Monitoring and evaluation are fundamental to informing whether habitat restoration is necessary (chapter 2), what type of intervention may be required (chapters 4 and 5), and what the objectives for restoration might be (chapter 4). They are also critical for assessing a project’s effectiveness once it has begun. Together, this information allows for adaptive management and improvements at both the project level (e.g., refining the methods or design) and across multiple projects. The latter can be especially useful for improving general recommendations for kelp restoration practices, such as site selection, restoration methods, or appropriate goals and timelines. These ‘programmatic’ improvements are especially important for kelp forest restoration, given the relative infancy of the practice and the need to better understand and refine the efficacy, effort, and cost-effectiveness of the various restoration approaches (chapter 5).
6.1 Why monitor?
The monitoring of restoration projects is generally conducted to achieve two separate but interconnected objectives. First, monitoring is needed to understand whether the initial restoration action was implemented as intended: so called implementation monitoring. The core purpose is to evaluate the immediate restoration method or action (chapter 5). Implementation monitoring may therefore focus on kelp-specific metrics (e.g., the number of kelp transplanted) but might also assess other organisms (e.g., the area harvested of urchins) and even technical/construction metrics (e.g., kilograms of seeded gravel deployed; number of artificial reefs installed). This type of monitoring may seem obvious; however, it can be valuable when working with commercial contractors (especially for larger-scale projects) or communicating early milestones and deliverables to funders and regulators. This type of monitoring is also critical in adaptively improving restoration techniques: for example, adjusting the volume/density of seeded gravel based on kelp growth and survival from previous efforts. Accordingly, the second monitoring type is performance monitoring, which is used to evaluate the trajectory of kelp forest recovery and whether the restoration activity is achieving its desired objective(s). Performance monitoring tends to focus on ecosystem and environmental conditions (e.g., fish assemblages; sedimentation/turbidity) in addition to the kelp themselves (e.g. recruitment of juvenile kelp).
6.2 Key considerations when designing and implementing a monitoring program
Performance monitoring needs to occur iteratively and over a longer period than implementation monitoring and ideally should begin before the restoration action (further details below). Monitoring should also occur long enough to encompass both short (<1 year) and medium term (>1–5 years) goals, especially since the recovery of kelp forests and their ecosystem services can take many years 1 2 3. Implementation monitoring typically occurs over shorter timeframes but may be iterative and over extended periods when projects have multiple phases or staged restoration actions.
Regardless of the monitoring type, several key concepts are important for any habitat restoration monitoring program. These include planning, establishing clear objectives, and using systematic and standardised monitoring protocols before and after the restoration intervention 4 5.
6.2.1 The importance of objectives and of systematic monitoring
The primary motivation for restoration is to improve or enhance a degraded habitat towards some preferred state, as defined by a ‘reference’ ecosystem and the primary objective(s) (chapter 4). A critical, but often overlooked, first step of any restoration project is therefore to identify the objectives or reference conditions that determine restoration success. A reference ecosystem or site would ideally be a healthy, local, natural kelp forest that is representative of the restoration objectives, although a reference model may be developed in those instances when a physical site is not available (chapter 4).
Essentially, these objectives and reference conditions become targets for the restoration program and are critical in guiding what monitoring criteria or metrics will be measured (Table 1). Having clear objectives also ensures the most efficient use of monitoring resources, and can aid adaptive management or flexible decision-making, which allow for modifications and improvements to restoration programs already underway.
To fully assess whether a project is meeting its objectives, it is necessary to conduct systematic monitoring before and after the restoration action at the restoration or impact site itself, but also at a control site. For kelp restoration projects, the control site would likely be an unrestored area that represents the before or pre-restoration conditions (e.g., bare or degraded reef). This is a so called Before-After-Control-Impact (BACI) design (e.g. Northern California: Bull kelp & Haida Gwaii: Gwaii Haanas), and together with the reference site as a target, allows fair comparison between different sites and their conditions over time and accurate evaluation of restoration effectiveness 6 7. Fundamentally, this approach allows the evaluation of any improvements at the restoration site relative to the control location(s), but also the trajectory of recovery and how the restored site is performing compared to the reference site or conditions. To enable fair comparisons, control and natural reference sites should have physical characteristics similar to the restoration site (e.g., flow, wave action, tidal range, salinity, water temperature, substrate type, water depth). Lastly, when pre-restoration monitoring is not possible, the comparisons between the restored and control site(s) become even more critical, as is the need to supplement any findings with comparisons to a reference kelp forest, where possible.
Implementing systematic, reproducible, pre- and post-restoration monitoring also allows for comparison of results across projects, since it eliminates the potential that observed changes are simply due to a difference in monitoring methods. Comparisons across projects can aid assessment of programmatic and/or landscape scale outcomes and help untangle the reasons behind success and failure across different locations. Well-planned monitoring programs also enable data collection to address research questions, which can apply across broader spatial scales and promote general improvements in kelp forest restoration methods and outcomes.
6.2.2 Seasonality and monitoring
An important, but sometimes overlooked, consideration for habitat monitoring programs is the seasonality and frequency of sampling. Seasonal changes can cause natural variations in environmental and biological metrics that might be misinterpreted as impacts from a restoration activity (e.g., animal presence/absence, changes in water quality). Seasonality also directly influences the kelp themselves, including their reproductive cycles, and patterns of growth and perennial/annual survival. As such, it is important that monitoring of specific metrics is relevant and aligned with seasonal patterns. Likewise, the frequency or regularity of monitoring must also be considered relative to the metric being assessed. For some metrics (e.g., adult canopy cover) lower-frequency sampling (e.g., monthly or seasonal) may be sufficient, whereas higher-frequency sampling (e.g., weekly) might be needed where responses may be more rapid or unknown (e.g., survival of outplanted juvenile kelp) (also see chapter 5).
