Environ Monit Assess DOI 10.1007/s10661-011-1993-y

The BIOTA Biodiversity Observatories in Africa—a standardized framework for large-scale environmental monitoring

Norbert Jürgens · Ute Schmiedel · Daniela H. Haarmeyer · Jürgen Dengler · Manfred Finckh · Dethardt Goetze · Alexander Gröngröft · Karen Hahn · Annick Koulibaly · Jona Luther-Mosebach · Gerhard Muche · Jens Oldeland · Andreas Petersen · Stefan Porembski · Michael C. Rutherford · Marco Schmidt · Brice Sinsin · Ben J. Strohbach · Adjima Thiombiano · Rüdiger Wittig · Georg Zizka

Received: 4 August 2010 / Accepted: 23 February 2011 © Springer Science+Business Media B.V 2011

Abstract The international, interdisciplinary bio- BIOTA Biodiversity Observatories, that meet the diversity research project BIOTA AFRICA ini- following criteria (a) enable long-term monitor- tiated a standardized biodiversity monitoring ing of biodiversity, potential driving factors, and network along climatic gradients across the relevant indicators with adequate spatial and tem- African continent. Due to an identified lack of poral resolution, (b) facilitate comparability of adequate monitoring designs, BIOTA AFRICA data generated within different ecosystems, (c) developed and implemented the standardized allow integration of many disciplines, (d) allow

Electronic supplementary material The online version of this article (doi:10.1007/s10661-011-1993-y) contains supplementary material, which is available to authorized users. N. Jürgens · U. Schmiedel (B) · D. H. Haarmeyer · A. Koulibaly J. Dengler · M. Finckh · J. Luther-Mosebach · Laboratoire de Production et Amélioration Végétales, G. Muche · J. Oldeland U.F.R. Sciences de la Nature, Université Biodiversity, Evolution and Ecology of Plants, d’Abobo-Adjamé, URES Daloa, 02, Biocentre Klein Flottbek and Botanical Garden, BP 150 Daloa 02, Côte d’Ivoire University of Hamburg, Ohnhorststr. 18, 22609 Hamburg, Germany A. Petersen e-mail: [email protected] Department of Research Management and Funding, University of Hamburg, Moorweidenstr. 18, D. Goetze · S. Porembski 20148 Hamburg, Germany Department of Botany, Institute of Biological Sciences, University of Rostock, Wismarsche Str. 8, M. C. Rutherford 18051 Rostock, Germany Applied Biodiversity Research Division, South African National Biodiversity A. Gröngröft · J. Luther-Mosebach · A. Petersen Institute (SANBI), Kirstenbosch, Institute of Soil Science, University of Hamburg, Rhodes Avenue, Newlands, Cape Town 7700, Allende-Platz 2, 20146 Hamburg, Germany South Africa

D. H. Haarmeyer · K. Hahn · R. Wittig M. C. Rutherford Chair of Ecology and Geobotany, Institute of Ecology, Department of Botany and Zoology, Stellenbosch Evolution and Diversity, J. W. Goethe-University, University, Private Bag X1, Matieland 7602, Siesmayerstr. 70, 60323 Frankfurt am Main, Germany South Africa Environ Monit Assess spatial up-scaling, and (e) be applicable within DFG Deutsche Forschungsgemeinschaft a network approach. A BIOTA Observatory en- DIVERSITAS An International Programme of compasses an area of 1 km2 and is subdivided Biodiversity Science into 100 1-ha plots. For meeting the needs of EBONE European Biodiversity Observa- sampling of different organism groups, the hectare tion Network plot is again subdivided into standardized sub- GEO BON Group on Earth Observations plots, whose sizes follow a geometric series. To Biodiversity Observation Network allow for different sampling intensities but at the GEOSS Global Earth Observation Sys- same time to characterize the whole square kilo- tem of Systems meter, the number of hectare plots to be sampled GLORIA Global Observation Research depends on the requirements of the respective Initiative in Alpine Environments discipline. A hierarchical ranking of the hectare ILTER International Long-Term Eco- plots ensures that all disciplines monitor as many logical Research hectare plots jointly as possible. The BIOTA Ob- ÖFS (German) Ökologische Flächen- servatory design assures repeated, multidiscipli- stichprobe nary standardized inventories of biodiversity and ROSELT/OSS Réseau d’Observatoires de Sur- its environmental drivers, including options for veillance Écologique à Long Terme/ spatial up- and downscaling and different sam- Observatoire du Sahara et du pling intensities. BIOTA Observatories have been Sahel installed along climatic and landscape gradients WMO World Meteorological Organization in Morocco, West Africa, and southern Africa. In regions with varying land use, several BIOTA Observatories are situated close to each other to Introduction analyze management effects. Biodiversity loss is one of the most complex chal- Keywords Diversity · Global change · lenges for science and society world-wide (Sala Permanent plot · Sampling scheme · et al. 2000; World Resources Institute 2005). Its Transect · Vegetation negative effects on ecosystem services and hu- man welfare are well documented (Dobson et al. 2006; Costanza et al. 2007; Turner et al. 2007). Abbreviations Therefore, the 193 countries that are parties to the Convention on Biological Diversity (CBD) ALTER-Net A Long-Term Biodiversity, Eco- agreed to the political goal of reducing the rate system and Awareness Research of biodiversity loss by 2010, an ambitious and Network scientifically problematic goal (Mace et al. 2010). BDM (Swiss) Biodiversity Monitoring One of the major weaknesses of this “2010 goal” BIOTA Biodiversity Monitoring Tran- was and still is the inadequate data on changes in sect Analysis in Africa biodiversity in space and time. Benchmarking of CBD Convention on Biological Diversity success or failure of the “2010 goal” is possible

M. Schmidt · G. Zizka B. J. Strohbach Research Institute Senckenberg and J. W. National Botanical Research Institute (NBRI), Goethe-University, Senckenberganlage 25, P/Bag 13184, Windhoek, Namibia 60325 Frankfurt am Main, Germany A. Thiombiano B. Sinsin Laboratoire de Biologie et d’Écologie Végétales, Laboratoire d’Ecologie Appliquée, Faculté des Sciences Unité de Formation et Recherche en Sciences Agronomiques, Université d’Abomey-Calavi, de la Vie et de la Terre, Université de Ouagadougou, 01 B. P. 526, Cotonou, Bénin 03 BP 7021 Ouagadougou 03, Burkina Faso Environ Monit Assess only for certain aspects of biodiversity. For exam- on the processes driving the changes in patterns ple, remote sensing techniques allowing the mea- of diversity at the community level at different surement of the change in spatial representation spatial scales. of certain ecosystems (e.g., forests, wetlands), and One of the reasons for this gap of empirical measurements of population size of some species medium and small-scale information is the (mostly large mammals) are amongst the few ex- lack of global or even regional methodological amples of successful monitoring. standards in biodiversity research, in particular This discussion highlights the need for stan- regarding analyzed spatial scales. As nearly all dardized methods to measure rates of biodiversity aspects of biodiversity are scale-dependent change, as mandated by the “2010 goal” (Pereira (Wiens 1989;Noss1990;Storchetal.2007; and Cooper 2006). We point out the difficulty Goetze et al. 2008; Dengler et al. 2009; see Fig. 1) in verifying whether projections of future species and ecological processes have cross-scale effects losses (Thomas et al. 2004; McClean et al. 2006; (Carpenter et al. 2006), multi-scale indicators for van Vuuren et al. 2006; Sommer et al. 2010) accu- biodiversity are required. If measured repeatedly, rately depict trends. Additionally, we emphasize they will provide evidence for the extent and that these projections still critically lack empirical rate of the biological response to environmental baseline data on local patterns of biodiversity and change at different spatial scales and thus provide their dynamics and interactions within communi- evidence that is required to validate, support, ties and habitats (Scholes et al. 2008). and improve current projections of biodiversity The lack of empirical biodiversity observation change (Araújo et al. 2005). In analogy to the long data is obvious at several levels of complexity; history of meteorological monitoring according even basic inventories of present (global to local) to standards of the World Meteorological Orga- biodiversity are missing despite their eminent role nization (WMO; see http://www.wmo.int/), which as baseline and reference data for changes over informs climate change projections, standardized time. Species richness, the central “currency” for biodiversity monitoring is required to inform biodiversity (Gaston and Spicer 2005), is reason- projections of global biodiversity change. ably documented at a large scale only (grain = The need for biodiversity research and moni- approx. 1,000–100,000 km2, e.g., Gaston 2000; toring network(s) that develop and implement a Mutke and Barthlott 2005), and only for well- concept and methodology for a standardized or at studied organism groups like vascular plants and least harmonized design for measurement of bio- vertebrates. Medium-scale information (grain = diversity change within ecosystems and real land- approx. 1–100 km2) on biodiversity is available, scapes has long been understood (Noss 1990; but only for selected taxa in some well-studied Yoccoz et al. 2001; Balmford et al. 2005; Carpenter regions like Europe, while standardized small- et al. 2006; Pereira and Cooper 2006; Grainger scale data (grain = 1m2–10 ha) covering larger 2009). Such a standardized design should be suit- areas (i.e., spatial extents) remain underutilized able for different biomes, allow spatial up-scaling (for review, see Dengler 2009b). For the majority and long-term monitoring as well as facilitate in- of the global surface, and particularly the most terdisciplinary approaches within a regional to biodiverse regions, comparable data are missing global network. Various environmental monitor- even on vascular plants and vertebrates, not to ing initiatives have been suggested and imple- mention the more specious groups of non-vascular mented, particularly during the last decade in plants, invertebrates, fungi, and microbes (con- order to understand and quantify environmental stituting 83% of the described species on Earth changes, e.g., ILTER (Kim 2006), ROSELT/OSS according to the figures in Lecointre and Le (Aïdoud et al. 2008), GLORIA (Pauli et al. Guyader (2006). Even if such information on local 2004;see“Comparison with other monitoring to regional biota were fully available, we would networks”), RAINFOR (Malhi et al. 2002), and not be able to project reliably the effect of global TEAM (see www.teamnetwork.org). More re- environmental change on global, regional, or even cently, several EU-funded projects (ALTER- local biodiversity, owing to a lack of information Net, see http://www.alter-net.info; EBONE, see Environ Monit Assess http://www.ebone.wur.nl/UK/) have been devel- and as ground-truth data for validations of models oped as a response to the need for consoli- and projections. They are thus critical for the de- dated biodiversity data. The bioDiscovery Core velopment of adaptation and mitigation strategies Project(Ashetal.2009) of the international for resource management under global climate DIVERSITAS Program (Loreau and Olivieri change. 1999) was launched to facilitate the process. For this purpose, BIOTA AFRICA developed However, in 2000, when the research project the BIOTA Observatories as a monitoring tool BIOTA AFRICA (Biodiversity Monitoring Tran- that covers the different levels of complexity (i.e., sect Analysis in Africa; Jürgens 2004,seehttp:// genes, species, ecosystems) and dimensions (i.e., www.biota-africa.org) initiated a standardized composition, structure, function and evolution) of biodiversity monitoring network across Africa, biodiversity (sensu Noss 1990; see Fig. 1). The no monitoring design that met all the above- design of these permanent observation sites within mentioned criteria was available (for details, see typical landscapes should: “Lessons learned and potential improvements”). • BIOTA AFRICA therefore developed and imple- Allow long-term monitoring of biodiversity mented the required standardized design, to be as well as of indicators and potential driving applied along climatic gradients across the African factors of biodiversity change with adequate continent, the so-called BIOTA Biodiversity Ob- spatial and temporal resolution. • servatories (further: BIOTA Observatories). Enable the monitoring of a broad range of In this article, we describe the BIOTA Obser- different taxa. • vatory design, which so far has been only partially Include measurement of important potential published, mostly in sources not easily or wide- abiotic (e.g., climate, soil characteristics) and ly accessible (Jürgens 1998, 2006; Schmiedel and biotic (e.g., land use, demography, biotic inter- Jürgens 2005;Krugetal.2006). We will focus on actions) drivers of change. • the design of the monitoring of vascular plants Analyse several different spatial scales, thus for which a standardized regular monitoring has supporting spatial up-scaling. • been furthest developed and implemented. We Allow comparison of data gained from then give an overview where BIOTA Observa- different biomes, ecosystems, and land use tories have so far been implemented covering regimes. major biomes throughout the African continent. Further, the philosophy of BIOTA AFRICA con- Further, we discuss the lessons learned and siders as important: suggestions for improvements that arose from our practical experience of 9 years of biodiver- • Integrating approaches of different scientific sity monitoring. Finally, we briefly discuss the disciplines (from remote sensing to ground advantages of the BIOTA Observatory design truth, social and natural sciences, empirical in comparison to other long-term monitoring and modelling approaches). frameworks. • Involving local stakeholders into the develop- ment and implementation of the observation program within a participatory process. Aims and criteria of the BIOTA AFRICA • Integrating the BIOTA Observatories into re- approach gional to global observation networks, with the aim to implement and adjust the required The basic aims of the project BIOTA AFRICA standards and long-term perspective globally. were to provide scientifically sound data on bio- diversity, its environmental driving factors, and BIOTA Observatories meeting these require- its changes in time for selected observation sites ments have been established and tested in north- representing major biomes and ecosystems of the ern, western and southern Africa along trans- African continent. Such data are needed urgently continental transects following major climatic gra- for ecological research, conservation planning, dients (for details, see “The BIOTA biodiversity Environ Monit Assess

