood’s cycad, Encephalartos woodii); D. & 1. Gordon (Mali landscape; Plant
study, Ghana; Thai forest; Rock hyrax, Procavia capensis); \UCN/J. McEachern (Diver and
fish); WCMC (Ecoregion and Africa maps; GIS work; Workshop facilitation).

Available from IUCN Publications Services Unit
219 Huntingdon Road, Cambridge, CB3 ODL, UK
Tel: +44 1223 277894; Fax: +44 1223 277175
Email: iucn-psu@wemc.org.uk
http://www.iucn.org/bookstore/

Contents

ACKNOWLEDGEMENTS hae ober, DOs SPR aot Coens eee Vv
BACKGROUND 3G: o* 4 iniiy: mouse, te foto of ae oe Vii
LRUINTRODUGTIONT 1 Ghcough. ihe, Jee OL pomleneinaie atic: 2 1
2RBASICPRINCIPEES:, 101) Asay .ANeRind: (pera, See 3
spn EUNCFIONSOF A'CUSTODIAN:«.. Sig 18 eae. eee 5
3c» Responsibilities and Rights 5). bos. 2s ae Sa eee eee: 5
52a Custodians and Qwiersie)>. 263 As, bene eRe, ce. SEC Oeces | MAIR 5
Sag Custodians/and Users; Mocept@iealh iQue Pree, seme eree OFF 7
ANY MANAGING CUSTODIANSHIP: .nc' ayes, Suporte oes he, roe 9
AsieAssioning:CuStodianship:.;).%9 21. he dey See, Ba, Cee. ie ee 9
Ash Managing Custodianship™\.. 5.°25)\.08 Ue ied 18 Caton, WED. pals 11
4:3) OReviewing Custodianship =. “0.0 SE Oe eee 12
4 Aoi Setting Priorities: ste. Gyo RYO, 10 POC Een, Mae: 13
5 CATALOGUES AND METADATABASES................. 14
6) (DATA ACCESS AGREEMENTS» 4) «2:4 HRIN2) Br bile) ese eee aie cere 15
GliM@verview outs O82 Oe QOglh, AOCUKE AE OTE 2 RE SRO B 15
62°" Roles of the Hub; Custodians:and Users) % «) 4186. 6. ARs Se 18
6.3! PCostRecaveryere OF 0. (RP 2UAL UR BAe BCU S. eer eee 19

7 CASE STUDY: COMMONWEALTH CUSTODIANSHIP
GUIDELINES; AUSTRALTAO Wess, CLS GL BR a PERSE, 21

SxUREFERENCES Oe, 210 Sz [rp See ROG, ce, OGY. OF VER Ce 24

7 \
te ee mir rn eg Shand ia Chuetirinign Sitbisce pois bone a “4

. ee ey ya pe, Sie vot, RE reanqay Coming for sae Enraks AC i Tie Wien ae: ,
AULA > heugett Maye ing: Pi 4 ah Pb nape, ond "ALY Wow Wisy Far) fre? Von ine, THEE: . ea
igallaapis: oo Op ms giana tne, OW SEA AE Hye awd cleo. wep - oe.

Ptisn fa me * wi seers sa inv a aie cs

Cone cusps ei, ues wba ne Wig Sen bea pl Spat, bc » (G6 Hoth Same

ry , cles gad hg aan By rm fee Aw Ly Pchpgrcn’ § SSE By ec ‘e cal, os
SacmbGaweanteeneeian Teer ween 2) Jenproinedsny ngsyl pre, Raae EE OE
“ieee. I, Moser e Selma Sateate NG pit Mra Rabat astneelt Mion pen iaaeey cnet ha
Gale” te aR itary ator gin fauter Pi, Pveriy CREASES Aes bale liye iataginnt i:
eae trslcke preety: .iaboiewar sve. ys dea tprdern st vir re sey, Pare ie pads
” ila aaa erane (ieee eater cadence Tel) Mh Dd Fiver 25 Mi 46

“ pos
“pee: ete entre ‘iG ae as se pipe tee Aas hhoyveet ie Thav@ age WT
en ai xh " nee y Ot ati Se Woden ee cag igi iy Des Se oe

mae | ‘anise vik “Btn Sang! es ihe Sa -gorey Wises » write) Ser

it he

ile : coe we ae **

Pid reg! te ae bale 0, com , ieaiilen nl aqeon . f
wie} a. ‘ Fiaoa a | eA, 4 pil ene ‘é * etbive bisa iidtbosaee

ad vis Oey, eOMR | ot ee a

oh & f ew ® ve ioe ’ fs , . san) SOR ERROR
r eee My e -
nee We i do serginy re 4 Mil emerge
rose ; * : » canbe
ee = yout. an * ; ea forth haberd al ong mga 4 er it fiery
e : ; les a ae dey of hand, San nue 2 - ; lene
Ce Wit. “A Wad ae Le ye iil = rh . er ae A

: :4 y ee eG dg
De 73 Paty i at ‘ation? caton Petnatin’ at
Se RB nih <) Sionderiaotin By oot

eee S| if be, - oy hie maanedell an. gyre: a nce See

“Det a se on, vaeaps saab

f ' ; bres EN). ; ' E ;
ms , % ; aes Sib hae 4 Bioreeis |
eu ASH set ericeag re, ine ; rate
ve Ps er . aero te Le eee a so hi
{ Che i wee
r)

et
; ath i Alt
a, Me aca Le (ee Da

ACKNOWLEDGEMENTS

The generous support of the United Kingdom’s Darwin Initiative for the Survival of
Species has provided for the development of a comprehensive programme of training
in biodiversity information management. This programme comprises an international
training team, drawing on expertise from collaborating organisations around the
world; the preparation of a training resource in the form of a handbook series and
related materials; and the development of computer-based demonstration tools.
Training is being promoted through the delivery of post-graduate modules, and
through regional and national workshops which have received additional support
from The British Council, British Airways Assisting Conservation Scheme, and
contributions from participating organisations. The programme has_ been
appropriately titled Darwin Initiative Training in Biodiversity Information
Management.

