Multidisciplinary baselines quantify a drastic decline of mussel reefs and reveal an absence of natural recovery

The onset of the Anthropocene has coincided with enormous global declines in natural ecosystems, leading to losses in the valuable goods and ecosystem services they provide. This global decline, in conjunction with growing recognition of the ecological importance of natural ecosystems, has generated a pressing need for restoration. Effective ecosystem restoration relies on accurate identification of the cause of decline and clear metrics of success, which are only possible with baseline data of both the pre-degradation and pre-restoration ecosystems. However, the establishment of these baselines can be difficult as different potential information sources each have benefits and drawbacks. Determining an efficient method to balance these diverse information sources and generate robust baselines is vital to achieving the United Nations ’ goal of massively scaled-up ecosystem restoration. Here we expand on the concept of multidisciplinary base-lines, or the combined use of sources and methods across a wide disciplinary spectrum to establish comprehensive and reliable ecosystem baselines, and use mussel reefs in the South Island of New Zealand as a test case. Using a combination of comprehensive historical review, extensive shoreline surveys, and local ecological knowledge, we demonstrate that local mussel abundances decreased by 97% since the mid-1960s as a result of overharvesting, leaving the extant populations scattered, small, and without recovery. This study demonstrates that harnessing multidisciplinary baselines allows for the consolidation of qualitative and quantitative estimates of ecosystem change over hundreds of years, as well as confirmation of causes

tion of the cause of decline and clear metrics of success, which are only possible with baseline data of both the pre-degradation and pre-restoration ecosystems.
However, the establishment of these baselines can be difficult as different potential information sources each have benefits and drawbacks. Determining an efficient method to balance these diverse information sources and generate robust baselines is vital to achieving the United Nations' goal of massively scaled-up ecosystem restoration. Here we expand on the concept of multidisciplinary baselines, or the combined use of sources and methods across a wide disciplinary spectrum to establish comprehensive and reliable ecosystem baselines, and use mussel reefs in the South Island of New Zealand as a test case. Using a combination of comprehensive historical review, extensive shoreline surveys, and local ecological knowledge, we demonstrate that local mussel abundances decreased by 97% since the mid-1960s as a result of overharvesting, leaving the extant populations scattered, small, and without recovery. This study demonstrates that harnessing multidisciplinary baselines allows for the consolidation of qualitative and quantitative estimates of ecosystem change over hundreds of years, as well as confirmation of causes of ecosystem degradation, and clear documentation of current ecosystem state beyond what is possible from any individual source. This approach to establishing ecosystem baselines also provides valuable avenues for the advancement of restoration by quantifying the temporal and geographic scales of ecosystem decline, identifying areas for intervention, and establishing clear metrics of success.
K E Y W O R D S interviews, local ecological knowledge, mussels, New Zealand, Perna canaliculus, shellfish INTRODUCTION Earth's natural ecosystems have suffered devastating declines in recent centuries including the loss of over 30% of forests, 85% of wetlands, and 50% of coral reefs (IPBES, 2019) largely as a result of anthropogenic impacts including land-use changes, resource extraction, and climate change (Hautier et al., 2015;Jackson et al., 2001). The loss of these natural ecosystems underpins the global biodiversity crisis (Dirzo et al., 2014) and has compounding consequences for the goods and services on which humans rely (Dobson et al., 2006). Increased recognition of the magnitude of these losses and the importance of natural ecosystems has generated urgent calls for conservation efforts that effectively combat decline including large-scale ecosystem restoration as highlighted by the declaration of 2021-2030 as the UN Decade of Ecosystem Restoration (UN Resolution 73/284).
