64 GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned Peter T. Harris1, Elaine K. Baker2 1 Marine and Coastal Environment Group, Geoscience Australia, Canberra, ACT, Australia 2 UNEP/GRID- Arendal School of Geosciences, University of Sydney, Australia Abstract This chapter presents a broad synthesis and overview based on the 57 case studies included in Part 2 of this book and on questionnaires completed by the authors. The case studies covered areas of seafloor ranging from 0.15 km2 to over 1,000,000 km2 (average of 26,600 km2) and a broad range of geomorphic feature types. The mean depths of the study areas ranged from 8 to 2,375 m, with about half of the studies on the shelf (depth ! 120 m) and half on the slope and at greater depths. Mapping resolution ranged from 0.1 to 170 m (mean of 13 m). There is a relatively equal distribution of studies across the four naturalness categories: near pristine (n " 17), largely unmodified (n " 16), modified (n " 13), and extensively modified (n " 10). In terms of threats to habitats, most authors identified fishing (n " 46) as the most significant threat, followed by pollution (n " 12), oil and gas development (n " 7), and aggregate mining (n " 7). Anthropogenic climate change was viewed as an immediate threat to benthic habitats by only three authors (n " 3). Water depth was found to be the most useful surrogate for benthic communities in the most studies (n " 17), followed by substrate/sediment type (n " 14), acoustic backscatter (n " 12), wave-current exposure (n " 10), grain size (n " 10), seabed rugosity (n " 9), and bathymetric/topographic position index (BPI/TPI) (n " 8). Water properties (temperature, salinity) and seabed slope are less useful surrogates. Multiple analytical methods were used to identify surrogates, with ARC GIS being by far the most popular method (23 out of 44 studies that specified a methodology). Of the many purposes for mapping benthic habitats, four stand out as preeminent: (1) to support government spatial marine planning, management, and decision making; (2) to support and underpin the design of marine protected areas (MPAs); (3) to conduct scientific research programs aimed at generating knowledge of benthic ecosystems and seafloor geology; and (4) to conduct living and nonliving seabed resource assessments for economic and management purposes. Out of 57 case studies, habitat mapping was intended to be part of an ongoing monitoring program in 24 cases, whereas the Seafloor Geomorphology as Benthic Habitat. DOI: 10.1016/B978-0-12-385140-6.00064-5 © 2012 Elsevier Inc. and Crown Copyright. Published by Elsevier Inc. All rights reserved. 872 Seafloor Geomorphology as Benthic Habitat mapping was considered to be a one-off exercise in 33 cases. However, out of the 33 one-off cases, the authors considered that their habitat map would form the baseline for monitoring future changes in 24 cases. This suggests that governments and regulators generally view habitat mapping as a useful means of measuring and monitoring change. In terms of the perceived clients and users of habitat maps, most authors considered industry, marine conservation, and the scientific community to be the primary users of habitat maps. However, the overwhelming majority of habitat surveys were funded by government or government-funded agencies/institutions (n ! 49), with only minor funding from private industry (n ! 7) or nongovernment organizations (n ! 4). A gap analysis (i.e., geomorphic features and habitats not included in the case studies) illustrates that whereas shelf and slope habitats are well represented in the case studies, estuarine and deltaic coastal habitats plus deep ocean (abyssal–hadal) environments were described in only a few case studies. Geographically, about half of the case studies were from waters around Western Europe, while the margins of the continents of Africa, Asia, and South America were not represented in any case study. Given the intense pressures facing benthic habitats and broad regional differences in ecosystems, species, and habitats, future case studies from these regions should be specifically sought for future editions of the Atlas. Key Words: Naturalness, surrogates, spatial marine planning, habitat classification schemes, habitat mapping clients, environmental data Introduction This chapter presents a broad synthesis and overview based on the 57 case studies presented in Part 2 of this volume. Authors were asked to prepare their case studies using a template that specified required information. To be accepted, case studies had to contain both geomorphic and biologic data, contain a clear description of at least one geomorphic feature type, describe the oceanographic setting, and provide an assessment of the naturalness of the environment. The spatial comparison of biological data with spatial physical data is a key element of every case study. Authors were given the opportunity to describe surrogacy relationships and methods used to identify and quantify them. On submitting their final papers, authors completed a questionnaire1 to provide specific details of their study area and other issues ranging from surrogates to data storage. Based on the content of the case studies, together with the authors’ responses to the questionnaire, the following topics are considered in the present synthesis of information: 1. 2. 3. 4. 5. 6. 1 study area attributes, including naturalness and anthropogenic threats; surrogates used (and not used) and spatial analysis techniques; the use of published habitat classification schemes; the purpose of habitat mapping (sectors and clients); consultation in habitat survey planning; data storage and accessibility; The questionnaire is reproduced in Appendix 1 of this chapter. GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 873 7. gap analysis; 8. best practices for habitat mapping. Not all parts of the template, nor every question in the questionnaire, are applicable to every case study, and there is more than one possible response to questions in some cases (e.g., multiple mapping methods were used, several or no surrogacy methods used, multiple stakeholders, multiple funding sources). Nevertheless, the authors’ responses provide an insight from the perspective of the GeoHab community2 on the purpose, scope, and value of habitat mapping for the benefit of the broader community. In the following, we present the responses from the case study authors as a series of graphs and statistical statements. These are followed by our interpretation of the findings with