Development of the handbooks has also benefited from experiences gained
through the Biodiversity Data Management (BDM) Project, administered by the
United Nations Environment Programme (UNEP) and funded by the Global
Environment Facility (GEF), and related initiatives supported through the European
Union (EU) and European Environment Agency (EEA). Indeed, Volume 6 draws
extensively on one of the key outputs of the BDM Project, the Guide to National
Institutional Survey (UNEP/WCMC 1998), developed in consultation with partici-
pating countries, the BDM Advisory Committee and the UNEP management team.
The concept of an information cycle was developed in collaboration with the
International Institute for Environment and Development (ITED) with support from
the UK Department for International Development (DFID). The handbooks have
been published through the generous support of the Commonwealth Secretariat.

Fundamental to the development of this programme have been the partnerships
established with training organisations around the world. These organisations have
worked collaboratively in hosting workshops, in reviewing the handbook materials,
and in providing guidance on how regional and national training needs can be met
most effectively. The training programme has significantly benefited from the input
of numerous individuals working in the field of biodiversity information
management. Among these individuals, particular mention goes to Professor Ian
Crain and Gwynneth Martin of the Orbis Institute, Ottawa, Claire Appleby, an
independent consultant, and to Drs Jake Reynolds and John Busby of WCMC for
their insightful work in developing the handbook series. Thanks are also extended to
Laura Battlebury for her tireless administrative and logistical support. The series

Volume 5 Data Custodianship and Access v

editor for the handbooks was Jake Reynolds, while Donald Gordon managed the
overall project.

To the many individuals, both within and outside WCMC who have contributed to
the development of materials and the delivery of training in biodiversity information
management, a profound debt of gratitude is owed. It is through this collaborative
effort that a service is being developed to contribute to the conservation and
sustainable use of living resources.

vi WCMC Handbooks on Biodiversity Information Management

BACKGROUND

The purpose of the WCMC Handbooks on Biodiversity Information Management is
to support those making decisions on the conservation and sustainable use of living
resources. The handbooks form part of a comprehensive programme of training
materials designed to build information-management capacity, improve
decision-making and assist countries in meeting their obligations under Agenda 21
and the Convention on Biological Diversity.

The intended audience includes information professionals, policy-makers, and
senior managers in government, the private sector and wider society, all of whom
have a stake in the use or management of living resources. Although written to
address the specific need for improved management of biodiversity-related
information at the national level, the underlying principles apply to environmental
information in general, and to decision-making at all levels. The issues and concepts
presented may also be applied in the context of specific sectors, such as forestry,
agriculture and wildlife management.

The handbooks deal with a range of issues and processes relevant to the use of
information in decision-making, including the strengthening of organisations and
organisational linkages, data custodianship and management, and the development of
infrastructure to support data and information exchange. Experience suggests that
some of the greatest challenges in information management today are concerned with
organisational issues, rather than technical concerns in the delivery of information
which supports informed decision-making. Consequently, topics are addressed at
management and strategic levels, rather than from a technical or methodological
standpoint, and alternative approaches are suggested from which a selection or
adaptation can be made which best suits local conditions. Nevertheless, in adopting
this framework approach, we have tried to adhere to recognised conventions and
formalisms used in information management and trust that in producing a ‘readable’
set of handbooks the integrity of the materials has not been compromised.

Volume 5 Data Custodianship and Access vii

Overall, the handbook series comprises:

Companion Volume

Volume 1 Information and Policy

Volume 2 Information Needs Analysis
Volume 3 Information Product Design
Volume 4 Information Networks

Volume 5 Data Custodianship and Access
Volume 6 Information Management Capacity
Volume 7 Data Management Fundamentals

Collectively, the handbook series promotes a shift from tactically based
information systems, aimed at delivering products for individual project initiatives,
to strategic systems which promote the building of capacity within organisations and
networks. This approach not only encourages data to be managed more effectively
within organisations, but also encourages data to be shared amongst organisations for
the development of the integrated products and services needed to address complex
and far-reaching environmental issues.

The handbook series can be used in a number of ways. Individual handbooks can
be used to guide managers on specific aspects of information management; they can
be used collectively as a reference source for strategic planning and project
development; they can also provide the basis for a series of short courses and training
seminars on key challenges in information management.

The companion volume provides the background to the handbook series. It also
assists readers in deciding which handbooks are most relevant to their own priorities
for strengthening capacity.

A second series of handbooks is planned to provide more detailed guidance on
information management methodologies, including the areas of data and technology
standards, database design and development, application of geographic information
systems (GIS), catalogues and metadatabases, and the development of decision-
support systems. The current series deals only briefly with formal system
development methodologies, and for more detailed treatments the reader is
encouraged to access the wide range of published and electronic resources available
in libraries and on the Internet, some of which are alluded to in individual handbooks
and reference sections.

Vili WCMC Hanabooks on Biodiversity Information Management

A number of computer-based training tools have been developed to accompany the
handbook series and are used in the training programme. These are based on a
protected areas database, a tree conservation database, a GIS demonstration tool and
a metadata directory. They aim to demonstrate key aspects in the collection,
management and analysis of biodiversity data, and the subsequent production and
delivery of information. They also illustrate practical issues such as data standards,
data quality-assurance, data access, and documentation. Each training tool is
supported by a user guide, together with a descriptive manual which traces the
evolution of the tool from design, through development to use.

