run quite separately at different sites to answer separate ..... enced in Britain; the cool sites in the north of ... the technology for remote database access is.
_ _ _ _ ~
Rotunno, R and Klemp, J. B. (1985) On the rotation and propagation of simulated supercell thunderstorms. J. Amos. Sci., 42, pp. 271-292 Thorpe, A. J. and Miller, M. J. (1978) Numerical simulations showing the role of the downdraught in cumulonimbus motion and splitting. Q. 3. R MetemL SOC.,104, pp. 873-4493 Weisman, M. L. and Klemp, J. B. (1984) The structure
and classification of numerically simulated convective storms in directionally varying wind shears. Mon. WU.Rev., lU, pp. 2479-2498 (1986) Characteristics of isolated convective storms. In: Ray, P. (Ed.) Mesoscale meternology and jmecazikg, American Meteorological Society, pp. 331-358
The Environmental Change Network: Integrating climate and ecological monitoring M. D. Morecroft, J. K. Adamson and A. M. Lane NERC Institute of Terrestrial Ecology, Merlewood Research Station, Cumbria
Meteorologists take monitoring for granted: some recording stations have been doing it for 200 years or more, our present understanding of climate systems could not have been gained without it, and no forecast is made without reference to measurements of recent conditions. In most of the other environmental sciences, monitoring has not had such a fundamental r6le and experimental approaches have offered a more rigorous and usually quicker path to understanding natural processes. The last ten years or so have, however, seen a profound change in attitudes to monitoring, driven by the realisation that the natural environment is changing and that human activities have a major r61e in these changes. By the late 1980s Britain had networks of sites monitoring both air and water pollution and a few national schemes recording specific biological variables such as numbers of bird, butterfly and moth species. These were, however, being run quite separately at different sites to answer separate questions. Whilst each issue is certainly interesting in its own right, knowledge of the interactions between them is vital to understanding the processes involved and making predictions for the future. The Environmental Change Network (ECN)was therefore set up to monitor different aspects of environmental change and their impacts at a range of sites throughout the UK.
ECN was conceived by the Natural Environment Research Council (NERC) and officially launched by the Minister for the Environment and Countryside in London in January 1992. At that time the first eight sites in the network had been identified although it was not until late 1992 that physical measurements began, with biological measurements following in spring 1993. The founding ECN sites have now been joined by a further three sites making a total of 11 where terrestrial monitoring takes place. Freshwater measurements have also been introduced and are currently made at 21 river sites and 16 lake sites. The network is not centrally funded but relies on sponsoring bodies that contribute by supporting monitoring at one or more sites, and that also define the direction of development of the network.
Climate change and ecology Climate change is potentially the most farreaching environmental change at the global level. The fact that climate has an impact on ecosystems is not hard to see - the global distributions of plants and animals clearly reflect climatic differences (Woodward 1987) and more local variations in climate such as a change in altitude are usually paralleled by changes in vegetation, soils and animal populations. The mechanisms which underlie these 7
observations are varied and complex but some general principles can be identified. At the most fundamental level, all organisms have climatic limits beyond which they cannot survive; these are normally determined by extremes of temperature or water supply. In very cold and very dry regions, the poles, high mountains and deserts, the species that occur there are determined simply by the ability to persist. In less severe regions, many more species are capable of survival but different organisms have different climatic optima for growth and reproduction and the distribution of species tends to reflect these. This is because species performing at or near their optima are likely to be most successful in competition for limiting resources such as light and nutrients (in the case of plants) or food (in the case of animals). Similar patterns can be seen in the case of agricultural animals and plants, with different species giving better yields in different climatic regions. Historically, wheat and barley have been the predominant cereals in much of Europe, but oats or rye have been favoured in the cooler north. A greenhouse warming of the magnitude predicted by the Intergovernmental Panel on Climatc Change (1995), approximately 2 degC by 2100, would be enough to cause movements of several hundred kilometres in the geographical location (or 200-300m in altitude) of climatic optima and boundaries for most plant and animal species. We would therefore expect to see shifts in distributions. In reality this might not happen in many species as dispersal and colonisation of new habitats can be a slow process; some may be able to persist where thev are for a considerable time, but others are likely to die out. Climate change would also be expected to affect the interactions between different components of ecosystems. For example, some plants flower or produce leaves earlier in a warm spring (Fitter et ul. 1995); in the event of climatic warming the animal species which feed on such plants will tend to decline if their life cycles are not regulated by the same environmental signals.
