1
The General Dynamic Model for the Hawaiian archipelago
Quantifying topographic complexity and questioning to what extend does island age
indeed positively correlate to topographic complexity and biodiversity
A Bachelor Thesis Earth Sciences
University of Amsterdam
Institute for Biodiversity and Ecosystem Dynamic
By:
Linde Berbers
10766618
Admission date:
07-07-2018
Amsterdam
Supervised by:
Dhr. Dr. K.F. (Kenneth) Rijsdijk
2
Abstract
Whittaker’s General Dynamic Model (GDM) on island biogeography is an important
framework for understanding island systems and processes with implications for species distributions. As models such as these are widely used as the basis for nature conservation, it is imperative that they do not make false assumptions. The GDM predicts that when an island is half way through its life cycle, it is topographically the most complex, has the largest elevational range and area, and holds the most biodiversity. However, several aspects of the GDM have not yet been empirically tested,
topographic complexity has not been quantified for the Hawaiian archipelago, nor has a link been made to island age. This research aims to evaluate the GDM for the Hawaiian archipelago, quantifying topographic complexity and questioning to what extend island age indeed positively correlates to topographic complexity and biodiversity. Topographic complexity is calculated in ArcGIS as the standard deviation of slope and absolute maximum elevation. The amount of Single Island Endemics (SIEs) is used as the metric for biodiversity. Furthermore island age and island area are used to test the model. The results show that topographic complexity and biodiversity indeed generally slope upwards in correlation to island age. Area and elevation however only decrease with island age. An anomaly was found in these trends for the middle island of Molokai, showing a minimum amount of SIEs. Possible explanations can be found in past geological events such as glacial configuration and associated climate shifts and surface area changes. This paper concludes that the GDM is too hypothetical to be the direct basis of future nature conservation and ecosystem
management. However, the model could be used as a guideline only if past and present processes which influences the system are thoroughly investigated.
3
Table of contents
Introduction p. 4
- The General Dynamic Model p. 5
- The Hawaiian Archipelago p. 6
Methods and data p. 8
Island analysis results p. 10
Discussion p. 12 - Geologic Factors p. 12 - Climatic Factors p. 15 - Edaphic Factors p. 17 Conclusion p. 18 References p. 20 Appendices p. 27 - Appendix A p. 27 - Appendix B p. 33 - Appendix C p. 36 - Appendix D p. 40
4
Introduction
When Darwin was departing from the Galapagos islands in 1853 he wrote ‘when I see these Islands in sight of each other, & possessed of but a scanty stock of animals, tenanted by these birds, but slightly differing in structure & filling the same place in Nature, I must suspect they are only varieties... if there is the slightest foundation for these remarks the Zoology of Archipelagos will be well worth examination;
for such facts would undermine the stability of species’
(Seen in Jorgensen & Fath, 2014; Wilson & MacArthur, 1967). Island biogeography is a subject of importance on many levels, greater than Darwin might have expected when he wrote this. Islands account for 5.3% of the surface area of the Earth, yet maintain 17% of all flowering plants found on Earth. Furthermore, island species are disproportionately endangered compared to continental species (figure 1). Namely, 37% of island species are endangered and 61% are already extinct (Whittaker et al., 2008; Tershy et al., 2015).
Islands allow us to see processes and factors of ecology, biology and biogeography and to see their specific effects on their surroundings. This can be applied to nature conservation, because by understanding how these processes work, we can manage and even predict the effects that possible future events will have on these systems (Groom et al., 2006). Several scholars have developed theoretical models aiming at describing and interpreting island systems and processes, with possible implications for species distributions (Wilson & MacArthur, 1967; Whittaker & Fernández-Palacios, 2007; Whittaker et al., 2008). The most recent model, called the General Dynamic Model (GDM) proposed by Whittaker et al., (2008) hypothesizes that when an island is half way through its life cycle, it is topographically the most complex, has the largest elevational range and area, and holds the most biodiversity. As these models are widely used as the basis for nature conservation (Ibáñez & Effland, 2011), it is imperative that they do not make false assumptions. Several aspects of the GDM however have not yet been empirically tested, and topographic complexity has not yet been quantified for the Hawaiian archipelago. Nor has a link been made to island age. This is where this research proposes to fill this research gap. This paper aims to evaluate the GDM for the Hawaiian archipelago by quantifying topographic complexity and questioning to what extend island age indeed positively correlates to topographic complexity and biodiversity. The model is tested by firstly defining topographic complexity as surface roughness, using the standard deviation of slope and absolute maximum elevation as metrics, calculated using ArcGIS. Single Island Endemics (SIEs) are used as
Figure 1; showing time series of extinctions of species, from Whittaker & Fernández-Palacios (2007).
5 the metric for biodiversity. Furthermore, absolute maximum elevation, island surface area and finally island age are used to test the GDM.
Firstly this proposal will elaborate on the GDM in order provide a theoretical framework, after which the choice for the Hawaiian archipelago and its general geology is discussed. Furthermore it will elaborate on the methods used for the quantification and evaluation of the model, after which the results of the GDM for the archipelago are put forward. In order to achieve the aim and explain the results of the analysis of the GDM, geography, soil and climate are researched in de appendix as they are the most vital factors influencing geodynamic processes on volcanic islands (Aplet et al., 1998; Otto et al., 2001; Seijmonsbergen et al., 2018). Hereafter in the discussion, the most distinctive observed geologic, climatic and edaphic features of the islands are considered, examining how they might justify the results of the analysis of the GDM and what these results may mean compared to other research.
The General Dynamic Model
Robert H. MacArthur and Edward O. Wilson proposed an Equilibrium Model of Island Biogeography (EMIB) in 1967, a framework used for understanding processes which have impact on insular processes and populations (Wilson & MacArthur, 1967). This model however has not been successful when applied to oceanic islands or processes and systems which operate on geological and evolutionary timescales, and has failed in integrating within island speciation and the life cycle dynamics of islands (Lomolino, 2000; Heany, 2007; Whittaker et al., 2008). Whittaker et al., (2008) saw these flaws in the EMIB model, after which they built upon this model and proposed their own General Dynamic Model (GDM) by combining it with a simple life cycle model of oceanic islands (figure 2). The model implies that when an island is half of its age, it has reached its maximum topographic complexity, elevational range and area, and occupies its maximum amount of biodiversity, in terms of individual species, biomass,
and environment heterogeneity (figure 3).
Figure 2; Idealized relationships between the age (x-axis, time) and area (dotted line), elevational range (dashed line), and topographic complexity (solid line) of a hypothetical oceanic island. Proposed by and taken from Whittaker et al., 2008.
Figure 3; The graphical representation of the key rates and properties of the GDM of oceanic island biogeography showing the postulated relationships between the biological characteristics where, for key rates, I is the immigration rate, S is the speciation rate, and E is the extinction rate; and, for species number, K is the potential carrying capacity, and R is the realized species richness. Proposed by and taken from Whittaker et al., 2008.
6 The Hawaiian archipelago
Remote island archipelagos are also referred to as ‘natural laboratories’ (Whittaker & Fernández-Palacios, 2007; Whittaker et al., 2017). Due to their spatial isolation, relatively small area size and geologically young age, their environments and biotas are relatively simple (Trantis et al.,
2016; Borregaard et al., 2016; Shaw & Gillespie, 2016), making them perfect ‘laboratories’ to study
evolutionary dynamics. As the islands of a typical archipelago such as Hawaii emerge, they are initially influenced by the same factors, such as geology, climate and soil type. Also, they are subject to the same species immigration and colonization from neighboring older islands (Whittaker et al., 2008). And as these islands arise in consecutive years after each other, each going through their own life cycle, archipelagoes can show us millions of years of evolution right now. For other archipelagos such as the Canaries or the Azores, the formation and the definition of island age is subject to more discussion since they are influenced by other and more complex geographic factors (Carracedo, 1994). Therefore, a volcanic archipelago such as Hawaii is the ideal laboratory to study processes of evolution.
