CHARACTERIZING SPATIAL LINKAGES BETWEEN INLAND AND COASTAL 

HABITATS FOR IMPROVED CONSERVATION IN MAUI, HAWAII 

 

 

By 

 

Janet Hsiao  

 

 

 

 

 

 

 

 

 

A THESIS 

Submitted to  

Michigan State University 

in partial fulfillment of the requirements 

for the degree of  

 

Fisheries and Wildlife —Master of Science 

2018 

CHARACTERIZING SPATIAL LINKAGES BETWEEN INLAND AND COASTAL 

HABITATS FOR IMPROVED CONSERVATION IN MAUI, HAWAII 

ABSTRACT 

 

 

By 

 

Janet Hsiao 

 

Characteristics of inland landscapes draining into the coastal environment influence 

condition and availability of coastal habitats. While many studies have shown linkages between 

inland landscapes and coastal regions, few have evaluated such relationships over large spatial 

extents for the purpose of aiding in efforts to conserve coastal habitats. The research described in 

this thesis addresses that need by testing spatial relationships between coastal habitat condition 

and characteristics of proximate draining catchments of Maui, Hawaii. Ecologically meaningful 

inland natural, inland anthropogenic, and coastal factors that influence coastal habitats were first 

identified, and relative influences of each group of variables on coastal habitat metrics were 

quantified. Results indicated that the strength of relationships between catchment characteristics 

and coastal habitat conditions varied with distance. These findings were then used to inform an 

assessment of coastal habitats based on intensity of inland anthropogenic disturbances. This step 

identified coastal areas at varying risk of degradation from inland disturbances, and we combined 

this result with a map depicting locations of ecosystem services and benefits around the island of 

Maui including locations for recreation, areas supporting fishing and/or fisheries and habitats for 

biodiversity. An inventory of those services and benefits most at risk from inland disturbance 

were provided, and this information may prove useful for resource managers working to 

conserve Maui’s coastal habitats. Outcomes of this research further demonstrate an approach that 

could be applied in other regions to understand linkages between inland and coastal habitats, 

improving efforts to conserve coastal habitats from current and future threats. 

Copyright by 
JANET HSIAO 
2018

 

To my family and friends, you are in all of these pages. 

 

iv 

ACKNOWLEDGEMENTS 

The research described in this thesis was made possible by the financial support provided 

by the USGS Aquatic GAP Program, as well as data support provided by the National Fish 

Habitat Partnership and the National Oceanic and Atmospheric Administration Pacific Islands 

Fisheries Science Center. I would also like to thank the Department of Fisheries and Wildlife at 

Michigan State University (MSU), Robert C. and Betty A. Ball, and Equal Opportunities Section 

of the American Fisheries Society for awarding me with travel grants and scholarships to share 

findings from this research with various stakeholders.  

I want to especially thank my advisor, Dr. Dana Infante, for introducing me to the field of 

aquatic landscape ecology and providing a safe space that allowed me to develop as a scientist. 

Dana, thank you for your guidance and enabling my entire graduate school experience. I aspire 

to your kindness and patience, and hope to one day be a mentor that facilitates opportunities for 

others as you have been for me.  

I want to acknowledge the guidance and invaluable feedback provided by my committee 

members, Dr. William Taylor and Dr. Abigail Lynch. Additionally, I would like to thank the 

individuals that has offered professional support, including data analytics, GIS assistance, and 

editorial feedback, in the completion of this research: Ralph Tingley, Kyle Herreman, Wesley 

Daniel, Arthur Cooper, Elizabeth Throckmorton, Jared Ross, and Marlene Üblacker. Finally, I 

am thankful for my family and friends from various communities for their encouragements and 

immense emotional support through this process, especially those from the Aquatic Landscape 

Ecology Lab, Great Lakes Leadership Academy, MSU Center for Systems Integration and 

Sustainability, and the Lansing eastside neighborhood. This project is a collective effort and I am 

genuinely grateful for all the incredible people I have had the privilege of knowing. 

 

v 

PREFACE 

The research chapters in this thesis have been prepared and formatted for publication. Therefore, 

there is some repetition in concept, study site descriptions, and methods among chapters.

 

vi 

TABLE OF CONTENTS 

LIST OF TABLES ........................................................................................................................x 

LIST OF FIGURES ................................................................................................................... xii 

OVERVIEW ................................................................................................................................1 
LITERATURE CITED ..............................................................................................................4 

CHAPTER 1: UNDERSTANDING SPATIAL LINKAGES BETWEEN INLAND AND 
COASTAL SYSTEMS: IDENTIFYING AND QUANTIFYING INLAND INFLUENCES 
ON COASTSAL HABITATS ....................................................................................................7 
Abstract ............................................................................................................................7 
Introduction .....................................................................................................................8 
Methods ..........................................................................................................................11 
Study area ..........................................................................................................11 
Data collection and variable creation ..............................................................12 
Inland landscape data ......................................................................................12 
Spatial units ..................................................................................................12 
Inland natural variables ...............................................................................13 
Inland anthropogenic variables ....................................................................13 
Coastal data .....................................................................................................15 
Coastal habitat data ......................................................................................15 
Spatial units ..................................................................................................15 
Spatial variables ...........................................................................................16 
Natural coastal variables ..............................................................................16 
Characterizing spatial linkages between inland and coastal data ..................17 
Windward and leeward designations ...............................................................17 
Analyses .............................................................................................................18 
Variable reduction ..........................................................................................18 
Redundancy analysis (RDA) ...........................................................................18 
Stepwise multiple linear regression (MLR) ....................................................19 
Results ............................................................................................................................19 
Description of study region ..............................................................................19 
Inland and coastal variables ..........................................................................19 
Reef habitat metrics ........................................................................................20 
Identifying variables important to coastal habitats .......................................21 

Relationships between environmental variables and coastal habitat      
metrics  ...............................................................................................................22 
Redundancy analysis results ..........................................................................22 
Stepwise multiple linear regression results ....................................................23 
Discussion.......................................................................................................................24 

Relative influence of coastal, inland natural, and inland anthropogenic    
groupings ...........................................................................................................26 
Influence of landscape factors on individual reef habitat metrics ...............27 
Development of a coastal spatial framework..................................................30 
Conclusion .........................................................................................................31 

 

vii 

APPENDICES ...........................................................................................................................33 
APPENDIX 1.A: Tables ...............................................................................................34 
APPENDIX 1.B: Figures ..............................................................................................46 
APPENDIX 1.C: Supplemental tables ........................................................................49 
LITERATURE CITED ............................................................................................................55 

CHAPTER 2: ASSESSING THE INFLUENCES OF INLAND LANDSCAPE FACTORS 
ON COASTSAL HABITATS FOR IMPROVED CONSERVATION ................................62 
Abstract ..........................................................................................................................62 
Introduction ...................................................................................................................63 
Methods ..........................................................................................................................66 
Study area ..........................................................................................................66 
Data collection and variable creation ..............................................................67 
Inland landscape data ......................................................................................67 
Spatial units ..................................................................................................67 
Inland disturbances .......................................................................................67 
Coastal data .....................................................................................................68 
Spatial units ..................................................................................................68 
Coastal ecosystem services and benefits .......................................................68 
Associating pour point catchments to coastal grid cells ................................69 
Characterizing influences of cumulative inland disturbances on coastal     
habitats ...............................................................................................................69 
Characterizing risk of degradation from inland disturbances to ecosystem  
services and benefits .........................................................................................70 
Results ............................................................................................................................71 
Associating pour point catchments to coastal grid cells ................................71 
Characterizing influences of cumulative inland disturbances on coastal     
habitats ...............................................................................................................71 
Characterizing risk of degradation from inland disturbances to ecosystem  
services and benefits .........................................................................................72 
Discussion.......................................................................................................................73 
Associating pour point catchments to coastal grid cells ................................74 
Characterizing influences of cumulative inland disturbances on coastal 
habitats ...............................................................................................................75 
Characterizing risk of degradation from inland disturbances to ecosystem 
services and benefits .........................................................................................75 
Study limitations ...............................................................................................77 
Utility of the approach for conserving coastal habitats .................................77 
APPENDICES ...........................................................................................................................79 
APPENDIX 2.A: Tables ...............................................................................................80 
APPENDIX 2.B: Figures ..............................................................................................83 
LITERATURE CITED ............................................................................................................89 

MANAGEMENT IMPLICATIONS .......................................................................................95 
Chapter 1 .......................................................................................................................95 
Chapter 2 .......................................................................................................................97 

 

viii 

LITERATURE CITED ..........................................................................................................100 

 

ix 

LIST OF TABLES 

Table A1.1. Code, units, date, and source for inland natural variables. ........................................35 
 
Table A1.2. Code, units, date, and source for inland anthropogenic variables .............................36 
 
Table A1.3. Code, units, date, and source for all compiled coastal variables ...............................38 
 
Table A1.4. Minimum (min.), maximum (max.), mean, and standard deviation (SD) of inland 
natural variables across all (n = 182), windward side (n = 100), and leeward side (n = 82) pour 
point catchments draining the Maui landscape. Variable descriptions are included in the Methods 
section.. ..........................................................................................................................................39 
 
Table A1.5. Units, minimum (min.), maximum (max.), mean, and standard deviation (SD) values 
of inland anthropogenic disturbance variables across all pour point catchments (n = 182) draining 
the Maui landscape.. ......................................................................................................................40 
 
Table A1.6. Minimum (min.), maximum (max.), mean, and standard deviation (SD) of coastal 
habitat metrics of the reef units. Variable descriptions are included in the Methods section. .......41 
 
Table A1.7. Minimum (min.), maximum (max.), mean, and standard deviation (SD) of coastal 
habitat metrics of the reef units. Variable descriptions are included in the Methods section. .......42 
 
Table A1.8 Results of the PCA of 9 coastal wave variables. PC1 shows the weight of each 
variable, and it explains 86.5% of the variation in wave variables. Variable descriptions are 
included in Table A1.3... ................................................................................................................43 
 
Table A1.9. Percent of total explained variance by RDA predicting habitat variables from the 3 
groups of variables (inland natural, inland anthropogenic, and coastal). Results include unique 
contributions from each group as well as their shared variances. Buffer widths indicate the 
distance between reef units and pour points, and inland factors were summarized for all pour 
point catchments within specified distances. .................................................................................44 
 
Table A1.10. Stepwise multiple linear regression results of best fit model characterizing 
relationships between environmental variables (Tables A1.1, A1.2, A1.3) and coastal habitat 
metrics (Table A1.7). Standardized β is shown in table, and predictor significance is noted for 
p<0.05 (bolded) and p<0.1 (italicized). Results shown are for reef units across the whole island 
(A), on the windward side (W), and on the leeward side (L) of Maui, Hawaii... ..........................45 
 
Table C1.1. Units, minimum (min.), maximum (max.), mean, and standard deviation (SD) values 
of inland anthropogenic disturbance variables across windward side pour point catchments (n = 
100) draining the Maui landscape. .................................................................................................50 
 

 

x 

Table C1.2. Units, minimum (min.), maximum (max.), mean, and standard deviation (SD) values 
of inland anthropogenic disturbance variables across leeward side pour point catchments (n = 82) 
draining the Maui landscape. .........................................................................................................51 
 
Table C1.3. Pearson pairwise correlations among inland natural variables. Bolded variables were 
included in the RDA. Variable descriptions are included in Table A1.1. .....................................52 

Table C1.4. Pearson pairwise correlations of inland anthropogenic variables. The bolded 
variables are included in the RDA. Variable descriptions are included in Table A1.2. ................53 
 
Table C1.5. Pearson pairwise correlations among coastal variables included in the RDA. 
Variable descriptions are included in Table A1.3. .........................................................................54 
 
Table A2.1. Proportion of coastal ecosystem services and benefits at various risk levels of 
degradation from inland disturbances; the greatest proportion for each group is bolded. ............81 
 
Table A2.2. Units, minimum (min.), maximum (max.), mean, and standard deviation (SD) values 
of inland variables within pour point catchments (n = 182) draining the Maui landscape. ...........82 
 
 
 

 

 

 

xi 

LIST OF FIGURES 

Figure B1.1. Location of Maui in the Hawaii archipelago (A), and location of 52 reef units and 
streams draining into the coastal environment around Maui (B). ..................................................47 
 
Figure B1.2. The local and network catchments of a river reach draining at a pour point into the 
coastal environment. ......................................................................................................................48 
 
Figure B2.1. Location of Maui in the Hawaiian archipelago (A), and pour points of catchments 
(n = 182) draining the Maui landscape (B). ...................................................................................84 
 
Figure B2.2. Coastal spatial units around Maui, Hawaii, in relation to pour point catchments of 
the landscape draining into the coastal environment. ....................................................................85 
 
Figure B2.3. Risk of degradation from inland disturbances of coastal units around Maui, 
Hawaii. ...........................................................................................................................................86 
 
Figure B2.4. Location of all ecosystem services overlaid with characterized inland disturbances 
in coastal spatial units. ...................................................................................................................87 
 
Figure B2.5. Location of ecosystem services by group: Recreation (A), Fishing/Fisheries (B), 
and Habitats for biodiversity (C). ..................................................................................................88 

 

xii 

OVERVIEW 

Coastal habitats offer a diversity of benefits that are ecologically and socially important 

(Martínez et al. 2007; Barbier et al. 2011), yet their conditions and availability are in decline 

globally due to anthropogenic disturbances (Myers et al. 2000; Sala and Knowlton 2006; UNEP 

2006). While many anthropogenic disturbances result from human activities in and uses of the 

marine environment (Thrush and Dayton 2002; Waycott et al. 2009; Halpern et al. 2007), 

another major contributor to the decline of coastal habitats originates from the terrestrial 

landscape (Crain et al. 2009), including human land use such as urbanization and agriculture. 

Across the globe, studies have shown links between degradation of coastal habitats and 

developed inland landscapes. Examples include eutrophication in Chesapeake Bay in the Mid-

Atlantic (Boesch 1996), hypoxia in the Gulf of Mexico (Alexander et al. 2008), coral reef 

bleaching of the Great Barrier Reef of Australia (Wooldridge 2009), and algae blooms of Lake 

Erie in the Laurentian Great Lakes (Michalak et al. 2013). These studies emphasize 

consequences of inland landscapes affecting conditions in coastal habitats. 

Despite wide acknowledgement of linkages between inland and coastal habitats, an 

established approach for associating units in the coastal environment with units in proximate 

terrestrial landscapes to consistently account for spatial relationships has yet to be developed. 

This is in part due to the lack of defined boundaries in the open water environment that identifies 

ecologically meaningful units, which hinders efforts to describe relationships between inland 

systems and the coastal environment. Additionally, drivers of coastal habitat condition are 

diverse. Different factors may promote or limit linkages between habitats, and important factors 

may vary by region (i.e., climate, geology, Talley et al. 2006). Understanding spatial 

relationships between characteristics of inland landscapes, including disturbances, and coastal 

 

1 

habitats is integral to identify sources of degradation to coastal ecosystem services (Halpern et al. 

2008; Allan et al. 2013). Such information could potentially inform conservation efforts through 

identification of areas that are highly threatened as well as factors contributing to their decline. 

The goal of this thesis is to improve understanding of spatial linkages between inland and 

coastal systems. Our study was conducted in Maui, Hawaii. The nearshore coastal environment 

supports a variety of ecosystem services and benefits including recreation, recreational and 

commercial fisheries, and habitats for biodiversity. Further, landscape characteristics vary across 

the island’s catchments that drain to the coast, ranging from natural to urbanized to agricultural 

lands. In Chapter 1, we begin by identifying how a set of metrics reflecting reef habitat condition 

throughout the nearshore region of Maui may be affected by variables in three groupings likely 

to influence coastal habitat conditions: inland natural landscape variables, inland anthropogenic 

landscape variables, and coastal variables. We assess the relative contribution of each variable 

grouping in explaining variation in coastal habitat condition, and we conduct analyses by 

considering catchment influences varying distances from coastal habitats to determine how 

proximity affects the strength of relationships.  Additionally, we predict specific coastal habitat 

metrics from a subset of landscape factors to determine important influences on various metrics 

and to assess the role of hydrology in affecting strength. In Chapter 2, we characterize how 

inland anthropogenic influences may be expressed in the coastal environment based on findings 

from Chapter 1. Our analytical approach accounts for inputs from multiple catchments by 

considering their drainage areas and distances to coastal spatial units. We identify areas in the 

coastal environment that are at greatest risk of degradation from inland disturbances in relation to 

coastal ecosystem services and benefits. Then, we assess the coastal ecosystem services and 

benefits that are the most threatened. Collectively, outcomes of this research increase our 

 

2 

understanding of how inland disturbances affect coastal habitats, provide information to support 

resource management decisions, and demonstrate a potential application for assessment by 

enabling the characterization of inland influences on coastal habitats. 

 

3 

LITERATURE CITED

 

4 

LITERATURE CITED 

Alexander, R. B., R. A. Smith, E. Gregory, E. W. Boyer, J. V Nolan, and W. Brakebill. 2008. 

Differences in phosphorus and nitrogen delivery to the Gulf of Mexico from the Mississippi 
River Basin. Environmental Science & Technology 42(3):822–830. 

Allan, A. J. D., P. B. Mcintyre, S. D. P. Smith, B. S. Halpern, and G. L. Boyer. 2013. Joint 

analysis of stressors and ecosystems services to enhance restoration effectiveness. 
Proceedings of the National Academy of Sciences 110(1):372–377. 

Barbier, E. B., S. D. Hacker, C. Kennedy, E. W. Kock, A. C. Stier, and B. R. Sillman. 2011. The 
value of estuarine and coastal ecosystem services. Ecological Monographs 81(2):169–193. 

Boesch, D. F. 1996. Science and management in four US coastal ecosystems dominated by land-

ocean interactions. Journal of Coastal Conservation 2(1):103–114. 

Crain, C. M., B. S. Halpern, M. W. Beck, and C. V. Kappel. 2009. Understanding and managing 

human threats to the coastal marine environment. Annals of the New York Academy of 
Sciences 1162:39–62. 

Halpern, B. S., K. a. Selkoe, F. Micheli, and C. V. Kappel. 2007. Evaluating and ranking the 

vulnerability of global marine ecosystems to anthropogenic threats. Conservation Biology 
21(5):1301–1315. 

Halpern, B. S., S. Walbridge, K. a Selkoe, C. V Kappel, F. Micheli, C. D’Agrosa, J. F. Bruno, K. 
S. Casey, C. Ebert, H. E. Fox, R. Fujita, D. Heinemann, H. S. Lenihan, E. M. P. Madin, M. 
T. Perry, E. R. Selig, M. Spalding, R. Steneck, and R. Watson. 2008. A global map of 
human impact on marine ecosystems. Science 319(2008):948–952.Martínez, M. L., A. 
Intralawan, G. Vázquez, O. Pérez-maqueo, P. Sutton, and R. Landgrave. 2006. The coasts 
of our world : Ecological , economic and social importance. Ecological Economics of 
Coastal Disasters 63(2–3):254–272. 

Martínez, M. L., A. Intralawan, G. Vázquez, O. Pérez-maqueo, P. Sutton, and R. Landgrave. 
2007. The coasts of our world: Ecological, economic and social importance. Ecological 
Economics of Coastal Disasters 63(2–3):254–272. 

Michalak, A. M., E. J. Anderson, D. Beletsky, S. Boland, N. S. Bosch, T. B. Bridgeman, J. D. 

Chaf, K. Cho, R. Confesor, I. Dalo, X. Liu, M. R. Mcwilliams, M. R. Moore, D. J. Posselt, 
R. P. Richards, and D. Scavia. 2013. Record-setting algal bloom in Lake Erie caused by 
agricultural and meteorological trends consistent with expected future conditions. 
Proceedings of the National Academy of Sciences of the United States of America 
110(16):6448–6452. 

Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, and J. Kent. 2000. 

Biodiversity hotspots for conservation priorities. Nature 403(6772):853–858. 

 

5 

Sala, E., and N. Knowlton. 2006. Global marine biodiversity trends. Annual Review of 

Environment and Resources 31:93–122. 

Talley, D. M., G. R. Huxel, M. Holyoak, D. M. Talley, G. R. Huxel, and M. Holyoak. 2006. 

Connectivity at the land-water interface. Pages 97–129in K. R. Crooks and M. Sanjayan, 
editors.Connectivity Conservation. Cambridge University Press. 

Thrush, S. F., and P. K. Dayton. 2015. Disturbance to Marine Benthic Habitats by Trawling and 

Dredging: Implications for Marine Biodiversity. Annual Review of Ecology and 
Systematics 33(2002):449–473. 