Environmental change Climate is not the only ecologically important environmental change at the present time and it 8
is neither possible nor desirable for a programme like ECN to restrict itself to one issue. During the 1980s a great deal of international research was concentrated on the topic of ‘acid rain’ and more recent work has examined the effects of ‘dry deposition’ of gases such as nitrogen dioxide and ammonia. These studies have shown that air pollution can cause soil acidification and changes in the composition of plant communities. The biggest effects that people have had on the British countryside to date have, however, resulted from direct changes in land use and management. Britain’s forest cover was largely cleared in prehistoric times, but the rate of change has never been greater than during this century with an increase in urban areas and the intensification of agriculture (Ban er ul. 1993). Climate, air pollution and land-management changes may be regarded as the main ‘driving variables’ which ECN monitors and studies, together with their impacts. ECN
measurements
ECN measurements are based on carefully prepared standard protocols which ensure that operator and instrument bias between sites and over time are kept to a minimum. Qualitycontrol and assurance procedures are built into the protocols. Many of the protocols have been prepared specifically for ECN by acknowledged experts in appropriate fields, but some (such as precipitation chemistry and butterflies) are based on existing methods used by other ‘seccoral’ monitoring schemes. This means that ECN data can be placed in a wider context and that the sectoral networks have at least some sites where more information is available to help interpret their results.
Meteorology in ECN All terrestrial sites monitor weather using an automatic weather station (AWS), commercially produced by Didcot Instruments and originally developed by the Institute of Hydrology. Hourly summary data are collected for a wide range of variables (Table 1) including total solar radiation and net radiation, as well as the more common ones of temperature, rainfall,
~
Table 1 Meteorological monitoring in ECN - variables measured by automatic weather stations Variable (units) Method specifications Wind speed (ms-') Wind direction (") Wet-bulb temperature ("C) Non-aspirated screen Dry-bulb temperature ("C) Non-aspirated screen Soil temperature at lOOmm ("C) Under bare soil Soil temperature at 300mm ("C) Under grass Rainfall, total per hour (mm) Tipping bucket (in pit at upland sites) Wet surface indicator (min h-l) Soil water potential at 200mm (bars)* Gypsum block* Solar radiation (Wm-2) Kipp solarimeter Net all-wavelength radiation (Wm-2) Albedo Downward- and upward-facing silicon cell solarimeters
Resolution 0.1ms-I 1" 0.ldegC 0. ldegC 0.1degC 0.1degC 0.5mm 1min 0.0 1bar* 1Wm-2 1Wm-2 1Wm-2
*Soil water measurement is being reviewed and methodology may change. and wind speed and direction. The automatic system offers a number of advantages over a traditional climate station, especially when it comes to investigating relationships with other variables and identifying specific local changes in climate. Data for the whole 24 hours of the day are useful in, for example, interpreting data on nocturnal animals such as bats and moths. True means can be calculated for whatever period of time is of interest and will give a more sensitive indication of change than daily spot readings; energy balance and potential evapotranspiration can also be calculated from daily means. From a practical point of view, using an A W S means that remote sites do not need to be visited daily, which may not even be possible in some weather conditions. Many of the ECN sites are also Meteorological Office climate stations and daily readings are taken with manual equipment; others provide weekly readings from, at least, thermometers, a raingauge and an anemometer. This provides a useful check for the AWS and facilitates comparison with historic datasets - at Rothamsted meteorological records go back to 1853.