Before any restoration effort can begin, it is firstly vital to establish what has been lost and what remains of the targeted ecosystem. To determine historical losses, restoration typically relies on the use of a reference model or pre-impact level, which can be compared with the degraded ecosystem to determine the extent of degradation and to the restored ecosystem to monitor recovery (Gann et al., 2019;SER, 2004). This pre-impact model establishes specific historical benchmarks or baselines (e.g., area covered, total abundance, or biomass) to facilitate direct, quantifiable comparisons (Wortley et al., 2013). In addition to historical or pre-impact baselines, restoration also relies on pre-restoration baselines, which reflect the state of the ecosystem after degradation but before the initiation of any restoration efforts (Gann et al., 2019). Utilizing these two baseline models, pre-impact and pre-restoration, provides concrete data and allows for clear, replicable measurements of success. In turn, these metrics of success facilitate the assessment of restoration effectiveness and influence how future resources are allocated, including funding and research efforts (Ruiz-Jaen & Aide, 2005). Ultimately, ensuring that baseline data are accurate, comprehensive, and accessible advances the effectiveness of restoration through the assessment of success and the facilitation of evidence-based resource allocation.
Traditionally, pre-impact baseline data for restoration are obtained from historical records or prior scientific surveys of the area of interest (SER, 2004;Wortley et al., 2013). In areas with robust historical data, this can be an effective method; however, restoration is often planned in areas where existing historical data are incomplete, scattered, or nonexistent. For example, deforestation in Western Europe was underway well over a millennium ago (Kaplan et al., 2017) and even smaller scale ecosystems can cover large timescales like oyster reefs in Chesapeake Bay, which began to decline almost 300 years ago (Jackson et al., 2001). In situations like these, pre-impact scientific data are nonexistent and even nonscientific historical records are rare. Relying solely on written historical data in such circumstances risks developing an incomplete baseline and impeding restoration efforts (Gann et al., 2019).
Local ecological knowledge, or information garnered from extensive observation of an area (Huntington et al., 2004), is increasingly being used as a valuable data source for establishing both pre-impact and pre-restoration environmental baselines (Lee et al., 2019;Taylor et al., 2011;Thornton & Scheer, 2012). Local ecological knowledge is a useful tool for establishing baselines as it provides data beyond scientific or historical records, builds connections with community members, and allows for consultation with groups traditionally excluded from scientific record-keeping (Fitzsimons et al., 2020). Despite these benefits, local ecological knowledge is often underutilized in restoration initiatives. For example, fewer than 5% of studies in a systematic review of shellfish reef restoration incorporated local ecological knowledge or any other social data into the study design (Toone et al., 2021). The use of local ecological knowledge can also have drawbacks, including difficulty in the identification of suitable local knowledge "experts" and the reliability of data across timescales (Davis & Wagner, 2003;Martinez-Levasseur et al., 2017). Shifting baseline syndrome, or the phenomenon by which environmental information is lost over time as people do not notice or accurately quantify changes in their surroundings (Jones et al., 2020;Pauly, 1995), can be a particularly challenging aspect of using local ecological knowledge to trace environmental decline. Such phenomena, like interview participants inaccurately recalling the true extent of historical ecosystems of interest, appear to occur commonly in practice (Papworth et al., 2009;Venkatachalam et al., 2010).
A range of methods can be used for establishing environmental baselines (e.g., historical records, scientific surveys, local ecological knowledge) and restoration projects typically select a method that suits the experience of their team, local setting, or logistical ability. However, each method in isolation represents only one portion of the ecological past and present of an area, resulting in incomplete baselines that are unable to comprehensively guide restoration projects. Terrestrial ecologists interested in mapping historical vegetation have demonstrated that multidisciplinary combinations of these data sources can provide robust historical baseline data (Duncan et al., 2010;Fensham, 1989;Trueman et al., 2013); however, this method remains underutilized, particularly in the marine realm (Shackelford et al., 2021). This study hypothesizes that multidisciplinary baselines generated from comprehensive historical review, local ecological knowledge, extensive surveys of current populations, and other complementing methods can be combined to broaden our understanding of degraded ecosystems ( Figure 1A). This hypothesis is tested in Kenepuru Sound, a drowned river valley at the top of New Zealand's South Island that historically supported extensive mussel reefs that are reportedly largely absent now (Handley, 2017). Mussel reefs are ecosystem engineers, building complex habitats (Benjamin, Handley, Hale, et al., 2022;Borthagaray & Carranza, 2007;Sea et al., 2022) and providing valuable ecosystem services, including water filtration (Bayne, 1976), sediment stabilization (Meadows et al., 1998), denitrification Sea et al., 2021), and the production of food for humans (Wijsman et al., 2018). The decline of local wild mussel populations and these ecosystem services has F I G U R E 1 Multidisciplinary baselines methodological framework (A) and application to Kenepuru Sound (B). Multidisciplinary baselines should cover a wide range of sources and methods but are adaptable to the local context of the targeted ecosystem and the examples provided here are not exhaustive. In Kenepuru Sound, nonscientific historical reports established qualitative baselines for over a century prior to harvesting, historical scientific surveys were used just prior to and during harvesting, harvesting records were used to determine the extent of overharvesting, local ecological knowledge provided baseline information pre-, during, and post-harvesting, and recent surveys provide a current, post-harvesting and pre-restoration baseline.