Volume 5 Data Custodianship and Access ix

sere, pieiedelamayiace a
eemeate yr dated uated: OF aie ¢odT evoronib wre a
tin rota eagle e bee.gieb wizariboid To rings bao twomegammtt
. uae Sage ‘Satie ohn i odie yodT: -noitareattal to gail é

‘a 100K gga ak: aeeneeyy stot Sonswrcd- leap, ce
> ont piaebatgiatey Lamas yriraienaets 4 Ahir orrdinact ching wen» yd bonegge

Sait esis nearest sian =

a ‘ier Plnaehy aie tf Pookhyacrtnis ae: ee
- \ ae ‘i Pa i ue =

si)!

hte A hiiiante: matieh openers a oetatt fro tatically’ aged

ai arb ie’ AMS:, ‘site tee ie? Pee IC a wilan) protect ae = 4
M4 tyr an pale ae th Dae ee thea baci dime: si cagaaly within org nike > eo
nme erat Pha yap tase cedly: ap aera as : dukes Weal Whe ofttotitely ne
meni ca ES Whe eit ages tata er he Ghared arqinias! ofp nicht mt,

We Arete a oF dhe amore prcnhuile tind evi vader 4 ockcieey,- Otay ia
pase hip si fi tid ah ints Lagggnd Fi "hen ie

fe ES Hatnaaenetied SMe ae “at genie of weve th sore Sb sh
ee + . judlie veri Ninionsie wh secur rey WN: at Savi: shay mabage Piet they Gee
i aie Ne a go taininees ame fie mindless Manpinty sid =
ita eee dey ah eles ide thee tera fey aemipcha th 1s vorrei a
a iki ivi I ek atic silicic a lode
few ee tions verborve Fe as tie ‘ech patetvo Sar Neseleoods a:
wee rales ih nstrPagy ws reer xicer ties rere re hr yfinw. « >it eu me :
¥ jie ahaa deh: eA otha OP i i eee, ee a

cs , Te) ee,
ese ee sansa einen dsonis
a a ate 14 IN we delta abet wise ith oo ee
; wr! 9 pe Min, Oe ym ant & ecient og kovierape sate
ie ao sabilorgses tick yapteditabiowen, snd the yice ebopiagen gall
peat iistota, “Pee courant: inonied oath ainly eietty width
sel aineae ca ogi, ct
Oa ca BA aaa ss wt ggONO pablo at sora
gibt la! ean Vache m4 “

1 INTRODUCTION

Many organisations are beginning to realise that knowledge is one of their greatest
assets. They are also discovering that the costs of maintaining this asset, and
exploiting the opportunities it creates, are significant. Given these costs, which may
be significant, organisations are looking to gain maximum value from their data
holdings. Increasingly, they are using data for multiple purposes and are recognising
the need to develop documentation and ensure compliance with established
standards. They are also beginning to understand the benefits of sharing data and
collaborating with others within information networks (see Volume 4).

The development of policy-relevant information on environmental concerns often
requires access to a wide variety of data sources, from numerous organisations and
disciplines (see Volume 3). If the process of information production is to be efficient
and cost-effective, such sources need to be readily accessible, as should the people
and tools necessary to convert them into information for decision-making.

Preferred sources of data are those organisations (occasionally individuals) which
are in the best position to ensure the quality and accessibility of their datasets,
and to advise on appropriate uses. These are referred to as custodians.’ If
custodianship is not assigned and managed carefully, then users may face a
bewildering set of incomplete and incompatible datasets, with inadequate
documentation and poorly defined and inconsistent access procedures. As a result,
they may find it virtually impossible, within the limited time available, to integrate
data into information that will usefully support decision-making.

Custodianship is the means by which responsibility for the management of a
dataset (or part thereof) is assigned to and accepted by the most appropriate
organisation. Its principal aims are as follows:

e@ To minimise duplication of effort.

@ To ensure that data are available for use (i.e. they exist and are accessible).

1 Although most references in this handbook concern the custodianship of data, the term applies also
to tools, applications and other technologies which transform data into information, or communicate
that information to users (see Janzen, 1993).

Volume 5 Data Custodianship and Access 1

e To ensure that data are quality-assured (i.e. they are valid, maintained,
documented and secure).

Custodianship provides a mechanism to ensure that important datasets exist, are
maintained and are accessible to legitimate users. It ensures accountability for and
reliability of datasets within a specific jurisdiction (e.g. a sector, discipline or
theme), thus ensuring that information products used by governments and other
decision-makers are accurate, complete, identifiable and auditable. In summary,
custodianship is the core of an efficient, responsive information infrastructure,
capable of serving the interests of individual organisations or networks.

WCMC Handbooks on Biodiversity Information Management

2 BASIC PRINCIPLES

As with other complex issues relating to the management of scientific information,
effective progress requires the underlying principles of data custodianship to be
understood and widely accepted. These principles, which are presented below, can
then guide operational developments:

e@ Data should be managed by the organisation in the best position to do so.

@ Data should be managed cost-effectively by staff who understand the data: what it
purports to represent, what its characteristics are, how it was collected, what
quality-assurance procedures have been applied and its limitations.

e@ Data should not be duplicated or fragmented in different places.

@ Data should be widely available to those that have a need to make better informed
decisions.

The above principles are not meant to be interpreted as rigid rules. There may be
sound operational reasons why, for example, a copy of a dataset may be duplicated in
another place, such as in a network hub (see Volume 4) for purposes of distribution.
This may be done because the custodian is not accessible on-line, which could
impede access to the original dataset. However, the risk (and cost) of violating
custodianship principles needs to be understood. In the above example, the risk is that
the distributable copy will become out of date, while the cost implications are that the
hub will have to monitor developments in the original dataset and periodically obtain
an updated copy.

The key is to manage data in such a way that they can be converted into a
variety of information products, for a variety of users, thus ensuring that they are
flexible enough to respond to the demands of decision-making. Ideally, every
dataset has a recognised custodian although, if priorities need to be made, the
custodians of essential datasets should be determined first. These datasets are
sufficiently vital to the day-to-day operation of an organisation or network that
they justify the effort and expense incurred in their collection, storage and
quality-assurance (see Volume 3).