Other measurements at terrestrial sites Precipitation samples are collected at weekly intervals so that pH, conductivity, and the concentrations of 11 major nutrient and pollutant ions, including sulphate, nitrate and ammonium can be determined. At present, only a single pollutant gas, nitrogen dioxide, is mea-
sured directly, using passive diffusion tubes which contain a steel mesh coated with a solution which absorbs the gas. The meshes are exposed to the atmosphere for a fortnight and are then rinsed to produce a solution for chemical analysis. ECN aims to extend the range of gases measured in the near future. Chemical analyses of soils and soil water are also undertaken at each site to determine changes in soil fertility and acidity. Where terrestrial sites have suitable streams or rivers, discharge and the same range of chemical determinants as for precipitation are measured, as drainage water is a major route for nutrient export from ecosystems. Biological recording covers a wide range of organisms. Sampling strategy is an important issue for all groups - populations often have a very patchy distribution within a site and it is necessary to cover a sufficiently large area to give a representative picture; transects or predetermined quadrats are used to facilitate this. Vegetation recording at each site begins with systematically located 2 m by 2 m quadrats for which species lists are recorded, which allows the site to be characterised in terms of the UK National Vegetation Classification (Rodwell 1991). These plots are located at the intersections of a grid laid over a map of the site, the spacing of the grid being adjusted according to the area of the site, to give approximately 500 quadrats. Change will be identified by rerecording subsets of these quadrats at threeand nine-year intervals. Special adaptations are 9
made to deal with linear features, permanent grass, cereals and woodland. In contrast to the plants, only certain selected animal groups are monitored - it is impossible to devise a sampling strategy which would encompass all species. The reasons for selection differ; amongst the vertebrates, birds, bats and frogs are of interest because of their conservation value, whereas rabbits and deer are important for their ability to modify an ecosystem by grazing. Methodologies for actually detecting species differ. Birds can be recorded by direct observation and recognition of song. Bats are recorded at dusk and fly rapidly and silently so identification relies on an electronic device which converts their ultrasonic calls to audible sounds. Rabbits and deer are also difficult to accurately count directly by eye so counts of droppings are made along a transect. Frog numbers are assessed by observations of spawn in selected ponds. Invertebrate animals are also important for a variety of reasons including pollination of plants and as food sources for higher animals, as well as in their own right. Those monitored at terrestrial ECh' sites are moths (using a light trap), butterflies (observation from a transect), spittle bugs (spring and autumn collections from randomly placed quadrats), predator ground beetles (by 'pitfall' trapping along transects) and crane-fly larvae (extracted from soil taken from randomly selected quadrats). The management at terrestrial sites is carefully recorded so that any impact that this may have can be identified when the data are being analysed. Aerial photography is used to detect large-scale changes not only on the sites but also on adjacent areas. Measurements at freshwater sites Plankton, diatoms, invertebrate fauna and submerged plants are being monitored although there are differences in methods according to whether a site is a river or a lake. Water chemistry is a particularly important measurement at freshwater sites with up to 35 variables being measured on each sample. These include all the variables at terrestrial sites plus chlorophyll and heavy metals. In addition, water pH, temperature, conductivity and turbidity 10
are measured continuously with on-site sensors. Discharge is measured at river sites using a weir or flume appropriate to the nature of the river, and water level is recorded at lake sites. AWSS are installed at some freshwater sites and will be introduced more widely in the future. Sites Sites have been chosen so as to maximise the geographical coverage and provide a good representation of the variation in climate, land use, soil and vegetation types in the UK. Sites continue to join the network and various criteria are also considered alongside location when a new site is proposed, to ensure that full use can be made of it. One important factor is stability of ownership and management to try to ensure that purely local changes do not distort the national picture. Another is the existence of long-term data from the past which allow past versus present comparisons at the outset, without having to wait for decades of ECN data to accumulate. Terrestrial sites range
1 Drayton
Fig. 1 Location of E m sites. Terrestrial sites are named and marked with a circle; freshwater sites are marked by triarigles, indicating rivers, and squares, indicating lakes.
