Geodiversity as a potential predictor of ecosystem diversity in Ecuador

(1)

Geodiversity as a potential predictor of ecosystem

diversity in Ecuador

1)

Darryl Holsboer, 11051957

02-07-2018, University of Amsterdam

Under supervision of Dr. A.C. Seijmonsbergen and Dr. K.F. Rijsdijk

(2)

2 1) A geodiversity map of continental Ecuador with a grid cell size of 1000m

(3)

3

Abstract

Geodiversity is the natural diversity of pedology, geomorphology, geology, hydrology and

topography. It is seen as the abiotic equivalent of biodiversity and has risen in relevance in recent decades. Studies pertaining to geodiversity have been conducted in Colombia, Hawaii, Austria, Brazil and Finland among others and have shown promising results. In this thesis the aim is to develop a geodiversity index for Ecuador to determine if there is correlation between geodiversity and

ecosystem diversity and if this correlation is strong or weak. Zonal statistics is applied to calculate the individual diversities of geology, geomorphology, hydrology, pedology and topography. These are reclassified and combined with the raster calculator into one map. This was performed for the Sucumbíos region, continental Ecuador and for varying grid cell sizes ranging from 500m to 4000m. From the results, the Andes mountains are concluded to be an important source of geodiversity. This is noticeable in the case study area and in continental Ecuador, where most of the geodiversity is found along the Andes. Rivers are another potential source of geodiversity in Ecuador with much of the geodiversity in Eastern Ecuador being located near rivers. Moreover, pedological diversity, hydrological diversity and topographical diversity explain most of the ecosystem diversity. Grid cell size is observed to be positively related to the overall correlations between geodiversity and ecosystem diversity. A cell size dependency for Ecuador at a national scale is thus suggested. Lastly, Geodiversity is perceived as a moderate predictor for ecosystem diversity at higher than 4000m grid cell sizes for continental Ecuador. For the case study area geoconservation and land use management remains a viable alternative application of geodiversity, as geodiversity is only weakly correlated with ecosystem diversity.

(4)

4

Content

1. Introduction ... 5

1.1 Context and relevance ... 5

1.2 Theoretical Framework ... 6

1.2.1 Geodiversity in other studies ... 6

1.2.2 Diversity in Ecuador ... 7

1.3 Aim and research question... 8

2. Methods and data ... 9

2.1 Preprocessing ... 9

2.2 Analysis ... 11

2.3 Deliverables ... 12

3. Results ... 13

4. Discussion ... 17

4.1 Geodiversity in the case study area ... 17

4.2 Geodiversity in continental Ecuador and its relation to the case study area ... 18

4.3 The effects of varying grid cell sizes ... 19

4.4 Improving methodology ... 20

5. Conclusion ... 21

6. Acknowledgements ... 22

7. References ... 23

8. Appendices ... 26

8.1 Appendix A: Input maps of Ecuador ... 26

8.2 Appendix B: Geodiversity maps of the case study area ... 31

8.3 Appendix C Thematic diversity maps of the case study area ... 33

8.4 Appendix D: Geodiversity maps continental Ecuador ... 45

8.5 Appendix E: Thematic maps continental Ecuador ... 48

8.6 Appendix F: Correlation matrices of continental Ecuador ... 72

(5)

5

1. Introduction

1.1 Context and relevance

Ecuador is diverse country, divided by the Andes mountains and possessing one of the most

biodiverse regions on earth in the Amazonia region. Still, much remains unknown, as Eastern Ecuador is home to large territories of undisturbed rainforest. Moreover, no geodiversity related study has been performed in this country yet. This makes it an interesting research area, especially due to the Western Amazonia region in particular possessing the greatest floral diversity of Amazonia (Hoorn et al., 2010).

Geodiversity can be interpreted in multiple ways, however for the purpose of this thesis geodiversity is defined as the natural diversity in pedology, geomorphology, geology, hydrology and topography. This combines the definition of Gray (2008), which includes geomorphology, geology and pedology, with newer additions proposed by Serrano & Ruiz-Flanõ (2007). It is often seen as the abiotic counterpart of biodiversity, with biodiversity being the range of flora, fauna and other living organisms Gray (2008).

Moreover, geodiversity has a wide range of applications and its relevance has risen in the recent decades, as is apparent in the rise of geoparks and geotourism. The latter, if handled well is a useful tool for protecting and promoting geosites, but also to educate people, or to provide jobs for local communities (Newsome et al., 2012).

The protection of geodiversity is important, as similarly to biodiversity, geodiversity is sometimes compromised by anthropogenic influences and unlike biodiversity where the use of DNA depositories and zoos facilitates restoration, geodiversity is often lost after destruction Gray (2008).

Biodiversity has long been an important element of conservation in ecosystems. This is not likely to change as climate change is believed to play a significant role in the loss of biodiversity in the future (Bellard et al., 2012). The significance of biodiversity is not exclusive to science, as ecosystem services also depend on biodiversity, which are not only essential for the provisioning of food and water, but also for recreation, cultural inspiration, waste treatment and protection against extreme events among others (Bellard et al., 2012; Crossman et al., 2013).

Geodiversity plays an important role in maintaining this biodiversity. The Aichi biodiversity targets set by the convention of biological diversity help to illustrate this fact (CBD, 2018). Targets such as the protection of ecosystem services, the minimalization of anthropogenic influences and the reduction of habitat loss for instance are targets that are just as important for geodiversity. Protecting

geodiversity will thus also protect biodiversity. Furthermore, geodiversity is an active contributor to ecosystem services and forms the basis to many of them. Some of the services that geodiversity provides: provisioning of water, coastal protection, cultural heritage, knowledge of natural processes and resources for recreational or economical purposes (Gordon et al., 2012).