some possible explanations. Unless otherwise stated, the interpretations and explanations provided are from the authors of this chapter and are not necessarily the views of the case study authors. Attributes of Case Studies The case studies covered areas of seafloor ranging from 0.15 km2 to over 1,000,000 km2 (average of 26,600 km2) and a broad range of geomorphic feature types. The mean depths of the study areas ranged from 8 to 2,375 m, with about half of the studies on the shelf (depth !120 m) and half on the continental slope and at greater depths (depth "120 m). Mapping resolution ranged from 0.1 to 170 m (mean of 13 m). The naturalness of the environment examined in each case study was described by the authors in their own terms. For the purposes of this discussion, we assigned each study to one of four categories: near pristine, largely unmodified, modified, and extensively modified. The results (Figure 64.1) illustrate that there is a good distribution of studies among the four naturalness categories. As might be expected, the modified and extensively modified case studies tended to be located in shallow water depths, on the shelf or in the coastal zone, whereas largely unmodified and near-pristine environments were located in more remote deep-water (slope and abyssal) areas. In describing the naturalness of study areas, authors specified the main anthropogenic threats that they perceived as having an immediate impact upon the habitats they studied. In several cases, habitat mapping was specifically undertaken in order to manage specific pressures from one or more human uses. The case studies in this book identified the impacts of fishing as the most significant threat to benthic habitats (Figure 64.2). Fishing has four times the number of citations as the second most perceived threat (pollution). Pollution includes the effects of marine debris, which was the most commonly cited form of pollution. Oil and gas development, aggregate mining, and coastal development are perceived as also being among the more significant threats (Figure 64.2). 2 The “GeoHab community” for the purposes of this discussion encompasses all participants in past and future GeoHab conferences and publications (including this book) and those engaged in habitat-mapping activities. 874 Seafloor Geomorphology as Benthic Habitat Anthropogenic climate change is viewed as an immediate threat to benthic habitats by only three authors (Chapters 21, 24, and 34; Figure 64.2). This is not necessarily evidence that climate change is not a significant threat to the marine environment in the longer term, but it does indicate that the experts working in the field of habitat mapping are currently documenting the harmful consequences of other human pressures (particularly fishing and pollution) on the condition and status of the marine environment. Future impacts of anthropogenic climate change will Naturalness Near pristine Largely unmodified Modified Extensively modified 0 5 10 15 20 Number of studies Figure 64.1 Graph showing the naturalness of environments in the case studies. 47 Fishing Threats to habitats Pollution 12 Aggregate mining 7 Other mining, oil and gas 7 Coastal development 6 Tourism 6 5 Cables Shipping 4 Invasive species 3 Climate change 3 Wind farms 3 0 10 20 30 40 Number of times cited in studies Figure 64.2 Graph showing the anthropogenic pressures cited in the case studies. 50 GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 875 be in addition to damage already suffered by marine habitats due to the cumulative impacts of pressures (fishing, pollution, mining, etc.) that have occurred over the past century (see Chapter 3). Surrogates Used (and Not Used) and Spatial Analysis Techniques Three questions posed in the questionnaire were related to the variables that were mapped in each study area and their usefulness as surrogates for the occurrence of benthic biota. The extent to which a particular surrogate performed was generally assessed via various statistical tests and multivariate analysis techniques. The most commonly measured variables in the case studies were (in decreasing order of citation frequency): depth (measured in all studies), sediment grain size, acoustic backscatter, seabed slope, substrate or seabed type, and rugosity (Figure 64.3). Each of these parameters is cited by over 20 authors as having been measured as part of the case study. However, not all measured variables were found to be useful surrogates for the occurrence of benthic organisms. This was determined using a range of analytical methods to identify surrogates and their predictive power. ArcGIS is by far the single most popular tool (23 out of 44 studies that specified a methodology). However, a Surrogates Success Rate (%) All 57 studies Depth 30 Substrate/sediment type 61 Backscatter 35 Wave-current speed, exposure 67 Grain size 29 Rugosity 39 TPI, BPI 57 Aspect 100 Primary production 25 Light attenuation 100 Sorting 100 Slope Water properties Emergence time Morphology Sedimentation rate Sea ice Curvature Volcanology Most useful Measured 0 5 10 15 20 25 30 35 40 Number of citations Figure 64.3 Comparison of the number of times variables were measured (blue) versus variables found to be the most useful surrogates for the occurrence of biota (yellow). Water depth was measured in all 57 case studies. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this book.) 876 Seafloor Geomorphology as Benthic Habitat number of different multivariate analysis methods are also used (24 studies), often in parallel with GIS. Of the various multivariate analysis software packages, PRIMER software and its incorporated tools were most commonly used to find relationships between physical and biological data, although other software was used to a lesser extent (see Table 5.2 in Chapter 5). The most useful surrogate for benthic communities was found to be water depth, followed by substrate/sediment type, acoustic backscatter, wave-current exposure, grain size, seabed rugosity, and BPI/TPI. It is interesting to note the variability in the “success rate” of surrogates (i.e., the number of times a parameter was found to be useful divided by the times it was measured). Some surrogates were measured in only a few studies but were found to be useful in every case [100% success rate; e.g., aspect (Chapter 29), light attenuation (Chapter 24), and sorting (Chapter 42); Figure 64.3]. Wave-current speed was measured in 15 studies and was found to be a useful surrogate in 10 of them (67% success rate), giving this parameter a higher