Ecological process indicators used for nature protection scenarious in agricultural landscapes of SW

This article is also available online at:www.elsevier.com/locate/ecolind

Ecological Indicators 7 (2007) 396–411

Ecological process indicators used for nature protection

scenarios in agricultural landscapes of SW Norway

Annette Bar a,b, Jorg Loffler a,*

a University of Bonn, Department of Geography, Germanyb University of Oldenburg, Institute of Biology and Environmental Sciences, Germany

Received 1 November 2005; received in revised form 5 April 2006; accepted 6 April 2006

Abstract

Conflicts between changing landscapes and static nature protection concepts were addressed as an example of the

agricultural landscape of SW Norway. We aimed to deduce indicators for spatio-temporal landscape changes to draw scenarios

for future protection perspectives of a RAMSAR and nature reserve. To estimate the variability of bird diversity, changes in

vegetation patterns were analysed to predict bird occurrence. We obtained a differentiated analysis of present landscape

dynamics by measuring landscape structure, vegetation, hydrology and nutrient concentration. Multivariate statistics were used

to extract the main driving forces for changes in vegetation patterns out of a complex landscape ecological data set.

Subsequently, we compared the measured data with those of past landscape stages to determine landscape changes and their

mechanisms at different spatio-temporal scales. Ecological process indicators (EPI) were derived, and three different indicator

constellations were used for scenario descriptions. These scenarios were chosen as to the current assumptions of typical

contrasting nature protection strategies. Concluding, we used EPIs to evaluate nature protection aims and to assess scenarios of

changing landscapes. This approach will be transferable to other examples of nature protection conflicts in the agricultural

landscape in general.

# 2006 Elsevier Ltd. All rights reserved.

Keywords: Biodiversity; Landscape dynamics; Landscape management; Landscape metrics; RAMSAR reserve; Sustainable development;

Vegetation mapping; Wetland ecosystems

1. Introduction

Cultural landscapes show spatio-temporal changes

of landscape structures according to intensive human

impact and natural succession (Lundberg, 2000;

* Corresponding author. Tel.: +49 228 73 7239.

E-mail address: [email protected] (J. Loffler).

1470-160X/$ – see front matter # 2006 Elsevier Ltd. All rights reserved

doi:10.1016/j.ecolind.2006.04.001

Waldhardt et al., 2004). Thus, current discussions

about changing landscapes and process-oriented

protection strategies (Pickett et al., 1992; Lundberg,

1996; Skanes, 1997) have to consider the question of

appropriate protection. As an example of the

agricultural landscape of SW Norway, we chose a

study area at Lake Grudevatn (Jæren). High diversity

in plant and bird life led to the designation of the site as

.

mailto:[email protected]

http://dx.doi.org/10.1016/j.ecolind.2006.04.001

A. Bar, J. Loffler / Ecological Indicators 7 (2007) 396–411 397

nature reserve in 1974 and as a RAMSAR reserve

(convention on wetlands signed in Ramsar, Iran 1971)

in 1985 (Wetlands International, 2004). High plant and

bird diversity as a reason for protection was not further

specified in the protection aims, so it remained unclear

what kind of species composition was desired:

a species composition of traditional agricultural

landscapes, in coincident with wetland protection

(RAMSAR), a species composition of minimal

anthropogenic disturbed areas, or a mosaic of both

which favors high total species numbers but of more

common species. Consequently, nor monitoring of

plant and bird species occurrence neither landscape

management measures have been carried out to

observe landscape changes or to maintain a desired

status of species composition within the protected

area. Thus, the landscape character has strongly

changed since 1974 caused mainly by shifts in

distribution and composition of vegetation units.

Due to the lack of specified protection aims and

continuing monitoring data the challenge was to use

the approach of ‘‘space-for-time substitution’’ (Pickett

et al., 1992). This approach should be combined with

available external data about past landscape stages in

order to determine ongoing and past changes. Our

study was framed by a general increasing interest in

deducing indicators to monitor environmental change

(Tiner, 2004). Our approach was focused on devel-

oping ecological indicators that represented different

vegetation transforming processes at different spatial

scales and time perspectives, hereafter referred to as

ecological process indicators (EPI). In detail, these

indicators should be extracted by mapping recent

vegetation patterns and measuring corresponding

hydrological and nutritional conditions. The structural

status quo of the landscape should then be compared

with past stages of spatial vegetation succession over

the past decades. As it is known that bird occurrence,

abundance and habitat selection is strongly influenced

by vegetation structures (Cody, 1987; Jones, 2001;

Tonu et al., 2005), the information about changes in

vegetation patterns should be used to estimate the

variability of bird diversity.

Using the derived EPIs to adjust different variable

constellations we aimed to suggest scenarios to

illustrate landscape changes and possible future

developments with respect to their relevance for bird

life at Lake Grudevatn. We intended to focus on major

principles of medium- and long-term landscape

changes; precise predictions of short-term configura-

tions of species and species numbers were not

intended. Our sub-goals were:

(a) t
o test a combination of multi-scale environmental
variables and methods, including a short inves-

tigation period combined with different historic

data sources in order to determine the complex of

medium-term and long-term landscape transform-

ing processes,

(b) t
o use vegetation changes as predictor for bird
diversity, and

(c) t
o derive EPIs for scenarios of landscape
development in a RAMSAR and nature reserve

by addressing conflicts between changing land-

scapes and static nature protection concepts.