ECOSPHERE
3 of 12 led to calls for restoration from residents and the wider community; however, there are inadequate baseline data from which to direct restoration efforts (Handley, 2017). In this study, the historic decline of mussel reefs in Kenepuru Sound is evaluated as a test case for the use of multidisciplinary baselines, ultimately characterizing historical mussel populations prior to degradation, determining the status of remnant populations, and building a baseline against which to measure progress of future restoration efforts.

Area description
Kenepuru Sound is an inner branch of the Pelorus Sound, which is one of a series of drowned river valleys that form the Marlborough Sounds region in the north of New Zealand's South Island (Urlich & Handley, 2020). Kenepuru Sound once supported extensive intertidal and subtidal reefs of the endemic green-lipped mussel (Perna canaliculus; Handley et al., 2017). Green-lipped mussels are found throughout New Zealand and occur in intertidal and shallow subtidal waters, typically below mean sea level where they most often attach to rocky reefs, boulders, and rubble (Jeffs et al., 1999). They are often found in high-density aggregations, known as mussel reefs or beds. In Kenepuru Sound, mussel reefs were intensively commercially harvested by handpicking from the shore and dredging from the seabed during the 1960s and 1970s (Dawber, 2004;Hickman, 1980). This harvesting is anecdotally reported to have been responsible for depleting wild mussel reefs, although the exact extent of this decline is unknown.

Historical population
A combination of scientific and nonscientific historical sources was used to establish pre-harvesting baseline mussel population data for Kenepuru Sound ( Figure 1B). To establish qualitative baselines prior to the 20th century, historical reports specific to mussels were extracted from a literature review of ecosystem decline in the Marlborough Sounds (Urlich & Handley, 2020) and from digitized newspaper articles held by the National Library of New Zealand and published in the Marlborough Daily Times, Marlborough Press, Marlborough Express, or Pelorus Guardian and Miners' Advocate, which were in print between 1860 and 1920. Articles were identified using the keywords "mussel(s)," "shellfish," or "k utai," the term used for mussels in Te Reo M aori, the indigenous language of New Zealand.
For more recent pre-impact data, a series of scientific surveys of mussel populations undertaken in Kenepuru Sound in 1968 and 1969 were analyzed (Flaws, 1975;Stead, 1971aStead, , 1971b. These previous surveys recorded subtidal and intertidal mussel abundances and sizes through dredging, shoreline observations, and scuba diving. Collectively, the surveys covered about half of the shoreline of Kenepuru Sound; however, they were conducted concomitantly with commercial mussel harvesting and, therefore, do not provide comprehensive pre-degradation baseline mussel abundances. To address this omission and supplement these surveys, mussel harvesting data for Kenepuru Sound were obtained from annual fishery reports including both the subtidal dredging industry and intertidal handpicking (King, 1985;Paul, personal data, presented as regional data in Francis & Paul, 2013;Paul, 2012). These data were originally recorded as sacks of harvested mussels, which were converted to tonnage following the 79 kg per sack ratio for the South Island used by Francis and Paul (2013). Finally, this tonnage was transformed into number of mussels harvested following the conversion for Kenepuru Sound mussels during the peak of the harvesting industry in 1969 of one million mussels per 434 t (Stead, 1971b). This conversion allowed for an estimation of the minimum pre-harvesting mussel abundance in Kenepuru Sound as any mussels harvested during this period would have been sourced from wild populations.