success rate than depth, substrate type, rugosity, TPI/BPI, backscatter, or grain size (Figure 64.3). There are a number of possible explanations as to why there is a discrepancy between the variables measured versus those considered by authors to be the most useful surrogates (Figure 64.3). Indirect variables such as depth and backscatter correlate with several direct variables, which possibly explain their overall good performance as surrogates. For example, backscatter correlates with sediment grain size, seabed rugosity, and substrate type, which are direct variables that individually performed well as surrogates (Figure 64.3). Some variables, such as emergence time, sea-ice presence, and light attenuation, are not broadly applicable to all marine habitats; they apply only to a small subset. Hence, they were not given an equal opportunity in this survey, which covers all of the case studies. It may also be argued that different surrogates will be more useful in some environments than in others. For example, water properties such as salinity are more likely to be important surrogates in estuaries than in the open marine environment (and only five of the case studies in this book were from estuarine or fjord environments; Chapters 11, 12, 18, 19, and 20). In the open marine environment, water properties vary most significantly over broad spatial and temporal scales, so unless the case study encompasses a broad area (or data are collected over a long time series) variability in water properties may not be large enough to register as a useful surrogate. Some data sets that have already been collected by different agencies over a period of time prior to a survey may be freely available at no extra cost and can be incorporated into the surrogacy analysis. Examples are commercial fisheries data, oceanographic time series observations, and modeled ocean currents, waves, and tides. These data can supplement the measurements and observations collected during the habitat mapping survey, but they may not have been collected at an appropriate resolution or in exactly the right location, thereby affecting their performance as a surrogate. It must be accepted, however, that some variables are simply not good surrogates for the occurrence of biota. Overall, the survey results suggest that water properties and seabed slope are not particularly useful in the environments studied; these variables were measured in more than 15 and 30 case studies (respectively) but not reported as being a useful surrogate in any of them. GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 877 Use of Published Classification Schemes GeoHab has not adopted or endorsed any particular habitat classification scheme, but much effort has gone into the design of various schemes, as reviewed in Chapter 4. Among the several advantages of using standardized classification schemes is that they enable comparisons to be made between different studies, by providing a standard framework and terminology which prompts the user for information through the course of the study. Also, their intrinsic predictive power of describing the relationships between physical habitats and their associated biological communities is useful for analysis, interpretation, and dissemination of results. The part of the questionnaire dealing with habitat classification schemes was completed by all 57 authors. Several authors used more than one scheme (Chapters 18, 26, 28, 29, and 40). At least one scheme was used in 21 of the case studies, but no scheme was used in 36 studies (i.e., most of the case studies, 36 out of 57, did not use any previously published habitat classification scheme; Figure 64.4). Most schemes were applied in only one or two studies. No single scheme was used by more than five studies. The scheme of Greene et al. [1] is the most cited. Several schemes listed in Figure 64.4 were developed by national or regional science agencies, and these were sometimes used in the case studies carried out in the relevant jurisdiction. However, the rate of uptake and utilization of schemes is problematic. For example, although 24 studies were carried out in European waters, only four of those studies applied the EUNIS habitat scheme (Figure 64.4). There may have been some confusion among the GeoHab community as to what constitutes a classification scheme. Some classification schemes nominated by No scheme Greene et al. [1] EUNIS IHO CMECS MNCR NOAA OSPAR BALMAR Riegl and Piller [9,10] Mumby and Harborne [11] Last et al. [12] 0 5 10 15 20 Number 25 30 35 40 Figure 64.4 List of different habitat classification schemes and the number of times they were used in the case studies. The schemes are as follows: Greene et al. [1]; EUNIS [2]; IHO [3]; CMECS [4]; MNCR [5]; NOAA [6]; OSPAR [7]; BALMAR [8]; Riegl and Piller [9,10]; Mumby and Harborne [11]; and Last et al. [12]. 878 Seafloor Geomorphology as Benthic Habitat authors in their questionnaires might be better described as definitions of terms and nomenclature. For example, the IHO [3] report is simply a list of geomorphic feature names and broad generic descriptions, rather than an actual classification scheme (with decision rules, etc.). Several authors cited the IHO definitions but did not nominate it as a habitat classification scheme in the questionnaire. The OSPAR [7] list of threatened species and habitats is similarly not an actual classification scheme. Apart from not fully understanding classification schemes, there may be other reasons why authors generally avoided using them. The extra effort required to apply a published classification scheme may come at too high a price for projects that are on a limited budget. The advantages of using a scheme are mostly only realized downstream in a project, where comparison of the results with other studies and communication become a priority. It is clear that further investigation is needed to fully understand the low rate of habitat classification scheme usage among GeoHab scientists. The Purpose of Habitat Mapping (Sectors and Clients) As was reviewed in Chapter 1, there are many purposes for mapping benthic habitats, but four of these stand out as being preeminent: (1) to support government spatial marine planning, management, and decision making; (2) to support and underpin the design of MPAs; (3) to conduct scientific research programs aimed at generating knowledge of benthic ecosystems and seafloor geology; and (4) to conduct living and nonliving seabed resource assessments for economic and management purposes (see Table 1.1 in Chapter 