The Hawaiian archipelago consists of reefs, shoals and islands and is characterized by the emerging of a chain of shield volcanoes (Landowski, 2015). As the Pacific Plate moves over a stationary hotspot towards the north-west, the hotspot pushes the crust up, exposing the crust above sea level, and volcanoes emerge over the hotspot (Whittaker & Fernández-Palacios, 2007). As the crust moves further along, the first island is removed from the hotspot and another island emerges.
Figure 4; (Redrawn from Price & Clague, 2002, Fig. 1.) Features of the Hawaiian ridge. Ages given in parenthesis in millions of years (Ma). The main islands are detailed in the inset. Outlined in red in the inset are the main islands researched in this paper, with altered ages from Eckstut et al., 2011.
7 Figure 5; (Taken from Price & Clague, 2002, Fig. 3): Island
configurations with 5 million years (Ma) intervals. The islands that are of interest: Ka, Kauai; Oa, Oahu; MN, Maui Nui; Ha, Hawaii.
This produces a linear array of volcanos, each increasing in age towards the north-west (Wilson, 1963; Price & Clague, 2002; figure 4). After reaching their maximum heights, moving further away from the hotspot, the islands subside and are subject to harsh erosion and weathering, flattening the islands. Finally they may sink into the ocean again or remain small atolls (Whittaker & Fernández-Palacios, 2007). The Hawaiian archipelago chain of islands is thus characterized by the emerging of volcanos, each going through a life cycle with well distinguished stages (Price & Clague, 2002). Furthermore, the total number, size and spacing of the Hawaiian islands have varied over time, due to sea level changes such as glacial configuration (Rijsdijk et al., 2014). The current landscape which consists of large islands which are closely
spaced together, preceded a period where the islands were smaller and farther away from each other (figure 5). Currently, the Hawaiian archipelago is a chain of islands stretching 2400km towards the north-west. There are five main islands in the Hawaiian archipelago, which will be the ones this paper will research. These are Kauai (5.3 Ma), Oahu (4 Ma), Molokai (2.1 Ma), Maui (1 Ma) and Hawaii (0.6 Ma) (Eckstut et al., 2011; see also figure 4).
25 Ma 20 Ma Ku PH PH Li La 15 Ma 10 Ma Ma Li La Ma LP Ne 5 Ma Ni Ka LP Ne Ni Ka Oa MN
8
Methods and data
In order to quantify topographic complexity, serval types of data had to be obtained. First however, topographic complexity and biodiversity needed to be well defined. Topographic complexity can be interpreted in various ways. Therefore, topographic complexity will firstly be
divided into two variables; surface roughness and elevation. For the elevation variable the absolute
maximum elevation of each island will be used, providing a clear unbiased number of maximum height (Grohmann et al., 2011). Surface roughness is a meaningful geomorphological variable which is often used in planetary and earth sciences to derive information concerning current or past
processes, material properties, and elapsed time since formation (Olaya, 2009). However, a single definition of surface roughness does not exist. According to Grohmann et al., (2011), in the geomorphometry context, surface roughness is used as expression for the variability of any topographic surface at some given scale. This scale then depends on the size of the geomorphic features or surface of interest. In this research the definition of Grohmann et al., (2011) is used for surface roughness. Next, there are several ways to measure surface roughness, such as slope variability, relative topographic position, standard deviation of slope and standard deviation of elevation. Grohmann et al., (2011) have made an assessment of these measures, and concluded that
‘Standard deviation of slope correctly identified smooth sloping areas and breaks of slope, providing the best results for geomorphological analysis.’ Based on their analysis, surface roughness is defined
here as the standard deviation of slope. A note has to made however on the use of this specific methodology. The standard deviation of slope is only the best way of quantifying surface roughness when the data has a spatial resolution of 10m (Grohmann et al., 2011). Therefore, a spatial resolution of 10m was used.
More specifically, surface roughness (as the standard deviation of slope) has been measured in ArcGIS, a Geographic Information System. This is a system designed for working with geographic information and maps in order to store, capture, manipulate, manage, analyze and present geographic or spatial data (Esri, 2018). The data used for the quantification are Digital Elevation Models (DEMs) of the Hawaiian islands. DEMs are 3D representations of the surface of a terrain, based on the
elevation of the terrain. They provide an objective quantification of surface relief and elevation, and are accordingly ideally suitable for the parameterization of features of surfaces (Grohmann et al., 2011). First the DEMs were loaded into ArcGIS, then through the use of Arc Toolbox, Spatial Analyst Tools, Neighborhood and lastly the Focal Statistics tool, a 5x5 cell neighborhood was created with the standard deviator as operator. Hereafter the mean was taken to retrieve one single representative number per island, for the reason that otherwise the values might be influenced by the presence of a caldera for example. This was done through the Get Raster Properties tool, with mean as Property type. The 10m DEMs have been retrieved from the United States Geological Survey, provided by the
9
Department of Earth and Space Sciences at the University of Washington (Greenberg, 2018). These DEMs have also been used to quantify absolute maximum elevation per island in ArcGIS.
As for biodiversity, the amount of Single Island Endemics (SIEs) will be used as metric. The
use of SIEs as a proxy for biodiversity is widely approved (Price, 2004; Emerson & Kolm, 2005;
Whittaker et al., 2008; Triantis et al., 2008; Otto et al., 2016; Borregaard et al., 2017). Specifically,
Price (2004) made a quantification of the amount of SIEs for the Hawaiian island, which is the data that is used.
Furthermore, geography, soil and climate are the most vital factors influencing geodynamic processes on volcanic islands (Aplet et al., 1998; Otto et al., 2001; Seijmonsbergen et al., 2018). Therefore, the study of these topics is indispensable in order to explain the results of the GDM analysis. These topics have been thoroughly researched through an extensive literature study and can be found in the appendix for the reason that they are secondary results which do not directly answer the research question.
Moreover, the specific age of each island and subsequently the age of each volcano is a complex concept. Following Triantis & Whittaker (2016), when performing an analysis which considers more than one oceanic island in an archipelago, the age of the individual islands should be
treated with considerable caution. After an extensive literature study, it seems that close to all research
concerning the Hawaiian archipelago use the island and volcano ages defined by Clague and Dalrymple (1987) (Moore & Clague, 1992; Ziegler, 2002; Eckstut et al., 2011; Landowski, 2015). These authors then refer to various other authors per island. However, Clague and Dalrymple (1987) provide the overview of the most widely used volcano ages, which therefore will also be used in this paper.
In sum, DEMs have been used to quantify the standard deviation of slope and absolute maximum elevation in ArcGIS. Data from a biodiversity study in the form of SIEs on the Hawaiian archipelago will be used as the proxy for biodiversity. Geography, soil and climate influencing the Hawaiian islands are researched in order to illuminate the results of the GDM analysis. Island and volcano ages are based on Clague and Dalrymple (1987).
