A new method to analyze seafloor geodiversity around the Hawaiian and Canarian archipelagos and the New Zealand subduction zone

(1)

A new method to analyze seafloor

geodiversity around the Hawaiian and

Canarian archipelagos and the New

Zealand subduction zone

Author:_{Floris Veloo} UvAnetID:_10628150

Thesis supervisor:_{Dr. A.C. Seijmonsbergen} Second supervisor: _{Dr. W.M. de Boer}

Date:_03-07-2017 Location:_Amsterdam Words: ₄₆₂₄

(2)

Abstract

The concept of geodiversity covers the natural variation in abiotic factors. Areas with a large variety of abiotic factors have the potential to maintain an high level of biodiversity. Due to changing environmental conditions, the assessment and quantification of geodiversity becomes increasingly important in geoconservation practices. As a result of digitally mapping geodiversity it is possible to easier locate geodiversity hotspots and therefore potential biodiversity hotspots.The assessment and mapping of these abiotic factors is still in an early stage. In particular the mapping of submarine geodiversity.

The aim of this study is to develop a geodiversity index for the seafloor that can be extrapolated to broader areas. Several datasets are used to create fifthteen sub-index maps. With these maps three submarine geodiversity index maps are made in ArcGIS for the Hawaiian and Canarian archipelagos and the New Zealand subduction zone. The developed geodiversity formula is a summation of abiotic factors such as geomorphology, bathymetry, range/std of slope angle and light penetration. In order to test the geodiversity index, the results are analyzed by using the input data. The geodiversity index shows large differences in geodiversity patterns between the study areas. Therefore, the variation in abiotic factors within and between the study areas is large. In general, the geomorphology- and bathymetry index show the largest effect on the geodiversity index. The analysis also reveals a positive correlation between both the slope indices. However, the light penetration index has a negligible influence on the geodiversity index in all study areas. These results are reflected in the input data and the final geodiversity maps. The analysis shows that geodiversity hotspots are more present in the subduction zone than in the archipelago's. This is mainly caused by the bathymetry diversity. The analysis also reveals that a relation between the geodiversity scores and the age of the seafloor give insignificant results.

(4)

1. Introduction

1.1 _{The relevance of geodiversity}

The term ‘geodiversity’ has first been introduced by Wiedenbein (1993). It can be defined as ‘the natural range of geological, geomorphological, soil and hydrological features’ (Gray, 2013). The concept covers the diversity of abiotic factors that make up the geosphere. The importance of preserving certain ‘geodiverse’ areas has only been acknowledged in recent years and is becoming more widely spread (Gray, M. 2013).

An explanation for the growing recognition of the importance of geodiversity is the interrelationship of the biology with its environment. In order to thrive in changing environmental conditions life has always had the tendency to adapt to its surroundings. This process of evolution has altered species into new species and shaped the biodiversity. According to Comer et al. (2015) environments with a large variety of abiotic factors have the potential to maintain an high level of biodiversity. Such ecosystems are also considered to be more resistant to external pressure as a result of the micro-environments. Therefore, geodiversity could potentially increases the adaptation time of the local biodiversity which makes these ecosystems more resilient.

Besides, geodiversity is of great societal and scientific relevance. The value of geodiversity hotspots include intrinsic, economic, educational, cultural and aesthetic values (Crofts and Gordon, 2014). It contains, for instance, information about past climatic conditions and species, and attracts geotourism. However, anthropological influences negatively affect these values. Both through being the force behind current environmental changes and by disrupting geodiverse sites. When the geology is removed, it can not be retrieved. Therefore, conservation of these geodiversity hotspots becomes more important for decision-makers. This so-called ‘geoconservation’ can help to preserve geodiversity nowadays and for future generations.