Volume 5 Data Custodianship and Access 3

Essential datasets underpin the development of multiple information products, for
multiple users, and are, thus, permanent, or at least of a lasting nature. Non-essential
datasets, on the other hand, are produced by ad hoc, undocumented processes for
quick results, and may be transient or of uncertain quality (after Janzen 1993, 1995).
Unfortunately, many organisations manage essential datasets as if they were
non-essential datasets.

4 WCMC Handbooks on Biodiversity Information Management

3 FUNCTIONS OF A CUSTODIAN

3.1 Responsibilities and Rights

As outlined in Volume 4, custodianship of a dataset carries with it certain
responsibilities (see Box 1). Each of these contributes to the well-being of a dataset
and, thus, to internal productivity gains within the organisation as well as an
increased capacity to collaborate with others. Naturally, custodians may harbour
concerns at the prospect of providing access to their data. For this reason they are also
invested with certain rights, consistent with broader government, corporate and other
applicable policies and agreements, which determine the conditions under which the
dataset can be used (see Box 1). Such rights are not intended to prevent legitimate use
of a dataset. Indeed, the aim is to foster an environment in which data access is
straightforward and encouraged.

Where a custodian does not have sufficient resources to undertake all of its
responsibilities, certain of these may be delegated or contracted to other
organisations, known as stewards (Janzen 1993). Custodianship, however, remains
with those responsible for the content of the dataset.

3.2 Custodians and Owners

Copyright does not protect facts, so it is not clear whether ‘data ownership’ has any
legal status. However, the concept is useful when describing those individuals or
organisations that have some claimed intellectual property rights (whether
enforceable or not) over certain data. In the majority of cases these ‘rights’ are
claimed on the grounds of original collection of the data.

Although, in most cases, data owners are also the custodians, it is important to
recognise that data custodianship differs from data ownership. Custodianship
does not necessarily signify ownership, although the distinction is important only
when the data custodian is a different entity from the owner. Box 2 presents a variety
of situations in which this occurs.

Volume 5 Data Custodianship and Access 5

Box 1 Responsibilities and rights of custodians

Responsibilities
© To build a dataset (with partners, as appropriate).

e@ To maintain a dataset (i.e. keep it up to date, abreast of standards, structured
as necessary).

To ensure the quality of a dataset (i.e. ensure that it is valid, maintained,
documented, secure).

To provide access to a dataset (to legitimate users).

To provide advice on appropriate uses of a dataset (e.g. suggested/unwise/
improper uses).

e@ To coordinate the development of a dataset (with appropriate partners).
Rights
@ To regulate access to a dataset (depending on category of user).

e To safeguard intellectual property (e.g. acknowledgement, regulation of
copying).

@ To recover costs (e.g. recover market value, investment, cost of supply).

The owner retains intellectual property rights over the data, although these rights
can, in practice, be virtually non-existent, as is the case with public-domain data. The
owner may choose to delegate some of these rights to a custodian. The custodian may
be likened to a trustee in terms of its relationship with the data. The degree of freedom
that the custodian has to either use or distribute the data depends on how rights are
delegated by the owner. In general, there will be a formal agreement between the
owner and the custodian which specifies what the custodian is allowed to do with the
data and the circumstances in which the owner needs to be consulted.

6 WCMC Handbooks on Biodiversity Information Management

Box 2 Where data ownership and custodianship differ

e In most government jurisdictions, data may be ‘owned’ by the supreme
executive authority, whereas the custodian may be just one of many agencies
created by that authority, designated to act on its behalf.

Data may be in the public domain but managed by a custodian organisation
acting, at least to some extent, ‘in the public interest’.

The owner of a dataset may be one or more persons, for example scientists or
naturalists, who originally collected the data, and who maintain ownership
tights, but who may be unable or unwilling to manage the data effectively
over the long term.

The owner may be an organisation that built a dataset for a particular purpose
but has since lost interest in managing the data. It may then delegate its
custodial responsibilities to another organisation, which becomes the
custodian. The original owner may still retain some residual intellectual
property rights over the data, which the new custodian would be obliged to
respect.

In some cases, the custodian may choose to contract some of their responsibilities
to another organisation, perhaps one of its partners in a network. Provided the
contracted organisation has no authority to use or distribute the data without approval
by the custodian, this arrangement would not ordinarily affect the custodian’s rights
or responsibilities over the data.

3.3 Custodians and Users

Properly organised custodianship is beneficial to users of both data and information.
For example, confusion over where to obtain accurate data is minimised, and reliable
advice on the source, currency and completeness of information products is
forthcoming. In return, users should assist custodians by providing feedback on
the usefulness of data, and by keeping them informed of their future requirements
(e.g. quality-assurance requirements). This helps the custodians plan their data
collection and management strategies.

Volume 5 Data Custodianship and Access 7

Where a user collects data on behalf of a custodian, with the intention of
submitting it for entry into a dataset, this should be done according to the standards
and procedures established by the custodian. Users should also return any data that
they have corrected or otherwise upgraded and, in turn, the custodian needs to ensure
that the upgraded data is made available to subsequent users.

& WCMC Handbooks on Biodiversity Information Management

4 MANAGING CUSTODIANSHIP

4.1 Assigning Custodianship

Custodianship of a dataset is normally accepted by the organisation most familiar
with its history, special management requirements and potential uses. Within any
particular network, such organisations may be obvious to the network’s partners,
allowing custodianship to be confirmed, rather than negotiated, without issue.
However, where several organisations claim custodianship of the same dataset, or no
custodian is apparent, the network hub, through its steering committee, may decide to
commission a review. This could be applied to specific datasets as the need arises or,
more ambitiously, it could be extended into a network-wide review (see Volume 6).