1). Only the final category (conduct living and nonliving seabed resource assessments) is directly relevant to an economic outcome; all others are related to government management and planning (Figure 64.5). One of the commercial purposes of habitat mapping specified in the list (Figure 64.5) relates to field testing of equipment. It is an interesting observation that seafloor mapping technology has developed over a number of decades through a close partnership between scientists and engineers. Habitat mapping is inextricably linked to developments in marine technology (especially sonar technology), and advances made in one field almost always lead to advances in the other. Habitat mapping was intended to be part of an ongoing monitoring program in 24 out of 57 cases (two case studies specified that part of their purpose was an ongoing monitoring program; Chapters 10 and 18). By contrast, habitat mapping was considered to be a one-off exercise in 33 case studies. This statistic is consistent with the usual government funding of marine research, where support can often be obtained for a one-off survey to collect new data about a particular environment or ecosystem process. Getting support for ongoing environmental monitoring is more difficult to obtain. However, out of the 33 one-off cases, 24 reported that their habitat map would form the baseline for monitoring future changes. This suggests that governments and regulators generally view habitat mapping as a useful means of measuring and monitoring environmental change, even where a decision to support ongoing monitoring has not been taken. There appears to be a disconnect between the stated purpose of habitat mapping in most case studies (i.e., government management), and the end users and clients GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 879 Purpose Spatial marine planning and management MPA design Scientific research and knowledge Fisheries resource assessment and management Assessment of nonliving resources Fisheries reserve design Baseline mapping for management Geological mapping Monitoring MPAs and environments Assessment of the cold-water coral distribution Testing existing national bioregions Hazard assessment Input to UNESCO world heritage application Equipment testing Hydrographic charting 0 5 10 15 20 Number of citations Figure 64.5 Purposes of habitat mapping specified by case study authors and the number of times they were cited. perceived by case study authors to be the main stakeholders (industry). Most authors considered marine conservation to be the biggest single user (Figure 64.6), followed by the fishing industry, government regulators, the scientific community, the tourism industry, navigation, other industry (deep-sea minerals, wind farms, etc.), oil and gas industry, and aggregate mining. Grouping all industry sectors together shows that most authors believe their mapping work is more relevant to industry in a general sense than to either conservation or government and science applications (Figure 64.6). Although the GeoHab scientists considered their data to be useful to industry (Figure 64.6), the overwhelming majority of habitat surveys were funded by government or government-funded agencies/institutions (n ! 49), with only minor funding from private industry (n ! 7) or nongovernment organizations (n ! 4). The role of government agencies as a catalyst for encouraging investment and interest from industry may be a key to understanding this apparent contradiction. For example, government agencies which deal with the management of natural resources often have a role of collecting environmental data to support and encourage industry investment in resource development; hence a government-funded survey may actually have industry as a primary purpose. The multiple potential uses of habitat mapping information may also be an explanation, as the data collected by one government agency to encourage industry investment might also be used by a different agency to make decisions about zoning and conservation. Consultation in Habitat Survey Planning Habitat mapping surveys are complex and expensive enterprises, requiring extensive planning and consultation to ensure successful outcomes. Consultation with all 880 Seafloor Geomorphology as Benthic Habitat Client groups Conservation Fisheries Goverment regulation Science community All industry Tourism Conservation Navigation Other industry Government and science Oil and gas 0 Aggregate mining 0 10 20 30 10 20 40 30 40 50 60 50 Number of citations Figure 64.6 Users and clients of habitat mapping specified by case study authors and the number of times they were cited. Inset shows that industry-related users grouped together actually form the largest single group. stakeholders is essential. Stakeholders comprise the survey participants (including the ship’s crew) and may also include industry practitioners, government managers, members of the scientific community, and (where relevant) representatives of indigenous or traditional users of the environment being studied. The latter group was consulted mainly in shallow water surveys along the coast or inner shelf areas. In general terms, the GeoHab science community has clearly embraced consultation, with most case studies consulting with at least three of the five broad stakeholder groups listed in Figure 64.7. All five stakeholder groups were consulted in eight case studies and four groups were consulted in 10 case studies. All case studies consulted with at least one stakeholder group. In the two studies that consulted only one group, it was the government that was engaged in planning the surveys. Data Storage and Accessibility The data collected during a habitat mapping survey is composed of many types, including electronic data (multibeam sonar, underwater video or camera data, current and water property measurements, modeled data, etc.) together with physical samples (sediments and biota) that must first be analyzed before numerical or classification type data are obtained. Government-funded agencies or institutions commonly have an obligation to store the data and samples they collect (for a period of time at least), which is consistent with the response to the question, “Is the habitat mapping GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 881 Participants Industry Government Science community Indigenous people–traditional users 0 10 20 30 Number consulted 40 50 Figure 64.7 Stakeholder groups consulted in preparation of habitat mapping surveys specified by case study authors and the number