he following recommendations are extracted directly from the report. The first number in parenthesis indicates the chapter of origin, the second indicates the order of appearance in each chapter.

Recommendation (1.2): Some of the sites should be in wilderness areas dedicated to LTERM studies. Other sites should be in areas containing managed ecosystems and in areas containing highly stressed or already degraded ecosystems. Special attention should be given to the preparation of guidelines for operating LTERM facilities in such areas.

Recommendation (2.1): Research into non-equilibrium thermodynamics ought to be one of the centrepiece themes of a Canadian LTERM.

Recommendation (2.2): In the context of biodiversity, habitat destruction and fragmentation should be monitored on a long-term basis.

Recommendation (2.3): The Canadian LTERM program should seek close working relations with the paleoecology and remote sensing communities in Canada. In both cases, the human resources and technical support already available will be invaluable.

Recommendation (2.4): Ecosystems need to be studied on many spatial and temporal scales (from local to continental, and from decades to centuries). Of particular importance is the patchiness of ecosystems, so that the landscape scale needs to be emphasized.

Recommendation (2.5): The ecosystem approach described in this chapter allows us to recommend a number of key research topics for a Canadian LTERM program:

• Studies of the relationships amongst biodiversity, ecosystem functioning and ecosystem integrity;
• The application of non-equilibrium thermodynamics to ecosystem functioning and integrity. This ought to be a centre-piece theme of the Canadian LTERM program;
• Studies of ecotones. These provide special opportunities to examine the effects of discontinuities on ecosystem functioning. They also may be particularly useful as early indicators of global change;
• Environmental time bombs. Ecosystem functioning may sometimes be threatened over the long term by minor but accumulating stresses, and the ultimate result may be a sudden and surprising disfunctioning of the ecosystem involved.

Recommendation (3.2): It is further recommended that the development of early warning indicators of long-term global change be given particular attention within this program. This will of course involve model development, field experiments and the establishment of long-term monitoring programs. Special attention should be given to UV-B monitoring in the arctic.

Recommendation (4.1): There have been many short-term studies of urban ecology, but there have been few LTERM-related programs in the urban environment. Therefore, it is recommended that urban ecosystems research be given attention comparable to that given to agricultural, forest and other less modified ecosystems.

Recommendation (4.2): Considering the importance of sustainable agriculture and forestry to Canadians and Canadian export markets, and considering the importance of soils and soil processes to ecosystems, it is recommended that the Canadian Soil Quality Benchmark Sites program be incorporated into a Canadian LTERM program.

Recommendation (4.3): It is recommended that forestry departments, forest companies and schools collaborate to institute a national system of tree sample and permanent plot data repositories. To support this initiative, research should be undertaken on criteria and indicators.

Recommendation (4.4): It is imperative to establish and support a long-term monitoring and research program that would follow structural and behavioural changes in several selected tundra sites. Information obtained will establish a reference base, against which all future changes will be compared. It will also assist in formulation of northern social, demographic and economic policies, and in management of renewable resources.

Recommendation (4.5): Marine areas should be included in a Canadian LTERM network. This shoud be done by building on existing marine ecosystem monitoring programs and by developing them to reflect the hierarchies within Canada's various marine ecozones.

Recommendation (4.6): There has been a wealth of ecosystem information collected for national park management, but the localized nature of the studies has resulted in a lack of standardization in methodology and data structures, both between parks and between study repeats. It is therefore recommended that Parks Canada institute a process of ecological data standardization and dissemination, in the context of LTERM.

Recommendation (4.7): National Parks are established in law to protect representative examples of Canada's natural diversity. This makes them ideal sites for LTERM, especially in light of the prospect of being able to repeat studies decades or centuries hence. However, landscape modifications for research purposes are not permitted in national parks. Therefore, it is recommended that portions of national parks be set aside for the purpose of establishing benchmark or reference LTERM sites where ecological manipulations could be undertaken.

Recommendation (4.8): Canada contains a significant fraction of the Earth's land and water surface. Thus Canadian participation in international terrestrial and marine ecological programs, particularly IGBP/GCTE, should be given high priority.

Recommendation (4.9): It is recommended that the Canadian LTERM program maintain both a continental and a hemispheric perspective. Close liaison should be developed between the Canadian LTERM and various relevant U.S. hemispheric national and state programs, and also circumpolar research studies. In the case of marine LTERM programs, the spatial scales involved must be sufficiently large to correspond to those that include the life cycles of the species involved.