Consequences Consequences for monitoring for monitoring

COMPOSITION STRUCTURE

e.g. annual relevés on Beta + gamma Plant formations, e.g. remote sensing different spatial scales diversity, habitats land cover change detection with ground truthing

Alpha diversity, e.g. annual relevés and Growth forms, biogeography e.g. hyperspectral population counts functional traits remote sensing

Population, e.g. SSR monitoring clade, … Genetic diversity, e.g. monitoring genetic mutation, diversity of cultivars Gene recombi- regulation Genesnation Phylogeny, e.g. phylogeography e.g. seed production, adaptation, Biotic interaction predation, plasticity, (competition, facilitation) selection soil fertility Species e.g. biomass & Biomes, trophic nets Co-evolution, functional links, e.g. parasite evolution forage value ESF&S, threats & impact evolutionary environments Ecosystems FUNCTION EVOLUTION

Fig. 1 Levels of complexity (genes, species, ecosystems) the conceptual design of the BIOTA monitoring approach and dimensions (composition, structure, function, and evo- (amended after Noss 1990) lution) of biodiversity as well as their consequences for

observatories in Africa”). The BIOTA Observa- Sampling layout of the BIOTA Biodiversity tories provide in-situ measured evidence for bio- Observatories diversity changes and their driving factors across the African continent. The resulting time series Spatial layout of ground-based data are complemented by mod- ern remote sensing-based biodiversity monitoring A BIOTA Observatory encompasses an area of tools (Oldeland et al. 2010). While the BIOTA 1km2 (1,000 m × 1,000 m) with boundaries Observatory data are critical to provide ground- oriented along cardinal directions (Fig. 2). The truthing for the remote-sensing monitoring, the spatial layout and subplot numbering of BIOTA latter can help to extrapolate the information Observatories is mirrored across the equator. The into a larger domain (Duro et al. 2007). For a 1-km2 area is divided into 100 1-hectare plots of better understanding of the processes that drive 100 m × 100 m, with corner points well marked spatial and temporal patterns of biodiversity, and geo-referenced. The hectare plots are num- BIOTA AFRICA also established several “aux- bered from 00 to 99 starting in the upper left iliary observatories” of deviating design, where corner (i.e., the North-West corner in the southern specific observation tasks or specific experiments hemisphere) and running from the left (West) were carried out (e.g., Jürgens 2006; Musil et al. to the right (East) and pole-wards through the 2009). BIOTA Observatory. The hectare plots constitute Environ Monit Assess

Fig. 2 Schematic layout 100 m of a BIOTA Observatory (in the southern hemisphere) and arrangement of different N Vegetation sampling sampling areas within one 50 m of the hectare plots

10 m 20 m 10 m 100 m

Soil Profile 1000 m

Zoological sampling 1000 m Automatic weather station the largest replicated sampling unit within the 1986) and developed a ranking method based BIOTA Observatory. on the d’Hondt divisor rules procedure (Balinski In general, the number of investigated hectare and Ramirez 1999; Palomares and Ramirez 2003; plots should be large enough to allow a statisti- Taagepera and Shugart 1989). For this purpose, cally robust description of the BIOTA Observa- tory. However, within the multidisciplinary team, one problematic issue became evident. The activ- ities of the involved disciplines had to be aligned 0 1 2 3 4 5 6 7 8 9 spatially in order to achieve the best possible inte- 0 88 83 56 94 67 92 27 45 21 50 gration of data, while the risk of artefacts caused 1 43 73 84 8 38 4 22 60 2 87 by the various scientific activities had to be min- imized. Our procedures for selecting the hectare 2 19 58 46 5 80 55 53 24 76 29 plots and assigning areas within hectare plots for 3 54 33 41 28 10 15 36 31 96 98 different purposes aim at the best possible com- promise between the two conflicting aims. 4 85 11 99 32 49 61 74 65 86 63

Some of the disciplines apply techniques that 5 100 69 75 59 78 52 91 39 34 90 are too laborious to allow investigation of a larger number of hectare plots. This fact could result 6 18 77 14 17 81 42 82 26 93 1 in spatial “fragmentation” of activities and hence 7 66 44 37 47 57 25 95 30 62 48 cause a lack of integration of these disciplines and their respective organism groups. To overcome 8 72 89 40 13 79 51 16 9 64 7 this problem, all disciplines agreed to do their 9 12 35 3 70 20 71 23 6 97 68 sampling work on hectare plots following a pre- defined sequence. This sequence is defined by a Fig. 3 Example of the grid system of a BIOTA Observa- ranking procedure, which assigns to each hectare tory (S08, Niko North, Namibia, see Table 2) with ranking plot one of the ranks 1 to 100 (Fig. 3). numbers. Line and row numbers are given at the left and top margins, respectively; the plot numbers are derived by Complete randomization of plots in heteroge- the combination of the line number (f irst f igure)andthe neous environments involves the risk of disre- row number (second f igure). Different shades of grey rep- garding rare habitat types (Wildi 1986; Ruxton resent the four habitat types distinguished on this BIOTA and Colegrave 2006;Rolecekˇ et al. 2007). To Observatory (white stony plain; light grey wash (plain); medium grey stony slope; dark grey wash-rivier). The num- ensure a representative randomized sampling, bers of the 20 highest ranked grids (those on which vascular we employed a stratified sampling design (Wildi plants are analyzed) are printed in bold letters Environ Monit Assess each hectare plot is characterized by a combi- trapping area (often involving digging or interfer- nation of its predominant vegetation and geo- ence with plants) on the one hand are placed in the morphological structures, further referred to as pole-ward halves of the hectare plots whereas the “habitat type”. The basic principle of the d’Hondt area for botanical small-scale monitoring as well divisor method is to divide the number of hectare as lichenological and microbiological sampling are plots of each stratum (i.e., habitat type) by natural arranged in its equator-ward part (see Fig. 3). numbers (1, 2, 3, 4, ... n). The resulting quotients Similarly, the standardized position of the destruc- of all habitat types are sorted in descending order. tive soil profile is set 4 m pole-ward of the centre The highest quotient and therefore its habitat type point of the hectare (Fig. 3). would be allocated to rank number 1, the habi- tat type of the second quotient would be ranked Sampling of vascular plants second, and so on. With this method, a sequence of the habitat types can be compiled according to The vegetation monitoring is done on sets of their real proportions. three nested plots, with plot sizes following a geo- Deviating from the original method of d’Hondt, metric series (100 m2, 1000 m2, and 10,000 m2). we up-weighed the proportion of rare habitat The plots are permanently marked with common types by using the square root of habitat frequency metal fencing poles or buried magnets. While the as basis for the determination of the ranking or- 100- and 10,000-m2 plots are squares, the 1000- der. This modification has the effect of positioning m2 plots have a rectangular shape of 20 m × the hectare plots of less common habitat types 50 m (Fig. 2). These plot dimensions are fre- on higher ranks. This effect is desirable because quently used among vegetation and biodiver- it ensures a minimum of replicates even for rare sity studies worldwide (Shmida 1984; Peet et al. habitat types in small sample sizes. Given the 1998; Stohlgren 2007) and are regularly applied allocation of the habitat types, the rank number for vegetation mapping in Namibia (Strohbach of the hectare plots is selected randomly within 2001; Hüttich et al. 2009). The type of sampled each. This results in a ranking of all hectare plots, data differs between the different plot sizes (see each with a ranking number between 1 and 100. Table 1). For the estimation of cover, the vegeta- Figure 3 shows an example of the grid struc- tion is divided in vertical strata according to stan- ture and ranking order of a BIOTA Observa- dardized height categories. Total species cover as tory. Besides achieving representativeness among well as cover of each vertical layer is estimated the plots sampled, the application of this ranking as accurately as possible in percent. It should be method allows for interdisciplinary studies as the stressed that deviating from phytosociological tra- same ranking, and therefore the same sequence of dition, where cover values below 1% are usually plots, is applied across all disciplines. not further differentiated (e.g., Dierschke 1994), Each discipline defines the number of hectare accuracy up to two decimal points is advisable for plots to be analyzed (x) and then carries out its re- the estimation of very low cover values as they search on the hectare plots of ranks 1 to x. Accord- frequently occur in arid environments (see similar ingly, the highest-ranked hectare plots are jointly sampled by all disciplines working on the BIOTA Observatory. This obviously involves the risk of disturbance, interference, and artefact. To reduce Table 1 Types of vascular plant data sampled within the different plot sizes such problems, experimental studies are allowed only outside the 1-km2 boundary of the BIOTA Plot size Presence Cover Abundance 2 a Observatory and the researchers are asked to 100 m XXX 1,000 m2 XX walk primarily along the lines from corner to 10,000 m2 = 1ha X corner of the hectare plots and leave its central a parts undisturbed. Further, the more intensive-use Only perennial plant species may be included, depending on limitations of data gathering. In a few BIOTA Obser- study sites of the different disciplines are spatially vatories with relatively dense vegetation, graminoid species separated within each hectare plot. The zoological were excluded from abundance counts Environ Monit Assess recommendation in Pauli et al. 2004). Each plot Soil studies is photographed in a standardized way for visual documentation. Soil studies are conducted in the ranking se- Individual-based quantitative and spatial mon- quence and at the pre-defined position within the itoring allows the recording of population dynam- hectare plots (Fig. 2); for details, see Petersen ics and thus a highly sensitive set of indicators (2008) and Petersen et al. (2010). A soil profile of for change. Such a monitoring is generally applied 0.6–1.2 m depth is described with respect to the in the 1000-m2 and 100-m2 plots of highest rank. following parameters: stratification, texture, rock There, each individual of all occurring peren- fragments, color, humus content, lime content, nial species (in the 1000-m2 plot only for nano- soil and surface structure, crusting, bulk density, phanerophytes and larger life forms) is measured penetration resistance, and distribution of roots (i.e., height: difference between ground to highest (FAO 2006). Each profile is documented with a living part of the plant; first diameter: longest photograph of the profile and the surrounding diameter; second diameter: longest diameter or- habitat. thogonal to first diameter). Additionally, its rela- Mixed samples for laboratory analyses are tive position is drawn on a plot map or registered taken from all horizons of the profile. Addition- in a 0.5-m grid. In humid biomes of West Africa, ally, defined sample volumes are taken using core woody species regeneration is also monitored in samplers in order to determine the bulk density. plots of 25 m2 size where size (height, diameter) The depth of the individual samples corresponds and position of each woody plant individual is to horizon boundaries. If there is no distinct hori- recorded. zon in the topsoil, the depth of the first sample All vegetation monitoring is repeated at regu- is 10 cm. While assessing the soil profile, the soil lar intervals (i.e., annually in arid and semi-arid material is deposited on a large sheet. The profile regions and typically in three-year intervals in is refilled after sampling to minimize distur- tropical regions) and at about the same time of bance of the site and its vegetation. The profiles the year during the phenological phase of maximal are described according to widely accepted stan- vegetation development. dards (e.g., FAO 1990; Schoeneberger et al. 2002) and then classified according to a globally applicable system (IUSS Working Group WRB Sampling other organism groups 2006).