10 Figure 6; the General Dynamic Model proposed by Whittaker et al., 2008. Edited here by adding the biodiversity trend from figure 3 for the purpose of a more comprehensive overview. 0 50 100 150 200 250
Hawaii Maui Molokai Oahu Kauai
SD_slope x10/ Max_elevation x100m/ Species/ Ma/ Area x100km2
The GDM for Hawaii
SD_slope Max_elevation SIE Age Area
Results
In the following section the result of the GDM analysis for the main islands is presented. The results from the literature study can be found in the appendix, as these do not directly answer the research questions yet are needed in order to explain the results. This includes a detailed description of the geography for each island and volcano (appendix A), the different climate zones (appendix B) and soil orders per island (appendix C), put into context with the geography, climate and elevation and an overview table of the main findings (appendix D).
The General Dynamic Model for the Hawaiian archipelago To restate, the GDM hypothesizes a parabolic
relationship between island age, topographic complexity, area and biodiversity for a volcanic island (figure 6). Expected was that the standard deviation of slope, absolute maximum elevation, the amount of SIEs and surface area, increase coinciding with island age. The result of the GDM analysis for the archipelago can be seen in figure 7 and the corresponding table 1 for the specific data. Starting with the observations which coincide to the hypothesized GDM, what can be observed from the result is first of all a general upward sloping trend in relation
to islands age for the standard deviation of slope and the amount of SIEs per island. These
observations are labeled as trends for the reason that not each island exhibits increasing data for these variables, however from the youngest to the oldest islands, the data generally increases (figure 8). In contrast to the hypothesized GDM, absolute maximum elevation only decreases with increasing island age and island area exhibits a general decreasing trend regarding island age (figure 8). A specific notable observation from the analysis are the results seen from the island of Molokai. Although the SIEs per island follow a general upward trend, the amount of SIEs reach a minimum on the island Molokai. Correspondingly, although the general trend for the surface area is downward sloping, Molokai is the smallest island.
Figure 7; The General Dynamic Model for the Hawaiian archipelago. Showed is the standard deviation of slope (SD_slope) times 10, absolute maximum elevation (Max_elevation) per 100m, Single Island Endemics (SIE) in amount of endemic species per island, age in Ma (million years ago), area per 100km2.
11 Features
Island
SD_slope Max_elevation
in m
SIEs Age (Ma) Area in km2
Hawaii 0.97703 4200.45 82 0.6 10432 Maui 2.47722 3053 81 1 727 Molokai 2.52876 1508 34 2.1 260 Oahu 2.71554 1233 128 4 597 Kauai 3.30294 1599 198 5.3 552 0 50 100 150 200 250
Hawaii Maui Molokai Oahu Kauai
SD_slope x10/ Species/ Max_elevation x100m/ Area x100km2
Trend lines
SD_slope SIE Max_elevation Area
Table 1; The island features for the GDM analysis of the Hawaiian archipelago. Standard deviation of slope (SD_slope) calculated using ArcGIS; absolute maximum elevation (Max_elevation) in meters calculated using ArcGIS; SIE as Amount of Single Island Endemics according to Price (2004); Age as island age in million years ago (Ma) according to Eckstut et al., (2011); Island area in km2 according to Eckstut et al., (2011).
Figure 8; The General Dynamic Model for the Hawaiian archipelago. Showed are the trend lines for the standard deviation of slope (SD_slope) times 10, Single Island Endemics (SIE) in amount of endemic species per island, absolute maximum elevation (Max_elevation) per 100m, area per 100km2.
12
Discussion
As this research aims to evaluate the GDM in order to strengthen it for further ecosystem management and nature conservation, the most notable observation from the results is that Molokai has the lowest amount of SIEs. The fact that Molokai is also the smallest island, might be a
contributing factor to the exceptional low amount of SIEs found on Molokai. Important to realize is that all native organisms which occur naturally in a particular habitat are the products of their
physical, biological and geological environments and interactions. The relative role however in which these factors contribute to the patterns we observe in species diversity and richness remain divided (Harrison et al., 2006). Historical justifications for patterns in species richness find their merits in the way an island emerged, (the type of) volcanic activity, influences of glaciation periods, geological hazards or island and species isolation (Wilson & MacArthur, 1967; Qian & Ricklefs, 2000; Stephens & Wiens, 2003; Hawkins et al., 2003b; Ricklefs, 2004). More modern explanations find their answers more in the physical environment, such as productivity and climate (Huston & DeAngelis, 1994; Hawkins et al., 2003a; Francis & Currie, 2003). The challenge lies not in including each element, but in evaluating certain factors and in recognizing that each factor influences the other. Naturally, countless factors could be observed; and according to the first law of ecology; ‘in nature, everything is connected to everything else’ (Kruckeberg, 2002). For the scope of the research however, here only the most distinctive features observed from the research are discussed which may help to explain the low amount of SIEs seen on Molokai. Discussed are first the geologic factors, then the climatic factors and finally the edaphic factors. The research on which the discussion is based can be found in the appendix.
Geologic factors
The chemical compositions of the specific lava flows are not taken into consideration as they
are not decisive for differences in biodiversity between the islands. First of all, the type of explosive
material and the chemical composition of lava flows do not seem to explain variations in biodiversity (Fernández-Palacios & de Nicolas, 1995; Otto et al., 2001). Secondly, as the Pacific Plate moves over a stationary hotspot with a stable mantle plume, the chemical compositions between the different lavas of each island do not differ significantly from each other (Seijmonsbergen et al., 2018).
As was clear from the results of the GDM analysis, the island of Molokai deviates from the overall trend and does not follow the expected theory (Fernández-Palacios, 2007; Whittaker et al., 2008). Likewise, Ibáñez & Effland (2011) state that Molokai has relief and an area which is not in line with wat is expected theoretically for archipelagos. From the research it became clear that Molokai used to be part of the much larger Maui Nui complex due to glacial configuration (Rijdsijk et al., 2014), only now due to the current see level is it the smallest island (figure 11). In fact, in a study on
13 Figure 10; General Dynamic Model proposed by Whittaker et al., 2008. Edited here by adding the biodiversity trend from figure 3 for the purpose a more comprehensive overview. Also added is the hypothetical point where the Hawaiian archipelago might currently be in.
the biogeographical implications of the Pleistocene sea level cycle, Rijsdijk et al., (2014) showed that former connected islands which are now isolated due to current sea levels, could explain endemism patterns on volcanic islands. Therefore, it has to be considered that Molokai should not be
distinguished as a separate island, but as the Maui Nui complex. In figure 9 the resulting GDM analysis can be seen, where Maui and Molokai are considered together as the Maui Nui complex. In figure 10, the hypothesized GDM is depicted. What can be noticed from the comparison of both figures is that the model for the archipelago is more similar to the hypothesized model than before considering Maui Nui. In order to illustrate, added to the hypothesized model is the here theorized position on the GDM in which the archipelago might currently be in. Namely, for the archipelago considering Maui Nui, altitude exhibits a downward slope, which can also be observed on the intersect with the hypothesized model. Next, the topographical complexity (as SD_slope) is upward sloping, as it is on the hypothesized model.
Furthermore, total area is decreasing with age for the archipelago, similar to the trending slope seen on the hypothesized slope for area. Lastly,
biodiversity (as SIEs) in general exhibits an upward slope, the same as the hypothetical biodiversity trend. However, considering Maui Nui does not entirely explain the irregularity for the amount of SIEs on Molokai as they are still lowest on Maui Nui. Therefore more factors should be taken into consideration in order to explain the exhibited anomaly for the amount of SIEs on Molokai.