Little attention has been paid to land-sea geodiversity relationships, e.g. through erosion/deposition. Coastal ecosystems and marine environments cover a large area of potential and still unexplored and unknown geodiversity. The geomorphology dataset (Harris et al. 2014) is one of the few maps which describe the global seafloor landscape. This lack of information could result in a neglecting of the existence of potential marine reserves, which therefore makes marine geodiversity of utmost importance. Due to the large area and inaccessibility of the seafloor, digital mapping of seafloor geodiversity will make it possible to easier locate and protect marine geodiversity sites. This should result in the localization and conservation of marine habitats with high biodiversity due to their interrelationship with the submarine environment. Research focused on these environments will become less time-consuming and less expensive as well.

1.2 The importance of geodiversity mapping

In order for policymakers to adequately implement geoconservation policies methods are in development for the localization of geodiversity sites. Gray (2013) states that ‘’Not all of the geodiversity of the planet needs to be conserved. Only those elements that are seen as being significant for one or more of their values’’. By using digitized maps with ArcGIS geodiversity index maps can be created, which visualize hotspot areas of specified abiotic factors. GI’s have been developed for terrestrial study areas including mountainous (Seijmonsbergen et al., 2014), flatlands (Vregelaar, 2015) and islands (Chambers, 2014). However, a geodiversity map for marine environments still has not been developed. In order to make an sufficient geodiversity map choices on the selection of legend categories and scale as well as abiotic factors have to be made. Which abiotic

(5)

factors should be included, depend on the study area, the available datasets and the aim of the research. In order to calculate a geodiversity map for a marine environment a geodiversity index will be created. With such an index large and inaccessible study areas can easily be analysed. Developing a geodiversity index for the marine environment will support policy-makers and scientists to easily locate hotspots and to implement policies more accurate. Due to the simplification of the expert-based maps, even users who lack earth scientific knowledge are able to interpret and understand these maps.

1.3 Research aim

The research aim is to develop a geodiversity index for the seafloor that can be used to analyse geodiversity between different marine environments. Therefore, the proposed index should be transferable to different marine areas. The analysis focusses on the relationship of the geodiversity index with the sub-indices and the geological setting. In order to achieve this goal the following research questions needs to be answered:

Research question:

- Which abiotic factors should be selected for the calculation of a geodiversity index for the seafloor in regard to the datasets available?

- What formula should be used for the calculation of a geodiversity index for the seafloor? - Is the geodiversity index score linked to the age of the seafloor?

1.4 Study area

The development of the submarine geodiversity index is tested in three case studies. The case studies are divided over three separate locations, 1. the seafloor surrounding the main islands of the Hawaiian and 2. Canarian archipelagos and 3. the seafloor surrounding the subduction zone at the southeast coast of the northern island of New Zealand (Fig. 1). In order to make a large enough analysis, the extent of the study areas of both archipelagos have an approximate distance of 100 kilometers from the coast. The study area extent of the subduction zone has an approximate distance of 200 kilometers from the coast, because the subduction zone is only present at one side of the coast.

These locations have been selected on the basis of their particular geological settings. The Hawaiian and Canarian archipelagos are the result of the movement of the marine tectonic plates over an active tectonic hotspot. As a contrast, the New Zealand subduction zone has a different origin. The analysis of the geodiversity in these contrasting geological settings will contribute to our understanding of seafloor development and showcases the usability of the marine geodiversity index.

(6)

(7)

2. Methods

2.1 General overview

Geodiversity maps present trends and patterns of potentially high geodiversity in large and inaccessible areas. These maps can then be used in fine-scale case studies to find local geodiversity hotspots for, for instance, geoconservation purposes (Gray, 2013). Like indices developed for terrestrial environments (Seijmonsbergen et al., 2014; Vregelaar, 2015; Chambers, 2014), the marine geodiversity index map is a combination of several sub-indices. These sub-indices represent the variety within the selected abiotic factors that make up the submarine geodiversity. The sub-indices were selected on the base of their relevance and influence on the submarine geodiversity index, and on datasets availability. The calculation of the geodiversity index (GI) is based on five sub-indices of abiotic factors: geomorphological diversity (Gdi), bathymetry diversity (Bdi), range of slope angle (Sr), standard deviation of slope angle (Ss) and light penetration diversity (Ldi).