2. Material and methods

Field work was carried out at Lake Grudevatn, Klepp

community (Jæren, SW Norway). The wetland is

protected as bird and nature reserve, covering 110 and

47 ha of land, respectively, and is surrounded by

intensively used arable land and pasture resulting in

considerable nutrient supply caused by drainage ditches

(Molværsmyr et al., 1989; Molværsmyr, 1990). The

water table shows high oscillations and flooding occurs

regularly. During the 20th century substantial anthro-

pogenic changes were caused by drainage measures in

order to extend the amount of agricultural land. Water

table lowering and an increased nutrient supply caused

rapid vegetation succession and accelerated sedimenta-

tion at the lake shores (Olafsrud, 1993).

Fig. 1 describes our methodological approach at

different spatio-temporal scales. The present status

quo of the bird and nature reserve was investigated

between May and August 2002. Vegetation was

mapped at 115 single square plots (5 m � 5 m)

aggregated to 18 transects along different topographic

and moisture gradients around the lake. The distance

between the plots was determined for each transect

according to the change of plant species compositions.

Plant species abundance was recorded using the

Domin-scale (Kent and Coker, 1992). On the micro-

scale, vegetation types were defined after Fremstad

(1997) whose scheme was often used as a national

A. Bar, J. Loffler / Ecological Indicators 7 (2007) 396–411398

Fig. 1. Methodological approach used in the study at different spatio-temporal scales.

reference in Norway. For each vegetation plot one soil

sample was collected. Soil profile features, soil

moisture, and soil structure were recorded after Boden

(1996). The pH was measured at 10, 20 and 30 cm

depth in H2O and CaCl2 as a volumetric sample-liquid

proportion of 1:2.5 in the field. Additionally, type and

degree of land use (e.g. kind of grazing animals) were

mapped and classified into five utilisation intensities

per grazing type. Classification was done by estima-

tion and internal ranking of trampling and browsing.

Hydrological investigations were carried out by

measuring the water table variations at 17 sites

weekly using tide gauges. At these points conductivity

was determined within the same intervals. Due to the

spatial array of measuring points in transects at

different vegetation units around the lake it was

possible to analyse nutrient fluxes subjected to

structural site conditions and hydrological regimes.

Changes in vegetation patterns were studied over a

period of 65 years. Historic data sources were used

such as aerial photos, vegetation type maps and long-

term measurements of hydrology and nutrient con-

centrations in the catchment area (Table 1). On the

meso-scale, land use units were classified which were

based on aerial photos from 1937, 1954 and 1989 and

GIS-based maps were derived for each year. These

maps were based on land use units and provide

information about landscape structures focusing on

both the reserve and its surroundings. Land use units

consisted of classified single patches characterized by

the same land use type. Based on these, landscape

metrics enabled a quantification of landscape struc-

tures (McGarigal and Marks, 1994). By comparing the

same area using three time sequences of aerial photos

structural landscape changes were studied. ‘‘Mean

patch size’’ (MPS) and ‘‘mean nearest neighbour’’

(MNN) were determined using the ArcView extension

‘‘patch 2.0’’. Landscape changes were expressed by

e.g. increasing/decreasing MPS for each year and land

use category. The MPS reflected changes especially of

the agricultural landscapes quite well due to distinct

land use borders (McGarigal and Marks, 1994) and

indicated landscape uniformity. Wetland connected-

ness which is important for migratory birds was

analysed by MNN distance calculations between

patches of the same land use type.

Finally, our newly produced vegetation type map

from 2002 was compared with the one from 1992

(Olafsrud, 1993) allowing to study qualitative changes

over a medium-term period. This comparison pro-

vided the option to differentiate vegetation changes

within the protection area in addition to the aerial

photo analyses with a focus on land use changes. For

argumentation concerning vegetation changes we


Table 1

Measured variables and external environmental data sources

Environmental

variables

Methods External sources References

Land use GIS-based analysis Aerial photos from 1937,

1954, and 1989

Fjellanger Widerøe (1937, 1954, 1989),

and personal communication with land

owners and administration authorities

Vegetation Vegetation mapping Vegetation type map 1992 Olafsrud (1993)

Soils Soil profile description,

pH and soil moisture

measurements

Soil type map 1962 Semb (1962), Boden (1996)

Hydrology Water table measurements

(wells and gauges)

Water table measurement

(Figgjo-River 1984–2002)

Molværsmyr et al. (1989), Molværsmyr

(1990), Fylkesmannen i Rogaland (1996),

Vikse (1998)

Nutrients Water conductivity

measurements (wells and

gauges)

Nitrogen and phosphor

concentration (1984–2002)

Fylkesmannen i Rogaland (1996), Vikse

(1998), Solberg (personal communication)

Breeding birds Reports, personal communication Løvbrekke (1992a,b, 1995, unpublished data)

referred to Nedkvitne et al. (1995), Pott (1995),

Fremstad (1997), Pott and Remy (2000) and Succow

and Joosten (2001).

External data of nutrient amounts and water table

fluctuations from the Figgjo-River were used to set the

current measurements into a longer time perspective

(K. Solberg, unpublished data).

For data on breeding bird communities in the study

area, we could not rely on own fieldwork results.

Species composition, species abundance and trends in

population development between 1979 and 2002

could coarsely be characterised evaluating Løvbrekke

(1992a,b, 1995) and using comments of local

ornithologists (Løvbrekke, unpublished data and

personal communication, 2002). Following the

‘‘guild’’ concept (Simberloff and Dayan, 1991;

Wilson, 1999), a set of ‘‘typical’’ species was assigned

to the following vegetation/land use units each:

eutroph lake and lakeshore, reed, fen and mesophile

grassland, arable land, and willow shrub and similar

early succession stages.

Since there is no accepted guild concept for

northern European bird communities, we used the

species assignment by Flade (1994) developed for

northern German vegetation units, but adapted it to

local conditions by adding/removing species to guilds

based on our own field experience. Hence, our

modification lacks the strictly statistical approach

but does reflect local circumstances much better.