0 50 100 150 200 250
Hawaii Maui Nui Oahu Kauai
SD_slope x10/ Max_elevation x100m/ Species/ Area x100km2
The GDM for Hawaii with Maui Nui
SD_slope Max_elevation SIE Age Area
Figure 9; The General Dynamic Model for Hawaiian archipelago considering Molokai and Maui together as the Maui Nui complex. Showed is the standard deviation of slope (SD_slope), times 10; absolute maximum elevation (Max_elevation) per 100m, Single Island Endemics (SIE) in amount of endemic species per island, age in Ma (million years ago), area per 100km2. Variables for Mui Nui are calculated by using the average of Molokai and Maui for SD_slope, Max_elevation, SIEs and age. Area is based on Price& Elliot-Fisk, 2004.
14 Another distinctive feature from the geologic research which sets Molokai apart from the other islands is that it is the smallest island. It is widely accepted and investigated that endemism increases with increasing area (Anderson, 1994; Losos & Schluter, 2000; Lovette et al., 2002; Roos et al., 2004; Losos & Ricklefs, 2009). According to Losos & Ricklefs (2009), there are two explanations which might account for the correlation between speciation and island area. Firstly, a larger island could give more opportunities for geographic speciation due to their usual higher ecological and topographical complexity and also fragmentation opportunities due to geological events or high sea levels (Gifford & Larson, 2008). Secondly, ecological diversity is frequently associated with a large area, meaning that bigger islands have extend niche spaces and can allow habitation of more species (Lovette et al., 2002). Similarly, Anderson (1994) poses that the number of endemic species in an area will increase together with the surface area. All of the above argues that due to the relatively small size of Molokai, SIEs had less opportunities to develop.
On the whole, from the research it became clear that Molokai used to be part of the much larger Maui Nui complex. Therefore, the GDM with Maui Nui was considered. This proved to be a better fit for the hypothesized GDM, but did not attest for everything. Therefore, more factors where considered, including the distinctive feature of the small area of Molokai. As research suggests that endemism increases with area, this could be an element explaining the minimum amount of SIEs seen on Molokai. More research should be performed however on how to quantify the variables used in the GDM for Maui Nui, as area and the amount of SIEs on the complex is not yet thoroughly researched. Above all, it seems that when taking into consideration past geological events such as glacial
configuration which alters the current structure of the study area, the analyzed GDM is fairly consistent to the hypothesized GDM.
15 Figure 11; From Price & Elliott-Fisk (2004); Summary of Maui Nui history. For each 0.2-myr interval, the approximate range of variation in the area and configuration of islands around that time is summarized. Added in red the names of the mentioned islands.
Climatic factors
As was researched, climate varies significantly on each Hawaiian island due to the sharp altitude gradients and trade winds. What can be seen from the table A2 is that Molokai receives the lowest amount of annual precipitation compared to the other islands. Various research links the amount of annual precipitation to biodiversity, which might imply that precipitation is correlated to the presence of the anomaly in the amount of SIEs on Molokai. For example, Aplet et al., (1998) in a research on biomass and species composition in Hawaii, found that the wet sides of volcanoes supported the most native species. In addition, they concluded a strong correlation between total species richness and precipitation. Also Otto et al., (2001) on a research in Tenerife concluded that mean annual precipitation is the factor which best explained variation in the structure, distribution, composition and richness of plant species. Equally, Otto et al., (2006) proposed that vegetation and climate favour species endemism. Fernández-Palacios and Nicolás (1995) confirmed that wind exposure and altitude, and as a consequence precipitation, are the greater environmental causes of vegetation variation. However, Swaine (1996) concluded that a highly fertile soil is correlated to areas of lower rainfall.
Another key point when discussing the climate of Molokai is that it used to be part of the much larger Mui Nui complex. Maui Nui had a flatter surface than the island of Hawaii currently does, and therefore probably had a warmer temperature. Also, because the isthmuses connecting the
volcanoes of Maui Nui were lower in elevation, precipitation caused by the trade winds may have been more discontinuous than it is now. In addition, due to its large size, it is probable that localized orographic precipitation could have developed. The trade winds would have been obstructed by the mountains, allowing the land to heat up, drawing moist air upwards, and creating precipitation (Price & Elliott-Fisk, 2004). Moreover, Hotchkiss (1998) and Hotchkiss & Juvik (1999) show that in glacial periods, the complex cooled with almost 5°C due to a decrease in the trade wind intensity. After this, the complex was divided into multiple islands (figure 11), changing the local climate again. Therefore, Molokai has experienced more climate shifts than the other islands. With this in mind, various research points out that endemism is negatively correlated to climatic shifts (Fjeldsa, 1995; Cronk, 1997; Fjeldsa & Lovett, 1997; Jansson, 2003; Fordham & Brook, 2010). Jansson (2003) argues that the less extreme the climatic shifts are, the higher the probability of endemics surviving. In particular, the paper concludes that the higher the temperature changes are
16 in an area since the last glacial period, the less endemics it accommodates. Fordham & Brook (2010) share the same conclusions, and argue that a prerequisite for island endemics is a long term stable environment by promoting adaptive radiation. These arguments suggest that the climatic changes Molokai experienced could be a factor the low amount of SIEs. However, it is also argued that the larger an area, the more seasonal changes there are (Stevens, 1989). And since Molokai is the smallest island, this argument would favor endemism on the island. Another point to be made is that Maui also used to be part of the Maui Nui complex, but does not exhibit any irregular trends.
On the whole, although precipitation and altitude are not clearly distinct from one another for which more research should be carried out to reduce uncertainties, it seems that it is highly probable that precipitation positively influences species diversity, vegetation diversity and endemism. Since Molokai receives the least precipitation, this might be a factor contributing to the low amount of SIEs found on Molokai. Furthermore, research shows that climatic shifts negatively impact endemism. As Molokai has experienced multiple climatic shifts, being part of a much larger island complex, this could also be a factor contributing to the minimum amount of SIEs on Molokai. By way of contrast however, the island of Maui was also part of this complex, and does not show any irregularities in the result.
17 Edaphic factors
What can be noticed about the variation in types of soils on the islands is that the amount of area coverage of most abundant soils and the second most abundant soils generally lie close together (table A1). However, on Molokai, the second most abundant soil order represents only half of the area that the most abundant soil does. What is significant about this fact is that this dominating soil order is the Oxisol, an extremely infertile soil. The second most abundant soil, the Inceptisol, is a relatively infertile soil as well. This might imply that the types of soil present on the islands best explains the anomaly in the amount of SIEs. This is in agreement to Anacker (2011), who pleads that endemism to a certain type of soil is an important and a specifically common type of habitat specialization.
Similarly, Kruckeberg (2002) points out that ‘endemism is the hallmark of specialized edaphic habitats’, and describes that plants are bound to their inanimate surrounding environment, more than animals. According to Kruckeberg (2002), all relatively complex plants adapted to live on land are linked to a certain kind of supporting structure; other plants, water availability, rock or soil. However, these supporting structures are, in turn, the products of materials and processes of biological and physical origin. A large part of the characteristics and origins of specific habitats are found in the geology. Therefore, the formation of a specific landform is formed or lead by geological processes, meaning that soil and species endemism cannot be directly linked, because soils are the products of the underlying historic geology. For example, in a study on species richness in insular environments by Harrison et al., (2006), it was found that 73% variation in regional endemic richness was based on historical factors, such as a characteristic geomorphology and geologic age and the regional
environment such as soil productivity and climate. While 66% of the variation was accounted for by the local environment, such as rocks, vegetation and soils. Nevertheless, since Kauai is also
dominated by the Oxisol, with a second most abundant soil only occupying a third of the most abundant soil, some may argue that soil type does not influence endemism. The difference however is that the second most abundant soil on Kauai is the fertile Mollisol, instead of the relatively infertile Inceptisol.