6.2.3 Citizen science monitoring programs
Well-managed citizen science programs can help alleviate some of the financial and resource burdens of monitoring, while also facilitating valuable public engagement. When citizen scientists are trained and involved in monitoring, they can provide critical support for evaluating project performance and can serve as project ambassadors who help build community support and understanding of restoration projects. This can be especially useful for marine restoration projects, where limited accessibility and visibility can often cultivate an out-of-sight and out-of-mind mentality among the public. There are several citizen science projects linked with kelp forest restoration and monitoring programs, including urchin control for kelp restoration (e.g. Northern California: Bull kelp 8 9 and mapping of kelp forest loss and recovery (Kelp Tracker, 2019; 10 11).
6.3 Approaches to monitoring
Several resources already describe a range of standardised monitoring methods for rocky reefs, kelp forests, and marine restoration projects (Box 6.1). Here we highlight some basic approaches to kelp forest monitoring that we consider most relevant for restoration practitioners. Nonetheless, no single monitoring approach or method is ideal for every application and circumstance, and so when developing a monitoring program, restoration projects should adopt several approaches that best suit their needs.
| Box 6.1 Kelp forest and reef habitat monitoring resources |
|---|
| Effective monitoring of restoration guidebook 12 PISCO kelp forest sampling protocols 13 Puget Sound kelp forest ecological surveys 14 Reef Life Survey methods manual 15 |
6.3.1 In-water
In-water surveys are perhaps the most widely used, detailed, and valuable of the monitoring approaches. These surveys can be conducted on the surface by snorkellers or SCUBA divers. Each method depends on a range of factors, including site access, water depth and clarity, and the biological/physical metrics being assessed (Table 6.1). In-water approaches do have limitations and restrictions, however, and can be difficult to scale to large areas due to their time- and resource-intensive nature. Divers may also require additional training and qualifications when using specialised breathing and technical equipment.
6.3.2 Remote sensing
Remote sensing uses technology to remotely monitor and survey habitats. Previously, this primarily relied on aerial imaging from aircraft and satellites 16 17 18. However, remote sensing and monitoring from drones is ever-increasing, as are remote approaches for in-water surveys such as Remotely Operated Vehicles (ROV) and Autonomous Underwater Vehicles (AUV) 19 20 21. Due to the remote nature of these approaches, they can be very effective at monitoring and observing very large areas (e.g., hundreds of kilometres), although they can have drawbacks regarding the level of detail and inability to survey many subtidal environments. Remote sensing approaches also typically require access to specialised technical equipment, training, and analyses, although the accessibility and availability of this equipment is rapidly improving.
6.3.3 Coastal or on-water surveys
These cover a suite of on-the-water methods that do not involve in-water surveys. Often, these approaches use watercraft, such as kayaks or powered vessels, to survey kelp canopies or biological and physical metrics from the surface. Similar observations can also take place from land in areas where kelp forests (especially those with floating canopies) grow close to the coast. These might also incorporate beach surveys to assess drift seaweed or ‘wrack.’ These surveys can be low-cost and are not especially resource intensive, and they can also be conducted by relatively unskilled observers, which means they can be effective at covering wide spatial scales. However, these benefits can come at the expense of the level of detail, and there may be physical limitations in assessing some critical kelp forest parameters from the surface.
| Box 6.2 Recording project data |
|---|
| Projects must ensure that they archive their project survey information in order to track their progress over time, learn from past mistakes, and share information. Selecting which variables to record and the appropriate formatting can be a difficult process, but consistent data archives can help advance the field of restoration and promote knowledge exchange between projects. We stress that all project outcomes should be recorded, since we can often learn just as much from our failures as we do from successes. Data storage and sharing also allows for formal analysis of project outcomes. As with some other aspects of kelp restoration, data analysis can require specialised or technical skills, but these can likely be facilitated and provided by local regulators and managers, environmental consultants, or universities. The Kelp Forest Alliance has a standardized data sheet, and we recommend that projects use it as a template and upload the results of their project on the kelp forest alliance website (kelpforestalliance.com). |
Table 6.1 Examples of monitoring metrics for kelp forest restoration. These are broadly organised as either implementation or performance monitoring metrics, but this categorisation is not strict and can often become mixed. Moreover, some metrics (marked with *) can be considered ecosystem services (e.g., fisheries’ benefits, carbon cycling), which themselves may also be objectives or even methods of restoration (e.g. improvements in water quality) (chapter 4).
| Examples of common monitoring metrics |
|---|
| Implementation monitoring |
| Area and/or amount of kelp transplanted Area and/or amount of seeded material deployed Area and/or amount of artificial substrate deployed Area and/or number of urchins removed |
| Performance monitoring |
| Kelp-specific Area or percent cover of kelp canopy Height, density, biomass, or survivorship of individuals Recruitment of juveniles Presence/quantity of reproductive tissue (i.e., Sori/sporophylls) Indicators of health (e.g., fouling, pigmentation) |
| Associated community Mobile organisms (e.g., fishes, large invertebrates) Sessile and/or benthic organisms (e.g., other seaweeds, sessile invertebrates) Epiphytes, micro-organisms Particular species-of-interest Positive (e.g., commercially valuable species) Negative (e.g., destructive grazers/herbivores) Community production (i.e., nutrient and carbon cycling) |
| Environmental/physical factors Hydrodynamics (e.g., water flow, currents, wave action) Subcanopy light levels Sedimentation Turbidity Water quality (especially nutrient levels) Water temperature |