It is often the case that environmental datasets are significant to a wide range of
stakeholders, not just their custodians. This suggests that greater, perhaps national
needs should prevail over individual feelings of data ownership, particularly in the
case of essential datasets which may be depended upon for projects of national
importance. Difficult decisions may have to be made in the short term to guarantee
the quality and accessibility ofa dataset in the long term. As with all decisions of this
nature, it is imperative that they are arrived at transparently and with the full
participation of leading stakeholders. Box 3 lists a variety of criteria which the
steering committee of the hub could use to determine which organisation is the most
appropriate custodian for a particular dataset (note that the criteria are not equally
important).

One way to apply the criteria in Box 3 is to select those organisations thought to be
most relevant in the particular context, and assign numeric values to each criteria
according to its relative importance. Thus, statutory responsibility may be perceived
as being the most influential factor determining custodianship in a particular
network, whereas best financial position might be considered to be the least. This
allows the steering committee to ‘score’ potential custodians according to their
suitability for the role, as illustrated in Table 1 (custodians denoted by the letter C).
Analyses of this kind do not provide sufficient grounds for assigning custodianship in
their own right, but may serve to focus discussion.

Volume 5 Data Custodianship and Access 9

Box3 Potential criteria for determining custodianship

e Statutory responsibility for management of a dataset (beware of overlapping
mandates!).

Greatest operational need for a dataset (e.g. for decision-making).
Normally first to record changes to a data item.

Requires the highest integrity of a data item (e.g. military precision).
Most ‘competent’ to manage a dataset.

Best financial position to manage a dataset (beware short-term effects).
Most technical or physical resources to manage a dataset.

Confidence of users in continuing to manage and develop a dataset (e.g.
committed, no ‘conflict of interest’).

When deciding the custodianship of essential datasets, the overriding principle is
that each dataset should have one and only one custodian. This is a practical and
effective way of ensuring that management responsibility is assigned to every dataset
which is valuable to multiple organisations and users. However, some environmental
datasets are not easily packaged under a single label, and overlaps in organisational
jurisdiction will occur. This can be resolved by designating one organisation as the
overall custodian and encouraging others to maintain specific sub-components. An
example would be a protected areas agency which manages a dataset containing,
amongst other entities, data on the distribution and significance of species within its
estate. Whilst it is justified in managing this dataset, the list of names used to
reference the species would: be managed by a more specialist custodian, such as a
national museum or herbarium.

10 WCMC Handbooks on Biodiversity Information Management

Table 1 Determining custodianship

Criteria Relative
importance

Statutory responsibility

Greatest operational need

First to record changes

Most ‘competent’

Best financial position

Confidence of users

4.2 Managing Custodianship

Responsibility for data may need to be assigned at several levels. At the national
level, responsibility for data themes may be assigned to separate lead organisations,
such as government departments or research establishments. Land infrastructure, for
example, including administrative boundaries, topography, settlements, roads and
rivers, might be assigned to a national department of survey and mapping, which
other national-level organisations see as the natural custodian.

At the sub-national level, land infrastructure data may be managed at a higher
resolution by local authorities, and be dispatched upwards to maintain the survey and
mapping department’s datasets. Thus, in reality, the survey and mapping department
is the hub of a land infrastructure network, with responsibility for data management
devolved to a series of sub-national custodians (see Volume 4). As such, data
harmonisation — the ability to integrate the various sub-national datasets — needs to
be resolved, at the outset, by agreeing appropriate data standards and protocols.

Network hubs should ensure that they provide sufficient guidance and

coordination to custodians to enable them to contribute effectively to the network’s
objectives (see Volume 4). Some obvious examples of where guidance may be

Volume 5. Data Custodianship and Access Py |

provided are the agreement of standards for data collection, storage and
quality-assurance, and consistent procedures for data access (see Volume 7).

A further complication is that, while a theme may have been assigned to a
particular organisation, other organisations may need to develop datasets within that
theme to meet their particular objectives, and for which they would become the
custodian. For example, the theme vegetation may be allocated to a natural resource
management agency, yet the defence department may need to develop a subsidiary
vegetation dataset, with attributes that show which areas heavy vehicles may traverse
and which they should not because of unacceptable surface damage.

When a dataset is modified, for instance by enhancement or integration with other
data, then the responsibilities and rights of its original custodian become diluted.
Some management process should then define a clear point at which the original
intellectual property rights and custodial responsibilities will be deemed abandoned.
Essentially, it is a matter for negotiation between the parties concerned as to how to
balance the requirements of the organisation that created the derived dataset with the
legitimate interests of the original custodian. For instance, a decision must be made as
to whether the new dataset is repatriated to the original custodian or whether it should
be retained by the organisation that created it (presuming that they are capable of
performing their custodial functions) (CSDC 1995).

4.3 Reviewing Custodianship

In becoming a custodian, an organisation needs to consider its data management
responsibilities and ensure that it is able to meet them. If an organisation cannot meet
its obligations, then it may consider relinquishing custodianship to another
organisation. As with the initial assignment of custodianship, reassignment requires
active but sensitive management, especially with datasets which have been identified
as being fundamental to many organisations and users.

12 WCMC Handbooks on Biodiversity Information Management

It is a good idea to undertake periodic reviews, perhaps every two years, of:
@ the appropriateness of current custodians;
@ progress in building, managing and enhancing key datasets;
@ dataset documentation, including catalogues, directories and metadatabases; and

@ implementation of data standards (see Volume 7).

Network hubs could consider developing a series of performance indicators for
custodians, so that reviews can monitor progress against established benchmarks. In
turn, custodians could report progress to their network hubs on a regular (e.g. annual)
basis.