of times they were cited. A total of 54 case studies responded to the question about stakeholder consultation. data stored or saved for future use?” Out of 54 responses, 45 (83%) said yes and 9 (17%) said no. As mentioned above, the majority of habitat surveys are funded by government or government-funded agencies/institutions, which have an obligation to the taxpayers to make their data and survey results freely available and accessible. Indeed, it is commonly a justification for collecting the data in the first place that it will be used by the wider community for more than one purpose (map once, use many ways). This perception is confirmed by the responses received from the question as to “whether the habitat mapping data collected for each case study had been used for any other (additional) purpose by a different group?” Out of 54 responses, 36 (67%) said yes and 18 (33%) said no. However, when it comes to the question of making the data and information available over the Internet, a different response is received. Out of 57 responses to this question, 22 authors (39%) reported that at least some part of the data they collected was accessible over the Internet, whereas 35 authors (61%) reported that none of their data were accessible via the Internet. There are several reasons why most agencies do not make their data web accessible. In the first place, government agencies often hold confidential information and so are unable to make all of their data web accessible. The data may be kept confidential for commercial reasons. A two- to five-year moratorium on making data public is often imposed to allow scientists involved with the data collection time to publish their findings. The cost of implementing a security system that distinguishes between confidential and nonconfidential data may prohibit some agencies from making any data accessible over the Internet. On the subject of cost, the funding provided to agencies to conduct surveys to collect data may not include the added costs of storing the data and making it web accessible. 882 Seafloor Geomorphology as Benthic Habitat Another reason why most agencies do not make their habitat data accessible over the Internet is related to the nature of digital data itself. Simply putting raw data online is rarely useful. If parameters and standards are not published with the data, or if infrastructure (server space and a fast network connection) is not available, it becomes difficult to understand and use the data. In the case of multibeam sonar, data collected at sea are first cleaned of noise and artifacts to produce a processed data file. In order to view raw or processed files, specialized software is needed to create a georeferenced bathymetric map of the seabed. Unless the user owns the correct software, the data files are unreadable. Data volume is another reason. The files comprising a week-long multibeam sonar survey can be very large (several terabytes), well beyond the scope of what can reasonably be delivered over a standard Internet connection. Instead, agencies commonly make a gridded map product (at an appropriate resolution) available over the Internet. Interested persons may then view the map online and inquire directly to the agency about access to the processed multibeam data files if this is necessary. In fact, several authors noted that their data were partially available in a decimated form, as in the case of providing a gridded bathymetry product. A final possible explanation as to why most agencies do not make their habitat data accessible over the Internet is that there is no great demand for the raw data; most of the general public are content with the published reports, gridded map products, and papers that are produced from the survey. Although it may be true that scientists need access to raw data to do their work, they can usually get that access by directly contacting colleagues at the agency that hosts the data. Politicians and managers want information, explanations, and advice from scientists (not raw data); they want to know that the advice they receive is backed up by adequate data and peer-reviewed papers, but they rarely get involved with the technical details of how the information was reduced and interpreted. Hence, the cost of building and maintaining a web-based data delivery system to serve up raw or processed digital data may simply not be warranted. Gap Analysis and Priorities for the Future of Habitat Mapping: Geographic Areas, Geomorphic Features, and Environmental Variables A gap analysis has been carried out based on: (1) the geographic distribution of case studies; (2) the geomorphic features encompassed by case studies; and (3) the environmental variables considered by case studies. Geographically, about half of the case studies were from waters around Western Europe, with the remainder scattered around North America and the periphery of the Indian and Pacific Oceans (Figure 64.8). There are a few case studies from the northern margins of Africa (Mediterranean and Red Seas), but none from the Atlantic margin of the continent. The margins of Asia and South America were not represented in any case study (there is no case study from the South Atlantic Ocean). Given the intense pressures facing benthic habitats and broad regional differences in ecosystems, species, and habitats, future case studies from these regions should be specifically sought for future editions of the Atlas. GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 883 Figure 64.8 Map showing the distribution of case studies. The numbers refer to case study chapters. About half of the case studies are from waters around Western Europe, with the remainder scattered around North America and the Austral-Asian region; few are located in Africa, and there are none in Asia or South America. In terms of geomorphic features, shelf and slope habitats were well represented in the case studies (Figure 64.9). Sandbanks, sand waves, coral reefs, canyons, and glaciated shelves are particularly well covered. The case studies mainly focused on a single geomorphic feature type (36 out of 57 studies); however, 18 case studies focused on two feature types, and three studies focused on three or more types. Surprisingly (given their accessibility and overall importance to humans), estuarine and deltaic coastal habitats are poorly represented in the case studies. In particular, wave-dominated estuaries and all types of deltas seem to have been avoided. This may be a consequence of the shallow depths of these environments, which greatly