Prediction of occurrence and population trends was

based on the current bird fauna situation and trends

between 1992 and 2002. Species treatment was based

on Clements (2000) inclusive all supplements to-date.

In order to extract major vegetation patterns in

relation to environmental variables, we applied

multivariate ordination techniques (Ter Braak and

Prentice, 1988; Ter Braak and Smilauer, 1998). Plant

species abundance was used in combination with

nominal data for structural environmental variables

and different quantified environmental variables. The

main environmental variables were extracted out of a

set of 24 environmental variables based on their

significance and colinearity. A Canonical Correspon-

dence Analysis (CCA; Ter Braak, 1986) was applied

on 118 species. The position of the species represented

their relation to the involved ecological gradients (Ter

Braak, 1994; Jongman et al., 1995).

Based on ordination vegetation complexes were

delimited by differences in soil moisture, nutrient

supply and land use. Characteristic plant species could

be assigned. A schematic section along a gradient

from the lake to arable land illustrates the recent

ecological conditions based on the extracted environ-

mental variables for the nature reserve. We used data

about past landscape stages concerning landscape

structure, vegetation, hydrology and nutrient concen-

tration and combined them with our results of recent

landscape patterns and functioning. Thus, we were

able to determine landscape changes and their

mechanisms on different spatio-temporal scales to


determine EPIs. These were interpreted as the main

driving forces for vegetation changes, and thus for

changes in species composition of breeding birds.

Vegetation succession was used to visualize the

tendency of vegetation changes by EPIs.

For future protection perspectives three scenarios

were created distinguishing different configurations of

the EPIs. Three contrasting scenarios were chosen in

agreement with the most widespread current assump-

tions of nature protection strategies: (1) emphasis on

protection of natural succession, (2) emphasis on the

preservation of a static cultural landscape for the

protection of certain species including extensive

conservation measures, and (3) a mixture of both

approaches. Based on our temporal analyses land-

scape changes will occur at different temporal scales

and anthropogenic changes can be gradual e.g.

nutrient accumulation and abrupt e.g. land use

transformation. Causes and consequences of manage-

ment measures are different for several processes,

such as for plant species composition and vegetation

type changes up to 10 years contrasting to sedimenta-

tion processes lasting more than 10 years. Thus, the

time horizon for described scenario perspectives

depends on the processes. We illustrated different

effects of process indication:

� S
cenario 1 describes future development based on
the status quo. Applying this scenario, all current

transforming processes will continue the trends

observed during the last decades concerning both

stronger natural vegetation succession and the

strengthened impact of intensified and extended

agricultural use. In reality, this scenario often

results from conflicts in land use (nature protection

versus agricultural use) and a lack of landscape

management initiative. However, it can be desirable

under the perspective of mosaic-cycle protection

aims (Remmert et al., 1991). In cultural landscapes

it is not suitable to preserve current plant and bird

diversity in many cases.

� S
cenario 2 was geared to certain landscape
management interventions as well as to landscape

conservation targets. This scenario is a realistic

option in the implementation of management

measures concerning preservation of a certain

stage, and maintaining the current plant and bird

species diversity. Some of the measures supporting

this scenario were already partly introduced during

the campaign for an improvement of water quality

in Jæren (Vikse, 1998).

� T
he landscape management strategy suggested in
scenario 3 postulates substantial changes in land

cultivation. This scenario describes the most fictive

perspective for the area. There is no consensus for a

transition of intensive land use into an economically

less productive type under the current socio-

agricultural situation. In fact, this nature protection

strategy would be the most substantial one

concerning preservation of current plant and bird

diversity and the improvement of the characteristic

traditional landscape.

3. Results

On the meso-scale, aerial photo analyses showed

continuing processes of landscape changes between

past stages of the study area from 1937 to 1954 and to

1989 (Fig. 2). Especially arable land expanded

drastically where areas could be used for agricultural

purposes. The area of arable land in total was enlarged

from 25.5 ha in 1937 to 77.5 ha in 1954 and to 161 ha

in 1989. The transformation of grassland into arable

land took place in two different steps. While arable

land emerged from grassland within the same small

parcels of land at the first time sequence (from 1937 to

1954), the managed units did not persist from 1954 to

1989 any longer. During the second period, arable land

extended into areas without former agricultural use

and was suited to efficient mechanical soil cultivation

in size and shape. The increase in MPS of arable land

reflects kinds of changes due to more efficient

agricultural management on bigger and equalised

land units. Increase of fen-MPS and decrease in fen-

MNN at the same time is a result of the loss of smaller

wetland areas at the expense of arable land. MPS of

grassland and forest decreased in total area, drastically

between 1954 and 1989. Apart from the tendency

towards reduction in general, the number of patches

increased continuously while MNN between forest

patches decreased. This illustrates that widespread

forest were under fragmentation into smaller units.

Currently, all described processes of land transforma-

tion are still ongoing (own observations in 2002, and

personal communication with authorities and farmers)


Fig. 2. Spatio-temporal distribution of land use units and configuration of landscape structures in 1937, 1954 and 1989 analysed on the basis of

aerial photos. Analysis of landscape structure changes was based on landscape metrics mean patch size (MPS) and mean nearest neighbor

(MNN) (McGarigal and Marks, 1994) for the years 1937, 1954 and 1989. Results were differentiated for each land use unit (patch level) and year

(modified after Bar et al., 2004).

even causing more pronounced transition zone

between intensively used agricultural land and

remaining protected wetlands.

On the micro-scale, natural processes caused

vegetation succession e.g. in Kvernebekken bay first

and foremost combined with sedimentation (Fig. 3).