4.4 Setting Priorities

Resource constraints invariably mean that strict priorities for data development
are needed. For this reason, it is important to identify appropriate custodians for
essential datasets, and support these where specific investments in institutional
capacity are required (see Volume 6). Every country, for example, needs an accurate
and stable dataset representing its national boundary. This dataset must be
maintained at a resolution and accuracy suitable for all major organisations and
programmes, which may require the boundary in a range of scales and projections.
One solution is to assign a single organisation, normally a national mapping agency,
custodianship for the entirety of this dataset in its various forms.

Custodians generally build datasets for their own corporate objectives, rather than
for the wider benefit of the networks in which they operate. Thus, ideally, when
building datasets, organisations should take into account the needs of their fellow
partners, in order to increase the range of purposes to which the datasets can
eventually be applied. Where partners require datasets that are of finer resolution,
more elaborate or, in general, are of a higher quality than that required for the
custodian’s own purposes, then agreement needs to be reached on how to cover any
additional costs which may be involved.

Volume 5 Data Custodianship and Access 13

5 CATALOGUES AND METADATABASES

Custodians are normally expected to document their datasets and to provide summary
descriptions to potential users as required. It is generally considered to be good
professional practice to document datasets, and custodians should aim to undertake
this as a matter of routine (see Volume 7). Typical features to document include the
theme, scale, completeness, currency, reliability, precision and pricing strategy of the
dataset, plus details of how it was collected, its intended purposes, and the data
standards and quality-assurance procedures which have been applied.

Within an information network, custodians may be encouraged to submit details to
the hub on any datasets which are, at least potentially, available for use by other
organisations. The catalogues which result — known as metadatabases in their
computerised form — may be published and disseminated widely to assist users in
locating the data and information they require. Due to the potentially large number of
datasets available within a network, catalogues usually contain only a summary of the
dataset’s purpose and quality, plus the contact details of the custodian and advice on
access procedures, including any costs involved.

14 WCMC Handbooks on Biodiversity Information Management

6 DATA ACCESS AGREEMENTS

6.1 Overview

Issues relating to data access are some of the most challenging, yet important, for an
information network to manage. The challenge is rooted in the legitimate concerns of
both custodians and users, as summarised in Box 4. These concerns are held by
individuals, project groups, organisations and governments, and cover a wide range
of issues. Some of these are genuinely important,” but others are largely spurious or
secondary to other issues, for example mistrust of user intentions or lack of
understanding of what the user requires.

Unless fully addressed, such concerns have the potential to hinder data access and,
hence, reduce interest in cooperation. One way forward is to accompany transactions
with formal agreements between parties, offering tangible assurances that the
concerns of both sides will be met. A key principle is that data should be made
accessible through the custodian, or their nominated distribution outlet, not from a
secondary source.

Data access agreements need not be viewed as a defensive measure intended to
limit access to data. Rather, they are a positive means of increasing trust and fair
dealing amongst the network’s or other partners. As confidence grows and data begin
to be mobilised more easily, the need for formal agreements diminishes until,
ultimately, they may no longer be required.

Data access agreements are perceived to be difficult to negotiate, but this need not
be the case. A useful step is for the network hub to draft a generic agreement for
distribution and adaptation by custodians. In the interests of simplicity, agreements
should assume goodwill on the part of users, not bad faith. Similarly, administrative
and cost impediments to data access should be kept to an absolute minimum.

2 Custodians need to ensure that data which are genuinely sensitive for reasons of privacy,
confidentiality or security, are adequately protected. An example would be detailed descriptions of
the locations of threatened species that are at risk from exploitation.

Volume 5 Data Custodianship and Access 15

Custodians are responsible for advising users on the potential uses of a dataset.
They may also recommend specific permitted uses or, conversely, excluded uses of a
dataset. This may occur when, for example, the custodian believes the data are
unsuitable for certain purposes due to uncertainties or ambiguities in the dataset. It
also provides a means for custodians to safeguard their intellectual property. Varying
conditions may be applied to different classes of user (e.g. government, NGO,
research, commercial). Typical provisions to consider when drafting a data access
agreement are presented in Box 5.

Box 4 Typical concerns over data access
Custodians
Will the dataset be misused?
Will intellectual property be respected?
Will the cost of supply be recovered?
Who will be liable in the event of a problem?
Will professional credibility suffer from the release of the dataset?
Could confidentiality be breached by the release of the dataset?
Does the transaction comply with internal network guidelines?
Users

@ Is the dataset fit for its intended use?

Will it be available at the right time?

Will onerous conditions be imposed on its use?
Is it available in a form which can be easily handled?

What, if anything, will it cost?

16 WCMC Handbooks on Biodiversity Information Management

In the event that incorrect data are provided by a custodian, or they are used for an
inappropriate purpose, liability could fall on the owner of the data, the custodian, a
third party which has provided the data, or all of these. The situation is most serious
when ‘negligence’ is detected, for instance when it is established that data were
poorly maintained or falsely documented. Exposure to liability is an emerging issue.
The most likely grounds for liability are where dataset documentation is misleading,
deceptive or negligent, or where there has been a violation of the provisions of a data
access agreement (see Onsrud 1989 for a discussion).

Box 5 Typical provisions in a data access agreement

e@ Whether data are available for single or multiple uses, and what those uses .
are.

Whether data can be retained by the user following its designated use
(risking independent updating or duplication), or whether they should be
destroyed (incurring subsequent costs when the data are needed again).

Whether the data are available for non-profit or profit-generating use and, if
the latter, whether any royalties would flow back to the custodian.

Whether users may give the data to third parties (the latter would normally
be referred back to the custodian unless prior consent was arranged).

Whether copies of any publications, products and other outputs derived
(even in part) from the data should be sent to the custodian.

What form of acknowledgement the user should employ (in cases where data
have been interpreted several times before use, a long list of acknow-
ledgements — known as an ‘audit trail’ — may be necessary).