Within BIOTA AFRICA, many other organ- Weather stations ism groups have been sampled with standard- ized designs on some or on nearly all BIOTA In close vicinity to each BIOTA Observatory, Observatories, with the sampling design and an automatic weather station is installed to re- intensity adapted according to disciplinary needs. late the time series of biodiversity data to local In BIOTA Southern Africa, for example, invento- weather conditions and long-term climatic trends. ries have included lichens on all substrata (Zedda Measured weather parameters include rainfall, air et al. 2008, 2011) and biological soil crusts with temperature, relative air humidity, leaf wetness, their constituent organisms such as cyanobacte- solar radiation, wind speed, and direction. ria and algae (Büdel et al. 2009). Further, fungi, ground-dwelling beetles (Henschel et al. 2010), termites (Vohland and Deckert 2005), ants, but- Additional data on driving factors terflies and moths, as well as small mammals (Giere and Zeller 2005) have been recorded ac- In BIOTA AFRICA, the ecological monitoring cording to uniform sampling protocols. The de- on the BIOTA Observatory was complemented tailed description of the specific sampling methods by studies on economic, legal, administrative, so- is not a subject of the present publication (for cial, and cultural driving factors of the local to details, see Haarmeyer et al. 2010). regional land use (e.g., Falk 2008; Pröpper 2009; Environ Monit Assess

Vollan 2009; Vollan et al. 2009). Additional stud- The BIOTA Biodiversity Observatories in Africa ies, which are located in the vicinity of the BIOTA Implementation Observatory, complement the monitoring data by experimental approaches (e.g., grazing exclosures BIOTA Observatories form the “backbone” of or active restoration treatments), to study ecolog- the large African–German, interdisciplinary re- ical processes underlying the observed changes. search project BIOTA AFRICA. Since 2001, they These studies do not use the BIOTA Obser- have been arranged along mega-transects that vatory design but relate to respective biodiver- follow important large-scale environmental gra- sity baseline data or infrastructure (e.g., weather dients in Africa (Fig. 4), thereby covering many stations). major ecosystem types of the biomes sampled

Fig. 4 Map of the BIOTA Maroc BIOTA West Africa established BIOTA W01 W02 Niger Observatories in Africa Mali Burkina Faso W03 (as of 4 April 2010). For Morocco W04 further information on W05 M02 Guinea the individual BIOTA W06 M01 W07 Nigeria Observatories, see Benin Ghana Ivory Coast Table 2 To go Algeria Liberia

0250 500 0500250 km km

BIOTA Southern Africa Zambia Angola

S42 S01 S43 S02 Namibia S03

S04 S05 Botswana S41 S16 S06 S38 S34 S36 S39 S40 S35 S37 S08 S09 S10 S11

S12 S17 S18 S21 S20 S24 South Africa S22 S25 S26 S27 S45 S28 S29 N S31 S32 S33 0250 500 km Environ Monit Assess 2006–2009 2006–2009 Woodland Savanna Woodland Savanna Woodland SavannaThornbush Savanna Thornbush Savanna Thornbush 2008, Savanna 2009 Sagebrush Steppe de la Comoé Overview of the established BIOTA Observatories in Africa (as of 4 April 2010) with their main characteristics S09S10S11 Niko SouthS12 Gellap Ost Nabaos Namibia Karios Namibia Hardap Namibia Karas Namibia Karas near Gibeon Karas Keetmanshoop 1076 Keetmanshoop 1099 Nama Karoo Karasburg 1045 Nama Karoo Nama Karoo 909 – Nama Karoo x 2001, 2002, 2004, 2001–2009 x 2001–2009 x 2001–2009 S01 Mile 46 Namibia Kavango near Cove 1180 Southern African x 2001–2003, 2005–2009 S02S03 MutompoS04 SonopS05 Namibia ToggekryS06 Otjiamongombe KavangoS08 Namibia Namibia Namibia Okamboro Otjozondjupa Niko Otjozondjupa North Otjozondjupa near Cove Namibia near Maroelaboom Erichsfelde Omatako 1236 Ranch Namibia Otjozondjupa 1180 1519 Southern African Hardap 1495 Ovitoto Southern African Southern African Southern African x near Gibeon x 1490 x 2001–2003, 2005, 2006, x 2001–2003, 2005–2009 Southern 2001–2009 African 1070 2004–2009 Nama Karoo x 2004–2009 – 2001, 2002, 2004, W01W02W03 YomboliW04 KolelW05 KikideniW06 NatiabouaniW07 Alibori Burkina Faso Terroire Villageois Sahel Comoé Benin Burkina Faso Burkina Burkina Faso Faso Est Est Sahel Benin Atacora Yomboli Ivory Coast Atacora Zanzan Kolel Natiabouani Kikideni Péhunco 286 Péhunco 258 Parc National Sahel Grass- and Shrubland 304 295 327 North – Sudanian Savanna 220 North Sudanian Savanna Sahel South 2001, Grass- Sudanian 2002 x and Savanna 297 Shrubland – South x Sudanian Savanna 2001–2003 x South Sudanian Savanna 2001 2001, 2003 – 2002 x 2001 2002 M01M02 El Miyit Taoujgalt Morocco Morocco Souss-Massa-Draâ Souss-Massa-Draâ Tizi n Taoujgalt Tafilalet 1869 736 Saharan Semidesert Ibero-Mauritanian – 2002, 2003, 2005–2009 – 2002–2005, 2008, 2009 BIOTA Southern Africa: North–South Transect, Namibia BIOTA West Africa Observatory Observatoryno.BIOTA Maroc Country name Primary Locality Altitude administrative unit Biome (m a.s.l.) Soil Years with data vascular plant data Table 2 Environ Monit Assess Thornbush Savanna 2005–2007 Thornbush Savanna Thornbush Savanna Thornbush Savanna Woodland Savanna Woodland Savanna National Park S45 Nieuwoudtville South Africa Northern Cape Nieuwoudtville 722 Fynbos – 2007 S18S20 KoeroegavvlakteS21 NumeesS22 GrootdermS24 South Africa SoebatsfonteinS25 PaulshoekS26 Northern Cape RemhoogteS27 GoedehoopS28 Sendelingsdrif South Africa South South Africa Africa RatelgatS29 MoedverlorenS31 Northern Cape Northern Northern Cape Cape South Rocherpan AfricaS32 South Africa Riverlands SoebatsfonteinS33 Alexander Sendelingsdrif Bay South Africa Northern Cape Elandsberg Northern Cape 635 South Africa Cape of Western Good Cape Leliefontein South Hope Africa Leliefontein Succulent Karoo Western Cape South Africa South Van Western Africa Rhynsdorp Cape South Africa 392 193 Western Vredendal 362 South Cape Africa Western Cape Van Western x Rhynsdorp Succulent Cape Karoo Succulent Succulent Karoo Karoo Western Piquetberg Cape Table 1048 2001–2009 Mountain 1027 MalmesburyS38 245 Wellington Succulent Karoo x Succulent Karoo x x Succulent Claratal KarooS39 239 2001–2009 140 2001–2006,S40 2001–2003, 2008 2006 Succulent Narais Karoo xS41 83 Succulent x Karoo Duruchaus x 35 140 2001–2009 Fynbos Sandveld 2001–2009 2001–2009 Fynbos Fynbos 95 x Namibia – Fynbos 2001–2009 2001–2009 Namibia Namibia Khomas Namibia x Khomas Khomas – – 2008, 2009 Omaheke Windhoek 2005 2004, x 2008, 2009 near 2009 Rehoboth near Rehoboth near Drimiopsis 1865 1614 1624 1523 Southern African Nama Karoo Nama Karoo Southern African x x 2005, 2007, 2009 x x 2005, 2008, 2009 2004–2009 2004–2009 S17 Alpha South Africa Northern Cape AskhamS34 896S35 KleinbergS36 Southern African GobabebS37 Ganab Rooisand x Namibia 2002, 2003, Namibia Erongo Namibia Namibia Erongo Erongo Khomas Walfis Bay Gobabeb Rooisand Desert Namib-Naukluft Ranch Park 1160 995 Southern 188 African Namib Desert 419 Namib Desert – Namib Desert 2005 – – 2010 – 2010 2004 S43 Omano go Ndjamba Namibia Omusati Ogongo 1100 Southern African x 2007–2009 S16S42 Wlotzkasbaken Ogongo Namibia Erongo Namibia Omusati Swakopmund Ogongo 73 Namib Desert 1103 Southern x African 2001–2003, 2005, 2010 x 2007–2009 BIOTA Southern Africa: North–South Transect, South Africa BIOTA Southern Africa: East–West Transect, Namibia More detailed information is available in Online Resource 1 Environ Monit Assess

(Fig. 4). They are situated within relatively ho- Zeller 2005; Vohland and Deckert 2005; Büdel mogeneous sites representative of the respective et al. 2009; Hahn-Hadjali et al. 2006; Wittig et al. region. Mostly, they correspond to zonal vegeta- 2007; Schmidt et al. 2010; Schmiedel et al. 2010a; tion, i.e., vegetation that is mostly determined by Zedda et al. 2011) as well as between different regional climate and hardly modified by soil and land use types (e.g., Hoffmann and Zeller 2005; geomorphological properties or human influence Vohland et al. 2005; Koulibaly et al. 2006; Mayer (see Walter and Breckle 1983). Where land use et al. 2006; Pufal et al. 2008) have been ana- differed significantly within a region (e.g., regard- lyzed with data from the BIOTA Observatories. ing grazing intensities), two or more BIOTA Ob- Time series provided insights into population and servatories were placed close to each other to ecosystem dynamics (Jürgens 2006). Data from cover a wider section of this variability. the BIOTA Observatories and their surround- Up to now, 46 BIOTA Observatories (Fig. 4, ings were fed into ecological models on mainte- Table 2, Online Resource 1) have been es- nance of biodiversity (Reineking et al. 2006), gene tablished and studied by researchers from ap- flow in fragmented landscapes (e.g., Blaum and proximately 50 institutions from six African Wichmann 2007), and hydrological processes countries (Benin, Burkina Faso, Ivory Coast, (Popp et al. 2009a, b; Tietjen et al. 2010). Many Morocco, Namibia, South Africa) and Germany. more analyses of the African BIOTA Observatory The BIOTA Observatories belong to BIOTA data are in progress, in particular, on biodiversity Maroc (n = 2), BIOTA West Africa (n = 7), patterns and their drivers at different spatial scales and BIOTA Southern Africa (n = 37). Year across the African continent as well as on nine- of implementation, frequency of repetition, and year time series of in-situ monitoring data. intensity of sampling varied depending on project priority settings of the respective BIOTA Ob- servatory and technical constraints (Table 2,On- Lessons learned and potential improvements line Resource 1). The monitoring data gained are stored centrally at the BIOTA Data Fa- Over a period of 9 years, the BIOTA Observa- cility of the BIOTA Head Office at the Bio- tory approach has been applied in various loca- centre Klein Flottbek, University of Hamburg, tions and biomes throughout the African conti- Germany (Muche et al. 2010). Data storage em- nent, involving numerous scientists of different ploys BIOTABase, a software specifically de- disciplines. Here, we report practical hints, and signed to store, process, and facilitate analyses suggestions for potential improvements. of long-term, multi-scale biodiversity data (http:// www.biota-africa.org/biotabase_a.php;seeMuche Ranking system and Finckh 2009). The vegetation databases of the three regional BIOTA projects are all For practical reasons, in southern Africa the habi- registered in the Global Index of Vegetation- tat classification for each single BIOTA Obser- Plot Databases (GIVD; Dengler et al. 2011; vatory, which was required for the ranking at see www.givd.info), with the unique IDs AF- the beginning of the project, was carried out by MA-001 (Morocco), AF-00-001 (West Africa), different scientists from various disciplines. Due and AF-00-003 (Southern Africa). The data to the number of people involved with differing are available on request for scientific research disciplinary perspectives and due to the wide vari- purposes. ety of biomes covered, the classifications achieved are appropriate for the respective BIOTA Obser- Results achieved vatory but hardly support habitat-specific compar- isons along the entire southern Africa transects. So far, soil (e.g., Mills et al. 2006; Petersen 2008; For future applications, we suggest a priori agree- Medinski et al. 2010) and biodiversity patterns and ment on a list of clearly defined, transect-wide processes along climatic gradients (e.g., Uhlmann applicable habitat types as has been done in West et al. 2004; Zedda and Rambold 2004;Giereand Africa. Environ Monit Assess

Further, the square-root weighting of the habi- implementation, we suggest assigning sufficient tat types within the ranking procedure might time and effort that allows for comprehensive be worth reconsidering. It indeed ensures the sampling or, alternately, sampling the 1-ha scale inclusion of rare habitat types in the sampling only for a subset of hectare plots. but at the same time complicates the deriva- As the BIOTA Observatory spatial layout is tion of parameter means for a whole BIOTA designed to allow for unlimited up- and downscal- Observatory and causes problems in statistical in- ing of plots, the addition of smaller plot sizes (e.g., ference (e.g., Botta-Dukát et al. 2007; Lájer 2007). 10, 1, 0.1, 0.01 m2) could easily be accomplished From a statistical point of view, it might be favor- and has already been tested on single BIOTA able to use an unmodified d’Hondt approach to Observatories (e.g., Peters 2010). The implemen- determine the hectare plots to be sampled for the tation of these small plots sizes with some repli- overall characterization of a BIOTA Observatory cation within the hectare plots (see Shmida 1984; and to complement this basic sampling—where Stohlgren et al. 1995; Peet et al. 1998; Dengler necessary—with additional hectare plots of rare 2009b) as a general standard, would add valu- habitats. The latter could be used for between- able information with little additional effort. This Observatory comparisons within transect-wide could, for example, characterize β-diversity via accepted habitat types, while the mean values the z-values of the species–area relationship, or al- calculated without these “additional” hectare low the extrapolation of species richness to larger plots are unbiased spatial means and thus al- areas (e.g., Dengler 2009a; Dengler and Oldeland low statistical comparison between whole BIOTA 2010). The need for a stronger focus on smaller Observatories. plot sizes was also identified by researchers work- ing in ecosystems with high vegetation density Spatial scales and species richness at small spatial scales (humid biomes of West Africa). On the other hand, larger The design of the BIOTA Observatories is plot sizes (e.g., 10 and 100 km2) might be appro- inherently multi-scaled because the analysis of priate for the sampling of some highly mobile ani- biodiversity at multiple spatial scales is fundamen- mals (e.g., large mammals, dragonflies), presently tal for the understanding of patterns and their not well covered by the BIOTA Observatory driving factors (see “Introduction”). For exam- design. ple, species inventories of vascular plants were explicitly sampled at 100-, 1,000-, and 10,000-m2 Representativeness scales, and are available as cumulative data of the 20-ha plots sampled plus occasional observa- At the start of the BIOTA AFRICA project, tions on the 1-km2 scale. As the deviation from the BIOTA Observatories as monitoring tools the square shape at one of the spatial scales and thus indicators for biodiversity change were (20 m × 50 m) causes difficulties in analyses such placed at sites that were regarded as representa- as species–area relationships (Stohlgren 2007; tive for the larger landscape. However, as a tool Dengler 2008), future implementations should for assessing biodiversity patterns at different spa- consider using square-shaped plots throughout tial scales in the landscape, the density of BIOTA (see case study in Namibia by Peters 2010). A Observatories should be further increased in or- squareshapealsowouldbetteralignwithremote- der to document spatially restricted patterns and sensing data, which uses square pixels (Alexander processes within a network of BIOTA Observa- and Millington 2000; Richards and Xiuping 2006). tories. The representativeness within the square Regarding the sampling of the hectare plots, kilometer was achieved with the ranking sys- it turned out that the time needed for complete tem. However—under shifting cultivation or fire sampling of vascular plant species on an entire events—a hectare plot may change within a hectare plot often exceeded the allocated time- year from semi-natural vegetation to an arable frame, especially for BIOTA Observatories with field. For within-Observatory data, such spatio- relatively dense vegetation. Therefore, for future temporal changes do not at all invalidate the Environ Monit Assess