The formula for the calculation of the geodiversity index per cell uses default settings of equal importance (weight = 1) and is formulated as (1):

(1) GI = Gdi + Bdi + Sr + Ss + Ldi

The final model (Fig. 2) has three parameters that can be changed; study area, cell size and weight of the sub-indices. If the model will be used for different seafloor locations, it is possible to use the same method but with adjusted weights for the sub-indices. For instance, study areas with a flat seafloor can have a slope angle diversity weight of zero and thus will not count for the geodiversity index.

Figure 2 presents an overview of the geodiversity model, which is based on eight submodels (see. Appendix A). The pre-processing model (1) defines the extent of the study areas and cell size of the indices. The five sub-index models geomorphology (2), bathymetry (3), range of slope angle (4), standard deviation of slope angle (5) and light penetration (6) are used for the calculation of the separate sub-index scores. The geodiversity index model (7) calculates the geodiversity index while the statistics model (8) calculates specific statistics for this research.

(8)

Figure 2._{General overview of the geodiversity model}

The processing of the datasets and the calculations is done in ArcGis 10.4.1. All produced maps have the WGS_84 geographic coordinate system. The analysis will be done in ArcGis 10.4.1 as well as in Matlab R2015b. Figure 3 shows the workflow used in this research. An extended description of this workflow is given in the following paragraph.

(9)

2.2 Data collection

Table 1 provides information on the input datasets used to create the sub-indices. The information is categorized by dataset and contains the source of the datasets, a description, year of publication, original scale or cell size and data type.

Table 1. _{Metadata of the input datasets}

* H = Hawaiian archipelagos. * C = Canarian archipelagos.

* NZ = New Zealand subduction zone

Dataset Source dataset Contains Year Scale/cell size Type Geomorphology Harris et al.

(2014)

Seafloor geomorphic features

2014 1:500.000 polygon

Bathymetry Ryan et al. (2009) Bathymetry data 2009 *H: 114m *C: 53m *NZ: 91m raster Range of slope angle Calculated from bathymetric data Range of slope angle per cell

2009 H: 114m C: 53m NZ: 91m raster Standard deviation of slope angle Calculated from bathymetric data Variation of slope angle per cell

2009 H: 114m C: 53m NZ: 91m raster Light penetration Calculated from bathymetric data Light penetration per cell 2009 H: 114m C: 53m NZ: 91m raster Age of the seafloor Muller et al. (2008)

Seafloor age 2008 1:500.000 polygon

The development of the geodiversity index is dependent on the availability of input datasets for the three study areas. The geomorphology dataset (Harris et al, 2014) is still the only available expert-based dataset for the seafloor, which describes submarine features by category. Geological maps for the whole seafloor are still unavailable and therefore not included. The range of slope angle describes the range between minimum and maximum slope angle within a cell. Therefore, it determines the potential diversity of slope angle within a cell. The standard deviation of slope angle describes the variation of slope angles within a cell. This variation is a measure for the degree of roughness and thus informs about the shape and formation of the landscape. Both slope-based maps are derived from the bathymetry dataset (Ryan et al., 2009) as well as the light penetration map. The geomorphology dataset is also based on the bathymetry and slope angle datasets as it was derived from seafloor depth differences and geometry. The bathymetry dataset consists of three separate bathymetry dataset classified by study area and extracted from a global bathymetry dataset (Ryan et al., 2009). Light penetration describes the diversity of available solar radiation at the seafloor per cell. The amount of available solar radiation in turn determines the growth rate of photosynthetic species which affect the deposition rate of organic and calcareous compounds (Hader et al. 2007).

(10)

The digital appendix (see. Appendix C.) contains all the original datasets including legends, classifications and metadata. The list of references includes the original reports which contain extensive information on the development of the input datasets.