Some form of disclaimer that protects the custodian from legal liability in
the event that their data prove to be unreliable or are used for an
inappropriate purpose (legal advice should be obtained).

Volume 5 Data Custodianship and Access 17

6.2 Roles of the Hub, Custodians and Users

The primary role of the hub is to enable the network’s partners to realise the benefits
of cooperation. In terms of data access, these benefits comprise value to the user
through access to essential data; value to the custodian for being of service (paving
the way for future exchanges and access to value-added products); and value to the
network for having enabled the transaction.

The hub can help to realise these benefits by facilitating the negotiation and
management of data access agreements. This could be achieved through the
preparation of a generic agreement embodying the principles of cooperation it wishes
to invoke, i.e. that environmental data are available to all stakeholders, for
non-commercial purposes, at minimum cost and with minimum administrative or
other impediments to access and use, whilst protecting the legitimate interests of
custodians.

In cases where access to an important dataset is not being provided satisfactorily,
the hub may wish to scrutinise the restrictions imposed by the custodian and
recommend new operating procedures. Experience has shown that many restrictions
imposed by custodians are unnecessary, even arbitrary, and can impose severe
constraints on data access.

The hub’s advice will need to be consistent with government policies relating to
information access and exchange. Where current policies are inconsistent with the
principle of wide information availability, with environmental decision-making
likely to be less effective as a consequence, network hubs have a role in advising the
government accordingly.

Custodians are responsible for developing policies on data access which are
consistent with the broader policy frameworks in which they operate. They need to
ensure that the provisions in their data access agreements are necessary to protect
their legitimate interests, and do not have the unintended consequence of inhibiting
reasonable use of the data by other users.

Users are obliged to comply with conditions prescribed by custodians. This is
crucial if the spirit of trust, so necessary to the successful operation of a network, is to
be built and maintained. Users may also provide feedback to custodians in the form of
advice of any errors or deficiencies encountered in the data, and an indication of their
future requirements, enabling custodians to continually improve their service.

18 WCMC Handbooks on Biodiversity Information Management

Finally, there are two responsibilities which users would ordinarily uphold even if
not specified by agreement. Where practicable, copies of any information products
derived fully or in part from a dataset are provided to the custodian; and the source(s)
of the data used should be fulsomely acknowledged.

6.3 Cost Recovery

Very few organisations have a corporate objective to collect data merely to sell data.
Data-rich organisations are generally in the business of selling the value they add to
data. In other words, the collection and management of data is a means to a larger
corporate end. As such, the cost of building and managing data is a cost of doing
business or, more simply, a business overhead. Certainly these costs need to be
recovered by selling products and services to clients, but not necessarily by charging
total costs to external data users.

When raw data are provided to a user, without any significant added value, then the
question arises as to what proportion of the data management overhead to charge to
that user. Normally, there is no argument that the marginal costs, to the custodian, of
providing those data can be charged to the user; the issue is what is a reasonable
price for data that already exist (particularly where collected at public expense), but
which were not created for the user or in the expectation that the user would require
them.

Unfortunately, the development of information networks is sometimes inhibited
by market-driven ideologies which authorise the cost recovery in data transactions.
The consequence of this is that organisations which could contribute to the resolution
of important environmental concerns are unable to do so because they lack the
resources to purchase the relevant data. The resources available to environmental
organisations, particularly NGOs, seldom reflect the significance to the community
of the issues at stake. This is largely due to a failure of the market to internalise
environmental costs and benefits.

Within a network, cost barriers should be kept as low as practicable to facilitate
data exchange. However, depending on the financial positions of the organisation
concerned, and the types of user who require access to the data, different strategies for
cost recovery may be applied. Not all of these are monetary-based, as illustrated in
Box 6.

Volume 5 Data Custodianship and Access 19

One way of facilitating data access in hostile policy environments (e.g.
over-bureaucratic or market-driven) is to develop datasets in partnership with other
organisations, or to barter data and other services (see Volume 4). There is no reason
why such arrangements should not include commercial partners.

Box 6 Strategies for cost recovery

Supply data totally free.
Supply data free to reciprocating users (i.e. barter).

Recover immediate cost of supply (i.e. time, consumables, energy etc. to
process the request).

Recover incremental costs (i.e. of managing the data and combating
depreciation).

Recover development costs (i.e. the initial investment in building the
dataset).

Recover “market value’ (i.e. the maximum cost which the user will pay).

20 WCMC Handbooks on Biodiversity Information Management

7 CASE STUDY: COMMONWEALTH
CUSTODIANSHIP GUIDELINES, AUSTRALIA

Australia has formed the Commonwealth Spatial Data Committee (CSDC) as the
peak coordinating body (network hub) for spatial data management at the national
level. It consists of senior governmental representatives. One of its tasks has been to
draft the Commonwealth Custodianship Guidelines setting out the Rights and
Responsibilities of Spatial Data Custodians and Lead Agencies (http://www.auslig.
gov.au/pipc/csdc/csdceguid.htm), the executive summary of which notes:

“The principle of custodianship should be applied . . . as a matter of good
practice, convention and convenience. It is simply the only way that data can be
managed in an orderly fashion. . .

The identification of custodians and the effective operation of custodianship are
necessary for the successful management of government spatial data...

To help improve [the management of . . . spatial data] the Committee will
implement a system of lead agencies and custodians of Commonwealth spatial
data, and outline their rights and responsibilities . . .

These guidelines are designed to assist lead agencies and data custodians in
developing improved practice in spatial data management .. .

A lead agency will be responsible for government-wide coordination of a data
class or category. This includes development of standards, and coordination of
data acquisition so as to avoid duplication. The CSDC will provide policy
guidance and oversee the lead agency system. Lead agencies will be identified
for broad categories of data after negotiation within CSDC...