On the whole, soil type may explain the irregularity in the SIE trend seen in the analysis of the GDM. However, separating soil type from other factors such as underpinning geological and physical processes is not possible in the scope of this research. Therefore, more specific research should be conducted in order to distinguish the influences of these factors on each other and reduce uncertainties.
18
Conclusion
To summarize, this research has evaluated the GDM for the main islands of the Hawaiian archipelago, quantifying topographic complexity and questioning to what extend island age indeed positively correlates to topographic complexity and biodiversity. Aiming to strengthen the GDM for future nature conservation and ecosystem management strategies, since models such as these are widely used as the basis for nature management. In ArcGIS, the standard deviations of slope and absolute maximum elevations were calculated per island and used as proxy for topographic
complexity. The amount of SIEs per island was used as proxy for biodiversity and absolute maximum elevation, island surface area and finally island age were also used to test the GDM. In line with the hypothesized model, the results showed a general upward trend for the standard deviation of slope and the amount of SIEs. Not in line was a general downward trend observed for the absolute maximum elevation and area. However, an anomaly in these trends was found on the island of Molokai, showing a minimum amount of SIEs and area. In order to explain the results a literature study was performed on the geologic, climatic and edaphic processes influencing the islands. From the geologic research it became clear the Molokai used to be part of the much larger Maui Nui complex. This proved to be a better fit for the hypothesized GDM, yet did not attest for the entire anomaly. Furthermore, the distinctive small surface area of Molokai was discussed as a possible reason for the inverted trend on the SIEs on Molokai, as research suggests that endemism increases with area. From the climate study it could be concluded that although precipitation and altitude are not clearly distinct from one another, it seems highly probable that precipitation positively influences species diversity, vegetation diversity and endemism. Since Molokai receives the fewest amount of precipitation, this might explain that Molokai has the lowest amount of SIEs. Secondly, as Molokai has experienced multiple climatic shifts by being part of the much larger Maui Nui complex, this could also be a factor contributing to the low amount of SIEs seen on Molokai as research points to a negative correlation between endemism and climatic shifts. The discussion on the edaphic factors proposes that on the whole, soil type could explain the minimal amount of SIEs on Molokai. However, separating soil type from other factors such as underpinning geological and physical processes is not possible in the scope of this research.
In conclusion, this paper proposes that the GDM is too hypothetical to be the direct basis of future nature conservation and ecosystem management. However, the model could be used as a guideline, only if the formation of the island system is thoroughly investigated, taking into
consideration past geological events such as glacial configuration which alters the current structure of the study area. Furthermore, it should be noted that topographic complexity and biodiversity could be interpreted in many ways, which might change the final analysis results in other research if these are defined differently. In addition, geographic, climatic and edaphic processes influence each other.
19 Therefore more specific research should be carried out, in order to determine the particular influences of each of these processes. Moreover, as this research only took into consideration some specific factors which might be of influence on the islands, more research should be performed to elucidate other influencing processes, ultimately in order to further strengthen the GDM for future ecosystem management and nature conservation.
20
References
Anacker, B. L. (2011). Phylogenetic patterns of endemism and diversity. Serpentine: The Evolution and Ecology of a Model System.University of California Press, Berkeley, , 49-70.
Anderson, S. (1994). Area and endemism. The Quarterly Review of Biology, 69(4), 451-471.
Bergmanis, E. C., Sinton, J. M., & Trusdell, F. A. (2000). Rejuvenated volcanism along the southwest rift zone, east maui, hawai ‘i. Bulletin of Volcanology, 62(4-5), 239-255.
Borregaard, M. K., Amorim, I. R., Borges, P. A., Cabral, J. S., Fernández‐Palacios, J. M., Field, R., et al. (2017). Oceanic island biogeography through the lens of the general dynamic model:
Assessment and prospect. Biological Reviews, 92(2), 830-853.
Borregaard, M. K., Matthews, T. J., & Whittaker, R. J. (2016). The general dynamic model: Towards a unified theory of island biogeography? Global Ecology and Biogeography, 25(7), 805-816.
Buol, S., Southard, R., Graham, R., & McDaniel, P. (2011). US soil taxonomy. Soil Genesis and Classification, Sixth Edition, , 207-232.
Carracedo, J. (1994). The canary islands: An example of structural control on the growth of large oceanic-island volcanoes. Journal of Volcanology and Geothermal Research, 60(3-4), 225-241.
Clague, D. A., & Dalrymple, G. B. (1987). The hawaiian-emperor volcanic chain. part I. geologic evolution. Volcanism in Hawaii, 1, 5-54.
Clague, D. A., & Moore, J. G. (2002). The proximal part of the giant submarine wailau landslide, molokai, hawaii. Journal of Volcanology and Geothermal Research, 113(1), 259-287.
Cronk, Q. (1997). Islands: Stability, diversity, conservation. Biodiversity & Conservation, 6(3), 477-493.
Deenik, J., & McClellan, A. (2007). Soils of hawai'i Cooperative Extension Service, University of Hawai'i at Mānoa.
Eckstut, M. E., McMahan, C. D., Crother, B. I., Ancheta, J. M., McLennan, D. A., & Brooks, D. R. (2011). PACT in practice: Comparative historical biogeographic patterns and species–area relationships of the greater antillean and hawaiian island terrestrial biotas. Global Ecology and Biogeography, 20(4), 545-557.
21 Emerson, B. C., & Kolm, N. (2005). Species diversity can drive speciation. Nature, 434(7036), 1015.
ESRI. 2018. ArcGIS Desktop; https://www.esri.nl/nl-nl/producten/arcgis-for-personal-use. Visited on 4/4/2018.
Faichney, I. D., Webster, J. M., Clague, D. A., Kelley, C., Appelgate, B., & Moore, J. G. (2009). The morphology and distribution of submerged reefs in the maui-nui complex, hawaii: New insights into their evolution since the early pleistocene. Marine Geology, 265(3-4), 130-145.
Fernández‐Palacios, J. M., & Nicolás, J. P. (1995). Altitudinal pattern of vegetation variation on tenerife. Journal of Vegetation Science, 6(2), 183-190.
Fjeldsa, J. (1995). Geographical patterns of neoendemic and older relict species of andean forest birds: The significance of ecologically stable areas. New York, USA.
Fjeldsa, J., & Lovett, J. C. (1997). Geographical patterns of old and young species in african forest biota: The significance of specific montane areas as evolutionary centres. Biodiversity & Conservation, 6(3), 325-346.
Fordham, D. A., & Brook, B. W. (2010). Why tropical island endemics are acutely susceptible to global change. Biodiversity and Conservation, 19(2), 329-342.
Francis, A. P., & Currie, D. J. (2003). A globally consistent richness-climate relationship for angiosperms. The American Naturalist, 161(4), 523-536.
Garza, J. A., Chu, P., Norton, C. W., & Schroeder, T. A. (2012). Changes of the prevailing trade winds over the islands of hawaii and the north pacific. Journal of Geophysical Research: Atmospheres, 117(D11).
Giambelluca, T. W., Chen, Q., Frazier, A. G., Price, J. P., Chen, Y., Chu, P., et al. (2013). Online rainfall atlas of hawai ‘i. Bulletin of the American Meteorological Society, 94(3), 313-316.
Gifford, M. E., & Larson, A. (2008). In situ genetic differentiation in a hispaniolan lizard (ameiva chrysolaema): A multilocus perspective. Molecular Phylogenetics and Evolution, 49(1), 277-291.
Greenberg, H. (2018). 10-meter digital elevation model of Hawaii. University of Washington, 2018;
22 Grohmann, C. H., Smith, M. J., & Riccomini, C. (2011). Multiscale analysis of topographic surface
roughness in the midland valley, scotland. IEEE Transactions on Geoscience and Remote Sensing, 49(4), 1200-1213.