This was forced by eutrophic site conditions. The

vegetation type maps provide more detailed informa-

tion about these changes for the period between 1992

and 2002. Especially in bays, changes ran fast and reed

vegetation expanded remarkably. Consequently, spe-

cies-poor reedbeds dominated by Phragmites australis

increased in area and currently dominate the

sedimentation zones. Growth of bouncing lawns

strengthened the fast sedimentation in the bays. Along

streaming exposed sites riparian vegetation was built

by dense Schoenoplectus lacustris stands. Between

1992 and 2002 the fen vegetation belt changed in plant

species composition due to higher nutrient concentra-

tions. These areas were invaded by Phalaris arundi-

nacea to a large extent. Vegetation formerly either

characterized by sedge species such as Carex rostrata

or mosaics of Potentilla palustris, Galium palustre,

Lysimachia thyrsiflora, and Myosotis palustris turned


Fig. 3. Spatial distribution of vegetation types presented for the Kvernebekken-bay. The maps are based on vegetation mapping in 1992 by

Olafsrud (1993) and in 2002 by Bar et al. (2004). Vegetation types were recorded after Fremstad (1997) (modified after Bar et al., 2004).

into species-poor patches. The fast and widespread

expansion of Salix-shrubs and tall forbs along the

transition zones between the organic and mineral soils

was remarkable.

Measurements of the hydrologic variability and the

nutrient dynamics gave more detailed information

about the current spatio-temporal dynamics within the

nature reserve (Fig. 4). The groundwater table in the

present fen areas proved to be permanently high.

Oscillation was mainly caused by flooding. Yet, fen

edge and mineral soils were characterised by

fluctuations of the groundwater table. In comparison

with long-term water table measurements of the

Figgjo-River over the past 20 years amplitudes

increased since the 1980s. This indicates more

frequent and intensive flooding than in former times.

One reason is land transformation into arable land in

the entire catchment which caused a reduced water

storage capacity and led directly to a run-off and thus

to more intensive flooding.


Fig. 4. Results of water table and conductivity variations based on weekly measurements between 27th May and 21st August 2002 are presented

in the diagrams. Conductivity concentration is illustrated in columns while the water level is given as lines. The diagrams represent ground water

measurements (B) including two measuring points at different distances to the waterside. The data gap in the middle of the period was caused by

extremely high flooding events when measuring was not possible anymore. Site variables as moisture and pH values at different depth as well as

soil type and grazing were recorded once within the investigation period in order to characterize site conditions (modified after Bar et al., 2004).

The nutrient regime was closely connected with

hydrologic conditions (Fig. 4). Current measurements

showed that large quantities of nutrients were

transported from canals into the area and the

groundwater was permanently nutrient enriched.

Agricultural land is well drained so that precipitation

water could run-off to lower parts, e.g. the fen areas.

Measured conductivity in ground water tables strongly

depended on the current hydrologic situation and on

the distance between river bank and arable land. A low

ground water table caused higher nutrient concentra-

tion close to the riverbank. This situation reversed

when the water table rises close to or above surface.

Sites close to the riverbank were flooded which had a

thinning effect on nutrient concentrations. As a result

measured conductivity was higher at the upper sites

with greater distance to the river bank. The

conductivity of soils in bays hardly varied as to lake

shore distance, but was generally found to be very

high. No streaming potentiated the fact that nutrient-

enriched sediments could depose over the preceding

years causing eutrophic site conditions. Long-term

measurements showed that nitrogen concentration

hardly decreased since the 1980s. Phosphorus con-

centrations were high until 2000, since then a distinct

decrease was observed (Molværsmyr, 1990; K.


Solberg, personal communication, 2002). Nitrophile

riparian vegetation spread along the riverbank. Dense

stands of tall herbs reduced the hydraulic stream flow

capacity, resulting in backwaters and thus intensify

flooding.

The hydrologic-nutritional situation in combina-

tion with grazing pressure showed direct impact on the

establishment of shrubs and forbs as well. These

benefited from a good nutrient supply caused by a

lower groundwater table which allows organic matter

to be mineralised. In addition, trampling of grazing

animals as mainly cattle caused severe damages

shaping an undulating surface and opened ground

patches. At such patches, germination is facilitated,

thus spreading and growth of shrubs and forbs was

enforced. Intensity of agricultural land use present as

Fig. 5. Results of a canonical correspondence analysis (CCA) for 118 plan

(shown as arrows and quadrats) were extracted. Based on these variables, c

intensive grazing within the protection area and high

nutrient supply from fertilisation mainly from outside

the reserve were figured out to be the main driving

forces for the changes.

Six main vegetation complexes could be classified

based on variable constellations and ordination

(Fig. 5). The ordination diagram shows how recent

plant compositions were determined by the environ-

mental variables. Plant species classified as reed/lake

shore vegetation (2) were more influenced by nitrogen

than the fen plants species (3). The lake shore complex

represents sites of reed vegetation that receives

nutrients from the river and canals. Salix species

(marked in black triangles) mainly occurred along the

transition zone (4) between the fen area (3) and pasture

land and illustrated the invasion of shrubs in fen areas.

t species and 24 environmental variables. Major ecological variables

urrent vegetation patterns were classified into vegetation complexes.


The two grassland complexes (5a and 5b) strongly

differed in their dependence on nutrient supply and

grazing pressure. Arable land (6) presented the sixth

complex. For each vegetation complexes representa-

tive plant species are shown. The impact of all

investigated environmental variables were aggregated

in a schematic landscape ecological concept (Fig. 6).

