Cyrtosperma Merkusii); in Vitro and in Vivo

Salinity Tolerance of Giant Swamp Taro (Cyrtosperma merkusii); In vitro and In vivo

Shiwangni RAO

A thesis submitted in fulfilment of the requirements of the Degree of Master of Science

School of Biological and Chemical Sciences

Faculty of Science, Technology and Environment

The University Of the South Pacific

September, 2011

©Shiwangni Rao 2011

Author Declaration

I Shiwangni Rao, declare that this thesis is my own work and that, to the best of my knowledge, it contains no material substantially overlapping with material submitted for the award on any other degree at any institution, except where due acknowledgement is made in the text.

Supervisor Declaration This is to declare that this thesis titled “Salinity Tolerance of Giant Swamp Taro Cyrtosperma merkusii” submitted in fulfilment for the Degree of Master of Science in Environmental Science to the University of the South Pacific, is the original research work of Miss Shiwangni Rao conducted under our supervision and guidance. It contains no material that is overlapping with material submitted for the award on any other degree at any institution, except where due acknowledgment is made in the text.

Principle Chief Supervisor;

Dr. Anjeela Jokhan

Dean Faculty of Science Technology and Environment

University of the South Pacific

P.O. Box 1168

Suva

Principle Co-supervisors:

Dr. Mary Taylor

Genetic Resources Coordinator/Centre of Pacific Crops and Trees (CePaCT) Secretariat of the Pacific Community Private Mail Bag Suva Signature: Date: ____02/12/11______

Dr. Arthur Webb Division of Science and Technology Secretariat of the Pacific Community Private Mail Bag Suva Signature:

iii

Acknowledgement

I would like to acknowledge the very important people that have helped me throughout this research. First and foremost the Australian Government, who provided the funding for this research through the International Climate Change Adaptation Initiative (ICCAI).

Dr. Anjeela Jokhan the chief principle supervisor for introducing me to the sponsors of this research. For being the academic guiding light throughout the research in terms of building the research and write-up.

Dr. Mary Taylor the co-supervisor, for accepting me as a candidate for this research and allowing me to conduct my research at the Centre for Pacific Crops and Tress (CePaCT), Narere. For giving her full support despite her busy schedules and encouraging my exposure in the field of scientific research. Also for introducing me to the resource personnel’s and for her continuous academic and technical support.

Dr. Arthur Webb co-supervisor, for his scientific and technical input and advice during the research, especial for the ground water salinity survey in Tuvalu. The staff of CePaCT, for teaching me tissue culture and for their moral and technical support during the experimental phase of the research.

University of the South Pacific Research committee for considering this research as significant and the developments that can be achieved through it and giving their approval. The technical staff in the chemistry and biology departments namely Rosely Sharm, Shelvin Singh, Dinesh Sharma and Roselyn Lata.

The Tuvalu Agriculture Minister Mr. Itai Lausaveve for guiding my stay in Tuvalu and for his technical support. The Kaupule members and agriculture officers of Nanumaga, Nanumea, Niutao, Nui and Nukulaelae for providing me with

iv information and technical support. The very enlightening farmers of the six surveyed islands of Tuvalu for sharing their knowledge on the giant swamp taro.

The Federated States of Micronesia, Pohnpei State Department Chief Agriculture officer Mr. Adelino Lorens, the office staff and the agriculture field technicians at the pilot farm. For all their technical support in the two day workshop in Pohnpei, helping in translation, collection of information in the farm. The farmers that participated in the workshop for sharing their bulk of knowledge and contributing in the development of the Giant swamp taro Cyrtosperma merkusii Descriptor List.

I would also like to acknowledge my husband Mr. James Chand, for his continuously encouragement and pushing me to go the extra mile in my research. For being understanding, for his prayers and advice during the research. My Parents Mr Mahendra Prakash, Mrs. Shanti Mani and my sisters for their moral support and prayers.

Finally, I would like to thank the many people that have not been listed above but have in their very own little way contributed to the smooth running and the successfully completion of this research project

Thank you all.

Abstract Climate change related sea level rise together with adjustments to wind and wave patterns may cause an increase in the incidence of salt water intrusion into fresh ground water lenses, particularly in atoll islands. This saline intrusion may end up in Giant swamp taro (Cyrtosperma merkusii) cultivation pits, a crop which is a major food on these atoll islands and also a great part of their cultural identity. Hence, an increase in the salinity levels of the fresh ground water not only threatens the food security of these atoll island communities, but also their identity. In past literature, Giant swamp taro has been referred to as slightly salt tolerant. It has been seen to survive at a salinity level of around 2-3ppt (Dunn, 1976; Manner, 2006: Webb, 2007). However, these salinity levels are only claims and have not been tested in controlled trials. Therefore, there is an urgent need to investigate the salt tolerance potential of Giant swamp taro and utilize it as a buffer against increases in ground water salinity. Furthermore, the documentation along with sustainable conservation of giant swamp taro is also essential to prevent the loss of traditional knowledge and diversity of this important crop. Given that climate change is expected to increase the incidence of salt intrusion into giant swamp taro pits, the fundamental endeavours of this project were threefold (a) to investigate the incidence of salt water intrusion in Tuvalu (b) to develop the knowledge base of giant swamp taro through a descriptor list (c) to develop a rapid in vitro screening methodology for salt tolerance screening which could be used to assess the salinity tolerance of two groups of swamp taro cultivars, Ikaraoi and Katutu from Kiribati. A preliminary in vivo method for screening for salt tolerance was also developed. The ground water salinity survey in Tuvalu was carried out over a period of five weeks and six islands were visited. Measurements were taken using a salinity meter. The survey showed ground water salinity levels between 2006 and 2010 increased on Funafuti but not on the other islands. An in vitro method was developed for rapid screening. Using this methodology, the two cultivar groups from Kiribati were shown to survive salinity concentrations of

0% (0ppt), 0.5% (5ppt), 1.0% (10ppt), 1.5% (15 ppt) and 2% (20ppt) salt. However, the in vivo plants were only able to tolerate up to 0.5% (5pp) salt concentrations, possibly due to stress imposed by other environmental factors. However, further research is needed for both the ground water salinity survey and the salt tolerance screening. Further investigations would give both a clearer idea of the incidence of increase in ground water salinity levels and also the variation that might exist in salinity tolerance with different cultivars. A more clear understanding of the extent to which genetic diversity can affect salt tolerance would assist the selection of hardier cultivars. There is much significant scientific value to be gained from this project, as limited studies have been carried out generally on giant swamp taro and determining the level of salinity tolerance is essential so that farmers and communities know which cultivars they can use as they try to manage climate change.

VII

Contents Author Declaration ...... i

Supervisor Declaration ...... ii

Acknowledgement ...... iv

Abstract ...... VI

1.0 INTRODUCTION ...... 1

2.0 LITERATURE REVIEW ...... 5

2.1 ATOLL GROUND WATER LENS ...... 5

2.1.1 Threats to fresh ground water lens ...... 7

2.2 SALINITY TOLERANCE IN PLANTS ...... 9

2.2.1 Effects of increased soil salinity ...... 11

2.2.2 Plant Response ...... 13

2.2.3 Approaches to salt tolerance ...... 17

2.2.4 Salinity Testing ...... 21

2.3 GIANT SWAMP TARO (Cyrtosperma merkusii) ...... 24

2.3.1 Origin ...... 27

2.3.2 Current Distribution ...... 29

2.3.3 Physiology ...... 30

2.3.4 Morphology ...... 38

2.3.5 Cultivar Descriptor List ...... 41

2.3.7 Utilization ...... 48

2.3.8 Nutrition ...... 50

2.3.9 Conclusion ...... 52

3.0 TUVALU GROUND WATER FIELD STUDY ...... 54

3.1 INTRODUCTION ...... 54

VIII

3.2 PRE SURVEY ...... 58

3.3 SURVEY ...... 61

3.4 RESULTS & DISCUSSION ...... 62

3.4.1 Nanumea ...... 62

3.4.2 Nanumaga ...... 68

3.4.3 Niutao...... 73

3.4.4 Nui ...... 77

3.4.5 Funafuti ...... 82

3.4.6 Nukulaelae ...... 88

3.4.7 Rainfall- Ground Water Recharge ...... 93

3.4.8 Tuvalu Ground Water Salinity ...... 94

.3.5 CONCLUSION ...... 98

4.0 DEVELOPMENT OF A RAPID IN VITRO SCREENING METHOD ...... 101

4.1 INTRODUCTION ...... 101

4.2 METHOD ...... 101

4.2.1 Multiplication...... 102

4.2.2 In Vitro ...... 103

4.2.3 In Vivo ...... 107

4.2.4 Evaluation Parameters ...... 109

4.2.5 Data Analysis ...... 110

4.3 RESULTS ...... 111

4.3.1 In Vitro ...... 111

4.3.2 In Vivo ...... 119

5.0 DISCUSSION ...... 125

6.0 CONCULSION ...... 128

Bibliography ...... 130

ANNEX ...... 140

X List of Figures 1.1 Atoll islands in the Pacific 1.2 Giant swamp Taro Cyrtosperma merkusii plantation 1.3 Man hidden by Giant swamp taro leaf that is more than 1m in length and width 2.1 Ground water lens as defined by the Ghybe-Herzberg Principle 2.2 Giant Swamp taro Cyrtosperma merkusii 2.3 Giant Swamp taro corm. 2.4 Giant Swamp taro farm in FSM. 2.5. Phylogenic classification of Giant Swamp taro. 2.6. Map of Malesia, including the possible origins of Giant swamp Taro Cyrtosperma merkusii. 2.7. Map of the cultural spheres and the current distribution of giant swamp taro expect for New Zealand 2.8. Sunken cultivation in Tuvalu, Nanumea. 2.9 ’Opened’/’ Bottomless’ Padanus Pandanus tectorious leaf woven baskets in Kiribati 2.10 Coconut Cocus nucifera leaf woven bottomless basket in Funafuti, Tuvalu. 2.11 Giant swamp taro cultivation in Fiji. 2.12 Concrete cement pit Taro cultivation on Tuvalu in Funafuti 2.13 Giant swamp taro harvested corms 2.14 Giant swamp taro corms 2.15 Labelled giant swamp taro flower. 2.16 young spadix and flower. 2.17 Seeded berries on spadix. 2.18 Mature spadix. 2.19 young Pwh weitata flower. 2.20 Mature Pwh weitata flower. 2.21 leaves of giant swamp taro. 2.22 corm of giant swamp taro. 3.1. Giant Swamp Taro (Cyrtosperma merkusii) or Pulaka

3.2. Pulaka pit on Nui Island, Tuvalu. 3.3. Causeway constructed in the Tepela Pit area, which had greatly divested the crops due to highly saline water. 3.4. Abandoned Giant swamp taro pit on Nanumea. 3.5. A fully productive pit on Nanumea, depicting the Pulaka productivity level that can be attained on the island 3.6. The Tepela area where pulaka is being once again cultivated in hope of reviving the plantation. 3.7. Very healthy pulaka plants that grow in Funafuti. 3.8 Pulaka Kula, one of the cultivars of Giant swamp Taro found on Funafuti claimed to be highly salt tolerant 4.1 In vitro Experiment treatment combination structure 4.2 Overall morphological response to salt applications, plants from the left; 0%, 0.5%, 1.0% and 1.5% salt.

List of Tables 2.1 Number of studies done for genetic modifications per species for salt tolerance from 1998-2003 (Flowers, 2003) 2.2 Number of tests done for a particular gene to improve salinity tolerance in plants and the number of studies carried out in each one since 1993-2003 (Flowers, 2003). 2.3 Kiribati description of giant swamp taro growth stages (Manner, 2009). 2.4 List of pests, diseases and their impacts on giant swamp taro (Bradburry, 1988; Iese.V, 2005; Manner, 2009). 2.5 Revised descriptor list 2.6 Local recipes of giant swamp taro 2.7 Giant swamp taro nutritional value (SPC, 2006) 2.8 Giant swamp taro cultivar nutritional value (Englberger, 2005) 3.1 Nanumea ground water salinity 3.2 Ground water salinity on Nanumaga

XII

3.3 Pulaka pit salinity on Niutao 3.4 Pulaka pit salinity on Nui 3.5 Pulaka pit salinity on Funafuti 3.6 Pulaka pita salinity on Nukulaelae 3.7 Comparison of average rainfall for 2006 and 2010 3.8 Comparison of 2006 and 2010 ground water salinity levels 4.1 Salt solution mixtures 4.2 Salinity increment 4.3 In vitro experimental design 4.4 In vivo Experimental Design 4.5 Mean of cultivar group measured parameters when subjected to the five salinity levels 4.6 Mean of plant response measured parameters when subjected to the five salinity levels 4.7 Plant Response to ASW and NaCl 4.8 Contamination Rate according to application method 4.9 Pre and post experiment salinity levels 4.10Cultivar group response to the salinity levels 4.11 Percentage survival rate of the cultivar groups 4.12Plant response to the various salinity levels 4.13Percentage survival rate of the various salinity levels

List of Graphs 3.1. Pulaka pit salinity on Nanumea 2010 3.2. Ground water salinity on Nanumea 3.3. Pulaka pit salinity on Nanumaga 2010 3.4. Ground water salinity on Nanumaga 3.5. Pulaka pit salinity on Niutao 2010 3.6. Ground water salinity on Niutao 3.7. Pulaka pit salinity on Nui 2010

XIII

3. 8. Ground water salinity on Nui 3.9. Pulaka pit salinity on Funafuti 2010 3.10. Ground water salinity on Funafuti 3.11. Pulaka pit salinity on Nukulaelae 2010 3.12 Ground water salinity on Nukulaelae 3.13. 2010 Ground water salinity levels in Tuvalu 3.14. Comparison of the 2006 and 2010 ground water salinity levels 4. 1 Regression analysis graph of Suckers 4.2 Regression analysis graph of corm 4.3 Regression analysis graph of number of dying leaves List of Maps

3.1 Nanumea. 3.2 GPS located Pulaka pits on Nanumea 3.3 Nanumaga. 3.4 GPS located Pulaka pits on Nanumaga 3.5 Niutao. 3.6 GPS located Pulaka pits on Nanumaga 3.7 Nui 3.8 GPS located Pulaka pits on Nui. 3.9 Fongafale atoll. 3.10 GPS located Pulaka pits on Funafuti, Fongafale 3.11 Nukulaelae atoll. 3.12 GPS located Pulaka pits on Motutala Islet, Nukulaelae. 3.13 GPS located Pulaka pits on Nukulaelae main islet.

XIV

1.0 INTRODUCTION A bit less than half a century (1961-2003) ago the rate of sea level rise was at 1.8+/- 0.5 mm per year (Bindoff, 2007). In the 21st century it is has been predicted by the six scenarios presented in the 4th Assessment Report of the Intergovernmental Panel on Climate Change (IPCC, 2007), to increase to 5 mm per year. At this rate of increase low lying atolls such as in Kiribati, Tuvalu and the Federated States of Micronesia (FSM) (Figure 1.1) are faced with the extreme effects of climate change, which greatly threaten their food security and livelihoods.

Atolls are very fragile, in the sense that they are relatively small in size and have limited geography and topography. They are also very vulnerable to extreme changes in climate and natural disasters. Along with this, they have a low adaptive capacity to environmental changes (Bindoff, 2007). Sea level rise and possible changes in rainfall threaten these fragile ecosystems by the possible increased effects of salt water intrusion. Salt water intrusion into the fresh ground water lens of the atolls increases the ground water salinity, which can be lethal to atoll vegetation. Atoll islands such as Tuvalu have seen a decline in crop production and salt water intrusion is considered to be responsible. Webb (2007), in an attempt to investigate the incidence of salt water intrusion in Tuvalu concluded that further, monitoring and documentation is needed to fully understand the scenario of salt water intrusion.

The more obvious contributors to increases in ground water salinity on atolls are low rainfall and sea water inundation, which is caused by storm surges, sea swells and king tides. Sea level rise has increased the probability of occurrence of these inundations, as it has given added height to the already high waves generated during these occurrences (Liz, 2007; Aung and Prasad, 2009; White and Falkland, 2010). This sea water penetrates through to the ground water lens and increases the ground water salinity level contributing to the threat to food security (Woodroffe, 1989, 2008; White and Falkland, 2010).

Figure 1.1 Atoll islands in the Pacific (from http:// www.infoplease. com)

To ensure food security and to buffer the impacts of climate change on these fragile atolls there is a need for food crops that can withstand the challenging conditions imposed by climate change. Among these needs are food crops that are tolerant to increasing ground water salinity. Salinity tolerance is of key importance to the atolls with any of their crops, hence the need for a rapid screening process for salinity tolerance. Giant swamp taro (Cyrtosperma merkusii) (Figure 1.2 and 1.3) is one such food crop where information regarding its salinity tolerance would be very useful, but at present little research of any kind is carried out on this crop.

By Shiwangni Rao

Figure1.2 Giant swamp taro Cyrtosperma merkusii plantation

Figure1.3 Man hidden by Giant swamp taro leaf that is more than 1m in length and width.

By Shiwangni Rao

In past studies conducted in the Micronesian region and Tuvalu, giant swamp taro has been consistently highlighted as a crop with potential salt tolerance characteristics (Dunn, 1976; Lambert, 1982; Vickers, 1982; Brandburry and Holloway, 1988; Onwueme, 1999; Kazutaka and Michia. 2003; Covich, 2006; Deenik and Yost, 2006.; Iese, 2005). However, some atoll communities claim that increasing soil salinity due to salt water intrusion is affecting their giant swamp taro production (Lausaveve 2010, pers. comm.).

In some cases this has led to farmers abandoning their cultivation pits on the atolls of Tuvalu. Since there has been no specific investigation on the crop in this regard, the controversy still exists over the giant swamp taro’s actual salinity tolerance capacity. Giant swamp taro is not only a local staple but has also been woven into the traditions and cultures of Pacific atoll communities (Thaman, 2002; Iese, 2005; Manner, 2009). This plant with a large diversity of cultivars is adapted to the challenging natural environment of atoll islands unlike many other Pacific staple crops. Therefore it is an ideal crop for looking for salinity tolerance and for building on what tolerance that might exist.

This research aims to investigate the incidence of salt water intrusion by looking at the ground water salinity in the ‘Pulaka Pits’ (Giant swamp taro pits in Tuvaluan) on the atoll islands of Tuvalu. It further aims to test the possible salt tolerance level in giant swamp taro (Cyrtosperma merkusii) by subjecting two Kiribati cultivar groups the larger ‘Ikaraoi’ and the smaller ‘Katutu’ to various salinity levels and by doing this an in vitro screening method will be developed. An in vivo screening method will also be investigated to support the results from the in vitro methodology. It also aims to development the initial cultivar character descriptor list done by Iese (2005). Achieving all of the above will greatly add to the knowledge base on giant swamp taro and provide a foundation on which to base further research regarding the very important issues of salinity tolerance.

2.0 LITERATURE REVIEW

2.1 ATOLL GROUND WATER LENS

Coral atolls such as the Federated States of Micronesia (FSM), Marshall Islands, Tuvalu and Kiribati are built on the relics of volcanic craters consisting of two layers. A Pleistocene karst of limestone deposit forms the first layer; this is covered by a second layer from the Holocene period. The upper Holocene layer is composed of unconsolidated calcareous matter such as sediments, coral sand, and coral fragments (Metai, 2002; Metutera, 2002; Webb, 2007; White and Falkland, 2010). Ground water lens which is the life source for the flora and fauna on these atolls typically forms as a result of rainfall on these land masses, but it is difficult to define the process in a simple formula. This is due to the many factors which affect the ground water lens such as rainfall, composition of atoll soil ranging from Holocene to present day, atoll underground structures, tidal pressure and vegetation (Woodraffe, 1989; Mimura, 1999; Deenik and Yost, 2006; Rozell, 2007).

The classical model of ground water on atolls suggests ground water forms a lens shape, with a transition boundary from sea water to fresh ground water. This projection has been defined by the Ghyben-Herzberg principle (Woodraffe, 1989; Mimura, 1999; Rozell, 2007) (Figure 2.1).

In light of further research and better understanding of the ground water lens White and Falkland (2010), have proposed a steady state approximation. This takes into account the majority of contributing factors, however this model has some limitations as not all the atoll islands are the same. Another limitation to the classic model provided by White and Falkland (2010) is that ground water probably does not form an exact lens shape; rather it varies in shape according to the atoll properties. In this steady state approximation, a ground water lens will not have a

5 sharp boundary. It is actually a transition zone where salinity increases with depth and distance from the centre of the land mass.

Ghyben-Herzberg principle- Ground water lens

Figure 2.1: Ground water lens as defined by the Ghyben-Herzberg principle (Woodraffe, 1989).

According to this approximation and under conditions of similar rainfall and island condition, higher and wider islands should have a larger ground water lens compared to smaller and narrower islands. According to White and Falkland (2010), raised limestone atolls will have more net ground water recharge than low coral atolls for the same amount of rainfall since on low atolls large root trees such as coconut (Cocus nucifera) transpire directly from the ground water lens (Dunn, 1976).

As stated earlier atolls are generally made up of karst limestone, coral fragments and sediment deposits, however atoll deposits may differ in texture. Tuvalu has a coarse texture, hence a more permeable atoll structure compared to the Maldives and Kiribati. This coarse material is due to the deposition made by many storms that resulted in boulders and rubble to be embedded on the atoll’s Holocene layer such

6 occurred during cyclone Bebe in 1972 (White and Falkland, 1999). Since an increase in permeability is expected to reduce the height of the ground water lens, Tuvalu’s ground water lens is much thinner than the Maldives and Kiribati (White and Falkland, 1999).

In addition, a ground water lens is also affected by daily tidal fluctuations, as the movement of tides encourages mixing of the fresh ground water and sea water (Metai, 2002; Deenik and Yost, 2006; White and Falkland, 2010). Tidal efficiency is defined as the ratio of tidal amplitude of the ground water to that of the sea. According to the common theory tidal influence decreases with distance from the coast and the ground water lag of tidal force response should increase. However, this does not hold true for low lying atoll islands, where wells and bores show a decrease in tidal influence from shore. This is due to the permeable limestone karst, where rapid transmission of tidal pressure takes place resulting in a vertical propagation in the middle of the atoll and vertical along with horizontal tidal propagation towards to the coast (Rozell, 2007; White and Falkland, 2010). All these factors combined make the fresh ground water lens of atoll islands vulnerable, this in turn affects the flora and fauna on the atolls.

2.1.1 Threats to fresh ground water lens The rate of sea level rise is increasing, with a projected 5mm per year in the 21st century (Bindoff, 2007). This poses many threats to fragile low lying atoll ground water lens and an atoll island’s food security.

According to some researchers, natural disasters as a result of climate change will increase in frequency and/or intensity in the future (Bindoff, 2007; Rodgers, 2009;). This means that frequency and probability of sea water inundation due to natural disasters may increase; coupled with the high permeability of atolls this could result in increased ground water salinity (White and Falkland, 2010). Such natural disasters

7 include storm surges, storm sea swelling, tsunamis generated from marine land slide, earthquakes and volcanic eruptions.

Apart from inundation, some of these natural disasters such as earthquakes, marine landslides and volcanic eruption may directly affect the ground water lens by causing excessive disturbance and disrupting the delicate balance of the ground water lens. However, at the moment there is no significant research on these scenarios. Drought is another natural disaster that affects the ground water lens but in a different manner. In the event of extended periods of drought, there is a decrease in ground water recharge, hence a reduction in the ground water lens and a corresponding increase in salinity. This is due to the main body of the ground water lens decreasing in size and a widening of the transition zone where freshwater and sea water mixing takes place (White and Falkland, 1999; Woodraffe, 1989).

One popular theory related to climate change and sea level rise is that sea level rise increases coastal/land erosion (Eid and Huisbergen, 1992; Gerald, et al, 2007; Aung, et al 2009; Lal, 2009; Talia, 2009). If erosion does occur then island land mass is reduced pushing the ground water lens transition boundary inwards. This results in a decrease in the ground water lens and an increase in the ground water salinity levels.

However, recent studies of erosion conducted by Webb and Kench (2010) show otherwise. The research employed remote sensing images and historical aerial photography of 27 atoll islands from Kiribati, Federated States of Micronesia and Tuvalu. According to this study in spite of a rate of 2 mm per year sea level rise over the last 19-61 years, 43% of the islands surveyed showed no change in size, 43% increased in size and only 14% of the total islands assessed decreased in size. In many cases it was observed that where an island was eroded on one side, the opposite side accumulated sediments, equalizing the sediment movement giving a net zero change in size. In line with this finding White and Falkland (2010) state that it is more likely for coastal erosion to occur as a result of extreme events that

8 generate strong waves rather than just a gradual rise in sea level. They also state that while sea level rise and coastal erosion could result in a decrease in size of the ground water lens, it is more likely to be affected by a decrease in recharge (rainfall).