BIOTA Observatory design, but on the other to achieve a consistent taxonomy of plants and hand, the (stratified) random design actually al- animals in databases with multiple contributors lows the documentation of such small-scale pat- (Berendsohn 1997; Jansen and Dengler 2010). terns and determination of their frequency and However, this task becomes much more trou- their causes (see detailed discussion in Dengler blesome in transnational projects (with different 2009b). However, in BIOTA Observatories with national checklists) within regions where a large particularly high spatio-temporal variability, it proportion of the flora and fauna is still await- might be necessary to sample more than the ing classification or at least is not covered in usual 20 hectare plots in order to determine pat- identification keys. This underscores the need for terns and processes with sufficient accuracy and voucher specimens, which must be identified later resolution. and archived properly, a time-consuming process. The delay in identification may also cause tem- Field sampling porary “inflation” of taxon numbers in the joint database due to inconsistent use of preliminary From BIOTA fieldwork, several practical hints field names for unidentified species. emerged that might be helpful for future re- searchers. The size and type of the marking Time effort material for the plots should be adjusted to the environment: poles can become overgrown by Generally, the workload for biodiversity moni- vegetation if too short; while other poles might toring varies strongly between organism groups be removed by passers-by or run over by rang- and biomes. To give an idea of the time ac- ing animals (e.g., cattle, elephants). Searches for tually needed, we refer again to our experi- “lost” poles can cost precious time. In areas where ences from the monitoring of vascular plants. marking material might become lost, exact coordi- One hectare with three nested plots of 100 m2– nates (taken with differential GPS) for all corners 10,000 m2), sampled according to the BIOTA of each plot are critical for replacing the poles at standards (see “Aims and criteria of the BIOTA their exact position. For some Observatories near AFRICA approach”) by one researcher, re- the sea, a rust-resistant form of marking material quired between one (Namib Desert) and five or is recommended. For the annual monitoring of more hours (species-rich plots in the Succulent vascular plant vegetation, cumulative species in- Karoo, southern African Savanna or Fynbos Bio- ventories per plot as derived from the previous mes). Individual-based monitoring required be- years’ datasets were copied into the field data tween one (dwarf shrub-dominated habitats in sheets. This reduced nomenclatural inconsistency southern Morocco) and more than three working between the years and guaranteed consistency in days (dense savanna vegetation in West Africa) the use of unavoidable “field names” for species in per plot. The time effort strongly increases with regions where the flora is less well known. For the the size of the plot. Thus, vegetation surveys on assessment of the nested vegetation plots within 1,000 m2 take much longer due to the structural a hectare, it proved to be efficient to start with complexity and size of the plot, and the estimated the smallest plot and then to continue with the cover values are less reliable than on 100 m2. next larger sizes. Another important aspect when For the same reason, the species inventories per different observers are involved over time is to hectare plot are likely to be incomplete in many standardize methods of cover estimation as far cases (Peters 2010). In comparison, Dolnik (2003) as is possible. Calibration of estimates between reports that an experienced botanist needs up to different observers before the work starts would 14 h to compile a complete species list for a 900-m2 be ideal. plot in structure-rich habitats even in temperate Determination, taxonomy, and databasing Europe despite the limited and well-known flora there. Even in industrialized countries where biodiver- As a consequence of the aspects mentioned, we sity is generally well-documented, it is not trivial recommend that for monitoring vascular plants Environ Monit Assess manpower allocated per BIOTA Observatory support of the biodiversity monitoring activities should be increased, the number of monitored was highly valuable, but they were also invaluable plots per plot size reduced, or the temporal fre- in facilitating the communication of research quency of monitoring adapted. However, arid and findings with the local stakeholders (Araya et al. semi-arid biomes with high inter-annual variabil- 2009; Schmiedel et al. 2010b). ity of rainfall require annual monitoring in order Also for data storage and maintenance per- understand the response of plant populations to manent staff should be responsible as it requires various driving factors (climate vs. land use). In infrastructure and institutional continuity that go tropical biomes with less variable rainfall pat- far beyond the resources of typical projects with terns, monitoring at lower frequencies (e.g., every limited funding periods. In addition to the local 3 years) may be sufficient. Further, it might be hosting and processing of the data, permanently worth considering carrying out individual-based implemented central data facilities should guaran- monitoring in several smaller rather than one tee good long-term data quality and security (see larger plot. also Scholes et al. 2008).

Institutional implementation Comparison with other monitoring networks The practical experience from a research project with temporarily employed project staff (pri- Meanwhile, various biodiversity monitoring ini- marily Ph.D. students) for the monitoring work tiatives have emerged as a result of the ongoing reveals that changes in personnel were the ma- debate on global biodiversity decline (Heywood jor source of inconsistency in monitoring data. and Watson 1995;Salaetal.2000;WorldRe- Further, BIOTA AFRICA, which was both a re- sources Institute 2005), most of them during the search and a monitoring project, faced the prob- last decade. Four important initiatives that, like lem that research, which is evaluated based mainly BIOTA, aim at facilitating long-term in-situ on short-term publication output, and monitor- monitoring of terrestrial biodiversity, are pre- ing, whose value increases with the number of sented in Table 3 and are compared with the consistent datasets over large spatial and tempo- BIOTA Observatories: i.e., ILTER = Interna- ral scales, often require different sampling ap- tional Long-Term Ecological Research (Kim proaches. As the potential publication output of 2006), ROSELT/OSS = Réseau d’Observatoires such large-scale, medium- to long-term monitor- de Surveillance Écologique à Long Terme/ ing typically is beyond the time available to in- Observatoire du Sahara et du Sahel (Aïdoud dividual researchers, there tends to be a conflict et al. 2008), GLORIA = Global Observation between the researcher’s short-term need for sci- Research Initiative in Alpine Environments entific output and the project’s need for time- (Pauli et al. 2004, 2009), and the DFG Bio- consuming long-term monitoring data. This also diversity Exploratories (Fischer et al. 2010). Some is the reason why the long-term monitoring on others have recently been reviewed by Dengler the Observatories only played a minor role in the (2009b), namely the European forest monitor- academic capacity development (i.e., involvement ing (e.g., Aamlid et al. 2007), the Swiss Bio- of Ph.D. candidates or Postdocs) at the African diversity Monitoring (BDM; e.g., Hintermann research institutes. As a consequence, we recom- et al. 2000), and the German “Ökologische mend employing permanent staff for the monitor- Flächenstichprobe” (ÖFS; e.g., Hoffmann-Kroll ing fieldwork. et al. 2000). As required by the CBD, most In order to support the time-consuming other countries have implemented some kind of fieldwork on the BIOTA Observatories, BIOTA national biodiversity monitoring (see Bischoff Southern Africa employed at full-time and trained and Dröschmeister 2000), but there is little stan- eight members of local land user communities dardization among countries, and comparabil- as BIOTA para-ecologists. This turned out to ity of data is thus low. In Europe, as a response be a promising approach: the para-ecologists’ to this lack, the EU-funded projects EBONE Environ Monit Assess in forest 2 100 m grid; of these grid area subdivided with a × 2 in grassland, 400 m 2 biodiversity of different taxa and levels management for biodiversity for ecosystem processes The understanding of The role of land use and The role of biodiversity 100 are intensive plots and 16 very are included and fluxes, remote sensing • • 16 m • 2 ) 2 2 and 150 m 2 , approx. 100 m 2 permanent plots andeight larger sectors ofvariable size (depending intensive plots (theon higher the work the relief; “intensity”, together approx. the on more one50 disciplines plot); m various experimental approaches are sampled Écologique à Long Terme/Observatoire du Sahara et du Sahel) ecological monitoringsites the relationship between (International LongTerm EcologicalResearch) (Réseau d’Observatoires de Surveillance Research Initiative (Global Observation in Alpine Environments) ) 2 , Not standardized Not standardized 1 m 2 ,(1km 2 , 1000 m 2 , subdivided into Not standardized Three-scale sampling: A site is constituted by 400-km 2 climate and land usechanges on biodiversity any kind networks of that long-term comprise effects of desertification around the Sahara. change effects on mountain environments for studies dealing with: 20 are selected accordingto a stratified-randomprocedure; within these,the smaller plots are nested exploitation unit plot or herd summits of different points, 500 in grassland and altitude 500 in the same in forest vegetation are region; selected for summits, each 16 of 1-m the (not random); of these 1000 plots, lichens), zoology (varioustaxa), soil sciences, mostlyclimatology, botany, remote zoology,sensing, socio-economy geomorphology, soil soil sciences, nutrient cycling, obligatory, socio-economy bryophytes climatology, agronomy, sciences, bryophytes, hydrology, lichens; including abiotic environment and lichens optional), mycology, microbiology, soil genetics), zoology (various taxa), (e.g., soil, climate) sciences, ecosystem pools BIOTA Observatories ILTER ROSELT/OSS GLORIA DFG Biodiversity Exploratories Comparison of major long-term monitoring frameworks with the BIOTA Observatories (as of 4 April 2010) monitoring design 100 1-ha plots, of which administrative unit, at least four mountain 100 m vascular plant 10,000 m Objective Analysing effects of Network of 32 national Tackling the causes and Assessment of climate Open research platform Plot sizes for 100 m General sampling 1 km Disciplines Botany (vascular plants, Not standardized, but Botany, zoology, Botany (vascular plants Botany (vascular plants, Table 3 Environ Monit Assess ) 2010 temperatures and humidity of soil and air ; biomass, growth, 2 ) Fischer et al. ( 2009 , 2004 plots and 8 sectors 100 2 rough 5-point scalecover-abundance estimate of vascular plants species in the 8 primary sectors productivity ) Pauli et al. ( 2008 of the respectiveobservatory 4 summits per site ) Aïdoud et al. ( 2006 329 Europe, 12 Africa,114 Asia, 4 Oceania) sites now included in ILTER have been monitored for a even longer period in setup, 12 planned lakes, coral reefs, ocean, agricultural and urban ecosystems) , adapted to vegetation per summit; at least 2 ) 2 , 1000 m 2 abundance of allvascular plantspecies standardized abundance of vascular estimates; intensive plant, bryophyte, and study sites: lichen Standing species on 1 m forest-savannamosaic, semi-deciduous tropical tundras,forest, prairie fynbos and hot and cold deserts, other grassland ecosystems wetlands, coastal ecosystems, ecotone upwards) semi-natural to intensively used) but aiming atglobal application industrialized countries concentrated in and Sahel in Africa www.biota-africa.org www.ilternet.edu/ www.oss-online.org www.gloria.ac.at www.biodiversity-exploratories.de biodiversity change of vegetationmonitoring data monitoring presence, of cover and vegetation not observatory; of presence, cover and (every 5–10 years) abundance, cover surveys, presence, monitoring 10,000 m already implementedimplementation 15 South America,covered 25 in total States) some of the humid savannas, and tropical forests, been resurveyed), 15 the Sahara (fromthe treeline ecosystems (from information onpotential drivers of type stations, and land intensity use (not in every site) intensive plots per site: per site for vegetation (100 m Website Reference The present paper Kim ( Frequency and type Annual to three-year Not standardized Species lists per Irregular monitoring “Repeated” vegetation Number of sites 46Year of first 2001Vegetation 523 types (49 North America; Desert, arid and 12 pilot observatories, 1980 62 (LTER active in Any (21 United (boreal, of , temperate these have 3 1992 Arid ecosystems around Alpine ecosystems Grassland and forest 2001 2006 Regions covered Africa (at present), Global, but sites are North Africa Global, but so far no site Germany Additional site Automatic weather Not standardized Not standardized Soil temperature Weather station on 100 Number of subplots 20 of each size Not standardized Not standardized, 16 1-m Environ Monit Assess