UNEP, U. N. E. P. 2006. Marine and coastal ecosystems and human well-being: A synthesis 

report based on the findings of the Millennium Ecosystem Assessment. Page UNEP. 

Waycott, M., C. M. Duarte, T. J. B. Carruthers, R. J. Orth, W. C. Dennison, S. Olyarnik, A. 

Calladine, J. W. Fourqurean, K. L. Heck, a R. Hughes, G. a Kendrick, W. J. Kenworthy, F. 
T. Short, and S. L. Williams. 2009. Accelerating loss of seagrasses across the globe 
threatens coastal ecosystems. Proceedings of the National Academy of Sciences of the 
United States of America 106(19):12377–12381. 

Wooldridge, S. A. 2009. Water quality and coral bleaching thresholds: Formalising the linkage 

for the inshore reefs of the Great Barrier Reef, Australia. Marine Pollution Bulletin 
58(5):745–751. 

 

 

 

 

 

 

 

 

 

 

 

 

 

6 

CHAPTER 1: 

UNDERSTANDING SPATIAL LINKAGES BETWEEN INLAND AND COASTAL 

SYSTEMS: IDENTIFYING AND QUANTIFYING INLAND INFLUENCES ON COASTAL 

HABITATS 

Abstract 

Characteristics of landscapes including catchment area, topography, and land cover draining into 

coastal habitats influence the condition and availability of those habitats. While linkages between 

inland landscapes and coastal habitats are widely acknowledged, few studies have evaluated their 

spatial relationships to aid in conservation. To address this need, we tested for relationships 

between reef habitat condition and characteristics of proximate draining catchments on Maui, 

Hawaii, USA. We identified inland natural, inland anthropogenic, and coastal factors that 

influence coastal habitats and quantified the relative influence of each group of variables. Our 

findings showed evidence for linkages between inland and coastal systems, as well as how the 

strength of relationships varied with distance between catchment pour points and reef habitats. 

Our major findings include that catchment area, annual rainfall, and agricultural land cover 

having strong effects on multiple reef metrics. We also showed differential associations between 

reef habitat metrics and measures of catchment characteristics when considering the windward 

side (wet side) and the leeward side (dry side) of Maui separately. We additionally showed that 

significant relationships between the landscape variables and reef habitat metrics are more 

common on the windward side than the leeward side, underscoring the role of hydrology in 

facilitating linkages. These results illustrate complex relationships between inland systems and 

coastal habitats. Improved understanding of the drivers of coastal habitat condition have 

implications for land and coastal management, which can ultimately help inform management 

decisions that aid in the conservation of nearshore coastal habitats.  

 

7 

Introduction 

Condition and availability of coastal habitats are in decline globally (Myers et al. 2000; 

Sala and Knowlton 2006; UNEP 2006). While human activities worldwide have impacted 

marine ecosystems, consequences of their cumulative disturbances can be more severe in coastal 

regions than in the open ocean (Halpern et al. 2008; Lotze 2006). Due to human settlement in 

landscapes near coastal habitats, average human population density within 100 km of coastlines 

is approximately three times higher than the global average (Small and Nicholls 2003; UNEP 

2006). Intense use of coastal areas can result in generally higher demand for and access to 

coastal resources as well as specific impacts including shoreline development, shoreline 

hardening, and other infrastructure (Creel 2003), and these factors can collectively degrade 

proximate coastal habitats. Additionally, coastal habitats are remotely affected by landscapes far-

removed from the coastal environment, with extensively developed watersheds also having 

negative influences on coastal habitat conditions. 

Influences of anthropogenic disturbances in terrestrial landscapes on coastal habitats are 

driven in part by hydrology. River discharge, surface runoff, and groundwater carry nutrients, 

sediments, and toxics into receiving coastal waters. Anthropogenic disturbances can disrupt 

natural hydrological connections between landscapes and the coast, leading to changes in, and 

potentially degrading, nearshore coastal habitats (Halpern et al. 2008; Wehrly et al. 2013). For 

example, impervious surfaces such as pavement and roads associated with urban land uses could 

prevent water from infiltrating into soils, subsequently reducing groundwater input and 

increasing surface runoff (Paul and Meyer 2001). Such changes in hydrology can have additional 

changes to nearshore coastal habitats including changing thermal regimes (Allan and Castillo 

2007); altering sediment dynamics (Milliman and Meaded 1983); and affecting delivery of 

 

8 

nutrients, toxics, and/or woody debris to coastal habitats (Kennedy and Woods 2012). Across the 

globe, studies have shown links between degradation of coastal habitats and developed inland 

landscapes, including agricultural and urbanized landscapes. Examples include eutrophication in 

Chesapeake Bay in the Mid-Atlantic United States (Boesch 1996), hypoxia in the Gulf of 

Mexico (Alexander et al. 2008), coral reef bleaching of the Great Barrier Reef of Australia 

(Wooldridge 2009), and algae blooms of Lake Erie in the Laurentian Great Lakes (Michalak et 

al. 2013). These studies emphasize consequences of inland landscapes affecting conditions in 

nearshore coastal habitats, in some cases, highlighting mechanisms by which these linkages 

occur in specific locations. However, as individual, regionally-specific studies, they offer little 

guidance for extrapolating results to other regions or for describing and comparing such linkages 

between inland and coastal habitats across large spatial extents (i.e., entire states, ecoregions, 

continents). These limitations reflect a gap in understanding of relationships between connected 

inland and coastal habitats and underscore a critical research need to aid in efforts to conserve 

coastal habitats from anthropogenic disturbances. 

While the importance of linkages between inland and nearshore coastal habitats are 

acknowledged, an approach that characterizes mechanisms by which they connect over large 

spatial extents has yet to be applied (Gorham 1996; Lamberti et al. 2010; Moser et al. 2012). 

Such an approach, however, exists for inland aquatic ecosystems. The landscape approach 

(reviewed in Allan 2004) asserts that conditions of inland lakes and streams are affected by 

conditions of their catchments, and interrelationships between inland aquatic habitats and 

landscape features within catchments have been described in many studies (e.g., Richards et al. 

1996, Wang et al. 2003, Danz et al. 2007, Infante et al. 2009, Esselman et al. 2011, Schinegger et 

al. 2016). However, associating specific areas within the coastal environment with landscape 

 

9 

characteristics is challenging. For rivers and inland lakes, boundaries of landscape influences can 

be determined largely through topography, yet the extent over which landscape influences affect 

coastal habitats is not well-understood. This lack of understanding hinders development of 

ecologically meaningful spatial units that would account for landscape influences on habitat in 

the coastal environment. Additionally, drivers of coastal habitat condition are diverse. Different 

factors may promote or limit linkages between habitats, and important factors may vary by 

region (Talley et al. 2006). In spite of these challenges, the need for a spatial framework that 

explicitly associates coastal habitats with inland spatial units is recognized. For example, in 

2013, the National Ocean Council released the National Ocean Policy Implementation Plan that 

emphasized the need for coastal and marine spatial planning, calling for a formalized method to 

inform and guide decision-making in the coastal and marine environments (NOC 2013). With 

diverse interests in coastal and marine resources (e.g., shipping, offshore energy, recreation), 

decision-support spatial planning tools based on a spatial framework would alleviate 

shortcomings of and complexity related to management decisions that consider conservation 

(Craig 2015). 

This study aids in addressing those challenges by investigating spatial linkages between 

inland and coastal systems of Maui, Hawaii, USA. Our first objective is to identify factors that 

most strongly influence coastal habitats in a region with substantial heterogeneity in both inland 

and coastal characteristics. In support of this objective, we first identify sets of coastal factors 

and of natural and anthropogenic landscape factors from the inland environment which may be 

important influences on coastal habitats. Our second objective tests for influences of those three 

groupings of factors on a set of reef metrics reflecting coastal habitat condition to identify the 

most influential grouping. Finally, our third objective predicts specific reef metrics from inland 

 

10 

and coastal factors to characterize the importance of different inland influences and identify 

those that explain the most variance in coastal habitat conditions. In support of this objective, we 

additionally test for the role of hydrology in affecting influences by conducting analyses on 

Maui’s catchments that receive substantially different amounts of precipitation. Greater 

understanding of spatial linkages between inland and coastal environments gained through this 

study can contribute to development of a spatial framework in the coastal environment. In this 

case, a spatial framework is composed of a set of spatial units that are ecologically-defined with 

shared characteristics. The explicit characterization of interrelationship of inland catchment 

characteristics and coastal habitat units can ultimately help inform management decisions and aid 

in conserving nearshore coastal habitats. 

Methods 

Study area 

This study was conducted in Maui, Hawaii, the second largest island of the Hawaiian 

archipelago with a land area of 1,884 km2 (Figure 1.1). The island was formed by the 

convergence of two volcanoes (West Maui Volcano, formed approximately 1.15 million years 

ago and East Maui Volcano, formed approximately 0.8 million years ago). Maui’s characteristic 

basalt-based geology is derived from cooled lava flows,which contribute to variability in 

geologic permeability across the island (Lau and Mink 2006). Due to prevailing trade winds and 

orographic effects, the windward side of Maui receives substantially more rainfall than the 

leeward side (Giambelluca et al. 2011), contributing to a higher prevalence of perennial streams 

on the windward side of the island and more intermittent streams on the leeward side (Timm et 

al. 2015, Giambelluca et al., 2011). Additionally, Maui exhibits distinct wet seasons (November–

April) and dry seasons (May–October). 

 

11 

Approximately 206 km2 of coral reef habitat occurs in the nearshore region around the 

coastline of Maui, supporting 12 species of corals (Grigg 1983). These corals provide habitat for 

a diversity of fish and invertebrates (Randall 1976, Friedlander et al. 2008), including endemic 

species that utilize the coastal habitats to complete their life cycle (McDowall 2003). Nearshore 

coastal habitats around Maui are hydrologically linked to the island’s landscape and are subject 

to both natural and anthropogenic influences. Rain events rapidly alter sediment inputs, turbidity, 

and salinity of coastal receiving waters (Grigg 1995, Lau and Mink 2006). While coastal 

organisms are adapted to natural extremes in conditions (e.g., Bittler et al. 2014, Moura et al. 

2016), anthropogenic activities can result in conditions that may be substantially different than 

those to which native organisms are adapted (DLNR DAR 2005, Storlazzi and Jaffe 2008). 

Additionally, both natural and anthropogenic influences originating within the marine 

environment (i.e., wave impacts, damage from ship anchors) may interact with and compound 

effects of inland influences on coastal habitats (Chabanet et al. 2005, Buma 2015). 

Data collection and variable creation 

Inland landscape data 

Spatial units  

To explore relationships between inland and coastal systems, we identified boundaries of 

all catchments draining to the coastal environment by perennial and/or intermittent streams and 

termed these spatial units “pour point catchments” (Figure 1.2). Pour point catchments were 

derived from a stream layer developed in support of the Hawaii Fish Habitat Partnership (HFHP) 

by Tingley et al. (In Review). Confluence-to-confluence stream reaches form the basis of the 

HFHP stream layer, and a local catchment is the topographically-defined land area that drains 

directly to the stream reach (following Wang et al. 2011). Pour point catchments include the 

 

12 

cumulative land area and river network upstream of the pour point draining into the coastal 

environment. 

Inland natural variables 

Inland natural variables summarized in pour point catchments were assembled from 

multiple sources to describe characteristics of the Maui landscape draining into the coastal 

environment (Table A1.1). Catchment slope was calculated by averaging slopes of stream 

reaches comprising networks within pour point catchments, and pour point catchment area is the 

sum of the area of all local catchments comprising a pour point catchment (methods described in 

Tingley et al., In Review). Mean annual rainfall data were summarized from Frazier et al. (2016), 

representing the area-weighted average depth of precipitation received across each local 

catchment comprising a pour point catchment. Soil and geologic characteristics across Maui 

were sourced from the Soil Survey Geographic Database (SSURGO; USDA 1995). SSURGO 

hydrologic soil groupings are estimates of relative soil infiltration rates. For this study, we used 

area-based weighting of values in all local catchments comprising pour point catchments. Values 

ranged from 1 to 4, where 1 represents the lowest infiltration rate and 4 represents the highest 

(see Tingley et al., In Review). SSURGO soil erodibility was similarly summarized by area-

based weighting of values in local catchments comprising pour point catchments; values ranged 

from 1 to 3 where 1 is not highly erodible and 3 is highly erodible (Tingley et al., In Review). 

Inland anthropogenic variables 

To identify potential disturbances to the coastal environment, anthropogenic variables 

assembled from multiple sources were also summarized in pour point catchments (Table A1.2). 

Urban landscape influences were characterized by multiple variables sourced from the Coastal 

Change Analysis Program (CCAP; NOAA 2005). These included high, medium, and low 

 

13 

intensity developed urban lands; open urban land; and a composite variable created by summing 

the above listed subcategories within pour point catchments. Additional urban landscape 

influences summarized in catchments included impervious surfaces (NOAA 2005); utility 

pipelines density (Hawaii Office of Planning 1983); percent golf course surface cover (Hawaii 

OP); population density (United States Census Bureau (USCB) 2002); and road density (USCB 

2002). These urban land variables represent potential sources of pollution and may alter 

hydrologic regimes, contributing to degradation of coastal habitat conditions or habitat loss. 

Agricultural landscape influences included percent of hay and cultivated crop land cover as well 

as a composite variable created by summing the two categories in pour point catchments (CCAP 

2005); these factors may also alter catchment hydrology or may result in inputs of excess 

nutrients, sediments, and/or toxics including herbicides and pesticides to coastal receiving 

waters. Percent of former plantation land cover accounts for pervasiveness of historical 

pineapple and sugarcane plantations within pour point catchments (Hawaii Office of Planning 

1989). This landscape factor was also considered a potential current disturbance that could result 

in discharges of nutrients, pesticides, and/or herbicides to coastal habitats. An additional 

disturbance, point-sources, include densities of Superfund National Priority sites 

(Comprehensive Environmental Response, Compensation and Liability Information System 

(CERCLIS; USEPA 2010), Permit Compliance System majors (PCS; USEPA 2010), Toxic 

Release Inventory sites (TRI; USEPA 2010), and Underground Injection Control sites (UIC; 

Hawaii DOH 2004) in pour point catchments. Stream fragmentation variables include densities 

of stream and road crossings (USCB 2002), densities of ditch intersections with streams (Hawaii 

Division of Aquatic Resources (DAR) 2004), and densities of dams (ACOE 2010). Length of 

ditches characterizes the relative intensity of water diversions within pour point catchments 

 

14 

(Hawaii DAR, 2004). Additionally, the percentage of upstream network classified as 303(d) 

listed streams (USEPA 2002) characterizes reaches that failed to meet state criteria for water 

quality, primarily due to sedimentation (Hawaii Department of Health 2012). 

Coastal data 

Coastal habitat data 

Coastal habitat data were obtained from the National Oceanic and Atmospheric 

Administration (NOAA) Coral Reef Ecosystem Program. Reef characteristics around the 

coastline of Maui were surveyed between 2005 and 2010. There were 888 benthic segment 

observations (each approximately 200 m in length) from 92 unique surveys that characterized 

coastal habitat factors following NOAA’s Benthic Towed-diver Survey protocol (Kenyon et al. 

2006). These factors include measurements of seafloor cover types throughout segments such as 

percent coral cover (C), stressed coral cover (SC), sand cover (S), macroalgae cover (MA), and 

crustose coralline algae cover (CCA). Habitat complexity (H) was also assessed in segments and 

is a measurement of structural rugosity that potentially reflects amount of refuge for reef 

organisms. Numbers of crown-of-thorns starfish (COTS) and urchins (U) were estimated to 

assess their respective densities (count per km2) in each segment. 

Spatial units 

Each of the 92 unique reef surveys were composed of multiple benthic segment 

observations collected within the same survey covering areas that ranged from 0.03 km2 to 0.59 

km2 around the island of Maui. Individual variables measured (described below) were averaged 

across observations from individual surveys, with the spatial region assessed by a survey 

hereinafter referred to as a reef unit. When reef units characterized areas that overlapped 50% or 

more, we retained information from the most recently-sampled unit. This step left us with 52 

 

15 

unique reef units for analysis (Figure 1.1), encompassing an area of approximately 21.84 km2 of 

the nearshore environment around Maui. 

Spatial variables 

Depth measurements were recorded at 5 minute intervals for each benthic segment. 

Multiple measurements were averaged for each benthic segment, and segment depths were again 

averaged to calculate a single measure of average depth for each reef unit. Shortest distances 

between end points of benthic segments to the shoreline and to the nearest pour point were 

calculated. Their respective measurements were averaged for each reef unit (Table A1.3). 

Natural coastal variables 

Natural coastal variables describe physical characteristics of the nearshore environment. 

Water temperature readings were recorded at 5 minute intervals for each benthic segment. Wave 

variables representing monthly environmental conditions of waters were assembled from the 

National Renewal Energy Laboratory (2011) and organized into 5 km by 5 km (25 km2) grid 

cells in the coastal environment (Table A1.3). Waves may be natural disturbances on the 

community structure of coral reefs, and their varying strength and patterns affect morphology 

and organismal assemblages of coastal habitats (Dollar 1982; Kaandorp and Kübler 2001). We 

considered multiple metrics to capture different wave characteristics. First, wave significant 

height is the distance from the trough to crest of waves, and greater wave height may represent 

greater potential disturbance to reef structures at greater depth (Gourlay 2011). Next, wave 

energy period is the amount of time it takes two consecutive wave crests to pass a fixed point, 

and shorter period may reflect more frequent impacts on reef structure (Gourlay 2011). Finally, 

wave power density measures the amount of energy generated by waves, with greater power 

potentially acting as a disturbance that may agitate reef communities (Gourlay 2011). We first 

 

16 

summarized values of wave measures by annual averages, by dry season (May – October), and 

by wet season (November – April) (Table A1.3) within each of the grid cells, resulting in a total 

set of 9 wave variables. Each benthic segment was attributed with the grid cell values in which 

they occurred, and values for segments were averaged to determine a single value for each reef 

unit.  

Characterizing spatial linkages between inland and coastal data 

To test the idea that greater proximity of reef units to pour points leads to greater 

influence of landscapes drained by rivers on reef units, we created buffers of varying widths 

around each reef unit and summarized landscape information for pour points in buffers following 

Jouffray et al. (2014). Pour point catchments were associated with a reef unit when pour points 

fell within distances defined by 1 km-, 3 km-, 5 km-, and 10 km-wide buffers. Areas drained by 

all catchments with pour points in buffer distances were summed to yield a total drained area, 

and other inland metrics were summarized by area-based weighting for further analysis. 

Windward and leeward designations 

Another idea we tested is that hydrology is a driver of inland and coastal linkages that 

may affect the extent of potential connections. We expected that influences of landscapes 

receiving more precipitation around Maui may be stronger and/or more easily detected in coastal 

environments. To account for varying precipitation levels experienced on windward and leeward 

sides of Maui, we used a demarcation based on stream discharges developed by the United States 

Geological Survey (USGS, Yamanaga 1972) that split the island into generally wetter 

(windward) and drier (leeward) sides (Figure 1.1). Some analyses (described below) were 

conducted within these regions to identify potential differences across the island. 

 

 

17 

Analyses 

Variable reduction 

To select minimally redundant variables that may explain coastal habitat conditions, we 

assessed groups of inland anthropogenic, inland natural, and coastal natural variables 

independently. We first assessed minimum, average, and maximum values of variables in each 

group and eliminated variables that did not vary substantially on Maui. Percentage variables 

were then arcsine square root transformed, and continuous variables were natural log 

transformed for additional analyses. For the inland anthropogenic and inland natural variable 

groups, Pearson pairwise correlations were assessed among remaining variables. When pairs of 

variables were found to be highly correlated (|r| ≥ 0.65), one was eliminated based on ecological 

interpretability. Principal component analysis (PCA, Legendre and Legendre 2012) was 

performed to identify major variables describing wave characteristics of the natural coastal 

variables. To aid in interpretation, Varimax rotation was applied to resulting axes. These steps 

resulted in a subset of variables used for further analysis. 

Redundancy analysis (RDA) 

To estimate the relative influences of the three variable groups (inland natural, inland 

anthropogenic, coastal) and to investigate overall inland influences compared to coastal 

influences on reef habitat metrics, we used redundancy analysis (RDA; e.g., Borcard et al. 1992) 

with the program CANOCO 5 (Leps and Smilauer 2003). We ran the analysis using summaries 

of landscape factors in multiple pour points catchments varying distances from reef units (1 km, 

3 km, 5 km, 10 km) to characterize how proximity of catchments could influence proportional 

variance explained by each variable group. 