Groom, M. J., Meffe, G. K., & Carroll, C. R. (2006). Principles of conservation biology Sinauer Associates Sunderland.
Harrison, S., Safford, H. D., Grace, J. B., Viers, J. H., & Davies, K. F. (2006). Regional and local species richness in an insular environment: Serpentine plants in california. Ecological Monographs, 76(1), 41-56.
Hawkins, B. A., Field, R., Cornell, H. V., Currie, D. J., Guégan, J., Kaufman, D. M., et al. (2003). Energy, water, and broad‐scale geographic patterns of species richness. Ecology, 84(12), 3105-3117.
Hawkins, B. A., Porter, E. E., Diniz-Filho, F., & Alexandre, J. (2003). Productivity and history as predictors of the latitudinal diversity gradient of terrestrial birds. Ecology, 84(6), 1608-1623.
Heaney, L. R. (2007). Is a new paradigm emerging for oceanic island biogeography? Journal of Biogeography, 34(5), 753-757.
Hotchkiss, S. C. (1998). Quaternary vegetation and climate of hawai'i University of Minnesota.
Hotchkiss, S., & Juvik, J. O. (1999). A late-quaternary pollen record from ka ‘au crater, O ‘ahu, hawai ‘i. Quaternary Research, 52(1), 115-128.
Huston, M. A., & DeAngelis, D. L. (1994). Competition and coexistence: The effects of resource transport and supply rates. The American Naturalist, 144(6), 954-977.
Ibáñez, J., & Effland, W. R. (2011). Toward a theory of island pedogeography: Testing the driving forces for pedological assemblages in archipelagos of different origins. Geomorphology, 135(3-4), 215-223.
Izuka, S. K., & Resig, J. M. (2008). Evidence of late Pliocene–early pleistocene marine environments in the deep subsurface of the lihue basin, kauai, hawaii. Palaios, 23(7), 442-451.
Jansson, R. (2003). Global patterns in endemism explained by past climatic change. Proceedings.Biological Sciences, 270(1515), 583-590.
23 Jefferson, A. J., Ferrier, K. L., Perron, J. T., & Ramalho, R. (2014). Controls on the hydrological and
topographic evolution of shield volcanoes and volcanic ocean islands. The Galapagos: A Natural Laboratory for the Earth Sciences, 204, 185.
Jorgensen, S. E., & Fath, B. (2014). Encyclopedia of ecology Newnes.
Kauahikaua, J., Cashman, K., Clague, D., Champion, D., & Hagstrum, J. (2002). Emplacement of the most recent lava flows on hualālai volcano, hawai ‘i. Bulletin of Volcanology, 64(3-4), 229-253.
Kottek, M., Grieser, J., Beck, C., Rudolf, B., & Rubel, F. (2006). World map of the köppen-geiger climate classification updated. Meteorologische Zeitschrift, 15(3), 259-263.
Kruckeberg, A. R. (2004). Geology and plant life: The effects of landforms and rock types on plants University of Washington Press.
Landowski, C. M. (2015). Characteristics of hawaiian volcanoes Government Printing Office.
Lomolino, M. (2000). A call for a new paradigm of island biogeography. Global Ecology and Biogeography, 9(1), 1-6.
Losos, J. B., & Ricklefs, R. E. (2009). Adaptation and diversification on islands. Nature, 457(7231), 830.
Losos, J. B., & Schluter, D. (2000). Analysis of an evolutionary species–area relationship. Nature, 408(6814), 847.
Lovette, I. J., Bermingham, E., & Ricklefs, R. E. (2002). Clade-specific morphological diversification and adaptive radiation in hawaiian songbirds. Proceedings.Biological Sciences, 269(1486), 37-42.
Macdonald, G. A., Abbott, A. T., & Peterson, F. L. (1983). Volcanoes in the sea: The geology of hawaii University of Hawaii Press.
Mark, R. K., & Moore, J. G. (1987). Slopes of the hawaiian ridge. US Geol.Surv.Prof.Pap, 1350, 101-107.
McMurtry, G. M., Campbell, J. F., Fryer, G. J., & Fietzke, J. (2010). Uplift of oahu, hawaii, during the past 500 ky as recorded by elevated reef deposits. Geology, 38(1), 27-30.
24 Moore, J. G., Clague, D., Holcomb, R., Lipman, P., Normark, W., & Torresan, M. (1989). Prodigious
submarine landslides on the hawaiian ridge. Journal of Geophysical Research: Solid Earth, 94(B12), 17465-17484.
Nelson, S. T., Tingey, D. G., & Selck, B. (2013). The denudation of ocean islands by ground and surface waters: The effects of climate, soil thickness, and water contact times on oahu, hawaii. Geochimica Et Cosmochimica Acta, 103, 276-294.
Olaya, V. (2009). Basic land-surface parameters. Developments in Soil Science, 33, 141-169.
Otto, R., Krüsi, B., Burga, C., & Fernandez-Palacios, J. (2006). Old-field succession along a precipitation gradient in the semi-arid coastal region of tenerife. Journal of Arid Environments, 65(1), 156-178.
Otto, R., Fernández‐Palacios, J. M., & Krüsi, B. O. (2001). Variation in species composition and vegetation structure of succulent scrub on tenerife in relation to environmental variation. Journal of Vegetation Science, 12(2), 237-248.
Otto, R., Whittaker, R. J., Gaisberg, M., Stierstorfer, C., Naranjo‐Cigala, A., Steinbauer, M. J., et al. (2016). Transferring and implementing the general dynamic model of oceanic island
biogeography at the scale of island fragments: The roles of geological age and topography in plant diversification in the canaries. Journal of Biogeography, 43(5), 911-922.
Peel, M. C., Finlayson, B. L., & McMahon, T. A. (2007). Updated world map of the köppen-geiger climate classification. Hydrology and Earth System Sciences Discussions, 4(2), 439-473.
Price, J. P. (2004). Floristic biogeography of the hawaiian islands: Influences of area, environment and paleogeography. Journal of Biogeography, 31(3), 487-500.
Price, J. P., & Elliott-Fisk, D. (2004). Topographic history of the maui nui complex, hawai'i, and its implications for biogeography. Pacific Science, 58(1), 27-45.
Price, J. P., & Clague, D. A. (2002). How old is the hawaiian biota? geology and phylogeny suggest recent divergence. Proceedings.Biological Sciences, 269(1508), 2429-2435.
Reardon, S. (2018). Hawaii volcano eruption holds clues to predicting similar events elsewhere. Nature, 557(7706), 477-018-05206-w.
25 Ricklefs, R. E. (2004). A comprehensive framework for global patterns in biodiversity. Ecology
Letters, 7(1), 1-15.
Rijsdijk, K. F., Hengl, T., Norder, S. J., Otto, R., Emerson, B. C., Ávila, S. P., et al. (2014). Quantifying surface‐area changes of volcanic islands driven by pleistocene sea‐level cycles: Biogeographical implications for the macaronesian archipelagos. Journal of
Biogeography, 41(7), 1242-1254.
Roos, M. C., Keßler, P. J., Robbert Gradstein, S., & Baas, P. (2004). Species diversity and endemism of five major malesian islands: Diversity–area relationships. Journal of Biogeography, 31(12), 1893-1908.
Sanderson, M. (1993). Prevailing trade winds: Weather and climate in hawaii. University of Hawai’i Press, Honolulu.
Seijmonsbergen, A.C. (2018). Exploring Hawaiian long-term insular geodiversity dynamics. Institute for Biodiversity and Ecosystem dynamics. Universiteit van Amsterdam, the Netherlands.