in area from 191 to 6500ha. Ecosystems include arable farmland, chalk grassland, deciduous woodland and moorland. The rivers range from large lowland rivers to mountain streams, and the lakes from the Norfolk Broads to remote Scottish lochs (Fig. 1). The range of climates at the various sites is a good sample of the diversity found in the British Isles, as Table 2 shows. There is a clear Table 2 Mean temperatures and total rainfall at ECN terrestrial sites during 1994 Site Mean Total temperature rainfall [OC) (mm) 10.2 880 Alice Holt 10.4 706 Drayton 782 7.8 Glensaugh 8.9 899 Hillsborough 5.6 2086 Moor House 10.4 1310 North Wyke 10.2 597 Rothamsted 7.9 882 Sourhope 10.1 670 Wytham
group of warm, dry, south-eastern sites with very similar conditions and another rather more varied group of northern and western sites which tend to be colder, wetter or both. If there are gaps in the coverage it is at the extremes of low temperature and high rainfall a situation which is a general problem in British climatology. It would also be good to have a
20
i
few more sites with intermediate temperatures and rainfall. As would be expected, differences in climate are reflected in differences in other aspects of the environment. Early ECN data on butterflies give a good example of this and show the potential of an integrated approach to monitoring. Butterflies are monitored by walking the same route once a week between April and September and the numbers of each species seen are recorded (further details can be found in Pollard and Yates 1993). Butterflies are sensitive to the range of temperatures experienced in Britain; the cool sites in the north of Britain consistently record fewer species than the warmer ones in the south. A relationship between diversity and summer temperature can be identified (Fig. 2). The sensitivity to temperature also becomes apparent in the dynamics of individual populations (Fig. 3). At Wytham, Oxfordshire, the peacock butterfly was abundant at two distinct times during 1994, indicating the hatching of two generations. In contrast, at Upper Teesdale, a cooler site in the north Pennines, the overall numbers were lower and there was only a single generation, towards the end of the summer. In the event of climatic warming we might predict that butterfly species would tend to move northwards and to higher altitudes, with species like the peacock having two generations over a
North Wyke
..
Alice Holt Wytham
Rotharnsted Drayton
.
Hillsborough
0 10
12
14
16
18
20
Mean July TemperaturePC)
Fig. 2 Numbers of butterfly species recorded at ECN sites during 1994 plotted against July 1994 temperatures
11
25
&
20
-
15
-
10
-
n
5
5 -
0 , : : t : i : : : :
:”: : : : : :
wider area of the country. The hot summer of 1995 gave support for this in that most species (including the peacock) were more abundant than in the preceding years.
Data management and the
ECN
database
The data collected for national long-term programmes like ECN should be stored and properly maintained in a structured database. The l:CN database aims to provide a complete representation of all information gathered over the duration of the programme. The data must also be accompanied by ‘meta-information’ information about the data, particularly now that the technology for remote database access is widely available, and use and interpretation of information relies less on personal contact. The development and maintenance of the central ECN database is one of the responsibilities of a central co-ordination unit which is run by the Institute of Terrestrial Ecology. The core database stores the raw data at resolutions specified by the protocols, and a summary database consists of monthly and/or annual summaries, generated at regular intervals. Remote access to the core database is available by means of SQI., a widely used ‘query language’. Summary data are more easily accessed by a specially constructed World Wide Web interface (details can be found at http:// www.nmw.ac.uklecn/). Use of ECN data is controlled by a licensing system; interested parties
should contact the E(:N Co-ordinator at the Institute of ’Terrestrial Ecology, Merlewood Research Station, Grange-over-Sands, LAI 1 6Jr; in the first instance. As time goes by, ECN data will become increasingly more valuable - not only increasing our understanding of interactions between environmental variables, but also giving early warning of changes we would wish to try to prevent or mitigate. They will also allow predictions to be tested and models to be refined, so that future management of the environment can be placed on a sounder footing.
~
12
References Rarr, C. J.>Runce. R. G. H., Clarke, R. -1.. Fuller, R. M.,Furse, .M. T., Gillespie, M. K., Groom, G. B., Hallam, C . J., Homung, M., Howard, I). C. and Ness, M. J. (1993) Cozoifrysde S U W ~ \ J1990: iriazn report. Corriiyszde 1990 Senes, Vol. 2. Department of thc Environment, London Fitter, A. H., Fitter, R. S. R., Hams, I. ’1.B. and Williamson, M. H. (1995) Relationship between first flowering date and temperature in the flora of a locality in central England. Funct. Eccol., 9, pp. 55-60
Intergovernmental Panel on Climate Change (1995) Second Assess~~wzt Repon, 1995. Cambridge University Press
Pollard, E. and Yates, T. J. (1993) Moriiruri,ig bunerflies ./or ecohgy and conservation;the British bunerflv monitorztig schenie. Chapman and Hall, London Rodwell, J. S. (1991) British plant wnznzunities, Vol. I (et seq.). Cambridge University Press Woodward, F. I. (1987) Climate and plant distribution. Cambridge University Press