restricts boat access and hence the application of sonar mapping technologies (but not LIDAR or other remote sensing techniques) and conducting marine fieldwork in general (e.g., video tows and sampling). The other obvious gap in geomorphic features is deep ocean (abyssal–hadal) environments, which were described in only a few case studies. Given that around 50% of the earth is located at abyssal depths, this knowledge gap is globally significant and represents a major challenge to the habitat mapping community. Some obvious reasons why the abyssal environment is avoided include: (1) the high cost of operating deep-water survey vessels; (2) the abyssal seafloor is largely located on the high seas and therefore falls outside the responsibility of national science agencies; and (3) the perception of these environments as being remote and hence not heavily impacted by human activities. The final aspect of the gap analysis relates to physical attributes of the environment that are likely to affect habitats but were not featured in the case studies for 884 Seafloor Geomorphology as Benthic Habitat Sand wave–sandbank Coral reef Canyon Glaciated shelf Seamount–guyot Plateau Shelf valley Temperate rocky reef Seagrass Ridge Fjord Trough–trench Hydrothermal vents Escarpment Pinnacle Estuary Peak Channel Cold seep Holes Platform Barrier Island Tidal inlet Embayment Sill Terrace Mound 0 1 2 3 4 5 6 7 8 9 10 Number of times feature cited Figure 64.9 Graph showing geomorphic features described in the case studies and the number of times each feature was cited as a major focus for the study (n ! 82). Some studies considered more than one type of feature. various reasons. A general observation is that most habitat mapping studies report a snapshot view in which a range of single observations are made over an area in order to produce a set of spatially complete, mostly temporally static, data layers that lend themselves to multivariate analysis techniques. But habitats are composed of many spatially and temporally dynamic elements that cannot be measured by a single observation. Often it is the variability of a physical attribute that characterizes habitats (e.g., the variation in temperature and salinity in an estuary, the frequency of storm disturbances to particular coastal and shelf habitats, or the migration rate of continental shelf dunes). The dynamic aspects of habitats include their connectivity in relation to adjacent habitats as sources or sinks of larvae and colonizers and the biological processes of predator–prey relationships, trophic levels, and gene flow. These temporally dynamic elements seem to be overlooked in many case studies. It is, of course, very expensive to collect a long time series of measurements needed to accurately represent the dynamic conditions of many environments. This is particularly a problem when the process is episodic in nature or has a long periodicity (“long” in terms of the duration of a typical research program may be only a few years!). However, the geologic record can often assist in such cases, by providing in situ environmental time series information. Sediment cores were collected as part of some case studies (Chapters 46 and 50), and changes in the down-core chemistry of sediments, fossil assemblages, sediment accumulation rates, or the physical character of sediment deposits (grain size, presence/absence of primary structures, etc.) are GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 885 all useful sources of information about changes in the environment [13,14]. Coral core records from the Great Barrier Reef document changes in river sediment supply since European settlement [15], for example. The resolution of sediment records is of course a function of accumulation and sediment mixing rates, and deposits suitable for palaeoenvironmental analysis may not always be available. However, where they occur, the integration of such geologic time series data sets into the characterization of habitats can provide insights into their naturalness, classification, and broader understanding of their temporal variability (and stability). Numerical modeling is another tool that can provide assistance in some cases where dynamic aspects of the environment are a primary factor characterizing habitats. Such is the case, for example, in many tide-dominated shelf and estuarine habitats and on tropical shelves influenced by storm events (hurricanes, typhoons, and cyclones). For example, repeat multibeam surveys combined with current measurements and modeling were used to understand the migration of large sand waves in Torres Strait, Australia, in relation to sea grass dieback [16]. Two case studies (Chapters 33 and 41) included modeled wave and current information in their analyses of habitat characterization, and both were located on the macrotidal, west European continental shelf. It is also important to note that wave and current energy are generally found to be among the most useful surrogates by the case study authors (Figure 64.3). Discussion: Best Practices for Benthic Habitat Mapping The case studies in this volume, and the responses to the questionnaires by authors reported in this chapter, provide valuable insights into what constitutes best practice for conducting a habitat mapping program (Figure 64.10). In the first instance, it is clear that thorough consultation and planning of habitat mapping surveys is standard practice among GeoHab scientists. Planning for most surveys will involve consultation with the scientists directly involved, industry practitioners holding a stake in the area to be surveyed, relevant government agencies, and the broader science community. In cases where indigenous people or traditional users have an interest in the proposed survey area, they are normally consulted by GeoHab scientists in designing the survey. The main objectives of conducting most habitat mapping surveys are to take stock of what benthos and habitats exist in a given area, measure the condition of the environment and its dependent communities, and try to identify trends in overall condition (or health). Therefore, understanding and quantifying the naturalness of the habitats present in an area is essential in order to establish baselines and measure change. This may be a challenge in areas where undisturbed reference sites are unavailable. Identifying trends in condition (improving or declining) may be accomplished through repeated mapping surveys and/or by taking a time series of measurements of previously identified surrogates. Several of