In a critique of the study carried out by Webb and Kench (2010), Schaeffer and Hare (2010) state that Webb and Kench (2010) focused on a time period where the rate of sea level raise was only 2mm/yr. They argue that with the projected increase in the rate of sea level rise and an increase in ocean acidity, the conditions of island stability may not persist. An increase in ocean acidity reduces calcification and growth of coral reefs; hence reducing the protection these coral reefs provide atoll islands in the face of increasing sea levels. Furthermore, the study by Webb and Kench (2010) was only conducted in relation to island horizontal mass (width) and has not taken the island elevation into account, which plays an important role in determining island ground water lens.

While there are a number of threats faced by ground water lenses, the human factor should not be ignored. According to Webb (2007) and White and Falkland (2010), humans have contributed significantly to impacting the quality of ground water lens on Pacific atolls. These contributions include pollution, over-extraction, population and development pressures. However as stated earlier the incidence of salt water intrusion and resultant increase in ground water salinity is a complex issue and needs further investigation. The same goes for the claims by farmers of the atoll communities that salt water intrusion is to blamed for the decrease in crop production.

2.2 SALINITY TOLERANCE IN PLANTS

Soil salinity in terms of dryland and wetland salinity, is one of the primary abiotic factors that hinders crop production not only in the Pacific but worldwide. It has been present since the pre agricultural times but more recently has been aggravated due to improper agricultural practices, deforestation, unsustainable living and

9 enhanced climate change (Zhao.M, et al., 2001; Yamaguchi and Blumwald, 2005). More than 6% of the earth’s entire land mass, which is 800 million ha of land, is currently affected by increased soil salinity (Munns and Tester, 2008). With the global population projected to increase by 2.8 billion over the next fifty years (United Nations 2004:4), increased soil salinity poses a threat to the food security systems of the world. Investing in development of salt tolerant food crops may hold some answers to these threats (Zhu, 2001; Arzani, 2008). This can be achieved by screening the diverse gene pool of crop plants, with their cultivars that vary in their response to environmental stress such as salinity (Arzani, 2008).

Salts are generally present in soil but at levels where they would be beneficial and not detrimental to plants. However, where soil salinity exceeds an electrical conductivity (EC) of 4000 dS/cm it is said to be saline (Munns and Tester, 2008). An electrical conductivity of 4000 µS/cm is equal to 40 mM of NaCl or 2.34 ppt which is the threshold of plant salinity tolerance. Only salt tolerant halophytes can survive such salinity levels, while salt intolerant glycophytes which form the majority of the earth’s mass flora cannot survive this level of salinity and the 0.2 MPa of osmotic pressure imposed by it (Munns and Tester, 2008). The composition of soil salinity includes sodium chloride along with other salts such as potassium chloride and magnesium chloride and so on. Despite the many salts present in soil, sodium chloride (NaCl) is the most researched salt. NaCl significantly affects plant health compared to other salts (Chen, et al., 2007; Arzani, 2008). The Na+ and Cl- ions are more detrimental to plants than other ions (Arzani, 2008); it causes ionic and osmotic stress that leads to early senescence of leaves, deteriorated plant health and even death. It is also the most soluble and wide spread salt (Munns and Tester, 2008).

2.2.1 Effects of increased soil salinity The effects of increased soil salinity on plant health are quite complex. Researchers are still struggling to clearly identify the causes and effects of plant responses due to salinity (Munns and Tester, 2008). For example, in the salt stress scenario, the cause- effect relationship between photosynthesis and growth is difficult to classify. Reduced photosynthesis can be a cause, or it can be the effect of reduced growth rate (Munns and Tester, 2008). Many researchers have concluded that salinity affects plants in two ways; osmotic and ionic stress (Yeo, 1998; Zhu, 2001; Flowers, 2003; Arzani, 2008; Yamaguchi and Blumwald, 2005).

The first effect increased salinity has on plants is that it induces osmotic stress. Similar to drought conditions, increased salt concentration in the soil causes a water deficit resulting from osmotic stress. This is evident minutes after exposure to increased salinity levels, above the 40 mM NaCl plant threshold limit. As a result of osmotic stress, stomatal conductance and shoot growth are significantly reduced (Zhu, 2001; Yamaguchi and Blumwald, 2005). Unlike roots which are in the frontline of contact with salts, shoots are more sensitive. Hence while root growth may be unaffected, shoot growth is significantly reduced. Followed by a decrease in expansion of growing leaves, slow emergence of new leaves and lateral buds, lateral buds may also remain dormant (Yamaguchi and Blumwald, 2005; Arzani, 2008). Furthermore, osmotic stress reduces tillering and causes curling of leaf sides in dicotyledons. The explanation given by Munns and Tester (2008) is that a reduction in stomatal conductance, shoot growth and leaf area by curling of leaves, allows plants to conserve water for longer. Reduction in stomatal conductance means that the stomata remains closed more often which reduces the amount of water being lost, while reduction in shoot growth and curling of leaf reduces the surface to volume ratio hence reducing the amount of area available for transpiration to occur. Conserving water is essential for plants in a salt stressed environment as this prevents salt accumulation in the cells and therefore allows longer survival.

When osmotic stress is induced, cells and leaves lose water. This loss may be quickly replaced within hours due to the plant’s internal osmotic adjustments but biomass growth may still be reduced. Over time these reductions contribute to the final smaller size of the plant and smaller and thicker dimension of leaves (Chen.Z, et al., 2007; Munns and Tester, 2008; Kader, et al., 2011). Munns and Tester (2008) explain that this response is due to some form of internal signalling which is initiated at the first instance of reduction in cell water potential. In the past, abscisic acid (ABA) had been seen as a potential initiator of the signal as it does play a role in root-to-shoot signalling, stomatal conductance and cellular signalling. However, ABA was ruled out when a study done in 2004 by Fricke found that while ABA levels increased initially at the time of osmotic stress, the effects of reducing biomass surface area and conductance still persisted long after. There is accruing evidence that a negative growth regulator protein called DELL of integrates a range of hormonal signals and gibberellins and is currently seen as the initiator of growth regulation in stressful conditions (Munns and Tester, 2008).

Meanwhile, the second effect of ionic stress subsequently comes into play as Na+ and Cl- ions accumulate to toxic levels over time, hence there is a delayed effect (Yamaguchi and Blumwald, 2005). This toxicity is evident with the increased senescence of old leaves. Salt ions “arrive” at the new and old leaves at the same rate. However, since the new younger leaves are still growing and expanding they are able to exceed the Na+ and Cl- ion accumulation and so avoid toxic effects. On the other hand, the older leaves are no longer growing or expanding hence the ions accumulate to toxic levels resulting in early leaf death and an increased rate of senescence (Yamaguchi and Blumwald, 2005; Munns and Tester, 2008). If the net pace at which old leaves expire exceeds the net pace at which new leaves emerge, growth of the whole plant is hindered, due to the reduction in the amount of photosynthesis occurring. Only in extremely high salt concentration does ion toxicity hinder plant growth more than osmotic stress (Arzani, 2008; Munns and Tester, 2008).

Apart from the toxic effects of ions on leaves, ion accumulation in the plant also hinders other physiological processes. Accumulation of Na+ in the cytosol causes an imbalance in the plant’s homeostasis by altering the Na+/K+ ratio (Zhu, 2001; Chen.Z, et al., 2007). This occurs due to the increased influx of Na+ ions through low-affinity K+ channels, as well as through low and high-affinity K+ carriers. The increase in concentration of the extracellular Na+ increases the -140mV potential difference across the plasma membrane (Yamaguchi and Blumwald, 2005; Chen.Z, et al., 2007). This results in an inward passive flow of Na+ ions through low-affinity K+ channels such as K+ Outward Rectifying Channel (KORCs), K+ Inward Rectifying Channels (KIRCs) and Non-Selective Cation Channels (NSCCs) (Zhu, 2001; Yamaguchi and Blumwald, 2005).

This ion accumulation also hinders enzyme reactions, and other processes such as photosynthetic parameters. This includes pigment composition, leaf osmotic and water potential, transpiration rate, temperature, leaf water content (Arzani, 2008; Munns and Tester, 2008) and respiration and nutrition acquisition (Zhu, 2001; Yamaguchi and Blumwald, 2005). Yet another indirect damage caused by increased uptake of salts is the production of Reactive Oxygen Species better known as ROS, produced mainly in the chloroplast. These substances can cause excessive cellar damage (Zhu, 2001).

2.2.2 Plant Response Glycophytes and halophytes are both susceptible to high salinities. However halophytes are better adapted to regulating uptake of salts than glycophytes (Yamaguchi and Blumwald, 2005). Plants adjust to an increase in soil salinity concentration in a number of ways. Glycophytes respond to the induced osmotic stress resulting from increased soil salinity by restricting the inflow of salts and by producing compatible solutes such as proline, glycinebetaine and sugars to adjust osmotic pressure (Zhu, 2001; Flowers, 2003; Yamaguchi and Blumwald, 2005). In

13 contrast halophytes use their cellular vacuoles to accumulate and compartmentalise the salts (Yamaguchi and Blumwald, 2005).

Responses to ionic stress are quite complex compared to osmotic stress. Plants do this in two ways; firstly by removing Na+ ions from the cells and secondly by compartmentalising of Na+ ions in vacuoles (Zhu, 2001; Flowers, 2003; Colmer, et al., 2006). Plants employ H+-ATPase and Na+/H+ antiporters present in the cell plasma membrane to actively ‘pump in’ H + and ‘pump out’ Na+. An electrochemical H+ gradient is formed by the H+-ATPase which allows coupling of the passive flow of H+ by the antiporters into cell and Na+ out of cell along the gradient (Yamaguchi and Blumwald, 2005).

The compartmentalization of Na+ into the cell vacuoles is mediated by a similar procedure whereby vacuole H+-trans locating enzyme such as the H+-PPiase and H+- ATPase generate an electrochemical gradient that allows the Na+/H+ antiporters to transport the Na+ into the vacuole (Yamaguchi and Blumwald, 2005). Furthermore, to address the damaging effects of Reactive Oxygen Species (ROS), plants under stress produce a number of proteins and osmolytes, many of which have unidentified roles but are assumed to help the plant alleviate the ROS produced and reduce damage (Zhu, 2001).

Twenty–five years ago Emanuel Epstein articulated aspects of the biological and technical challenges related to solving soil salinity tolerance (Yamaguchi and Blumwald, 2005). Since then many studies have been carried out but with little success. This is because salinity tolerance is affected by more than one gene and the identification of the key genetic codes is quite an intricate task (Yeo, 1998; Colmer, et al., 2006; Cuarteo, et al., 2006; Soumaya, et al., 2010). However some studies on the subject claim a small degree of improved salt tolerance, where genetic engineering subjected to a single gene, enzyme, antiporter or osmolyte can bring

14 about significant tolerance (Fatokun, et al., 2002; Yesayan, et al., 2008; Kchaou, et al., 2010).

On the other hand, Flowers (2003), in analysing 68 research papers on barley, citrus, rice, and tomatoes from 1993 to 2003 (Table 2.1 and 2.2), concluded that while these crop species were slightly improved by means of enhancements to a particular gene, enzyme or osmolyte they should not be presumed to be total salinity tolerance at a whole plant level. Improvements and enhancements may improve salinity tolerance at a cellular level, however this may be detrimental at a whole plant level as at the cellular level the complexities of a whole plant are not truly presented (Flowers, 2003; Cuarteo, et al., 2006). Munns and Tester (2008) are of a different opinion, in that they believe that by engineering a gene, responsible for a single aspect of salt tolerance such as intracellular compartmentation or osmolyte production, salt tolerance may be enhanced. They acknowledge that this enhancement may be small but can be improved by the development of more specific Quantitative Trait Loci (QTL) Markers. QTLs can identify the multiple genes in association with the specific trait and then using these traits to breed more tolerant cultivars (Colmer, et al., 2006; Cuarteo, et al, 2006).

Table 2.1. Number of studies done for genetic modifications per species for salt tolerance from 1998-2003 (Flowers, 2003) Species studied No. of studies Arabidopsis thaliana 14 Brassica napus and B. juncea 3 Citrus (Carrizo citrange) 1 Cucumis melo (melon) 2 Diospyros kaki (Japanese persimmon) 1 Lycopersicon esculentum (tomato) 5 Medicago sativa (alfalfa) 2 Nicotiana tabaccum (tobacco) 19

Oryza sativa (rice) 17 Solanum melongena (eggplant) 1 Solanum tuberosum (potato) 2 Triticum aestivum (wheat) 1

Table 2.2 Number of tests done for a particular gene to improve salinity tolerance in plants and the number of studies carried out in each one since 1993-2003 (Flowers, 2003). Tested Genes no. of Tests Apoplastic invertase, Apo-Inv 1 Arginine decarboxylase, ADC 1 Betaine aldehyde dehydrogenase, BADH; betB, choline dehydrogenase (CDH); 15 15 choline oxidase, codA (glycinebetaine) Ca2+-dependent protein kinase, CDPK 1 Ca/H antiporter, CAX1 1 Calcium-binding protein, EhCaBP 1 Calicneurin; protein kinase, CaN 1 Ca protein kinase, OsCDPK7 1 Glutathione S-transferase, GST and glutathione peroxidase, GPX 1 Glyceraldehyde-3-phosphate dehydrogenase, GPD 1 Glycogen-synthase kinase-3, AtGSK 1 Heat shock protein, DnaK/HSP70 1 1 High-af®nity potassium transporter, *HKT1a 3 3 Isopentenyl transferase, ipt (increased cytokinin) 1 1 Late embryo abundant protein, HVA1 (a LEA) 2 2 Mannitol 1-phosphate dehydrogenase, mt1D (mannitol) 6 6 Myo-inositol O-methyltransferase, IMT1 (ononitol) 1 1

Omega-3 fatty acid desaturase, fad7 (fatty acid processing) 1 1 Osmotin-like protein 1 1 Proline dehydrogenase; Delta (1)-pyrroline-5-carboxylate synthetase (proline) 4 4 Proline transporter, AhProT1 1 1 Proton sodium exchanger, *HNX1a 4 4 Putative transcription factor, Al®n1 2 2 Rare Cold Inducible gene 3, RCI3 1 1 Glutamine synthetase, GS 2 Rice Hal2 like, RHL 1 S-adenosylmethionine decarboxylase, SAMDC (spermine, spermidine) 1 Serine/threonine kinase, AT-DBF2 1 Sorbitol-6-phosphate dehydrogenase, SPD (sorbitol) 1 SR-like, putative splicing protein 1 Transcription factors, DREB1A; AhDREB1 2 Trehalose-6-phosphate synthase/phosphatase, TPSP (trehalose) 1 Yeast halotolerance gene, Hal2 3 Yeast halotolerance gene, Hal1 2 Yeast mitochondrial superoxide dismutase, Mn-SOD 1 Vacuolar H+-pyrophosphatase, AVP1 1

2.2.3 Approaches to salt tolerance With the rise in global population and the demand for better climate adapted food crops to meet the global food consumption needs, much effort and funding has gone into researching crops of significance globally, rather than regional or national importance.

One approach being used to address salinity tolerance employs Quantitative Trait Loci (QTLs), marker-assisted selections, and direct selection in stressful conditions to utilize the natural genetic variation in plants. The second approach includes breeding of transgenic plants where existing gene expressions have been altered and the introduction of novel genes (Yamaguchi and Blumwald, 2005; Arzani, 2008).

2.2.3.1 Approach 1

The first approach is basically selecting salt tolerant cultivars and lines using either conventional selection techniques or modern day molecular biology techniques, wherein DNA markers are used to identify QTLs. QTLs are able to identify several genes controlling a specific trait. It requires less time and the factor of environmental effects on a trait is eliminated, compared to conventional selection techniques (Yamaguchi and Blumwald, 2005). However, the drawback of this particular approach is that microsatellite markers using Restriction Fragment Length Polymorphism (RFLP) and Amplified Fragment Length Polymorphism (AFLP) in high density DNA maps are required, the development of this is costly and time consuming.

Conventional selection or direct selection techniques are less expensive and intensive. However they have their own limitations such as the time required and the effect of the environment on the organism. A recent experiment using this technique was conducted by Kchaou et al. (2010) on five Olive cultivars; Arbequira I18, Arbosana I43, Chetoui, Chemlali and Koroneiki. The assessment of salinity tolerance using direct selection found that overall ‘Chemali’ was the most tolerant and ‘Arbequina’ was the least tolerant (Kchaou, et al., 2010). It was observed that the induced salinity levels (0, 0.5, 50, 100, 200 mM) affected plant growth parameters differently. Kchaou et al. (2010) concluded this to be the result of the genetic variation within the species (Kchaou, et al., 2010). Direct selection has been used by many other researchers to select salt tolerant cultivars as in scented geranium (Garnett, et al., 2002), tomato (Colmer, et al., 2006; Cuarteo,et al., 2006), wheat

(James and Munns.R, 2003), taro (Tyagi, et al., 2009) rice (Pajuabmon, et al., 2009), potato (Aghaeri, et al., 2008), Barley (Fricke and Peters, 2002), pistachio (Banakar and Ranjbar, 2010) and alfalfa (Peel, et al., 2004). This type of selection sometimes forms the basis or foundation of salinity tolerance screening, which is then further, developed using the second approach.

Wild relatives have also been used to improve salt tolerance in domestic cultivars. The first approach is used to select the most salt tolerant wild cultivar of the crop which is then hybridized with the domesticated one. This approach useful for those crops where salt tolerant wild cultivars (Satoh, et al., 1998; Colmer, et al., 2006; Cuarteo, et al., 2006) are available , for example tomato and wheat. However progress has been slow despite the breakthroughs in genetic engineering and the successful identification of salt tolerant wild cultivars. This is due to the involvement of a large number of genes and the various environmental factors that affect the production of a salt tolerant transgenic using a wild cultivar (Cuarteo, et al., 2006). Furthermore, combining genes of distant relatives is quite difficult compared with those for close relatives due to the greater divergence in gene pools (Colmer, et al., 2006). It is also quite costly to recover the receptor cultivar’s genetic background (Cuarteo, et al., 2006).

2.2.3.2 Approach 2

Advancements in genetics and molecular biology have provided a boost to the second approach. Many researchers today focus on engineering a particular gene (Wu, et al., 2008; Subramanyam, et al., 2010), enzyme, osmolyte (Sakhanokho and Kelley, 2009) or antiporter (Rubio, et al., 2007; Hein'andez, et al., 2009; Wei, et al., 2010), that can improve the tolerance and resistance of plants against increased salinity (Fricke and Peters, 2002; Flowers, 2003; Yamaguchi and Blumwald, 2005). There have been quite a number of studies carried out on mutagenesis and enhancement of certain genes to produce tolerant cultivars of crop plants. One such study shows that the gene coding for cation H+ AtNHX1 mutants in Arabidopsis

19 thaliana integrated in yeast can enhance salinity tolerance in yeast (Hein'andez, et al., 2009). Similar studies have been conducted in other crops such as barley (Chen, et al., 2007), tomato (Rubio, et al., 2004), arabidopsis (Zhao, et al, 2007) and tobacco (Wu, et al., 2008) (Table 2.1 and 2. 2).

Research has also been carried out on halo-tolerant (salt tolerant) bacteria and algae as they are believed to be true salt tolerants (Liska, et al., 2004; Kader, et al., 2011). One such halo tolerant green algae is Dunaliella (Liska, et al., 2004) which has been found to have 76 salt induced proteins that aid in tolerance to salinity levels of up to 3M NaCl. In salt stress conditions normal plants limit transpiration in order to retain water, this restricts CO2 uptake and reduces plant photosynthetic productivity. Salt induced proteins in Dunaliella induce CO2 assimilation, which means CO2 is available for normal photosynthesis to take place and it also diversifies its energy resources for glycerol production which acts as an osmotic regulator to buffer the impact of osmotic stress.

More recent research has demonstrated an increase in the salinity tolerance of Maize by means of inoculation with the bacteria Geobacillus caldoylisilyticus IRD (Kader, et al., 2011). Maize plants inoculated with Geobacillus caldoylisilyticus showed higher growth and weight at 350mM of NaCl compared to the non-treated plants. Geobacillus caldoylisilyticus reduces the impact of salt stress in plants by regulating plant physiology, reducing the accumulation of toxic levels of Na+ and Cl- ions. It was also seen that plants inoculated with the bacteria had increased vascular bundles in leaves and decreased in roots.

In the last two decades salicylic acid (SA) has received much attention from scientists as not only does it provide possible solutions to salt tolerance but it also induces resistance against low temperatures, heat stress, toxic metals , pathogens and oxidative damage (Sakhanokho and Kelley, 2009; Zahra, et al., 2010). Salicylic acid

20 is phenolic in nature and acts as signalling proteins or hormone, it is also an endogenous growth regulator of physiological processes in plants such as photosynthesis, growth, ion transmission and absorption. Salicylic acid plays a significant role in the redox reaction across membranes; hence it counteracts ROS (produced due salt stress) negative effects by production of anti-oxidant enzymes which induces oxidative stress, example superoxide dismutase. Salicylic acid has so far been tested in a number of species such as tomato, maize and wheat. Unfortunately, as for most species tested for salinity tolerance, there has been little success in bringing these applications to crop production (Zhu, 2001).

However, some research does offer possible solutions to salinity problems. Sakhanokho and Kelley (2009) have slightly improved salinity tolerance in Hibiscus species by the in vitro application of salicylic acid to Hibiscus acetosella and Hibiscus moscheutos. Both of the species had higher survival when exposed to a salinity level of 0.5mM NaCl after application of salicylic acid. Tomato (lycopersicum esculentum Mill.) plants were treated with 0, 0.5, 1.0 and 1.5 mM of SA and tested against salinity concentrations of 0, 25, 50, 75 and 100 mM. The study showed that plants treated with SA reduced ROS during photosynthesis, increasing the chlorophyll a and b content (Zahra, et al., 2010).

2.2.4 Salinity Testing Various screening methodologies such as field, in vivo, in vitro tissue cultures and hydroponics have been employed in the past by researchers to study the different aspects of salinity tolerance. Of the three screening methodologies, field experiments are advantageous for testing plant salinity tolerance in their natural habitat. However, field testing means there are numerous variables that may affect the performance of giant swamp taro besides salinity; e.g. micro climate, soil fertility, pH, temperature, the amount of water in soil and the light intensity (Yamaguchi and Blumwald, 2005; James and Munns, 2003; Arzani, 2008). To avoid some of the problem of external

21 factors affecting a field experiment, it has been suggested (James and Munns, 2003; Arzani, 2008) that experiments should be done in a controlled environment. Green house (referred to as In vivo in this research) experiments allow us to do whole-plant experiments while controlling several environmental variables.

The majority of in vivo experiments employ the use of gardening pots with holes at the base for drainage. Some researchers prefer additional procedures of rearranging pots to ensure no one pot is excessively exposed or the contrary to sunlight (Nyman et al, 1983). Pots are filled with either potting mix/ soil (Nyman, 1983; Zhao, et al., 2007; Banakar and Ranjbar, 2010) or sand–perlite mixture of 1:1 (Kchaou, et al., 2010; Shaddad, et al., 2010). Plants are watered with various concentrations of Hoagland solution and additional fertilizers depending on plant requirement. When testing the highest salinity levels salt solutions are usually applied in increasing increments instead of one single application in order to acclimatise the plant and avoid toxic shock. In vivo plants may also benefit from periodic additions of fresh water that mimic flushing action of rain. For example Kchaou, (2010) applied 200 ml of de-ionized water weekly to plants in a salinity tolerance study carried out on five olive cultivars.

Peel (2004) employed Ray leach Cone-tainers, with a 70 mm layer of grit to hold enough moisture and silica sand. Silica is an inert medium and prevents accumulation of salts. A 10x10 cm2 square of capillary matting was used to confine sand in the cones and to ensure proper flow of nutrients and treatments. These cones were then arranged in flats of 98 cones each and submerged in nutrient solution and salt treatments as and when required. Although when using in vivo techniques it is possible to control light exposure, difference in soil microclimate and pH to some extent, it is not so easy to control the level of moisture present in the soil, the exposure to wind, cold and heat. These factors that are beyond control in an in vivo experiment but can be controlled in an in vitro system.