2.3 Grid definition

In order to obtain useful output data the correct cell size needs to be selected. According to Hengl (2006) the optimal cell size is dependent on the scale or cell size of the input data and the meaning behind it. Hengl (2006) recommends to calculate the coarsest (2km), finest (250m) and recommended (1km) cell size for this study from these input data. This study will use a 1 km cell size in order to maximize variation between the geodiversity index classes and to optimally visualize geodiversity patterns. For the creation of the vector grids the ‘create-fishnet’-tool is used. In order for all individual cells to have equal influence on the geodiversity index only whole cells are selected. The ‘polygon to raster’-tool converts the vector grid to a raster. With this newly produced grid template raster calculations can be done

2.4 Pre-processing

The pre-processing consists of the creation of the study areas and clipping the input data to the study area (Fig. 4). This is necessary because the rasters of the sub-indices need to have the same cell size and area extent. Firstly, the buffer-tool is used to create 2 study areas covering the seafloor with an approximate distance of 100 km from the coast (Hawaii & Canary) and one study area with 200 km from the coast (New Zealand). The ‘clip-tool’ clips the input data as well as the external grids to the research areas.

2.5 Calculating the sub-indices

Firstly, the fields ‘Shelf classification’ and ‘Classification’ are removed from the geomorphology dataset (Ryan et al. 2009), because these fields provide irrelevant information on the already included ‘Geomorphic features’ field for this research. For instance, the fields include a classification of the slope angle of geomorphic features. This would result in an additional weight of the slope and geomorphology indices on the geodiversity index. The dataset is converted into a raster file and overlain by the fishnet grid. The ‘zonal-statistics’-tool of ArcGIS calculates the number of geomorphic features per cell with the ‘VARIETY’-option. The maximum score of geomorphic features per cell is 5, which results in a

classification ranging from 0 to 5. In order to create the light penetration dataset the ‘area-solar-radiation’-tool is used and added to the Beer-Lambert law formula (Fig. 5). The diversity scores of the bathymetry, light penetration and standard deviation of slope angle are created with the ‘STD’-option in the ‘zonal-statistics’-tool. Then the ‘reclassification’-tool is used following

(11)

the geomorphology classification by using 6 equal size classes. The sub-index classes had to have equal size classes in order to have an equal effect on the geodiversity index. The same method was used for the creation of the range of slope angle diversity index, but with the ‘RANGE’-option in the zonal-statistics tool. The light penetration diversity index is the only sub-index which is reclassified according to the ‘natural breaks’-method.

2.6 Calculating the geodiversity index

The geodiversity index is calculated with the raster calculator-tool according to the GI-formula (1). The weight of the individual sub-indices is set to one by multiplying each sub-index with 1. The output raster contains the summation of the reclassified diversity scores of the sub-indices per cell ranging from 0-12_{. Then the geodiversity index is reclassified into 5 classes (Table 2.) with the} reclassification-tool based on natural breaks. The Jenks (1967) method is selected, because it reduces the variance within classes and maximizes variance between classes. The result is a reclassification method in which the classes are based on the average diversity score of a study area. So, the geodiversity classes represent geodiversity scores relative to a study area. As a result of this approach, the geodiversity index maps of different areas are better comparable. The final classes represent areas with potential geodiversity scores ranging from ‘Very low’ to ‘Very high’.

Table 2._{Classification of geodiversity scores per study area}

Geodiversity class

Geodiversity score without reclassification

Hawaii Canary New Zealand

1. _{Very low} 0 0 0-2 2. _Low 0-1 0-1 2-3 3. _Moderate 1-4 1-4 3-4 4. _High 4-7 4-8 4-6 5. _{Very high} 7-12 8-10 6-12 2.7 The analysis

After the development of the geodiversity index an analysis is carried out to test the index for the three study areas.

Firstly, in order to analyze the effect of the separate sub-indices on the geodiversity index the band collection statistics tool is used to create a correlation matrix per study area. The values in these matrices represent the degree of linear dependency between the indices. Then the geodiversity maps are visually examined and interpreted in order to understand the origin of the patterns.