Analysis of bird mapping between 1979 and 2002

led to the following conclusions (Table 2). Bird

species of fens and extensive grassland (‘‘meadow-

birds’’ cf. Beintema et al., 1995) experienced a large

decline. Out of 14 species assigned to this guild, 3

(Black Grouse Tetrao tetrix, Corncrake Crex crex,

Short-eared Owl Asio flammeus) were extinct already

in 1979, a further five species (Dunlin Calidris alpine,

Black-tailed Godwit Limosa limosa, Ruff Philoma-

chus pugnax, Yellow Wagtail Motacilla f. flava and

Whinchat Saxicola rubetra) disappeared until 2002.

Numbers of Common Snipe Gallinago gallinago and

Eurasian Curlew Numenius arquatus stayed stable

(eight and four pairs in 2002, respectively), on

Redshank Tringa totanus and Northern Lapwing

Fig. 6. The impacts of all investigated environmental variables were aggreg

environmental variables are presented below. Moisture and nutrient gradi

Vanellus vanellus, no information is available. Data

on birds breeding at the lakeshore (Great Crested

Podiceps cristatus and Little Grebe Tachybabtus

ruficollis, different dabbling ducks Anatinae, Coot

Fulica atra) is deficient, but probably no severe

declines took place, except in Garganey Anas

querquedula. Gull counts in 1992 and 2002 showed

a strong decrease in Black-headed Gull Larus

ridibundus numbers (125–150 pairs to 3 pairs), and

a marked increase in Common Gull Larus canus

numbers (2–5 pairs to 10–24 pairs). Species diversity

and numbers of breeding pairs increased in bird

communities characteristic for early succession stages

(willow shrub). Chiffchaff Phylloscopus collybita was

first recorded breeding in 1992, and numbers of this

species increased strongly until 2002. However,

numbers of further species indicating succession

processes stayed stable (e.g. Reed Bunting Emberiza

schoeniclus) or declined (Greater Whitethroat Sylvia

communis).

Based on ecological measurements in comparison

with former stages of the study area EPIs were derived

ated in a schematic landscape ecological concept. Major gradients of

ents were related to near-surface conditions.


Table 2

Main vegetation units, corresponding bird guilds and trends in species population development 1979–2002

Vegetation unit/land use type Species Trend 1979–2002

Eutroph lake and lakeshore

Little Grebe Tachybaptus ruficollis ?

Mute Swan Cygnus olor 0

Eurasian Wigeon Anas penelope ?

Garganey Anas querquedula ?

Northern Shoveler Anas clypeata ?

Common Gull Larus canus +

Black-headed Gull Larus ridibundus �

Reeds

Water Rail Rallus aquaticus ?

Sedge Warbler Acrocephalus schoenobaenus ?

European Reed Warbler Acrocephalus scirpaceus ?

Fen and mesophile grassland

Black Grouse Tetrao tetrix ��Corncrake Crex crex ��Northern Lapwing Vanellus vanellus ?

Dunlin Calidris alpina ��Common Snipe Gallinago gallinago 0

Eurasian Curlew Numenius arquata 0

Black-tailed Godwit Limosa limosa ��Common Redshank Tringa totanus ?

Ruff Philomachus pugnax ��Short-eared Owl Asio flammeus ��Eurasian Skylark Alauda arvensis ?

Yellow Wagtail Motacilla flava flava �Northern Wheatear Oenanthe oenanthe 0

Whinchat Saxicola rubetra ��

Arable land

Northern Lapwing Vanellus vanellus ?

Yellowhammer Emberiza citrinella ++

Willow shrub and earlier successions

Greater Whitethroat Sylvia communis ��Willow Warbler Phylloscopus trochilus 0

Chiffchaff Phylloscopus collybita ?

Reed Bunting Emberiza schoeniclus 0

��: strong decline; �: decline; 0: numbers stable; +: increase; ++: strong increase; ?: no population numbers available.

(Fig. 7). These were defined as the main driving

forces. Ecological process indication for the lake shore

areas was sedimentation in bays which led to nutrient

accumulation as to nutrient enriched sediments. The

decrease of the water storage capacity due to land use

transformation into arable land and their drainage led

to more frequent and intensive flooding. In addition,

nitrophile dense riparian vegetation spread and forced

flooding intensity by backwater. For vegetation

succession this meant that nitrophile reed vegetation

characterized by species-poor high and dense stands

expanded. Accordingly, population numbers of reed

inhabiting bird species such as Sedge Warbler

Acrocephalus schoenobaenus and Eurasian Reed

Warbler Acrocephalus scirpaceus might have

increased. Depending on the availability of mudflats

and reedbeds in shallow water, Water Rail Rallus

aquaticus might have become a regular breeder.

Nutrient accumulation was also evident in the fen

complex caused by changing flooding regimes and

nutrient supply from surrounding agriculture land and

as consequence of former water table lowering. Thus,

mineralization processes started, additionally forced

by trampling of grazing animals. The gaps in


Fig. 7. Based on the landscape scheme (Fig. 6) the status quo was defined by the most important environmental variables (moisture, nutrients,

and land use) and vegetation units with characteristic plant and bird species compositions were assigned. EPIs (grey boxes) were derived out of

the complex interactions of all environmental variables within each vegetation complex.

vegetation offered the opportunity for nitrophile

shrubs and forbs to germinate. Consequently, fen

vegetation was invaded in wetter parts by reed

vegetation, in drier parts shrubs and forbs germinated

and typical low and sparse fen plant species got lost.

This created a mosaic habitat consisting of swamps,

dense reed vegetation and pioneer shrub areas. Reed

bunting, Chiffchaff, Willow warbler Phylloscopus

trochilus and possibly some Sylvia warblers might

have benefited from the situation and increased in

territories.