Wilhem Roux first established the basic idea of tissue culture in 1885 as an in vitro system when he maintained the medullary plate of an embryonic chicken for a number of days using a warm saline solution (Steinhardt et.al., 1913). Since then tissue culture as an in vitro system has been commonly used for salt tolerance screening (Zhao, et al.,2007; Wu, et al., 2008; Tong, et al., 2010; Yifei, et al., 2009). In an in vitro system, all the environmental and growth medium factors are strictly controlled. Plants are cultured in a specific basal medium, in culture bottles and kept in rooms with controlled duration and intensity of light, temperature and humidity. Since the plants are in culture bottles they are not affected by the outside climate as with green house and infield system.

In an in vitro screening, modified Murashige and Skoog medium (1962) has been readily used as the basal medium, to which various NaCl concentrations are added (Nyman, 1983; D’Antonio and Weber, 1999; Hady.A, 2006). The cultures are then placed in a strictly controlled environment in labs. The use of in vitro techniques allows for salinity screening of large numbers of genotypes. This is due to the relatively short time taken for growth and multiplication in tissues culture compared to conventional screening. It is also to some extent comparable with field experiments and requires less area in which to conduct the experiment compared to field experiments (Arzani, 2008).

Yet another form of screening methodology used is a hydroponic system. One such study involving the use of hydroponics was carried out by Rush and Epstein in 1981 to study the relation of minerals such as potassium, sodium, and chloride to wild halophytic and domestic salt-sensitive tomato species. Similarly, hydroponics has been employed in present day, such as in the study of leaf growth in salt stress in barley (Fricke and Peters, 2002) and many others. Both the experiments applied salt treatments at incremental basis up to the highest salinity level of the experiment. These increments were undertaken at periods of 2-7 days.

Therefore, with the complexities of the impacts of increased soil salinity on plants and the need for development of salt tolerant cultivars, researchers have adopted and modified these salt screening methodologies. Adjusting their choice of procedure to the type of crop, level of growth assessment and extent of control over external variables.

2.3 GIANT SWAMP TARO (Cyrtosperma merkusii)

The increase in the ground water salinity level poses a threat to the food security system of the atoll island communities. Crop adaptation to increasing salinity levels presents one way of curbing the problem. Due to the significant role Giant swamp taro (Cyrtosperma merkusii) plays in food security and lives of the atoll island communities, it has been selected for research and development as a climate ready salt tolerant crop (Figure 2.2).

By Shiwangni Rao

Figure 2. 2 Giant Swamp taro Cyrtosperma merkusii

By Shiwangni Rao

Figure 2.3 (left) Giant Swamp taro corm (from http://www.pbase.com/ jamato8/image/ 127855767). Figure 2.4 (right) Giant Swamp taro farm in FSM.

Giant swamp taro is a large herbaceous perennial plant that can reach up to 5 metres in height and the corm can weigh up to 10 - 20 kg when harvested within a year or two (Ivancic, 1992) (Figure 2.4). Some have been seen to weigh up to 100 - 120 kg (Figure 2.3). However these values vary according to the genotype and with maturity, especially in the larger cultivars (Dunn, 1976; Covich, 2006). Vickers (1982) states giant swamp taro is the largest root crop with an edible corm. It has large leaf blades reaching up to 1m in width and quite similar in shape to Alocasia macrorrhizus that is it is saggitate to hastate with two long acute basal lobes (Manner, 2009). Some giant swamp taros have pricks on the petioles and the colours may vary according to the cultivar. While others have reduced leaves, ‘cataphylls’ on the underside of the leaf blade (Ivancic, 1992.; Iese, 2005; Cyrtosperma merkusii, 2006). The inflorescence is a cylindrical spadix with a large spathe (Hather and Weisler, 2000; Iese, 2005; Manner, 2009). The many cultivars of the giant swamp taro are generally classified into two groups in Kiribati, which have been adopted for the purpose of this study. Group one, the “Ikaraoi” is larger, takes much longer to mature (2-3 years) and has five or less suckers (shoots emerging from corm). The

“Katutu” is the smaller cultivar group; it matures much earlier than the Ikaraoi (within 6 months) and has more than five suckers (Iese, 2005; Manner, 2009).

Plantae

Tracheophyta Subphylum: Euphyllophytina Liliopisa Subclass: Aridae

Arales Superorder: Aranae

Araceae Subfamily: Lasioideae Cyrtosperma merkusii

Figure 2.5. Phylogenic classification of Giant Swamp taro. (Cyrtosperma merkusii, 2006)

The genus name Cyrtosperma is derived from the Latin for “Curved Seed” (Mayo, et al., 1997) (Figure2.5). The species name on the other hand is quite ambiguous, as the giant swamp taro has been confused with other taro species in the past, which has led to it being called by a number of species names. Consensus now favours Cyrtosperma merkusii. From the limited documentation available there are 18 known cultivars in Tuvalu (Iese, 2005), about 11 in Kiribati and 60 in Federated States of Micronesia (Englberger, et al., 2003.; Englberger, et al., 2005.; Englberger, et al., 2007) plus many unknowns. Iese (2005) used 27 characterization descriptors that were discussed and articulated with 21 farmers from Fiji, Pohnpei Federated States of Micronesia and Tuvalu. He chose farmers from these three countries to provide a better knowledge base of the diverse range of cultivars grown and known on the three types of islands. Tuvalu represented atolls that are low lying, Pohnpei for an intermediate between coralline and volcanic islands and the highlands of Rewa province in Fiji for volcanic islands.

Giant swamp taro (Cyrtosperma merkusii) cultivation can be quite a strenuous task, depending on soil fertility. In the highlands of Fiji, the soil is quite rich, hence little

26 work is required. However, on atolls such as Kiribati, Tuvalu and Federated States of Micronesia (FSM) the plants require much nurturing. Cultivation on the other atoll communities follows a similar pattern of cultivation with their own modifications.

Giant swamp taro is often referred to as being salt tolerant (Lambert, 1982; Brandburry and Holloway, 1988; Kazutaka and Michia., 2003; Covich, 2006; Deenik and Yost, 2006). However, the degree to which it is salt tolerant remains uncertain (Nyman, 1983; Webb, 2007). Webb (2007) in his survey showed that giant swamp taro grew well in soils with an electrical conductivity of 1000µScm-1 (0.67ppt) or less. It could also tolerate 2000µScm-1 (1.34ppt) or less but electrical conductivity between 2000µS cm-1 and 3000 µS cm-1 (2.01ppt) was fatal to the plant, meaning that it grew well in fresh and mildly salty water but died in brackish water. Wiens (1962) proposed that water in the giant swamp taro pits in Kiribati was particularly fresh and sometimes fresher than well water. Dunn (1976) supports giant swamp taro being salt tolerant. He found the tolerance to be within the range of 2-3ppt salinity as did Brown (2000) who stated that giant swamp taro grew quite well in brackish water. A 2004 Agroforestry in Micronesia report states that Cyrtosperma merkusii grew well in salty pits. The results from these studies suggest that the extent to which giant swamp taro is salt tolerant depends on the cultivar.

2.3.1 Origin Taro is a major staple of the Pacific islands. There are four types of taro which are normally found in the Pacific. These include the common taro Colocasia esculenta, Xanthosoma sagittifolium, Alocasia macrorrhizos and the giant swamp taro Cyrtosperma merkusii. Of these four types of taro Cyrtosperma is the largest, reaching up to 5 metres in height (Dunn, 1976; Hather, 2000; Iese, 2005) and takes the longest to reach maturity. It is also known for its hardy qualities of surviving in atoll environments; hence this spectacular crop has been found to be cultivated in large numbers in the atoll islands.

Giant swamp taro is one of the root crops that have spread across the Pacific reaching as far out as the Makatea Island on the northwest of Henderson island in the Tuamotu Archipelago (Hather, 2000). Some researchers have concluded Cyrtosperma to be of an Indonesian or Indo-Malayan origin. Lebot (1992) and Hay (1990) argue that the high lands of Papua New Guinea could be a place of origin (Figure 2.6). On the other hand, using archaeobotanical analysis Hather (2000) found that “…Cyrtosperma was an aboriginal introduction across Polynesia except for New Zealand and Easter Island where climate plus cultural preferences may have discouraged its growth...” he also found that Cyrtosperma merkusii was present as far back as 1451 A.D. While the uncertainty of origin may still exist, the Indo- Malayan region is certainly the region that holds the greatest diversity of the root crop (Bradburry, 1988; Iese, 2005).

Figure 2.6 .Map of Malesia, including the possible origins of Giant swamp Taro Cyrtosperma merkusii. (wiki/File:Malesia.png).

2.3.2 Current Distribution

Figure 2.7 Map of the cultural spheres and the current distribution of giant swamp taro expect for New Zealand (ANU Cartographic Services, 2008).

With the lapse of time from 1451 A.D to the twenty first century, modernization has played a pivotal role in shaping the present trends in lifestyle preferences from technology to traditions. Giant swamp taro has fallen victim to modernization. It once flourished in the Indo-Pacific region and was seen as a major root crop and a local food staple but it is now being replaced by western foods at an accelerating pace. Currently the distribution of giant swamp taro may still be the same as it was in the past but the intensity of cultivation / population has dropped drastically (Iese, 2005; Talia, 2009) (Figure 2.7). However, the full extent of this reduction is poorly characterised, along with the many probable causes of it, highlighting a need for further research and investigation.

According to Hather (2000) giant swamp taro is drought and salt tolerant in comparison to other root crops such as the common taro and yams. Due to these qualities modern cultivation is focused around areas where other crops do not do so well such as on atolls. Atoll soils lack the minerals and texture of good soil, such as those found on the volcanic islands. Atoll soils are dry and slightly saline and have a thin organic top layer made from the decomposition of fallen vegetation. The Micronesian and Western Pacific region are where most of these low lying atolls lay, hence it is this area that has the highest giant swamp taro population (Manner, 2009). The isolation of atoll islands from the main land and continents has resulted in cultivar divergence such as in the atolls of Federated States of Micronesia where approximately 60 cultivars exists. Giant swamp taro is also present on volcanic islands where it grows wild or with limited cultivation such as in Fiji around the Rewa province, where it is grown because the area is frequently flooded. The common taro which is more preferred in Fiji cannot survive (Iese, 2005) this frequent flooding.

2.3.3 Physiology

Giant swamp taro thrives in a tropical climate and tolerates occasional dry seasons with variation in rainfall. Manner (2009) notes it easily tolerates temperatures of 35- 38°C down to 15°C monthly mean temperature. In Yap it is planted in Mesei type of soil, it’s a soil with a pH of 4.5-5.5 which is high in organic matter and water logged, resulting in a mucky dark loamy soil that is overlaid with soil alluvial in origin. Dechel is another soil type similar to Mesei and Dechel is a type of soil that has a pH of 5.1-7.3 and is used in Guam for giant swamp taro cultivation. Ngedebus is another type of soil is used in Ulithi atoll, this is found in marshy land depressions with a gravel, loam and sand mixture resulting in pH of 6.6-8.4.

Farmers on Funafuti in Tuvalu have found that a particular cultivar the ‘Pulaka Kula’ is better adapted to saline conditions than other cultivars found on the island. According to farmers during high tides sea water causes the pulaka (Tuvaluan name for giant swamp taro) in the pits to wilt but the pulaka kula is unaffected. It also needs less attention compared to the Tuvaluan cultivar ‘Ikaraoi’ which has the most valuable corm. 2.3.3.1 Cultivation

Figure 2.8. (above, left) Sunken cultivation in Tuvalu, Nanumea.

By Shiwangni Rao

By Shiwangni Rao By: Tevita Kete

Figure 2.9 (above left).Coconut Cocus nucifera leaf woven bottomless basket in Funafuti, Tuvalu. Figure 2.10 (Right) Giant swamp taro cultivation in Fiji.

Figure 2.11 Concrete cement pit Taro cultivation on Tuvalu in Funafuti.

By Shiwangni Rao

Cultivation of the crop varies from island to island and also among cultivars and groups of Ikaraoi and Katutu. Very few cultivars of giant swamp taro produce viable seeds; hence farmers prefer vegetative propagation. One method of vegetative propagation is the planting of cuttings trimmed from harvested plants, with around a 30 cm petiole and a reasonable bit of corm attached (Manner, 2009). Another method is to plant suckers that emerge from the corm. The petioles of the suckers are cut in a downward diagonal direction for better growth of the plant (Iese, 2005) and it is believed by some Tuvaluan farmers that if the plant is left unattended it will produce suckers and if attended it does not.

Giant swamp taro is grown in natural or manmade swamp depressions. Many of the manmade swamps present on the atolls today were excavated by the first settlers on the island (Thaman, 2002; Iese, 2005) (Figure 2.8). The hard coralline soil was dug with primitive tools (for instance a digging stick) and these swampy pits were dug down until the ground water lens was reached. Then mulch and compost was applied to create an organic soil. In Kiribati bottomless baskets are woven using pandanus (Pandanus tectorious) or coconut (Cocus nucifera) leaves and the giant swamp taro is planted in these with constant composting. In Tuvalu farmers dig holes in the pulaka pits approximately 20-30 cm wide and 15-30 cm deep (Figure 2.8). This hole is referred to as ‘Knowledge of the hole’ or in Tuvaluan ‘Logo o tepoko” by some

Tuvaluan farmers (Iese, 2005). Young trimmed suckers are planted in this hole and tied to a stick called the ‘tokai’ for support.

There are two methods of composting seen in Tuvalu in regards to what is being used as compost and the timing of the compost application, both of which are regarded essential by the farmers for a good, high quality. The first and most popular cultivation method involves applying compost to the young plant when planting using the fresh leaves of pukavai (Pisonia grandis). However, a limited amount of compost is applied, as too much may cause overheating and eventual death of plant (Iese, 2005). This is why some farmers prefer the second method where compost is applied only after the first leaf appears. The leaves of the pukavai decay easily in about two weeks and support rapid growth of the plant. When decaying it also produces a pungent smell which deters insects and pests (Iese, 2005). After the emergence of three to four leaves the next compost is applied; this may contain cuttings of other plants including pukavai, gasu (Scaveola taccada), kanava (Cordia subcordata), and puavao (Guettarada speciosa) (Thaman, 2002; Iese, 2005). This compost along with some soil for air circulation is applied at a distance from the young plants, as digging and applying at the base would damage the tender young roots. After this, farmers check the decay of the compost by stamping around the plant; if it feels cushioned it means that the compost has decayed (Iese, 2005). Also when the base of the young plant starts to show it is taken as an indication that the last compost has decayed and more needs to be applied. After seven months of composting with the leaves of the above-stated plants, coconut husks and green leaves are then employed, as it turns the soil dark and the giant swamp taro corm becomes tasty with a good texture (Iese, 2005).

It was observed that similar to cultivation in Kiribati some farmers had woven coconut leaves in a circle around the plant and then applied compost to it (Figure 2.9). This along with platted mud walls around the plant is done to hold the compost, when the young plants have six to seven leaves. Using plaited coconut leaves also

33 supplied compost and a new one is plaited every time the old one decays. As stated earlier the mulch of giant swamp taro consists of surrounding vegetation, cuttings of leaves, soil, pumice and general compost. However, the contents of the mulch vary between farmer families, that have their own secret recipes, times and chants that are used during mulching; this knowledge is rarely revealed due to fear of competition (Thaman, 2002). Many forms of compost timing are employed by farmers. This includes taking into account moon phases, the number of leaves sprouting and when the base of the plant starts to show the trimming at neap and spring tides, and when the soil around the pit starts to feel soft (Iese, 2005).

In Guam farmers have the dechel cultivation system, where land is cleared and planted with the setts (cuttings of the stem with a bit of corm) or suckers of previous harvest (Manner, 2009). In Palau the Mesei system is employed where the soil is mulched and overturned; the outer island of the Federated States of Micronesia have cultivation methods similar to the Kiribati and Tuvalu. However on the main land Pohnpei, cultivation is less intense similar to the highlands of Fiji. The most simple and easiest cultivation of giant swamp taro can be found in the Rewa province in Fiji where holes are dug and a bunch of suckers tied with giant swamp taro petiole strip is planted ensuring that at least one sucker survives (Figure 2.10). This also gives larger corms and gives protection to the plants during flooding (Iese, 2005). Apart from this the most recent advancement seen in Tuvalu in giant swamp taro cultivation is planting it in concrete cement pits as this ensures that salinity in the soil does not affect the plant (Figure 2.11). The timing of planting giant swamp taro also varies. Atoll island farmers such as on Tuvalu prefer to plant when the tide is very low; usually during the cool and dry seasons when the pits are not flooded. Fijian farmers prefer to plant during the rainy season when the soil is softer (months of October to April).

2.3.3.2 Maturity and Harvest

Giant swamp taro in general matures in 2-4 yrs. However, the ‘Katutu’ cultivar of Kiribati and the Chuukese cultivar ‘Onou maram’ (Manner, 2009), can mature in six months. Plant height is directly related to its maturity age, in the sense that the longer a plant cultivar takes to mature the taller it grows and vice versa. In Tuvalu, farmers count the number of flowers as an indication of maturity; after seven flowers the plant is said to be mature and ready for harvest (Iese, 2005). Similarly in Yap, maturity depends on plant flowering, which begins in the second year of growth. Also when emerging leaves are smaller than usual and when the corm starts to rise above the ground (giant swamp taro corm grows both up and downwards) (Manner, 2009). In Kiribati, maturity is related to growth stage and time (Table 2.3). In Fiji harvesting takes place when suckers begin to scatter, growth is reduced and when all plant leaves turn yellowish. It is also believed by the Fijian farmers that the best time to harvest giant swamp taro is between June to September. At the time of harvest corm can weigh from 15-20 kg (Hather, 2000) to 100-120 kg when left for a long time in ground (Dunn, 1976; Iese, 2005; Manner, 2009) (Figure 2.12 and 2.13). Giant swamp taro also gives a good yield becasuae only a limited number of pests and diseases is know to affect the plant. Except for the rarely found Dasheen Mosaic Virus (DMV) which affects the whole plant, the rest of the pests are mostly nematodes that burrow into the corm (Table 2.4)

Figure 2.12 Giant swamp taro harvested corms.(from http://atinleparang.blo gspot.com/2009 _11_01_archive.html)

Table 2.3 Kiribati description of giant swamp taro growth stages (Manner, 2009). Growth stage Time Description Te Kunei 9 months This is the harvest time for the Katutu cultivar group. Corm is approximately half the size of forearm in length; at this size the corm is tender. Te 3 years Corm is the size of a full forearm in length namatanibura and fully matured. However some are very bitter at this stage. Etan 5 years Corm three quarters of an arm in length. tenamatanibura Te anga 7 years Corm is a full arm's length in size and used in certain rituals. Te bonaua 10+ Corm is up to breastbone in length and the corm becomes hard. This corm is mainly used as presentation on special occasions such as by the groom’s family to the brides during weddings.

Figure 2.13 Giant swamp taro corms (from http://bild- art.de/kpress/)

Table 2.4 List of pests, diseases and their impacts on giant swamp taro (Bradburry, 1988; Iese.V, 2005; Manner, 2009). Pest/diseases Impact on plant

Papauana huebneri Beetle burrows and makes the corm susceptible to other parasitic organism, which can result in plant death. Glover Ahis gossypii eat leaves of giant swamp taro Mealy bugs Pseudococcus, eat leaves of giant swamp taro Nr.adoniumL. Ferrisiana virgata Ck11 eat leaves of giant swamp taro Bag worm (unidentified) eat leaves of giant swamp taro Hippotion sp. Caterpillar, leaf eating Spodotera Litura Caterpillar, leaf eating Theretra pinastrina Caterpillar, leaf eating Criconemella denoudeni Nematode that burrows into the corm resulting in rot C. onoesis Nematode that burrows into the corm resulting in rot Helicotylenchus dihystera Nematode that burrows into the corm resulting in rot Meloidogyne Sp Nematode that burrows into the corm resulting in rot Pratylenchus coffeae Nematode that burrows into the corm resulting in rot Radopholus similis Nematode that burrows into the corm resulting in rot Pythium rot affects corm Dasheen mosaic virus affect whole plant

2.3.4 Morphology

Giant swamp taro produces bright and conspicuous flowers in many of its cultivars (Figure 2.14). The inflorescences consist of a reproductive spadix, a colourful leaf like spathe, seeds and flower stalk. The spadix is an elongated inflorescence and in certain cultivars is fertile producing seeds upon maturity, while in others stays sterile (Figure 2.15 and 2.16). The fertility rate of the inflorescences depends on the cultivar. Usually fertility rate is quite low and many of the seeds produced are sterile; hence many farmers use vegetative propagation (Figure 2.17). The spadix is approximately 20-25cm in length, is hermaphroditic. It is unlike the common taro Colocasia esculenta which has the female component at the top and the male component at the bottom usually enclosed, giant swamp taro has both the male and female component in each flower of the inflorescent and the spadix is usually exposed. The pollen produced on these tiny anthers varies in colour from white yellow to reddish orange (section 3.5).

Spadix

Spathe

By Shiwangni Rao By Shiwangni Rao

Figure 2.14 (left) labelled giant swamp taro flower. Figure 2.15 (middle left) young spadix and flower.

By Shiwangni Rao By Shiwangni Rao

Figure 2.16 (middle right) Seeded berries on spadix. Figure 2.17 (right) Mature spadix.

The spathe is a thick leaf like colourful covering that partially envelopes the spadix. The colour of the spadix and spathe depends on the cultivar of giant swamp taro ranging from green to yellowish green, pink to maroon at maturity. Pwh weitata a Federated States of Micronesia cultivar has a unique spathe as it turns from yellow to green at maturity which is rare (Figure 2.18 and 2.19). The flower stalk also has its own colour usually intermediate between spathe and petiole colours.

Figure 2.18 (left) young Pwh weitata flower. Figure 2.19 Mature Pwh weitata flower. By Shiwangni Rao By Shiwangni Rao

The leaf petioles have a diverse range of epidermis texture; they range from smooth to varying degrees of spines. Petiole colours range from maroon, yellowish pink to dark green. And have three types of petiole neck shape namely, straight, curved and swan neck. Leaf height and size depends on cultivar and maturity of the plant (Figure 2.20).

By Shiwangni Rao

Figure 2.20 (left) leaves of giant swamp taro. Figure 2.21 (right) corm of giant swamp taro.

By Shiwangni Rao

As in all taro the stem of the giant swamp taro is absent (acaulescent), but it has a very valuable corm (Figure 2.21). The swollen corm, can weigh from 10-150 kg depending once again on cultivar and maturity. The corm flesh may be completely white, pink to yellow in colour with brown to yellow corm fibres. The corm is slightly harder than the other three taros and the degree of hardness depends on the cultivars and length of stay the corm has been in ground. The Tuvaluan cultivars of smooth and thorny suwetena and the Mwhng seri of FSM are three cultivars known for the softness of their corms. The giant swamp taro corms have a small root system, with root fibres that are 5mm in diameter, short and thick. These roots are concentrated just below the leaf base area

2.3.5 Cultivar Descriptor List As stated earlier little research has been carried out on giant swamp taro, hence limited knowledge is available for development of the crop. In an attempt to broaden the knowledge on giant swamp taro a detailed cultivar descriptor list was developed (Table 2.5) by improving on the list prepared by Iese (2005). Iese (2005) had a number of traits and characteristics that were overly variable in response to environmental conditions and hence had low utility in distinguishing between cultivars. The revised descriptor list was developed during a two day workshop conducted in Pohnpei, Federated States of Micronesia with the help of local experts.

The workshop consisted of 37 participants, 27 farmers and 10 agriculture field technicians of Pohnpei Agriculture Department. There was an equal distribution of ages ranging from young to old framers, while a 3:1 gender ration for men to women was present in the workshop. Using a giant swamp taro descriptor list composed by Iese (2005) consisting of 27 descriptors, along with IPGRI (2007) full descriptor list for Taro Colocasia esculenta as a guide for characterisation a detailed draft descriptor list was prepared. This was then presented on a PowerPoint presentation and explained to the participants with translations in Pohnpeian from the Pohnpei Chief Agriculture Officer Mr. Adelino Lorens. Through an open discussion the participants at the workshop, worked through the various descriptors to select the most pertinent descriptors for giant swamp taro. There was common agreement among the participants for all the selected descriptors across both age range and gender. Mr. Adelino Lorens on previous trips to the outer islands of Federated States of Micronesia had collected approximately 50 proposed cultivars and had planted these in the Pilot farm in Pohnpei. The Pilot farm presented a collection of all Federated States of Micronesia cultivars; hence it was used for characterization of the giant swamp taro cultivars using the developed descriptor list with the help of four agriculture field technicians.