 

 

18 

Stepwise multiple linear regression (MLR) 

Stepwise multiple linear regression (MLR) was conducted to identify the most important 

environmental variables and their directionality in predicting each of the coastal habitat metrics. 

Model selection for each habitat metric was performed using Akaike Information Criterion (AIC) 

(Burnham and Anderson 2002). A standardized regression coefficient (β) was calculated to 

evaluate the relative strength of variables selected for inclusion in the models predicting coastal 

habitat metrics. Regressions were performed for the entire island, the windward (wet) side, and 

the leeward (dry) side to assess the role of hydrology in establishing linkages between inland and 

coastal systems. 

Description of study region 

Inland and coastal variables  

Results 

Areas of pour point catchments on Maui are relatively small, ranging from 0.08 km2 to 

160.29 km2 with a mean of 7.64 km2 (Table A1.4). Catchment slopes on Maui, in contrast, are 

steep, ranging from 0.27% to 31.20% with a mean of 14.77% (Table A1.4). Higher gradients 

may contribute to inland inputs being delivered to coastal environments quickly following 

precipitation events, leading to disturbances of greater magnitude and/or intensity than from 

streams with lower gradients. Hydrologic soil grouping values for catchments ranged between 

1.00 and 3.65 (on a scale of 1 to 4, 1 being the lowest infiltration rate), with a mean of 2.11 

(Table A1.4). Annual precipitation also varied across catchments, ranging from 279.63 mm/yr to 

8403.95 mm/yr, with a mean of 2497.48 mm/yr (Table A1.4). Means and ranges of inland 

natural landscape factors were similar between the windward and leeward sides of Maui, except 

 

19 

for precipitation. The average value for catchments on the windward side is 3175.64 mm/yr, and 

the average value for catchments on the leeward side is 1670.46 mm/yr (Table A1.4). 

Developed land cover in Maui catchments averaged 4.5% and ranged from 0.00% to 

40.07% (Table A1.5), and agricultural land cover in Maui averaged 9.49% and ranged from 

0.00% to 84.02% (Table A1.5). These patterns were similar on the windward vs. leeward sides of 

the islands, as were patterns for road crossing density. Ditch length density varied, however; 

catchments on the windward side of the island had greater ditch length density (average of 

391.74 m/km2, Table A1.1) than catchments on the leeward side (average of 213.53 m/km2, 

Table A1.2). Other inland anthropogenic disturbances were relatively similar in average values 

and ranges between windward and leeward sides (Table A1.1, Table A1.2). 

Reef units around Maui were on average 373.68 m from the shoreline, 1385.52 m from 

pour points, and 14.16 m deep (Table A1.6). While some coastal factors are similar for reef units 

on the windward and leeward sides, some metrics differed substantially. Most notably, annual 

wave power density was much higher on the windward side (11.53 kW/m) vs. the leeward side 

(4.14 kW/m, Table A1.6). While the differences in average wave power density between the wet 

season and dry season on the windward side is 9.73 kW/m, seasonal differences are less extreme 

on the leeward side (1.46 kW/m, Table A1.6). Distance to pour points is on average 858.84 m on 

the windward side but 1803.23 m on the leeward wide (Table A1.6), and greater distance could 

influence the strength of influences of inland landscape factors on coastal habitats on the leeward 

side vs. the windward side of Maui (Figure 1.1). 

Reef habitat metrics 

Coastal habitat metrics also varied around Maui. Around the entire island, percent coral 

cover in reef units averaged 14.05%, stressed coral cover averaged 1.46%, macroalgae cover 

 

20 

averaged 11.17%, and crustose coralline algae cover averaged 4.55% (Table A1.7). Some 

metrics also differed on the windward vs. leeward sides. Percent sand cover averaged 30.12% 

across the island, with less sand cover observed on the windward side than the leeward side 

(16.19% vs. 41.17% respectively, Table A1.7). Additionally, crown-of-thorns starfish density 

was higher (0.00014 vs. 0.000028 #/m2) and urchin density was lower (0.00014 vs. 0.0081 #/m2) 

on the windward vs. leeward side of the island (Table A1.7). 

Identifying variables important to coastal habitats 

Four minimally redundant natural inland variables were selected for further analysis 

based on investigation of correlations among five natural inland variables: catchment area, 

catchment slope, mean annual rainfall, and average and minimum hydrological soil grouping 

(Table A1.3). Minimum hydrological soil grouping was eliminated from further analysis due to 

its high correlation with average hydrological soil grouping. Investigation of correlations among 

14 inland anthropogenic variables also resulted in four minimally redundant anthropogenic 

variables retained for further analysis: urban land cover, agricultural land cover, ditch length 

density, and road crossing density (Table A1.4). Urban land cover and agricultural land cover 

were selected for their potential to alter catchment hydrology and to be sources of pollution and 

nutrient inputs. Road crossing density (which characterizes stream network fragmentation) and 

ditch length density (which characterizes water diversions in catchments) were selected for 

potential influence on hydrology. 

In the coastal factor grouping, PCA of nine variables representing multiple aspects of 

wave characteristics resulted in one axis which explained 86.5% of the total variation (Table 

A1.8). This outcome shows high redundancy among wave characteristics, and because of this, 

we selected a single variable, annual wave power density (Ann_wef), for further analyses based 

 

21 

on its ecological interpretability. Three other variables were used to characterize coastal 

conditions including distance from shore, distance from catchment pour point, and bathymetric 

depth.  These variables and annual wave power density were not highly correlated with each 

other (Table A1.5).  

Relationships between environmental variables and coastal habitat metrics 

Redundancy analysis results  

Redundancy analysis (RDA) showed that inland natural, inland anthropogenic, and 

coastal groupings of variables explained a substantial amount of variation in coastal habitat 

metrics. When pour points occurred within 1 km of reef units, 38.7% of total variation in reef 

metrics was explained; when pour points occurred within 3 km of reef units, 40.2% of total 

variation was explained; and when pour points occurred within 5 km of reef units, 39.6% was 

explained. Amount of explained variance dropped when catchments within 10 km of reef units 

were tested (32.4%), and because of this, regression analyses (described below) were conducted 

using landscape factors summarized within pour point catchments occurring 5 km from reef 

units.  

Although RDA results showed that coastal variables explained the most variation in 

coastal habitat metrics of all variable groups tested for all 4 buffer extents (Table A1.9), our 

findings emphasize the importance of inland influences on reef metrics. Almost six percent of 

the variance in metrics was explained collectively by inland natural and inland anthropogenic 

factors for catchments with pour points occurring within 5 km of reef units (Table A1.9). We 

showed that as distance between the catchment pour points and reef units increased, the 

contribution of inland natural influences to explained variance increased while the contribution 

of inland anthropogenic influences decreased (Table A1.9). The greatest inland anthropogenic 

 

22 

influence was detected 1 km from reef units; the greatest inland natural influence was detected 5 

km from reef units (Table A1.9). 

Stepwise multiple linear regression results 

Using stepwise multiple linear regression (MLR), individual reef habitat metrics were 

predicted from environmental factors. For all study sites occurring around Maui, the best 

predicted habitat metric was habitat complexity (R²adj = 0.54). Other well-predicted metrics 

(R²adj > 0.2) include: percent sand cover (R²adj = 0.44), crown-of-thorns starfish density (R²adj 

= 0.25), percent coral cover (R²adj = 0.23), and percent macroalgae cover (R²adj = 0.22, Table 

1.10). Reef metrics that were not well-predicted (R²adj < 0.2) include stressed coral cover, 

crustose coralline algae cover, and urchin density (Table A1.10). 

Coastal factors were selected more frequently as significant predictors of reef habitat 

metrics than inland natural factors and inland anthropogenic factors. Coastal factors significantly 

(p<0.1) predicted 7 metrics, inland natural factors predicted 4 metrics, and inland anthropogenic 

factors predicted 2 metrics (Table A1.10). Of all coastal factors, annual wave power density 

significantly predicted the most metrics. It was positively associated with habitat complexity and 

crown-of-thorns starfish density and negatively associated with percent sand cover (Table 

A1.10). Other important coastal predictors were depth and average distance from pour points. 

Depth was negatively associated with habitat complexity and percent coral cover and positively 

with percent sand cover (Table A1.10). Average distance from pour points was positively 

associated with percent coral cover (Table A1.10).  

Inland natural and anthropogenic factors were associated with several reef habitat metrics 

at the whole-island scale across Maui. Our best-predicted reef metric, habitat complexity, was 

negatively associated with rainfall and positively associated with pour point catchment area 

 

23 

(Table A1.10). Rainfall was also negatively associated with urchin density (Table A1.10). 

Agricultural land cover was a significant and positive predictor of two habitat metrics: percent 

macroalgae cover and crown-of-thorns starfish density (Table A1.10). In contrast, urban land 

cover, ditch length density, and road crossing density were never significant predictors of any 

reef habitat metrics at the whole-island scale (Table A1.10). 

We further assessed results for the windward vs. leeward sides of Maui to identify 

potential differences that may be associated with different amounts of precipitation. Overall, 7 of 

8 reef habitat metrics had more variance explained on the windward side than the leeward side of 

Maui; only percent sand cover was better explained on the leeward side (Table A1.10). Coastal 

variables significantly predicted 5 metrics on the windward side and 1 metric on the leeward side 

(Table A1.10), and inland natural variables significantly predicted 4 metrics on the windward 

side and 1 metric on the leeward side, suggesting that coastal habitat factors may be more 

strongly influenced by these landscape influences on the windward vs. leeward sides. In contrast 

to this, inland anthropogenic variables significantly predicted 4 metrics on both sides of Maui 

(Table A1.10). Agricultural land use was positively associated with coral cover on the windward 

side, and negatively associated with coral cover on the leeward side (Table A1.10). Agricultural 

land use was also positively associated with macroalgae cover, crown-of-thorns starfish density, 

and urchin density on the windward side (Table A1.10). Urban land use significantly predicted 3 

reef metrics on the leeward side; it was negatively associated with habitat complexity and 

positively associated with sand cover and macroalgae cover (Table A1.10). 

Discussion 

Our results identified factors that contribute to linkages between inland and coastal 

habitats. We showed how associating a unit in the coastal environment with potential inland 

 

24 

influences at various distances alters the strength of relationships, and additionally demonstrated 

how relationships can vary with amount of precipitation received by catchments. In support of 

our first objective, we selected a set of minimally redundant and ecologically meaningful 

variables likely to influence coastal habitats, and we addressed our second objective by 

quantifying the relative influences of three groupings of those variables (inland natural, inland 

anthropogenic, and coastal factors) on 8 reef habitat metrics that reflect coastal habitat 

conditions. Coastal factors explained the most variance in reef habitat metrics, yet inland 

influences derived from river pour points up to 5 km from reef units also explained substantial 

amounts of variation in coastal habitats. Inland anthropogenic influences are the greatest 1 km 

from reef units and decrease with the inclusion of more catchments farther away from reef units, 

while inland natural influences increased as more catchments were considered up to 5 km. For 

our final objective, we used multiple linear regression to independently predict reef metrics from 

inland natural, inland anthropogenic, and coastal variables to identify the most influential 

predictors and assess how hydrology (as reflected by substantially different amounts of 

precipitation on windward vs. leeward Maui) may affect associations between reef metrics and 

predictors. While 5 out of 8 reef habitat metrics were well-predicted across the entire island of 

Maui, we also observed that more reef habitat metrics were significantly predicted on the 

windward side than the leeward side of Maui by coastal and inland natural variables. This was 

not the case for inland anthropogenic variables, which predicted 4 out of 8 reef habitat metrics on 

both the windward side and leeward side. These findings underscore linkages between inland 

anthropogenic influences and coastal habitats, which has particular relevance for understanding 

and managing landscape-derived disturbances on reefs in Hawaii.  

 

 

25 

Relative influence of coastal, inland natural, and inland anthropogenic groupings 

Based on results of the redundancy analysis, coastal factors as a group explained more 

variance in the full set of reef habitat metrics than the inland factors in our study region. This 

result is not unexpected given proximity and the role of natural coastal factors, such as wave 

characteristics and bathymetric depth, in determining benthic habitat characteristics. For 

example, Dollar (1982) tested for effects of wave energy on coral reef communities throughout 

Hawaii and showed that physical disturbances from waves are dominant drivers of coral 

community structure. Additionally, depth is known to affect wave exposure of benthic habitats 

and light intensity through the water column, thus indirectly influencing morphology and biology 

of coastal habitats (Kaandorp and Kübler 2001).  

We also observed that inland influences explained a portion of reef habitat variation, a 

finding that also follows those of other studies describing relationship between coastal habitat 

conditions and inland landscape characteristics. In the US Virgin Islands, Oliver et al. (2011) 

found that percent impervious surface in watersheds was negatively associated with percent coral 

cover. In the main Hawaiian Islands, Rodgers et al. (2012) identified a strong, positive 

correlation between a watershed health index and reef habitat index that were developed 

separately. Both studies showed how measures of coral reef health are negatively associated with 

disturbed landscapes in proximate watersheds. Our findings build on that understanding, and we 

additionally show how distance may influence relationships. We showed that relationship 

strength between overall inland influences and coastal habitat conditions was largely consistent 

for pour points within 5 km of reef units, but the importance of different types of influences 

varied with distance. Most notably, amount of variance explained by inland anthropogenic 

influences was greatest 1 km from reef units. In contrast, the amount of variance explained by 

 

26 

inland natural influences was greatest 5 km from reef units. This could in part be due to inland 

anthropogenic influences having more localized impacts on habitats that are in close proximity, 

i.e., within 1 km. To our knowledge, no other studies have explicitly tested for distances in 

exploring landscape-scale effects of inland influences on coastal habitats for multiple watersheds 

over a large heterogeneous region. 

Influence of landscape factors on individual reef habitat metrics 

We used environmental variables to predict each of the reef metrics individually, and the 

findings illustrate complex interrelationships between landscape factors and coastal habitats. For 

the 5 best-predicted metrics at the whole-island scale, habitat complexity and macroalgae cover 

were predicted by inland natural variables, while macroalgae cover and crown-of-thorns starfish 

density were predicted by inland anthropogenic variables. This result shows the complexity of 

inland linkages to coastal habitats and highlights greater sensitivity of some metrics over others 

to inland disturbances. Coastal habitat conditions are often characterized by physical and 

biological metrics as synthesized by Diaz et al. (2004) in their review of the state-of-knowledge 

in classifying and evaluating coastal habitat quality. In addition, they emphasize that choice of 

most effective metrics should vary with research or management interests. In our case, metrics 

associated most strongly with inland anthropogenic variables are biological (i.e., macroalgae 

cover and crown-of-thorns starfish density), as opposed to metrics that are largely physical (i.e., 

habitat complexity, sand cover). Our identified association between biology of coastal habitats 

and anthropogenic land use follows that of Mallin et al. (1993); they showed increased primary 

production in coastal habitats with heavily agricultural catchments throughout the Neuse River 

Estuary in North Carolina, USA. Similarly, Babcock et al. (2016) reported on outbreaks of 

crown-of-thorns starfish after rain events at Australia’s Great Barrier Reef, with rain events 

 

27 

facilitating nutrient inputs from the landscape from various anthropogenic practices including 

fertilizer application. These results have particular relevance for understanding how reef habitat 

metrics that represent different aspects of the coastal environment may respond differently to 

various predictors. Our results highlighted the greatest inland disturbances to Maui’s coastal 

habitats and the most sensitive biological attributes of those habitats, providing useful 

information for conservation of Maui’s coastal habitats as well as guidance for efforts in other 

regions. Specifically, this result highlights greater sensitivity of some metrics. A manager of 

coastal habitats may consider those reef metrics associated most strongly with inland 

anthropogenic variables as indicators of inland disturbances. 

Of all variable groupings tested, coastal variables including annual wave power density, 

depth, and distance to pour point were significant predictors of 4 out of the 5 well-predicted reef 

habitat metrics across the island of Maui. Jokiel et al. (2004) showed the importance of wave 

characteristics on the main Hawaiian Islands as major natural factors influencing reef structure, a 

conclusion that follows our finding at the whole-island scale, where wave power density was 

associated with habitat complexity. Jokiel et al. (2004) also showed the importance of depth by 

characterizing how reef communities found at various depths differed in composition and in their 

response to changes in environmental conditions, which supports our observation of depth being 

negatively associated with habitat complexity and coral cover. This is explained by depth being a 

landscape control of the major environmental parameters (i.e., light and hydrodynamics) in 

determining community structure of coastal benthic organisms (Kaandorp and Kübler 2001). 

Glynn (1985) suggested that outbreaks of crown-of-thorns starfish in the tropical eastern Pacific 

region occur with major disturbances such as strong wave events associated with El Niño, which 

may partially explain the positive relationships we observed between wave power density and 

 

28 

crown-of-thorns starfish density. Greater average wave power density could suggest an area of 

the coastal environment being subjected to greater physical disturbances. Friedlander and Parrish 

(1998) found that coral community diversity and evenness in Hanalei Bay of Kauai, Hawaii, 

were related to distance to the river mouth, which is supported by the association we detected 

between distance to catchment pour points and coral cover.  

Inland characteristics also influenced coastal habitats, with rainfall, catchment area, and 

agricultural land cover identified as significant predictors of select habitat metrics. We showed 

that rainfall is negatively associated with habitat complexity, a finding that follows Milliman and 

Meade (1983) who estimated the quantity of inland-sourced sediments delivered to the coast was  

driven strongly by precipitation and area of the draining catchment. Our results additionally 

follow those from a synthesis from Fabricius (2005), which describes that runoff associated with 

agriculture and urbanization affects coral reef population dynamics –specifically, increases in 

macroalgae and crown-of-thorns starfish. 

We saw further evidence for connections between inland and coastal systems by looking 

at relationships between significant predictors and reef habitat metrics and assessing how 

findings differed between the windward and leeward sides of Maui. Coastal and inland natural 

variables were significant predictors of reef habitat metrics much more frequently on the 

windward side than leeward side, suggesting that reef habitats on the windward side are 

subjected to more influence from the coastal and inland natural variables that we tested in our 

study. This could be in part attributed to greater amounts of precipitation on the windward side 

that flushes potentially more catchment-derived materials into the open water environment 

(Rodgers et al. 2012). Additionally, all reef metrics were better predicted on the windward side 

except sand cover. Infrequent precipitation events can be associated with increased concentration 

 

29 

of sediments in runoff from rain events (Langbein and Schumm 1958), resulting in rain events on 

the leeward side transporting greater quantities of sediments to the coast. In addition, we noted 

that inland anthropogenic variables predicted the same number of reef metrics across the 

windward and leeward side of Maui. Our finding suggests that agricultural land cover may be the 

primary disturbance to coastal habitats on the windward side, and urban land cover may be the 

primary disturbance on the leeward side. Though specific mechanisms by which inland 

anthropogenic variables influence reef habitats are still unclear, it is worth noting the importance 

of urban land cover as a significant predictor on the leeward side despite that the windward 

catchments have greater amounts of urban land cover. The linkage between inland and coastal 

habitats are complex, but it is evident that the connection is in part driven by hydrology. 

Development of a coastal spatial framework 

Findings from our research could inform development of a spatial framework, which 

accounts for how inland influences are expressed in the coastal environment. First, inland natural 

and inland anthropogenic influences from pour points 5 km from reef units in our study explain 

the greatest amounts of variation in reef habitat metrics. A hypothetical spatial framework in the 

coastal environment of our study region could weigh inland natural characteristics of catchments 

with pour points falling within 5 km of a coastal spatial unit more heavily than catchment pour 

points that are farther than 5 km. Second, compared to inland natural factors, inland 

anthropogenic factors are best detected from pour points that fall within 1 km of reef units. 

Spatial units within 1 km of pour points would be subjected to the strongest inland anthropogenic 

influences by weighing them more heavily than those more than 1 km away. Lastly, we saw 

evidence of dominant drivers varying with hydrology, and that more variance was explained on 

the windward side and the leeward side. Therefore, catchments on the windward side would be 

 

30 

assigned greater weight than catchments on the leeward side. The spatial relationships we 

described would be considered when extrapolating the relative catchment influence in the coastal 

spatial units. Improved understanding of linkages between inland and coastal systems is the basis 

that would enable attribution of information to meaningful spatial units (Wang et al. 2016). As 

coastal and marine spatial planning becomes a growing priority in the United States for 

structured decision making in coastal resource management (Craig 2015), our study identified 

elements that would aid in the development of a framework that allows for the inventory and 

management of spatial information in the coastal environment.  