CSDC will maintain a register of spatial data custodians. The Register will
define the custodian agency or agencies and data category, data items,
geographic coverage and any other feature necessary to define the dataset ...”

The Guidelines discuss the distinction between lead agencies and custodians, noting
that the responsibilities of a lead agency are primarily those of policy-development
and coordination, and detailing a set of specific objectives for cooperation. Clearly, in
this example, lead agencies correspond to network hubs for the thematic areas in

Volume 5 Data Custodianship and Access 21

which they have expertise. The CSDC itself corresponds to an over-arching hub
seeking to coordinate the efforts of lead agencies.

The custodian of a dataset is described as ‘the nominated body, or person
responsible for the development and/or the management of that dataset, and who has
the right to determine the conditions on which those data may be used or released’.
The Guidelines stress that:

“All spatial data collected .. . forms part of the . . . corporate spatial data
resource. Individual agencies involved in the management of the spatial data
act as custodians on behalf of the Commonwealth. The Commonwealth
Government delegates operational responsibility for the data, and for
implementing Commonwealth policy with regard to it, but retains the ultimate
responsibility for the data.

Custodianship is not necessarily synonymous with ownership of either the data
nor the copyright of that data.”

Among the benefits of custodianship, the Guidelines note that improved coordination
assists in the avoidance of duplication of effort and the unnecessary costs which this
causes, thus achieving greater benefits from the investments in data development
which are made. The assignment of custodianship is recognised as being a complex
process, requiring formal negotiation and documentation of the outcome. A list of
essential datasets, together with suggested lead agencies and possible custodians is
located at http://www..auslig.gov.au/pipc/csdc/csdccust.htm.

Lead agencies are highlighted as a mechanism to facilitate higher-level
coordination of data themes, including developing broad initiatives, future projects
and agreement of standards. Each identified lead agency is responsible for
disseminating information about its category of data, particularly to potential users,
and for promoting access to the data. This includes seeking to remove impediments to
and improving the efficiency of data transactions.

A major role of the lead agency is to avoid duplication of effort in data collection,
by providing avenues for communication between the network’s partners and
mechanisms to identify and coordinate data collection activities. However, the lead
agency is not empowered to determine priorities for data acquisition; that is the right
of the custodians.

22 WCMC Handbooks on Biodiversity Information Management

The Guidelines further note that standards for recording and managing data need to
be considered, both by lead agencies and custodians. In particular, standard methods
for selecting features and attributes, for assigning meanings and values, for ensuring
data quality, and for regulating data transfer are noted. Indeed, it is recommended that
custodians seek compliance with established quality management standards in order
to ensure the quality of their datasets.

In this example from Australia, custodians are entitled to levy a charge for access
to their data. The amount charged depends on several factors, including Government
charging directives, the costs of collecting, storing and distributing the data, market
rates, the intended uses of the data, and the conditions of any pre-negotiated
arrangements.

More information about the Commonwealth Spatial Data Committee (CSDC) can
be obtained from http://www.auslig.gov.au/pipc/csdc/csdcmain.htm. Details of the
policy governing data transactions is available at http://www.auslig.gov.au/pipc/
csdc/csdcsdti.htm.

Volume 5 Data Custodianship and Access 23

8 REFERENCES

Janzen, J. 1993. Management Guide to Custodianship. Ministry of Forests, BC,
Canada. (http://mofwww. for.gov.bc.ca/isb/datadmin/s35.htm)

Janzen, J. 1995. Presentation Summary: Corporate Data Categories. Corporate
Spatial Database Meeting, November 20/21, 1995, Ministry of Forests, BC,
Canada. (http://mofwww. for.gov.bc.ca/isb/datadmin/summpres.htm)

Onsrud, H.J. 1989. Legal and Liability Issues in Publicly Accessible Land
Information Systems, Proc. GIS/LIS, Vol. 1. pp 295-300.

24 WCMC Handbooks on Biodiversity Information Management

ce ES incl ipa Siew Mal
or bars $5 atte ois: aetioniiaie 4

‘ Ty vA

—_eatatoge veil. ies Messiaen Foi: ; pgesiene a
: igen Sih Jeong: AEN yh eal Ny. Micwtey: at Foqal
a a hla mee ¥ hee wR Me eva id ee ee tent NT

i Rs Fraps tua vies cr = ‘i

! bebe: sage ta, | : oe t a ube att | ng ; ~ hs
abc,

: , su > , ; Ep VRS
eer Teen MIO ER ap Bee 1
- ’ .

7 4 ‘ he) " 4 i

. Ais ~ =

WCMC Handbooks on
Biodiversity Information Management

These handbooks have been developed for use by senior
decision-makers and mid-career professionals. They review
the issues and processes involved in the management of
biodiversity information to support the conservation and
sustainable use of living resources. They also provide a
framework for the development of national plans and
strategies and for meeting reporting obligations of
international programmes and conventions. Collectively, the
handbook series may be used as a training resource or, more
generally, to support institutions and networks involved in
building capacity in information management.

Companion Volume

Volume 1 Information and Policy

Volume 2 Information Needs Analysis
Volume 3 Information Product Design
Volume 4 Information Networks

Volume 5 Data Custodianship and Access
Volume 6 Information Management Capacity
Volume 7 Data Management Fundamentals

DARWIN INITIATIVE SECRETARIAT
4-A2 Ashdown House, 123 Victoria Street
London SWIE 6DE. United Kingdom

COMMONWEALTH SECRETARIAT
Marlborough House, Pall Mall
London SW1Y 5HX. United Kingdom

WORLD CONSERVATION MONITORING CENTRE
219 Huntingdon Road
Cambridge CB3 0DL
United Kingdom

Tel: +44 1223 277314
Fax: +44 1223 277136

E-mail: info@wemce.org.uk

WCMC Internet Home Page
http://www.wemce.org.uk