Remnant population
Intertidal green-lipped mussels were surveyed for distribution, size, and abundance along 73 km of Kenepuru Sound from July to November 2020. This survey distance reflects over 80% of the Kenepuru shoreline and all areas with available road and foot access. Surveys were conducted over 22 days, across 4 months, during spring tides when low tide elevation was at least 1 m below mean sea level (tidal range~3 m). Each survey took place over 2 h and centered around the time of low tide. The shoreline was divided into subsections during each survey with a new subsection beginning if the dominant substrate changed or if there was a substantial change to mussel density, resulting in a total of 331 subsections overall. Surveys were conducted by walking along the shoreline and counting every mussel within each subsection. The average mussel density for each subsection was calculated after the surveys by dividing the total mussels counted in each subsection by the length of the subsection. Across the entire survey, 400 mussels were haphazardly selected and measured with calipers along the anterior-posterior axis of the shell to provide mussel size data. To analyze changes in mussel sizes between current and historical populations, the size frequency distribution recorded in the 2020 survey was statistically compared with the distribution recorded in the 1969 survey using a two-sample Kolmogorov-Smirnov test conducted in R (R Core Team, 2022).
Subtidal green-lipped mussel presence was recorded via a drop camera survey conducted in May 2021. An underwater camera was lowered to the seafloor at 44 locations throughout Kenepuru Sound with preference given to areas with historically recorded mussel reefs. The locations represented a variety of benthic substrates from mud to rock and were between 3 and 18 m from the surface. At each location, a video recording of up to 5-min duration was made while the support vessel drifted and the recording was analyzed by counting visible mussels recorded over the duration of the video to enable quantification of mussel abundance.

Interviews
Interviews with 13 long-term (50+ years) residents of Pelorus Sound were conducted between September and December 2020. The greater Pelorus region was chosen for residents rather than strictly Kenepuru Sound to allow for additional interview participants who were familiar with the study area but currently lived elsewhere in the region. As this population was not concentrated in a single institution or industry, chain-referral or "snowball" sampling was used whereby primary participants suggested other potential participants (Drescher et al., 2013). Kenepuru Sound has a small, rural population, therefore the pool of potential participants was limited. However, the population is well connected; therefore, chain referral began with three individuals identified by community members familiar with the project and continued until no new individuals were suggested by participants. In total, the interview participants consisted of six former commercial or recreational mussel handpickers, four current or former mussel farmers, and three residents with no affiliation to the mussel aquaculture or harvesting industries. While chain-referral sampling has been criticized for potentially leading to selection bias of like-minded individuals (Davis & Wagner, 2003), it is the preferred method in situations where desired information is very localized with a limited number of individuals with in-depth knowledge (Bart, 2010;Drescher et al., 2013).
Participants were contacted and prescreened via initial phone calls or emails to determine suitability for a full interview (i.e., current or former residents of Pelorus Sound for at least 50 years). Full interviews were conducted in person except for two cases where logistical constraints (i.e., remote dwellings) necessitated phone interviews. Interviews lasted 60-90 min and followed a questionnaire (Appendix S1), although responses were left open-ended to allow for unexpected findings (Neuman, 2014). The questionnaire was used to confirm each participant's history in the area, detail their knowledge and memories of mussels in Kenepuru Sound, and ask about any other environmental changes they have observed. Relevant maps of Kenepuru Sound and mussel shells of assorted sizes (40-190 mm of shell length) were provided as visual aids. Interviews were recorded for later analysis. Participant responses were sorted into categories based on four primary questions: (1) whether they perceived a decline in mussel populations within their experience; (2) what they believed to be the causes of any decline; (3) whether populations were recovering or in continued decline; and (4) the perceived cause of this decline or recovery. Participants provided written consent (Appendix S2) and were offered the opportunity to withdraw from the study at any time. The interview design and process were approved by the University of Auckland's Human Participants Ethics Committee (UAHPEC) on 25 August 2020 (Ref: UAHPEC2564).