Furthermore, geodiversity is relevant for land use management, where it can be used as a tool by policymakers. In a study conducted by Santos et al. (2017) for instance urbanisation was discovered to be a considerable threat to areas of high geodiversity in the region of Armação dos Búzios in Brazil. Finally, geodiversity could also be used as a predictor of biodiversity or in this case ecosystem

(6)

6 has the disadvantage of being time consuming and expensive. Additionally, many of these surveys occur at a very local level, which means that extrapolation is a necessity, which in turn can generate questionable results when the quality of the original dataset is suboptimal (Parks & Mulligan, 2010). Using a geodiversity index as a surrogate is therefore seen as a viable alternative, as many elements of geodiversity are mapped on a very detailed scale and are easily obtainable. Datasets for

topographical and hydrological diversity can for instance be derived from digital elevation maps, which are available globally in a variety of resolutions (ibid).

The geodiversity index will be developed with zonal statistics. This method calculates diversities for thematic maps such as: geology and hydrology by counting the number of pixels in a predefined zone. These diversity maps are then combined to form the geodiversity index. This method has proven to be quite effective in geodiversity related studies, which is the reason why in this study it will be applied to Ecuador (Pereira et al., 2013; Seijmonsbergen et al., 2017; Santos et al., 2017).

1.2 Theoretical Framework

1.2.1 Geodiversity in other studies

Similar studies pertaining to geodiversity have been performed in Colombia, Hawaii, Austria, Brazil and Finland among others and have shown promising results (Park & Mulligan, 2010; Pereira et al., 2013; Seijmonsbergen et al., 2018a; Seijmonsbergen et al., 2018b; Tukiainen et al., 2017). These studies present a wide range of applications, in which geodiversity may be applied.

In Colombia a preliminary geodiversity index was created and compared with diversity in fauna, which appeared to correspond the most with bird diversity. No statistical analyses were however applied in this study (Park & Mulligan, 2010). In Hawaii, significant positive correlations were discovered for very high and high geodiverse areas in relation to the age of its islands, with the largest portion of variance being explained by the topographical diversity (Seijmonsbergen et al., 2018a). In Austria, two methods were used to determine geodiversity namely: an inventory

assessment on a regional scale and a geoconservation based approach on a local scale. Both methods have proven to yield useful results, with important factors on the local scale being morphological processes, soil properties and slope angles (Seijmonsbergen et al., 2018b). In Brazil a methodology was devised to achieve the maximum number of geodiversity units. Geological, hydrological and geomorphological diversity were perceived to be the most beneficial. Other interesting indices were paleontological and mineral diversity, however these require detailed local datasets and could lead to inflated geodiversity scores (Pereira et al., 2013). In Finland geodiversity was combined with climatic variables to analyse its relation with threatened species richness. Geodiversity contributed significantly in explaining the variation of species richness (Tukiainen et al., 2017).

In Ecuador research has been conducted for the relation between topography and tree diversity. However only a minor correlation was discovered and since topography is only one element of geodiversity further research should be conducted (Valencia et al., 2004). Furthermore, topography has been an important component in the geodiversity index of the previous case studies, which warrants a closer inspection in Ecuador.

(7)

7

1.2.2 Diversity in Ecuador

Geological diversity

Ecuador is very diverse in geology due to the Andes mountain Range, which has been integral in the geological history of Ecuador. The Andes divides Ecuador in three distinct areas. The western Costa, central Sierra and eastern Oriente. Rocks in Eastern Ecuador are mostly river sediment based such as sandstone, shale and conglomerates, while rocks in Western Ecuador are mainly marine sediment and volcanic based such as limestone and breccia (Fig. 12 in Appendix A). Metamorphic rocks and igneous rocks such as schist and andesite are found along the Andean mountains (Feiniger & Bristow, 1980).

Hydrological diversity

In Ecuador rivers are found almost everywhere (Fig. 8 in Appendix A). However, a distinction can be made between intermittent and perennial rivers, or between areas of high and low drainage densities. Western Ecuador is drier than Eastern Ecuador (Dodson & Gentry, 1991). Therefore, it likely also possesses more intermittent rivers, as more rivers in Western Ecuador are dependent on precipitation as opposed to Eastern Ecuador. Additionally, Eastern Ecuador is part of the Amazon river basin. The largest drainage basin on earth, so river densities are likely to be high in this area of Ecuador (Mora et al., 2010).

Topographical diversity

Topographical diversity is also significantly influenced by the Andes mountains. With the highest areas of elevation being found in Central Ecuador along the Andean Cordillera, while the lowest elevations are found in Eastern Ecuador and on the coastal plains of Western Ecuador (Fig. 11 in Appendix A; Jaillard et al., 2000).

Pedological Diversity

Similarly to geology, pedology is quite varied in Ecuador. Soils east from the Andes are mostly characterised by their weathered state, high water tables and weakly developed horizons. These include ultisols, inceptisols and entisols. Soils in central and Western Ecuador are more nutrient rich with Vertisols, Andisols and mollisols being some examples (Fig. 10 in Appendix A). The difference in fertility can also be observed in the land use of Ecuador with most of the crops cultivated in West and Central Ecuador (de Koning et al., 1998).

Geomorphological diversity

The geomorphological diversity of Ecuador is also present throughout the country. In the eastern Amazon region, processes such as erosion, inundation and sedimentation dominate the region, while central Ecuador is very diverse due to tectonic, volcanic and glacial processes. In western Ecuador oceanic, erosion and mass movement processes are more common (Fig.12 in Appendix A). Tectonic and volcanic activity is attributed to tectonic movement off the coast of Ecuador, where the Nazca plate subducts under the South-American plate. Over 200 active volcanoes are believed to be located in the Andes, of which 55 are found in Ecuador (Stern, 2014; Kumagai, 2007).

(8)

8

1.3 Aim and research question

The aim of this research is to develop a geodiversity index for Ecuador, which has not been done before. The index is compared with the existing ecosystem map of Ecuador to determine if there is a correlation between ecosystem diversity and geodiversity and how strong or weak this correlation is. Not only the geodiversity is compared but also its individual components namely: Geology,

geomorphology, hydrology, pedology and topography to examine which components explain the largest amount of variance. The geodiversity index will initially be tested on a regional level,

thereafter it will be upscaled to a national level. A comparison is then made between correlations on a regional level and correlations on a national level.