Table 2.5 Revised descriptor list No. Trait Variability 1.1 Plant span/ spread 1.Narrow Plant Span (<50cm) 2. Medium (50-100cm) Height 3. Large (>100cm)

1.2 Plant height at 1.short(3-4ft) maturity 2.medium(5- 10ft) Direct 3.long(>10ft) suckers

1.3 Number of suckers 1.many(<10) (direct shoots) 2.few(5-10) 3.less (>5)

1.3 Number of suckers 1.hastate (direct shoots) (Having the shape of an arrowhead but with the basal lobes pointing outward at right angles) Hastate 2.peltate Peltated (Having a flat circular structure attached to a stalk near the centre, rather than at or near the margin; shield- shaped

2.2 Spread of leaf lobes 1.Overlapping Wavy Margin Entire Margin 2. Acute angles (<45º) 3.Right angles (90º)

2.3 Leaf blade margin 1.Entire (not Overlapping lobes wavy) Acute lobes 2.Undulate (wavy) 3. Sinuate (Very wavy) 2.4 Leaf blade colour 1.Whitis 2.Yellow/ Yellow green 3. Light green 4. Dark green 5. Pinkish green 6. Reddish green 7. Purplish 8. Blackish 2.5 Leaf lamina 1.Absent appendages/ 2. Present cataphylls Cataphylls

2.6 Leaf main vein colour 1.Whitis 2.Yellow/ Yellow green 3. Light green Leaf main 4. Dark green vein 5. Pinkish green 6. Reddish green 7. Purplish 8. Blackish

2.7 Leaf arrangement(la) 1.Absent 2. Present Clockwise

Counter-clockwise

3.1 Colour of top third 1.Whitis (P/c/t/third) 2.Yellow/ Yellow green 3. Light green Top 4. Dark green 5. Pinkish Middle green 6. Reddish green 7. Purplish Bottom 8. Blackish

2.8 Number of leaves(nol) 1.few(<5) 2.normal (5- 10) 3.many (>10)

2.9 Leaf lamina length : describe width ratio

Length

Width

2.1 Petiole junction describe pattern

3 Petiole(leaf stalk) 3.1 Colour of top third 1.Whitis (P/c/t/third) 2.Yellow/ Yellow green 3. Light green 4. Dark green 5. Pinkish green 6. Reddish green 7. Purplish 8. Blackish 3.2 Colour of middle same as above third (P/c/m/third) 3.3 Colour of lower third same as above (P/c/l/third) 3.4 Petiole stripes 1.Absent 2. Present Petiole with fine printed stripes

3.5 Petiole 1.straight shape(top)(ps(top) 2.curved Swan neck 3.swan’s neck

Straight neck

Curved neck

3.6 Petiole 1.Absent throne/spine/(pthorns) 2. Present

3.7 Petiole spine 1.green color(p/t/(col) 2.dark green 3.yellow Thorns/ 4.red Spines 5.pink 6.purple

3.8 Spine size(p/t/size) 1.short(<2mm) 2.medium (2- 3mm) 3.long (3- 4mm) 4 Inflorescence/ Flower 4.1 Flower formation 1.Absent 2. Present

4.2 flower stalk color 1.Whitis 2.Yellow/ Stalk Yellow green 3. Light green 4. Dark green 5. Pinkish Spadix green 6. Reddish green Spathe 7. Purplish 8. Blackish 4.3 spathe (flower cover) Describe color top/ bottom and young/old 4.4 Spadix/ pollen color Same as 4.2 options 4.5 Berries color Same as 4.2 options

4.6 Seeds Viability (sv) 1.viable Berries/ seeds (grow) 2.non-viable Indicator of fertility (don't grow 4.7 Male portion of 1.Enclosed flower 2. Exposed

4.8 Fertility of the female 1.none part of the 2. Low (<40% inflorescence fertile flowers) 3. Intermediate (<80%) 4. High (almost 100%) 5 Corm 5.1 Corm Size 1.Small 2.Medium 3.Large 5.2 Corm Cortex Color 1.White 2. Yellow 3. Orange 4. Pink 5. Red 6. Purple Central part 7. Other Roots 5.4 Corm flesh color same as above central part 5.5 Corm flesh Fiber same as above color Cortex 6 Roots Describe

7 Taste 1.Very Hard 2. Itchy/ irritating 3. Good 4.Very good

8 Special Describe characteristics eg. Drought or salinity tolerance

2.3.7 Utilization

‘…for people of Tuvalu, pulaka appears to be the most important food crop next to the coconut....’ (Dunn,1976). Iese (2005) citing Nakon (2001) states giant swamp taro cultivation yearly was 8 million calories per hectare which is equal to the production of true taro in Oceania.

Apart from being consumed for its nutritious corm the giant swamp taro has other very essential uses as well. It has been woven into the island traditions and culture as a precious gift to be presented as an offering to the chief or chiefly person’s on special occasions such as family gatherings or marriage ceremonies. In Tuvalu the Kaupule (Island Councils) call for Fuauli (medium sized giant swamp taro) for the Fakaala (gathering such as weddings, funerals, village meeting) requiring every family to provide Pulaka. Nafa competitions are one of the competitions the people of Tuvalu look forward to. Partners for the competition are announced at the beginning of the year and they work hard in the intervening period to produce the best pulaka. The winner is announced according to the highest corm yield (Dunn , 1976; Iese, 2005; Manner, 2009).

Giant swamp taro is a very prestigious plant for Micronesians and especially for Tuvaluan and Kiribati people. It is a great contributor to the food security of the atolls as well as to traditions and customs. Apart from this its valuable corm, giant swamp taro has many other uses of its leaves. This is used to wrap food, cover stored ripening fruits, used as an umbrella, drinking bowls, dancing shirts ‘Titi’ for Tuvaluans, and garlands for Rotumans. Its petiole is used for fertilizer, mat weaving and medicines (Dunn, 1976; Hather, 2000; Englberger, 2005; Iese, 2005; Manner, 2009; Talia, 2009)

Table 2.6 Local recipes of giant swamp taro

Tuvaluan traditional dishes: Tao – peeled and earth oven baked pulaka Kofuga pulaka- cleaned pulaka is cooked in coconut cream wrapped in banana leaves. Faalifu- cleaned and boiled pulaka in coconut cream Taufagogo- pulaka is cleaned and cut into small pieces, this is then placed inside a scraped green coconut shell with some coconut cream and Toddi and the baked in an earth oven. Lipilipi/ tokotokoi- small pieces of baked pulaka is mixed with boiled coconut cream Fakapapulaka/ Tulolo – boiled pulaka is pounded until smooth then mixed with Toddi Fekei- grated pulaka is mixed with Toddi and cooked wrapped in pulaka leaves. This is then mixed with boiled coconut cream Nepo- scraped coconut and boiled pulaka is pound together until smooth and then mix with water first followed by toddy Solo/ Mafu- grated pulaka cooked covered with banana leaves ValuValu pulaka- grated pulaka is mixed with water and Toddi placed inside a scraped green coconut and baked. ______(Iese, 2005)

Marshallese Traditional Dishes; Wūden- giant swamp taro is mixed with nuts and grated coconut or cooked and boiled banana, breadfruit and Colocasia taro. Jebwater- mixed with coconut milk and grated Taro, baked in an oven wrapped in taro leaves. Totaimon – mixed with coconut oil and sap and grated Colocasia taro. KōmāKij – with mashed potatoes or Colocasia taro. Jukjuk- mixed with coconut and pounded Colocasia taro. ______(Manner, 2009)

Federated States of Micronesia Dishes; Women in FSM have learnt to make many exquisite dishes from giant swamp taro; Lihili - taro is boiled and mashed with a special pounding stone, and coconut milk is added to it. Mwael/ Piaia - Taro boiled with coconut milk. Rotama - Raw taro grated and mixed with cassava starch. Taro chips - Taro is cut into thin slices then fried in oil, salt is added on top according to taste. Taro donuts - Taro is pounded and then sugar is added. This mixture is then shaped into donuts and fried. Taro flour - Finely grated taro is sundried till it turns floury. This can then be used as normal flour for cooking and baking. Taro cooked in earth oven is another popular way of having taro in FSM.

2.3.8 Nutrition Giant swamp taro like other root crops has high starch and carbohydrate content and low protein. It has the second lowest protein content (0.8% dry weight) in comparison to the other three taro, of which the common taro has the highest protein content of 4.5% of dry weight (Iese, 2005). Also in comparison to the other three taro species giant swamp taro has a higher content of Vitamin B1, C, calcium and B- carotene which is directly related to the uncooked corm colour; the darker the shade of yellow for the corm the higher the carotene content (Englberger, 2003; Englberger, 2005; Iese, 2005). Apart from this, giant swamp taro also has mineral content as shown in table 2.7 and 2.8.

Giant swamp taro has a high content of oxalic acid crystals which causes irritation in the mouth and throat; this is 10 times higher than common taro, sweet potatoes and cassava. However, these crystals can be easily removed by proper cooking.

Table 2.7 Giant swamp taro nutritional value (SPC, 2006) Food Item K cal Fibre (g) Calcium Iron (mg) Zinc ß carotene Thimian Vitamin (mg) (mg) equiv. (µg) (mg) C (mg) Cyrtosperma corm 72 2.5 165 0.6 1.9 27 0.02 7.9 colour unspecific

-white/ cream coloured na na na na na 55-300 na na (Community, 2006)3,4 -yellow-coloured3-5 na na 240-1440 1.4-3.6 4.1-63 460-4486 na na

Table 2.8 Giant swamp taro cultivar nutritional value (Englberger, 2005) Food Sample a Nb Iron Zn Ca Mg P Mn Cu Na K Giant swamp taro, e3 0.1 7 103 24.7 15.5 1.6 0.2 52.8 130 fanal Giant swamp taro, e3 0.20 4.8 137 23.7 16.7 2.2 0.4 46.4 141 mwashei

2.3.9 Conclusion Increasing soil salinity is a pressing issue that needs to be addressed, especially with small islands where the rise in sea level is expected to cause salt water intrusion in the island’s fresh ground water lens. Crop failure due to increased soil salinity is reported to be an increasing problem in the Pacific region in atolls such as in Tuvalu and Kiribati as well as in other parts of the world, where prolonged drought and improper agricultural practices have degraded ground water quality (Toshio et al. 2005).

Giant swamp taro does have variation in their genome despite being mostly cultivated vegetatively by farmers (Iese, 2005). Current genetic diversity of the giant swamp taro should be preserved. It is essential to conserve cultivars, along with proper documentation and data collection so that invaluable information about these cultivars and the cultivars themselves will not be lost. Currently a handful of cultivars are threatened as many of these have evolved in isolation on the islands and are endemic. In addition, not all the cultivars are cultivated; selection is more to do with preference of taste, use and ease of cultivation (Iese, 2005). Apart from effects of climate change and globalization the changing preference of farmers may cause the disappearance of some of the cultivars.

Conserving crop biodiversity is one way of preserving the genetic diversity needed for future breeding efforts to adapt to rising soil salinity. At present the gremplasam centres around the world are working towards conserving genetic resources and finding solutions to the current threats to agricultural sectors. The Centre for Pacific Crops and Trees (CePaCT) established under SPC in Fiji is one such germplasm centre and is currently responsible for the conservation, duplication and documentation of Pacific crops and trees including giant swamp taro.

Lastly, apart from being a food crop giant swamp taro, as stated earlier, has been incorporated into the island culture and traditions, which makes this particular crop of major significance for Pacific island nations. If it is lost they will not only lose genetic diversity but also a big portion of their culture, traditions and part of their identity.

This research aims to establish a rapid salt tolerance screening methodology for giant swamp taro and to broaden its knowledge base that will prove helpful to researches and farmers in selecting salt tolerant cultivars. This is achieved by employing in vitro and in vivo techniques on two groups of cultivars of giant swamp taro, the larger cultivar group Ikraoi and the smaller cultivar group Katutu. In doing so this thesis attempts to answer questions such as “What is the current salinity of the ground water lens of atoll islands in Tuvalu?’, ‘Is Giant swamp taro truly salt tolerant?’, “what amounts of salinity can it tolerate’.

3.0 TUVALU GROUND WATER FIELD STUDY

3.1 INTRODUCTION

Atolls have come about due to ages of deposition of unconsolidated carbonate materials on the relics of karst limestone reef, atop volcanic craters (White, 2010). Where rainfall is adequate these Holocene deposits give rise to a freshwater lens, known as the ‘Ghyben-Herzberg lens’ (Dunn, 1976; Rozell, 2007; Woodraffe, 1989). In contrast, according to White (2010) the lower boundary between the fresh water and sea water is not the more or less classical lens shape given by the ‘Ghyben- Herzberg model’. Rather a wide transition or mixing zone where the ground water salinity increases with depth from freshwater to seawater. A well-developed Ghyben- Herzberg lens will have salinity that is in standard acceptance with drinking water guidelines of the World Health Organization, which is 250mg of chloride ion per litre of water or its respective electrical conductivity (White, 2010).

This fresh water lens is very fragile, as stated in the International Panel on Climate Change (IPCC) 2007 assessment report (Mimura, 2007).“…Owing to factors of limited size, availability, geology and topography water resources in small islands are extremely vulnerable to changes and variations in climate, especially in rainfall…”.

Figure 3.1. Giant Swamp Taro (Cyrtosperma merkusii) or Pulaka

By Shiwangni Rao

Since the freshwater lens consists of a transition zone, there is some degree of mixing taking place of sea and fresh water. However, the question is will the effects of climate change increase this mixing, causing salt water intrusion and a resulting rise in salinity level. In the case of Tuvalu, the people fear the country’s food security is being greatly threatened by the effects of climate change currently and may worsen in the future. Tuvaluan farmers suspect that salt water intrusion due to the increase in sea level is causing the decline in their crop production. Farmers claim taro is not able to tolerate the high salinities and dies out. However, these are mostly assumptions by the locals, and unfortunately there has been very little research carried out on the issue except for the study done by Webb (2007), which gives a brief account of the ground water salinity in taro pits for a number of islands on the nine atolls of Tuvalu, namely Nanumaga 744µS/cm, Nanumea 608µS/cm, Niutao 471µS/cm, Nui 209µS/cm, Funafuti 3774µS/cm and Nukulaelae 1236µS/cm.

One of the local food crops that are feared of being lost due to this expected increase in ground water salinity by the locals is the Giant swamp taro (Cyrtosperma merkusii) or ‘Pulaka’ in Tuvaluan (Figure 3.1). Pulaka not only plays a large role in the atoll communities’ food security system, but is also significant in their cultural and traditional systems. The plant depends heavily on the islands’ ground water supply, as it is cultivated in swampy areas. Manmade swamps are created digging a big pit until the ground water is reached (Dunn, 1976; Thaman, 2002; Iese, 2005; Manner, 2009) (Figure 3.2). In other words, the giant swamp taro is cultivated in a window dug into the ground water lens. Considering the concerns of the farmers, if there is an increase in the ground water salinity these taro plants may be affected.

Apart from the suspected seawater intrusion as a result of sea level rise, seawater inundation is another threat seen by the Tuvaluans in events of natural disasters, such as cyclones, storm surges and King Tides (Liz, 2007; Talia, 2009). For instance in 1997 waves from cyclone Keli washed over the island of Tepuka Savilivili after which much of the island vegetation was devastated (Liz, 2007; Talia, 2009).

Similarly the Northern Part of Nanumaga has a history of incursions that have damaged Pulaka pits in the area (Dunn, 1976; Webb, 2007).

Coastal erosion is another impact of climate change that poses a threat to ground water lens. The idea that increases in sea level will result in coastal erosion causing loss of land and decrease in the size of groundwater lens is now in question, as a recent report by Webb and Kench (2010) argue otherwise. The study was carried out on the aerial photography of the 27 islands in the three atoll groups of Federated States of Micronesia, Kiribati and Tuvalu. The study shows that with the total rise in sea level of 120mm over the last 20 - 60 years at a rate of 2 mm per year a total of 86% of the islands remained stable. A portion of this 86% actually increased in size, thus no land was lost as where shoreline erosion took place it was equalized by sediment deposition on the opposite side. Only 14% of the islands showed scenarios of erosion where land was lost. White (2011) in agreement with this states that shoreline erosion is more likely to occur due to extreme events such as king tides and cyclones, than just the gradual rise in sea level.

Figure 3.2. Pulaka pit on Nui Island, Tuvalu.

By Shiwangni Rao

However, a limitation with the study is that it looked at the pattern of shoreline movement only within the past 2mm/yr rate of increase. With the projected increase in rate of sea level rise and increased acidity of the sea the question raised is will these islands still be stable in the future (Schaeffer, 2010). Also with this argument of Webb and Kench (2010) in place, the assumptions of the fresh ground water transition boundary being pushed inland and decreasing in size with relative land loss is in question and needs much detailed investigation.

Apart from the possible threats imposed by climate change and sea level rise, there are other anthropogenic factors that may be affecting the decline in Pulaka production, which have not yet been investigated. Some of these factors include disturbance caused by construction, population pressure, and migration of pit owners within the atolls and abroad. Also land disputes and the fast shifting preference towards an easier lifestyle, imported food and white collar jobs. While climate change is a prime suspect behind the decline in crop production by the farmers, these anthropogenic factors also contribute.

This groundwater case study builds on the study carried out by Webb (2007), where by the salinity of ground water in Pulaka pits was measured. These pits form a window into the ground water lens of the island. Comparison with past data will show if any change in salinity has occurred and if it has, to what degree. These are only snapshot values that would aid in giving a brief idea of saltwater intrusion in Tuvalu.

3.2 PRE SURVEY

Salinity is normally measured in electrical conductivity (EC) values µS/cm, as salt water conducts electricity. Giant swamp taro has been considered a salt tolerant plant by some (Vickers; Lambert, 1982; Brandburry, 1988; Wagih, 1997; Onwueme, 1999; Kazutaka, 2003; Covich, 2006; Deenik, 2006) and not so by others (Nyman, 1983; Wagih, 1997; Webb, 2007). In his report Webb (2007) found that giant swamp taro grew well in electrical conductivity values of 1000µS/cm and less, which is relatively fresh water. It tolerates EC of around 2000 - 3000µS/cm that is slightly saline while EC values of more than 3000µS/cm may be lethal to the plant (Dunn, 1976; Webb, 2007). Webb (2007) states that the response of giant swamp taro to high salinity is likely complex and the duration and intensity of exposure play a significant role in this. Along with the duration, factors such as shade, soil composition, weather and planting may also contribute significantly. To fully analyse the situation these factors need to be taken into consideration.

Webb (2007) suggested that only the islands of Funafuti, Nukulaelae and Niutao had salinity concentrations too high for swamp taro cultivation. Funafuti holds the highest value at 5000µS/cm, Niutao at 4000µS/cm and Nukulaelae at 3000µS/cm. Of these three atoll islands only Funafuti showed little variation between the salinity concentrations of the different pits which suggested that high ground water salinity was generally consistent throughout Funafuti. For the other two islands, apart from a few pits, the majority had low salinity concentrations suggesting that high salinity was not constant throughout the island. The high salinity found in the Tepela area (Central region) on Niutao may have been due to causeway construction in the area. The rest of the islands had salinity values average ranging from 1321+/- 363 µS/cm to 161 +/- 90 µS/cm (Webb, 2007). Also Dunn (1976) in his report noted that a pit on Motutala Islet in Nukulaelae had weak points in the bedrock which allowed seawater to seep into the Pulaka pits.

Figure 3.3. Causeway constructed in the Tepela Pit area, thought to have changed hydrology and caused greater salinity in some taro pits.

By Shiwangni Rao

Before the salinity survey was conducted, discussions were held with the Agriculture Minister Mr. Lausaveve and farmers on the respective islands. These discussions shed light on some of the causes and issues facing Pulaka cultivation in Tuvalu.

Niutao is an island that appears to be at high risk of increasing ground water salinity (Webb, 2007). Niutao is renowned for its natural swamp called the Tepela, a large pool of organic soils which produce Pulaka of the highest quality. However, after the construction of a dyke/ causeway (Figure 3.3), farmers noticed a rapid decline in Pulaka yield and many later abandoned their pits. There has also been a bridge constructed across the large pool on the island, for ease of access and this may have disturbed salinity in the Tepela. Some farmers today are trying to revive the pulaka production of Niutao and have started to plant in the Tepela despite the risks of low yield. Development is slow though there seems to be progress (Nuitao Kaupule members 2010, pers. comm.).

On the islet of Motutala in Nukulaelae the ‘Kaupule’ island council have built a small structure to stop the seawater from affecting the Pulaka pits. While most of the planning information associated with this construction has been lost, the acting secretary of Kaupule, Mr. Leki described it as a burrow that reaches deep into the

59 ground water lens, similar to a well, essentially a square with concrete opening measuring 30cm x 30cm. The council had hoped that this would gather all the sea water entering the Pulaka pit during high tide, unfortunately this construction has been of no benefit but may have aggravated the situation (discussed further in the discussion section).

On the islands of Nanumaga and Nanumea (Lakena islet) the situation is worse with many of the Pulaka pits abandoned. On Nanumaga a large northern pit has been completely abandoned with only isolated Pulaka sprouting randomly. According to the farmers the pit soil was too saline to grow Pulaka due to waves during King Tides that washed over and inundated the pit with sea water. Nanumea has the same situation, but other reasons such as landowners migrating to overseas or families having no sons to cultivate pits were also identified as reasons for abandoning giant swamp taro cropping.

In many of these islands there is a preference for imported food such as rice, noodles, bread, biscuits as Pulaka requires more work to cultivate and to cook. The younger generation is shifting towards the white collar jobs and has no interest in farming. The preference of an easier and more relaxed lifestyle may be a threat to giant swamp taro cultivation as great as the rise in ground water salinity in Tuvalu.

During discussions with farmers it was noted that there were two groups of farmers. There were men who have been farming all their lives and men who were retired government workers and seaman who have returned to work on the fields. Apart from the issues of increase in soil salinity in regards to the giant swamp taro, the loss of traditional knowledge is also a threat to the giant swamp taro. Tuvalu has around 18 cultivars of giant swamp taro (Iese, 2005) and there are only a handful of farmers who are aware of all these cultivars and cultivate them. For new farmers such as the retired workers, this knowledge is limited. The knowledge to differentiate/classify between the cultivars is diminishing and only the common ones such as the

‘Paipialaliga’, ‘Ikalaoi’ and the ‘Ikaulalua’ are well known and grown. The others are greatly endangered and at risk of local extinction.

3.3 SURVEY

Of the nine islands of Tuvalu, six were studied, Nanumea, Nanumaga, Niutao, Nui, Funafuti, and Nukulaelae. Upon arrival on the island contact was established with the island agriculture officer and a guide was provided. Pulaka Pits were located with the help of Webb’s 2007 study, maps from Google Earth and the guide.

The following procedures were carried out to obtain data for GPS location, ground water salinity and general observations: Once the pits were located, the pit location was noted on a Garmin GPS 72H Handheld unit. General observations were made on the pit such as: - Plants present - Time of sampling - Pit health status - If it was still in use, or left fallow - Vegetation apart from Giant swamp taro and soil status

At a single GPS location (refer to Google maps in the result section) six salinity measures were taken around the area randomly using a DiST 4 EC and TDS tester (salinity meter) with an accuracy of +/-1µS/cm. The 15 cm long salinity meter was dipped in the swamp water. Special attention was paid to keep the swamp water as undisturbed as much possible, as this would cause mixing of the swamp water and the data would be compromised. Once the reading stabilized on the salinity meter the conductivity value was noted.

During these survey sessions on the various islands discussions were carried out with the farmers regarding the pulaka cultivation and so on. Also samples for the Centre

61 for Pacific Crops and Trees (CePacT) were collected for plant genetic resource conservation and duplication.

Data analysis was done using Microsoft Excel, Garmin device software combined with Google earth for maps. For data analysis surface salinity values collected in this study were compared to surface salinity values in Webb (2007) using paired T-test in Genstat software and an α =0.05.

3.4 RESULTS & DISCUSSION

Ground water salinity measurements were taken from a number of pulaka pits present on the six atoll islands in late July and early August, 2010. The atoll islands have been name tagged with a yellow square on the Google earth maps. These results were then compared and analysed against the results from the study conducted by Webb in January to April, 2006. The same pits were sampled as in Webb (2007) study. Since these measurements were taken only four years after Webb (2007) study, they do not reveal a long-term trend but they are useful to indicate how constant overtime the ground water salinity is.