Historical population
Early nonscientific reports confirm high mussel abundances in the Kenepuru Sound area. The earliest primary accounts of mussels in the Marlborough Sounds report "muscles [sic] and sorts of shellfish in great plenty" as early as 1770 (Parkinson, 1770, p. 114 cited in Beaglehole, 1962, p. 453), although archeological evidence and secondary reports confirm M aori subsistence harvesting as early as the 14th century (Seersholm et al., 2018;Wadsworth, 2015). Mussels were still abundant by the 19th century when there are reports of "enormous mussel shells which are to be found in such abundance in the Sound[s]" (Marlborough Press, 1865) and mussel reefs with "apparently no end" (Marlborough Express, 1889). Harvesting by local M aori populations was also still common including as gifts (koha) for presenting to inland tribes (iwi; Marlborough Express, 1902), and for feasting during celebrations (h akari) and for funeral gatherings (tangi; Marlborough Press, 1881), although reports of European consumption at the time are largely limited to stranded sailors (e.g., Pelorus Guardian and Miners Advocate, 1898).
The first scientific surveys indicated that in 1969 over 70 locations in Kenepuru Sound were recorded with "harvestable quantities" of dense mussel reefs defined as a minimum of 30 mussels m −2 (Figure 2; Flaws, 1975;Stead, 1971a). Mussel harvesting data drawn from annual fishery reports (Francis & Paul, 2013;King, 1985;Paul, 2012) of handpicking and dredging revealed that the commercial harvesting industry in Kenepuru Sound was largely restricted to a six-year period from 1968 to 1973 during which time 2159 t or over 4.9 million mussels were harvested from Kenepuru Sound. The industry declined rapidly thereafter as an average of only 130 t of mussels were harvested per year in the decade following 1973. In full, at least 4.9 million mussels were present in Kenepuru Sound prior to 1968 and the advent of widespread commercial harvesting.

Remnant population
In total, 107,932 intertidal mussels were recorded over the 73 km of the Kenepuru Sound shoreline surveyed in 2020. Average mussel densities were low throughout Kenepuru Sound with a median density of just 0.6 mussels m −2 . Overall, there were fewer than 10 mussels m −2 for 97.1% of the shoreline (71.1 km) and fewer than 1 mussel m −2 for 71.6% of the shoreline (52.5 km; Figure 3). Only 0.3% of the shoreline (0.2 km), representing three shoreline subsections, reached 30 mussels m −2 , historically regarded as the density suitable for commercial harvesting (Figure 3; Stead, 1971a). With regard to mussel shell lengths, the size frequency distribution was significantly different between the historical (1969) and current (2020) populations (p < 0.001). The mean shell length of the 400 intertidal mussels measured in the 2020 population was 106 ± 26 mm, a 28 mm decrease among adult mussels compared with historical measurements (Figure 4). Only 0.5% of measured mussels were juveniles, that is, below 30 mm in length, compared with 14.8% of historical mussels. Additionally, large mussels (i.e., over the historic legal harvest size of 140 mm) comprised only 11% of the surveyed mussels compared with 56% in the historical surveys. No mussels were recorded subtidally from any of the video recordings from the 44 sites within Kenepuru Sound (Figure 3).