the case studies described modified or extremely modified environments, yet theirs are the first systematic assessment of the condition and status of that particular habitat. The commencement of a habitat mapping and monitoring program in areas heavily impacted by human activities still requires establishing a baseline of relative naturalness so that improvements in 886 Seafloor Geomorphology as Benthic Habitat D Monitoring P n r ito Mo R ing Drivers for habitat surveys S I - Establish baseline for future monitoring - Condition assessment - Resource assessment Survey planning - Consultation - Selection of appropriate surrogates Use standard terminology and methods - Standard data collection methods - Review surrogacy methods Data storage and web access - Save data for future use - Maps and reports web accessible - Map once use many ways Make use of existing data -Oceanographic data, sediments, and biota - Revise survey plan to fill gaps Standard products - Shaded relief bathymetry - Bathymetry 3D fly-thrus - Grayscale backscatter - UW video and photography - Dimensionless ordination plots Consider habitat classification scheme - Scheme designed for region - Generic scheme Figure 64.10 Series of steps involved in the conduct of a benthic habitat mapping survey to achieve best practice. Habitat mapping and monitoring conform to the “state” part of the DPSIR framework (yellow shading in insert at top left). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this book.) condition can be measured and reported. Such improvements may become the performance indicators to justify the establishment of MPAs or fisheries reserves, for example. GeoHab scientists represent a range of disciplines, and it has been an objective of this book to try to bring clarity to the terminology used for describing habitats and the benthic marine environment. The glossary to this volume contains around 200 definitions contributed by the authors and adopted by mutual consent. While it remains a work in progress, the glossary provides a useful reference for habitat mapping terminology. Standard methods and protocols for marine data collection are well established for most data types (sediment grain size, taxonomy, water properties, depth, etc.). The case studies in this book also describe some new developments in video classification (Chapters 46 and 47) and automated methods for analyzing and classifying seabed morphology based on multibeam bathymetry and acoustic backscatter data (Chapters 8, 34, 36, and 58; see also technologies listed in Table 1.2, Chapter 1). The different methodologies described in a number of case studies to identify surrogacy GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 887 relationships are also among some of the more recent and exciting developments in habitat mapping science. In terms of habitat data products, there appears to be convergence in the use of 3D imagery and false-colored grids for representing bathymetry data. Fly-through video movies can be generated from bathymetry data sets, providing a powerful communication device, but are of no use in hard-copy-only literature. Maps of sediment type, grain size, and geomorphic features are also a common theme among the case studies. Dimensionless ordination plots are a popular way of displaying relationships between biological observations and physical surrogates. Although these data presentation methods are not standardized, there is some consistency among the case studies in their occurrence, which allows some comparisons to be made. GeoHab authors seem less certain about adopting any standard classification scheme in their studies. Most case studies in this book did not utilize any scheme in spite of the advantages for comparison of results and communication. The jury is still out as to whether any one scheme will become a habitat mapping standard in the years ahead. Selecting which biological and physical variables to measure is perhaps one of the most important decisions to be made at the outset of a habitat mapping program. It is obvious that not all variables are equally useful as surrogates (Figure 64.3). However, different variables will be more useful surrogates in particular habitats; no single variable is equally useful as a surrogate in all habitats. Since collecting and analyzing each data type has a direct cost to any program, science teams planning habitat mapping surveys would be well advised to review the literature to optimize the list of variables to be measured. The collected data should be archived and stored; this is standard practice for most government agencies (Figure 64.10). Where the data were collected using public funds, storage and archiving is generally a requirement. However, even in cases where data have been collected using private resources (where confidentiality may be a requirement), thought should be given to data storage and archiving lest the data be lost. In terms of making data accessible, it is the common practice of most GeoHab scientists to provide access to reports and publications describing the data over the Internet. This appears to be a high priority for stakeholders. Making people aware that the habitat mapping work has been completed and providing access to summary maps and reports is generally more important than making large volumes of raw digital data accessible over the Internet. If a moratorium is imposed on the release of the data (e.g., to allow scientists time to publish their findings), the length should be kept to as short a time as practicable and for no longer than 5 years. One final and very important best practice for habitat mapping is to make use of existing information and to make available for reuse data collected during a habitat mapping survey. Using existing data sets, such as commercial fisheries data, oceanographic observations, modeled wave and currents information, museum biological data, and so on, is a cost-effective means of supplementing and adding value to newly collected habitat survey data. Making available data collected for reuse by other groups is a policy adopted by many government departments in relation to any and all government-funded programs. For habitat mapping science, this policy translates into “Map once, use many ways.” 888 Seafloor Geomorphology as Benthic Habitat Acknowledgments Thanks to Geoffroy Lamarche (New Zealand Institute of Water and Atmospheric Science) for helpful comments and suggestions on an earlier draft of this chapter. This work was produced with the support of funding from the Australian Government’s