3.4.1 Nanumea Nanumea lies north of Funafuti, the main island of Tuvalu. It is a small elongated northwest-southeast running closed lagoon atoll. 12.07km in length and 2.41km in width with a total land area of 3.24km2 (Dunn, 1976). The low lying coral atoll accommodates a small population of 918 (Resture, 2008). The atoll group also consists of a smaller islet Lakena that was used as shelter for the people of Nanumea during the World War II. At present locals have chosen Nanumea as the island of residency and Lakena as the agricultural land where they plant their pulaka and other vegetables. Like all the other low lying atolls of Tuvalu the small islet of Lakena which rises barely 14.02m (Resture, 2008) above sea level is also facing the threat of

62 salt water intrusion and an increase in ground water salinity. Looking at the 2006 data the highest pulaka pit/ground water salinity was found to be 761.54 µS/cm +/- 186.79 and the lowest 386.67 µS/cm +/- 366.79. While for 2010 the highest salinity was found to be 864.33 µS/cm +/- 763.40 and the lowest was 511.67 +/-44.91 µS/cm. The average ground water salinity of the island in 2006 was found to be 607.64 µS/cm +/-136.69 and in 2010, 597.37 µS/cm +/- 165.32 (Table 3.1). Statistical analysis of the results showed that there is no significant difference in between the 2006 and 2010 salinity levels in Nanumea.

Nanumea was found to have high variability in the pulaka pits, at a standard deviation of 165.32, indicating that the pits experienced a range of salinity levels. Plants generally had diminished growth and showed heavy chlorosis on leaves compared to healthy plants on other islands. Many of the pits have been abandoned (Figure 3.4) and others were covered in weed. Only a handful of the pits showed cultivation of other crops such as common taro, banana, sugarcane but these were not at their optimum health. The soil on the island is light and poor. The pits are very exposed to sunlight; pulaka in pits that were shaded appeared healthier than the others. In addition, the highest salinity values lay in the region close to the small lagoon on the island, which had a salinity value of 8330 µS/cm+/-165.32. A similar peak in salinity is also seen in Webb’s 2006 study.

Figure 3.4.(Left) Abandoned Giant swamp taro pit on Nanumea.

By Shiwangni Rao

Figure 3.5. (Right) A fully productive pit on Nanumea, depicting the Pulaka productivity level that can be attained on the island

Hence for Nanumea it can be said that the ground water salinity is well below 1000 µS/cm which is more or less fresh water. If more effort is put in cultivating giant swamp taro the plant could likely grow quite well on the island (Figure 3.5).

Location: Nanumea - Lakena Date: 04/08/2010 Weather: Fine

Vaipulaka ate Faifeau

Vaipulaka a Ranford

Vaipulaka a Haumaafe & Lolua

Map 3.1 (Top) Nanumea. Map 3.2 GPS-located Pulaka pits on Nanumea

Table 3.1 Nanumea ground water salinity

Pit Name GPS Conductivity uS/cm Pit Area Notes/Observations point point poin point point point Mean SDE 1 2 t 3 4 5 6 V Vaipulak 05̊ 39’06.5” S 340 310 290 330 310 290 429 95 healthy pit with young a 176̊ 04’39.8”E plants , but diminished a growth Haumaef a and Lolua II 05̊ 39’04.8” S 380 420 400 330 570 450 average health of pit, well 176̊ 04’33.3”E cultivated

III 05̊ 39’03.8” S 560 500 450 580 450 530 176̊ 04’33.2”E

IV 05̊ 39’02.7” S 440 530 560 420 370 490 176̊ 04’32.1”E

Vaipulak 05̊ 39’02.3” S 350 410 460 300 520 560 864 763 pit is regularly visited but a a 176̊ 04’30.8”E plants have poor health Ranford II 05̊ 39’01.5”S 630 620 570 510 520 420 well cultivated plants in 176̊ 04’29.1”E good health, diminished growth III 05̊ 39’01.7” S 440 360 480 610 510 520 pit partially cultivated with dalo on raised islands

176̊ 04’28.3”E healthy pit

IV 05̊ 39’01.7” S 810 770 930 540 760 1200 average health of pit 176̊ 04’28.3”E

V 05̊ 39’01.4” S 3090 3330 2540 1120 1020 1030 poor pit has lots of weed 176̊ 04’25.8”E and some sugarcane

Vaipulak 05̊ 39’01.2”S 610 590 650 660 620 710 average health of pit, some a ate 176̊ 04’25.2”E banana present Faifeau II 05̊ 39’00.7”S 600 660 1250 480 150 630 625 217 poor pit with weak plants 176̊ 04’24.9”E and weeds

III 05̊ 38’58.7”S 620 620 540 450 930 480 half the pit is left fallow 176̊ 04’25.6”E

IV 05̊ 39’59.9”S 480 610 530 560 640 520 556 59 healthy pit with some taro 176̊ 04’27.5”E also

V 05̊ 39’00.6”S 540 470 540 440 540 540 511 44 healthy pit 176̊ 04’28.8”E

Total 597 165 The above table shows ground water salinity level data recorded in Nanumea and the observation made. Vaipulaka a Ranford had the highest salinity at 864.33 µS/cm and Vaipulaka ate Faifeau V had the least at 511.67 µS/cm. The mean salinity level on the island was 597.37 µS/cm with a standard deviation of 165.32, the Pulaka pits on the island were fairly poor with most left partially fallow

Pulaka pit salinity on Nanumea 2010 2000 1500 1000 500 0 Vaipulaka Vaipulaka a Vaipulaka ate Unknown 1 Unknown 2 a Haumaefa Ranford Faifeau & Lolua

Electrical Conductivity (μS/cm) Conductivity Electrical Pulaka Pits

Graph 3.1. Showing the mean ground water salinity levels and standard deviation bars in Nanumea Pulaka pits, with the Vaipulaka a Ranford with highest recording and Vaipulaka a Haumaefa and Lolua with the lowest.

Groundwater salinity on Nanumea 1600 1400 1200 1000 800 600 400 200 2010 0 2007 -200 Vaipulaka Vaipulaka a Vaipulaka Unknown 1 Unknown 2 -400 a Haumaefa Ranford ate Faifeau & Lolua

Electrical Conductivity (μS/cm) Conductivity Electrical Pulaka Pit

Graph 3.2. Comparisons of the mean ground water salinity levels and standard deviation bars of 2006 data (Webb, 2007) for each of the respective pulaka pits.

3.4.2 Nanumaga Running North-South, Namugama is an oval shaped reef island approximately 3.62 Km long and 1.61 Km wide with a land area of about 3.24Km2 (Dunn, 1976). Dunn (1976) in his report stated that the people of Nanumaga claimed the Northern pit on the island was completely washed by sea water, rendering it unproductive and people

68 stopped planting. The pit recorded salinity values of 47000 and 30000 µS/cm (Dunn, 1976). Dunn concluded that the salts in the pit were highly mobile moving in and out with the tide and as such he suggested that the pulaka plants were victims of ground water salinity and not soil salinity.

Nanumaga is a single island with a large lagoon in the middle; this particular island has been identified as one of the islands with serious salinity issues by Mr. Levusevu. The northern pit ‘Vaipulaka I Tokelau’ was found deserted. At the time of the survey the pit had a salinity measure of only 530 µS/cm which is quite fresh. The pit had a few remnant pulaka growing and appeared quite healthy. In the rest of the pits which are located at the southern end of the island the highest salinity recorded was 550 µS/cm+/- 44.2. In 2006 the highest pulaka pit salinity was found to be 1665µS/cm+/-304.06 with the mean ground water salinity of the island 744.28 µS/cm+/-5.48 and 473.33 µS/cm+/- 115.9 in 2010 (Table 3.2). Overall there was no statistical significant difference found between the 2006 and 2010 ground water salinity level.

The soil on the island is dark and relatively rich in organic matter (this could be a result of good mulching and composting in the past). There were lots of crop species apart from giant swamp taro growing in the pits such as banana and breadfruit trees which were all healthy and so were the pulaka plants in the pits. Vegetation such as breadfruit trees is generally intolerant to saline conditions. The presence of these in the pits is an indicator that the island has long profited from the good ground water. Hence overall the island ground water status has improved since the washing over by the king tides and the island at the time of this study can support a healthy flora. But as stated by Dunn (1976) salinity moves with the rise and fall of tides hence at high tides salinity may increase if high water ever occurs again. The southern pits which are not as vulnerable to the tides as the northern pits had less variability compared to 2006 suggesting the salinity level at the time of the study was stable.

Location: Nanumaga Date: 04/08/2010 Weather: Fine

Vaipulaka I Pit II Tokelau

Vaipulaka I Toga

Map 3.3 Nanumaga. Map 3.4 GPS located Pulaka pits on Nanumaga

Table 3.2 Ground water salinity on Nanumaga

Pit Name GPS Conductivity uS/cm Pit Area Notes/Observations

Point point point 3 point point point Mean SDE 1 2 4 5 6 V Vaipulak 06̊ 16’34.3”S 490 510 550 540 520 570 530 28 Pit has been abandoned a I 176̊ 19’21.7”E for quite some time, Tokelau covered with weeds and some pulaka growing wildly and are healthy Vaipulak 06̊ 18’00.5”S 300 370 350 310 350 360 340 28 healthy Pit with some a I Toga 176̊ 19’14.7”E Banana also present II 06̊ 17’59.8”S 590 500 560 510 530 610 550 44 healthy Pit 176̊ 19’11.1”E Total 473 115 The above table depicts the mean ground water salinity levels of Nanumaga and the other related observation. The highest ground water salinity was recorded at 550 μS/cm while the lowest was at 530 μS/cm. The Island’s mean salinity was found to be 473.33 μS/cm with a standard deviation 115.90. Except for the Vaipulak I Tokelau pit the other two pits were healthy with a number of other food crops apart from pulaka.

Pulaka Pit Salinity on Nanumaga

700 2010 600 500 400 300 200 100 0 Electrical conductivity μS/cm conductivity Electrical Vaipulak I Tokelau Vaipulak I Toga II Pulaka Pit

Graph 3.3. Showing the mean ground water salinity levels and standard deviation bars of Nanumaga Pulaka pits, with Pit 5 having the highest recording and Vaipulaka I Toga with the lowest.

Groundwater Salinity in Nanumaga 2500

2000

1500 2010 1000 2007

500 Electrical conductivity μS/cm μS/cm conductivity Electrical 0 Vaipulak I Tokelau Vaipulak I Toga II

Graph 3.4. A comparison of the 2006 mean ground water salinity levels (standard deviation bars) (Webb, 2007) and the 2010 salinity values of each pulaka pit on Nanumaga, pit II showing the greatest decrease in the salinity values

3.4.3 Niutao The atoll has a rectangular shape running in the east-west direction approximately 2.41 Km in length and 1.21 Km in width with a total land mass of 2.43 Km2 (Dunn, 1976). Niutao has a large natural swamp ‘Tepela’ that was a favourable spot for Pulaka farming in the past. However the dyke construction appears to have aggravated ground water salinity problems. Many of the farmers and officials infer this to be the result of the dyke construction that spiked increase in the pit salinity level some years ago (Dunn, 1976; Webb, 2007).

Today many of the ‘Kaupule’ members and the farmers have started to plant in the Tepela area again, hoping to revive the land and get it back to its old productivity potential (Figure 3.6). Looking at the data gathered in 2006 the ground water salinity in the Tepela area measured up to a 2450 µS/cm+/- 1098.12 and at present it sits at 1460 µS/cm+/- 860.98 (Table 3.3). Apart from the dyke/Causeway construction, a bridge has also been constructed to allow people to cross the large central lake without having to go around. Once again this large construction relative to the small island size may have contributed to changes in salinity levels of the island as explained above.

Figure 3.6. The Tepela area where pulaka is being once again cultivated in hope of reviving the plantation.

By Shiwangni Rao

As seen from graph 3.6, highest ground water salinity on the island is 1500 µS/cm +/- 860.98, which lies well in the fresh ground water zone of 1500-2500 µS/cm. Statistical analysis of the 2006 and 2010 ground water salinity levels showed there is no significant difference between the two readings.

Despite the relatively fresh ground water in the pulaka pits on Niutao, the pits were not being used up to its potential productivity, meaning the pits were poorly cultivated, but growing reasonably well. It was observed that the plants in the shady regions or on the shady outskirts of the pits were growing better than those plants exposed to full sunshine. Apart from pulaka vegetation other crops such as breadfruit, banana and sugarcane were also growing quite well in this area.

Location: Niutao Date: 05/08/2010 Weather: Fine

Matakakaka Talo Sualiki

Lololuli

Tepela

Vaipulaka Lasi Pua Te Talo

Map 3.5 (Top) Niutao. Map 3.6 (Bottom) GPS located Pulaka pits on Nanumag

Table 3.3 Pulaka pit salinity on Nuitao

Pit Name GPS Conductivity uS/cm Pit area Notes/Observations point point point point point point Mean SDEV 1 2 3 4 5 6 Vaipulak 06̊ 06’41.9”S 580 440 430 1170 630 450 616 283 poor pit plenty weeds a Lasi 177̊ 20’32.5”E and plants not in optimum health Tepela 06̊ 06’38.1”S 450 1190 2050 2060 2480 530 1460 860 average health of pit 177̊ 20’33.2”E

Pua te 06̊ 06’43.4”S 1930 350 460 330 410 410 648 629 plenty weeds talo 177̊ 20’49.6”E

Lolouli 06̊ 06’29.4”S 560 620 390 400 550 440 493 95 average pit, some sugar 177̊ 21’01.5”E cane also present on pit

Talo 06̊ 06’45.1”S 660 530 440 470 520 510 521 75 average pit Sualiki 177̊ 21’01.0”E

Matakak 06̊ 06’42.0”S 550 630 540 430 420 490 510 79 healthy pit a 177̊ 20’92.8”E

Total 708 373 The above table shows the recorded ground water salinity level of Niutao. The highest ground water salinity was found to be at the Tepela are at 1460 µS/cm while Matakaka had the lowest at 510 µS/cm. The mean island ground water salinity was found to be 708.33 µS/cm with a standard deviation of 373.47.

2500 Pulaka Pit Salinity on Niutao 2010

2000

1500

1000

500 Electrical conductivity μS/cm conductivity Electrical 0 Vaipulaka Tepela Pua te Lolouli Talo Matakaka Lasi talo Sualiki Pulaka Pit

Graph 3.5. The above graph shows mean ground water salinity levels (with standard deviation bars) of Nuitao in the various Pulaka pits. The Tepela has the highest salinity level while Lolouli and Matakakaka share the lower values

Ground water Salinity in Niutao 4000 3500 3000 2500 2000 1500 1000 2010 500 0 2007 Electrical conductivity μS/cm conductivity Electrical

Pulaka Pit

Graph 3.6. Graph showing the comparison of mean ground water salinity levels (with standard deviation bars) between 2006 (Webb, 2007) and 2010 data, with the Tepela pit showing a decrease in the salinity level while the rest show an overall increase.

3.4.4 Nui Nui is another island with healthy pits, and also a rich Pulaka diversity. Thirteen cultivars of pulaka can be found on this island, and the tradition of pulaka farming is very much alive. Running North-South the small island has an elongated oval shape, 8.05 km in length and 3.22 km width with a total landmass of 3.24km2 (Dunn, 1976). Ground water salinity was found to be 610.28 µS/cm+/- 250.55 in 2010 and 209.2 µS/cm+/-169.89 in 2006, (Table 3.4). Statistical analysis of the 2006 and 2010 ground water salinity showed that there was no significant difference between the two readings. Also ground water salinity level in Nui was found to be well within the fresh water zone.

Location: Nui Date: 23/07/2010 Weather: Fine

Tabontebike

Vaipulaka Lasi Vaipulaka Foliki Pit 5

Vaipulaka ate Faifeau

Map 3.7 (Top) Nui. Map 3.8 (Bottom) GPS located Pulaka pits on Nui.

Table 3.4 Pulaka pit salinity on Nui

Pit Name GPS Conductivity uS/cm Pit area Notes/Observations point point point point point point Mea SDE 1 2 3 4 5 6 n V Tabontep 07̊ 14’26.5”S 700 670 450 650 680 490 606 107 Dalo grown on raised islets. No ike 177̊ 08’53.6”E other crop, half fallow field. Soil dark and loamy in nature. Pit is extensively exposed to sunlight Vaipulak 07̊ 14’31.9”S 520 510 530 490 520 550 720 37 left fallow as no compost present a 177̊08’50.5”E and lots of weed. Has banana and Lasi I breadfruit trees which not at optimum health. Soil mainly calcareous and pit exposed II 07̊ 14’32.2”S 640 680 700 710 640 670 healthy pit 177̊08’51.6”E

III 07̊ 14’33.6”S 980 1190 1690 1300 1530 1030 left fallow, high chlorosis on 177̊08’57.7”E leaves of plants

IV 07̊ 14’34.1”S 400 420 390 430 400 380 healthy pit 177̊08’58.7”E

Vaipulak 07̊ 14’39.7”S 600 570 630 530 620 660 601 42 healthy pit but there was not a 177̊08’55.4”E much water Foliki present to do testing

Vaipulak 07̊ 14’39.0”S 780 750 720 690 760 800 636 56 healthy pit a 177̊08’55.1”E ate Faifeau II 07̊ 14’36.8”S 580 540 450 570 490 510 healthy pit 177̊08’54.1”E

Pit 5 07̊ 14’35.4”S 470 490 420 500 450 430 485 68 poor pit, lots of weed present and 177̊08’54.8”E no work seems to be done on it

II 07̊ 14’36.2”S 530 540 540 570 550 490 healthy pit 177̊08’54.6”E

III 07̊ 14’36.5”S 580 460 340 380 570 430 has been left fallow 177̊08’54.6”E

Total 610 250 The above table shows the ground water salinity levels on Nui. The highest ground water salinity was recorded at 720.83 µS/cm in Vaipulaka Lasi I while the lowest was recorded in Pit 5 at 485.56 µS/cm. The mean ground water salinity was found to be 610.28 µS/cm with a standard deviation of 250.55

Pulaka Pit Salinity on Nui 2010 1200 1000 800 600 400 200 0 Tabontepike Vaipulaka Lasi Vaipulaka Vaipulaka ate Pit 5

Electrical conductivity μS/cm conductivity Electrical I Foliki Faifeau Pulaka Pit

Graph 3.7. Mean ground water salinity levels (with standard deviation bars) of Pulaka pit on Nui showing low variability with Vaipulaka Lasi having the highest value and Pit 5 with the lowest.

Ground water Salinity in Nui 1200

1000

800

600

400 2010 2007 200

Electrical conductivity μS/cm conductivity Electrical 0 Tabontepike Vaipulaka Vaipulaka Vaipulaka Pit 5 Lasi I Foliki ate Faifeau Pulaka Pit

Graph 3.8. The above graph depicts a comparison of the mean ground water salinity levels (with standard deviation bars) of 2006 (Webb, 2007) and 2010 salinity values on Nui, the values show and overall increase in salinity except for Pit 5.

3.4.5 Funafuti The capital island of Tuvalu, Fongafale is the longest and the narrowest of all the nine atoll islands in the archipelago. It is a North-south running oval shaped atoll with a total landmass of 2.43 Km2 (Dunn, 1976). The high ground water lens is the product of the limited landmass of the islands. That makes this ground water lens highly vulnerable to any form of pollution, construction and earthworks and environmental change such as changes in rainfall and temperature. Fongafale having Funafuti as the capital has over the years experienced much construction of buildings and houses (Dunn, 1976; Webb, 2007).

Funafuti has survived bombing and military upheaval of World War II and the associated earth works and construction of the island airstrip. Funafuti is struggling to keep its pulaka cultivation culture alive with limited resources.

Figure 3.7. Very healthy pulaka plants that grow in Funafuti.

By Shiwangni Rao

The pulaka plantations area is clustered in the centre of the island where according to Ghyben-Herzberg principle (Rozell, 2007) (Woodraffe, 1989) the ground water is the deepest. However farmers claim that the daily tides do have a noticeable impact on the plants suggesting that the islands ground water lens may be shallow.

Fongafale happens to have the highest recorded ground water salinity levels (Table 3.5, Graph 3.9) compared to the other islands of Tuvalu. The highest salinity recorded in 2010 was 14790 µS/cm+/-1020.27 compared to 6570 µS/cm+/- 678 in

2006 giving a difference of 162 % in the ground water salinity, for the highest recorded salinity. The island has experienced a mean 79 % increase in the ground water salinity having an overall average of 6749 µS/cm +/- 3703.14 in 2010 compared to the 3774 µS/cm+/- 1749.94 in 2006 (Table 3.7). There is a significant difference in the mean ground water salinity levels between 2006 and 2010 (p=0.04).

This high salinity in Fongafale is in agreement with Webb (2007) in citing Falkland (1999) where it has been stated that ground water salinity levels are too high for human consumption and use. Also that Funafuti virtually has no fresh ground water lens present due to the course material constituents of the island.

Webb (2007) states that pulaka has been seen to tolerate up to 2000 µS/cm but dies out at 3000 µS/cm which in comparison to the recorded 14790 µS/cm is quite low. Looking at the high salinity values for Funafuti in both the 2006 and 2010 survey it is a possibility that pulaka in fact does have high salt tolerance or salinity tolerance has been induced in the plants over the number of years of high ground water salinity. Furthermore, table 3.5 shows that Pulaka is quite healthy in the pit of highest salinity level 14790µ S/cm+/-1020.27 (Figure 3.7) and according to the farmers on Funafuti a particular cultivar the ‘Pulaka Kula” seems to show higher tolerance in comparison to other cultivars (Figure 3.8).

Apart from this, some other factors that may be affecting the survival of these plants could be the amount of rainfall and evapotranspiration rate. However the above stated factors biological and climatic are only assumptions of the possibilities and needs a detailed investigation.

Figure 3.8 Pulaka Kula, one of the cultivars of Giant swamp Taro found on Funafuti claimed to be salt tolerant by farmers.

Location: Funafuti- Fongafale Date:19/07/2010 and 27/07/2010 Weather: Fine

Northern Pit

Central Pitt

Southern Pit

Map 3.9 (Top) Fongafale atoll. Map 3.10 (Bottom) GPS located Pulaka pits on Funafuti, Fongafale.

Table 3.5 Pulaka pit salinity on Funafuti Conductivity µS/cm Pit Area Pit GPS point point point point point point SDVE Notes/Observations Name Mean 1 2 3 4 5 6 08̊ 31’26.8’S average health of pit, highly Southern 7990. 7880 8290 7800 8130 7950 7890 184.07 exposed to sunlight, not much pit 179̊ 11’45.1”E 00 water in pit 08̊ 31’27.5’S 2040. 2240 2340 1540 1970 2370 1780 335.08 poor pit, lots of weed 179̊ 11’45.5”E 00

08̊ 31’27.8’S poor pits with half left fallow 4766. 4640 4780 4880 4750 4830 4720 84.30 and some other vegetation like 179̊ 11’45.1”E 67 banana and breadfruit.

08̊ 31’28.0’S poor pits with half left fallow 2280. 1499.7 1250 4240 1350 4190 1380 1270 and some other vegetation like 179̊ 11’46.1”E 00 1 banana and breadfruit. Very healthy pit, also has a 08̊ 31’17.5’S Central 1613 1390 14790 1020.2 large number of banana plants. 14340 14950 15800 13620 pit 179̊ 11’53.0”E 0 0 .00 7 located close to housing well attended 08̊ 31’17.8’S healthy pit, soil has a 1066 1043 8321. 2309.7 9810 5220 6230 7580 noticeable amount of white 179̊ 11’53.0”E 0 0 67 0 calcareous soil 08̊ 31’20.1’S 6555. 2608.7 small healthy pit, with a lot of 8340 8170 8500 7800 4030 2490 179̊ 11’52.4”E 00 3 banana trees

08̊ 31’24.8’S 7846. 2014.4 Very healthy pit, located in a 10800 9120 6130 6720 5610 8700 179̊ 11’47.9”E 67 8 very shady area 08̊ 31’23.7’S 1130 1003 10968 poor pit, exposed to sunlight 11050 11130 11330 537.14 179̊ 11’45.0”E 0 0 .00 constantly 08̊ 31’26.0’S 9557. 2659.5 healthy pit located too close to 8730 9600 6770 13130 179̊ 11’46.6”E 50 4 piggery

Northern 08̊ 31’16.4’S very healthy pit, located close 4600 3810 6320 5000 4100 5630 4910 946 pit 179̊ 11’54.3”E to housing well attended Half of the pit is thoroughly cultivated while the other is left 08̊ 31’15.9’S fallow. Well exposed to 3190 2490 2740 2820 2310 2560 2685 306 179̊ 11’53.6”E sunlight, no composting present at the time of visit. Loamy soil. 08̊ 31’17.0’S Average health of pit well 5090 5420 5270 5600 3900 4850 5021 607 179̊ 11’52.6”E composted. Total 6748 3703 The above table shows the recorded data of the ground water salinity and the observations made on Funafuti. The highest salinity was found in the Southern pit at 14790µS/cm while the lowest was found in the central pit at 2280 µS/cm. The mean ground water salinity was found to be at 6748.63 µS/cm with a standard deviation of 3703.14.