The main question in this thesis is therefore:

What is the relation between a geodiversity index and the existing ecosystem map of Ecuador? This main question is subdivided in the following sub questions:

How can a geodiversity index be constructed?

How much variation is explained by the geodiversity index and its individual components in the case study area?

What is the relation between geodiversity and ecosystem diversity after upscaling the geodiversity index?

What is the effect of varying fishnet cell sizes on the results of both the case study area and continental Ecuador?

(9)

9

2. Methods and data

In this thesis the methodology was divided in three parts, of which a workflow is presented (Fig.1). The preprocessing phase was primarily characterised by data acquisition and preparation, whereas in the analysis phase the geodiversity, ecosystem diversity and diversities of the sub-indices were calculated with zonal statistics. In the deliverables phase, maps were created and correlation matrices were presented in tables to visualise the results.

Figure 1: Workflow of the methodology divided in three phases

2.1 Preprocessing

The input data was acquired from various open sources (Tbl.1). The following input maps were used for the geodiversity index: geomorphology, river, pedology, digital elevation and an ecosystem map. From these maps geological, geomorphological, hydrological, topographical and pedological diversity was calculated, which together formed the geodiversity index (Tbl.1).

The geodiversity index is formulated as follows:

GI= Hydrological diversity index(Hdi)+ Topographical diversity index(Rdi) + Geological diversity index(Gdi) + Geomorphological diversity index(Gmdi) + Soil diversity index(Sdi).

Subsequently, ecosystem diversity was calculated to which geodiversity was compared. Every sub-index was weighted equally, as it permitted a fairer comparison between the individual indices and ecosystem diversity.

For the geological diversity map, no useable maps were available, so as a substitute the lithology field in the geomorphology map was used as a measure of geological diversity. Furthermore, the topographic diversity used slope as an indicator for diversity, which already has precedent in other geodiversity studies (Serrano & Ruiz-Flanõ, 2007). Lastly, drainage density was applied as a measure of hydrological diversity and has been utilised in multiple geodiversity studies already (Manosso & de Nóbrega, 2016; Argyriou et al., 2016).

(10)

10 Table 1: Metadata

In order to use zonal statistics a grid had to be created, in which calculations could take place. For this the create fishnet tool was utilised. An important setting in this tool is the cell size, which determines the surface area of each cell.

The optimal grid cell size was therefore calculated based on the method developed by Hengl (2006). With this method the coarsest (1), finest (2) and recommended grid resolutions are obtained.

(1)

(2)

SN stands for scale number and is determined by the least detailed input map, as the geodiversity index is only as detailed as its least detailed map. In this study the river map with a scale of

1:1.000.000 had the lowest resolution. Using the equation this leads to a recommended grid cell size of 1.000.000 * 0.0005 = 500 meters, as the recommended value is situated between the highest and lowest effective grid cell sizes.

In this case 500 meter was computationally manageable, in other cases a compromise would have to be made between resolution and rendering time.

Scale/ cell size

Pedology Soil units used for soil

diversity Polygon GCS_Provisio nal_S_Americ an_1956 1:25.000 2017 Ministerio de agricultura y ganaderia Geomorphology Geomorphological units used for geomorphological diversity Polygon GCS_SIRGAS 1:25.000 2017 Ministerio de agricultura y ganaderia

Geomorphology Lithology used for

geological diversity Polygon GCS_SIRGAS 1:25.000 2017

Ministerio de agricultura y ganaderia Hydrology Rivers used for

hydrological diversity Polyline

GCS_WGS_19 84 1:1.000.000 2006 WWF NASA/USGS/ JPL-CALTECH Gabriela Viteri, Ministerio del Ambiente ArcGIS online, dicamarco Provinces Ecuador and its

provinces used as extent Vector

GCS_WGS_19

84 Unknown 2016

Ecosystem Vegetation zones used

for ecosystem diversity Vector

GCS_WGS_19

84 1:100.000 2018

DEM Surface elevation used

for topographic diversity Raster

WGS_1984_U

TM_Zone_17S 50 x 50 2007

Map Description Data

type

Coordinate system

Publication

(11)

11 500 meters was thus taken as the initial grid cell size, however higher grid cell sizes were also used to compare the effects of cell size on the correlations between geodiversity and ecosystem diversity. These higher grid cell sizes were 1000m, 2000m and 4000m respectively.

For the case study area, the Sucumbíos region in North-East Ecuador was chosen, due to its elevation gradient from west to east, it’s computationally manageable size and the presence of the amazon rainforest. The strategy was to first test the methodology in this case study area. The methodology was subsequently upscaled at a national level to compare the potential differences. The Galapagos islands were left out for convenience. Finally, different grid cell sizes were compared with each other for continental Ecuador and the case study area.

After determining the case study area, the input maps were prepared for the analysis phase. The preparation started with reprojection of input data. SIRGAS 2000 UTM Zone 17S was used as the geographic coordinate system, which is suited for mapping continental Ecuador for scientific purposes (EPGS, 2018).

The reprojected maps were then clipped to the extent of Ecuador and the case study areas. The hydrological and topographical maps required some additional preparation to calculate drainage densities and slope angles, for this purpose the line density and slope tools were utilised. For the line density a search radius of 500 meters was taken, equivalent to the recommended grid cell size, while for the slope default settings were used.

The digital elevation map used to create a slope map required some modifications. As the map contained negative values way below what was deemed realistic. Consequently, all values below zero were changed to NoData, which for Ecuador made the most sense.

The pedology map required some editing in the attribute table before it could be used. This was due to two soil classification methods described in the attribute table. The United states department of agriculture soil taxonomy (USDA) and the world reference base (IGW, 2015). The latter is used worldwide, but was not complete in this dataset and was as such left unused. A new field was added to which the main and sub groups of the USDA soil taxonomy were joined, so that a larger variation in soil could be represented. These main and sub groups were represented by the Orden and Suborden fields in the used soil map.