Shaw, K. L., & Gillespie, R. G. (2016). Comparative phylogeography of oceanic archipelagos: Hotspots for inferences of evolutionary process. Proceedings of the National Academy of Sciences of the United States of America, 113(29), 7986-7993.
Sinton, J. M., Eason, D. E., & Duncan, R. A. (2017). Volcanic evolution of moloka ‘i, hawai ‘i: Implications for the shield to postshield transition in hawaiian volcanoes. Journal of Volcanology and Geothermal Research, 340, 30-51.
Soil Survey Staff. Official Soil Series Descriptions. USDA-
NRCS. http://soils.usda.gov/technical/classification/osd/index.html. Visited on 22/05/2018.
Stephens, P. R., & Wiens, J. J. (2003). Explaining species richness from continents to communities: The time-for-speciation effect in emydid turtles. The American Naturalist, 161(1), 112-128.
Stevens, G. C. (1989). The latitudinal gradient in geographical range: How so many species coexist in the tropics. The American Naturalist, 133(2), 240-256.
Swaine, M. (1996). Rainfall and soil fertility as factors limiting forest species distributions in ghana. Journal of Ecology, 419-428.
26 Tershy, B. R., Shen, K., Newton, K. M., Holmes, N. D., & Croll, D. A. (2015). The importance of
islands for the protection of biological and linguistic diversity. Bioscience, 65(6), 592-597.
Triantis, K. A., Mylonas, M., & Whittaker, R. J. (2008). Evolutionary species–area curves as revealed by single‐island endemics: Insights for the inter‐provincial species–area
relationship. Ecography, 31(3), 401-407.
Triantis, K., Whittaker, R. J., Fernández-Palacios, J. M., & Geist, D. J. (2016). Oceanic archipelagos: A perspective on the geodynamics and biogeography of the World’s. Issue: Frontiers of
Biogeography, 8 (2).
Whittaker, R. J., & Fernández-Palacios, J. M. (2007). Island biogeography: Ecology, evolution, and conservation. Oxford University Press.
Whittaker, R. J., Triantis, K. A., & Ladle, R. J. (2008). A general dynamic theory of oceanic island biogeography. Journal of Biogeography, 35(6), 977-994.
Whittaker, R. J., Fernandez-Palacios, J. M., Matthews, T. J., Borregaard, M. K., & Triantis, K. A. (2017). Island biogeography: Taking the long view of nature's laboratories. Science (New York, N.Y.), 357(6354), 10.1126/science.aam8326.
Wilson, E. O., & MacArthur, R. H. (1967). The theory of island biogeography. Princeton, NJ.
Wilson, J. T. (1963). A possible origin of the hawaiian islands. Canadian Journal of Physics, 41(6), 863-870.
Wolfe, E. W., & Morris, J. (1996). Geologic map of the island of Hawaii.
Xu, G., Frey, F. A., Clague, D. A., Abouchami, W., Blichert‐Toft, J., Cousens, B., et al. (2007). Geochemical characteristics of west molokai shield‐and postshield‐stage lavas: Constraints on hawaiian plume models. Geochemistry, Geophysics, Geosystems, 8(8).
27
Appendix A ~ The geology of the Hawaiian archipelago
Life stages of volcanoes
Idealized volcanic islands go through four explosive stages in their life cycles (Macdonald et al., 1983; Landowski, 2015). The stages are distinguished by the style and rate of the eruption, heat production towards the outside of the crust of the earth and lava composition (Wolfe & Morris, 1996). The first stage, called the pre-shield stage, happens entirely under water. No explosions take place due to the water pressure of the ocean water pushing on the volcano. Lava pours out underneath the sea surface as thick pillows, until the shield becomes thick enough that it almost reaches the surface, where the water pressure is no longer strong enough to prevent explosions. The subsequent shield stage is where the volcano produces almost all of its mass. It is marked by the repeated collapsing and filling of calderas. As the volcano moves farther away from the hotspot it reaches the post-shield stage, where the magma chamber cools and becomes less shallow, and the eruptions are less explosive. Here the top of the volcano is buried underneath pyroclastic material and lavas. The volcano slowly dies out. However, it is possible that after a long stage of volcanic inactivity a volcano may become active again. This is called the rejuvenated stage. Here the explosions are powerful, the result of increasing gas pressure within the lava (Macdonald et al., 1983, Moore & Clague, 1992 Landowski, 2015). Although there are five volcanoes in the Hawaiian archipelago which most probably have gone through each of these four stages, not one volcano shows us products of each of the stages. This is for the reason that several volcanoes have yet to complete their life cycle, and also because the products of some of the earlier stages are buried between products of later stages
(Landowski, 2015).
The island of Hawaii
The island of Hawaii is the youngest and also the biggest of the main islands. The island is made up of five separate volcanoes; Kohala, Mauna Kea, Hualalai, Mauna Loa and Kilauea (figure A1). Each volcano submerged from beneath the sea surface to form one single island (Wolfe & Morris, 1996). The center of volcanism has shifted increasingly towards the south-east, reflecting the movement of the Pacific Plate towards the north-west. Hence, Kohala, the volcano located most towards the north-west, is the oldest volcano in the island and Kilauea, located on the south-east is the youngest. The two volcanoes located most south; Kilauea and Mauna Loa, are the most volcanically active (Wolfe & Morris, 1996).
Kohala
The Kohala volcano is built along two rift zones (figures A1 & A7). The north-eastern flank is marked by high cliffs where the original slope is cut away by the erosive forces of the wind and sea (Macdonald et al., 1983). The western flank is gentle sloping, almost unaffected by erosion. Kohala is
28 Figure A1; Digital Elevation Model of the island of Hawaii. Triangles indicate volcanoes. Hawaii-dem Value is in meters. Data from the USGS (Greenberg, 2018), processed in ArcGIS.
in the post-shield stage, due to its relative small age but well defined form. Kohala had already reached its current volume before its southern hillside was covered by the Mauna Kea volcano, towards the south-east from Kohala (Landowski, 2015).
Mauna Kea
The original shield lavas of Mauna Kea are completely covered by pyroclastic material, due to the swift transition from the shield to the post-shield stage. Similarly as on Kohala, on the north-east slope deep canyons have emerged due to the trade wind and rains and the drier western slope is barely influenced by this erosion (figure A1). There is no evidence of any eruptive history for Mauna Kea in the last 2000 years, therefore most researchers classify Mauna Kea in the post-shield stage
(Macdonald et al., 1983; Wolfe & Morris, 1996; Landowski, 2015). Hualalai
The Hualalai volcano is formed along a single rift zone (figure A7). It has just begun the post-shield stage, with a recent eruption in 1801 and earthquakes in 1929, suggesting a still warm magma chamber (Macdonald et al., 1983). Mauna Loa and Hualalai were once active simultaneously, as there are traces found of both lavas twined together (Kauahikaua et al., 2002).
Mauna Loa
Mauna Loa is the most voluminous volcano in the archipelago (table A2), and is still in the shield stage. It is built along two rift zones (figure A7), has maintained its current size and volume since the last ice age and has erupted frequently in the last two centuries (Landowski, 2015). Erosive processes are modest (figure A1). Having
approximately the same age as Hualalai, the two volcanoes have been swelling and shrinking through the movements of faults, with earthquakes as the result (Macdonald et al., 1983).
Kilauea
The youngest volcano is only a small bump on the flank of Mauna Loa, yet completely independent (figure A1). The volcano is in the shield stage, with eruptions along the two rift zones taking place as this research is written (Reardon, 2018). However, the pumice lies on a layer of soil, indicating a period of volcanic inactivity long enough for soil formation to have taken place (Macdonald et al., 1983).
