The following criteria guide experimental design (Lovett et al. 2007).
(1) The program is designed around clear and compelling science questions
These determine the variables measured, spatial extent of sampling, intensity and duration of measurements, and usefulness of data.
(2) Review, feedback, and adaptation is included in the design
The key science questions and data collected need to be continually reviewed for relevance and validity without compromising continuity of core measurements.
(3) Measurements need to be efficient, cost-effective and amenable to long-term sustainability
Core measurements should provide statistically valid, basic measures of system function and dynamics. Power analysis should assist in determining replication requirements. Costs of measurements should not jeopardise the long term viability of the project. Photopoint monitoring, where used, should adhere to established guidelines (O’connor and Bond 2007).
(4) Quality and consistency of data are maintained
Quality assurance protocols need to be established at the outset. Sample collections and measurements must be rigorous, repeatable, well documented and employ accepted methods. Changes in methods should be rare and should involve overlap periods to ensure comparability.
(5) Long-term data accessibility and sample archivingis ensured
Metadata should provide all relevant details of survey, collection, analysis and data reduction. Raw data should be stored in accessible form for reanalysis. Raw data, metadata and descriptions of procedures should be stored in multiple locations.
Policies of confidentiality, data ownership, and data hold-back times should be established at the outset. Archiving of samples (soils, water, plant voucher specimens and animal materials) can be valuable resources for future analysis.
(6) Monitoring data is continually examined, interpreted and presented
Error and trend detection is optimised by frequent external review through publishing or sharing of data and results.
Adequate resources need to be committed to managing data and evaluating, interpreting, and publishing results. Often these are given low priority compared to actual data collection.
(7) Monitoring is part of an integrated research program
An integrated program may include qualitative observation, experimentation, modelling and cross-site comparisons.
Scientific questions determine experimental design: the type of data to be collected, the scale at which it is collected, the sampling strategy to be used and the analytical methods to test reliability and meaningfulness of results.
The following scientific questions are posed to frame assessment and monitoring programs relating to the establishment phase that could significantly guide early restoration activities in this project from seedling to later stages.
1. Are spatial distributions of species clumped, random, even, or absent (i.e. no native species present on the site)?
2. Are spatiotemporal distributions and abundances of species associated with specific environmental conditions (niche specificity, e.g relating to availability of light, water or nutrients), dispersal parameters (species pool, vectors, timing, e.g. priority and year effects) or disturbance regimes?
3. Are recruitment, growth and mortality rates of species (productivity) related to specific environmental associations or gradients, or to something else?
4. Are temporal changes in vegetation or functional groups at the site behaving according to normal linear successional dynamics or do threshold or other dynamics apply?
5. Is the system already in or heading towards a desirable or undesirable stable state?
6. What specific system parameters or leading indicators can give early warning of impending regime change?
7. What specific interventions (e.g. mowing, herbicide applications, removal of plants, direct seeding, planting etc) could most effectively impact on system parameters — the species pool, species interactions and feedbacks, resource availability, disturbance regimes — and hence the stability or resilience of a current stable state or the dynamics of successional processes? How will the interventions restart, redirect or speed-up the trajectory of succession?
8. What system parameters will suffice to indicate autogenic succession has been achieved?
9. What parameters (re. structure, function, stability) best define the desired end point of successional processes that are initiated by restoration activities?