(European Biodiversity Observation Network; inclusion of many experimental approaches. On see http://www.ebone.wur.nl/UK/) and ALTER- the other hand, due to the lack of a completely Net (ALong-Term Biodiversity, Ecosystem and random or systematic plot placement, they do not Awareness Research Network; see http://www. even allow derivation of spatially valid means of alter-net.info) now aim at coordinating the vari- biodiversity parameters within each of the three ous national monitoring initiatives. Exploratories, let alone valid extrapolations to The presently most comprehensive global bio- larger areas. diversity monitoring network, ILTER (Kim 2006; Regarding spatial scale, up to now, most mon- see Table 3), is actually a network of many inde- itoring schemes are restricted to only one scale, pendent national long-term observation networks neglecting the fact that biodiversity patterns, their lacking a functioning standardization in the moni- drivers and their responses to global change are toring protocol among the nations and often even likely to be scale-dependent (see Turner and within the national subnetworks. Therefore, IL- Tjørve 2005; Field et al. 2009; Dengler 2009b). TER, while comprising some of the most valu- While the European forest monitoring suggests able ecological long-term datasets worldwide, is applying a uniform plot size of 400 m2 (but in not comparable to the BIOTA Observatories that application among different countries plots of aim at generating standardized and thus glob- 10–1,200 m2 are used; see Dengler 2009b), the ally comparable data for the future. The sec- German ÖFS and the DFG Biodiversity Explo- ond global monitoring network, GLORIA, has a ratories, for example, use different plot sizes for highly standardized, very detailed sampling pro- individual habitat types. This makes comparison tocol(Paulietal.2004, 2009), but is restricted between studies impossible and does not allow for to plants in one specific habitat type (mountain spatial or temporal transitions between habitat summits), with climate solely as a driving factor types. GLORIA and BDM both have two differ- of biodiversity change (see Table 3). Two fur- ent spatial scales, but the bigger one (GLORIA: ther supranational monitoring schemes, the Euro- summit sector; BDM: transect within 1 km2) is not pean forest monitoring (Aamlid et al. 2007)and comparable between sites. With presently three ROSELT/OSS (Aïdoud et al. 2008;seeTable3) (or four if the 1 km2 is also counted) defined are both restricted to specific ecosystems. Among spatial scales for sampling, the BIOTA Obser- the national biodiversity monitoring schemes, the vatories are clearly outstanding in this respect. Countryside Survey in the United Kingdom (CS; Only schemes like the Carolina Vegetation Survey since 1978; e.g., Bunce 2000) and the Swiss BDM (Peet et al. 1998) and the approach proposed by (since 2000; e.g., Hintermann et al. 2000) belong Dengler (2009b) analyze more spatial scales, but to the most extensive and methodologically most both have been applied only regionally for one-off advanced programs. Within the BDM, for ex- inventories and not for monitoring so far. ample, complete vascular plant species lists are recorded every 5 years for 1,600 10-m2 perma- nent plots and for standardized transects within Conclusions and outlook 520 1-km2 areas alongside data on various ani- mal groups (Hintermann et al. 2000). These plots The BIOTA Observatories provide the following are distributed in a stratified-random approach basic features that have been identified as being over the whole of Switzerland and thus allow critical for a design of a global biodiversity mon- the calculation of mean biodiversity trends across itoring network: (a) they are standardized but at the territory of the country. Finally, the DFG the same time flexible regarding the minimum Biodiversity Exploratories are outstanding among number of hectare plots sampled; (b) they are all reviewed monitoring schemes regarding the designed for sampling at more than one spatial wide taxonomic coverage (including, apart from scale with down- and up-scaling options; (c) they vascular plants and vertebrates, also bryophytes, focus on time series; (d) they are suitable interdis- lichens, many above- and below-ground inverte- ciplinary approaches; (e) they are applicable in all brates as well as fungi and microbes) and the types of biomes; and (f) the highly flexible open- Environ Monit Assess access database software BIOTABase facilitates vation Network (GEO BON) within the Global the handling and processing of time-series data Earth Observation System of Systems (GEOSS; from nested plots as they are inherent to BIOTA Scholes et al. 2008). Such a global network of Observatories. With this combination of features, standardized BIOTA Observatories that provide BIOTA Observatories are unique among existing empirical, spatially explicit information on biodi- biodiversity monitoring schemes (see “Lessons versity and, more general, environmental changes, learned and potential improvements”). The ap- would be a crucial contribution for global change plicability of the BIOTA Observatory approach research in order to develop adaptation and miti- has been tested and proven for over 9 years of gation strategies from local to global scales. BIOTA research across a wide range of different biomes on the African continent. Many results Acknowledgements The basic idea of the BIOTA from this standardized sampling on BIOTA Ob- AFRICA sampling design was conceived by N.J. The spa- tial design was then developed in detail by N.J., U.S., A.G., servatories in Africa have already been published, and A.P. While the botanical sampling was discussed and with many more analyses in progress, in partic- elaborated by N.J., A.K., A.T., B.S., D.G., G.Z., K.H.-H., ular, those of the 9-year monitoring data (see M.F., M.S., S.P., R.W. and U.S., the soil sampling was “Sampling of vascular plants”). defined by A.G. and A.P. Practical experiences in sampling and analysing biodiversity data with the BIOTA Obser- The existing BIOTA Observatories provide im- vatory approach were contributed by J.O. and M.F. for portant information on biodiversity patterns on Morocco, by A.K., A.T., B.S., D.G., G.M., K.H.-H., M.S., different spatial scales—and therefore fill gaps S.P. and R.W. for West Africa, and by B.J.S., D.H.H., J.D., of critical baseline information that are missing J.L.-M., J.O., N.J., M.C.R. and U.S. for Southern Africa. The overview of the existing BIOTA Observatories was even for generally much better studied regions collated by G.M., J.L.-M., J.O., A.K., D.G., K.H.-H., M.S., of the Earth than Africa. The long-term in situ and M.F., and the map was produced by J.O. This article monitoring, commenced on BIOTA Observato- was planned by U.S., N.J., M.F., G.M., D.H.H., and J.D., ries and complemented by remote sensing, is crit- primarily written up by U.S., D.H.H., J.D., and N.J., and was critically revised by A.G., A.K., A.P., A.T., B.J.S., ical to identify long-term trends in biodiversity B.S., D.G., G.M., J.L.-M., J.O., K.H.-H., M.C.R., M.S., changes and to differentiate them from medium- R.W., S.P., and G.Z. The development and implementation term responses of biodiversity to climatic oscilla- of BIOTA Observatories in Africa has been funded by tions (O’Connor and Roux 1995; Anyamba and the German Federal Ministry of Education and Research under promotion numbers 01 LC 0601A (Morocco), and Eastman 1996; Hereford et al. 2006). We thus 01 LC 0017, 01 LC 0617D1 (West Africa), and 01 LC 0024, propose considering the BIOTA Observatory de- 01 LC 0024A, 01 LC 0624A2 (Southern Africa). We thank sign, potentially with those improvements sug- the respective provincial and national departments for al- gested in “The BIOTA biodiversity observatories lowing access and research in protected and unprotected areas. The authors are grateful to all BIOTA AFRICA in Africa”, for worldwide application. This ap- colleagues in various African countries and Germany for proach is applicable in any terrestrial (and discussion of the design and implementation of the BIOTA semi-terrestrial) biome and landscape worldwide, Observatories, to Curtis Björk and Will Simonson for im- irrespective of spatio-temporal heterogeneity or proving our English language usage, and to two anonymous reviewers for constructive comments on an earlier version degree of human influence, and provides a wide of this article. array of standardized indicators of biodiversity, its change and its driving factors in a time- and cost- efficient manner. The BIOTA Observatories in Africa have been References implemented and monitored by a team of African and German researchers. Based on our experi- Aamlid, D., Canullo, R., & Starlinger, F. (Eds.) (2007). Manual on methods and criteria for harmonized sam- ences, we promote the continuation and extension pling, assessment, monitoring and analysis of the of the monitoring on BIOTA Observatories in effects of air pollution on forests—Part VIII: As- Africa and the extension of this initiative into sessment of ground vegetation (Updated 10/2007). other parts of the world. The discussed strengths http://www.icp-forests.org/pdf/manual8.pdf. Accessed 26 July 2010. of the BIOTA Observatory design suggest them Aïdoud, A., Jauffret, S., & Sakona, Y., (2008). Long-term as a key component of the Biodiversity Obser- environmental monitoring in a circum-Saharan net- Environ Monit Assess

work: The ROSELT/OSS experience. OSS Synthesis Carpenter, S. R., DeFries, R., Dietz, T., Mooney, H. A., Collection 3. Tunis: OSS, Tunis. Polasky, S., Reid, W. V., et al. (2006). Millennium Alexander, R., & Millington, A. C. (Eds.) (2000). Vegeta- ecosystem assessment: Research needs. Science, 314, tion mapping: From patch to planet. Chichester: Wiley. 257–258. Anyamba, A., & Eastman, J. R. (1996). Interannual vari- Costanza, R., Fisher, B., Mulder, K., Liu, S., & Christopher, ability of NDVI over Africa and its relation to El Nino T. (2007). Biodiversity and ecosystem services: A Southern Oscillation. International Journal of Remote multi-scale empirical study of the relationship be- Sensing, 17, 2533–2548. tween species richness and net primary production. Araújo, M. B., Pearson, R. G., Thuiller, W., & Erhard, Ecological Economics, 61, 478–491. M. (2005). Validation of species–climate impact mod- Dengler, J. (2008). Pitfalls in small-scale species–area els under climate change. Global Change Biology, 11, sampling and analysis. Folia Geobotanica, 43, 269– 1504–1513. 287. Araya, Y. N., Schmiedel, U., & von Witt, C. (2009). Link- Dengler, J. (2009a). Which function describes the species– ing ‘citizen scientists’ to professionals in ecological area relationship best?—A review and empirical eval- research, examples from Namibia and South Africa. uation. Journal of Biogeography, 36, 728–744. Conservation Evidence, 6, 11–17. Dengler, J. (2009b). A flexible multi-scale approach for Ash, N., Jürgens, N., Leadley, P., Alkemade, R., Araújo, standardised recording of plant species richness pat- M. B., Asner, G. P., et al. (2009). bioDISCOVERY: terns. Ecological Indicators, 9, 1169–1178. Assessing, monitoring and predicting biodiversity Dengler, J., & Oldeland, J. (2010). Effects of sampling pro- change. bioDISCOVERY Science Plan and Imple- tocol on the shapes of species richness curves. Journal mentation Strategy. DIVERSITAS Rep. 7. DIVER- of Biogeography, 37, 1698–1705. SITAS, Paris, http://www.diversitas-international.org/ Dengler, J., Löbel, S., & Dolnik, C. (2009). Species con- uploads/File/bioDiscovery_sp_final.pdf. Accessed 26 stancy depends on plot size—a problem for vegetation July 2010. classification and how it can be solved. Journal of Veg- Balinski, M., & Ramirez, V. (1999). Parametric methods of etation Science, 20, 754–766. apportionment, rounding and production. Mathemati- Dengler,J.,Jansen,F.,Glöckler,F.,Peet,R.K., cal Social Science, 37, 107–122. De Cáceres, M., Chytrý, M., et al. (2011). The Global Balmford, A., Crane, P., Dobson, A., Green, R. E., & Index of Vegetation-Plot Databases (GIVD): A new Mace, G. (2005). The 2010 challenge: Data availabil- resource for vegetation science. Journal of Vegetation ity, information needs and extraterrestrial insights. Science, 22. doi:10.1111/j.1654-1103.2011.01265.x. Philosophical Transactions of the Royal Society B, 360, Dierschke, H. (1994). Pf lanzensoziologie—Grundlagen 221–228. und Methoden. Stuttgart: Ulmer. Berendsohn, W. G. (1997). A taxonomic information Dobson, A., Lodge, D., Alder, J., Cumming, G. S., Keymer, model for botanical databases: The IOPI Model. J., McGlade, J., et al. (2006). Habitat loss, trophic col- Taxon, 46, 283–309. lapse, and the decline of ecosystem services. Ecology, Bischoff, C., & Dröschmeister, R. (Eds.) (2000). European 87, 1915–1924. monitoring for nature conservation. Schriftenreihe für Dolnik, C. (2003). Artenzahl-Areal-Beziehungen von Lanschaftspf lege und Naturschutz 62 (199 pp.). Bonn: Wald- und Offenlandgesellschaften—Ein Beitrag Bundesamt für Naturschutz. zur Erfassung der botanischen Artenvielfalt unter Blaum, N., & Wichmann, M. (2007). Short term trans- besonderer Berücksichtigung der Flechten und Moose formation of matrix into hospitable habitat facilitates am Beispiel des Nationalparks Kurischen Nehrung gene flow and mitigates fragmentation. Journal of An- (Russland). Mitteilungen der Arbeitsgemeinschaft imal Ecology, 76, 1116–1127. Geobotanik in Schleswig-Holstein und Hamburg, 62, Botta-Dukát, Z., Kovács-Láng, E., Rédei, T., Kertész, M., 1–183. & Garadnai, J. (2007). Statistical and biological con- Duro, D., Coops, N. C., Wulder, M. A., & Han, T. (2007). sequences of preferential sampling in phytosociologiy: Development of a large area biodiversity monitoring Theoretical considerations and a case study. Folia system driven by remote sensing. Progress in Physical Geobotanica, 42, 141–152. Geography, 31, 235–260. Büdel, B., Darienko, T., Deutschewitz, K., Dojani, S., Falk, T. (2008). Communal farmers’ natural resource use Friedl, T., Mohr, K., et al. (2009). Southern African and biodiversity preservation—A new institutional eco- biological soil crusts are ubiquitous and highly diverse nomic analysis from case studies in Namibia and South in drylands, being restricted by rainfall frequency. Mi- Africa. Göttingen: Cuvillier. crobial Ecology, 57, 229–247. FAO (Ed.) (1990). Guidelines for soil description (3rd ed.). Bunce, R. G. H. (2000). The experience of the Country- Rome: FAO. side Survey in Great Britain for monitoring biodiver- FAO (Ed.) (2006). Guidelines for soil description (4th ed.). sity of the wider countryside. In C. Bischoff & R. Rome: FAO. Dröschmeister (Eds.), European monitoring for nature Field, R., Hawkins, B. A., Cornell, H. V., Currie, D. J., conservation. Schriftenreihe für Lanschaftspf lege und Diniz-Filho, A. F., Guégan, J.-F., et al. (2009). Spa- Naturschutz 62 (pp. 95–104). Bonn: Bundesamt für tial species-richness gradients across scales: A meta- Naturschutz. analysis. Journal of Biogeography, 36, 132–147. Environ Monit Assess