The final preparation step was the rasterization of the vector input maps, since zonal statistics requires the input to be of a raster filetype (Fig.1). For this the feature to raster tool was used. 50m was taken as a grid cell size, equal to the input digital elevation map (DEM). According to the

equation of Hengl, 12.5 meters would have been the recommended cell size to get the most amount of detail from the maps with high resolutions, but due to computational limitations, a higher cell size was chosen, which still falls in the range of 6.25 to 62.5 meters calculated by the equation of Hengl. This cell size was used for all maps, including the raster map of the drainage density created by the line density tool, which automatically turned the vector river map in a raster map.

2.2 Analysis

In the analysis phase zonal statistics was applied to calculate the diversities of the thematic maps (Fig.1). This tool calculates a predefined statistic for a given extent, which can be a mean, sum, standard deviation, min, max or variety. For the majority of the maps, all input maps excluding the drainage density and slope maps, variety were calculated with zonal statistics.

For the drainage map the mean density per grid cell was calculated. The choice of mean was made, because it indicates better which areas are affected the most by hydrology, however in practice the

(12)

12 mean and range give similar results. For the slope map the standard deviation was calculated as a measure of diversity, this in substitute of the range, which did not always give realistic results. However, both statistics are usable for measuring geodiversity (Seijmonsbergen et al., 2018a). The next step in the analysis was the reclassification of all maps, which is necessary because of the wide range of values between maps (Fig.1). This was performed with the reclassify tool. For this tool the default settings were used. The number of classes were set at five. The classification method used was Jenks natural breaks, which groups classes according to the least amount of variance within groups and the largest amount of variance between groups (ESRI, 2017).

After reclassification all thematic maps were summed into one geodiversity map with the raster calculator tool. The geodiversity map was thereafter reclassified like the thematic maps.

The final step was calculating correlations with the band collection statistics tool, which is a tool that can calculate regular statistics, covariance and correlation matrices of all input datasets. All thematic diversity maps and the geodiversity index were used as inputs in this tool.

2.3 Deliverables

The resulting correlation matrices were exported to excel and visualised (Fig.1). Additionally, the attribute tables of the geodiversity maps were exported to excel to calculate the areal change of all classes with increasing cell sizes. Lastly, maps were created of all thematic diversity maps and

geodiversity indices of the case study area and continental Ecuador. For the maps the labels very low, low, medium, high and very high were applied to the five classes grouped with the Jenks natural breaks method.

(13)

13

3. Results

In this section the results are presented in order of sub questions. First the results for the case study area are explored at the initial fishnet cell size of 500m. Hereafter, the results of continental Ecuador are presented and compared with the case study area. Lastly, the results of the different cell sizes are presented and examined for both the case study area and continental Ecuador.

Figure 2: Geodiversity map of the case study area at a grid cell size of 500m

Table 2: Correlations for all maps of the case study area at a grid cell size of 500m

In the case study area (Fig.2) the highest geodiversity was found in the west in the Andes mountains. In the east areas of high geodiversity could however be found along rivers.

In the statistical results, geomorphology, geology and pedology were perceived to explain the majority of the variation of the geodiversity index (Tbl.1). High positive correlations (>0.7) were also discerned between these thematic maps, with geology being strongly correlated to geomorphology and pedology being moderately correlated with geomorphology and geology. Hydrology and topography were furthermore recognized to be moderately positively correlated (>0.3 - <0.7) to the geodiversity index.

Layer Pedology Geology Geomorphology Hydrology Topography Geodiversity Vegetation

Pedology 1 Geology 0.67 1 Geomorphology 0.68 0.74 1 Hydrology 0.25 0.27 0.18 1 Topography 0.22 0.15 0.28 0.00 1 Geodiversity 0.76 0.76 0.78 0.47 0.55 1 Vegetation 0.17 0.11 0.06 0.14 -0.03 0.11 1

(14)

14 The correlations between geodiversity and ecosystem diversity on the other hand were low (<0.3) (Tbl.1). Among these correlations, pedology and hydrology explained most of the variation. Topography explained the least amount of variation and was slightly negatively correlated.

(15)

15 Table 3: Correlations for all maps of continental Ecuador at a grid cell size of 500m

For continental Ecuador most of the high and very high geodiversity was found in the west of the country and along the Andes mountain range (Fig.3). The statistical results were similar to the case study area. Geomorphology, geology and pedology still explained most of the variation (Tbl.2). The topographical index was also the sub-index which had the highest correlation with ecosystem diversity, while the hydrological index had the lowest correlation with ecosystem diversity.

Another interesting result was the correlation between geodiversity and ecosystem diversity, which was higher for continental Ecuador (0.17) than for the case study area (0.11).

Figure 4: Correlations for different grid cell sizes between geodiversity and ecosystem diversity for continental Ecuador and the case study area.

Pedology 1 Geology 0.57 1 Geomorphology 0.62 0.68 1 Hydrology 0.19 0.26 0.17 1 Topography 0.18 0.11 0.22 0.00 1 Geodiversity 0.71 0.71 0.75 0.38 0.58 1 Vegetation 0.12 0.08 0.09 0.06 0.17 0.17 1

(16)

16 Figure 5: Percental change of geodiversity classes by increasing grid cell size

Figure 6: Percental change of geodiversity classes by increasing grid cell size (continental Ecuador) -100 -50 0 50 100 150 200 250 300 1000m 2000m 4000m Perc en ta l ch an ge

Fishnet cell sizes

Very low Low Medium High Very High

-100 0 100 200 300 400 500 1000m 2000m 4000m Perc en ta l ch an ge

Fishnet cell sizes

(17)

17 For the different fishnet cell sizes a general increase in correlation was perceived for increasing cell sizes. Both the case study area and continental Ecuador had increasing correlations, however the increase in correlations were bigger for continental Ecuador, which could be due to the bigger contribution of topography (Fig.4,Tbl.3).