Conclusion 

The results of our study show that inland landscapes of Maui influence coastal habitat 

conditions, as observed in the amount of variance in reef habitat metrics explained by inland 

variables. In particular, urban and agricultural land cover are the most influential landscape 

disturbances that we tested. Also, the spatial relationships from our findings can be referenced to 

develop an ecologically meaningful spatial framework to characterize linkages between inland 

landscape and the coastal environment. Once established, spatial frameworks would enable the 

creation of spatial units that allow for the attribution and organization of information in 

consistent, comparable manners. Our study in Maui demonstrated the potential value of 

extending the application of the landscape approach from inland systems to the coastal 

environment, to aid in identifying areas of coastal reef habitats that are threatened by inland 

disturbances. The differential outcomes in the role of hydrology affecting the strength of 

relationship between inland characteristics and the various coastal habitat metrics suggest that 

further research on the relationships between the landscape and the coastal environment is still 

needed to fully characterize how inland influences affect coastal habitat across large spatial 

 

31 

extents. In particular, different regions and other coastal habitat types of the world may have 

similar or dissimilar trends. This could be due differential levels of sensitivity of the coastal 

habitat of interest, as well as different environmental conditions such as amount of mean annual 

rainfall received in proximate catchments. Our study improves understanding of spatial 

relationships between inland characteristics and coastal habitats in Maui and contributes to the 

development of a potential spatial framework in the coastal environment by highlighting the 

factors that could facilitate linkages between inland and coastal systems, and characterizing how 

that relationship may alter with proximity. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

32 

APPENDICES

 

33 

 

 

 

APPENDIX 1.A: 

Tables 

 

 

34 

Table A1.1. Code, units, date, and source for inland natural variables. 
Variable 

Code 

Date  Source 

Units 
km2 

% 

mm/yr 
 

2005  Tingley et al., in prep 
2005  Tingley et al., in prep  

2015  Frazier et al. 2015 
1995  SSURGO2 
1995  SSURGO2 

Catchment area 

Catchment slope 

Mean annual rainfall  
Average hydrological soil grouping1 
Minimum hydrological soil grouping1 

Area 

Slope 

Rainfall 
Soil 

Soil_min 

  

 

 
 

 

1. On a scale of 1 to 4, 1 being the lowest infiltration rate 
2. SSURGO (Soil Survey Geographic Database, 

http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_053627) 

 

 

35 

Table A1.2. Code, units, date, and source for anthropogenic variables. 

Variable 

Code 

Units 

Date 

Source 

Urban landscape influences 

Developed impervious surface 

Developed (high intensity) 

Developed (medium intensity) 

Developed (low intensity) 

Developed (open) 

Developed (all) 

Utility pipeline length density 

Percent golf course surface cover 

Population density 

Road length density 

Agricultural landscape influences 

Cultivated crops 

 

 

Imperv  % 

High 

Med 

Low 

Open 

Urban 

PipesL 

Golf 

Pop 

RoadL 

 

Crops 

% 

% 

% 

% 

% 
m/km2 

%  
#/km2 
km/km2 

 

% 

Pasture/hay 
Agriculture (combined cultivated crops 
and pasture/hay) 
Former plantations 

Pasture  % 

Ag 

 

% 

 

 

2005 

2005 

2005 

2005 

2005 

2005 

1983 

1993 

2010 

2014 

 

2005 

2005 

 
CCAP1 
CCAP1 
CCAP1 
CCAP1 
CCAP1 
CCAP1 
Hawaii OP2 
Hawaii OP2 
US Census3 
US Census3 

 
CCAP1 
CCAP1 

2005 

CCAP1 

 

 

Percent of surface that was once used 
for pineapple and/or sugarcane 
production 

Ex_ag 

% 

1989 

Hawaii OP2 

Point source disturbances 

 

 

 

 

Comprehensive Environmental 
Response, Compensation, and Liability 
Information System site density 
Permit Compliance System site density   PCS 
National Pollutant Discharge 
Elimination System (NPDES) site 
density 
Underground injection well density  

UIC 

CERLIS 

NPDES 

Toxic release inventory site density  

All EPA site density 

Stream fragmentation 

Road crossing density 

Ditch intersection density 

Dam density 

Ditch length density 

Ditch length density 

TRI 

EPA 

 

RoadsX 

DitchX 

Dams 

 

DitchL 

#/km2 

2014 

EPA4 

#/km2 

2014 

EPA4 

#/km2 

2014 

EPA4 

#/km2 
#/km2 
#/km3 

 
#/km2 
#/km2 
#/km2 

 
m/km2 

2010 

2014 

2014 

 

2014 

2004 

2010 

 

2004 

Hawaii DOH5 
EPA4 
EPA4 

 
US Census3 
Hawaii DAR6 
ACOE7 

 
Hawaii DAR6 

 
 
 

 

 

 

 

 

36 

Table A1.2. (cont’d) 

303D listed streams 

Percent of upstream network classified 
as 303D stream with measured TMDL 

 

303d 

 

% 

 

 

2012 

EPA4 

1.  CCAP (Coastal Change Analysis Program, 

http://www.csc.noaa.gov/digitalcoast/data/ccapregional/)  

2.  Hawaii OP (Hawaii Office of Planning http://planning.hawaii.gov/) 
3.  TIGER US Census (Topologically Integrated Geographic Encoding and Referencing, 

www.census.gov/geo/maps-data/data/tiger-line.html)  

4.  EPA (US Environmental Protection Agency, http://health.hawaii.gov/cwb/site-

map/clean-water-branch-home-page/integrated-report-and-total-maximum-daily-loads/) 

5.  Hawaii DOH (Hawaii State Department of Health, http://hawaii.gov/health/)  
6.  Hawaii DAR (State of Hawaii Department of Aquatic Resources, 

http://dlnr.hawaii.gov/dar/)  

7.  ACOE (US Army Corps of Engineers, http://www.usace.army.mil/)  

 

37 

 

 

 

Table A1.3. Code, units, date, and source for all compiled coastal variables. 
Variable 

Units 

Code 

Date 

Source 

Spatial variables 

Distance from shore 

 
Dist_shore 

Distance from pour point 

Dist_pp 

Bathymetric depth 

Physical characteristics 

Water temperature 

Wave significant height     
(annual average) 

Wave significant height      
(wet season average) 

Depth 
 

Temp 

Ann_ssh 

Wet_ssh 

Wave significant height       
(dry season average) 

Dry_ssh 

Wave energy period        
(annual average) 

Wave energy period         
(wet season average) 

Ann_wep 

Wet_wep 

Wave energy period            
(dry season average) 

Dry_wep 

 
m 

m 

m 
 

°C 

m 

m 

m 

s 

s 

s 

 
 

 

 
 

 

 
calculated 

calculated 

TDS; NOAA 
 

TDS; NOAA 

2011 

NREL1 

2011 

NREL1 

2011 

NREL1 

2011 

NREL1 

2011 

NREL1 

2011 

NREL1 

Wave power density      
(annual average) 

Wave power density         
(wet season average) 

Wave power density         
(dry season average) 

Ann_wef 

kW/m 

2011 

NREL1 

Wet_wef 

kW/m 

2011 

NREL1 

Dry_wef 

kW/m 

2011 

NREL1 

1.  NREL (National Renewable Energy Laboratory, http://www.nrel.gov/gis/data_mhk.html) 

 

38 

 

 

 

Table A1.4. Minimum (min.), maximum (max.), mean, and standard deviation (SD) of inland 
natural variables across all (n = 182), windward side (n = 100), and leeward side (n = 82) pour 
point catchments draining the Maui landscape. Variable descriptions are included in the Methods 
section.  
Variable 
(units) 
All (n = 182) 
Min. 
Max. 
Mean 
SD 

279.63 
8403.95 
2497.48 
1863.63 

0.27 
31.20 
14.77 
7.01 

1.00 
3.09 
1.60 
0.61 

1.00 
3.65 
2.11 
0.58 

Soil12        
.     . 

Soil_min12     
 

* 

Rainfall1 
(mm/yr) 

Area1      
(km2) 

Slope1     

0.08 

160.29 

7.64 
14.30 

 

 
 
 
 

 
 
 
 

(%) 

 

 
 
 
 

Windward (n = 100) 
Min. 

 

Max. 

Mean 
SD 

 
 
 
Leeward (n = 82) 
Min. 
 
Max. 
 
Mean 
 
SD 
 

0.22 

52.40 
6.57 
8.58 

0.08 

160.29 

8.94 
19.07 

3.19 

23.36 
11.54 
4.63 

0.27 
31.20 
18.71 
7.41 

592.62 

8403.95 
3175.64 
1925.65 

279.63 
7009.57 
1670.46 
1403.06 

1.16 

3.58 
2.10 
0.50 

1.00 
3.65 
2.12 
0.67 

1.00 

3.00 
1.64 
0.57 

1.00 
3.09 
1.54 
0.65 

 

 
 
 

 
 
 
 

 

 
 
 

 
 
 
 

 
1. Indicates variables retained for further analysis. 
2. On a scale of 1 to 4, 1 being the lowest infiltration rate 

 

39 

 

 

 

 

Table A1.5. Units, minimum (min.), maximum (max.), mean, and standard deviation (SD) values 
of inland anthropogenic disturbance variables across all pour point catchments (n = 182) draining 
the Maui landscape.  

Variable 

Urban landscape influences 

Developed impervious surface* 

Developed (high intensity)* 

Developed (medium intensity)* 

Developed (low intensity)* 

Developed (open)* 

Developed (all)* 

Utility pipeline length density* 
Percent golf course surface cover 
Population density* 

Road length density* 

Agricultural landscape influences 

Cultivated crops* 

Pasture/hay* 

Agriculture (combined Cultivated crops & 
Pasture/hay)* 

Former plantations 

Percent of surface that was once used for 
pineapple and/or sugarcane production 

Point source disturbances 

Comprehensive Environmental Response, 
Compensation, and Liability Information 
System site density 

Units 

Min.  Max. 

Mean 

SD 

 

% 

% 

% 

% 

% 

 

 

0.03 

13.93 

0.00 

4.68 

0.00 

8.04 

0.00 

15.07 

0.00 

23.30 

0.00 

% 
m/km2 
0.00 
0.00 
% 
#/km2 
0.00 
km/km2  0.00 

40.07 

1066.3 
24.42 
685.12 

 

2.10 

0.15 

0.52 

1.59 

2.23 

4.50 

54.00 
0.35 
34.33 

 

502.92 

2.68 

0.58 

1.22 

2.63 

4.13 

2.13 
2.13 
166.36 

42857 

5010.51 

88.38 

 

% 

% 

% 

 

% 

 

 

 

0.00 

81.73 

0.00 

56.12 

 

5.08 

4.40 

 

7.58 

12.19 

0.00 

84.02 

9.49 

8.45 

 

 

 

 

0.00 

77.75 

1.42 

15.10 

 

 

 

 

#/km2 

0.00 

0.26 

0.00 

0.45 

Permit Compliance System site density  

#/km2 

0.00 

2.63 

0.07 

0.13 

0.01 

#/km2 

0.00 

3.24 

#/km2 
#/km2 
#/km3 
 
#/km2 
#/km2 
#/km2 
 
m/km2 
 

0.00 

1.00 

0.00  <0.00 

<0.00 

0.00 
 
0.00 
0.00 
0.00 
 
0.00 
 

5.4 
 
5.23 
1.99 
1.38 
 
2871.80 
 

0.2 
 
0.96 
0.14 
0.02 
 
311.45 
 

National Pollutant Discharge Elimination 
System (NPDES) site density 

Underground injection well density 

Toxic release inventory site density  

All EPA site density 

Stream fragmentation 

Road crossing density* 
Ditch intersection density 
Dam density 

Ditch length density 

Ditch length density* 

303D listed streams 

0.00 

0.30 

0.32 

0.13 

0.02 
 
0.93 
0.32 
0.07 
 
486.04 
 

Percent of upstream network classified as 
303D stream with measured TMDL 

*Indicates variables retained for further analysis.

% 

0.00 

100 

5.36 

0.73 

 

40 

Ann_wef 
(kW/m)* 
 

Wet_wef 
(kW/m)* 
 

Dry_wef 
(kW/m)* 
 

0.40 

20.1 

7.41 

5.68 

3.80 

20.10 

11.53 

4.48 

 

 

0.35 

0.53 

30.08 

11.38 

9.95 

8.26 

4.84 

3.24 

 

 

5.32 

30.08 

16.38 

6.51 

 

 

2.23 

10.13 

6.65 

2.58 

0.40 

0.35 

0.53 

17.20 

23.02 

11.38 

4.14 

4.24 

4.86 

5.50 

3.40 

3.00 

Table A1.6. Minimum (min.), maximum (max.), mean, and standard deviation (SD) of coastal habitat metrics of the reef units. 
Variable descriptions are included in the Methods section. 
Variable 
(units) 

Wet_ssh      
(m) *   * 

Depth     
(m)* 

Temp  
(°C) 

Dry_ssh  

Ann_wep     

Wet_wep 

Dist_shore    

Dist_pp  

Ann_ssh 

(m)* 

Dry_wep 

(s)* 

(m)* 

(m)* 

(s)* 

(s)* 

(m)* 

All (n = 52) 

Min. 

Max. 

Mean 

SD 

23.55 

156.42 

6.16 

24.83 

2230.71 

8267.57 

20.27 

26.86 

373.68 

367.08 

1385.52 

14.16 

25.75 

1721.45 

2.60 

0.42 

Windward (n = 23) 

Min. 

Max. 

Mean 

SD 

121.71 

322.48 

12.42 

2230.71 

2430.54 

20.27 

431.15 

481.86 

858.84 

15.13 

604.45 

1.66 

24.97 

26.07 

25.47 

0.25 

Leeward (n = 29) 

Min. 

Max. 

Mean 

SD 

23.55 

988.60 

328.10 

241.12 

156.42 

6.16 

24.83 

8267.57 

16.64 

26.86 

1803.23 

13.40 

25.98 

2168.40 

2.96 

0.39 

*Indicates variables retained for further analysis. 
 

0.28 

2.02 

1.14 

0.57 

0.86 

2.02 

1.53 

0.35 

0.28 

1.99 

0.83 

0.52 

0.26 

2.44 

1.28 

0.69 

1.00 

2.44 

1.79 

0.41 

0.26 

2.26 

0.87 

0.59 

0.3 

1.72 

1.00 

0.46 

0.72 

1.61 

1.27 

0.30 

0.30 

1.72 

0.78 

0.44 

 

 

 

7.43 

9.68 

8.60 

0.63 

8.50 

9.68 

8.85 

0.31 

7.43 

9.63 

8.40 

0.75 

7.32 

10.42 

8.42 

0.94 

7.69 

8.49 

7.92 

0.20 

7.32 

10.42 

8.83 

1.10 

 

 

 

6.84 

11.22 

8.25 

1.58 

6.84 

7.30 

6.98 

0.12 

7.20 

11.22 

9.25 

1.48 

 

41 

 

C     
(%) 

SC   
(%) 

Table A1.7. Minimum (min.), maximum (max.), mean, and standard deviation (SD) of coastal 
habitat metrics of the reef units. Variable descriptions are included in the Methods section. 
Variable 
(units) 
All (n = 52) 
Min. 
Max. 
Mean 
SD 

0.00000 
0.00066 
0.00008 
0.00013 

0.00000 
0.15273 
0.00461 
0.02195 

0.05 
87.50 
30.12 
26.03 

H*     
* 

COTS  
(#/m2) 

S      

(%) 

MA  
(%) 

CCA 
(%) 

0.00 
27.25 
4.55 
5.40 

0.40 
60.00 
14.05 
16.03 

0.00 
13.75 
1.46 
2.72 

U      
(#/m2) 

0.00 
53.38 
11.17 
15.54 

1.00 
4.60 
2.53 
0.82 

Windward (n = 23) 
Min. 

0.70 

Max. 

Mean 
SD 

56.75 
10.97 
13.68 

Leeward (n = 29) 
Min. 
Max. 
Mean 
SD 

0.40 
60.00 
16.49 
17.52 

0.00 

8.55 
1.67 
2.51 

0.00 
13.75 
1.30 
2.91 

0.05 

66.50 
16.19 
18.48 

6.80 
87.50 
41.17 
26.07 

0.45 

50.42 
10.29 
13.99 

0.00 
53.38 
11.86 
16.88 

0.00 

15.50 
4.25 
4.75 

0.00 
27.25 
4.79 
5.95 

1.10 

4.60 
2.89 
0.74 

1.00 
3.80 
2.25 
0.78 

0.00000 

0.00000 

0.00066 
0.00014 
0.00016 

0.00000 
0.00022 
0.00003 
0.00006 

0.00221 
0.00014 
0.00046 

0.00000 
0.15273 
0.00815 
0.02912 

*On a scale of 1 to 6, 1 being the least complex and 6 being the most complex  
 
 

 

42 

Table A1.8 Results of the PCA of 9 coastal wave variables. PC1 shows the weight of each 
variable, and it explains 86.5% of the variation in wave variables. Variable descriptions are 
included in Table A1.3. 

Variable 
Ann_ssh 
Wet_ssh 
Dry_ssh 
Ann_wef 
Wet_wef 
Dry_wef 
Wet_wep 
Ann_wep 
Dry_wep 

PC1 
0.993 
0.991 
0.989 
0.988 
0.985 
0.984  
-0.417 
-0.934 
-0.937 

 

 

 

 

43 

Table A1.9. Percent of total explained variance by RDA predicting habitat variables from the 3 
groups of variables (inland natural, inland anthropogenic, and coastal). Results include unique 
contributions from each group as well as their shared variances. Buffer widths indicate the 
distance between reef units and pour points, and inland factors were summarized for all pour 
point catchments within specified distances.  

Variable Groups 

1km buffer 

3km buffer 

5km buffer 

10km buffer 

Inland anthropogenic (A) 
Inland natural (N) 
Coastal (C)  
A + N 
N + C 
A + C 
A + N + C 
Total explained 
*Negative values can occur when random normal variables explain more variation than 
explanatory variables, and are treated as zeros for interpretation (Legendre 2008).  

-0.8* 
2.9 
11.9 
2.0 
15.2 
11 
-9.8* 

0.2 
5.6 
26.0 
1.2 
-0.8* 
2.1 
5.3 

2.2 
3.0 
25.9 
0.6 
3.8 
3.0 
0.3 

38.7 

0.9 
3.7 
21.5 
3.2 
2.4 
7.0 
1.6 

40.2 

39.6 

32.4 

 

 

44 

Table A1.10. Stepwise multiple linear regression results of best fit model characterizing relationships between environmental 
variables (Tables A1.1, A1.2, A1.3) and coastal habitat metrics (Table A1.7). Standardized β is shown in table, and predictor 
significance is noted for p<0.05 (bolded) and p<0.1 (italicized). Results shown are for reef units across the whole island (A), on the 
windward side (W), and on the leeward side (L) of Maui, Hawaii. 

Adj. 
R2 

0.543 
0.485 
0.438 
0.225 
0.679 
0.377 

- 

0.183 

- 

0.443 
0.222 
0.305 
0.224 
0.386 
0.320 

- 

0.285 

- 

0.245 
0.444 
0.248 
0.095 
0.175 

- 
 
 
 

Metric 

Extent 

A 
W 
L 
A 
W 
L 
A 
W 
L 
A 
W 
L 
A 
W 
L 
A 
W 
L 
A 
W 
L 
A 
W 
L 
A 
W 
L 

H 

C 

SC 

S 

MA 

CCA 

 

COTS 

 
 
U 
 

Count 

of metrics 
affected 

 

 

 

Inland anthropogenic variables 

Inland natural variables 

Coastal variables 

Urban 

Ag 

DitchL  RoadX 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0 
0 
0 

 
 
 
 

0.79 
-0.67 

 
 
 
 
 
 

0.50 
0.62 

 
 
 
 

0.37 
0.63 

 
 

0.46 

 
2 
4 
1 

 
 

-0.68 

 
 
 
 
 
 
 
 

0.58 

 
 

0.59 

 
 
 
 
 
 
 
 
 
0 
0 
3 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0 
0 
0 

Soil  Rainfall  Area 
0.24 

-0.52 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0 
0 
0 

 
 
 
 

-0.81 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

-0.34 

 
 
2 
0 
1 

 
 
 
 
 
 

-0.47 

 
 
 
 

0.23 
0.30 

 
 
 
 
 
 
 
 
 
 
2 
2 
0 

45 

Slope  Dist_shore  Dist_pp  Depth  Ann_wef 

 
 
 
 

0.46 

 
 
 
 
 
 
 
 
 
 
 

0.38 

 
 
 
 
 
 
 
0 
2 
0 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0 
0 
0 

 
 
 

0.37 
0.28 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1 
1 
0 

-0.47 
-0.31 

 

-0.45 
-0.49 

 
 
 
 

0.38 

 
 
 
 
 
 

-0.48 

 
 

-0.53 

 
 
 
 
3 
4 
0 

1.26 
0.58 

 
 
 
 
 
 
 

-0.82 
-0.51 

 
 
 
 
 
 
 

0.50 

 

0.53 

 
 
 
3 
2 
1 

 

 

APPENDIX 1.B: 

Figures 

46 

   

 

Figure B1.1. Location of Maui in the Hawaii archipelago (A), and location of 52 reef units and 
streams draining into the coastal environment around Maui (B). 
 