Fischer, M., Kalko, E. K. V., Linsenmair, K. E., Pfeiffer, Hoffmann-Kroll, R., Benzler, A., Schäfer, D., & Seibel, S., Prati, D., Schulze, E.-D., et al. (2010). Explorato- S. (2000). Setting up national biodiversity monitoring ries for large-scale and long-term functional biodiver- for nature conservation in Germany—the Ecological sity research. In F. Müller, C. Baessler, H. Schubert, Area Sampling (EAS). Schriftenreihe für Lanschaft- & S. Klotz (Eds.), Long-term ecological research— spf lege und Naturschutz, 62, 79–94. between theory and application (pp. 429–443). Berlin: Hüttich, C., Gessner, U., Herold, M., Strohbach, B. J., Springer. Schmidt, M., Keil, M., et al. (2009). On the suitabil- Gaston, K. J. (2000). Global patterns in biodiversity. ity of MODIS time series metrics to map vegetation Nature, 405, 220–227. types in dry savanna ecosystems: A case study in the Gaston, K. J., & Spicer, J. I. (2005). Biodiversity: An intro- Kalahari of NE Namibia. Remote Sensing, 1, 620–643. duction (2nd ed.). Malden, MA: Blackwell. IUSS Working Group WRB (2006). World reference base Giere, P., & Zeller, U. (2005). Small mammal diversity and for soil resources 2006: A framework for international reproduction along a transect in Namibia (BIOTA S classification, correlation and communication. World 07). In B. Huber, J. Sinclair, & K. H. Lampe (Eds.), Soil Resources Reports, 103.Rome:FAO. African biodiversity: Molecules, organisms, ecosystems Jansen, F., & Dengler, J. (2010). Plant names in vegetation (pp. 305–313). Berlin: Springer. databases—a neglected source of bias. Journal of Veg- Goetze, D., Karlowski, U., Tockner, K., Watve, A., Riede, etation Science, 21, 1179–1186. K., & Porembski, S. (2008). Spatial and temporal di- Jürgens, N. (1998). Biodiversity monitoring transect analy- mensions of biodiversity dynamics. In W. Barthlott, sis. In W. Barthlott & M. Gutmann (Eds.), Bio- K. E. Linsenmaier, & S. Porembski (Eds.), Biodiver- diversitätsforschung in Deutschland. Potentiale und sity: Structure and function. In UNESCO (Ed.), Ency- Perspektiven (pp. 1–73). Bad Neuenahr-Ahrweiler. clopedia of Life Support Systems (EOLSS) (pp. 166– Jürgens, N. (2004). BIOLOG—Africa. Research towards 208). Oxford EOLSS. sustainable use and conservation of biodiversity in Grainger, A. (2009). Towards a new global forest science. Africa. Introduction. In E. Beck, W. G. Berendsohn, International Forestry Review, 11, 126–133. M. Boutros, M. Denich, K. Henle, N. Jürgens, et al. Haarmeyer, D. H., Luther-Mosebach, J., Dengler, J., (Eds.), Sustainable use and conservation of biological Schmiedel, U., Finckh, M., Berger, K., et al. (2010). diversity—A challenge for society. Proceedings of the The BIOTA Observatories. In N. Jürgens, D. H. International Symposium Berlin, 1–4 December 2003 Haarmeyer, J. Luther-Mosebach, J. Dengler, M. (pp. 130–131). Berlin: Federal Ministry of Education Finckh, & U. Schmiedel (Eds.), Biodiversity in south- and Research. ern Africa. Volume 1: Patterns at local scale— Jürgens, N. (2006). Recent change of flora and vegetation the BIOTA Observatories (pp. 6–801). Göttingen: in Namibia—A brief review of dynamics, drivers and Hess. scientific approaches. In H. Leser (Ed.), The chang- Hahn-Hadjali, K., Schmidt, M., & Thiombiano, A. (2006). ing culture and Nature of Namibia: Case studies. The Phytodiversity dynamics in pastured and protected Sixth Namibia Workshop Basel 2005. In Honour of Dr. West African savannas. In S. A. Ghazanfar & h. c. Carl Schlettwein (1925–2005) (pp. 91–108). Basel: H. J. Beentje (Eds.), Taxonomy and ecology of Basler Afrika Bibliographien. African Plants: Their conservation and sustainable Kim, E.-S. (2006). Development, potentials, and challenges use—Proceedings of the 17th AETFAT Congress Ad- of the International Long-Term Ecological Research dis Ababa 21.-26.09.2003 (pp. 351–359). Kew: Royal (ILTER) Network. Ecological Research, 21, 788–793. Botanic Gardens. Koulibaly, A., Goetze, D., Traoré, D., & Porembski, S. Henschel, J. R., Grohmann, C., Siteketa, V., & Linsenmair, (2006). Protected versus exploited savanna: Charac- K. E. (2010). Monitoring tenebrionid beetle biodiver- teristics of the Sudanian vegetation in Ivory Coast. sity in Namibia. African Study Monographs, Supple- Candollea, 61, 425–452. mentary Issue, 40, 117–128. Krug, C. B., Esler, K. J., Hoffman, M. T., Henschel, J., Hereford, R., Webb, R. H., & Longpre, C. I. (2006). Pre- Schmiedel, U., & Jürgens, N. (2006). North–South co- cipitation history and ecosystem response to multi- operation through BIOTA: An interdisciplinary mon- decadal precipitation variability in the Mojave Desert itoring programme in arid and semi-arid southern region, 1893–2001. Journal of Arid Environments, 67, Africa. South African Journal of Science, 102, 187– 13–34. 190. Heywood, V. H., & Watson, R. T. (Eds.) (1995). Global Lájer, K. (2007). Statistical tests as inappropriate tools for biodiversity assessment (1140 pp.). Cambridge: Cam- data analysis performed on non-random samples of bridge University Press. plant communities. Folia Geobotanica, 42, 115–122. Hintermann, U., Weber, D., & Zangger, A. (2000). Bio- Lecointre, G., & Le Guyader, H. (2006). Biosystematik— diversity monitoring in Switzerland. Schriftenreihe für Alle Organismen im Überblick. Berlin: Springer. Lanschaftspf lege und Naturschutz, 62, 47–58. Loreau, M., & Olivieri, I. (1999). Diversitas: An interna- Hoffmann, A., & Zeller, U. (2005). Influence of variations tional programme of biodiversity science. Trends in in land use intensity on species diversity and abun- Ecology and Evolution, 14, 2–3. dance of small mammals in the Nama Karoo, Namibia. Mace, G. M., Cramer, W., Díaz, S., Faith, D. P., Belgian Journal of Zoology, 135, 91–96. Larigauderie, A., Le Prestre, P., et al. (2010). Biodi- Environ Monit Assess

versity targets after 2010. Current Opinion in Environ- improve estimates of species diversity from remotely mental Sustainability, 2, 3–8. sensed spectral heterogeneity? Ecological Indicators, Malhi, Y., Phillips, O. L., Lloyd, J., Baker, T. R., Wright, 10, 390–396. J., Almeida, S., et al. (2002). An international network Palomares, A., & Ramirez, V. (2003). Thresholds of the to monitor the structure, composition and dynamics of divisor methods. Numerical Algorithms, 34, 405–415. Amazonian forests (RAINFOR). Journal of Vegeta- Pauli, H., Gottfried, M., Hohenwallner, D., Reiter, K., tion Science, 13, 439–450. Casale, R., & Grabherr, G. (Eds.) (2004). The Mayer, C., Soka, G., & Picker, M. (2006). The importance GLORIA Field Manual—Multi-Summit Approach. of monkey beetle (Scarabaeidae: Hopliini) pollination Luxembourg: Office for Official Publications of the for Aizoaceae and Asteraceae in grazed and ungrazed European Communities. areas at Paulshoek, Succulent Karoo. Journal of Insect Pauli, H., Gottfried, M., Klettner, C., Friedmann, B., Conservation, 10, 323–333. Laimer, S., & Grabherr, G. (Eds.) (2009). Amendment McClean, C. J., Doswald, N., Küper, W., Sommer, J. H., to the 4th-GLORIA Field Manual. Draft, July 2009. Barnard, P., & Lovett, J. C. (2006). Potential impacts Vienna: University of Vienna & Austrian Acad- of climate change on sub-Saharan African plant prior- emy of Sciences. http://www.gloria.ac.at/downloads/ ityareaselection.Diversity and Distribution, 12, 645– AMENDMENTS_GLORIA_Manual__DRAFT_2009- 655. 07_f.pdf. Accessed 26 July 2010. Medinski, T. V., Mills, A. J., Esler, K. J., Schmiedel, U., Peet, R. K., Wentworth, T. R., & White, P. S. (1998). A & Jürgens, N. (2010). Do soil properties constrain flexible, multipurpose method for recording vegeta- species richness? Insights from boundary line analysis tion composition and structure. Castanea, 63, 262–274. across several biomes in south western Africa. Journal Pereira, H., & Cooper, H. D. (2006). Towards the global of Arid Environments, 74, 1052–1060. monitoring of biodiversity change. Trends in Ecology Mills, A. J., Fey, M. V., Gröngröft, A., Petersen, A., & and Evolution, 21, 123–129. Medinski, T. V. (2006). Unravelling the effects of soil Peters, J. (2010). Plant diversity patterns at different properties on water infiltration: Segmented quantile spatial scales in a semi-arid savanna ecosystem regression on a large data set from arid south-west in central Namibia. Diplom thesis in Landscape Africa. Australian Journal of Soil Research, 44, 783– Ecology & Nature Conservation, University of Greifs- 797. wald. Also available from: http://www.biologie.uni- Muche, G., & Finckh, M. (2009). BIOTA Base short hamburg.de/bzf/syst/Diplom_thesis_Jan_Peters_2009. manual. Biocentre Klein Flottbek, University of pdf. Hamburg, Hamburg. http://www.biota-africa.org/ Petersen, A. (2008). Pedodiversity of southern African downloads/biotabase/BIOTABaseManual.pdf. drylands. Hamburger Bodenkundliche Arbeiten 62. Accessed 26 July 2010. Hamburg: Verein zur Förderung der Bodenkunde in Muche, G., Hillmann, T., Suwald, A., & Jürgens, N. (2010). Hamburg. Data access and availability: BIOTA data facility. In Petersen, A., Gröngröft, A., & Miehlich, G. (2010). Meth- U. Schmiedel & N. Jürgens (Eds.), Biodiversity in ods to quantify the pedodiversity of 1 km2 areas— southern Africa 2: Patterns and processes at regional results from southern African drylands. Geoderma, scale (pp. 337–342). Göttingen & Windhoek: Hess. 155, 140–146. Musil, C. F., van Heerden, P. D. R., Cilliers, C. D., Popp, A., Blaum, N., & Jeltsch, F. (2009a). Ecohydrological & Schmiedel, U. (2009). Mild experimental climate feedback mechanisms in arid rangelands: Simulating warming induces metabolic impairment and massive the impacts of topography and land use. Basic and mortalities in southern African quartz field succulents. Applied Ecology, 10, 319–329. Environmental and Experimental Botany, 66, 79–87. Popp, A., Vogel, M., Blaum, N., & Jeltsch, F. (2009b). Mutke, J., & Barthlott, W. (2005). Patterns of vascular Scaling up ecohydrological processes: Role of surface plant diversity at continental to global scales. In I. water flow in water-limited landscapes. Journal of Friis & H. Balslev (Eds.), Plant diversity and complex- Geophysical Research—Biogeoscience, 114, Article ity patterns—local, regional and global dimensions— G04013. doi:10.1029/2008JG000910. Proceedings of an international symposium held at Pröpper, M. (2009). Culture and biodiversity in central the Royal Danish Academy of Sciences and Letters Kavango, Namibia (p. 440). Berlin: Reimer. in Copenhagen, Denmark, 25–28 May, 2003 (pp. 521– Pufal, G., Mayer, C., Porembski, S., & Jürgens, N. (2008). 537). Biologiske Skrifter, 55. Copenhagen: Reitzels. Factors affecting fruit set in Aizoaceae species of Noss, R. F. (1990). Indicators for monitoring biodiver- species of the Succulent Karoo. Basic and Applied sity: A hierarchical approach. Conservation Biology, Ecology, 9, 401–409. 4, 355–364. Reineking, B., Veste, M., Wissel, C., & Huth, A. (2006). O’Connor, T. G., & Roux, P. W. (1995). Vegetation Environmental variability and allocation trade-offs changes (1947–1971) in a semi-arid, grassy dwarf maintain species diversity in a process-based model shrubland in the Karoo, South Africa: Influence of of succulent plant communities. Ecological Modelling, rainfall variability and grazing by sheep. Journal of 199, 486–504. Applied Ecology, 29, 247–260. Richards, J. A., & Xiuping, J. (2006). Remote sensing dig- Oldeland, J., Wesuls, D., Rocchini, D., Schmidt, M., & ital image analysis: An introduction (4th ed.). Berlin: Jürgens, N. (2010). Does using species abundance data Springer. Environ Monit Assess