Commonwealth Environment Research Facilities (CERF) program and is a contribution of the CERF Marine Biodiversity Hub and UNEP/GRID Arendal. This chapter is published with the permission of the Chief Executive Officers of Geoscience Australia and UNEP/GRID Arendal. Appendix 1 Questionnaire Completed by All Case Study Authors Name of first author: Contact e-mail: Case study number: Case study title: Science Questions: What was the size (spatial area, km2) of your study site? ___________km2 What was the water depth range of your study site (from most shallow to deepest)? _____m to ____m. What was the maximum resolution (minimum distance in meters between soundings) of acoustic mapping in your study? ______________m. Apart from geomorphology, what variables were measured in your study to characterize habitats? (Tick the relevant boxes) ● ● ● ● ● ● ● ● ● ● ● Sediment grain size Sediment composition (carbonate content, TOC, etc.) Acoustic backscatter Water properties (temperature, salinity, dissolved oxygen, etc.) Primary production Slope Rugosity Topographic position index (TPI) Current strength or bed shear stress Wave characteristics or wave-induced bed stress Other (specify) ________________________ Which of the above (if any) did you find to be the most useful surrogates for biota (communities, species, etc.)? _______________________________ Did you use any spatial analysis techniques in your study (e.g., multivariate analysis, GIS tools)? If so please specify ____________________________________ GeoHab Atlas of Seafloor Geomorphic Features and Benthic Habitats: Synthesis and Lessons Learned 889 Did you use any previously published classification scheme in your study (e.g., EUNIS)? If yes, please specify here: _________________ Socioeconomic Questions: What was the main user group that your study was serving? (Tick the relevant boxes) ● ● ● ● ● ● ● Fisheries Conservation Oil-gas Aggregate mining Navigation Tourism Other (specify) ________________________ What was the main purpose of the survey (e.g., input to spatial marine planning, marine protected area design, fisheries reserve, seabed resource assessment)? ________________________________________________________ After the survey was completed, was the data used for a different purpose by another group? If so, please explain: ______________________________ Who paid the costs for your study? (Government research body, private sector, other) _________________ Was your study a one-off study or was it part of an ongoing monitoring program? (Circle one): One-off study/ongoing monitoring If it was a one-off study, is there any current plan to use the outcome of your study to establish a baseline for future assessment? (Circle one): YES/NO Who was consulted in planning your survey? (Tick the relevant boxes) ● ● ● ● ● ● Survey participants Industry practitioners Government Science community Traditional users (indigenous people) Other (specify) ________________________ Are the data collected in your study stored (or planned to be stored) in a national database? (Circle one): YES/NO Are the data collected in your study currently web accessible? (Circle one): YES/NO References [1] H.G. Greene, M.M. Yoklavich, R.M. Starr, V.M. O’Connell, W.W. Wakefield, D.E. Sullivan, et al., A classification scheme for deep seafloor habitats, Oceanol. Acta 22 (1999) 663–678. 890 Seafloor Geomorphology as Benthic Habitat [2] C.E. Davies, D. Moss, The EUNIS Classification, European Environment Agency, http:// www.eea.europa.eu/data-and-maps/data/nationally-designated-areas-national-cdda-3/ eunis-habitat-classification. [3] IHO, Standardization of Undersea Feature Names: Guidelines Proposal Form Terminology, International Hydrographic Organisation and International Oceanographic Commission, Bathymetric Publication No. 6., fourth ed., IHO, Monaco, 2008. [4] C.J. Madden, D.H. Grossman, A framework for a coastal/marine ecological classification standard (CMECS), in: B.J. Todd, G. Greene, (Eds.), Mapping the Seafloor for Habitat Characterisation, Geological Association of Canada, St. Johns, NL, 2007, pp. 185–210. [5] D.W. Connor, J.H. Allen, N. Golding, L.K. Howell, L.M. Lieberknecht, K.O. Northen, et al., The Marine Habitat Classification for Britain and Ireland Version 04.05, JNCC, Peterborough, http://jncc.defra.gov.uk/page-1584, 2004. [6] NOAA National Centers for Coastal Ocean Science, Shallow-water benthic habitats of American Samoa, Guam, and the Commonwealth of the Northern Mariana Islands, NOAA Technical Memorandum NOS NCCOS 8, Silver Spring, MD, http://ccma.nos. noaa.gov/products/biogeography/us_pac_terr/index.htm, 2005. [7] OSPAR, Quality Status Report 2010, OSPAR, London, http://qsr2010.ospar.org/en/ index.html, 2010, 176 pp. [8] Z. Al-Hamdani, J. Reker, U. Alanen, J.H. Andersen, J. Bendtsen, U. Bergström, et al. (Eds.), Towards marine landscapes in the Kattegat and Baltic Sea, BALANCE Interim Report No. 10, 2007, 118 pp. [9] B. Riegl, W.E. Piller, Distribution and environmental control of coral assemblages in northern Safaga Bay (Red Sea, Egypt), Facies 36 (1997) 141–162. [10] B. Riegl, W.E. Piller, Coral frameworks revisited—reefs and coral carpets in the northern Red Sea, Coral Reefs 18 (1999) 241–253. [11] P.J. Mumby, A.R. Harborne, Development of a systematic classification scheme of marine habitats to facilitate regional management and mapping of Caribbean coral reefs, Biol. Conserv. 88 (2) (1999) 155–163. [12] P.R. Last, V.D. Lyne, A. Williams, C.R. Davies, A.J. Butler, G.K. Yearsley, A hierarchical framework for classifying seabed biodiversity with application to planning and managing Australia’s marine biological resources, Biol. Conserv. 143 (2010) 1675–1686. [13] A. McMinn, G.M. Hallegraeff, P. Thomson, A. Jenkinson, H. Heijnis, Cyst and radionucleotide evidence for the recent introduction of the toxic dinflagellate Gymnodinium catenatum into Tasmanian waters, Mar. Ecol. Prog. Ser. 161 (1997) 165–172. [14] K. Morris, J.C. Butterworth, F.R. Livens, Evidence for the remobilization of Sellafield waste radionuclides in an intertidal salt marsh, West Cumbria, U.K., Estuarine Coast. Shelf Sci. 51 (2000) 613–625. [15] M.T. McCulloch, S. Fallon, T. Wyndham, E. Hendy, J.M. Lough, D. Barnes, Coral record of increased sediment flux to the inner Great Barrier Reef since European settlement, Nature 421 (2003) 727–730. [16] J. Daniell, P.T. Harris, M. Hughes, M. Hemer, A. Heap, The potential impact of bedform migration on seagrass communities in Torres Strait, Northern Australia, Cont. Shelf Res. 28 (2008) 2188–2202. View publication stats