Secondly, the index maps of the Hawaiian and Canarian archipelagos and the New Zealand subduction zone are compared. This analysis provides more insight on the differences and similarities in abiotic factors amongst the different geological settings.

Lastly, an analysis of the age of the seafloor in relation to the geodiversity present is carried out in order to find age/geodiversity patterns. This will be done with the ‘MEAN’-option in the zonal statistics tool. This tool calculates the average geodiversity score per class of the seafloor age dataset.

(12)

3. Results

This paragraph highlights the final geodiversity index maps per study area and the results of the analyses. The original data (Table 1.), preprocessing maps, sub-indices as well as the metadata can be found in the digital appendix (see. Appendix C). The sub-index maps are also shown in appendix B. Figure 6, 7 and 8 show the spatial distribution of geodiversity within the study areas of Hawaii, Canary & New Zealand. The geodiversity index map is based on 5 sub-index maps; geomorphology, bathymetry, range of slope angle, std of slope angle & light penetration. Geodiversity classes range from ‘Very low’ (1) to ‘Very high’ (5) values for geodiversity.

Hawaii:

(13)

Canary:

(14)

New Zealand:

(15)

Table 3._{Distribution of the geodiversity index classes per study area} Geodiversity

class

Distribution of classes (%)

Hawaii Canary New Zealand

1. _{Very low} 40.40 8.35 48.48

2. _Low 31.00 82.67 19.32

3. _Moderate 27.88 8.04 14.82

4. _High 0.54 0.63 16.87

5. _{Very high} 0.18 0.31 1.51

Table 3 shows the distribution of geodiversity scores per study area. The New Zealand study area has the highest geodiversity values. Figure 9 (A, B and C) displays the boxplots of the geodiversity indices and sub-indices per study area.

Figure 9. Boxplots of the geodiversity indices and sub-indices per study area. The purple dots are outliers and the asterisks are extreme outliers. _{A: Hawaiian archipelagos. B: Canarian archipelagos. C: New Zealand subduction zone.}

(16)

In tables 4, 5 and 6 the correlation matrices per study area are presented. The values represent the degree of linear dependency between the indices. The tables reveal large differences in correlation between the geodiversity index and the various sub-indices. Most notable is the difference in high/low correlation between the geodiversity index and the geomorphology and bathymetry indices per study area (Tables 4, 5 and 6). Furthermore, both slope derivatives are highly correlated.

Table 4._{Correlation matrix of Hawaii}

Index Geodiversity Geomorphology Bathymetry Range of

slope angle STD of slope angle Light penetration Geodiversity 1.000 Geomorphology 0.839 1.000 Bathymetry 0.429 0.101 1.000 Range of slope angle 0.208 -0.062 -0.037 1.000 STD of slope angle 0.182 -0.052 -0.013 0.836 1.000 Light penetration -0.005 -0.005 -0.002 0.010 0.004 1.000

Table 5._{Correlation matrix of Canary}

slope angle STD of slope angle Light penetration Geodiversity 1.000 Geomorphology 0.267 1.000 Bathymetry 0.914 0.014 1.000 Range of slope angle 0.271 0.003 -0.005 1.000 STD of slope angle 0.216 0.005 -0.003 0.697 1.000 Light penetration 0.006 -0.014 0.000 -0.000 -0.000 1.000

(17)

Table 6. Correlation matrix of New Zealand

slope angle STD of slope angle Light penetration Geodiversity 1.000 Geomorphology 0.089 1.000 Bathymetry 0.984 0.006 1.000 Range of slope angle 0.106 -0.007 -0.012 1.000 STD of slope angle 0.042 -0.015 -0.015 0.544 1.000 Light penetration 0.003 -0.011 -0.003 0.008 0.021 1.000

(18)

Figure 10. Mean geodiversity scores per age of the seafloor per study area. A: Hawaiian archipelagos B: Canarian archipelagos C: New Zealand subduction zone

(19)

Discussion

4.1 The geodiversity index

The research aim is to develop a geodiversity index for the seafloor that can be used to analyse geodiversity between different marine environments. The index differs from other indices due to its focus on marine environments. Therefore, this study had to use a slightly different approach. First of all, there were not much input datasets available. Most geodiversity studies have the ability to make their own selection of land surface parameters. For instance, Pereira et al. (2013) uses geology and pedology expert-based maps and Seijmonsbergen et al. (2014) uses soil and hydrology datasets. However, seafloor parameters are hard to find due to the lack of available datasets or the lack of usefulness. After all, using an hydrology index in a marine environment is irrelevant. This resulted in a narrow selection of potential seafloor parameters to be included in the geodiversity index.