Moreover, after examining the individual classes, the very high, high, medium and low classes in the case study area were observed to increase in area with cell size as a whole. Of these classes the very high and high classes increased the most (Fig.5).

The results for continental Ecuador were similar except that the low class decreased as a whole in addition to the very low class. Moreover, the maximum areas for the very high and high classes were perceived for the 2000m fishnet cell size instead of the 4000m fishnet cell size (Fig.6).

4. Discussion

4.1 Geodiversity in the case study area

In the case study area zones with high and very high geodiversity were found along rivers. The importance of rivers is reflected in the correlations of the individual sub-indices, as the correlations were highest for the pedological and hydrological diversity indices (Tbl.2).

The reason why geodiversity and ecosystem diversity is higher along rivers, could be explained by the input maps. Data collection often starts in the field. It is possible that more data was collected along existing natural and artificial infrastructure such as rivers and roads, which would introduce some sampling bias in favour of river areas (Reddy & Dávalos, 2003).

Alternatively, rivers form natural barriers and transform landscapes, which may contribute to geodiversity and ecosystem diversity. According to Ward et al., (1999) floodplain river ecosystems are among the most species rich, due to the variety of ecotones, which are transitional zones between ecosystems. It is believed that high disturbance due to flooding and fragmentation leads to increases in biodiversity in these ecosystems where resources are abundant.

Rivers furthermore transform landscapes due to erosion and sedimentation, these processes play a role in geomorphological diversity, soil diversity and topographical diversity. Floodplains, alluvial fans and V shaped valleys, products of rivers are examples of geomorphology (Latrubesse et al., 2005), fluvisols and gleysols are soils characterised by fluvic processes or stagnation of water (IGW, 2015), and erosion changes relief which influences slopes (Montgomery & Brandon, 2002).

The higher correlation in comparison to the other sub-indices between pedological diversity and ecosystem diversity is explained by the important role of soils in supporting biodiversity. Soils contain a significant portion of the total biodiversity on land with regard to fauna. Soil diversity plays an important role in soil biodiversity (Havlicek & Mitchell, 2014). This soil biodiversity is an important driver of floral diversity, as soil microbes are essential in providing resources for vegetation (Van Der Heijden et al., 2008).

The lowest variation in ecosystem diversity explained by a sub-index was topographical diversity, which approached 0. This is interesting and could be explained by the spatial distribution of the diversity classes. For topography a clear gradient is present from west to east with decreasing slope angles, while ecosystem diversity is more irregularly distributed, especially for the medium diversity

(18)

18 class, which is present in the entire case study area. Studies examining the effects of topography on tree and habitat diversity in tropical forests, including one study in Eastern Ecuador, do suggest that topography only contributes slightly to tree and habitat diversity, as only 25% of the examined species were believed to be specifically adapted for topography (Valencia et al., 2004).

The overall and individual correlations between geodiversity and ecosystem diversity were relatively low, which was unexpected. This could be attributed to a variety of reasons. First, because of the calculated ecosystem diversity, which did not have many areas of high and very high diversity in the case study area (Fig.36 in Appendix B). In addition, most of these high and very high diversity areas were located along rivers in the east of the case study area in the amazon rainforest. This in contrast to high and very high levels of geodiversity, which were primarily found along rivers in the west of the case study area in the Andes mountains (Fig.2).

Moreover, ecosystems host a broad collection of plants and animals, thus a comparison with a geodiversity index consisting of more detailed and abundant soil types for instance might logically yield lower correlations.

Lastly, other factors not taken into consideration could have contributed to the predicting power of the geodiversity index. Such as mineral occurrences or palaeontology, which have for instance been used in geodiversity studies in Brazil (de Paula Silva et al., 2015;Pereira et al., 2013). The former is a measure of natural resources, while the latter measures the variation in fossils (Pereira et al., 2013). In the geodiversity index most of the variation was explained by the geomorphological, pedological and geological diversity indices. Unsurprising, since these maps are quite detailed and possess a lot of classes and thus variation. These indices however also correlated strongly with each other, which can be attributed to the fact that soils are often dependent on geology and geomorphology. Geology determines parent material, while geomorphological processes such as erosion, sedimentation and inundation changes soil (Jenny, 1994; IGW, 2015).

4.2 Geodiversity in continental Ecuador and its relation to the case study area

For continental Ecuador the overall correlation between geodiversity and ecosystem diversity was higher than for the case study area. This was expected, since the variability is much greater on a national scale, as western and eastern Ecuador have differing ecosystems due to the Andes mountains (Peters et al., 2010).

An interesting difference between the case study area and continental Ecuador was the relation between hydrology and ecosystem diversity and topography and ecosystem diversity. Topography was a relatively important index for continental Ecuador, while hydrology was the least important. This is an almost opposite result of the case study area. This difference might be explained by the influence of the Andes, which is more significant for continental Ecuador than for the case study area, where the Andes only makes up a fraction of the total area.

Additionally, in Western Ecuador less of a gradient is perceivable in slope and the overall

topographical diversity is higher than in the east, which includes the case study area. Even though topography is believed to have minimal influence on tree and habitat diversity in tropical rainforests, elsewhere topography is often cited as one of the more important indicators of ecosystem diversity,

(19)

19 for instance in South-Western Ecuador where the dominant ecosystems are more influenced by topography (Coblentz & Keating, 2008).

4.3 The effects of varying grid cell sizes

The correlation between geodiversity and ecosystem diversity was determined to be higher for continental Ecuador than for the case study area. This was performed for a grid cell size of 500m, which for a country of roughly 280.000 km2_{is a fine sized grid (Food and Agriculture Organization,} 2018).The entire methodology was therefore subsequently tested for 1000m, 2000m and 4000m fishnet cell sizes, which according to expectation would lead to an increase in correlations.