 
 
 
 
 
 

 

47 

 
 
 
 
 

 

 

Pour point 
catchment 

  Local 
catchment 

Coastal 

environment 

Figure B1.2. The local and network catchments of a river reach draining at a pour point into the 
coastal environment. 

48 

 

 

 

APPENDIX 1.C: 

Supplemental tables 

 

49 

Table C1.1. Units, minimum (min.), maximum (max.), mean, and standard deviation (SD) values 
of inland anthropogenic disturbance variables across windward side pour point catchments (n = 
100) draining the Maui landscape.  

Units 

Min.  Max. 

Mean 

SD 

 

% 

% 

% 

% 

% 

 

0.08 

0.00 

0.00 

0.03 

0.00 

 

11.06 

4.50 

7.09 

15.07 

23.30 

0.04 

0.00 
0.00 

0.00 

40.07 

% 
m/km2 
% 
#/km2 
611.90 
km/km2  69.20  42856 
 

594.95 
24.42 

 

 

0.05 

0.06 

81.73 

28.82 

0.11 
 

84.02 
 

 

2.59 

0.14 

0.52 

2.10 

3.18 

5.95 

32.10 
0.47 

37.71 

5839 
 

6.86 

3.57 

10.43 
 

 

2.74 

0.54 

1.07 

3.07 

5.13 

8.84 

97.47 
2.74 

82.85 

8117 
 

14.51 

5.41 

16.24 
 

% 

% 

% 

 

% 

 

0.00 

0.00 

0.00 

0.00 

 

 

 

 

#/km2 

0.00 

0.00 

0.00 

0.00 

#/km2 

#/km2 

#/km2 
#/km2 
#/km3 
 
#/km2 
#/km2 
#/km2 
 
m/km2 
 

0.00 

2.63 

0.07 

0.32 

0.00 

0.00 

0.00 

0.00 
 
0.00 
0.00 
0.00 
 
0.00 
 

2.99 

0.00 

0.00 

5.26 
 
4.86 
1.99 
1.38 
 
2295.25 
 

0.11 

0.00 

0.00 

0.18 
 
0.98 
0.17 
0.03 
 
391.74 
 

0.43 

0.00 

0.00 

0.74 
 
0.92 
0.36 
0.15 
 
469.41 
 

% 

0.00 

100.00 

6.87 

24.90 

Variable 

Urban landscape influences 

Developed impervious surface 

Developed (high intensity) 

Developed (medium intensity) 

Developed (low intensity) 

Developed (open) 

Developed (all) 

Utility pipeline length density 
Percent golf course surface cover 
Population density 

Road length density 

Agricultural landscape influences 

Cultivated crops 

Pasture/hay 

Agriculture (combined Cultivated crops & 
Pasture/hay) 

Former plantations 

Percent of surface that was once used for 
pineapple and/or sugarcane production 

Point source disturbances 

Comprehensive Environmental Response, 
Compensation, and Liability Information 
System site density 

Permit Compliance System site density  

National Pollutant Discharge Elimination 
System (NPDES) site density 

Underground injection well density 

Toxic release inventory site density  

All EPA site density 

Stream fragmentation 

Road crossing density 
Ditch intersection density 
Dam density 

Ditch length density 

Ditch length density 

303D listed streams 

Percent of upstream network classified as 
303D stream with measured TMDL 

 

50 

Table C1.2. Units, minimum (min.), maximum (max.), mean, and standard deviation (SD) values 
of inland anthropogenic disturbance variables across leeward side pour point catchments (n = 82) 
draining the Maui landscape.  

Units 

Min.  Max. 

Mean 

SD 

 

% 

% 

% 

% 

% 

 

 

0.03 

13.93 

0.00 

4.68 

0.00 

8.04 

0.00 

11.63 

0.00 

9.86 

30.61 

1066.27 
7.32 

685.12 

0.00 

0.00 
0.00 

% 
m/km2 
% 
#/km2 
0.00 
km/km2  0.00 
 

 

1.50 

0.17 

0.53 

0.97 

1.07 

2.74 

80.70 
0.20 

30.20 

 

2.48 

0.64 

1.40 

1.80 

1.87 

5.20 

221.18 
0.95 

95.05 

38680.79 
 

3999.24 
 

6247.97 
 

 

% 

% 

% 

 

% 

 

0.00 

47.31 

0.00 

56.12 

0.00 

56.53 

 

 

2.91 

5.42 

8.33 

 

8.11 

11.04 

13.59 

 

0.00 

77.75 

3.16 

11.58 

 

 

 

 

#/km2 

0.00 

0.26 

0.00 

0.03 

#/km2 

#/km2 

#/km2 
#/km2 
#/km3 
 
#/km2 
#/km2 
#/km2 
 
m/km2 
 

0.00 

2.16 

0.07 

0.27 

0.00 

3.24 

0.00 

1.00 

0.00 

0.00 

0.00 
 
0.00 
0.00 
0.00 
 
0.00 
 

5.40 
 
5.23 
1.56 
0.81 
 
2871.80 
 

0.16 

0.01 

0.00 

0.23 
 
0.94 
0.10 
0.01 
 
213.53 
 

0.47 

0.11 

0.00 

0.73 
 
0.95 
0.26 
0.09 
 
490.83 
 

Variable 

Urban landscape influences 

Developed impervious surface 

Developed (high intensity) 

Developed (medium intensity) 

Developed (low intensity) 

Developed (open) 

Developed (all) 

Utility pipeline length density 
Percent golf course surface cover 
Population density 

Road length density 

Agricultural landscape influences 

Cultivated crops 

Pasture/hay 

Agriculture (combined Cultivated crops & 
Pasture/hay) 

Former plantations 

Percent of surface that was once used for 
pineapple and/or sugarcane production 

Point source disturbances 

Comprehensive Environmental Response, 
Compensation, and Liability Information 
System site density 

Permit Compliance System site density  

National Pollutant Discharge Elimination 
System (NPDES) site density 

Underground injection well density 

Toxic release inventory site density  

All EPA site density 

Stream fragmentation 

Road crossing density 
Ditch intersection density 
Dam density 

Ditch length density 

Ditch length density 

303D listed streams 

Percent of upstream network classified as 
303D stream with measured TMDL 

% 

0.00 

100.00 

3.51 

17.54 

 

51 

Table C1.3. Pearson pairwise correlations among inland natural variables. Bolded variables were 
included in the RDA. Variable descriptions are included in Table A1.1. 

  

Area 

Slope 

Rainfall 

Soil 
Soil_min 

Area 

1.00 
-0.09 
0.13 
-0.01 
-0.06 

Slope  Rainfall 

Soil 

Soil_min 

 
1.00 
0.07 
-0.38 
-0.52 

 

 
1.00 
-0.27 
-0.12 

 

 

 
1.00 
0.68 

 

 

 

 
1.00 

 

52 

 
 
 
 

 

Table C1.4. Pearson pairwise correlations of inland anthropogenic variables. The bolded variables are included in the RDA. Variable 
descriptions are included in Table A1.2. 
  

Imperv  High  Med 
 

 

Low  Open  Urban  Crops 
 

 

 

 

Pasture 
 

-0.02 

-0.01 

-0.06 

-0.04 

-0.04 

-0.05 

0.22 

-0.02 

0.11 

0.21 

0.17 

0.18 

Ag 
 

Pop  RoadL  DitchL  PipesL 
 

 

 

 

RoadX 
 

 

 

 

 

 

1.00 

0.23 

0.82 

0.51 

0.52 

0.46 

0.01 

0.20 

 

 

 

 

 

 

1.00 

0.73 

0.33 

0.22 

0.06 

 

 

 

 

 

 

 

1.00 

0.50 

0.45 

0.33 

 

 

 

 

 

 

 

 

1.00 

0.44 

0.35 

-0.08 

-0.05 

-0.22 

0.11 

0.19 

0.05 

 

 

 

 

 

 

 

 

 

1.00 

0.56 

0.17 

0.30 

 

 

 

 

 

 

 

 

 

 

1.00 

0.35 

0.11 

 

 

 

 

 

 

 

 

 

 

 

1.00 

0.11 

 

 

 

 

 

 

 

 

 

 

 

 

1.00 

53 

1.00 

0.68 

0.88 

0.94 

0.87 

0.95 

0.62 

0.37 

0.59 

0.78 

0.57 

0.39 

1.00 

0.84 

0.52 

0.48 

0.61 

0.48 

0.17 

0.42 

0.67 

0.42 

0.27 

 

1.00 

0.82 

0.74 

0.86 

0.55 

0.34 

0.54 

0.82 

0.50 

0.31 

 

 

1.00 

0.94 

0.98 

0.56 

0.42 

0.56 

0.75 

0.53 

0.38 

 

 

 

1.00 

0.97 

0.63 

0.44 

0.64 

0.70 

0.51 

0.41 

 

 

 

 

1.00 

0.62 

0.43 

0.61 

0.77 

0.54 

0.40 

Imperv 

High 

Med 

Low 

Open 

Urban 

Crops 

Pasture 

Ag 

Pop 

RoadL 

DitchL 

PipesL 

RoadX 

 
 
 
 
 

 

 

Table C1.5. Pearson pairwise correlations among coastal variables included in the RDA. 
Variable descriptions are included in Table A1.3. 

  

Dist_shore 

Dist_pp 

Depth  Ann_wef 

Dist_shore 

Dist_pp 

Depth 

Ann_wef 

1.00 

0.34 

0.26 

-0.13 

 

 

 

 

 

 

1.00 

0.53 

1.00 

0.23 

-0.13 

1.00 

 
 

 

 

 

 

 

54 

LITERATURE CITED

 

55 

LITERATURE CITED 

Alexander, R. B., R. A. Smith, E. Gregory, E. W. Boyer, J. V. Nolan, and W. Brakebill. 2008. 

Differences in phosphorus and nitrogen delivery to the Gulf of Mexico from the Mississippi 
River Basin 42(3):822–830. 

Allan, D., and M. Castillo. 2007. Stream ecology: Structure and function of running waters, 2nd 

edition. Springer. 

Allan, J. D. 2004. Landscapes and riverscapes: The influence of land use on stream ecosystems. 

Annual Review of Ecology, Evolution, and Systematics 35:257–284. 

Babcock, R. C., J. M. Dambacher, E. B. Morello, É. E. Plagányi, K. R. Hayes, H. P. A. 

Sweatman, and M. S. Pratchett. 2016. Assessing different causes of crown-of-thorns starfish 
outbreaks and appropriate responses for management on the Great Barrier Reef. PLoS ONE 
11(12):1–20. 

Bittler, K. M., L. P. Scheef, and E. J. Buskey. 2014. Freshwater inflows and blue crabs: The 
influence of salinity on selective tidal stream transport. Marine Ecology Progress Series 
514:137–148. 

Boesch, D. F. 1996. Science and management in four US coastal ecosystems dominated by land-

ocean interactions. Journal of Coastal Conservation 2(1):103–114. 

Borcard, D., P. Legendre, and P. Drapeau. 1992. Partialling out the spatial component of 

ecological variation. Ecology 73(3):1045–1055. 

Buma, B. 2015. Disturbance interactions: characterization, prediction, and the potential for 

cascading effects. Ecosphere 6(4):1–15. 

Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: A 

practical information-theoretic approach, 2nd edition. Springer-Verlag New York, Inc., 
New York. 

Chabanet, P., M. Adjeroud, S. Andréfouët, Y.-M. Bozec, J. Ferraris, J.-A. Garcìa-Charton, and 

M. Schrimm. 2005. Human-induced physical disturbances and their indicators on coral reef 
habitats: A multi-scale approach. Aquatic Living Resources 18:215–230. 

Cooper, A. R., D. M. Infante, K. E. Wehrly, L. Wang, and T. O. Brenden. 2016. Identifying 
indicators and quantifying large-scale effects of dams on fishes. Ecological Indicators 
61:646–657. 

Craig, R. 2015. An historical look at planning for the federal public lands: Adding marine spatial 

planning offshore. George Washington Journal of Energy & Environmental Law:1–20. 

 

56 

Crain, C. M., B. S. Halpern, M. W. Beck, and C. V. Kappel. 2009. Understanding and managing 

human threats to the coastal marine environment. Annals of the New York Academy of 
Sciences 1162:39–62. 

Creel, L. 2003. Ripple effects: Population and coastal regions. Page Population Reference 

Bureau. 

Danz, N. P., G. J. Niemi, R. R. Regal, T. Hollenhorst, L. B. Johnson, J. M. Hanowski, R. P. 

Axler, J. J. H. Ciborowski, T. Hrabik, V. J. Brady, J. R. Kelly, J. a. Morrice, J. C. Brazner, 
R. W. Howe, C. a. Johnston, and G. E. Host. 2007. Integrated measures of anthropogenic 
stress in the U.S. Great Lakes Basin. Environmental Management 39(5):631–647. 

DLNR DAR (Department of Land and Natural Resources, Department of Aquatic Resources). 

2005. Marine Protected Areas in Hawai’i. Honolulu, Hawaii. 

Dollar, S. J. 1982. Wave stress and coral community structure in Hawaii. Coral Reefs 1:71–81. 

Done, T. J. 1992. Phase-shifts in coral reef communities and their ecological significance. 

Hydrobiologia 247:121–132. 

Esselman, P. C., D. M. Infante, L. Z. Wang, D. Y. Wu, A. R. Cooper, and W. W. Taylor, 2011. 

An index of cumulative disturbance to river fish habitats of the conterminous United 
States from landscape anthropogenic activities. Ecological Restoration 29:133-151. 

Fabricius, K. E. 2005. Effects of terrestrial runoff on the ecology of corals and coral reefs: 

Review and synthesis. Marine Pollution Bulletin 50:125–146. 

Fonseca, V. P., M. G. Pennino, M. F. de Nóbrega, J. E. L. Oliveira, and L. de Figueiredo 

Mendes. 2017. Identifying fish diversity hot-spots in data-poor situations. Marine 
Environmental Research 129:365–373. 

Frazier, A. G., T. W. Giambelluca, H. F. Diaz, and H. L. Needham. 2016. Comparison of 
geostatistical approaches to spatially interpolate month-year rainfall for the Hawaiian 
Islands. International Journal of Climatology 36(3):1459–1470. 

Friedlander, A. M., E. Brown, and M. E. Monaco. 2007. Defining reef fish habitat utilization 

patterns in Hawaii: Comparisons between marine protected areas and areas open to fishing. 
Marine Ecology Progress Series 351:221–233. 

Friedlander, A. M., M. K. Donovan, K. A. Stamoulis, I. D. Williams, E. K. Brown, E. J. Conklin, 

E. E. DeMartini, K. S. Rodgers, R. T. Sparks, and W. J. Walsh. 2017. Human‐induced 
gradients of reef fish declines in the Hawaiian Archipelago viewed through the lens of 
traditional management boundaries. Aquatic Conservation: Marine and Freshwater 
Ecosystems:1–12. 

 

57 

Friedlander, A. M., and J. D. Parrish. 1998. Habitat characteristics affecting fish assemblages on 

a Hawaiian coral reef. Journal of Experimental Marine Biology and Ecology 224:1–30. 

Giambelluca, T. W., J. K. DeLay, M. A. Nullet, M. A. Scholl, and S. B. Gingerich. 2011. 

Canopy water balance of windward and leeward Hawaiian cloud forests on Haleakalā, 
Maui, Hawai’i. Hydrological Processes 25(3):438–447. 

Gorham, E. 1996. Linkages among diverse aquatic ecosystems: A neglected field of study. Pages 

203–217Freshwater Ecosystems: Revitalizing Educational Programs in Limnology. 
National Research Council, National Academy Press, Washington, DC. 

Gourlay, M. R. 2011. Wave Set-Up. Springer. 

Grigg, R. W. 1995. Coral reefs in an urban embayment in Hawaii: a complex case history 

controlled by natural and anthropogenic stress. Coral Reefs 14(4):253–266. 

Grigg, R. W. 1983. Community structure, succession and development of coral reefs in Hawaii. 

Marine Ecology Progress Series 11:1–14. 

Halpern, B. S., S. Walbridge, K. a Selkoe, C. V Kappel, F. Micheli, C. D’Agrosa, J. F. Bruno, K. 
S. Casey, C. Ebert, H. E. Fox, R. Fujita, D. Heinemann, H. S. Lenihan, E. M. P. Madin, M. 
T. Perry, E. R. Selig, M. Spalding, R. Steneck, and R. Watson. 2008. A global map of 
human impact on marine ecosystems. Science 319(5865):948–952. 

Johnson, L. B., and G. E. Host. 2010. Recent developments in landscape approaches for the 

study of aquatic ecosystems. Journal of the North American Benthological Society 29:41–
66. 

Jokiel, P. L., E. K. Brown, A. Friedlander, S. K. Rodgers, and W. R. Smith. 2004. Hawai’i Coral 

Reef Assessment and Monitoring Program: Spatial patterns and temporal dynamics in reef 
coral communities. Pacific Science 58(2):159–174. 

Jouffray, J. B., M. Nyström, A. Norström, I. Williams, L. Wedding, J. Kittinger, and G. 

Williams. 2015. Human and natural drivers of multiple coral reef regimes across the 
Hawaiian Archipelago. Philosophical Transactions B 370:1–8. 

Kaandorp, J. A., and J. E. Kübler. 2001. The Algorithmic Beauty of Seaweeds, Sponges and 

Corals. Berlin: Springer Verlag. 

Kennedy, D. M., and J. L. D. Woods. 2012. The influence of coarse woody debris on gravel 

beach geomorphology. Geomorphology 159–160:106–115. 

Kenyon, J. C., R. E. Brainard, R. K. Hoeke, F. A. Parrish, and C. B. Wilkinson. 2006. Towed-
diver surveys, a method for mesoscale spatial assessment of benthic reef habitat: A case 
study at Midway Atoll in the Hawaiian Archipelago. Coastal Management 34(3):339–349. 

 

58 

Lamberti, G. a., D. T. Chaloner, and A. E. Hershey. 2010. Linkages among aquatic ecosystems. 

Journal of the North American Benthological Society 29(1):245–263. 

Langbein W.B., and S. A. Schumm. 1958. Yield of sediment in relation to mean annual 

precipitation. Eos, Transactions American Geophysical Union 39:1076–84. 

Lau, L. S., and J. F. Mink. 2006. Hydrology of the Hawaiian Islands. University of Hawai‘i 

Press, Honolulu. 

Legendre, P., and L. Legendre. 1988. Numerical ecology: developments in environmental 

modelling, 2nd edition. Elsevier, Amsterdam. 

Leps, J., and P. Smilauer. 2003. Multivariate analysis of ecological data using CANOCO. 

Cambridge University Press, New York. 

Lotze, H. K., H. S. Lenihan, B. J. Bourque, R. H. Bradbury, R. G. Cooke, M. C. Kay, S. M. 
Kidwell, M. X. Kirby, C. H. Peterson, J. B. C. Jackson, and M. Bay. 2006. Depletion, 
degradation, and recovery potential of estuaries and coastal seas. Science 312:1806–1809. 

Mann, K. H. 2000. Ecology of coastal waters: With implications for management, 2nd edition. 

John Wiley & Sons, Dartmouth, Nova Scotia. 

McDowall, R. M. 2003. Hawaiian biogeography and the islands’ freshwater fish fauna. Journal 

of Biogeography 30(5):703–710. 

Michalak, A. M., E. J. Anderson, D. Beletsky, S. Boland, N. S. Bosch, T. B. Bridgeman, J. D. 