The process regime in the transition complex is

rather similar to what was found in the fen areas. Due

to drier soil conditions and trampling mineralization

was advanced. Shrubs and forbs were already

established. Typical pasture determined plant species

could spread towards the transition zone because of

drier soil conditions and higher grazing pressure.

An EPI for the wet and drier grassland was the

invasion of shrubs from the transition zone where

vegetation could not be heavily grazed due to

comparatively wet site conditions. In addition further

drainage measures lowered the water table and led to

shrub growth and soil mineralization. Nutrient supply

was high because of direct fertilisation and eluvia-

tions from arable land. Thus, high productive grass-

land species spread at drier parts and changed natural

grassland communities. Where it was economical

reasonable grassland was transformed into arable

land.


Fig. 8. Scenarios based on different EPI constellations for future developments. Main aspects are shown and changes are exemplarily illustrated

for each scenario.

Fig. 8 describes consequences of different con-

stellations of EPI-based scenarios for future develop-

ments.

Scenario 1 implicates that nutrient supply and

cultivation pressure could still be high. This is evident

from a current ongoing expansion of arable land and

may stop directly at the nature reserve border.

Accordingly, drainage measurements might be accom-

plished which would also have an impact on the nature

reserve itself. Nutrient supply also continues to be high.

The nature reserve could continue to loose its

attractiveness for breeding wetland and meadow bird

species. Numbers of Common Snipe and Eurasian

Curlew would probably drop down, a re-colonisation by

more sensitive species such as Corncrake or Short-eared

Owl is excluded in this scenario. In general, birds

specialised on different habitat features would be

replaced by more ubiquitous species.

In scenario 2, cultivation pressure might stop.

Consequently, no further transformation into arable

land and attached drainage would take place. A few

constructed wetlands might be built in some afferent

canals causing a reduction of nutrient outfluxes. In fact,

the nutrient supply would be reduced to a certain

degree, and especially phosphorus would be reduced by

up to 61% (Fylkesmannen i Rogaland, 1996). However,

the reduction might not stop further sedimentation and

the expansion of shrubs and forbs. If conservation of

meadow birds was a goal in this scenario, management

measures would have to be implemented. Removal of

shrubs, restoration of the natural water table and

appropriate extensive land use are minimum require-

ments to maintain the current situation (cf. Beintema

et al., 1995). Extinct species would probably not return

to the area in this scenario. The implementation of this

kind of landscape management would only have a

medium-term effect, because nutrient supply would

still be high and would cause fast re-growth. On the

other hand, these management interventions might help

to maintain the current diversity in bird species of open

landscapes.

Scenario 3 presents a substantial change in land

cultivation which aims at a broad reduction of

nutrients achieved by constructed wetlands in all

streams. In addition, strong fertilisation restrictions

might be implemented within the reserve. Compulsory


edge strips of non-fertilised land would function as

buffer zones and prevent direct nutrient flux into surface

water. These regulations of land use would have to be

compensated by a substantial financial support. Within

the nature reserve the use of wet grasslands as meadows

rather than as pastures would have to be supported. The

conversion into meadows or extensive pastures might

prevent the disturbance of wetland breeding birds. Also,

fens would be protected against additional eutrophica-

tion, mineralization, and invading shrubs, since

trampling would not destroy the upper organic layers.

Nevertheless, succession might not be avoided com-

pletely, as changes would still be natural as a

consequence of lake ageing. Bird life diversity would

be expected to be high. Numbers of species of the fen/

grassland group (especially not very sensitive ones, like

Northern Lapwing and Eurasian Curlew) might

stabilise or even increase. A return of several extinct

breeders such as Black-tailed Godwit to the area seems

possible with an improvement in habitat quality by

more extensive agricultural use, but strongly depends

on overall population numbers and trends in the region.

Species preferring agriculturally used areas (such as

Yellowhammer) would probably not show a further

increase in numbers and might even decline.

4. Discussion

The attractiveness of the nature reserve Lake

Grudevatn evolved from its complex natural conditions,

diverse landscape structures and moderate agricultural

impact resulting in a diversity of vegetation units as well

as faunistic habitats (Wetlands International, 2004).

The interactions between anthropogenic and natural

processes are complex and vary greatly in direction, rate

and scale (Jones, 1991). Formerly, bird diversity was

high including species of open landscapes (i.e. L.

limosa), as well as those depending on dense vegetation

structures such as reeds, shrubs and forests (Løvbrekke,

1995). But during the last decades many areas in the

agricultural landscape were strongly influenced by

anthropogenic variables and landscape transformation

happened rapidly (Moss, 2000). Thus, bird species

composition was changed although bird species

richness remained high at Lake Grudevatn (Løvbrekke,

1992a, b, 1995). But this rather resulted from the mosaic

of agricultural land and small remaining protected

wetlands. Our results confirmed that this patch-context

effect (Dunning et al., 1992; Thies and Tscharntke,

1999) had a strong impact on species composition in the

SW Norwegian landscape. Within this context, the

protection of bird diversity following a static protection

philosophy failed under a perspective of protecting bird

species compositions of the characteristic wetlands

within an extensively used traditional agricultural

landscape. Agricultural activities were intensified

outside the protected area and nutrient supply caused

substantial changes in the vegetation (Bar et al., 2004).

Thus, even the effort of landscape management

measures is limited and could neither prevent sedi-

mentation processes nor reverse vegetation succession

completely. Nevertheless, this fact should not be

interpreted as an argument against protection efforts

in general. High bird species diversity will be desirable,

if protection aims focus on extensive anthropogenic

impact in coincident with wetland protection main-

tained by drainage prohibitions and extensive pastures

(Wetlands International, 2004). To apply sustainable

management measures successfully, it will be important

to develop an integrative management plan based on

changing natural conditions, changing human impact,

and realistic future options (Opdam et al., 2002;

Lundberg, 2004; Loffler and Steinhardt, 2004).