Furthermore, the research aims to develop a geodiversity index for the seafloor that is transferable. Therefore, the index includes more general seafloor parameters, which are easy available. As a result all the selected sub-index datasets are derivatives of the bathymetry dataset (Ryan et al. 2009) with the exception of the geomorphology dataset (Harris et al. 2014). Also the study areas are relatively large in contrast to other studies (Chambers, 2014; Seijmonsbergen et al., 2014; Vregelaar, 2015). On the whole it gives a more comprehensive representation of the geodiversity index and the sub-indices within the chosen type of landscapes.

4.2 Interpretation of the results

4.2.1 Interpretation of the geodiversity index

The developed geodiversity index shows patterns and hotspots of geodiversity on the seafloor. It should be kept in mind that these patterns are the result of the selected input datasets. The geodiversity patterns in the Hawaiian study area are mainly caused by the diversity in geomorphic features (Table 4.). The variation in bathymetry is similarly spread across the study area, but influences the geodiversity index less strongly. These two sub-indices determine most of the geodiversity scores and therefore cause most of the patterns. The individual slope angle sub-indices only have little influence on the geodiversity index. However, both sub-indices show high diversity scores for similar locations and are often found in small clusters (see. Appendix B). As a result geodiversity hotspots mainly follow the distribution of the slope angle sub-indices.

Similar results have been found in the Canary study area. Most geodiversity hotspots follow the same pattern as both slope angle sub-indices (see. Appendix B). The average geodiversity score is caused by the variation in bathymetry in contrast to the Hawaiian study area (Tables 4 and 5). The geomorphology sub-index tends to follow the distribution of the bathymetry diversity, but covers less area (see. Appendix B). Therefore, the influence of the variation in geomorphic features on the geodiversity index is less strong. Average geodiversity patterns are mainly caused by the variation in bathymetry.

The results show an even larger difference between the influences of the sub-indices on the geodiversity index for the New Zealand study area (Table 6.). Geodiversity scores are mainly caused by the variation in bathymetry and only receives little influence of the other sub-indices.

(20)

By examining the results some general outcomes are noteworthy. Firstly, the light penetration index has negligible influence on the geodiversity index for all study areas (Tables 4, 5 and 6). For the Hawaiian study area the light penetration index is even negatively correlated with the geodiversity index (Table 4.). This can be explained by the extent of the study area and the way the light penetration index is calculated (Fig. 5). Light intensity decreases exponentially with depth in the ocean. Therefore, most light is already dispersed and absorbed before it arrives at the seafloor. This effect is reflected through the locations of high light penetration diversity scores, which are all concentrated at the coast (see. Appendix B). The seafloor depth is relatively low at the coast in contrast to most of the study area.

Secondly, in all three study areas both slope angle sub-indices are highly positively correlated with one another (Tables 4, 5 and 6). This means that cells with a large difference between minimum and maximum slope angle also contain many different slope angles. This may indicate that these locations have a rough landscape, which may contain many micro-environments.

Lastly, in general the geomorphology and bathymetry index have the largest effect on the geodiversity index (Fig. 9 and Tables 4, 5 and 6). These indices cause the average diversity patterns. However, geodiversity hotspots are largely created by the combination of the geomorphology, bathymetry and the slope angle indices.