Chaf, K. Cho, R. Confesor, I. Dalo, X. Liu, M. R. Mcwilliams, M. R. Moore, D. J. Posselt, 
R. P. Richards, and D. Scavia. 2013. Record-setting algal bloom in Lake Erie caused by 
agricultural and meteorological trends consistent with expected future conditions. 
Proceedings of the National Academy of Sciences of the United States of America 
110(16):6448–6452. 

Milliman, J. D., and R. H. Meade. 1983. World-wide delivery of river sediment to the oceans. 

The Journal of Geology 91(1):1–21. 

Moser, S. C., S. Jeffress Williams, and D. F. Boesch. 2012. Wicked challenges at land’s end: 

Managing coastal vulnerability under climate change. Annual Review of Environment and 
Resources 37:51–78. 

Moura, R. L., G. M. Amado-Filho, F. C. Moraes, P. S. Brasileiro, P. S. Salomon, M. M. 

Mahiques, A. C. Bastos, M. G. Almeida, J. M. Silva, B. F. Araujo, F. P. Brito, T. P. Rangel, 
B. C. V. Oliveira, R. G. Bahia, R. P. Paranhos, R. J. S. Dias, E. Siegle, A. G. Figueiredo, R. 
C. Pereira, C. V. Leal, E. Hajdu, N. E. Asp, G. B. Gregoracci, S. Neumann-Leitao, P. L. 
Yager, R. B. Francini-Filho, A. Froes, M. Campeao, B. S. Silva, A. P. B. Moreira, L. 
Oliveira, A. C. Soares, L. Araujo, N. L. Oliveira, J. B. Teixeira, R. A. B. Valle, C. C. 

 

59 

Thompson, C. E. Rezende, and F. L. Thompson. 2016. An extensive reef system at the 
Amazon River mouth. Science Advances 2(4):e1501252–e1501252. 

Muscatine, L., and J. W. Porter. 1977. Reef corals: Mutualistic symbioses adapted to nutrient-

poor environments. BioScience 27(7):454–460. 

Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, and J. Kent. 2000. 

Biodiversity hotspots for conservation priorities. Nature 403(6772):853–858. 

NOAA (National Oceanic and Atmospheric Administration), Office for Coastal Management. 

2005. C-CAP high resolution land cover state of Hawaii. U.S. Department of Commerce, 
Washington, DC. http://maps.coast.noaa.gov/arcgis/rest/services/CCAP_HR 
/Hawaii_HR_2005/MapServer/ Accessed 10.05.01. 

NOC (National Ocean Council). 2013. National Ocean Policy Implementation Plan, Washington 

DC. 

Oliver, L. M., J. C. Lehrter, and W. S. Fisher. 2011. Relating landscape development intensity to 

coral reef condition in the watersheds of St. Croix, US Virgin Islands. Marine Ecology 
Progress Series 427:293–302. 

Olker, J. H., K. E. Kovalenko, J. J. H. Ciborowski, V. J. Brady, and L. B. Johnson. 2016. 
Watershed land use and local habitat: Implications for habitat assessment. Wetlands 
36:311–321. 

Randall, J. E. 1976. The endemic shore fishes of the Hawaiian Islands, Lord Howe Island and 

Easter Island. Colloque Commerson, 1973. O.R.S.T.O.M. Travaux et Documents 47:49–73. 

Richards, C., L. B. Johnson, and G. E. Host. 1996. Landscape-scale influences on stream habitats 

and biota. Canadian Journal of Fisheries and Aquatic Sciences 53(S1):295–311. 

Rodgers, K. S., P. L. Jokiel, C. E. Bird, and E. K. Brown. 2009. Quantifying the condition of 

Hawaiian coral reefs. Aquatic Conservation: Marine and Freshwater Ecosystems 20(1):93–
105. 

Sala, E., and N. Knowlton. 2006. Global marine biodiversity trends. Annual Review of 

Environment and Resources 31:93–122. 

Small, C., and R. J. Nicholls. 2003. A global analysis of human settlement in coastal zones. 

Journal of Coastal Research 19(3):584–599. 

Storlazzi, C. D., and B. E. Jaffe. 2008. The relative contribution of processes driving variability 

in flow, shear, and turbidity over a fringing coral reef: West Maui, Hawaii. Estuarine, 
Coastal and Shelf Science 77(4):549–564. 

 

60 

Talley, D. M., G. R. Huxel, M. Holyoak, D. M. Talley, G. R. Huxel, and M. Holyoak. 2006. 

Connectivity at the land-water interface. Pages 97–129 in K. R. Crooks and M. Sanjayan, 
editors. Connectivity Conservation. Cambridge University Press. 

Timm, O. E., T. W. Giambelluca, and H. F. Diaz. 2015. Journal of Geophysical Research: 

Atmospheres. Journal of Geophysical Research: Atmospheres 120:92–112. 

Tingley et al. In Review. Influences of natural landscape characteristics on distributions of 

tropical stream organisms: Stream reach classification in the Hawaiian Islands 
Hydrobiologia.  

UNEP (United Nations Environment Programme). 2006. Marine and coastal ecosystems and 

human well-being: A synthesis report based on the findings of the Millennium Ecosystem 
Assessment.  

USDA (U. S. Department of Agriculture, Natural Resources Conservation Service). 1995. Soil 

Survey Geographic (SSURGO) Data Base, Report 1527. U.S. Department of Agriculture, 
Washington, DC. 

Wang, L., D. Infante, P. Esselman, A. Cooper, W. Taylor, D. Beard, G. Whelan, A. Ostroff, L. 
Wang, D. Infante, D. Beard, and G. Whelan. 2011. A hierarchical spatial framework and 
database for the National River Fish Habitat Condition Assessment. Fisheries 36(9):37–41. 

Wang, L., J. Lyons, P. Rasmussen, P. Seelbach, T. Simon, M. Wiley, P. Kanehl, E. Baker, S. 

Niemela, and P. M. Stewart. 2003. Watershed, reach, and riparian influences on stream fish 
assemblages in the Northern Lakes and Forest Ecoregion, U.S.A. Canadian Journal of 
Fisheries and Aquatic Sciences 60(5):491–505. 

Wehrly, K., L. Wang, D. Infante, C. Joseph, A. Cooper, L. Stanfield, and E. S. Rutherford. 2013. 

Landscape change and its influences on aquatic habitats and fisheries in the Great Lakes 
Basin. Pages 81–104. Great Lakes Fisheries Policy and Management: A Binational 
Perspective. 

Wooldridge, S. A. 2009. Water quality and coral bleaching thresholds: Formalising the linkage 

for the inshore reefs of the Great Barrier Reef, Australia. Marine Pollution Bulletin 
58(5):745–751. 

Yamanaga, G. 1972. Evaluation of the streamflow-data program in Hawaii. United States 
Department of the Interior. Geological Survey: Water Resources Division. Honolulu, 
Hawaii. 

 

 

 

 

61 

CHAPTER 2: 

ASSESSING INFLUENCES OF INLAND LANDSCAPE FACTORS ON COASTAL 

HABITATS FOR IMPROVED CONSERVATION 

 

Abstract 

Coastal habitats offer a diversity of ecosystem services, but their condition and 

availability are in decline in part due to anthropogenic disturbances originating from the 

terrestrial landscape. Improved understanding of how inland anthropogenic disturbances affect 

coastal habitats is essential for their conservation. This chapter builds on findings from previous 

work that explores linkages between inland and coastal habitats to assess coastal habitat 

condition and characterize threats to ecosystem services and benefits around the island of Maui, 

Hawaii. We first identified catchments associated with discrete areas of the coastal environment 

(i.e., coastal grid cells) and then assessed the influence of landscape disturbances on grid cells by 

extending a measure of inland disturbances to coastal habitats. We accounted for catchment 

drainage area and proximity of multiple catchments to estimate the risk of degradation to coastal 

habitats, steps that have not been achieved in other studies. We next evaluated locations of 

ecosystem services and benefits provided by coastal habitats and assessed how they may be 

affected by the inland disturbances. Our findings show that a majority of coastal ecosystem 

services and benefits around Maui, including recreation, fishing/fisheries, and habitats for 

biodiversity are at risk of degradation from inland disturbances. We assessed relative risk of 

degradation of coastal areas and identified locations of high and low risks that could be used by 

resource managers for development of conservation actions. The spatially-explicit products (e.g., 

coastal spatial units and their associated cumulative disturbance scores) and our approach 

demonstrate a framework for assessing the condition of coastal habitats based on landscape 

disturbances that may be applicable for other regions.  

 

 

62 

Introduction 

Coastal habitats offer a diversity of benefits that are ecologically and socially important 

(Martínez et al. 2007; Barbier et al. 2011). Habitat-forming organisms found in coastal areas 

such as coral, kelp, and mangroves serve as critical habitat for many taxa including fish, sea 

turtles, marine mammals, and waterfowl, contributing to high biodiversity in areas where these 

organisms are found (UNEP 2006, Allsopp et al. 2008). Many marine species utilize coastal 

habitats as foraging, spawning, and nursery areas (e.g., Weng et al. 2007, Erwin 1996), and 

amphidromous and anadromous fishes require access to coastal habitats to complete their life 

cycles (e.g., Bell 1999, McDowall et al. 2003). Coastal habitats provide ecosystem services that 

benefit humans, such as provision of fish for consumption, reduction of sediment erosion from 

beach fronts through mitigation of wave impacts, support of primary production and nutrient 

cycling, and creation of multiple recreational opportunities such as diving and surfing to 

residents of and visitors to coastal areas. 

The ability of coastal habitats to provide ecological benefits and ecosystem services are 

in decline globally because they are increasingly subjected to anthropogenic disturbances (UNEP 

2006). Many disturbances result from human activities in the marine environment including 

dredging and trawling of benthic habitats for commercial fisheries (Thrush and Dayton 2002; 

Waycott et al. 2009) and degradation of habitats (such as coral reefs) from direct swimmer and 

diver contact (Halpern et al. 2007; Smith and Edgar 2014). However, another major source of 

disturbances to coastal habitats originates from the terrestrial landscape (Crain et al. 2009), 

including human land uses such as urbanization and agriculture. 

Influences of anthropogenic disturbances in terrestrial landscapes on coastal habitats are 

driven in part by hydrology. Natural hydrological connections are maintained by rainfall, but 

 

63 

anthropogenic disturbances can disrupt linkages between landscapes and the coast, leading to 

changes in and potentially degrading nearshore coastal habitats (Halpern et al. 2008; Wehrly et 

al. 2013). Impervious surfaces like pavement, roads, and building roof tops could prevent water 

from infiltrating into soils, subsequently reducing groundwater input and increasing surface 

runoff (Paul and Meyer 2001). Such changes in hydrology as well as other consequences 

resulting from conversion of natural to developed landscapes can additionally alter coastal 

habitats. In rivers draining urbanized or agricultural landscapes, studies have documented altered 

thermal regimes (Allan and Castillo 2007); altered sediment dynamics (Milliman and Meaded 

1983); and modified delivery of nutrients, toxics, and/or woody debris (Kennedy and Woods 

2012), and these impacts to rivers can be directly transmitted to coastal habitats. 

Across the globe, studies have shown links between degradation of coastal habitats and 

developed inland landscapes. Examples include eutrophication in Chesapeake Bay in the Mid-

Atlantic (Boesch 1996), hypoxia in the Gulf of Mexico (Alexander et al. 2008), coral bleaching 

of the Great Barrier Reef of Australia (Wooldridge 2009), and algae blooms of Lake Erie in the 

Laurentian Great Lakes (Michalak et al. 2013). These studies show consequences of inland 

landscapes affecting conditions in nearshore coastal habitats, in some cases, highlighting 

mechanisms by which these linkages occur. However, as individual, regionally-specific studies, 

they offer only modest guidance for extrapolating results to other regions or for describing and 

comparing such linkages between inland and coastal habitats across large spatial extents (i.e., 

entire states, ecoregions, continents). These limitations reflect a gap in understanding of 

relationships between connected inland and coastal habitats and underscore a critical research. 

The explicit characterization of linkages between inland and coastal systems can aid in efforts to 

conserve coastal habitats from anthropogenic disturbances. 

 

64 

An established approach for associating units in the coastal environment with units in 

proximate terrestrial landscapes to consistently account for spatial relationships has yet to be 

developed, though a few recent, regional examples have attempted to do this. Finlayson and 

Shipman (2003) identified discrete areas throughout Puget Sound based on current patterns and 

assumed sources of sediments delivered from terrestrial sources. Also, Wang et al. (2015) 

created units in the Laurentian Great Lakes based on proximity to the shoreline and bathymetric 

depth. These studies, however, do not explicitly account for how the coastal environment could 

be affected by inland anthropogenic disturbances, or how that influence could vary from 

different hydrologic inputs. Additionally, drivers of coastal habitat condition are diverse. While 

factors such as proximity to inland anthropogenic disturbances may result in fairly similar 

degrees of influence globally, other factors, such as climate and geology, may promote or limit 

linkages between draining catchments and coastal habitats and may vary by region (Talley et al. 

2006). Better understanding of spatial relationships between inland disturbances and coastal 

habitats as well as factors that can modify these relationships is integral to fully assess sources of 

degradation to coastal ecosystems (Halpern et al. 2008; Allan et al. 2013).  

This study aids in meeting those needs by characterizing how coastal habitats around the 

island of Maui, Hawaii may be degraded by inland disturbances delivered by river catchments. 

Maui is an appropriate study system because it has wide variation in catchment characteristics, 

including a range in urban and agricultural land use, and its inland and coastal systems are 

hydrologically linked (Hsiao, this volume a). Additionally, Maui’s unique inland and coastal 

habitats support endemic species that are a focus of conservation efforts. Using our 

understanding of spatial relationships developed in Hsiao (this volume a), our first objective of 

this chapter characterizes how discrete areas of the coastal environment may be spatially linked 

 

65 

to sets of catchments based on proximity. Next, we characterize the influence of landscape 

disturbances on coastal habitats by extending a measure of inland river habitat disturbance 

resulting from anthropogenic land uses into coastal environments. Our final objective evaluates 

how ecosystem services and other benefits provided by coastal habitats may be affected by the 

inland-derived disturbances. This study shows an approach for assessing the condition of coastal 

habitats based on landscape disturbances, providing information that can aid in management and 

conservation of Maui’s coastal resources and potentially those of other regions. 

Study area 

Methods 

This study was conducted in Maui, Hawaii, the second largest island of the Hawaiian 

archipelago with a land area of 1,884 km2 (Figure B2.1). The island was formed by the 

convergence of two volcanoes (West Maui Volcano, formed approximately 1.15 million years 

ago and East Maui Volcano, formed approximately 0.8 million years ago). Approximately 206 

km2 of coral reef habitat occurs in the nearshore region around the coastline of Maui, supporting 

12 species of corals (Grigg 1983). These corals provide habitat for a diversity of fish and 

invertebrates, including many endemic species (Randall 1976, Friedlander et al. 2008). 

Nearshore coastal habitats around Maui are hydrologically linked to the island’s landscape and 

are subjected to both natural and anthropogenic influences. Rain events rapidly alter sediment 

inputs, turbidity, and salinity of coastal receiving waters (Grigg 1995, Lau and Mink 2006). 

While coastal organisms are adapted to natural extremes in conditions (e.g., Bittler et al. 2014, 

Moura et al. 2016), anthropogenic activities can result in conditions that may be substantially 

different than those to which native organisms are adapted (DLNR DAR 2005, Storlazzi and 

Jaffe 2008). Additionally, both natural and anthropogenic influences originating within the 

 

66 

marine environment (i.e., wave impacts, damage from ship anchors) may interact with and 

compound effects of inland influences on coastal habitats (Chabanet et al. 2005, Buma 2015). 

Data collection and variable creation 

Inland landscape data 

Spatial units 

Our inland spatial units are called “pour point catchments,” which include all catchments 

draining from headwaters to the coastal environment by perennial and/or intermittent streams 

(Figure B2.2). Pour point catchments were derived from a stream layer developed in support of 

the Hawaii Fish Habitat Partnership (HFHP) by Tingley et al. (In Review). Confluence-to-

confluence stream reaches form the basis of the HFHP stream layer, and a local catchment is the 

topographically-defined land area that drains directly to the stream reach (following Wang et al. 

2011). Pour point catchments include the cumulative land area and river network upstream of the 

pour point draining into the coastal environment. 

Inland disturbances 

Inland disturbances are characterized by a cumulative habitat condition index (HCI) 

developed as a part of the 2015 National Assessment of Fish Habitats (http://fishhabitat.org/, 

Crawford et al. 2016). The HCI incorporates a combination of seven sub-indices that integrate 

major categories of disturbances to rivers and their catchments including: urban land use, 

agricultural land use, point source pollution, stream fragmentation, presence of former pineapple 

and sugarcane plantations, and ditch diversions within catchments. Subindices were composed of 

a variety of anthropogenic disturbance variables in each category, and the HCI used in this 

analysis is the linear combination of subindex scores. Specific variables and methods used to 

create each subindex and the HCI are described in Tingley et al. (In Review). Higher HCI scores 

 

67 

indicate greater intensity of anthropogenic disturbance within catchments and to rivers draining 

catchments (Crawford et al. 2016), and values at catchment pour points were used to reflect 

potentially greater risk of disturbance to nearshore coastal habitats. 

Coastal data 

Spatial units 

Coastal spatial units used in this analysis are 1 km2 grid cells that are guided to 18.079° 

N, 160.89° W. The 1 km x 1 km dimension was a resolution determined from previous work 

which tested for the strength of associations between metrics reflecting reef habitat condition 

with inland influences over varied distances (Hsiao, this volume a). These coastal grid cells 

extend from the coastline of Maui to the 100-fathom depth, an estimate used by National 

Oceanic and Atmospheric Administration's Office of the Coast Survey as a surrogate for 

potential coral ecosystem distribution throughout the United States (Rohmann et al. 2005). 

Coastal ecosystem services and benefits 

Point locations of coastal ecosystem services and benefits observed around the island of 

Maui were digitized by the Hawaii Office of Planning (1989) from the U.S. Army Corp of 

Engineers “Maui Resource Atlas.” While some services and benefits provided by coastal habitats 

may have changed around Maui since 1989, this dataset is the only comprehensive and 

consistent coverage available to discriminate between different services provided by Maui’s 

coastal habitats. We grouped all services and benefits into three main categories to aid in 

interpretation: recreation, fishing/fisheries, and habitats for biodiversity (Table A2.1). 

Recreational activities include excursion boating, sailing, board surfing, body surfing, canoe 

paddling, and sport diving (e.g., snorkeling and scuba diving). Types of fishing include gill 

netting, pole and line fishing, troll and bottom fishing, as well as cultural practices such as torch 

 

68 

fishing and spear fishing. Habitats for biodiversity include locations that support organisms 

including octopus, sea urchin, shark, and opihi –a shellfish that is a staple of the traditional 

Hawaiian diet. We also identified a subset of ecosystem services and ecological benefits that 

may be dependent upon coral reefs (Table A2.1), and by dependent, we suggest that the presence 

and health of coral reefs is very important to providing this service or benefit. These include 

sport diving, spear fishing, and habitat for octopus and opihi. 

Associating pour point catchments with coastal grid cells 

For this study, we assigned influences of pour point catchments to coastal grid cells based 

on proximity, an approach supported by findings from Hsiao (this volume a). We created 1 km, 5 

km, and 10 km buffers around centroids of each coastal grid cell (Jouffray et al. 2014). Pour 

point catchments were associated with a coastal grid cell when they fell within the varied buffer 

distances, and we used these varied distances to modify the strength of inland disturbances on 

coastal grid cells (process described below). 

Characterizing influences of cumulative inland disturbances on coastal habitats 

  A cumulative disturbance score was calculated for each coastal grid cell using the HCI 

scores of pour point catchments falling within 1 km-, 5 km-, and 10 km-buffer distances. This 

approach follows Halpern et al. (2008), a study that estimated impacts of inland disturbances on 

marine habitats based on proximity. We enhanced this concept to more directly account for 

distance between catchments and coastal habitats and for catchment area (i.e., greater catchment 

area is associated with greater river discharge and potentially greater input of disturbance to 

coastal habitats, following Wehrly et al. 2013). First, disturbance scores were calculated for 

coastal grid cells with pour point catchments that fall within 1 km buffers, pour point catchments 

between 1 km and 5 km buffers, and pour point catchments that fall between the 5 km and 10 km 

 

69 

buffers. Scores from pour points within specified distances were combined into single scores 

based on weighting by catchment areas to yield up to three potential scores for each coastal grid 

cell. Next, a decay factor was applied to the catchment combined scores more than 1 km away 

from a coastal grid cell to account for the weakening impact of inland disturbances further away 

from coastal habitats (following results of Hsiao, this volume a). The decay factor was 0.1 for 

catchments from 1 km to 5 km away, and 0.01 for catchments greater than 5 km away from a 

coastal grid cell. Finally, separate scores were combined into the cumulative grid cell score, with 

influences standardized by area-based weighting. We rescaled the final values to range between 

0 and 100, with low values indicating low levels of potential disturbances. We then assigned risk 

categories to each grid cell, by assigning equal numbers of cell in each of five categories: very 

low, low, moderate, high, and very high. 