The results confirmed this hypothesis and can be explained by how aggregation is performed in zonal statistics. If the cell size of individual cells in a grid is increased, more values are taken into account (Fig.7). In areas where diversity is clustered, this leads to aggregation, as smaller clusters combine into larger clusters. Additionally, the detail of the input maps determines the diversity. With a 500m grid, some maps might be less diverse than others, because of the resolution. In the rasterization phase all maps were standardized at a cell size of 50 m, however, many maps were more detailed than that, so some detail was omitted. Decreasing the cell size of the raster maps might yield better results, but is also more computationally expensive.

Figure 7: The effect of aggregation on a geomorphological raster map after calculating the variety in zonal statistics. Left a grid cell size of 4000m is used and right a grid cell size of 500m.

Diversity Very Low Low Medium High Very High

(20)

20 The higher correlations for continental Ecuador as opposed to the case study area could also be explained by the greater variation in geodiversity and ecosystem diversity for continental Ecuador. Ecosystem diversity is already relatively minor in the case study area, so further aggregation would not improve the correlations by much (Fig.39 in Appendix B). Correlations for the case study area even appear to reach an optimum (Fig.4), however more fishnet cell sizes would have to be

examined to confirm this observation. Greater fishnet cell sizes would also perhaps lead to moderate correlations for continental Ecuador, if the linear relation in Fig.4 is extrapolated.

Another result was the areal increase in geodiversity classes with increasing cell size (Fig.5). The highest increases were attributed to the high and very high classes. This is logical, since it is a relative increase. High and very high classes have smaller areas, so the relative increase is bound to be bigger, compared to other classes. However, for continental Ecuador the largest increase was not found for the 4000m grid, but for the 2000m grid (Fig.6). It is unclear whether this indicates an optimal grid size for high and very high geodiversity in continental Ecuador, or whether this occurred due to an error in the analysis phase.

It is however clear that the choice of grid size is dependent on the application of use. In ecological assessments scales ranging from continental to regional are applied, while for climate vulnerability assessments are conducted at landscape and ecoregion scales (Comer et al., 2015). The effect of grid size on geodiversity was also examined in Great Britain, where higher scales led to less geodiversity (Bailey et al., 2017).

In this study geodiversity has some potential use as a predictor of ecosystem diversity in continental Ecuador at higher grid cell sizes. At the case study area level geodiversity is more appropriate for land use management and geo-conservation, since the explanatory power for ecosystem diversity is relatively minor in this area.

4.4 Improving methodology

This methodology is a useful tool in determining areas of high geodiversity and has been tried and tested in a multitude of studies already with promising results. Most of the uncertainty in this thesis stems from the input maps, but is not easily resolved, as for Eastern Ecuador especially data remains quite scarce. The utilised ecosystem map could nonetheless be enhanced by using biodiversity data from databases such as the Global Biodiversity Information Facility (GBIF). This data could be used to create boosted regression trees, which are models that predict species richness (Bailey et al., 2017). These models can then be applied on a raster and be used in spatial analyses.

Furthermore, validity could be improved by choosing geodiversity indicators with the least amount of bias. This can be difficult as many commonly used indicators in geodiversity are also correlated with each other, as was the case in this thesis. This can be resolved by using weights for the individual indicators, however this does require some expert knowledge. Lastly, this study will form the basis for future research. The influence of greater cell sizes, using data of species diversity and upscaling to an even larger research area are examples of subjects, which will require further examination.

(21)

21

5. Conclusion

In this thesis geodiversity was compared with ecosystem diversity in Ecuador to determine the correlation and strength of this correlation. Interesting results were produced and interpreted. Firstly, the Andes mountains are concluded to be an important source of geodiversity. This is

noticeable in the case study area and in continental Ecuador, where most of the geodiversity is found along the Andes. Rivers are another potential source of geodiversity in Ecuador with much of the geodiversity in Eastern Ecuador being located near rivers. For the case study area specifically however sampling bias has to be taken in to consideration, as Eastern Ecuador is still relatively unexplored.

Secondly, pedological diversity, hydrological diversity and topographical diversity explain most of the ecosystem diversity. For continental Ecuador topography and pedology explain most of the

ecosystem diversity, while for the case study area hydrology and pedology explain most of the ecosystem diversity. Topography thus appears to be less influential for ecosystem diversity in the river dominated region of the Oriente.

Thirdly, grid cell size is observed to be positively related to the overall correlations between

geodiversity and ecosystem diversity. For continental Ecuador the effect of grid cell size is stronger, due to the potential higher variation in geodiversity and ecosystems. In the case study area the effects of different grid cell sizes are less noticeable. This suggests a cell size dependency for Ecuador at a national scale.

Lastly, Geodiversity is perceived as a moderate predictor for ecosystem diversity at higher than 4000m grid cell sizes for continental Ecuador. There is potential for stronger correlations, but more research should be conducted to verify this hypothesis. For the case study area geoconservation and land use management remains a viable alternative application of geodiversity, as geodiversity is only weakly correlated with ecosystem diversity.

(22)

22

6. Acknowledgements

The author would like to thank Dr. A.C. Seijmonsbergen for the invaluable feedback and supervision, which proved essential in the completion of this thesis. Gabriel Muñoz for providing the necessary datasets and metadata, Alex de Meyer for his datasets and preliminary research, Dr. W.M de Boer for the use of his GIS studio and Paolo Tasseron for correcting format and language.

(23)

23

7. References

Argyriou, A. V., Sarris, A., & Teeuw, R. M. (2016). Using geoinformatics and geomorphometrics to quantify the geodiversity of Crete, Greece. International journal of applied earth observation and geoinformation, 51, 47-59.

Bailey, J. J., Boyd, D. S., Hjort, J., Lavers, C. P., & Field, R. (2017). Modelling native and alien vascular plant species richness: At which scales is geodiversity most relevant?. Global Ecology and

Biogeography, 26(7), 763-776.

Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W., & Courchamp, F. (2012). Impacts of climate change on the future of biodiversity. Ecology letters, 15(4), 365-377.

Coblentz, D., & Keating, P. L. (2008). Topographic controls on the distribution of tree islands in the high Andes of south‐western Ecuador. Journal of Biogeography, 35(11), 2026-2038.