Combining EPI-based estimations of changes in

vegetation patterns at different spatial scales and time

perspectives with the variability of bird species

diversity proved to be applicable. Moreover, we

succeeded to test a combination of multi-scale

environmental variables and methods, including a

short investigation period combined with different

historic data sources. By using vegetation changes the

complex of medium- and long-term landscape

transforming processes could be determined and

helped illustrating bird species changes better than

solely studies based on fragmentary bird data.

However, general negative trends in bird species

occurrence have to be investigated within a wider

context. The different constellations of EPIs for

scenarios partially consider this fact and address

conflicts between changing landscapes and static

nature protection concepts. Since other studies showed

that the ecological mechanisms behind the more

common ‘‘non-scientific’’ indicator approaches

needed further validation within their geographical

context (Cousins and Lindborg, 2004), our EPI-based


approach might be used as a tool to solve similar

nature protection problems.

Prognostic indicator-based studies are directed to

landscape planning and management and thus of

particular interest for local administrators (Venturelli

and Galli, 2006). Since northern European wetlands

and extensive pastures are one of the most threatened

habitats in the rural landscape and thus in the focus of

protection plans, it is necessary to assess effects of

management and landscape changes (Cousins and

Lindborg, 2004). The crucial point is that nature

protection areas in agricultural landscapes of northern

Europe are seldom found and the few small ones are

very isolated and still influenced by their agricultural

surroundings (Bar et al., 2004). This was also found

for central European cultural landscapes (Jedicke,

1994). Thus, the interest to maintain these remaining

areas increased in order to protect a large number of

ecotypes and species diversity (Arler, 2000). Further-

more, it is important to define protection aims and

introduce management measures if necessary. Our

scenarios are no predictions but they show in which

direction development might take place and how

protection aims have to be adjusted within a realistic

frame. The approach is usable for protection areas

where protection aims are neither specified nor fit into

the recent protection strategy, and where long-term

monitoring data do not exist. Such scenarios might be

used to describe potentials and limits of calls for

actions for different protection strategies.

Acknowledgements

We would like to thank Johannes Kamp (Old-

enburg, Germany) for support on ornithological data

interpretation, Professor Anders Lundberg (Bergen,

Norway) for hospitality, cooperation, botanical and

nature protection expertise, and two reviewers for

helpful suggestions to improve the paper.

References

Boden, A.G., 1996. Bodenkundliche Kartieranleitung. Schweizer-

bart’sche Verlagsbuchhandlung, Stuttgart.

Arler, F., 2000. Aspects of landscape or nature quality. Landscape

Ecol. 15, 291–302.

Bar, A., Loffler, J., Lundberg, A., 2004. Spatio-temporal changes of

a nature protection area in the agricultural landscape of Jæren,

SW-Norway. Norwegian J. Geogr. 58, 147–157.

Beintema, A.J., Moedt, O., Ellinger, D., 1995. Ecologische Atlas van

de Nederlandse Weidevogels. Schuyt & Co., Haarlem.

Clements, J.F., 2000. Birds of the World: A Checklist. IBIS Publish-

ers, Temecula, CF.

Cody, M., 1987. Habitat selection in birds. Academic press, London.

Cousins, S.A.O., Lindborg, R., 2004. Assessing changes in plant

distribution patterns—indicator species versus plant functional

types. Ecol. Indic. 4, 17–27.

Dunning, J.B., Danielson, B.J., Pulliam, H.R., 1992. Ecological

processes that affect populations in complex landscapes. Oikos

65, 169–175.

Fjellanger Widerøe, A.S., 1937, 1954, 1989. Aerial photos of Lake

Grudevatn, Fornebu.

Flade, M., 1994. Die Brutvogelgemeinschaften Mittel- und Nord-

deutschlands. Grundlagen fur den Gebrauch vogelkundlicher

Daten in der Landschaftsplanung. DLV, Eching.

Fremstad, E., 1997. Vegetasjontyper i Norge. Norsk institutt for

naturforskning, NINA Temaheft 12, Trondheim.

Fylkesmannen i Rogaland (Landbruksavdelinga), 1996. Storskal-

prosjekt ureining. Biologiske Reinseparkar i Frøylandsana.

Fylkesmannen i Rogaland, Stavanger.

Jedicke, E., 1994. Biotopverbund: Grundlagen und Maßnahmen

einer neuen Naturschutzstrategie. Ulmer, Stuttgart.

Jones, M., 1991. The elusive reality of landscape. Concepts and

approaches in landscape research. Norwegian J. Geogr. 45, 229–

244.

Jones, J., 2001. Habitat selection studies in avian ecology: a critical

review. Auk 118, 557–562.

Jongman, R.H.G., Ter Braak, C.J.F., Van Tongeren, O.F.R.

(Eds.), 1995. Data analysis in community and landscape

ecology. University Press, Cambridge.

Kent, M., Coker, P., 1992. Vegetation description and analysis: a

practical approach. CRC Press, London.

Loffler, J., Steinhardt, U., 2004. Herleitung von Landschaftsleitbil-

dern fur die Landschaftsbewertung. Beitrage fur Forstwirtschaft

und Landschaftsokologie 38, 147–154.

Løvbrekke, H., 1992a. Vadefugl ved Grudevatn naturreservat. Falco

3, 168–184.

Løvbrekke, H., 1992b. Hekkefugler ved Grudevatn naturreservat.

Falco 4, 202–211.