4.2.2 Geodiversity scores and geological settings

Table 3 shows the distribution of geodiversity classes per study area. The New Zealand subduction zone stands out in area that is covered with an ‘high’/’very high’ geodiversity score. The geodiversity index for New Zealand is mainly characterized by bathymetry diversity, which is logically explained by the presence of the subduction zone. The Hawaiian and Canary study areas have mutually similar distribution of high geodiversity classes. The lower classes (1-3) differ more. An explanation is found by the geomorphology dataset. The geomorphology index in the Hawaiian study area shows a large area with a very low diversity score, while in the Canary study area most of the study area is covered with a low diversity score. As a result of the on average low geodiversity score, the geomorphology index can influence low scoring geodiversity cells enough to rise an index class.

Figure 10 shows the mean geodiversity score per age classification of the seafloor per study area. Figure 10A and 10C show an overall decrease in geodiversity score with increasing age. Figure 10B does not show a degree of linear relationship between the geodiversity score and the seafloor age. This might be explained by the distance between the study area locations, the conditions within the study areas and the lack of information between the geological timelines of the study areas. The basaltic plates of the study areas have a different origin and destination and are therefore affected by different environmental processes. So, the seafloors of the study areas are not a concatenation of the same basaltic plate. Therefore, the formation of geodiversity over time in a study area and the associated scores may not be the result of an accumulation of the same processes. This results in an incomplete comparison between the seafloor ages of the study areas. Furthermore, the seafloor ages of the study areas are not a concatenation of one time-series. So, there are large caps in information about geodiversity scores per intermediate seafloor age. This means that the age of the seafloor in relation to the geodiversity score can only be compared within a study area and not between.

(21)

4.3 Points of attention

The proposed geodiversity index for the marine environment shows geodiversity patterns and hotspots based on several selected seafloor parameters. Because there has never been a geodiversity index developed for the seafloor, there is a lack of information on seafloor geodiversity. Therefore, it is uncertain which parameters could best be used for such a index. As a result the selected seafloor parameters are based on a subjective choice and on the input datasets available. The geodiversity index could be improved by, for instance, a marine geologic or tectonic feature dataset. Testing more seafloor parameters would also give more information about the effect that these sub-indices have on the geodiversity index in different marine environments. More research opens up the potential to develop a more complete and general marine geodiversity index as well as several selectable geodiversity indices suitable for specific marine environments. For instance, the outcome of the light penetration index indicates that the sub-index is not suitable for large and deep seafloors. However, it could become an important parameter in shallow seas or fine-scale case studies.

It is also important to notice that new input data should be carefully selected. For instance, the geomorphology dataset (Harris et al. 2014) is derived from a global geomorphic feature dataset. This means that similar methods have been used to create the dataset and that all features are created on the basis of the same choices. Therefore, the dataset is suitable for every submarine location. However, using multiple datasets from different sources for the calculation of a sub-index results in an incorrect assessment of the sub-index.

Furthermore, the geodiversity index and light penetration classes are based on the natural breaks method of classification. The other sub-indices classes are based on the equal intervals method of classification. This last method is chosen because it classifies the classes equally over the possible range of diversity scores. The advantage is that classes are not adjusted to the average diversity scores of an area and therefore represent real diversity values. By the reclassification of the geodiversity index, these real diversity values are converted in diversity classes that represent geodiversity scores relative to the area. As a result of this approach, the geodiversity index maps of different areas are better comparable. However, it still should be kept in mind that the difference in minimum and maximum value of diversity differs per study area.

(22)

5. Conclusion

The developed geodiversity index has successfully created geodiversity maps for the seafloor. The geodiversity patterns and hotspots can be explained by the input datasets. The advantage of these datasets is the global extent of the data and the free availability on the world-wide-web.

The influence of the individual sub-indices on the geodiversity index differs largely within and between study areas. On average, the bathymetry diversity index was most correlated with the geodiversity index followed by the geomorphology diversity index. These indices also cause most of the geodiversity patterns. Both slope angle indices had similar influence on the geodiversity index and were highly positively correlated to one another. These indices cause most of the geodiversity hotspots. However, the light penetration diversity index had negligible effect on the geodiversity index and should therefore not be included in study areas with large and deep seafloors.