Pulaka Pit Salinity on Funafuti 2010 16000 14000 12000 10000 8000 6000 4000 2000 0 Electrical Conductivity μS/cm Conductivity Electrical Southern pit Central pit Northen pit Pulaka Pit

Graph 3.9. Mean ground water salinity levels (with standard deviation bars) on Funafuti, with one of the central pit s with highest value and southern pit with the lowest. Funafuti has the highest ground water salinity values compared to the rest of the islands.

Ground water Salinity in Funafuti 16000 14000 12000 10000 8000

6000 2010 4000 2007

Electrical conductivity μS/cm μS/cm conductivity Electrical 2000 0 Southern pit Central pit Northen pit Pulaka Pit

Graph 3.10. The above graph is a comparison of the mean ground water salinity levels (with standard deviation bars) of 2006 (Webb, 2007) and 2010 data of ground water salinity values on Funafuti showing an overall increase in the ground water salinity levels on the island.

3.4.6 Nukulaelae Nukulaelae consists of 14 islets of which Fangaua is the chosen residential island while another significant island Motutala is used for pulaka cultivation. This atoll is 11.26 Km long and 4.02 Km wide with a total land area of approximately 1.6 Km2 (Dunn, 1976). Pulaka cultivation takes place on the southern end of the island and on another smaller islet. Motutala has natural swamps present such as those found in the Tepela area on Niutao. The people of Nukulaelae have put in place a small structure to prevent salt water from entering the pulaka pits. Unfortunately this construction does not address the issues of ground water lens in fact it might have disturbed the transition zone of the fresh ground water lens and the sea water.

From the data gathered it was found that there was no significant difference in the ground water salinity levels measured in 2006 and 2010. 2010 ground water salinity levels ranged from 8530 µS/cm+/- 1681.62 to 725 µS/cm+/- 114.85, with a mean ground water salinity of 2823.39 µS/cm+/- 2606.54. While in 2006 survey recorded the highest value as 2446 µS/cm+/- 1072.51 and the lowest at 398.57 µS/cm+/- 369.11 with mean ground water salinity at 1235.58 µS/cm+/-783.09. Webb (2007) in his survey found that the pit, Viapulaka mataafale on Nukulaelae had extremely high salinity unlike the rest of the salinity values which rested well between 444+/-378 to 908+/- 384 µS/cm. Looking at the 2010 data the Vaipulaka a Toe pit on Nukulaelae has elevated salinity followed by Vaipulaka Mataafale. Except for the two pits Vaipulaka a toe and Vaipulaka ite Fakai the rest of the values fall below the 2006 salinity levels or just slightly above.

Location: Nukulaelae Date: 11/08/2010 Weather: Cloudy

Map 3.11 (Top) Nukulaelae atoll. Map 3.12 (Bottom) GPS located Pulaka pits on Motutala Islet, Nukulaelae. Map 3.13 (next page) GPS located Pulaka pits on Nukulaelae main islet.

Vaipulaka ite Fakai

Vaipulaka I mataafale

Vaipulaka a Toe

Vaipulaka a Uputaua

Table 3.6 Pulaka pita salinity on Nukuklaelae Notes/ Conductivity uS/cm Pit Name GPS Observations point point1 point 2 point 3 point 5 point 6 Mean SDEV 4 Nukulaelae Vaipulaka a 9̊ 22”36.31”S 10410 9550 9060 8060 5520 8580 8530 1681 average pit toe 179̊ 48”38.74”E Vaipulaka a 9̊ 22”37.33”S 1110 1480 1370 1790 1150 1440 990 451 Uputaua 179̊ 48”38.12”E 9̊ 22”39.26”S II 620 660 690 530 520 530 179̊ 48”37.69”E Vaipulaka I 9̊ 22”27.52”S 740 880 560 640 810 720 725 114 mataafale 179̊ 48”31.44”E Vaipulaka 9̊ 22”18.65”S 2620 2390 3160 3200 2970 3170 2918 338 ite Fakai 179̊ 48”29.46”E Motutala Islet Vaipulaka I 9̊ 21”02.81”S 1110 780 720 740 790 1160 952 161 Healthy pit Motulata 179̊ 48”59.59”E 9̊ 22”05.03”S II 870 860 880 840 860 880 179̊ 48”59.91”E 9̊ 22”06.49”S III 1190 1160 1170 1040 1020 1080 average pit 179̊ 48”38.74”E Total 2823 2606 The table above is the collected data on pulaka pit salinity on Nukulaelae and other observations. The highest salinity was found in the Vaipulaka a toe at 8530 µS/cm and the lowest in Vaipulaka I mataafale at 725 µS/cm. The mean ground water salinity of the island was found to be 2823.39 µS/cm with a standar deviation of 2823.39.

Pulaka pit salinity on Nukulaelae 2010 12000.00 10000.00 8000.00 6000.00 4000.00 2000.00 0.00 Vaipulaka a Vaipulaka a Vaipulaka I Vaipulaka Vaipulaka I

Electrical conductivity μS/cm conductivity Electrical toe Uputaua mataafale ite Fakai Motutala Pulaka Pit

Graph 3.11. The above graph depicts the mean ground water salinity levels (with standard deviation bars) in the Pulaka pit on Nukulaelae, Vaipulaka a toe with the highest salinity values and Vaipulaka Mataafale with lowest.

Ground water Salinity In Nukulaelae 12000.00 10000.00 8000.00 2010 2007 6000.00 4000.00 2000.00 0.00 Electrical conductivity μS/cm conductivity Electrical Vaipulaka Vaipulaka Vaipulaka I Vaipulaka Vaipulaka I a toe a Uputaua mataafale ite Fakai Motutala Pulaka Pit

Graph 3.12. Comparison of the mean ground water salinity levels (with standard deviation bars) of 2006 (Webb, 2007) and 2010 data of the ground water salinity levels on Nukulaelae showing an overall increase in the salinity levels with Vaipulaka a Toe with the highest recorded increase.

3.4.7 Rainfall- Ground Water Recharge Rainfall plays a significant role in the recharge of the ground water lens. The Tuvalu Metrological Centre based in Funafuti, records weather data for Funafuti only. Data obtained from the Tuvalu Metrological Centre shows that the mean rainfall for the duration of the study was 118+/- 76mm in Funafuti. While average rainfall for the 2006 study was 294+/- 150mm, giving an overall 59.94 % difference (Table 3.7). Meaning that at the time of survey for the 2006 study there was more rainfall compared to the present 2010 survey in Funafuti, this is one factor that can result in lower salinity values for 2006 compared to 2010.

Table 3.7 Comparison of average rainfall for 2006 and 2010

Month Mean Rainfall SDEV Monthly Rainfall Average - 2006 January 383 170 February 393 189 March 407 196 April 59 133 May 231 124 Total 294 150 Monthly Rainfall Average - 2010 July 172 86 August 64 48 76 Total 118 Average monthly rainfall record for the duration of study in 2006 (Webb, 2007) and 2010 with their calculated standard deviation.

3.4.8 Tuvalu Ground Water Salinity There was no significant difference in ground water salinity levels in Tuvalu between 2006 and 2010 (p=0.18). However, looking at the six islands separately Funafuti had significant difference at α=0.05 in the ground water salinity levels between the 2006 and 2010 readings. Nanumea and Nanumaga experienced seawater inundation according to the locals caused by cyclones and king tides over the years. However, rainfall since these events has resulted in recovery of the fresh ground water lens. However, it is likely that if another wave event inundates these pits they will again become saline.

Nui and Nukulaelae also have ground water salinity values that are within the boundaries of fresh water to mildly brackish. For Nukulaelae having its ground water salinity values falling within the mildly brackish zone is largely due to the effects of two pits, the rest of the pits have salinity values in the fresh water range. Funafuti which has the highest ground water salinity places it in the brackish water range. It also has the only increase in ground water salinity between 2006 and 2010 with the mean ground water salinity of 6749 µS/cm+/-3703.14 and an overall increase of 78.83 %.

Although Funafuti has brackish ground water giant swamp taro is still persisting on Funafuti, ranging from very healthy to very poor pits. Funafuti gives an interesting opportunity to investigate the apparently high salinity tolerance level of Pulaka here, as the island not only has brackish ground water but the soil present on the island is fairly poor as well.

In Niutao, the Tepela area was subjected to heavy construction and earthworks following which the island had an elevated level in the Tepela ground water salinity and the pits in the area became unproductive. Tepela area had the highest salinity values above the standard fresh water limit of 1500-2500 µS/cm (Webb, 2007). The rest of the recorded values fell in the fresh water zone.

While this survey has given a glimpse of the status of ground water salinity in Tuvalu, showing that there was no change in ground water salinity except for one island between 2006 and 2010. Further monitoring and research is needed, as this change accounts only for ‘snap-shot’ measurements made at two points in time 2006 and 2010. Continuous monitoring of ground water salinity levels would enable a far better understanding of the relationship between salinity and other important variables such as tide, weather, storms and rainfall to be developed.

Table 3.8 Comparison of 2006 and 2010 ground water salinity levels Island 2006 2010 F probability

Nanumea 608+/- 137 597+/-165 1.00

Nanumaga 744+/-634 473+/-116 0.25

Niutao 471+/-947 708+/-373 0.44

Nui 209+/-170 610+/-251 0.13

Funafuti 3774+/-1750 6749+/-3820 0.02

Nukulaelae 1236+/-783 2823+/-2606 0.63

The above table contains the islands average (total all pit means divided by total number of pits) salinity and the f probability of statistical difference on the islands surveyed

2010 Groundwater Salinity level in Tuvalu 16000

14000

12000

10000

8000

6000

4000 Electrical Conductivity μS/cm μS/cm Conductivity Electrical

2000

0 Nanume Nanumaga Niutao Nui Funafuti Nukulaelae

-2000

Graph 3.13. Ground water salinity levels in the six islands of Tuvalu, Funafuti having the highest salinity levels followed by Nukulaelae Niutao and so on.

Comparision of 2006 & 2010 Ground water salinity levels

16000 2010 14000 2007 12000

10000

8000

6000

4000

Electrical conductivity (μS/cm) conductivity Electrical 2000

0 II Pit 5 Lolouli Tepela Vaipulaka … Matakaka Central pit Central Pua te talo Talo Sualiki Talo Unknown 1 Unknown 2 Northen pit Northen Southern pit Tabontepike Tabontepike Vaipulaka Lasi Vaipulak I Toga Vaipulaka Lasi I Vaipulaka a toe Vaipulaka Foliki Vaipulak I Tokelau Vaipulaka ite Fakai Vaipulaka a Ranford Vaipulaka a Uputaua Vaipulaka I Motutala Vaipulaka ate Faifeau Vaipulaka ate Faifeau Vaipulaka I mataafale Nanumea Nanumaga Niutao Nui Funafuti Nukulaelae Location

Graph 3.14. Comparison of the 2006(Webb, 2007) and 2010 data for all the Pulaka pits on all the islands, with Funafuti having the highest salinity measures followed by Nukulaelae

3.5 CONCLUSION

Giant swamp taro (Cyrtosperma merkusii) or Pulaka cultivation is a very strenuous task, it requires much care and attention and takes time to grow, but if well attended the plant can still be grown in many areas in Tuvalu. It is interesting to note the apparent high salinity tolerance of cultivars grown in Funafuti.

Residents fear the rise in sea level will lead to a rise in ground water salinity. However, there are other factors that contribute to increase in salinity. These are human induced factors such as construction, earthworks, population pressure, and development pressure (engineering, ground water pumping, etc.). Factors which may exacerbate saline intrusion include natural disasters and rainfall variability.

Furthermore, as seen from the discussions with the local farmers, high ground water salinity is not the sole reason for the decreased pulaka production in Tuvalu. Other contributing factors exist such as land issues, presence of a male farmer in the family, migration, the preference of modern food due to its accessibility and convenience, or the general desire for a modern lifestyle and social status. These factors have a significant impact on the traditional framing throughout the Pacific and Tuvalu is no exception.

As seen from the results ground water salinity levels have shown no significant difference in Nanumea, Nanumaga, Nui, Niutao and Nukulaelae. Funafuti recorded a significant increase in ground water salinity between 2006 to 2012. Except for Funafuti and Nukulaelae the rest of the islands have ground water salinity level fall below the limit of fresh water zone of 1500- 2500 µS/cm (Webb, 2007). On the other hand, with the factors affecting this salinity it is hard to say that if this difference in ground water salinity would persist, decrease or increase even further.

Solving the issue of increasing ground water salinity presents major obstacles; Tuvalu has small islands with small land masses which results in small and fragile ground water lens that can be easily disturbed. Engineering methods to protect the fresh ground water lens would be very costly and not likely to be successful.

While at the moment there is no perfect solution to the issue of ground water salinity, food security on the island can be supported by sustainable use of agro-biodiversity. Evaluating the different cultivars of pulaka will provide farmers with the information as to variation in salinity tolerance. However before they can evaluated, they must be conserved.

A holistic approach to agro biodiversity includes:

- Promoting cultivation of all the cultivars and encouraging farmers to plant as many cultivars of pulaka as possible Competitions that promote the diversity of local food crop. Agricultural programs that aid farmers by promoting local food crops Solving land issues and allocating land to farmers Exchange of cultivars with other countries in the Pacific

- Traditional knowledge conservation involves conserving all the knowledge in relation to giant swamp taro such as cultivation, preparation, recipes, traditional usage and values. Also includes myths legends, stories, songs, poems, cultivars of giant swamp taro, how to differentiate between cultivars, its harvest time and its maturity age.

Conservation using traditional knowledge entails: Documentation Exchanging traditional knowledge possibly through workshop programs Passing the knowledge to the younger generations by the elders

Incorporating the knowledge in the school curriculums Awareness through competitions, posters, broadcasting, workshops and speeches.

Furthermore, one of the scientific approaches to addressing the threat to food security in Tuvalu and the other Pacific atoll islands in the region is development of a salt tolerant cultivar of giant swamp taro. There is an urgent need to develop this particular crop as salt tolerant, not only to ensure food security but also the preservation and conservation of the atoll communities’ culture, tradition and identity.

As a final point, to find the perfect solution the problem has to be perfectly understood. Ground water salinity and its resulting effect on pulaka production, needs more exhaustive investigation to be fully understood. As it’s a complex phenomenon and research into the plant physiology and plant-soil relation is required to determine the physical and biological aspects involved in pulaka growth.

100

4.0 DEVELOPMENT OF A RAPID IN VITRO SCREENING METHOD VERSUS IN VIVO SCREENING

4.1 INTRODUCTION

Screening for potential salt tolerance traits in giant swamp taro cultivars presents a door of opportunity for developing a buffer to the effects of climate change and sea level rise. In vitro testing allows giant swamp taro cultivars that show salt tolerance in situ such as the Pulaka Kula to be screened and fully utilized. It is also faster than conventional methods and is not as expensive or complex as the DNA marker assisted screening methods. Investigating screening methodologies such as these, present the best way forward into understanding the dynamics of island ecology and solving the big problems, small islands and countries such as Tuvalu face.

The development of a screening methodology for salinity tolerance included the evaluation of different salinity levels, the nature of the salt solution applied and the method of application. Two types of salt solutions were used, the standard NaCl that is used in salinity screening experiments and the Artificial Sea Water (ASW) to mimic the effects of sea water inundation and ground water lens intrusion. This experiment was conducted on two group of giant swamp taro cultivars from Kiribati, namely the larger cultivar group ‘Ikaraoi’ and the smaller ‘Katutu’.

4.2 METHOD

As the available planting material was limited, different accessions of the same group of cultivars were combined, namely accessions CM/KB 05 and 06 for Ikaraoi and CM/KB 9 and 10 for Katutu. These accessions were obtained from the Secretariat of the Pacific Community (SPC), Centre for Pacific Crops and Trees (CePaCT). These accessions were imported by CePaCT from Kiribati in 2009. Suckers were prepared for importation by trimming the plant to 10x10cm at the acaulescent stem where the petiole and corm fused of which 5cm is corm.

101

The trimmed suckers were washed thoroughly with water, dead plant tissues removed and then tightly wrapped in paper and sealed in cartons for import. Upon arrival these suckers were trimmed and a meristem of 0.5cm removed for initiation. This meristem was then washed in 70% ethanol for one minute, followed by 10% bleach for 10 minutes, and then 5% bleach for 5 minutes and finally sterile water twice for 15 minutes. After this sterilization process the meristem was then cultured on solid basal MS (Murashige and Skoog, 1962) media of pH around 5.6-5.8. The accessions were selected based on the number of plants available.

4.2.1 Multiplication The initial number of plants obtained from the SPC-CePaCT, were multiplied to obtain the desired number of plants for the in vitro and in vivo testing this began in April, 2010. Multiplication/ bulking-up were achieved using taro tissue culture multiplication technique established by CePaCT. The methodology consists of a three step cycle whereby plantlet (of at least 1.5cm height) initiation is carried out in agar basal MS (Murashige and Skoog, 1962 with 30g/L sugar). After a standard three to four weeks of initiation the plantlet was transferred to MS medium containing 0.50mg/L thidiazuron (TDZ). Then after another three to four weeks of culture the plants were transferred to MS medium containing 0.80 mg/L 6- benzylaminopurine (BAP). This procedure effectively increased the number of plants available for experimentation in three months at an average of 30% increase per month on the initial number. All transfers of plant material were done in a Laminar airflow cabinet using sterilized forceps and scalpel. Each plant was removed and cleaned by removing dead and old plant tissues then the plants were trimmed down to a height of 3cm with 0.5cm corm attached. After cleaning the plants were subcultured into growth medium; care was taken to avoid unnecessary movement and therefore contamination. After multiplication the in vitro experiment began in August, 2010.

102

4.2.2 In Vitro Five different levels of salinity were tested namely 0.5% (5ppt), 1.0% (10ppt), 1.5% (15ppt) and 2.0% (20ppt) salt plus the control, 0% (0ppt) salt. These salinity levels were based on the salinity level giant swamp taro has been seen to tolerate which is up to 5ppt and the extreme it would have to encounter in cases of seawater inundation which may be up to 20-30ppt. Two types of salt solution were compared in this experiment, namely Artificial Sea Water (ASW) and Sodium Chloride (NaCl) salt solution. For the ASW, the different salinity levels under evaluation were prepared from a stock solution. While the NaCl solutions were made directly from pure NaCl. ASW was used as one salt solution, as a substitute for sea water in an attempt to be more representative of the atoll situation where intruding sea water is the cause of ground water salinity increment. The ASW was prepared by adding 13.96g of NaCl, 0.39g . KCl, 2.25g MgSO4 7H2O and 3.80g MgCl2 to 100ml double distilled water (Nyman, 1983). Aliquots from this ASW stock were used to make the various salinity concentrations. For the NaCl solutions, salt was weighed respectively and added to double distilled water (Table 4.1). This ASW recipe is based on the four major salts found in sea water, namely sodium chloride, potassium chloride, magnesium sulfate and magnesium chloride.

Application of these salt solutions was carried out in two ways. With the first approach salt solutions were added to the medium as it was prepared and then autoclaved. With the second approach salt solutions were applied to the top of the medium right after the plants had been subcultured on to medium (Figure 4.1). Plants of at least 3-4cm height were used for the in vitro experiment.

103

cultivars Type of treated with solution salt the 5 application solution salinity levels

Ikaraoi adding NaCl salt solution Katutu while making Ikaraoi medium ASW In vitro Katutu multiplic ation Ikaraoi adding NaCl salt solution Katutu after subcultu Ikaraoi ASW ring Katutu

Figure 4.1 In vitro Experiment treatment combination structure

Table 4.1 Salt solution mixtures Salinity ppt % salinity NaCl ASW Stock

5 0.50% 5g/L 5ml/L

10 1.00% 10g/L 10ml/L

15 1.50% 15g/L 15ml/L

20 2.00% 20g/L 20ml/L

104

Once prepared, the salt solutions were added to the medium according to the two methods described, namely, either mixed with the basal MS (Murashige and Skoog, 1962 with 30g/L sugar) then autoclaving at 1.05kg/cm2 (15psi) and 121ºC, or applied to the autoclaved medium after it has been prepared. For the latter, 5ml of each salt solution was pipetted into separate 6cm x 2cm tin screw cap glass bottles, which were autoclaved and the salt solution in it was added on top of the basal MS after subculturing. For both approaches liquid MS (without agar) with a pH of approximately 5.6-5.8 and 6cm x 7cm glass bottles with polycarbonate screw lids were used.

The control for the two experiments was the same 0% salt containing basal MS. Plants were subject to salinity treatments on an incremental basis with intervals of 0.5% salt (ASW / NaCl) per week, until the final salinity levels was reached. This prevented the plants from suffering shock due to a sudden increase in salinity. The salinity increments were applied in a staggered fashion so that once the final salinity level was reached for each treatment the duration of that treatment would be the same (Table 4.2). For example for the 2.0% salt treatment, culture in 0.5% salt increment started in week one and continued until week 4. While for the 1.5% salt treatment salinity the increment started on the second week and continued for three weeks whereby it reached its final salinity level in the fourth week. Similarly for 1.0% salt treatment, salinity applications began in the third week and ended in the fourth week (Table 4.3). Plants were subcultured into fresh media for each incremental increase of 0.5% using the two salt solution application approaches.

105

Table 4.2 Salinity increment weeks Final Salinity %

1 2 3 4

0.5% 0.5% 0.5% 0.5% 2.0

0.5% 0.5% 0.5% 1.5

0.5% 0.5% 1.0

0.5% 0.5

For each treatment combination, five replicates were prepared (Table 4.3). Plants were placed out randomly in blocks according to replicate number. The experiment proper began after each treatment had reached the desired salinity level and was conducted for eight weeks. The in vitro experiment cultures were kept in CePaCT’s growth room, illuminated for 16 hour photoperiods with Gro Lux tube lights at a light intensity of 4.4mW cm-2 and 25+/-2ºC room temperature. Plant response to the various salinity treatments was evaluated by taking morphology measurements on a weekly basis, namely plant height, number of leaves emerging, number of suckers emerging, corm size, root size and number of dying leaves. The initial and final plant weights were measured at the beginning and at the end of the eight week experimental period. The toxicity of the salinity levels was assessed by measuring the chlorophyll content of leaves at the end of the experiment. Number of contamination (fungal and bacterial colonies) on individual plant cultures was also recorded during the eight week experimental period.

106

Table 4.3 In vitro experimental design Treatment Cultivar Method T1 T2 T3 T4 T5 Group 0%salt 0.5%salt 1%salt 1.5%salt 2.0%salt (0g/L) (5g/L) (10g/L) (15g/L) 2(0g/L) NaCl Ikaraoi 5 5 5 5 5 (M1) Katutu 5 5 5 5 5 no. of plants 10 10 10 10 10 NaCl Ikaraoi 5 5 5 5 5 (M2) Katutu 5 5 5 5 5 no. of plants 10 10 10 10 10 ASW Ikaraoi 5 5 5 5 5 (M1) Katutu 5 5 5 5 5 no. of plants 10 10 10 10 10 ASW Ikaraoi 5 5 5 5 5 (M2) Katutu 5 5 5 5 5 no. of plants 10 10 10 10 10 Total plants per treatment 40 40 40 40 40 Total plants = 200

4.2.3 In Vivo Plants were subjected to five different levels of salt, namely 0.5%, 1.0%, 1.5%, 2.0% plus the control 0% salt (Figure.4.4). As with the in vitro method artificial sea water was used to mimic sea water as closely as possible. Plants were potted in 50 10x10cm black pots which stood in saucers to avoid run-off of the applied salt solutions. Pots were filled with 250 mL of Yates’s advance seedling common potting mix and CePaCT’s procedure for transferring the tissues cultured plants into the pots was followed. This included washing the growth medium gently off the plants ensuring that no part of the plant was damaged. The plants used in this in vivo experiment were at least 6cm in size from the base to the apex of the plant when removed

107 from the tissue culture bottles. The plants were firmly planted in the pots and a clear plastic bag was used to cover the plants to allow them to acclimatize to the environment and prevent shock and dehydration. Plastic bags were removed progressively over a one month period. For example plastic bags were removed for an hour the first day, the following day they were removed for two hours and so on until it was completely removed. Plants were watered with 40ml of tap water three times a week. From the time of transfer into the pots, the plants were kept in a shaded green house in the CePaCT at approximately 25+/-2ºC and after three months transferred to a typical green netted green house at approximate temperature of 27+/-2 ºC. In this green house plants were watered with 50ml of tap water three times a week. Plants were allowed to acclimatize for another month in this green house, and then 50ml of salt solution was applied on an incremental basis of 0.5% salt per week for four weeks. The experiment was carried out for two months after the five months of transfer and acclimatization phase of the plants.