Comer, P. J., Pressey, R. L., Hunter Jr, M. L., Schloss, C. A., Buttrick, S. C., Heller, N. E., ... & Shaffer, M. L. (2015). Incorporating geodiversity into conservation decisions. Conservation Biology, 29(3), 692-701.

Crossman, N. D., Burkhard, B., Nedkov, S., Willemen, L., Petz, K., Palomo, I., ... & Alkemade, R. (2013). A blueprint for mapping and modelling ecosystem services. Ecosystem services, 4, 4-14.

De Koning, G. H. J., Veldkamp, A., & Fresco, L. O. (1998). Land use in Ecuador: a statistical analysis at different aggregation levels. Agriculture, ecosystems & environment, 70(2-3), 231-247.

de Paula Silva, J., Rodrigues, C., & Pereira, D. I. (2015). Mapping and analysis of geodiversity indices in the Xingu river basin, Amazonia, Brazil. Geoheritage, 7(4), 337-350.

Dodson, C. H., & Gentry, A. H. (1991). Biological extinction in western Ecuador. Annals of the Missouri Botanical Garden, 273-295.

Esri. (2017). Data classification methods. Retrieved from:

http://pro.arcgis.com/en/pro-app/help/mapping/layer-properties/data-classification-methods.htm. Accessed on: May 8 2018. Feininger, T., & Bristow, C. R. (1980). Cretaceous and Paleogene geologic history of coastal Ecuador. Geologische Rundschau, 69(3), 849-874.

Food and Agriculture Organization. (2014). Ecuador. United Nations. Retrieved from: http://www.fao.org/countryprofiles/index/en/?iso3=ECU. Accessed on: 30-06-2018

Gordon, J. E., Barron, H. F., Hansom, J. D., & Thomas, M. F. (2012). Engaging with geodiversity—why it matters. Proceedings of the Geologists' Association, 123(1), 1-6.

Gray, M. (2008). Geodiversity: developing the paradigm. Proceedings of the Geologists' Association, 119(3-4), 287-298.

Havlicek, E., & Mitchell, E. A. (2014). Soils supporting biodiversity. In Interactions in Soil: Promoting Plant Growth (pp. 27-58). Springer, Dordrecht.

(24)

24 Hoorn, C., Wesselingh, F. P., Ter Steege, H., Bermudez, M. A., Mora, A., Sevink, J., ... & Jaramillo, C. (2010). Amazonia through time: Andean uplift, climate change, landscape evolution, and biodiversity. science, 330(6006), 927-931.

IWG, W. (2015). World reference base for soil resources 2014, update 2015. International soil classification system for naming soils and creating legends for soil maps. World soil resources reports. FAO, Rome Google Scholar.

Jaillard, E., Hérail, G., Monfret, T., Díaz-Martínez, E., Baby, P., Lavenu, A., & Dumont, J. F. (2000). Tectonic evolution of the Andes of Ecuador, Peru, Bolivia and northernmost Chile. Tectonic Evolution of South America, 31, 481-559.

Jenny, H. (1994). Factors of soil formation: a system of quantitative pedology. Courier Corporation. Kumagai, H., Yepes, H., Vaca, M., Caceres, V., Naga, T., Yokoe, K., ... & Vasconez, F. (2007). Enhancing volcano‐monitoring capabilities in Ecuador. Eos, Transactions American Geophysical Union, 88(23), 245-246.

Latrubesse, E. M., Stevaux, J. C., & Sinha, R. (2005). Tropical rivers. Geomorphology, 70(3), 187-206. Manosso, F. C., & de Nóbrega, M. T. (2016). Calculation of geodiversity from landscape units of the Cadeado Range region in Paraná, Brazil. Geoheritage, 8(3), 189-199.

Montgomery, D. R., & Brandon, M. T. (2002). Topographic controls on erosion rates in tectonically active mountain ranges. Earth and Planetary Science Letters, 201(3-4), 481-489.

Mora, A., Baby, P., Roddaz, M., Parra, M., Brusset, S., Hermoza, W., & Espurt, N. (2010). Tectonic history of the Andes and sub-Andean zones: implications for the development of the Amazon drainage basin. Amazonia: Landscape and species evolution, 38-60.

Newsome, D., Dowling, R., & Leung, Y. F. (2012). The nature and management of geotourism: A case study of two established iconic geotourism destinations. Tourism management perspectives, 2, 19-27.

Parks, K. E., & Mulligan, M. (2010). On the relationship between a resource based measure of geodiversity and broad scale biodiversity patterns. Biodiversity and Conservation, 19(9), 2751-2766. 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.

Peters, T., Diertl, K. H., Gawlik, J., Rankl, M., & Richter, M. (2010). Vascular plant diversity in natural and anthropogenic ecosystems in the Andes of Southern Ecuador: Studies from the Rio San Francisco valley. Mountain Research and Development, 30(4), 344-352.

Reddy, S., & Dávalos, L. M. (2003). Geographical sampling bias and its implications for conservation priorities in Africa. Journal of Biogeography, 30(11), 1719-1727.

Santos, D. S., Mansur, K. L., Gonçalves, J. B., Junior, E. R. A., & Manosso, F. C. (2017). Quantitative assessment of geodiversity and urban growth impacts in Armação dos Búzios, Rio de Janeiro, Brazil. Applied Geography, 85, 184-195.

Seijmonsbergen, A. C., Guldenaar, J., & Rijsdijk, K. F. (2018a). Exploring Hawaiian long-term insular geodiversity dynamics. dynamics, 35, 31-43.

(25)

25 Seijmonsbergen, A. C., De Jong, M. G., Hagendoorn, B., Oostermeijer, J. G., Rijsdijk, K. F., Hoorn, C., ... & Antonelli, A. (2018b). Geodiversity Mapping in Alpine Areas. Mountains, Climate and Biodiversity, 155.