Characterizing risk of degradation from inland disturbances to ecosystem services and 

benefits 

Locations of observed ecosystem services and benefits were spatially overlaid with the 

cumulative inland disturbance scores calculated for coastal grid cells. We then calculated the 

percentage of each ecosystem service and benefit subjected to the five risk categories of 

degradation from inland influences. Results were summarized by groups of ecosystem services 

and benefits (i.e., recreation, fishing/fisheries, and habitats for biodiversity) and by the subset 

that may be highly dependent on coral reefs. Finally, we calculated percentages of individual and 

groups of ecosystem services and benefits that fall in each of the five risk categories to assess 

which are the least and most at risk from inland disturbances.  

 

 

70 

 

Results 

Pour point catchment areas of Maui range between 0.08 km2 and 12.89 km2 (Table A2.2). 

While the HCI ranges from 0.00 to 1.00 for catchments throughout the five main Hawaiian 

Islands, the index ranges from 0.00 to 0.41 for Maui catchments, suggesting low to moderate 

levels of anthropogenic disturbances within these pour point catchments. 

Associating pour point catchments with coastal grid cells 

A total of 1,461 coastal grid cells were created between the shoreline and 100-fathom 

depth around Maui (Figure B2.2). There are more grid cells on the west side than the east side of 

Maui as a result of generally shallower depths on the west and steeper drop offs on the east. Of 

the grid cells, 239 have catchment pour points within 1 km, 789 have closest catchment pour 

points within 5 km, and 402 have closest catchment pour points within 10 km (Figure B2.2). 

Forty coastal grid cells are further than 10 km from the closest pour point catchment (Figure 

B2.2).  

Characterizing influences of cumulative inland disturbances on coastal habitats 

Figure B2.3 shows locations of urban and agricultural land use on Maui, as well as the 

cumulative potential risk of inland disturbances on coastal habitats. Together, these provide 

insights as to how and why impacts vary throughout Maui’s coastal habitats. Most notably, a 

majority of coastal units subjected to very high risk of degradation from inland influences are 

located near West Maui, where urban and agricultural land use are concentrated along the shore 

(Figure B2.3). Grid cells due west of the West Maui landscape range from very high to moderate 

risk of degradation from inland influences (Figure B2.3). In contrast, more coastal units in the 

southern and eastern coast of Maui are at low and very low risk of degradation from inland 

influences, where urban and agricultural land use are less common (Figure B2.3).  

 

71 

Characterizing risk of degradation from inland disturbances to ecosystem services and 

benefits 

Ecosystem services and benefits are widely distributed around the nearshore coastal 

environment of Maui (Figure B2.4). While fishing and fisheries occur throughout nearshore 

waters, recreation-related ecosystem services are concentrated on the western vs. eastern side of 

Maui (Figure B2.5A and B). Locations of habitat for biodiversity, in contrast, are much more 

spatially disparate. Fewer locations in general occur on the southern vs. northern coast. In terms 

of specific organisms, octopus are predominantly on the northern and western coast, and opihi 

are concentrated on the northern and southeastern coast (Figure B2.5C). Services and benefits 

that may be directly dependent on coral reefs (i.e., sport diving, spear fishing, and habitats for 

octopus and opihi) are found widely around the island.  

Our results also show that most ecosystem services and benefits occurring around Maui 

are at some risk of degradation from inland disturbances (Table A2.1). Across groups of services 

and benefits, 74% of locations that support recreation-related ecosystem services, 67% of 

fishing/fisheries locations, and 72% of habitats for biodiversity are at high or very high risk of 

degradation (Table A2.1). Within the recreation group, all locations of excursion boating and 

sailing, more than half of locations for board and body surfing, and half of locations for canoe 

paddling are in areas of very high risk of degradation (Table A2.1). Shell collecting is the only 

recreation service that predominately falls in areas of moderate risk (67%, Table A2.1). Within 

the fishing/fisheries group, all locations for bait fishing and most locations for torch fishing 

(95%), specialized fisheries (59%), and gill netting (58%) are at very high risk. Nearly half of all 

spear fishing locations and one-third of throw netting locations are at very high risk from inland 

disturbances. Most crabbing locations (56%) occur in areas of moderate risk, while troll and 

 

72 

bottom fishing locations are distributed across areas of moderate (34%) and high (34%) risk 

(Table A2.1). Of habitats for biodiversity, those supporting sea urchin and shark, as well as a 

majority of octopus (94%) and seaweed (87%) habitat locations are at very high risk of 

degradation (Table A2.1). Opihi habitats are generally distributed across areas of moderate 

(31%), high (26%), and very high (27%) risk of degradation from inland disturbances (Table 

A2.1). The subset of ecosystem services and benefits that may be most dependent on coral reefs 

are largely at high (19%) and very high (47%) risk of degradation from inland disturbances 

(Table A2.1). 

Discussion 

In this chapter, risk of degradation to coastal habitats from inland disturbances was 

assessed around the island of Maui, Hawaii. We first characterized spatial linkages between units 

on the landscape (pour point catchments) and units in the coastal environment (coastal grid 

cells). This allowed us to quantify and visualize how influences from inland landscapes may 

change with distance from coastal regions. Next, we used condition of inland catchments and our 

understanding of spatial relationships between inland and coastal habitats to characterize risk of 

disturbances from inland sources on coastal habitats. This step accounted for inputs from 

multiple catchments and catchment drainage area. By incorporating area, a surrogate for the 

magnitude of river flows, we accounted for greater influences from larger catchments with more 

potential to facilitate disturbance to the coast (Wehrly et al. 2013), and we are currently not 

aware of other studies that have accounted for these factors. Last, we used this information with 

spatial locations of coastal ecosystem services and benefits around Maui to demonstrate how 

they may be at risk from inland disturbances. This step is similar to one from a study done 

throughout the Laurentian Great Lakes to assess sources of anthropogenic stressors in relation to 

 

73 

ecosystem services (Allan et al. 2013), but we aren’t aware of other efforts in marine systems. 

Results showed many locations supporting ecosystem services and benefits related to recreation, 

fisheries, and habitats for biodiversity at risk of degradation from inland influences. Collectively, 

our results demonstrate a framework for assessing condition of coastal habitats based on inland 

landscape disturbances as well as an application of the approach. Our outcomes tell us which 

habitats around Maui are most and least threatened, and by extension, which ecosystem services 

and benefits are at greatest risk. This knowledge can be used to improve our ability to conserve 

Maui’s coastal habitats into the future. 

Associating pour point catchments to coastal grid cells 

  Development of coastal grid cells from the Maui shoreline to the 100 fathom depth is 

similar to work done in the Laurentian Great Lakes, where researchers defined the region of the 

coastal environment that would be most strongly affected by inland influences (i.e., the tidal 

riverine coastal and tidal riverine open water zones in FGDC 2012; the coastal margin and 

nearshore zone in Wang et al. 2015). However, our designations are unique in that we subdivided 

this region to account for coastal habitats that may be variably affected by catchment pour points. 

Furthermore, pour point catchments associated with coastal grid cells were developed based on 

findings from Hsiao (this volume a). Specifically, we defined distances to weight catchment 

influence on grid cells based on the fact that the most variance was explained in coastal habitat 

metrics by inland anthropogenic influences within 1 km of those habitats while overall variance 

explained by both natural and anthropogenic inland influences was greatest when catchments 

were within 5 km of habitats (Hsiao, this volume a). 

 

 

 

74 

Characterizing influences of cumulative inland disturbances on coastal habitats 

We also showed how the nearshore area of Maui is variably affected by inland 

influences. Because urban and agricultural land cover are concentrated on the western and 

northern coast of Maui, coastal grid cells adjacent to those areas are at greater risk of 

degradation. This outcome follows findings of both Halpern et al. (2008) and Wehrly et al. 

(2013) who characterize disturbance in coastal habitats from inland sources based on proximate 

developed landscapes. Our approach differs from these studies by additionally weighing 

catchment area and pour points at varying distances away from a coastal grid cell based on 

known relationships derived from Hsiao (this volume a).  

We also accounted for magnified risk of degradation by the concentrated input of 

multiple catchments into the coastal environment, such as the coastal region around West Maui, 

where many catchments drain into the coast. Currently, we are not aware of another study that 

explicitly accounts for multiple catchments and weighs their relative influence by catchment 

area, a surrogate for stream flow that reflects a larger rivers’ ability to drain larger areas of the 

landscape and potentially carry more materials into receiving waters than a smaller river.  

Characterizing risk of degradation from inland disturbances to ecosystem services and 

benefits 

We showed that a majority of the ecosystem services and benefits around Maui are at 

very high risk of degradation from inland disturbances. This finding follows the Millennium 

Ecosystem Assessment that reported a decline in ecosystem services within coastal habitats 

across the globe (UNEP 2006). Specifically, we considered three broad groups of ecosystem 

services and benefits in our assessment: recreation, fishing/fisheries, and habitats for 

biodiversity. First, regarding recreation, tourism is the largest sector of the Hawaiian economy, 

 

75 

and recreation by tourists is heavily dependent on the condition and health of coral reefs.  

Marine-based recreation is estimated to contribute $304 million annually to the state’s economy 

(Cesar and van Beukering 2004), with over 80% of visitors participating in some form of 

recreation that involves the coastal environment such as surfing and snorkeling (Friedlander et al. 

2005). Next, commercial fisheries in Hawaii's nearshore environment, which target at least 57 

species of fish including multiple species of tuna, mahimahi, and ono, are valuated at 

approximately $16.4 million annually (Grafeld et al. 2017). Recreational fishing such as throw 

netting and pole and line fishing are activities also highly-valued by locals and visitors to 

Hawaii, and it is estimated that more individuals participate in recreational vs. commercial 

fishing (Smith 1993). Additionally, nearshore fisheries are culturally important in Hawaii, and 

Native Hawaiians have traditionally relied on catch from torch fishing and spear fishing for 

subsistence (Carl 2009). Finally, we also found that habitats for biodiversity are at risk of 

degradation from inland influences, a fact that will challenge the ability to conserve organisms 

that depend on these habitats from both current and future stressors. Reduced biodiversity infers 

the possibility of decreased resilience of an ecosystem to recover from disturbances as well as a 

reduction in their ability to support other services (Worm et al. 2006). With a projected reduction 

in coral cover expected to occur throughout the Hawaiian Islands (Bruno and Selig 2007), we 

may also expect a co-occurring reduction in availability of all the ecosystem services and 

benefits that depend on coral reefs, and this will impact aspect societal well-being unless actions 

are taken to conserve these systems (UNEP 2006, Barbier et al. 2011). 

Our finding suggests that management actions that reduce the impacts of disturbance on 

the landscape draining to Maui’s coastal habitats could lessen the risk of degradation and protect 

the ecosystem services and benefits that those habitats provide. Based on results of the 

 

76 

characterization of inland influences on coastal habitats, ecosystem services and benefits are 

disproportionately located in areas that are at risk of degradation from inland disturbances. This 

is in part due to their locations being near the coast which are more strongly influenced by inland 

influences due to greater intensity of human settlement. Intense use of coastal areas can result in 

generally higher demand for and access to coastal resources as well as specific impacts including 

shoreline development, shoreline hardening, and other infrastructure (Creel 2003).  

Study limitations  

Our analytical approach was based on findings from Hsiao (this volume a), which tested 

for spatial associations between inland landscape factors and metrics characterizing coastal 

habitat condition. Outcomes of that study identified distances over which influences of inland 

landscape factors on coastal habitats changed, and these findings informed how we chose to 

associate pour point catchments with coastal grid cells as well as the various distance categories 

that we used to apply decay functions. This process could be improved with specific discharge 

estimates for catchments, allowing for more precise estimates of the magnitude of catchment 

influences based on stream flow vs. catchment area. It would also allow for greater 

understanding of effects of rain events as drivers that could promote or limits linkages around the 

island. We also did not have information to account for how currents may affect linkages, and 

how inland influences dissipate in the coastal environment. Our step to assign a decay factor 

could be improved with the availability of current information to account for the directionality of 

how inland influences travel throughout the coastal environment.  

Utility of the approach for conserving coastal habitats 

The analytical approach highlighted in this chapter demonstrates the utility of the 

development of consistent spatial units in the coastal environment attributed with information on 

 

77 

their linkages to catchments. The outputs yielded from this process enabled our identification of 

coastal areas that are at greater risk of degradation. This information would support decisions 

related to identification of target areas, allocation of management resources, and the designation 

of protection and regulations to consider inland linkages to the coastal environment. 

While our particular exercise evaluated the status of ecosystem services and benefits, the 

coastal spatial units could facilitate assessments of various other factors of interest for improved 

coastal conservation strategies (i.e., assessment of key fisheries locations, identification of 

marine protected areas). Other regions could also follow the framework laid out by this approach 

to test for relationships between inland and coastal systems and to extend inland disturbances 

into the coastal environment. Conservation efforts are needed to ensure that coastal habitats 

continue to provide ecosystem services and benefits, but their effectiveness is in part driven by 

strategic placements. Allan et al. (2013) highlight compelling opportunities in restoration efforts 

that target areas that offer many ecosystem services at low risk of degradation, as well as those 

areas at high risk that are affected by just a few stressors (as opposed to multiple stressors). Our 

analytical approach yields spatially-explicit products that can support decisions related to 

mitigating inland disturbances for the conservation of coastal habitats. Specifically, the creation 

of coastal spatial units enables the attribution and organization of information in consistent, 

comparable manners. Their cumulative disturbance scores disseminates the relative risks from 

inland disturbances and identifies highly threatened areas and sources of degradation.

 

78 

APPENDICES

 

79 

 

 

 

APPENDIX 2.A: 

Tables 

 

80 

Table A2.1. Proportion of coastal ecosystem services and benefits at various risk levels of 
degradation from inland disturbances; the greatest proportion for each group is bolded. 

 
Recreation 

Excursion Boating 
Sailing 
Aquatic Recreation 
Board Surfing 
Body Surfing 
Canoe Paddling 
Sport Diving* 
Shell Collecting 

Fishing/fisheries 

Bait Fishing 
Torch Fishing 
Specialized Fisheries 
Gill Netting 
Pole and Line Fishing 
Spear Fishing* 
Throw Netting 
Crabbing 
Troll and bottom fishing 

Habitats for biodiversity 

Sea Urchin 
Shark 
Octopus* 
Seaweed 
Lobster 
Opihi* 

Services dependent on coral reefs 
Other 

Anchorage 

Very Low 

0.06 
0.00 
0.00 
0.00 
0.00 
0.11 
0.00 
0.13 
0.00 
0.05 
0.00 
0.00 
0.05 
0.04 
0.05 
0.04 
0.08 
0.00 
0.09 
0.01 
0.00 
0.00 
0.02 
0.00 
0.04 
0.00 
0.03 

 

0.00 

Low  Moderate  High  Very High 
0.01 
0.00 
0.00 
0.00 
0.00 
0.00 
0.00 
0.00 
0.11 
0.06 
0.00 
0.00 
0.00 
0.03 
0.07 
0.07 
0.10 
0.00 
0.11 
0.08 
0.00 
0.00 
0.00 
0.00 
0.04 
0.15 
0.08 

0.19 
0.00 
0.00 
0.00 
0.12 
0.06 
0.33 
0.26 
0.67 
0.22 
0.00 
0.00 
0.09 
0.13 
0.23 
0.24 
0.25 
0.56 
0.34 
0.19 
0.00 
0.00 
0.02 
0.06 
0.35 
0.31 
0.23 

0.20 
0.00 
0.00 
0.33 
0.26 
0.33 
0.17 
0.18 
0.00 
0.20 
0.00 
0.05 
0.27 
0.21 
0.18 
0.19 
0.24 
0.11 
0.34 
0.15 
0.00 
0.00 
0.02 
0.06 
0.17 
0.26 
0.19 

0.54 
1.00 
1.00 
0.67 
0.62 
0.50 
0.50 
0.44 
0.22 
0.47 
1.00 
0.95 
0.59 
0.58 
0.47 
0.46 
0.34 
0.33 
0.13 
0.57 
1.00 
1.00 
0.94 
0.87 
0.39 
0.27 
0.47 

 

0.00 

 

0.00 

 

 

0.00 

1.00 

*Services or benefits included in the subset that may depend directly on coral reefs.  

 

 

 

81 

Table A2.2. Units, minimum (min.), maximum (max.), mean, and standard deviation (SD) values 
of inland variables within pour point catchments (n = 182) draining the Maui landscape.  

Variable 
Catchment area 
Habitat condition index*  
*Index ranges from 0 (best condition) to 1 (worst condition) across the five main Hawaiian 
Islands. 

Units  Min. 
km2 
0.08 
0.00 

Max. 
12.89 
0.41 

 

Mean 
1.98 
0.08 

SD 
2.34 
0.10 

 

 

 

 

82 

 

 

 

APPENDIX 2.B: 

Figures 

 

83 

Figure B2.1. Location of Maui in the Hawaiian archipelago (A), and pour points of catchments 
(n = 182) draining the Maui landscape (B). 
 

 

  

 

84 

 
Figure B2.2. Coastal spatial units around Maui, Hawaii, in relation to pour point catchments of 
the landscape draining into the coastal environment. 
 
 

 

 

 

 

85 

Figure B2.3. Risk of degradation from inland disturbances of coastal units around Maui, Hawaii. 
 
 

 

 

 

86 

 
Figure B2.4. Location of all ecosystem services overlaid with characterized inland disturbances 
in coastal spatial units. 
 

 

87 

Figure B2.5. Location of ecosystem services by group: Recreation (A), Fishing/Fisheries (B), 
and Habitats for biodiversity (C). 

 

 

88 

LITERATURE CITED

 

89 

LITERATURE CITED 

Alexander, R. B., R. A. Smith, E. Gregory, E. W. Boyer, J. V Nolan, and W. Brakebill. 2008. 

Differences in phosphorus and nitrogen delivery to the Gulf of Mexico from the Mississippi 
River Basin. Environmental Science & Technology 42(3):822–830. 

Allan, A. J. D., P. B. Mcintyre, S. D. P. Smith, B. S. Halpern, and G. L. Boyer. 2013. Joint 

analysis of stressors and ecosystems services to enhance restoration effectiveness. 
Proceedings of the National Academy of Sciences 110(1):372–377. 

Allan, D., and M. Castillo. 2007. Stream ecology: Structure and function of running waters, 2nd 

edition. Springer. 

Allsopp, M., R. Page, P. Johnston, and D. Santillo. 2009. State of the World’s Oceans. 

Barbier, E. B., S. D. Hacker, C. Kennedy, E. W. Kock, A. C. Stier, and B. R. Sillman. 2011. The 
value of estuarine and coastal ecosystem services. Ecological Monographs 81(2):169–193. 

Bell, K. N. I. 1999. An Overview of Goby-Fry Fisheries. International Center for Living Aquatic 

Resources Management 22(4):30–36. 

Bittler, K. M., L. P. Scheef, and E. J. Buskey. 2014. Freshwater inflows and blue crabs: The 
influence of salinity on selective tidal stream transport. Marine Ecology Progress Series 
514:137–148. 

Boesch, D. F. 1996. Science and management in four US coastal ecosystems dominated by land-

ocean interactions. Journal of Coastal Conservation 2(1):103–114. 

Bruno, J. F., and E. R. Selig. 2007. Regional decline of coral cover in the Indo-Pacific: Timing, 

extent, and subregional comparisons. PLoS ONE 2(8). 

Buma, B. 2015. Disturbance interactions: characterization, prediction, and the potential for 

cascading effects. Ecosphere 6(4):1–15. 

Carl, A. R. 2009. Method is Irrelevant : Allowing Native Hawaiian Traditional and Customary 

Subsistence Fishing to Thrive. University of Hawai’i Law Review 32(1):203–235. 

Cesar, He. S. J., and P. J. H. van Beukering. 2004. Economic Valuation of the Coral Reefs of 

Hawai’i. Pacific Science 58(2):231–242. 