Table 4.4 In vivo Experimental Design Treatment T1 T2 T3 T4 T5 0% salt 0.5% salt 1.0% salt 1.5% salt 2.0% salt Cultivar Group (0g/L) (5g/L) (10g/L) (15g/L) (20g/L)

Katutu 5 5 5 5 5

Ikaraoi 5 5 5 5 5 no. of plants per treatment 10 10 10 10 10 Total no. of plants 50

108

4.2.4 Evaluation Parameters For both the in vitro and the in vivo experiments the effects of the treatments were assessed mainly on the morphology of the plant except for the chlorophyll content measurement. The morphological traits assessed were height, number of leaves emerging, the number of suckers emerging, the corm size, root size and number of leaves dying. These traits were measured weekly, while weight was measured monthly due to the low degree of change. The chlorophyll content was measured at the end of the experiment. Contamination rates for the in vitro plants were assessed weekly to determine the effectiveness of method of application. Parameters such as height, corm size and weight were measured with a ruler or weighed on a bench top scale. For the in vitro tissues, measurements were taken from outside the culture bottles as both the liquid media and bottle are transparent and removing the plants would have caused contamination and plant death. Height was measured from the tip of the tallest leaf to the base of the petiole/stem, where corm growth begins. Corm was measured from the base of the petiole/stem, where corm growth begins to the corm base. Measurement of the roots was more difficult unlike that of plant height and corm size, as roots tended to coil in the small culture bottles, once they grew beyond 3cm. Hence roots were measured from base to root tip until they became too coiled to be measured in which case estimates of length were recorded. A digital bench top scale was used for weight measurements. The average weight of a culture bottle with the required media was subtracted from the weight of the culture bottles with media and plant, to obtain the actual plant weight at an accuracy of +/-0.001g.

The criteria used for counting an emerging leaf were when a leaf’s full leaf blade length (though not fully opened) became visible. When visually estimated leaves displayed 50% chlorosis, the leaf was counted as dead. The number of suckers was recorded if a sucker measured at least 0.5cm in length from base to tip. Contamination was recorded when spots of fungal or bacterial growth became visible at a colony diameter of 1-2mm.

109

4.2.4.1 Chlorophyll Content

After two months of exposure to the various treatments, leaves were analysed for chlorophyll content. 1 gram of leaf was soaked for 2 minutes in 5 ml of 80% acetone (80 ml acetone plus 20 ml distilled water). This was done to soften the leaves and allow for easy extraction of the chlorophylls. After two minutes the acetone was drained and the leaves ground using a mortar and pestle with 2 ml of 80% acetone. The extracted juice of 1.5 ml was then centrifuged in a micro centrifuge at 8000 revolutions per minute for 15 minutes at 5 °C. The supernatant was then transferred using a micropipette into a cuvette. The chlorophyll content was determined by measuring absorbency at two wavelengths Abs 664 and Abs 647 in spectrophotometer. The spectrophotometer was first loaded with the blank or controls (80% acetone) then the other supernants were loaded and readings recorded. After each reading the cuvettes were rinsed with 80% acetone twice and once with the next supernant that was to be loaded; this prevented contamination of the supernetants from the residue left in the cuvettes. From this reading chlorophyll content was calculated by;

Total Chlorophyll (µg/ml/g) = [7.04(Abs 664)+20.27(Abs 647)] x V/W V= volume of the leaf extract (ml) W = fresh weight of leaf (g)

4.2.5 Data Analysis The data was analysed using the Genstat statistics software and SPSS. The total plant response and the two cultivar group responses were compared against the five salinity levels using ANOVA followed by a Tukey’s post hoc test and liner regression analysis for the in vitro and in vivo. This included measurements of plant height, number of leaves, and number of suckers, root size, corm size, and number of dying leaves, weight and chlorophyll content.

110

A non-parametric Mann Whitney U (Wilcoxon rank sum) test was used to analyse the number of fungal contaminations recorded with the two methods of salinity application to the growth medium for the in vitro.

For the in vivo a percentage survival analysis was also done looking at the number of plants that were alive when the experiment ended.

4.3 RESULTS

Plants can tolerate up to 2.38 ppt of salt which is mildly brackish water, anything above 2.38 ppt maybe fatal (Munns and Tester, 2008). This experiment tested 0% , 0.5%, 1.0%, 1.5% and 2.0% salt concentration which translate to 0 ppt, 5 ppt, 10 ppt, 15 ppt and 20 ppt of salt concentration. This is in line with sea water that has a salinity level of 35-36 ppt. As stated earlier giant swamp taro Cyrtosperma merkusii has been reported to tolerate salinities slightly more than 2 ppt (Webb, 2007) however, until now there have been no specific investigations carried out to confirm this.

4.3.1 In Vitro Comparisons were done of the different salinity levels, the nature of the salt solution applied and the method of salt solution application. These comparisons were done in respect to the two cultivar groups individually and combined. A 100% survival rate was recorded for the range of salinity levels tested, whether using NaCl solution or ASW.

For the in vitro salt tolerance screening experiments, plants were replanted to a new medium on a weekly basis for three weeks to allow for the incremental increasing of the salinity levels until the treatment level was attained. In addition, no water is lost to the environment outside the culture bottles in an in vitro system; hence plants were subjected to constant salinity levels.

111

4.3.1.1 Cultivar Group Response

To assess the plant response of the two groups of cultivars Analysis Of Variance (ANOVA) was employed as this allowed the incorporation and accountability of the various factors affecting the experiment. For example, in this particular case the two cultivar groups were assessed in their response to the five salinity levels. Their individual response to the five salinities was compared for significant difference that may exist. For the two groups tested in the experiment, their response of biomass and toxicity was analysed this showed that, no significant difference existed in majority of the tested parameters. Tukey’s post hoc analysis was carried out on those parameters that did show significant difference. Table 4.5 shows the result of the analysis that indicates that significant difference existed randomly between salinity levels. Tukey’s post hoc analysis showed the Ikaraoi mean difference in corm size of 1.5% salt was higher in comparison to 0% and 0.5% salt, followed by 2% salt in comparison to 0% salt (p<0.05). Also the mean difference of 2% salt was significantly more than 0.5% salt for number of dying leaves. For Katutu significant differences existed in height where 0.5%salt showed a higher significant difference in mean than 0% salt. 1% salt showed higher mean difference in number of suckers and corm size compared to 0% salt. For weight 0.5% salt had a higher mean difference compared to 1.5% salt (p<0.05).

Table 4.5 Mean of cultivar group measured parameters when subjected to the five salinity levels. Values are mean+/-standard error of 20 replicates. Mean values in each column not sharing a common letter differ significantly (p<0.05) from each other (Tukey’s Post Hoc test) Katutu % Salt Height No. Suckers Corm Size Weight 0 -0.78+/-0.28a 0.80+/-0.17a 0.82+/-0.09a -0.42+/-0.05ab 0.5 0.33+/-0.23b 1.45+/-0.31ab 0.87+/-0.08ab -0.16+/-0.11ab 1 -1.95+/-0.23ab 2.7+/-0.39b 1.28+/-0.11b -0.20+/-0.21ab 1.5 -0.63+/-0.28ab 2.00+/-0.44ab 1.12+/-0.13ab -0.24+/-0.19ab 2 -0.17+/-0.2ab 2.00+/-0.44ab 1.11+/-0.13ab -0.33+/-0.08a

112

F 3.79 3.73 0.249 2.49 df 99 99 95 99 p 0.007 0.007 0.034 0.048

Ikaraoi %salt Corm Size No. Dying leaves 0 0.74+/-0.05a 2.6+/-0.11ab 0.5 0.93+/-0.09ab 1.7+/-0.24b 1 1.08+/-0.12ac 2.0+/-0.25ab 1.5 1.36+/-0.11c 2.4+/-0.28ab 2 1.21+/-0.12bc 2.7+/-0.25a F 5.79 3.14 df 90 99 p 0 0.018

4.3.1.2 Plant Response

As for the combined response of the varieties, from table 4.6 it can be seen that for the eight measured parameters for biomass and toxicity only plant height, number of suckers and corm size show significant difference (p<0.05). Further analysis through Tukey’s test showed that for height 0.5% salt had higher mean difference compared to 0% salt followed by 1.5% salt (p<0.05). While for number of suckers 1% salt had more suckers followed by 1.5% salt , then 2% salt in comparison to 0% salt (p<0.05). For corm size 1.5% salt had the highest mean difference compared to 0% and 0.5% salt, followed by 1% and 2% salt compared to 0% salt. To see if these significant difference in mean show any significant trend a liner regression analysis was carried out. This showed that height and root had no significant trend, while number of suckers, corm size and number of dying leaves showed an increase with increase in salt concentration (Graph 4.1, 4.2, 4.3).

The number of suckers increased with increases in salinity levels from 0 to 2% in both ASW and NaCl, similar to the results achieved by Abdel-hardy in 2006. Munns and Tester (2008)

113 outline the benchmark of salt tolerance in plants in two stages, tolerance of osmotic stress and tolerance of ionic stress. According to them tolerance to osmotic stress is achieved when plants are able to produce new leaves in response to increased salinity levels, while ionic tolerance is achieved when plants do not have early senescence of leaves. The experiment showed that in the presence of salinity levels up to 2% salt or 20 ppt plants were able to produce suckers and increase corm size with the trade-off of a nonlethal decrease in number of leaves content. Elimination of old leaves and emergence of new suckers may be a means of salt tolerance strategy for giant swamp taro. By eliminating old leaves where NaCl accumulates to toxic levels and investing in new suckers that have young leaves which are expanding and hence diluting NaCl concentrations (Munns and Tester, 2008) the plant reduces its net NaCl concentration. This is similar to the results achieved by Nyman (1983) where the taro callus was pale yet the callus produced leaves and survived.

The experiment had a 100% survival rate at all salinity concentrations with no prominent wilting. This is in line with the bench mark of salinity tolerance in plants by Munns and Tester (2008) which indicates that in vitro both the Ikaraoi and the Katutu group of cultivars were able to survive salinity levels of up to 2% salt which is approximately 56% sea water. The above response of plants to the various salinity levels is also in line with Zhu (2001) of plants exhibiting salt tolerance; according to him plants respond to increased salinity by detoxification, homeostasis and growth regulation. ROS which is produced in the chloroplast, under salt stress conditions damages the photosynthetic apparatus, enzymes and cellular membrane, detoxification is seen in the in vitro experiment as the plants under these conditions were able to photosynthesis and produce new suckers. Plants also did not show wilting which indicates that salt ions did not accumulate to a toxic level, hence exhibiting homeostasis. Also, with no reduction in height and emerging of new suckers the plants showed good growth regulation. These positive responses to the salt levels tested has also been seen in salt tolerance screening of salt tolerant plants by Fatokun et al (2002) in cowpeas and in wild einkom wheat by Yesayan et al (2009) and Colmer et al (2006) .

114

Table 4.6 Mean of plant response measured parameters when subjected to the five salinity levels. Values are mean+/-standard error of 40 replicates. Mean values in each column not sharing a common letter differ significantly (p<0.05) from each other (Tukey’s Post Hoc test) No. Dying %Salt Height No. Suckers Corm Size Root Size leaves 0 0.29+/-0.17a 0.70+/-0.13a 0.78+/-0.05a 3.64+/-0.18a 2.5+/-0.01ab 0.5 0.49+/-0.17b 1.35+/-0.24ab 0.90+/-0.6ab 2.76+/-0.22b 1.9+/-0.18b 1 0.15+/-0.15ab 1.9+/-02.8b 1.18+/-.08bc 3.12+/-0.25ab 2.2+/-0.18ab 1.5 0.17+/-0.17a 1.8+/-0.31b 1.24+/-0.08c 3.14+/-0.24ab 2.45+/-0.17ab 2 0.33+/-0.15ab 1.78+/-0.26b 1.15+/-0.08bc 3.45+/-0.22ab 0.99+/-0.15a F 1.09 3.89 6.91 2.36 3.66 df 199 199 186 186 199 p 0.002 0.005 0.000 0.055 0.007

Fitted and observed relationship

7 variety=CM910 6 Graph 4.1 shows variety=CM56 significant positive 5 trend in number of suckers in relation to 4 salt concentrations in liner regression 3 analysis

variety=CM910 1 variety=CM56

-0.5 0. 0 0. 5 1. 0 1. 5 2. 0 %salt

115

Fitted and observed relationship

2. 5 variety=CM910 Graph 4.2 shows variety=CM56 significant positive 2. 0 trend in corm size and salt concentrations in 1. 5 liner regression analysis.

1. 0 variety=CM910 variety=CM56

0. 5

0. 0

-0.50. 0 0.5 1. 0 1. 5 2. 0

%salt

Fitted and observed relationship Graph 4.3 shows 5 variety=CM910 significant positive trend in number of variety=CM56 4 dying leaves and salt concentrations in liner

3 regression analysis.

variety=CM910 variety=CM56 2

-0.50. 0 0.5 1. 0 1. 5 2. 0

%salt

116

4.3.1.3 Comparison of Salt Solutions Applied

NaCl which is the most common and abundant salt (Arzani, 2008) was tested against artificial sea water which is the main source of salt in the rise of the ground water salinity levels in the small atoll islands (White and Falkland, 2010). Despite the difference in the two salts constituents both had the same impact on the plants (Table 4.7).

This similarity of plant response to NaCl and ASW may be due to the percentage of the constituent salts present in the artificial seawater. NaCl dominates the constituents of seawater as it makes up an average 68.43% of the salts present in it. While the other salts namely calcium chloride, magnesium chloride and potassium chloride form a small fraction (Bedjaian and Loukhovitskaya, 2011). This high percentage of NaCl and its ions Na+ and Cl- are the most detrimental of the four major salts present in sea water (Arzani, 2008), hence NaCl would have had the most impact. Consequently the NaCl used in isolation and the NaCl present in the artificial seawater which is 80.6% of its total molecular weight, would have similar impact on the plants, as such either ASW or NaCl can be used in a salt screening experiment without compromising the experiment.

Table 4.7 Plant Response to ASW and NaCl Parameter f Mean Probability ASW NaCl Height 0.61 0.11+/-0.16 0.05+/-0.16 no. leaves 0.51 12.09+/-0.15 1.99+/-0.15 no. suckers 0.26 1.86+/-0.27 1.55+/-0.27 corm 0.21 1.08+/-0.09 0.97+/-0.09 root 0.23 2.72+/-0.25 3.02+/-0.25 No. dying leaves 0.88 2.36+/-0.17 2.34+/-0.17

117

Weight 0.46 -0.42+/-0.12 -0.51+/-0.12 chlorophyll content 0.48 15.71+/-1.71 14.48+/-1.71 With respect to the two salt solutions tested ASW and NaCl, table 4.7 summarizes the ANOVA results of the response plants had to the two types of salt solutions (ASW and NaCl) applied in the various concentrations. The table also shows the f probabilities of the measured parameters.

4.3.1.4 Comparison of the Method of Salinity Application

Use of tissues culture and in vitro techniques requires great care as the cultures can be easily contaminated and as a result die out. Therefore, in any screening methodology conducted in vitro it is essential that proper techniques such as subculturing and solution application be critically analysed to avoid contamination and loss of samples. This experiment tested two approaches to the salt (ASW and NaCl) solution application; the first approach was direct application of the salt solution to the medium while mixing, followed by autoclaving. The second approach was salt solution application to the medium following subculture of the plant in the medium. Within the duration of the experiment a total of four contaminations were recorded where salt solution was mixed in medium and 14 contaminations where salt solution was applied from on top after subculturing. The contaminations were mainly white fungus that appeared on the medium; the fungus initially present as spots grows and covers the plant resulting in deteriorated plant health and death.

Where salt solution was applied from “on top” there is significantly more contamination compared to where solution was “made with media”. A total of 77.8% more contamination results when salt solution was poured from on top after subculturing at significance of statistical difference of 0.013. Post hoc analysis could not be carried out as there are only two groups for comparison (Table.4.8).

The second approach allows for human handling errors to occur increasing the chance of contaminants to enter the culture. In the first approach the salt solution is already added to the

118 medium which is then sterilized, after which plants are subcultured onto the medium. In the second approach, plants are subcultured into the prepared medium and then the salt solution is poured on top. This requires extra movement and increases the time window during which the nutrient rich medium is exposed to the outside atmosphere in the Laminar airflow and hence the increased contamination rate. Thus mixing testing solutions with the medium during preparation is a safer approach that does not compromise the experiment samples or the results.

Table 4.8 Contamination Rate according to application method Application of Solution Contamination Approach 1 ( salt solutions mixed with media) 4 Approach 2 (salt solutions poured from on top after subculturing) 14 Significance probability 0.013

4.3.2 In Vivo In vivo treatments were planned to be conducted for eight weeks. However, after three weeks the plants were affected by the salinity and started to die, hence the experiment was concluded and results recorded. Due to the approach used in the building up to the final salinity level only the first four treatments were established with 2% omitted. Of the four salinity treatments 1% and 1.5% had zero percent survival hence limited parameters were available for measurement such as the number of leaves emerging, rooting, dying leaves and chlorophyll content. There were no significant measurements available for height, corm, and weight due to the premature conclusion of the experiment and no suckers had emerged in the short period of time. This was the same for all the four treatments.

119

Unlike in vitro, the plants in the in vivo did not grow so well. After a successful acclimatization phase of four months in the green house, plants were not able to tolerate the salinity levels. From the second week of the incremental phase plants started to wilt with heavy chlorosis of leaves, by the third week of the increment phase plants had died out. This allowed only three increments to occur, the highest being 1.5% salt. Therefore the final salinity levels tested in vivo were 0%, 0.5%, 1.0% and 1.5%. Artificial seawater was used as the salt solution, to mimic the actual scenario faced by giant swamp taro on the atoll islands.

Conducting experiments in vitro allows control over the environmental factors such as temperature, light, heat and moisture. While with in vivo experiments in the green house, these factors cannot be controlled, leaving plants exposed to the elements of nature (Munns and James, 2003). The green house was located in an elevated area where it was quite windy; however one side of the green house was against an excavated hill, shielding it form the wind and sun. In the first week it was noticed that plants on this shielded side were fresher and had water remaining in their saucers while plants on the side exposed to wind and sun were not. Water loss due to evaporation and transpiration is one of the factors that might have resulted in the low survival rate of the plants in the green house. Evaporation and transpiration reduces the amount of water in the pots and moisture in the soil, resulting in an increase in the initial salt concentrations. Table 4.9 reports the salinity levels measured in the pots at the end of the experiments which shows a significant increase in the salt concentrations from the applied 0.5% to 1% , from the applied 1% to 3% and from 2% to 5% salt.

Table 4.9 Pre and post experiment salinity levels

Pre salinity % Post salinity % 0 0 The table shows the pre salinity percentage 0.5 1 which is the initial salinity levels applied to 1 1.5 the plants and the post salinity percentage which is the salinity level recorded from the 1.5 1.5 pot saucers or effluent after the experiment.

120

Furthermore, in vitro plants are subjected to optimum conditions such as temperature, light intensity, wave length and duration (Hughes, 1981). While in vivo, plants are exposed to variable temperatures, humidity, light intensity and duration (Arzani, 2008). In an in vitro culture complex reactions with temperature, heat, light and growth medium result in a gaseous phase in the culture bottles (Hughes, 1981). An in vivo system lacks this and instead the porous soil gives greater surface area to volume ratio for evaporation to take place. All of these factors not only add to water loss from plant and soil, but also affects the physiological mechanisms operating to maintain plant ionic and osmolic homeostasis (Zhu, 2001).

Looking at the 100% survival rate of the control where no salt was applied it can be seen that the plants had successfully acclimatized, hence salinity application before plants could acclimatize can be ruled out as a cause of the low survival rate. After the application of the three salinity levels plant health started to deteriorate resulting in death, as can be seen by the 60% survival rate in 0.5% salt and the zero percent survival rate of 1.0% and 1.5% salt (Table 4.10). This was the same for both the cultivars and when combined the same results were achieved for the overall plant response (Table 4.12).

4.3.2.1 Cultivar group response

Since the plants of the 1.0% and 1.5% salt had died out analysis was done on the control as “no salt” to the 0.5% salt as “salted”. Comparison of the two cultivar groups showed that they both had the same response to the applied salinity levels of 0% and 0.5% salt (Table 4.10). There was no increase in corm size or production of suckers in any of the treatments for either of the cultivars in the three weeks, thus these parameters were not evaluated. Of the four evaluated parameters of number of leaves, rooting size, number of dying leaves and chlorophyll content, none recorded f probabilities less than 0.05. Indicating that, none of the evaluated parameters of the two group of cultivars have any significant difference. Both had the same response to the salinity levels tested. This similarity in response also stands with their rate of survival when subjected to the increased levels of salinities. Both the cultivars

121 have the same response to the subject salinity levels with 100% survival at 0% salt, decreasing to 60% survival at 0.5% and 0% survival rate at 1.0% and 1.5% salt (Table 4.11).

Table 4.10 Cultivar group response to the salinity levels Parameter f probability Ikaraoi Katutu No. Leaves 0.437 2.23+/-0.69 2.28+/-0.71 Root 1 1.750+/-0.47 1.75+/-0.47 dying leaves 0.681 3.9+/-0.79 3.7+/-0.82 Chlorophyll 0.828 20.26+/-1 22.26+/-1 The above table is the summary of the calculated ANOVA values with comparison of the two cultivar groups Ikaraoi and Katutu. It has the f probabilities or the probability of any significant difference of the number of leaves, rooting, and number of dying leaves and chlorophyll content of the two cultivar groups.

Table 4.11 Percentage survival rate of the two group of cultivars Cultivar group % salt no alive no. dead % survival Ikaraoi 0 5 0 100 0.5 3 2 60 1 0 5 0 1.5 0 5 0 Katutu 0 5 0 100 0.5 3 2 60 1 0 5 0 1.5 0 5 0 The above table represents the percentage survival rate of the five replicate of the two group of cultivars Ikaraoi and Katutu to the various salinity levels

122

4.3.2.2 Plant response

With a f probability of less than 0.05, rooting size, number of leaves emerging and number of leaves dying has significant differences in the response to the two salinity levels. Number of leaves emerging and number of leaves dying increases with increase in salinity, while rooting decreases. Roots in the 0% salt treatment were fibrous, healthy and well developed while the roots in the 0.5% salt treatment had died/melted towards the root tips but the majority of the root system was alive and healthy. However, there was no significant difference in the chlorophyll content (p>0.05). In plants subjected to 0.5% it was seen that while old leaves wilted and died, new leaves had sprouted. Furthermore, despite the wilting of old leaves the chlorophyll content of the new leaves was the same as the control plants (p= 0.863) (Table 4.12).

With wilting of leaves it can be said that the plants had experienced the first stress, the osmotic stress, this response is consistent with many experiments such as the work done by Antonio and Weber in 1999 to Aghaeri in 2008. Following the osmotic stress, ionic stress takes its toll resulting in early senescence of old leaves. However at 0.5% salt some level of ionic and osmotic homoeostasis might have been obtained resulting in new shoots. Thus, it can be seen that all the plants subjected to 0% salt survived while those subjected to higher salinities died out such as 0.5 % salt where 4 plants died, while all the plants of 1.0 % salt and 1.5% died giving a percentage survival rate of 0% (Table 4.13). The two cultivars Ikaraoi and Katutu cannot survive salinity levels of more than 0.5% (5ppt) of artificial sea water in vivo. Therefore more experiments need to be done on these groups of cultivars of giant swamp taro to find cultivars that can tolerate salinity levels of more than 0.5% (5 ppt) (Figure 4.2).

Table 4.12 Plant response to the various salinity levels Parameter f probability 0% 0.50% No. Leaves 0.04 2+/-0.47 2.5+/-0.85 Root 0.002 2+/-0.47 1.5+/-0.53

123 dying 0.001 1.6+/-0.52 6+/-0.47 leaves Chlorophyll 0.863 19.48+/-1 23.05+/-1 The above represents the calculated Anova values of the number of leaves emerging, rooting, number of dying leaves and chlorophyll content with their respective mean values. These calculated f probabilities and means are of the total plant response to the various salinity levels tested. Post hoc analysis could not be carried out as there are only two groups for comparison.

Table 4.13 Percentage survival rate of the various salinity levels

% salt no alive no. dead % survival 0 10 0 100 0.5 6 4 60 1 0 10 0 1.5 0 10 0 The above shows the percentage survival rate of the ten replicates of the plants to the subjected salinity levels of 0%, 0.5%, 1.0% and 1.5% salt.

