diagramming relationships within organisations (and to the community and stakeholders) [soft systems methodology] or between actions in achieving an outcome [fishbone diagrams];
showing the ecological system we are interested and how we think it operates so we can think about what we need to measure and how [conceptual diagrams],
maps of the problem space showing habitat, sources of pollutants (etc.), topography, visual indicators, and the dimensions of the problem and issues of scale; and
sometimes just exploratory doodles to help get a handle on the issues you are dealing with.
Some models can be very pictorial and could be designed so non-experts can conceptualise the system.
Others can be more mechanistic or process oriented and may be intended to assist in understanding key parts of the system and to allow discussion of monitoring design and the implications for analysis.
There needs to be some explicit link between your outcomes and what you will measure. Simple cause-effect relationships are often illusory but by using models you can make clear your understanding of processes and relationships to ensure that the most appropriate parameters are chosen. Pictures can clarify the complexity of the situation and, therefore, provide insights into the limits to management of the system and what we can achieve. Working with ucertainty is integral to environment management and how we respond to information as it becomes available.
Do not just do the same old tests because that is what everyone else is doing (or "you" have done in the past)! Recognise also that you might need to include data on operational inputs if you are going to be able to understand how your actions are effecting the environment. If you have measurable steps between the action and the final effect then look at these. This is particularly the case where you are looking at policies or higher level strategies; look to see that they have actually been implemented (and are working) and don't just rely on the environmental (or other) endpoint.