Interviews
All 13 interview participants (hereafter "residents") confirmed a local decline in mussel populations and mussel sizes within the time frame of their memories of the area, primarily in the late 1960s and the 1970s. Residents F I G U R E 2 Kenepuru Sound and abundant mussel locations. The black outline delineates the area surveyed in the present 2020 study. Triangles indicate areas with >30 mussels m −2 in historical surveys (Flaws, 1975;Stead, 1971a). Circles indicate areas with >30 mussels m −2 in current shoreline surveys. Inset maps display the location of Kenepuru Sound within the greater Marlborough Sounds and New Zealand. F I G U R E 4 Mussel shell lengths recorded in shoreline surveys. The dashed line reflects mussel shell length from historical surveys in 1969 (n = 168; Flaws, 1975); the solid line reflects mussel shell length from the current shoreline survey in 2020 rounded to the nearest centimeter (n = 400). ECOSPHERE 7 of 12 reported that mussels were once widespread in the area, covering extensive areas of Kenepuru Sound with dense reefs throughout the subtidal and intertidal zones. Residents unanimously identified overharvesting, specifically commercial handpicking and dredging, as the cause of this mussel decline. Residents also recalled that overharvesting and competition with farmed mussels drove commercial handpicking to lose profitability and largely cease by the mid-1970s. Residents did not perceive any recovery of mussel reefs in the 45 years following the cessation of wild mussel harvesting. Residents identified many potential causes for this lack of recovery and often suggested multiple causes, but six overarching factors emerged as being commonly identified ( Figure 5). Loss of settlement surfaces (i.e., the extirpation of algae that enable early settlement and the burial of hard substrates that facilitate later settlement) was the most commonly cited reason for the lack of recovery, being identified by 6 of the 13 residents. Sedimentation of the shoreline preventing reestablishment of mussel reefs and competition with farmed mussels for phytoplankton and other resources were the next two most frequently identified reasons, each proffered by five residents. Predation, specifically from fish, like snapper (Pagrus auratus), on mussel larvae and juveniles, was also proposed by four residents as a leading barrier to recovery. Finally, a lack of existing reef structures and pollution including plastics and chemicals were each suggested by three residents as the primary obstacles to recovery.

DISCUSSION
Effective conservation and restoration rely on an understanding of ecosystem changes over time. A comprehensive understanding of these changes allows managers to determine whether ecosystems are in decline and which interventions are needed for their maintenance or recovery (Gann et al., 2019). Multidisciplinary baselines provide a method to establish these ecosystem changes by consolidating historical information across a range of sources that can be combined in a manner to best fit the local context. For example, by combining historical scientific and nonscientific reports, harvesting records, and local ecological knowledge from interviews with residents, this study has uncovered qualitative descriptions of abundant local mussel reefs dating back centuries as well as quantitative estimates of a historical population F I G U R E 5 Potential factors behind lack of mussel recovery in Kenepuru Sound as identified through local ecological knowledge. The size of each box corresponds to the number of interviewed residents identifying that factor as a barrier to mussel recovery. Residents were unprompted and could identify any number of potential factors.
of close to 5 million mussels prior to degradation and firsthand accounts of the extent of ecosystem decline. Determining restoration success is challenging and often complicated by unclear definitions of success (Zedler, 2007) and a lack of reference systems (Suding, 2011), but pre-impact multidisciplinary baselines like these provide clear and quantifiable metrics for restoration managers and facilitate accurate measurements of ecosystem recovery.
Understanding the underlying causes of ecosystem degradation is also essential to ensuring current and future populations are protected from further declines (Gann et al., 2019). Multidisciplinary baselines not only allow for the identification of causes of ecosystem decline but also allow for multiple confirmations of the same information, which improves reliability and minimizes the impact of biased data. For example, the use of multidisciplinary methods in this study demonstrated that overharvesting from dredging and handpicking in the 1960s and 1970s was responsible for the decimation of mussel reefs not only through historical descriptions of harvesting but also through detailed landing records and interviews with former mussel harvesters. Pinpointing the direct cause of ecosystem decline has resulting implications for ecosystem recovery. For example, in Kenepuru Sound, commercial wild mussel harvesting ended over four decades ago, suggesting that future efforts can focus on restoration and factors preventing ecosystem recovery rather than on eliminating the initial cause of decline. Ultimately, this aspect of multidisciplinary baselines is most useful in areas where the initial cause of decline is only suspected or entirely unknown, for example, ecosystems facing multiple forms of degradation (Paine et al., 1998) or those for which thresholds have been exceeded leading to hysteresis (Baskett & Salomon, 2010;Cowan et al., 2008).