Serrano, E., & Ruiz-Flaño, P. (2007). Geodiversity: a theoretical and applied concept. Geographica Helvetica, 62(3), 140-147.

Stern, C. R. (2004). Active Andean volcanism: its geologic and tectonic setting. Revista geológica de Chile, 31(2), 161-206.

Tukiainen, H., Bailey, J. J., Field, R., Kangas, K., & Hjort, J. (2017). Combining geodiversity with climate and topography to account for threatened species richness. Conservation Biology, 31(2), 364-375. Valencia, R., Foster, R. B., Villa, G., Condit, R., Svenning, J. C., Hernández, C., ... & Balslev, H. (2004). Tree species distributions and local habitat variation in the Amazon: large forest plot in eastern Ecuador. Journal of Ecology, 92(2), 214-229.

Van Der Heijden, M. G., Bardgett, R. D., & Van Straalen, N. M. (2008). The unseen majority: soil microbes as drivers of plant diversity and productivity in terrestrial ecosystems. Ecology letters, 11(3), 296-310.

Ward, J. V., Tockner, K., & Schiemer, F. (1999). Biodiversity of floodplain river ecosystems: ecotones and connectivity. Regulated rivers: research & management, 15(1), 125-139.

Wu, J. (2013). Landscape sustainability science: ecosystem services and human well-being in changing landscapes. Landscape ecology, 28(6), 999-1023.

(26)

26

8. Appendices

8.1 Appendix A: Input maps of Ecuador

(27)

27 Figure 9: Input map for the geomorphological diversity in Ecuador

(28)

28 Figure 10: Input map for the pedological diversity in Ecuador.

(29)

29 Figure 11: Input map for the topographical diversity in Ecuador

(30)

30 Figure 12: Input map for geological diversity in Ecuador. No legend was added, due to the large number of unique values.

(31)

31

8.2 Appendix B: Geodiversity maps of the case study area

Figure 13: Geodiversity map of the case study area with a 1000m grid cell size

(32)

32 Figure 15: Geodiversity map of the case study area with a 4000m grid cell size

(33)

33

8.3 Appendix C Thematic diversity maps of the case study area

Figure 16: Geological diversity map of the case study area with a 500m grid cell size

(34)

34 Figure 18: Geological diversity in the case study area with a 2000m grid cell size

(35)

35 Figure 20: Geomorphological diversity in the case study area with a 500m grid cell size

(36)

36 Figure 22: Geomorphological diversity in the case study area with a 2000m grid cell size

(37)

37 Figure 24: Hydrological diversity in the case study area with a 500m grid cell size

(38)

38 Figure 26: Hydrological diversity in the case study area with a 2000m grid cell size

(39)

39 Figure 28: Pedological diversity in the case study area with a 500m grid cell size

(40)

40 Figure 30: Pedological diversity in the case study area with a 2000m grid cell size

(41)

41 Figure 32: Topographical diversity in the case study area with a 500m grid cell size

(42)

42 Figure 34: Topographical diversity in the case study area with a 2000m grid cell size

(43)

43 Figure 36: Ecosystem diversity in the case study area with a 500m grid cell size

(44)

44 Figure 38: Ecosystem diversity in the case study area with a 2000m grid cell size

(45)

45

8.4 Appendix D: Geodiversity maps continental Ecuador

(46)

46 Figure 41: Geodiversity map of continental Ecuador with a 2000m grid cell size

(47)

47 Figure 42: Geodiversity map of continental Ecuador with a 4000m grid cell size

(48)

48

8.5 Appendix E: Thematic maps continental Ecuador

(49)

49 Figure 44: Geological diversity in continental Ecuador with a 1000m grid cell size

(50)

(51)

(52)

52 Figure 47: Geomorphological diversity of continental Ecuador with a 500m grid cell size

(53)

(54)

(55)

(56)

56 Figure 51: Hydrological diversity in continental Ecuador with a 500m grid cell size

(57)

(58)

(59)

(60)

60 Figure 55: Pedological diversity in continental Ecuador with a 500m grid cell size

(61)

(62)

(63)

(64)

64 Figure 59: Topographical diversity in continental Ecuador with a 500m grid cell size

(65)

(66)

(67)

(68)

68 Figure 63: Ecosystem diversity in continental Ecuador with a 500m grid cell size

(69)

(70)

(71)

(72)

72

8.6 Appendix F: Correlation matrices of continental Ecuador

Table 4: Correlation matrix geodiversity of continental Ecuador with a 1000m grid cell size

Pedology 1 Geology 0.62 1 Geomorphology 0.66 0.71 1 Hydrology 0.26 0.33 0.22 1 Topography 0.26 0.20 0.34 0.01 1 Geodiversity 0.77 0.78 0.82 0.40 0.53 1 Vegetation 0.26 0.20 0.23 0.12 0.26 0.24 1

Pedology 1 Geology 0.64 1 Geomorphology 0.68 0.70 1 Hydrology 0.23 0.30 0.19 1 Topography 0.32 0.26 0.43 -0.02 1 Geodiversity 0.78 0.71 0.82 0.36 0.59 1 Vegetation 0.32 0.26 0.31 0.10 0.32 0.29 1

Pedology 1 Geology 0.62 1 Geomorphology 0.63 0.69 1 Hydrology 0.15 0.17 0.13 1 Topography 0.31 0.36 0.48 -0.08 1 Geodiversity 0.73 0.80 0.77 0.32 0.64 1 Vegetation 0.33 0.27 0.40 0.06 0.37 0.34 1

(73)

73

8.7 Appendix G: Correlation matrices of the case study area

Table 7: Correlation matrix geodiversity of the case study area with a grid cell size of 1000m

Pedology 1 Geology 0.65 1 Geomorphology 0.67 0.75 1 Hydrology 0.26 0.29 0.17 1 Topography 0.28 0.22 0.36 -0.02 1 Geodiversity 0.78 0.80 0.82 0.43 0.55 1 Vegetation 0.22 0.14 0.06 0.17 -0.03 0.14 1