Rolecek,ˇ J., Chytrý, M., Hájek, M., Lvoncík,ˇ S., & Tichý, L. Strohbach, B. J. (2001). Vegetation survey of Namibia. (2007). Sampling design in large-scale vegetation stud- Journal of the Scientif ic Society of Namibia, 49,1– ies: Do not sacrifice ecological thinking to statistical 31. purism! Folia Geobotanica, 42, 199–208. Taagepera, R., & Shugart, M. S. (1989). Seats and votes: Ruxton, G. D., & Colegrave, N. (2006). Experimental de- The ef fects and determinants of electoral systems sign for the life sciences (2nd ed.). New York: Oxford (p. 292). New Haven: Yale University Press. University Press. Thomas, C. D., Cameron, A., Green, R. E., Bakkenes, Sala, O. E., Chapin, F. S. III, Armesto, J. J., Berlow, E., M., Beaumont, L. J., Collingham, Y. C., et al. (2004). Bloomfield, J., Dirzo, R., et al. (2000). Global biodi- Extinction risk from climate change. Nature, 427, 145– versity scenarios for the year 2100. Science, 287, 1770– 148. 1774. Tietjen, B., Jeltsch, F., Zehe, E., Classen, N., Gröngröft, A., Schmidt, M., Agonyissa, D., Ouédraogo, A., Hahn-Hadjali, Schiffers, K., et al. (2010). Effects of climate change K., Thiombiano, A., Koulibaly, A., et al. (2010). on the coupled dynamics of water and vegetation in Changes in plant species composition following a cli- drylands. Ecohydrology, 3, 226–237. matic gradient in West Africa. In X. van der Burgt, J. Turner, W. R., & Tjørve, E. (2005). Scale-dependence in van der Maesen, J.-M. Onana (Eds.), Systematics and species–area relationships. Ecography, 28, 721–730. conservation of African Plants. Proceedings of the 18th Turner, W. R., Brandon, K., Brooks, T. M., Costanza, R., AETFAT Congress, Yaoundé, Cameroon (pp. 823– da Fonseca, G. A. B., & Portela, R. (2007). Global 828). Kew: Royal Botanical Gardens. conservation of biodiversity and ecosystem services. Schmiedel, U., & Jürgens, N. (2005). Biodiversity Obser- BioScience, 57, 868–873. vatories. A new standardised monitoring tool for bio- Uhlmann, E., Görke, C., Petersen, A., & Oberwinkler, diversity studies. Basic Applied Dryland Research, 1, F. (2004). Comparison of AMF species diversity in 87–91. winter-rainfall areas of South Africa and summer- Schmiedel, U., Dengler, J., Luther-Mosebach, J., rainfall areas of Namibia. Mycological Progress, 3, Gröngröft, A., Muche, G., Petersen, A., et al. (2010a). 267–274. Patterns and dynamics of vascular plant diversity van Vuuren, D. P., Sala, O. E., & Pereira, H. M. (2006). The along the BIOTA transects in southern Africa. In future of vascular plant diversity under four global U. Schmiedel & N. Jürgens (Eds.), Biodiversity in scenarios. Ecology and Society, 11(2), Article 25. southern Africa. Volume 2: Patterns and processes at http://www.ecologyandsociety.org/vol11/iss2/art25/. regional scale (pp. 118–135). Göttingen: Hess. Vohland, K., & Deckert, J. (2005). Termites (Isoptera) Schmiedel, U., Mtuleni, V. S., Christiaan, R. A., Isaacks, along a north–south transect in Namibia and South R. S., Kotze, D., Lot, M. J., et al. (2010b). The BIOTA Africa. Entomologische Zeitschrift, 115, 109–115. para-ecologist programme towards capacity develop- Vohland, K., Uhlig, M., Marais, E., Hoffmann, A., & ment and knowledge exchange. In U. Schmiedel & N. Zeller, U. (2005). Impact of different grazing sys- Jürgens (Eds.), Biodiversity in southern Africa. Vol- tems on diversity, abundance and biomass of beetles ume 2: Patterns and processes at regional scale (pp. (Coleoptera), a study from southern Namibia. Mit- 319–325). Göttingen: Hess. teilungen aus dem Museums für Naturkunde in Berlin, Schoeneberger, P. J., Wysocki, D. A., Benham, E. C., Zoologische Reihe, 81, 131–143. & Broderson, W. D. (2002). Field book for describ- Vollan, B. (2009). Co-operation for common pool re- ing and sampling soils. Version 2.0. Lincoln (NE): sources: An experimental perspective. München: Hut. USDA Natural Resources Conservation Service, Na- Vollan, B., Prediger, S., & Frölich, M. (2009). The influence tional Soil Survey Center. of collective property rights on grazing manage- Scholes, R. J., Mace, G. M., Turner, W., Geller, G. N., ment in a semi-arid region. http://www.escholarship. Jürgens, N., Larigaudrie, A., et al. (2008). Toward org/uc/item/8j9521t1. Accessed 10 April 2010. a global biodiversity observing system. Science, 321, Walter, H., & Breckle, S.-W. (1983). Ökologie der Erde— 1044–1045. Band 1: Ökologische Grundlagen in globaler Sicht. Shmida, A. (1984). Whittaker’s plant diversity sampling Stuttgart: Fischer. method. Israel Journal of Botany, 33, 41–46. Wiens, J. A. (1989). Spatial scaling in ecology. Functional Sommer, J. H., Kreft, H., Kier, G., Jetz, W., Mutke, J., Ecology, 3, 385–397. & Barthlott, W. (2010). Projected impacts of climate Wildi, O. (1986). Analyse vegetationskundlicher Daten— change on regional capacities for global plant species Theorie und Einsatz statistischer Methoden. richness. Proceedings of the Royal Society B-Biological Veröf fentlichung des Geobotanischen Institutes Science, 277, 2271–2280. der Eidgenössischen Technischen Hochschule, Stiftung Stohlgren, T. J. (2007). Measuring plant diversity—lessons Rübel in Zürich, 90. Zurich: Geobotanisches Institut, from the f ield. Oxford: Oxford University Press. ETH. Stohlgren, T. J., Falkner, M. B., & Schell, L. D. (1995). Wittig, R., König, K., Schmidt, M., & Szarzynski, J. (2007). A modified-Whittaker nested vegetation sampling A study of climate change and anthropogenic impacts method. Vegetatio, 117, 113–121. in West Africa. Environmental Science and Pollution Storch, D., Marquet, P. A., & Brown, J. H. (Eds.) (2007). Research, 14, 182–189. Scaling biodiversity. Cambridge: Cambridge Univer- World Resources Institute (Ed.) (2005). Ecosystem and hu- sity Press. man well-being: Biodiversity synthesis—A report of the Environ Monit Assess

Millennium Ecosystem Assessment. Washington, DC: Zedda, L., Köhler, T., & Rambold, G. (2008). The World Resources Institute. project BIOTA Southern Africa lichens: Meth- Yoccoz, N. G., Nichols, J. D., & Boulinier, T. (2001). Moni- ods. http://biota-africa.uni-bayreuth.de/wiki/BIOTA_ toring of biological diversity in space and time. Trends Lichens_meth. Accessed 24 March 2010. in Ecology and Evolution, 16, 446–453. Zedda, L., Gröngröft, A., Schultz, M., Petersen, A., Mills, Zedda, L., & Rambold, G. (2004). Diversity change of soil- A., & Rambold, G. (2011). Distribution patterns of growing lichens along a climate gradient in Southern soil lichens across different biomes of southern Africa. Africa. Bibliotheca Lichenologica, 88, 701–714. Journal of Arid Environments, 75, 215–220. Online Resource 1 of the following article:

The BIOTA Biodiversity Observatories in Africa – A standardized framework for large-scale environmental monitoring

submitted to Environmental Monitoring and Assessment Norbert Jürgens a, Ute Schmiedel a,*, Daniela H. Haarmeyer a, Jürgen Dengler a, Manfred Finckh a, Dethardt Goetze b, Alexander Gröngröft c, Karen Hahn-Hadjali d, Annick Koulibaly e, Jona Luther-Mosebach a, Gerhard Muche a, Jens Oldeland a, Andreas Petersen c,f, Stefan Porembski b, Michael C. Rutherford g,h, Marco Schmidt i, Brice Sinsin j, Ben J. Strohbach k, Adjima Thiombiano l, Rüdiger Wittig d, Georg Zizka i

* Corresponding author. Tel.: +49 40 42816 548; fax: +49 40 42816 539, email: [email protected]‐hamburg.de (U. Schmiedel)

Field(s) Notes

Latitude (°) WGS84; north-west corner of BIOTA Observatory Longitude (°) WGS84; north-west corner of BIOTA Observatory Mean annual Modelled data for 1950-2000; WorldClim Bioclim dataset BIO1; Source: Hijmans, R.J., Cameron, S.E., Parra, temperature (°C) J.L., Jones, P.G., Jarvis, A., 2005. Very high resolution interpolated climate surfaces for global land areas. Int. J. Climat. 25, 1965-1978.

Mean annual Modelled data for 1950-2000; WorldClim Bioclim dataset BIO12; Source: Hijmans, R.J., Cameron, S.E., Parra, precipitation (mm) J.L., Jones, P.G., Jarvis, A., 2005. Very high resolution interpolated climate surfaces for global land areas. Int. J. Climat. 25, 1965-1978.

Biome Classification within BIOTA AFRICA WWF ecoregion Source: Olson, D.M., Dinerstein, E., Wikramanayake, E.D., Burgess, N.D., Powell, G.V.N., Underwood, E.C., D’Amico, J.A., Itoua, I., Strand, H.E., Morrison, J.C., Loucks, C.J., Allnutt, T.F., Ricketts, T.H., Kura, Y., Lamoreux, J.F., Wettengel, W.W., Hedao, P., Kassem, K.R., 2001. Terrestrial ecoregions of the World: a new map of life on Earth. BioScience 51, 933–938.

Intensity of land Landuse intensity classe (none = no grazing or other land use; low = e.g. low-intensity grazing by game in a use nature reserve; medium = sustainable land use; high = degrading land use)

Years with x = data available in BIOTABase; (x) = data in process (will be available soon) vascular plant data

4 Sandveld Duruchaus S41 Narais S40 Claratal S39 Rooisand S38 Ganab S37 Gobabeb S36 Kleinberg S35 S34 BIOTA SouthernAfrica:East-WestTransect, Namibia Nieuwoudtville CapeofGoodHope S45 Elandsberg S33 Riverlands S32 Rocherpan S31 Moedverloren S29 Ratelgat S28 Goedehoop S27 S26 Paulshoek Soebatsfontein S24 Groot Derm S22 Numees S21 Koeroegap Vlakte S20 S18 Omano goNjamb Ogongo S43 Wlotzkasbaken S42 S16 Nabaos S12 Gellap Ost S11 Niko South S10 Niko North S09 Okamboro S08 Otjiamongombe S06 Toggekry S05 Sonop S04 Mutompo S03 Mile 46 S02 S01 BIOTA SouthernAfrica:North-SouthTransect,Namibia W07 W06 W05 W04 W03 W02 W01 BIOTA WestAfrica M02 M01 BIOTA Maroc 2 Remhoogte S25 S17 BIOTA SouthernAfrica:North-SouthTransect,SouthAfrica Observatory No. iieiBriaFs s oraKkdn 18310342342. 4 NorthSudanianSavanna SahelGrass-andShrubland 842 374 27.9 NorthSudanian Savanna 29.0 855 304 SahelGrass- andShrubland 295 SouthSudanianSavanna 364 1050 28.2 0.37442 27.0 -0.38439 258 29.0 220 286 14.54518 11.87341 0.51929 -3.83333 -0.38129 11.64648 8.76667 14.61480 Ibero-MauritanianSagebrushSteppe Saharan Semidesert 298 57 21.6 22.5 736 1869 -6.32585 -5.63236 31.38864 30.36741 - - - - - ParcNational delaComoé A Kikideni Natiabouani Kolel - Karios - Bouna Yomboli Gourma Gourma Oudalan Zanzan Oudalan IvoryCoast TizinTafilalet Taoujgalt Sahel Est BurkinaFaso Est BurkinaFaso BurkinaFaso Benin Sahel Comoé Ouarzazate Zagora BurkinaFaso Terroire Villageois Souss-Massa-Draâ Souss-Massa-Draâ A Natiabouani Morocco Kikideni Morocco Kolel Yomboli Taoujgalt El Miyit