29
Figure A2; From Price & Elliott-Fisk (2004); Summary of Maui Nui history. For each 0.2-myr interval, the approximate range of variation in the area and configuration of islands around that time is summarized. Added in red the names of the mentioned islands.
Figure A3; Digital Elevation Model of the island of Maui. Triangles indicate volcanoes. Maui_dem Value is in meters. Data from the USGS (Greenberg, 2018), processed in ArcGIS.
The island of Maui
Due to glacial configurations, the island of Maui together with Molokai, Lanai and Kahoolawe used to be one large mountain range having a surface area of 14600km2 during the last glacial maximum (figure A2) (Price & Elliot-Fisk, 2004). Maui consists of two volcanoes, the older West Maui and the younger Haleakala. Having otherwise been separated by water, the two islands are connected by the Maui isthmus, a thin streak of land originated by the lava flows of both volcanoes (figure A3) (Faichney et al., 2009).
West Maui
Having completed the first three stages and having performed post-erosional eruptions (eruptions which have taken place after a long period of volcanic
inactivity), the older West Maui volcano is classified in the rejuvenated stage. Having reached a height of 2km, the cap of the volcano collapsed along the three rift zones (figure A7) into a 3km wide caldera, explaining its relative low elevation and the rounded topographic structure as seen from figure A3. From the caldera downwards, lava slopes down at an angle of 10-20 degrees, which is sharper than usual in the archipelago, indicated in red in figure A3 (Macdonald et al., 1983). Lava flows are separated by 1.5m thick soil, suggesting that the volcano did not have one long period of volcanic inactivity, instead it erupted at steady infrequent intervals (Price & Elliot-Fisk, 2004).
Haleakala
The Haleakala volcano is formed along three rift zones (figure A7) (Bergmanis et al., 2000). During a period where erosion was the main force shaping the landscape, two valleys melted together forming one large depression in the landscape (figure A3). When the volcano became active again, lava from the rift zones started to fill up the valley, making it look like a filled caldera, when actually, the Haleakala caldera is the product of erosion. Evidence of eruptions within the last two centuries puts Haleakala in the post-shield stage (Macdonald et al., 1983).
30 Figure A4; Digital Elevation Model of the island of Molokai. Triangles indicate volcanoes. Molokai_dem Value is in meters. Data from the USGS (Greenberg, 2018), processed in ArcGIS.
The island of Molokai
Just as Maui, Molokai used to be part of the Maui Nui complex. Similarly, Molokai is the product of two volcanoes, connected by an isthmus joining the two volcanoes into one island (Price & Elliot-Fisk, 2004). The most distinctive geological feature is that the island is the smallest island. West Molokai
The older West Molokai volcano was built by eruptions along one rift zone (figure A7). Unlike most of the rest of the Hawaiian volcanoes discussed, West Molokai does not show a recognizable summit or caldera (figure A4) (Xu et al., 2007). As can be seen from figure A4, the volcano has a gentle slope and low elevation. The trade winds blow sand inland from the north-west isthmus, forming a dune belt, also called the Desert Strip, extending towards the north-west nearly all the way across West Molokai (Macdonald et al., 1983). The volcano has not showed any renewed signs of activity, and is consequently characterized as being in the post-shield stage.
East Molokai
The island of Molokai is built for two thirds by the East Molokai volcano (Sinton et al., 2017). The north coast of East Molokai is marked by the sharpest slope in the archipelago (figure A4) (Mark & Moore, 1987). Built along two rift zones (figure A7), a caldera caused a depression at the summit of the volcano. Now removed by the force of erosion, this caldera used to be filled by flows of lava (Clague & Moore, 2002). The north shore of the East Molokai volcano is entirely wiped away and is now a sea cliff (figure A4), the result of the immense power of wave erosion fueled by the trade-winds and rains across the Pacific Ocean (Macdonald et al., 1983). However, Moore et al., (1989) suggest that the sea cliff is not just a product of erosion, but that it is the main face of the 40km wide Wailau landslide, which
occurred 1.5 million years ago. High rainfall and active post-shield and post-shield volcanism, as was the case on Molokai, are needed to cause such large landslides (Clague & Moore, 2002). After the formation of the northern cliffs, the East Molokai volcano resumed explosive activity, characterizing it as currently in the rejuvenated stage (Macdonald et al., 1983).
31 The island of Oahu
The two volcanoes comprising Oahu are also called ‘ranges’, expressing their resemblance to a mountain range, instead of a volcano (figure A5). Now both volcanoes are lengthy ridges and owe their shape and topography mostly to erosion. Unlike the isthmus forming Molokai and Maui, the two volcanoes are united by a mildly sloping plateau, built from the lavas of both volcanoes (McMurty et al., 2010).
Waianae
The older Waianae volcano is built along three rift zones (figure A7), which used to form a caldera until the lava spilled over the western flank, resulting in sharp edges where erosion has produced valleys and removed portions of the original shield (figure A5). The eastern flank is little influenced by erosion (Nelson et al., 2013). As the volcano has not shown any renewed signs of activity, it is still in the post-shield stage (Macdonald et al., 1983).
Koolau
The Koolau ridge used to be an exceptional lengthened shield, built on a linear rift zone (figure A7), consequently with a caldera twice as long as it was wide. As seen from figure A5, the caldera lies on the shoreline, showing that almost half of the original shield is no longer present (Macdonald et al., 1983). Most likely this is the work of wave erosion, fueled by trade winds (Nelson et al., 2013). As the volcano showed volcanic activity after an extensive period of inactivity, it is in the rejuvenated stage.
Figure A5; Digital Elevation Model of the island of Oahu. Triangles indicate volcanoes. Circle indicate caldera. Oahu_dem Value is in meters. Data from the USGS (Greenberg, 2018), processed in ArcGIS.
32 The island of Kauai
The oldest island of the main Hawaiian islands the product of one volcano and three rift zones (figure A7), shaped by erosion and later volcanic activity in the rejuvenated stage (figure A6). At the top of the volcano lied the largest caldera of the Hawaiian islands with a diameter of 8 to 9km. Two other small calderas formed on the east and the southeast flanks of the shields. These are the only flank calderas of which we know exist on the Hawaiian islands (Izuka & Resig, 2008).
. Figure A6; Digital Elevation Model of the island of Kauai. Circle indicate caldera. Kauai_dem
Value is in meters. Data from the USGS (Greenberg, 2018), processed in ArcGIS.
33
Appendix B ~ Prevailing climate of the Hawaiian archipelago
The general prevailing climate of the Hawaiian archipelago is tropical. However, the islands experience a variety of other climates as well (figure A8), much depending on factors such as island shape, elevation, and especially by strong trade winds from the Pacific Ocean (Garza et al., 2012). Coming from the north-east towards the south-west, this wind pattern prevails consistently during the whole year (Sanderson, 1993). To illustrate, figure A9 shows the effects these trade winds have on the spatial distribution of precipitation. Overall, the islands experience highest precipitation on the higher elevated and northern and western sides, due to the strong trade winds. Accordingly, the west, south, and lower elevated parts of the islands are much drier (Giambelluca et al., 2013). The youngest island exhibits the most climate orders and the oldest island the least.
Figure A8; Koppen climate classification from Kottek et al., 2006. Climate types calculated from data from WorldClim.org
Figure A9; best estimates of the mean rainfall for the 30-yr base period 1978–2007. Data from Giambelluca et al., (2013). Figure processed in ArcGIS.