Løvbrekke, H., 1995. Avifaunaliste for Grudevatn naturreservat.

Falco 1, 40–42.

Lundberg, A., 1996. Environmental change and nature management

in Norway. Norwegian J. Geogr. 50, 143–155.

Lundberg, A., 2000. Die historische Entwicklung der westnorwe-

gischen Kulturlandschaft im Licht der neuen Paradigmen der

Okologie. Eine Diskussion des Einflusses des Gerateeinsatzes

auf die Gestaltung der Kulturlandschaft, Norden 13, 77–99,

Bremen.

Lundberg, A., 2004. Environmental change, human efforts and

nature management. Norwegian J. Geogr. 58, 145–146.

McGarigal, K., Marks, B., 1994. FRAGSTATS—Spatial Pattern

Analysis Program for Quantifying Landscape Structure. Oregon

State University, Corvallis.


Molværsmyr, A., 1990. Samlerapport: Forurensnings-undersøkelse i

vassdrag og sjøomrader i Rogaland. RF-132/90, Stavanger.

Molværsmyr, A., Sanni, S. and Tyvold, T., 1989. Basisundersøkelse

av Figgjovassdraget 1984–1987, RF-219/89, Stavanger.

Moss, M.R., 2000. Interdisciplinarity, landscape ecology and the

‘‘transformation of agricultural landscapes’’. Landscape Ecol.

15, 303–311.

Nedkvitne, J.J., Garmo, T.H., Staaland, H., 1995. Beitedyr i kul-

turlandskap. Landbruksforlaget, Oslo.

Olafsrud, K., 1993. Grudevatn naturreservat, Klepp kommune,

Rogaland. Verneverdier og forslag til skjøtselstiltak. Hovedopp-

gave (master thesis), Institute for biology and nature, University

of As (NHL), As.

Opdam, P., Foppen, R., Vos, C., 2002. Bridging the gap between

ecology and spatial planning in landscape ecology. Landscape

Ecol. 16, 767–779.

Pickett, S.T.A., Parker, V.T., Fiedler, P.L., 1992. The new paradigm

in ecology: implications for conservation biology above the

species level. In: Fiedler, P.L., Jain, S.K. (Eds.), Conservation

Biology. The Theory and Practice of Nature Conservation,

Preservation and Management. Chapman & Hall, New York.

Pott, R., Remy, D.,2000. Gewasserdes Binnenlandes.Ulmer, Stuttgart.

Pott, R., 1995. Die Pflanzengesellschaften Deutschlands. Ulmer,

Stuttgart.

Remmert, H., Golley, F.B., Billings, W.D., Lange, O.L., Olson, J.S.

(Eds.), 1991. Mosaic-Cycle Concept of Ecosystems. Springer,

New York.

Semb, G., 1962. Jorda pa Jæren. Beskrivelse til jordbunnskarter over

en del av Jæren. Jordbunnbeskrivelse 41 (42) fra Statens Jor-

dundersøkelse. Melding fra Norges Landbrukshøgskule.

Simberloff, D., Dayan, T., 1991. The guild concept and the structure

of ecological communities. Ann. Rev. Ecol. Syst. 22, 115–143.

Skanes, H.M., 1997. Towards an integrated ecological-geographical

landscape perspective. A review of principal concepts and

methods. Norwegian J. Geogr. 51, 145–171.

Succow, M., Joosten, H. (Eds.), 2001. Landschaftsokologische

Moorkunde. E. Schweizerbartsche Verlagsbuchhandlung, Stutt-

gart.

Ter Braak, C.J.F., 1986. Canonical correspondence analysis: a new

eigenvector technique for multivariate direct gradient analysis.

Ecology 67, 1167–1179.

Ter Braak, C.J.F., 1994. Canonical community ordination. Part I.

Basic theory and linear methods. Ecoscience 1, 127–140.

Ter Braak, C.J.F., Prentice, I.C., 1988. A theory of gradient analysis.

Adv. Ecol. Res. 18, 271–317.

Ter Braak, C.J.F., Smilauer, P., 1998. CANOCO Reference Manual

and User’s Guide to CANOCO for Windows. Center for Bio-

metry, Wageningen.

Thies, C., Tscharntke, T., 1999. Landscape structure and biological

control in agroecosystems. Science 285, 893–895.

Tiner, R.W., 2004. Remotely-sensed indicators for monitoring the

general condition of ‘‘natural habitat’’ in watersheds: an appli-

cation for Delaware’s Nanticoke River watershed. Ecol. Indic. 4,

227–243.

Tonu, O., Alamets, K., Parnamets, H., 2005. Modelling bird habitat

suitability based on landscape parameters at different scales.

Ecol. Indic. 5, 314–321.

Venturelli, R.C., Galli, A., 2006. Integrated indicators in environ-

mental planning: methodological considerations and applica-

tions. Ecol. Indic. 6, 228–237.

Vikse, L. (Ed.), 1998. Aksjon Jærvassdrag. Handlingsplan 1998–

2002. Plandokument. Rogaland Fylkeskommunes regionalpla-

navdeling, Stavanger.

Waldhardt, R., Simmering, D., Otte, A., 2004. Estimation and

prediction of plant species richness in a mosaic landscape.

Landscape Ecol. 19, 211–226.

Wetlands International, 2004. http://www.wetlands.org/RSDB/

default.htm, April 29, 2004.

Wilson, J.B., 1999. Guilds, functional types and ecological groups.

Oikos 86, 507–522.

http://www.wetlands.org/RSDB/default.htm

http://www.wetlands.org/RSDB/default.htm

Documents

Ecological process indicators used for nature protection scenarious in agricultural landscapes of SW