Recommendation (4.10): It is recommended that the Canadian LTERM program maintain a long-term (50 years or longer) perspective. Accepting the precautionary principle and assuming that ecological changes will indeed occur, considerable effort should be given to developing early warning indicators and improving our understanding of ecosystem functioning under changing conditions.

Recommendation (5.1): An ecosystem approach must be taken in the LTERM program. Ecosystems should be studied inter alia as non-linear systems with multiple states between which the system may flip.

Recommendation (5.2): Ecosystems need to be studied on a variety of temporal scales, from decades to centuries.

Recommendation (5.3): In studying ecosystems, efforts should be made to capture the effect of ecosystem patchiness on a variety of scales.

Recommendation (5.4): Each terrestrial LTERM site should preferably include within its bounds, a small calibrated catchment, providing basic information on hydrological and biogeochemical cycling.

Recommendation (5.5): Special attention needs to be given within LTERM to stressed and already degraded ecosystems, within the context of ecological integrity and sustainability.

Recommendation (5.6): A main recommendation of the Panel is that long-term institutional and financial support be provided by the government of Canada, in collaboration with the provinces and the community of Canadian ecologists.

Recommendation (5.7): Because of the importance of Canada's renewable resource base, it would probably be desirable to ensure that an LTERM Program be supported at senior levels of the Federal and Provincial governments (i.e. Cabinet level).

Recommendation (5.8): The federal government should situate responsibility for a Canadian LTERM program in a single national agency, whether that agency be pre-existing or created for this purpose.

Recommendation (5.9): In any cost-cutting exercise, the Panel urges LTERM sponsors to accord the greatest possible protection to on-going, long-standing programs. In return, LTERM operators must accept the responsibility to maintain high levels of ecological relevance, to maintain standard methods, and to report to their peers and the public at large.

Recommendation (5.10): Ecosystems should be studied from as many perspectives as possible, including hydrological cycling, biogeochemical cycling, food webs, overall energetics, geomorphology, and soils, population dynamics, life cycle assessments, landscape patchiness and corridors, stress-response assessments and ecosystem integrity.

Recommendation (5.11): The objectives for LTERM programs should be carefully specified for each station. These objectives need not be the same across all stations in the national network.

Recommendation (5.12): Site continuity should be ensured for existing sites and shielded from undue disturbances.

Recommendation (5.13): Considerable effort should be given to improved understanding of ecosystem functioning under changed conditions, and of the potential for extrapolating findings from local to regional and even larger scales.

Recommendation (5.14): The Canadian LTERM program must establish a national data management strategy, including the development of a metadata system. This strategy should map out means for keeping up with evolving instrumentation and scientific understanding, the use of citizen monitoring programs, and the design of repositories for biological samples and data archives. About 10-15% of the LTERM budget should be spent on quality control, data storage and documentation.

Recommendation (5.15): Individual Canadian LTERM facilities should be integrated into a national network. Networking among researchers and administrators should be encouraged and facilitated through electronic mail, regular meetings of station managers and research scientists, and the above-mentioned data management strategy. [In some cases, of course, networking has been detrimental by adding bureaucracy. This tendency should be avoided.]

Recommendation (5.16): Noting that networks of LTERM stations will be required to assess a wide range of public issues and research questions, some of which cannot yet be imaged, the principle of "nested networks" should be followed, in which tiers of stations are established, each to be activated as required. These networks should be supplemented by archived remotely sensed ecological data covering the areas between stations.

Recommendation (5.17): In choosing sites, attention should be given to the kinds of questions that each site will be capable of answering, and not answering,.

Recommendation (6.1): It is recommended that Environment Canada (EC) be officially recognized as the LTERM lead agency in Canada. This recognition must be backed by a commitment for long-term funding of LTERM activities in Canada and should include specific reference to the necessary reporting and advisory mechanisms that will ensure that EC plays its role of lead agency.

Recommendation (6.2): It is recommended that the Canadian Global Change Program appoint an LTERM Standing Committee that would receive reports from and provide advice to EC and other organizations involved in LTERM activities in Canada. The costs associated with the operation of this Committee would have to be covered by EC, other government departments, provincial ministries and other bodies supporting the Program.

Recommendation (6.3): It is recommended that a report on Canadian LTERM activities be prepared periodically by the proposed Standing LTERM Committee.