The New Zealand subduction zone has the highest values for geodiversity. If geodiversity is related to the geological setting can not be concluded. In this study the geodiversity index is largely influenced by the bathymetry diversity index. Bathymetry diversity values are in general higher in subduction zones which may have a relation to the geological setting.

The geodiversity index gives a general overview of submarine geodiversity patterns and hotspots. However, the geodiversity index does not necessarily represent actual geodiversity in practice. The methods used for the establishment of the input data and the methods used in this research may not result in a geodiversity index, that is useful for all applications. In order to make the index applicable further research is needed. The index can be used for further research on submarine geodiversity.

(23)

6. References

Chambers, R. (2014). Development of a geodiversity index on a volcanic island: Tenerife, Canary Islands.

Crofts, R., & Gordon, J.F. (2014). Geoconservation in protected areas._Parks, 20 (2), 61-76. Gray, M. (2013). Geodiversity: Valuing and Conserving Abiotic Nature, 2nd Edition. Ch. 15.1. Valuing and conserving geodiversity. _{Wiley-Blackwell.}

Hader, D.P., Kumar, H.D., Smith, R.C., Worrest, R.C. (2007). Effects of solar UV radiation on aquatic ecosystems and interactions with climate change. Photochemical & Photobiological Sciences. 6, 267– 285.

Harris, P.T., Macmillan-Lawler, M., Rupp, J., Baker E.K. (2014). Geomorphology of the oceans.

Marine Geology, 352, 4–24.

Hengl, T. (2006) Finding the right pixel size. _{Computers & Geosciences}, 32(9), 1283-1298. Müller, R.D., Sdrolias, M., Gaina, C., Roest, W.R. (2008). Age, spreading rates and spreading symmetry of the world's ocean crust. _{Geochem. Geophys. Geosyst.,} 9.

Pereira, D. I., Pereira, P., Brilha, J., & Santos, L. (2013). Geodiversity assessment of paraná state (brazil): An innovative approach. Environmental Management, 52(3), 541-552.

Ryan, W.B.F., Carbotte, S.M., Coplan, J.O., O'Hara, S., Melkonian, A., Arko, R., Weissel, R.A., Ferrini, V., Goodwillie, A., Nitsche, F., Bonczkowski, J., Zemsky, R. (2009). Global Multi-Resolution Topography synthesis. _{Geochem. Geophys. Geosyst.,} 10.

Seijmonsbergen, A.C., de Jong, M.G.G., de Graaff, L.W.S., Anders, N.S. (2014). Geodiversitat Von Voralberg Und Liechtenstein. Geodiversity of Vorarlberg and Liechtenstein. Zurich, Switzerland: HauptVerlag.

Vregelaar, M., (2015). Developing a geodiversity index for the Netherlands. _{University of Amsterdam.}

Wozniak, B., Dera, J. (2007). Light absorption in seawater. Atmospheric and Oceanographic Sciences Library. _Springer. 33.

(24)

7. Appendix

A: Models

(25)

(26)

Geomorphology model

(27)

Slope models

(28)

Geodiversity index model

(29)

B: Sub-index maps

Hawaii:

Sub-index maps of the Hawaiian archipelagos. Classes range from ‘ Very low’ (0) to ‘Very high’ (5) for diversity scores.

(30)

Canary:

Sub-index maps of the Canarian archipelagos. Classes range from ‘ Very low’ (0) to ‘Very high’ (5) for diversity scores.

(31)

New Zealand:

Sub-index maps of the New Zealand subduction zone. Classes range from ‘ Very low’ (0) to ‘Very high’ (5) for diversity scores.

(32)

C: Digital appendix

Map:_{Veloo_Thesis} Geodatabases: - Hawaii.gdb - Canary.gdb - NewZealand.gdb Mxd-file: Veloo_Thesis.mxd

A new method to analyze seafloor geodiversity around the Hawaiian and Canarian archipelagos and the New Zealand subduction zone