lpha Benin libori Observatory name a aii rnoSaomn oae oae 2.34 5069492. 1NamibDesert 21 NamibDesert 16 21.0 17.6 419 188 15.04689 14.72518 -23.53241 -22.98984 SucculentKaroo 162 18.7 Fynbos 251 140 17.6 18.43906 35 SucculentKaroo 159 -31.46093 Gobabeb Kleinberg 19.1 18.30560 SucculentKaroo 239 -32.60085 162 NamaKaroo 98 NamaKaroo 18.60031 19.0 146 -31.28588 245 19.1 NamaKaroo NamaKaroo 20.8 6 238 153 NamaKaroo 909 238 18.59152 1045 SouthernAfrican WoodlandSavanna 16.3 536 19.5 20.6 WalfisBay Gobabeb -31.27671 17.99548 17.80396 19.5 83 1076 1099 22.2 -26.39064 -27.67652 1070 17.84897 18.00484 Swakopmund 1180 RocherpanNatureReserve Omaheke 18.39229 17.83909 Moedverloren208 Swakopmund -25.33341 -26.40120 19.25833 Khomas SouthernAfricanThornbushSavanna -34.26020 -25.34251 WalfisBay 394 Khomas -18.30178 SouthernAfricanThornbushSavanna Namibia SouthernAfricanWoodlandSavanna Khomas 396 497 Luiperskop 211 (Ratelgat) 19.2 Namibia Khomas Flaminkvalkte 111(Goedehoop) Namibia 1519 19.4 Erongo 21.8 Namibia Erongo 16.72933 1495 1236 CapeofGoodHope-Peninsula Namibia Erongo 16.93511 -21.50144 TableMountainNationalPark Namibia 18.90390 Piquetberg Karios-Fish RiverCanyon Namakw Namibia Vredendal -21.59683 -19.07382 CityofCapeTown Namibia VanRhynsdorp CapeWinelands VanRhynsdorp NorthernCape WestCoast WesternCape South Africa WestCoast WesternCape South Africa WestCoast Niko377 reserve WesternCape South Africa WestCoast Nabaos7 Niko377 grazing WesternCape South Africa WestCoast GellapOst3 WesternCape South Africa Karasburg WesternCape South Africa Namakw WesternCape South Africa Namakw Mutompo OtjiamongombeWest 44 South Africa OmatakoRanch305 Namakw NorthernCape Namakw NorthernCape Keetmanshoop South Africa nearGibeon Namakw NorthernCape Keetmanshoop South Africa nearGibeon Siyanda NorthernCape South Africa Karas NorthernCape South Africa Keetmanshoop Sonop903 NorthernCape South Africa Keetmanshoop South Africa Mariental nearCove Erichsfelde Mariental Omatako Namibia Omusati Karas Namibia Omusati nearMaroelaboom Karas near Namibia Erongo Karas Namibia Okahandja Hardap Namibia Cove Okahandja Hardap Rundu Okahandja Namibia Grootfontein Namibia Otjozondjupa Namibia Otjozondjupa Namibia Otjozondjupa Rundu Otjozondjupa Namibia Kavango Namibia Namibia Namibia Namibia Kavango ot fiaNrhr aeNamakw NorthernCape South Africa Country A A

aoaPhnoPhno-1.46 .46 2 6412 SouthSudanianSavanna 1121 SouthSudanianSavanna 1085 26.4 27.0 327 297 2.24961 2.24676 10.14469 10.41762 - - Péhunco Péhunco Péhunco Péhunco tacora tacora Primary administrative unit idokWindhoek Rooisand DesertRanch Gobabis Windhoek Windhoek Windhoek Windhoek OmanogoNjamb Ogongo Ogongo Uutapi Uutapi A

randis Secondary administrative unit a a a a a a a erRhbt aas-31081.95 641. 8 NamaKaroo 289 SouthernAfricanThornbushSavanna NamaKaroo 397 287 18.6 19.1 18.6 1624 1523 1614 16.89658 19.13390 16.90000 -23.12038 -22.04335 -23.13318 Dipcadi389 Duruchaus Narais near Drimiopsis near Rehoboth near Rehoboth edlnsrfNme 2.90 6931321. 7SucculentKaroo 57 17.6 362 SucculentKaroo 73 16.95381 -28.29309 17.0 635 17.02567 -28.22665 NamibDesert 12 Namib-Naukluft Park 16.9 73 14.45948 Nieuwoudtville (Paulshoek) Leliefontein624 -22.31500 Wellington Malmesbury Numees KoeroegapVlakte Leliefontein Leliefontein Soebatsfontein A Wlotzkasbaken-Mile14 Sendelingsdrif Sendelingsdrif A Swakopmund Ovitoto

eadrByGotDr elwDn 2.13 6643131. 7SucculentKaroo 57 17.6 193 16.65433 -28.61234 GrootDerm-Yellow Dune lexander Bay skham Locality lrtl1 2.73 674516 8035SouthernAfricanThornbushSavanna 335 SouthernAfricanThornbushSavanna 200 18.0 19.5 1865 1160 16.77485 16.10498 -22.77938 -23.29454 Claratal 18 Chausib 27 Southern AfricanWoodlandSavanna 469 22.7 1103 15.30521 -17.69720 Ogongo AgriculturalCollege A aa 2.28 5584951. 2 Namib Desert 129 19.5 995 15.53844 Fynbos 301 -23.12182 16.3 Fynbos 560 722 Fynbos 453 18.0 19.12880 95 16.8 -31.39002 140 19.03111 18.58012 33.41416 -33.48934 SucculentKaroo 247 15.5 Succulent Karoo 146 Ganab 1027 17.0 18.27570 392 -30.39536 17.54337 Glenlyon -30.18650 Elandsberg Riverlands SouthernAfricanThornbushSavanna 398 Remhoogte 416(Rooiwal) 19.7 Soebatsfontein Commonage 1490 17.06248 -22.01918 A Okamboro Ovitoto55 e oeg ietc eeomn ete-83131.43 102. 3 SouthernAfricanWoodlandSavanna 536 22.2 1180 19.24731 -18.30183 lex MorengaLivestockDevelopmentCentre pa-67022.10 9 0115Southern AfricanThornbushSavanna 185 20.1 896 20.61402 -26.76092 lpha Farm name and number a

Latitude (°) 3.86 825914 5322SucculentKaroo 252 15.3 1048 18.27579 -30.38568 SouthernAfricanWoodland Savanna 467 22.7 1100 15.26480 -17.70803

Longitude (°)

Altitude (m a.s.l.)

Mean annual temperature (°C)

Mean annual

9Fynbos 89 precipitation (mm)

Biome A A A A A A A onlyexplorativesoilprofiles A - A onlyexplorativesoilprof - A A A A A A A A A A A A A A A A A A rangelandforsheep,goatsanddromedaries A A rangelandfordromedariesandgoats A high collective high collective PA1214 –Mediterraneanwoodlandsandforests PA1321 –NorthSaharansteppeandwoodlands A A A A A A A A A A A A A A

11 aiinsvnawolnspiaemdu atefrigadtuim- Experimentalfarming withcattle medium farmingwithcattleandgoats, somegame naturereserve farmingwithcattle, sheep andgoats naturereserve medium none low state Farmingwithcattle and game none medium private private state naturereserve private nature reserve state Cattlefarmingandtourism none medium none private - naturereserve state T1309 –Kalaharixericsavanna state - T1309 –Kalaharixericsavanna T1309 –Kalaharixericsavanna low nature reserve T1309 –Kalaharixericsavanna farmingwithsheep,fewhorses T1316 –Namibiansavannawoodlands farmingwithsheep naturereserve farmingwithsheep none naturereserve medium state T1316 –Namibiansavannawoodlands medium naturereserve low communal T1315 –Namibdesert farmingwithgoatsandsheep none naturereserve none private T1315 –Namibdesert state none private high private farmingwithsheepand goats low communal medium state T1322 –SucculentKaroo communal state farmingwithcattle,sheep andgoats T1203 –Montanefynbosandrenosterveld private medium farmingwithcattle,sheep andgoats T1202 –Lowlandfynbosandrenosterveld communal medium T1202 –Lowlandfynbosandrenosterveld recreatationarea communal farmingwithspringbokandoryx naturereserve,game farming T1202 –Lowlandfynbosandrenosterveld medium none T1322 –SucculentKaroo T1322 –SucculentKaroo low - farmingwithcattleandsmallstock private T1322 –SucculentKaroo Demonstrationfarming withcattle state T1322 –SucculentKaroo private high T1322 –SucculentKaroo low T1322 –SucculentKaroo T1322 –SucculentKaroo state T1322 –SucculentKaroo state T1322 –SucculentKaroo farmingwithgoatsandsheep experimentalfarmingandbreedingwithsheepgoats T1309 –Kalaharixericsavanna medium farmingwithcattle,goatsandsheep high farmingwithcattleand gameforhunting communal medium farmingwithgame for hunting,cattlefarming high T0702 –AngolanMopanewoodlands rotationalfarmingwith goats state medium communal medium T0702 –AngolanMopanewoodlands rotationalfarmingwith goats communal private medium T1315 –Namibdesert communal private T1314 –NamaKaroo T1316 –Namibiansavannawoodlands protectedsince1926,annualsavannaburningandpoaching breedingexperimentsandstandardrotationalgrazing,farmingwithgoats andcattl T1316 –Namibiansavannawoodlands experimentalfarmingwithcattle,game T1316 –Namibiansavannawoodlands medium high low farmingwithcattle T1316 –Namibiansavannawoodlands medium protected T1309 –Kalaharixericsavanna communal medium protected pasture state T1309 –Kalaharixericsavanna medium state state medium communal T1309 –Kalaharixericsavanna T0709 –KalahariAcacia-Baikiaeawoodlands state pasture state T0726 –ZambezianBaikiaeawoodlands medium pasture communal T0726 –ZambezianBaikiaeawoodlands medium pasture communal medium communal T0713 –SahelianAcaciasavanna T0722 –WestSudaniansavanna T0713 –SahelianAcaciasavanna T0722 –WestSudaniansavanna T0722 –WestSudaniansavanna T0722 –WestSudaniansavanna T0722 –WestSudaniansavanna WWF ecoregion

Tenure

Intensity of land use

Type and history of land use e x x x x x x x x x x x x x x x - - - - x x - - - x x x x x x x x x x x x x x - - Soil data plus ranks30,50 plus ranks39 plus ranks26-40,70 plus transect(ranks51,55,57,59,61) only ranks1-15,plustransect(ranks20,22,31,90) only ranks1-15,plustransect(ranks19,24,25,32,76,79,88,90) with differentsamplingdesign with differentsamplingdesign with differentsamplingdesign with differentsamplingdesign plus ranks27,29 only ranks1-15 plus rank65 only ranks1-24 soil sam plus ranks26,68,69 Deviations from standard soil

p sampling les arefromthenon-establishedBIOTAObservator iles y S13nearb y 0420 ------2005, 2008,2009 2004-2009 - 2004-2009 2005, 2007,2009 - - 2010 2007 2008, 2009 2009 2004, 2008,2009 2005 2001-2009 2001-2009 2001-2009 2001-2009 - 2001-2009 2001-2006, 2008 - 2001-2003, 2006 - - 2001-2009 - - 2002, 2003,2005-2007 - 2007-2009 2007-2009 2001-2003, 2005,2010 2001-2009 2001-2009 - 2001-2009 2001, 2002,2004,2006-2009 2001, 2002,2004,2006-2009 2004-2009 2004-2009 2001-2009 2001-2003, 2005,2006,2008,2009 2001-2003, 2005-2009 2001-2003, 2005-2009 2002 - 2001-2003 - 2001, 2003 2001 2001, 2002 2002-2005, 2008,2009 2002, 2003,2005-2009 04--- - - 2005 2010 2004 2001-2009 - 2001 2002 Years with vascular plant data ------(x)------x x x x x x x x x ------x x x x x x x x x x x x x x x ------(x) 2001 x x x x x x x x x x x x x x x x x x x x ------x x x ------x x x Years withvascularplantdata 2002 ------x x x x x x x x x x x x x x x x - - x x x x ------x x 2003 x x x x x x x x x x x -- x - - x x x x x x x x ------x - 2004 x x x x x x x ------x x x x x x x x x x x x x x - - x x x x x x 2005 -- x x ------x x x x x x x x x x ----(x)x x x x x x x x x x x -- x 2006 x x x x x x x x x - ---- x x x x x x x x x x x x - x x x 2007 x x x x x x x - (x) (x) --- x x ( x x x x x x x x x x x x x x x - x x x

) 2008 x x x x x x x x x x x x x x x -- x x x x x x x x x x x x x 2009 - - - - x - - - ranks1-10 - ranks1-9,30 ------ranks1-40;without1000m² ranks1-40;without1000m² - - 2010 ranks 1,2,5,8on100m2 ranks 1-15 ranks 1-10 ranks 1-10 ranks 1-10 ranks 1-10 ranks 1-100,only10,000m² ranks 1-10 ranks 1-10 Deviations from standard vascular plant sampling