In addition to documenting historical knowledge and initial causes of ecosystem decline, multidisciplinary baselines allow researchers to assess the current state of the targeted ecosystem. In Kenepuru Sound, extensive surveys of the seabed and shoreline revealed a current population of just under 108,000 mussels spread throughout the Sound. These data, combined with the historical estimates, revealed that current mussel populations in Kenepuru Sound are at less than 3% of historic levels and only reach historic densities on less than 1% of the intertidal shoreline with no populations recorded in the deeper subtidal. This multidisciplinary approach can be utilized to gather information beyond just population numbers, for example, the surveys conducted in Kenepuru Sound revealed a 28 mm decrease in adult mussel shell lengths between 1969 and 2020. This decrease is potentially due to the lack of existing reef structures that protect mussels from heat and predator stress resulting in more energy available for shell growth (Bertolini et al., 2018). Information on the current status of ecosystems is central to restoration efforts as it provides insight into the suitability of the current environment to support the targeted ecosystem, the locations of remnant populations, and estimates of total ecosystem decline. Upon securing this information, conservation and restoration managers can establish evidence-based protective areas around remnant populations, quantify the necessary extent of restoration, and identify other important characteristics (like shell length for mussels in Kenepuru Sound) to monitor over time. For example, mussel restoration efforts in Kenepuru Sound have used these baseline data to select restoration sites and determine metrics of success .
Finally, multidisciplinary baselines generate valuable data beyond historical and present timescales by providing data on the potential future of an ecosystem, specifically by identifying potential barriers to ecosystem recovery and methods to ameliorate these barriers. Natural recovery is the ultimate goal of restoration as it is more efficient and cost-effective than restoring an entire ecosystem via transplantation or reseeding (SER, 2004). However, natural recovery can be elusive as post-degradation environments may no longer be capable of supporting mechanisms like dispersal or survival that were possible pre-degradation (e.g., de Paoli et al., 2015;Filbee-Dexter & Scheibling, 2014;Moksnes et al., 2018). In Kenepuru Sound, this study revealed a lack of natural recovery despite commercial harvesting ceasing four decades ago. Local residents reported that mussel reefs have not returned to historic densities as confirmed by surveys of current populations, which also found almost no juvenile mussels, indicating the absence of recruitment events in at least the preceding two years and potentially much longer. Using local ecological knowledge in concert with these surveys of current populations allowed residents to suggest potential causes for lack of recovery, specifically loss of settlement surface, competition from marine farms, sedimentation, predation, lack of reef structure, and pollution. Consolidating local knowledge and environmental observations like this provides conservation and restoration managers with clear starting points to initiate natural recovery. Potential barriers to natural recovery uncovered by generating multidisciplinary baselines can each be validated through experimental testing to determine which factors are most detrimental to recovery and which techniques effectively resolve them.
Multidisciplinary baselines can be considered a toolbox to uncover the past, present, and future of a degraded ecosystem. Just as every construction project does not require the use of every tool in a toolbox, every ecosystem does not need every potential method for establishing baselines. Techniques can be added or removed based on the specific context of the localized situation, but it is ultimately the combination of sources and methods that makes multidisciplinary methods comprehensive and reliable. Local ecological knowledge, historical analysis, literature reviews, current scientific surveys, indigenous knowledge, and archeological evidence are all examples of potential sources of valuable baseline data that should be assessed in conjunction to form multidisciplinary baselines far more robust than possible from any individual method. Selecting which methods to use in a specific context should be motivated by first determining which sources are available in a given area. For example, in understudied areas prior scientific surveys may not exist, but local ecological knowledge may be able to supplant them. Next, potential methods should be assessed for their robustness and extent of geographic and temporal coverage. The ultimate selection of methods should prioritize sources that either cover the largest geographic and temporal scales or provide coverage of regions and/or times not covered by other sources. Finally, reporting the results of these disparate data sources together rather than as siloed efforts allows for the findings to be compared and contrasted directly and encourages dialogue between different disciplines and knowledge holders. If the UN Decade on Ecosystem Restoration is to be a success then scientists and practitioners need to learn from this research and draw upon the full toolbox of available resources to generate comprehensive baselines and, in turn, a firmer foundation for restoration.