Chabanet, P., M. Adjeroud, S. Andréfouët, Y.-M. Bozec, J. Ferraris, J.-A. Garcìa-Charton, and 

M. Schrimm. 2005. Human-induced physical disturbances and their indicators on coral reef 
habitats: A multi-scale approach. Aquatic Living Resources 18:215–230. 

 

90 

Crain, C. M., B. S. Halpern, M. W. Beck, and C. V. Kappel. 2009. Understanding and managing 

human threats to the coastal marine environment. Annals of the New York Academy of 
Sciences 1162:39–62. 

Crawford, S., G. Whelan, D. M. Infante, K. Blackhart, W. M. Daniel, L. Pam, T. Birdsong, D. J. 

Wieferich, R. Mcclees-funinan, S. M. Stedman, and P. Ruhl. 2015. Hawaii Streams - 
Detailed Methods. Page Through a Fish s Eye: The Status of Fish Habitats in the Hawaii 
Streams. 

Creel, L. 2003. Ripple effects: Population and coastal regions. Page Population Reference 

Bureau. 

Department of Land and Natural Resources (DLNR), D. of A. R. 2005. Marine Protected Areas 

in Hawai’i. Honolulu, Hawaii. 

Erwin, R. M. 1996. Dependence of Waterbirds and Shorebirds on Shallow-Water Habitats in the 

Mid-Atlantic Coastal Region: An Ecological Profile and Management Recommendations. 
Estuaries 19(2):213–219. 

Federal Geographic Data Committee (FGDC). 2012. Coastal and Marine Ecological 

Classification Standard (CMECS). 343pp. 

Finlayson, D. 2003. Puget Sound Drift Cells: The importance of waves and wave climate. Puget 

Sound Notes (47):1–12. 

Grafeld, S., K. L. L. Oleson, L. Teneva, and J. N. Kittinger. 2017. Follow that fish: Uncovering 

the hidden blue economy in coral reef fisheries. PLoS ONE 12(8):1–25. 

Grigg, R. W. 1983. Community structure, succession and development of coral reefs in Hawaii. 

Marine Ecology Progress Series 11:1–14. 

Halpern, B. S., K. a. Selkoe, F. Micheli, and C. V. Kappel. 2007. Evaluating and ranking the 

vulnerability of global marine ecosystems to anthropogenic threats. Conservation Biology 
21(5):1301–1315. 

Halpern, B. S., S. Walbridge, K. a Selkoe, C. V Kappel, F. Micheli, C. D’Agrosa, J. F. Bruno, K. 
S. Casey, C. Ebert, H. E. Fox, R. Fujita, D. Heinemann, H. S. Lenihan, E. M. P. Madin, M. 
T. Perry, E. R. Selig, M. Spalding, R. Steneck, and R. Watson. 2008. A global map of 
human impact on marine ecosystems. Science 319(5865):948–52. 

Hsiao, J. This volume a. Understanding spatial linkages between inland and coastal systems: 

Identifying and quantifying inland influences on coastal habitats in Characterizing spatial 
linkages between inland and coastal habitats for improved conservation in Maui, Hawaii, 
Masters Thesis. Michigan State University, East Lansing, MI. 

 

91 

Jouffray, J.-B., M. Nyström, A. Norström, I. Williams, L. Wedding, J. Kittinger, and G. 

Williams. 2015. Human and natural drivers of multiple coral reef regimes across the 
Hawaiian Archipelago. Philosophical Transactions B 370:1–8. 

Kennedy, D. M., and J. L. D. Woods. 2012. The influence of coarse woody debris on gravel 

beach geomorphology. Geomorphology 159–160:106–115. 

Lau, L. S., and J. F. Mink. 2006. Hydrology of the Hawaiian Islands. Page Hydrology of the 

Hawaiian Islands. University of Hawai‘i Press, Honolulu, Hawaii. 

Martínez, M. L., A. Intralawan, G. Vázquez, O. Pérez-maqueo, P. Sutton, and R. Landgrave. 
2007. The coasts of our world: Ecological, economic and social importance. Ecological 
Economics of Coastal Disasters 63(2–3): 254–272. 

McDowall, R. M. 2003. Hawaiian biogeography and the islands’ freshwater fish fauna. Journal 

of Biogeography 30(5):703–710. 

Michalak, A. M., E. J. Anderson, D. Beletsky, S. Boland, N. S. Bosch, T. B. Bridgeman, J. D. 

Chaf, K. Cho, R. Confesor, I. Dalo, X. Liu, M. R. Mcwilliams, M. R. Moore, D. J. Posselt, 
R. P. Richards, and D. Scavia. 2013. Record-setting algal bloom in Lake Erie caused by 
agricultural and meteorological trends consistent with expected future conditions. 
Proceedings of the National Academy of Sciences of the United States of America 
110(16):6448–6452. 

Milliman, J. D., and R. H. Meade. 1983. World-wide delivery of river sediment to the oceans. 

The Journal of Geology 91(1):1–21. 

Moura, R. L., G. M. Amado-Filho, F. C. Moraes, P. S. Brasileiro, P. S. Salomon, M. M. 

Mahiques, A. C. Bastos, M. G. Almeida, J. M. Silva, B. F. Araujo, F. P. Brito, T. P. Rangel, 
B. C. V. Oliveira, R. G. Bahia, R. P. Paranhos, R. J. S. Dias, E. Siegle, A. G. Figueiredo, R. 
C. Pereira, C. V. Leal, E. Hajdu, N. E. Asp, G. B. Gregoracci, S. Neumann-Leitao, P. L. 
Yager, R. B. Francini-Filho, A. Froes, M. Campeao, B. S. Silva, A. P. B. Moreira, L. 
Oliveira, A. C. Soares, L. Araujo, N. L. Oliveira, J. B. Teixeira, R. A. B. Valle, C. C. 
Thompson, C. E. Rezende, and F. L. Thompson. 2016. An extensive reef system at the 
Amazon River mouth. Science Advances 2(4):e1501252–e1501252. 

Rohmann, S. O., J. J. Hayes, R. C. Newhall, M. E. Monaco, and R. W. Grigg. 2005. The area of 

potential shallow-water tropical and subtropical coral ecosystems in the United States. Coral 
Reefs 24(3):370–383. 

Small, C., and R. J. Nicholls. 2003. A global analysis of human settlement in coastal zones. 

Journal of Coastal Research 19(3):584–599. 

Smith, S. D. A., and R. J. Edgar. 2014. Documenting the Density of Subtidal Marine Debris 

across Multiple Marine and Coastal Habitats. PLOS ONE 9(4):e94593. 

 

92 

Storlazzi, C. D., and B. E. Jaffe. 2008. The relative contribution of processes driving variability 

in flow, shear, and turbidity over a fringing coral reef: West Maui, Hawaii. Estuarine, 
Coastal and Shelf Science 77(4):549–564. 

Talley, D. M., G. R. Huxel, M. Holyoak, D. M. Talley, G. R. Huxel, and M. Holyoak. 2006. 

Connectivity at the land-water interface. Pages 97–129in K. R. Crooks and M. Sanjayan, 
editors.Connectivity Conservation. Cambridge University Press. 

Tingley et al. In Review. Influences of natural landscape characteristics on distributions of 

tropical stream organisms: Stream reach classification in the Hawaiian Islands 
Hydrobiologia.  

Thrush, S. F., and P. K. Dayton. 2015. Disturbance to Marine Benthic Habitats by Trawling and 

Dredging: Implications for Marine Biodiversity. Annual Review of Ecology and 
Systematics 33(2002):449–473. 

United Nations Environment Programme (UNEP). 2006. Marine and coastal ecosystems and 

human well-being: A synthesis report based on the findings of the Millennium Ecosystem 
Assessment. 76pp. 

Wang, L., D. Infante, P. Esselman, A. Cooper, D. Wu, W. Taylor, D. Beard, G. Whelan, and A. 
Ostroff. 2011. A Hierarchical Spatial Framework and Database for the National River Fish 
Habitat Condition Assessment. Fisheries 36(September 2014):436–449. 

Wang, L., C. M. Riseng, L. A. Mason, K. E. Wehrly, E. S. Rutherford, J. E. McKenna, C. 

Castiglione, L. B. Johnson, D. M. Infante, S. Sowa, M. Robertson, J. Schaeffer, M. Khoury, 
J. Gaiot, T. Hollenhorst, C. Brooks, and M. Coscarelli. 2015. A spatial classification and 
database for management, research, and policy making: The Great Lakes aquatic habitat 
framework. Journal of Great Lakes Research 41(2):584–596. 

Waycott, M., C. M. Duarte, T. J. B. Carruthers, R. J. Orth, W. C. Dennison, S. Olyarnik, A. 

Calladine, J. W. Fourqurean, K. L. Heck, a R. Hughes, G. a Kendrick, W. J. Kenworthy, F. 
T. Short, and S. L. Williams. 2009. Accelerating loss of seagrasses across the globe 
threatens coastal ecosystems. Proceedings of the National Academy of Sciences of the 
United States of America 106(19):12377–12381. 

Wehrly, K., L. Wang, D. Infante, C. Joseph, A. Cooper, L. Stanfield, and E. S. Rutherford. 2013. 

Landscape change and its influences on aquatic habitats and fisheries in the Great Lakes 
Basin. Pages 81–104. Great Lakes Fisheries Policy and Management: A Binational 
Perspective. 

Weng, K. C., A. M. Boustany, P. Pyle, S. D. Anderson, A. Brown, and B. a. Block. 2007. 
Migration and habitat of white sharks (Carcharodon carcharias) in the eastern Pacific 
Ocean. Marine Biology 152(4):877–894. 

 

93 

Wooldridge, S. A. 2009. Water quality and coral bleaching thresholds: Formalising the linkage 

for the inshore reefs of the Great Barrier Reef, Australia. Marine Pollution Bulletin 
58(5):745–751. 

Worm, B., E. B. Barbier, N. Beaumont, J. E. Duffy, C. Folke, B. S. Halpern, J. B. C. Jackson, H. 
K. Lotze, F. Micheli, S. R. Palumbi, E. Sala, K. A. Selkoe, J. J. Stachowicz, and R. Watson. 
2006. Impacts of Biodiversity Loss on Ocean Ecosystem Services. Science 314(5800):787–
790. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

94 

MANAGEMENT IMPLICATIONS 

Coastal habitats are ecologically and socially important, but their conditions and 

availability are imperiled due to anthropogenic disturbances (UNEP 2006). While many 

anthropogenic disturbances result from human activities in and uses of the marine environment 

(Thrush and Dayton 2002; Waycott et al. 2009; Halpern et al. 2007), another major contributor to 

the decline of coastal habitats originates from inland landscapes (Crain et al. 2009), including 

human land use such as urbanization and agriculture. While linkages between inland landscapes 

and coastal habitats are widely acknowledged, few studies have evaluated their spatial 

relationships to inform development of a coastal spatial framework that would account for 

mechanisms by which inland variables influence coastal habitats. This thesis attempts to improve 

understanding of spatial relationships between inland and coastal systems. Specifically in this 

section, we highlight the principal findings described in Chapter 1 and Chapter 2, which provide 

information that could contribute to an approach that associates units in the coastal environment 

with units in proximate terrestrial landscapes to consistently account for spatial relationships. We 

then synthesize findings in the context of potential management implications, which can inform 

decisions that support the conservation of coastal habitats.  

Chapter 1 

In Chapter 1, we explored how coastal habitat conditions are influenced by landscape 

factors. The coastal habitat conditions are reflected by a set of reef metrics summarized by 

locations of reef units that described benthic habitats: habitat complexity, percent coral cover, 

percent stressed coral cover, percent sand cover, percent macroalgae cover, percent crustose 

coralline red algae cover, crown-of-thorn-starfish density, and urchin density. The landscape 

factors are represented by 3 main groupings of variables: inland natural, inland anthropogenic, 

 

95 

and coastal. The inland variables are summarized in pour point catchments, and the coastal 

variables are attributed to their respective reef units.  

Results of this chapter indicated that the strength of relationships between inland 

characteristics and coastal habitat conditions varied with distance between pour point catchments 

and reef units. In particular, we found that the greatest amount of variance explained by inland 

anthropogenic variables in coastal habitat condition was observed at 1 km. This suggests that 

when managing for inland anthropogenic disturbances, decisions should account for influences 

of catchments with pour points within 1 km of the coastal habitats of interest. 

We also noted that the amount of variance explained in coastal habitat conditions 

decreased when pour point catchments greater that 5 km away from reef units were included in 

the study. This finding suggests that habitat conditions within 5 km of catchment pour points are 

sensitive to inland catchment characteristics. Inland natural variables in our analysis included 

mean annual rainfall. Managers of coastal resources should anticipate the extent to which inland 

influences may affect coastal habitat conditions to be altered in the face of climate change, 

depending on whether a draining catchment receives more or less precipitation than its historical 

annual averages.  

Our predictions of individual reef metrics from environmental variables illustrate 

complex interrelationships between inland factors and coastal habitats. Of the best-predicted reef 

metrics, habitat complexity and macroalgae cover were predicted by inland natural variables, 

while macroalgae cover and crown-of-thorns starfish density were predicted by inland 

anthropogenic variables. This result shows the complexity of inland linkages to coastal habitats 

and highlights greater sensitivity of some metrics to inland disturbances. A manager of coastal 

habitats may consider those reef metrics associated most strongly with inland anthropogenic 

 

96 

variables (i.e., percent macroalgae cover and crown-of-thorns starfish density), as indicators of 

inland disturbances from proximate catchments. This is an important step forward with the 

National Ocean Council setting the need for coastal and marine spatial planning as a national 

priority in the United States (Craig 2015). The major findings of this chapter contribute to the 

development of a coastal spatial framework, which characterizes ecologically-meaningful spatial 

units and their interrelationships with inland influences. This improved understanding would 

facilitate could guide decision-making in the coastal and marine environment by accounting for 

disturbances that originate from the landscape. 

Chapter 2 

In Chapter 2, we characterized spatial relationships between inland pour point catchments 

and spatial units in the coastal environment based on Chapter 1 findings (Hsiao, this volume a). 

We first created spatial units consisting of 1 km2 grid cells and categorized them into 3 major 

groups: those with closest pour point catchments that fall within 1 km buffers, between 1 km and 

5 km buffers, and between the 5 km and 10 km buffers. We calculated a cumulative disturbance 

score from inland influences that accounts for proximity of the catchment pour points associated 

with each coastal spatial unit, catchment area, and measures of inland habitat condition. Inland 

habitat condition is a composite index showing relative risk of inland disturbances to inland 

habitats (i.e., streams), and we applied the index to evaluate potential risk of inland disturbances 

to coastal habitats downstream of proximate catchments. We mapped the cumulative disturbance 

score calculated for each coastal spatial unit to visualize areas of coastal habitats at varying risk 

of degradation from inland disturbances (Figure B2.3). We additionally mapped locations of 

ecosystem services to assess those that are most threatened by inland disturbances (Figure B2.4). 

 

97 

The outputs could facilitate decisions that seek to identify areas of concern and for allocation of 

management resources.  

Coastal areas are experiencing rapid ecological changes (Goussard and Ducrocq 2017). 

Specifically in the context of coral reef management, our findings of relative low and high risk of 

habitat degradation to inland influences (Hsiao, this volume b) could offer guidance that 

managers seek. Conservation resources are often limited, and the prioritization of areas with low 

levels of unmanageable threats or high levels of manageable threats would facilitate effective 

allocation (Harris et al. 2017). Through the characterization of threats from inland disturbances 

on coastal habitats, the spatially-explicit maps from this chapter enable the identification of 

target areas that are at the greatest and least risk of habitat degradation from inland disturbances 

that accounts for the linkages between inland and coastal systems. Our maps enable the 

identification of areas that are not degraded, which supports the designation of Marine Protected 

Areas (MPAs) of habitats that are still intact. The maps also enable the identification of areas 

that are degraded, as well as their associated catchments, for targeted mitigation of 

anthropogenic threats that could be remotely influencing the coastal habitats of interest.  

The mapped locations of ecosystem services in relation to the relative risk of degradation 

of coastal habitats from inland disturbances is a proof-of-concept exercise that highlighted the 

utility of our approach for assessments. For example, the relative risk of degradation of coastal 

habitats could support existing policies, such as the management of Essential Fish Habitat (EFH) 

that is a part of the 1996 Magnuson-Stevens Fishery Conservation and Management Act 

(Rosenberg et al. 2000). EFH are designated habitat areas that are actively managed; they 

support important fisheries (e.g., serve as spawning grounds and nursery) and warrant special 

protection. Their locations can be mapped similarly to the ecosystem services, and the analytical 

 

98 

process described (Hsiao, this volume b) could be applied, providing information regarding the 

relative risk of degradation of the designated EFH of interest from inland disturbances. 

The results of this chapter additionally highlights the utility of developing consistent and 

comparable spatial units in the coastal environment. Our attribution of information in the coastal 

spatial units with their associated inland catchments were enabled by the findings of spatial 

linkages from an earlier study (Hsiao, this volume a). As improved understanding between 

inland systems and coastal habitats become available, the process of data attribution can be 

refined to more accurately reflect inland disturbances in the coastal environment. The extent of 

linkages between inland and coastal systems may also be altered in the face of climate change 

(Goussard and Ducrocq 2017), with anticipated shifts in coastal community structure and trophic 

interactions (Harley et al. 2006). Our study approach allows for the comparison of how inland 

influences may be expressed on coastal habitats when the parameters that affect linkages 

changes. Such information would facilitate better assessments of various management metrics of 

interest (e.g., reef health metric, resilience measures, and ecosystem services) for improved 

coastal conservation strategies.  

 

 

 

 

 

 

 

 

 

 

99 

LITERATURE CITED

 

100 

LITERATURE CITED 

Craig, R. 2015. An historical look at planning for the federal public lands: Adding marine spatial 

planning offshore. George Washington Journal of Energy & Environmental Law:1–20. 

Crain, C. M., B. S. Halpern, M. W. Beck, and C. V. Kappel. 2009. Understanding and managing 

human threats to the coastal marine environment. Annals of the New York Academy of 
Sciences 1162:39–62. 

Goussard, J. J., and M. Ducrocq. 2017. Facing the future: Conservation as a precursor for 

building coastal territorial cohesion and resilience. Aquatic Conservation: Marine and 
Freshwater Ecosystems 27(S1):151–161. 

Halpern, B. S., K. a. Selkoe, F. Micheli, and C. V. Kappel. 2007. Evaluating and ranking the 

vulnerability of global marine ecosystems to anthropogenic threats. Conservation Biology 
21(5):1301–1315. 

Harris, J. L., E. Estradivari, H. E. Fox, O. S. McCarthy, and G. N. Ahmadia. 2017. Planning for 
the future: Incorporating global and local data to prioritize coral reef conservation. Aquatic 
Conservation: Marine and Freshwater Ecosystems 27(S1):65–77. 

Hsiao, J. This volume a. Understanding spatial linkages between inland and coastal systems: 

Identifying and quantifying inland influences on coastal habitats in Characterizing spatial 
linkages between inland and coastal habitats for improved conservation in Maui, Hawaii, 
Masters Thesis. Michigan State University, East Lansing, MI. 

Hsiao, J. This volume b. Assessing the influences of inland landscape factors on coastal habitats 

for improved conservation in Characterizing spatial linkages between inland and coastal 
habitats for improved conservation in Maui, Hawaii, Masters Thesis. Michigan State 
University, East Lansing, MI. 

Rosenberg, A., T. E. Bigford, S. Leathery, R. L. Hill, and K. B. 2000. Ecosystem approaches to 
fishery management through essential fish habitat. Bulletin of marine science (November 
2014):66:535-542. 

Thrush, S. F., and P. K. Dayton. 2015. Disturbance to Marine Benthic Habitats by Trawling and 

Dredging: Implications for Marine Biodiversity. Annual Review of Ecology and 
Systematics 33(2002):449–473. 

UNEP (United Nations Environment Programme). 2006. Marine and coastal ecosystems and 

human well-being: A synthesis report based on the findings of the Millennium Ecosystem 
Assessment.  

Waycott, M., C. M. Duarte, T. J. B. Carruthers, R. J. Orth, W. C. Dennison, S. Olyarnik, A. 

Calladine, J. W. Fourqurean, K. L. Heck, a R. Hughes, G. a Kendrick, W. J. Kenworthy, F. 

 

101 

T. Short, and S. L. Williams. 2009. Accelerating loss of seagrasses across the globe 
threatens coastal ecosystems. Proceedings of the National Academy of Sciences of the 
United States of America 106(19):12377–12381. 

 

102