Figure 4.2 Overall morphological responses to salt applications, plants from the left; 0%, 0.5%, 1.0% and 1.5% salt.

124

5.0 DISCUSSION

Giant swamp taro is a local food crop of the Pacific and an everyday food source for the atoll islands, and is also a neglected and underutilized crop species in the region. Neglected in the sense that only the traditional farmers in the atolls islands cultivate it and even here its use is now fast being eroded by changing preferences of food and life style. Coupled with the effects of climate change, giant swamp taro is threatened through loss of its diverse range of cultivars and traditional cultivation knowledge. It is disturbing to see that despite the importance of this unique food crop in terms of food security, traditions and identity, the scientific and common knowledge of the crop is limited and restricted by the lack of research.

This research reveals some of the very important factors that have led to the drop in giant swamp taro diversity and production. As seen in the Tuvalu case study climate change may be one such factor as increases in sea level and the increased frequency and intensity of storms likely threaten this crop. Disturbance in the ground water lens or seawater inundation results in increased ground water salinity levels. Coastal erosion due to sea level rise may be another contributing factor, however the effect of coastal erosion or displacement on the fresh ground water lens is relativity under-investigated and needs further research.

Apart from climate change, anthropogenic factors are also large contributors to the decline in giant swamp taro production and the erosion of traditional knowledge. These anthropogenic factors consist of direct impacts on the fresh water lens and impact on the giant swamp taro. Direct impact on the fresh water lens such as in the case of Funafuti includes disturbances caused by construction, pollution, development pressure, extraction and population pressure (Webb, 2007). This causes enhanced mixing in the transition zone of the lens. While impacts on the giant swamp taro include land allocation and disputes, preference of an easier lifestyle, white collar jobs and western foods that are easier to cook and does not require the strenuous task of cultivation.

125

The case study survey on Tuvalu gives a glimpse of the incidence of ground water salinity levels. It was seen that in overall comparison to 2006 salinity levels in 2010 were slightly higher on one island. From the survey it was found that giant swamp taro was able to tolerate salinity values well above those stated in past literature works (Dunn, 1976; Webb, 2007; Manner, 2009) which was also seen in the in vitro and in vivo experiment. Giant swamp taro cultivars from Tuvalu had been collected during the survey and so were not available for the salt screening experiments as giant swamp taro takes a long time to grow and to multiple. This was especially evident on Funafuti where the highest ground water salinity levels and the second highest percentage increase in salinity were found. Nui ground water salinity values fell well in between the fresh water zone of 1500-2500 µS/cm. Similarly for Niutao which had quite fresh overall ground water, while Nanumaga and Nanumea had no change in salinity levels.

Atoll islands in Tuvalu have a small low lying topography, making their ground water lens highly vulnerable, variation in groundwater salinity levels existed. This is due to the variation in size, amount of development pressure, population pressure and natural disasters experienced by the atoll islands. For example Funafuti has the highest groundwater salinity levels as it has the most vulnerable topography, it has also experienced the most development and populations pressures. While, Nanumaga has experienced no increase in ground water salinity as it has a wider topography and has not experienced extensive development and population pressures.

Moreover, apart from the finds of the salinity survey, the in vitro and in vivo salt tolerance screening also provided significant information. Both the in vitro and in vivo experiments showed that no significant difference in response to rise in salinity levels exist between the two cultivar groups Ikaraoi and Katutu. It also revealed that this two groups of cultivars could tolerate all the four salinity levels up to 2% or 20ppt in vitro. While in vivo the two group of cultivars could only tolerate up to 0.5% or 5ppt. In vitro results also revealed that no

126 significant difference existed between the impacts of sodium chloride and artificial sea water on the plants. Also that mixing salt solution while making the growth medium resulted in 77.8% less contamination compared to applying salt solutions after subculturing.

The difference in salinity tolerance response of plant in vitro and in vivo is due to the type of exposure the plants get in the two systems. In an in vivo system plants are exposed to variable temperature, light intensity, photoperiod duration, humidity, heat and wind, while in vitro systems are enclosed with optimum growth conditions. This exposure in an in vivo system causes unaccounted increase in salinity as soil and plant water evaporates, dehydration, disturbances in plant ion and osmotic homeostasis, hence resulting in lower salt tolerance seen in plants in vivo than in vitro.

The giant swamp taro salt tolerance rapid screening methodology used in the research, is practical enough for the Pacific where there is lack of technical and financial resources. This rapid screening methodology basically involves screening in vitro and in vivo with either sodium chloride or artificial sea water at five equal interval levels of salinities which can be subjected to change depending on the aim of the experiment. While the methodology used gives a strong foundation for rapid salt tolerance screening, there is still a lot of improvement that needs to be done. The methodology can be improved by using larger sample populations and using cultivars instead of group of cultivars for both in vitro and in vivo. Also constant monitoring of soil salinity levels in an in vivo system is needed to avoid unaccounted increases in applied salinity levels.

The ground water salinity survey from Tuvalu revealed that the swamp taro can tolerate 8000µS/cm (5.36ppt), which is more than the 3000µS/cm (2.01ppt) (Dunn,1997, Webb, 2007) limit stated in past literature. Also both the in vitro and in vivo showed similar results, where swamp taro effectively tolerated 5ppt of salt. Hence it can be said that the giant swamp taro has the capacity to tolerate 5ppt of salt.

127

6.0 CONCULSION

Despite the many factors affecting the decline in traditional knowledge and production, the status of the giant swamp taro can be improved. Farmers need to be encouraged to pay more attention to their crops and take pride in cultivating the many cultivars. Local authorities need to take more initiative in working with farmers to revive the culture of giant swamp taro production. This action from farmers and governments would not only increase the diversity of giant swamp taro but also increase the variation of climate ready traits present in the crop gene pool. This variation is essential in development of improved crop cultivars that can buffer against not only climate change but also pest and diseases. This can be achieved by education, awareness, genetic resource sharing and conservation. For example the Tuvalu ground water salinity survey and giant swamp taro descriptor development in Pohnpei, carried out in the research acted as a mechanism for creating awareness, promoting genetic resource sharing and conservation. The giant swamp taro cultivars collected from Tuvalu and Pohnpei while doing the research are now being conserved and duplicated at CePaCT.

Government and the local authorities can promote education, awareness of giant swamp taro and its cultivation by incorporating this knowledge in the local school curriculums. Gardening competitions in schools that assess the diversity and health of giant swamp taro can help get the younger generation to get involved. This would result in students wanting to learn more about how to cultivate the crop, the cultivars present and how to distinguish between them, hence keeping the traditional knowledge alive.

Awareness can also be created by actively involving farmers and women in the atoll islands in creating farm gene banks and in workshops that promote cultivation of local food crops. Governments can promote and encourage, scholarship and funding providers such as AusAid and European Unions to focus on capacity building and research on giant swamp taro. This would ensure food security and expand the current scientific knowledge base of the crop in areas of drought and salt tolerance.

128

Furthermore, to encourage and expand the diversity range, sharing of genetic resources is important. Hence local government’s willingness towards sharing of giant swamp taro cultivars is essential, as this sharing of genetic resources also leads to ex suit conservation.

Finally, with the many key information provided on ground water salinity, cultivar descriptor and the rapid screening methodology atoll island communities can now better deal with the impacts of climate change. This can be done by classifying and screening the diverse range of cultivars for variation in salt tolerance. In doing so, not only can we develop salt tolerant crops but we can also ensure food security and security of the traditions of the Pacific island countries.

129

Bibliography Abdel-Hardy, M.S. 2006. In vitro prediction of salt tolerance in Wheat. 2nd International conference on water resources and arid environments. Retrieved from http://www.nrc.sci.eg/nepad/Nepad_Members/Wheat_Project

Aghaeri. 2008. In vitro screening of potato Solanum tuberosum L. cultivars for salt tolerance using parameters and RAPD analysis. American-Eurasian Journal of Agriculture and Environmental Science , 3 (2): 159-164.

Agroforestry in Micronesia. 2004. Traditional agroforestry in the high lands of Micronesia. Retrieved from Http://www.unu.edu/unupress/unubooks/80824e/ 80824Eof .html

ANU Cartographic Services, R. A. 2008. Oceanic encounters. Retrieved 11 17, 2010, from http://epress.anu.edu.au

Arzani, A. 2008. Improving salinity tolerance in crop plants: a biotechnological view. In vitro cell Development Biology plant, 44.

Aung, T., A. Singh, and C.L. Prasad. 2009. Sea level threat in Tuvalu. American Journal of Applied Sciences, 6 (6): 1169-1174.

Banakar, M., and Ranjbar, G. 2010. Evaluation of salt tolerance of Pistachio cultivars at seedling stage. American-Eurasian Journal of Agricultural and Environmental Science , 9: 115-120.

Bedjaian, Y and E. Loukhovitskaya. 2011. Absorbtion of water vapour on MgCl2x6H2O salt surface. Atmospheric Environment, 45 (14); 2373-2378.

Bindoff, N.J. 2007. Observations: Oceanic climate change and sea level. In S. Q. Solomon, Climate Change 2007: The Physical Science . Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press, United: 385-429.

130

Brandburry, J.H and W.D. Holloway. 1988. Chemistry of tropical root crops; significance for nutrition and agriculture in the Pacific. ACIAR Monograph , 6: 201.

Brown, D. 2000. Aroid plants of Arum family (2nd ed). USA: Timber press.

Chen. Z, I.I. Pottosin, T.A. Fuglsang, M. Tester, D. Jha, I .Zepeda-Jazo, M. Zhou, M.G. Palmgren, I. A. Newman and S. Shabala. 2007. Root plasma membrane transporters controlling K+/Na+ homeostasis in salt-stressed Barley. Plant physiology , 145: 1714-1725.

Colmer. T.D, T.J. Flowers, and R. Munns. 2006. Use of wild relative to improve salt tolerance in Wheat. Journal of Experimental Botany , 57 (5): 1059-1078.

Covich, A. 2006. Dispersal of limited biodiversity of tropical insular streams, 54 (4): 523- 547.

Cuarteo. J, M.G. Bolarin, M.J. Asins, and V. Moreno. 2006. Increasing salt tolerance in Tomato. Journal of Experimental Botany , 57 (5): 1045-1058.

Cyrtosperma merkusii. 2006. Retrieved Febuary 24, 2010, from http://zipcodezoo.com/plants/c/c-m/24/2/2010

D’Antonio, C., and E. Weber. 1999. Germination and growth resaponse of hybridizing Carpobrotus species (AIZOACEAE) from coastal California to soil salinity. American Journal of Botany , 86 (9): 1257–1263.

Deenik, J.H and R.S. Yost. 2006. Chemical properties of atoll soil in the Marshall Islands and Constrains to crop production. Geoderma, 136 (3-4), pp. 666-681

Deo, P., A.P. Tyagi, M. Taylor, D.K. Becker, and P.M. Hardy. 2009. Improving Taro (Colocasia esculenta) production using biotechnological approaches. The South Pacific Journal of Natural Sciences, 27: 6-13.

Dunn, E. 1976. Report on a visit to the island of Tuvalu. Noumea, New Caledonia: SPC.

131

Eid, E., and C.H. Huisbergen. 1992. Sea level rise and coastal zone management. Default Hydraulics: 471.

Englberger, L., G.C. Marks, M.H. Fiztgerald, and K. Lippwe. 2005. Food composition data from the Frederated States of Microneasia. Micronesica , 37 (2): 195-213.

Englberger, L., K. Albert, A. Levendusty, C.O. Hogrlima, M. Sackryas, and A. Lorens. 2007. Report on a visit to Pingelap atoll, Pohnpei, FSM: documentation and promotion of local food crops. Pohnpei, FSM: Island food community of Pohnpei.

Englberger, L., W. Aalbersberg, P. Ravi, E. Bonnin, G.C. Marks, M.H. Friztgerald, J. Elymore. 2003. Further analysis on Micronesian banana, taro, breadfruit and other foods for provitamin A carotenoids and minerals. Journal of Food Composition and Analysis , 16.

Falkland, T. 1999. Water management for Funafuti, Tuvalu. Australian Agency for International Development.

Fatokun, C., S.A. Tarawali, B.B. Singh, P.M. Karmawa, and M. Tamo. 2002. Challenges and opportunites for enhancing sustainable cowpea production. Ibadan, Nigeria: Proceeding of the world cowpea production III.

Flowers, T. 2003. Improving crop tolerance. Journal of Experimental Botany , 55 (396): 307- 319.

Fricke,W and W.S. Peters. 2002. The biophysics of leaf growth in salt-stressed Barley. Plant physiology , 129: 374-388.

Garnett, M.R, S.J. Murch, S. Krishnara, M.A. Dixon, and R.K. Saxena. 2002. The rizofliteration of sodium from hydroponic fluid using scented Geranium. Water, air and soil pollution , 140: 343-366.

132

Gerald, D., M.S. Fenster, B.A. Argow, and I.V. Buynevich. 2007. Coastal Impacts Due to Sea Level Rise. Annual Review of Earth and Planetary Sciences, 36:601-647.

Hather and Weisler. 2000. Prehistoric Giant Swamp taro (Cyrtosperma merkusii) from Herderson Island, Southeast Polynesia. Pacific Science , 54: 149-156.

Hein'andez. A, X. Jiang, B. Cubero, P.M. Nieto, R.A. Bressan, P.M. Hasegawa, and J.M Pardo. 2009. Mutants of the Arabidopsis thaliana cation/H+ antiporter AtNHX1 conferring increased salt tolerance in Yeast. Journal of Biological Chemistry , 284 (21): 1427-14285.

Hughes, K. W. 1981. In vitro ecology; Exogenesis in plant culture systems. Environmental and Experimental Botany, 21 (3-4); 281-288.

Ivancic, A. 1992. Breeding and genetics of Taro Colocasia esculenta (L.) schott.,. Honiara, soloman Islands: United nations Development Programme,Pacific Printers.

Jackson, G. 1980. Diseases and pest of taro South Pacific Commission. Noumea, New Caledonia: South Pacific commission.

James, R., and R. Munns. 2003. Screening methods for salinity tolerance; a case study with teraploid wheat. Plant and Soil , 253: 201-218.

Kader, A., M.Aly, and Esaway, M. 2011. Inoculation of Geobacillus cyldaylosilyticus IRD into Maize (Zea mays L.) to sustain tolerance against high salt stress. Journal of American Science , 7 (1): 71-79.

Kchaou, H., A. Lard, K. Gargouri, M. Chareb, F. Morales and M. Msallen. 2010. Assesment of tolerance to NaCl salinity of five cultivars, based on growth characteristics and Na+ and Cl-. Scientia Horticulturae , 124 : 306-315.

Lal, P. N., Kinch, J., and F. Wickham. (2009). Review of economic and livelihood impact assessments of, and adaptation to, climate change in Melanesia. Honolulu, HI: The Bishop

133

Museum. Retrieved from CAKE: http://www.cakex.org/virtual-library/Review-of-Economic- and-Livelihood-I... to-Climate-Change-in-Melanesia

Lambert, M. 1982. Taro cultivation in the South Pacific. Noumea: New celadonia, Maxwell printing.

Liska, A. J., A. Shevchenko, U. Pick and A. Katz. 2004. Enhanced photosynthesis and redox energy production contribute to salinity tolerance in Dunaliella as revealed by homology- based proteomics. Plant Physiology , 136: 2806-2817.

Liz. M. 2007. 'The Tides are getting higher and higher': A Pacific voice on Climate change. Just Change , 10: 22-23.

Manner, H.I. 2009. Farm and Forestry Production and Marketing Profile for Giant Swamp Taro (Cyrtosperma chamissonis). In: Elevitch, C.R. (ed.). Specialty Crops for Pacific Island Agroforestry. Permanent Agriculture Resources (PAR), Holualoa, Hawai‘i. http://agroforestry.net/scps.

Mayo, S., and J. Boyce, P. C. 1997. The genera of Araceae. Belgium: The Trustees, royal Botanic Gardens, Kew.

McGinley, M and J. E. Duffy. "Atoll". In: Encyclopedia of Earth. Eds. Cutler J. Cleveland. First published in the Encyclopedia of Earth March 29, 2010; Last revised Date March 29, 2010; Retrieved October 23, 2011.

Metai, E. 2002. Vunerability of freshwater lens on Tarawa-The role of hydrological monitoring in determinig sustainable yield. Kiribati: Pacific Regional Consultation on Water in Small Island Countries.

Metutera, T. 2002. Water Managment in Kiribati with special emphasis on groundwater development using infilteration galleries. Kiribati: Public Utilities board.

134

Mimura, N. 1999. "Vulnerability of island countries in the South Pacific to sea level rise and climate change," Climate Research,12: 137-143.

Munns, R., and M. Tester. 2008. Mechnisms of Sslinity tolerance. Annual Review of Plant Biology , 59: 651-681.

Nakano, K. 2001. The energy supply of food stuffs in the Yap Islands; A preliminary report for Kagoshima University Reserach center for the Pacific Islands. Occasional Paper (34): 95- 100.

Nawaz. K., A. I. 2010. Induction of salt tolerance in two cultivars of Sorghum (Sorghum bicolour L.) by exogenous application of Proline at seedling stage. World Applied Sciences Journal , 10 (1): 93-99.

Nyman, L. C.J. Gonzales and J. Arditti. 1983. In-Vitro selection for salt tolerance of Taro Colocasia esculenta var antiquorum. Annals of Botany , 51: 229-236.

Onwueme, I.C., 1999. Taro cultivation in Asia and the Pacific. [Internet] FAO Regional Office for Asia and the Pacific, Bangkok, Thailand. Retrieved from http://www.fao.org/DOCREP/005/AC450E/ac450e00.htm#Contents

Pajuabmon, A., P. Theerakulpiut, B. Kijwijan and N. Muangsan. 2009. In vitro investigation of salt tolerance characteristics of rice seedlings. Research Journal of Agricultural and Biological Sciences, 5 (4): 423-427.

Peel, M. D, B. L. Waldron, K. B. Jensen, N. J. Chatterton, H. Harton and L. M. Dudley. 2004. Screening for salinity tolerance in Alfalfa:A repeatable method. Crop Science , 44: 2049- 2053.

Rigby, G. 2008. A Climate for Change. New South Wales, Australia. Retrieved from http://www.nswchurches.org/Resources/Articles/A08062_Rigby_A_Climate_for_Change.pdf

135

Rodgers, J. 2009. Climate change and health; a Pacific perspective. Commonwealth Health Ministry. Retrieved from http://www.thecommonwealth.org/files/190386/ FileName /SPCregionalpaper_2009.pdf

Rotar, P. P., L.D. Plucknett and B.K. Bird. 1978. Bibliography of Taro and other edible aroids. Hawaii agricultural experiment sation.

Rozell, D. 2007. Quantifing the impacts of Gobal Warming on salt water intrusion at the Shutter Island, New York using groundwater flow Model. New York: Stony brook University.

Rubio. L., A. Rosada, A. L. Rueda, A. Borsani, M. J. Garcia-Sanchez, V. Valpuesta, J. A. Fernandes and M. A. Botella. 2004. Regulation of K+ transport in Tomatoe roots by TSS1 locus; implications in salinity tolerance. Plant Physiology , 134 : 452-459.

Sakhanokho. H. F and R. Y. Kelley. 2009. Influence of salicylic acid on in vitro propagation and salt tolerance in Hibiscuc acetosella and Hibiscus. African Journal of Biotechnology , 8: 1474-1481.

Satoh, K and N. Murata. 1998. Stress response of photosynthetic organisms. Amsterdam, Netherlands: Elsevier:133-148.

Schaeffer, M. 2010. Persistence of atoll islands under recent projected sea level rise. Climate Analytics :1-6.

Schmidt, C. 2005. Spheres of influence: keeping afloat: A strategy for small island nations. Environmental Health Perspective , 113 (9): 606-609.

Shaddad, H. M. 2010. Review; Salt tolerance of crop plants. Journal of Stress Physiology and Biochemistry , 6 (3): 64-90.

Shen, S. 2007. Noah’s ark to save Tuvalu. Just Change (10): 18-19.

Smith, A. 2009. Sea level rise and the vunerability of coastal people. Intersection, , 7. Retrieved from http://www.ehs.unu.edu/file/get/4097.

136

Soumaya. A. 2010. QTLs mapping of morphologicial traits related to salt tolerance in Medicago truncatula. Acta Plant physiology.

Staples, R. C., and G. H. Toenniessen 1984. Salinity tolerance in plants. Wiley-Interscience.

Star, C. 2004. Climate justice campaigns and Environmental Refugees. Nathan.: Griffith University. Retrieved from http://eprints.usq.edu.au/2530/1/climate_justice _campaigns_and_environmental_refugees.pdf

Subramanyam. K., K.V. Sailaja, K. Subramanyam, D. M. Rao and K. Lakshmidevi. 2010. Ectopic expression of an osmotin gene leads to enhanced salt tolerance in Transgenic chilli pepper (Capsicum annum L.). Plant Cell Organ Tissue Culture , 105: 181-192.

Krauke, Y., and H. Sychrová. 2010. Chimeras between C.glabrata Cnh1 and S.Cerevisiae Nha1 Na+/H+ antiporters are functional proteins increasing the salt tolerance of yeast cells. Folia Microbil , 55 (5): 435-441.

Talia, M. 2009. Towards fatele Theology:A Contextual Theological respoense in addressing threats to global warming in Tuvalu. Taiwan: Tainan Theological College and Seminary Tainan.

Temleton, J.K., and M.H. Lambert. 1975. FAO mission report on root crops; Development in the Pacific. South Pacific Commission.

Thaman, R. 2002. Threats to Pacific island biodiversity and biodiversity conservation in the Pacific Island. Suva,Fiji. University of the South Pacific. Retrieved from http://devnet.anu.edu.au/online%20versions%20pdfs/58/0758 Thaman.

Tong,Y and H. Dongxian. 2010. A photoautototrophic in vitro system for evaluating salt tolerance of soybean (Glycine max L) plants. Journal of Tropical Agriculture, 48 (1-2): 40-44.

Tyagi, A. D. 2009. Seed germination and seedling development in Taro (Colocasia esculenta). The South Pacific Journal of Natural Sciences , 27: 6-13.

137

Iese,V 2005. Characterisation of Giant swamp taro ( Cyrtosperma chamissonis (Merr) Schott) using morphological features, genetic finger printing and tarditional knowlegde from farmers. Suva, Fiji: University of the south Pacific.

Vickers, M. 1982. The agromony of Cyrtosperma chamission.(schott)Merr. In Kiribati.

Wagih, M. 1997. Eradication of taro viruses from seedlings via seed rescue culture coupled with thermotherapy. African Crop Science Journal , 5 (4): 419-424.

Wang, Y., J. Huang, R. Lu, Z. Chen, T. He and R. Zhou. 2009. Induction and in vitro selection of variants with salt tolerance from Brassica chinensis. Journal of Nuclear Agricultural Science , 23 (6): 951-957.

Webb, A. 2007. Assessment of salinity of groundwater in swamp taro (Cyrtosperma chamissonis) “Pulaka” pits in Tuvalu. SOPAC Technical Report 75. Suva, Fiji. SOPAC.

Webb, A., and Kench, P. 2010. The dynamic response of reef islands to sea-level rise: Evidence from multi-decadal analysis. Global and Planetary Change : 234-246.

Wei.Q., Y. Guo, H. Cao and B. Kuai. 2010. Cloning and characterization of an AtNHX2-like Na+/H+ antiporter gene from Ammopiptanthus mongolicus (leguminosae) and its ectopic experssion enhanced drought and salt tolerance in Arabidopsis thaliana. Plant Cell Tissues Organ Culture . DOI10.1007/S11240-010-9869-3.

White, I., and T. Falkland. 2010. Management of fresh water lenses on Small Pacific Islands. Hydrology Journal , 18: 227-246.

White, I., D. Scott, and T. Falkland. 1999. Droughts in small coral islands;case study, South Tarawa, Kiribati. International Hydrological Programme , Technical documents in hydrology , 26.

Wiens, H. 1962. Atoll environment and ecology. New Haven: Yale university press.. Retrieved 09 2010, from http://en.wikipedia.org/wiki/ File: Malesia.png

138

Woodraffe, C. 1989. Salt water intrustion into ground water:an assessment of effects on small island states due to rising sea level. Australia: University of Wollongong. Retrieved from www.islandvulnerability.org/slr1989/woodroffe2.pdf

Woodroffe, C. 2008. Reef-island topograhpy and the vulnerability of atolls to sea level rise. Global Planet Change , 62: 7-96.

Wu. L., Z. Zhang, H. Zhang, X. Wang and R. Huang. 2008. Transcriptional modulation of ethylene response factor protein JERF3 in the oxidative stress response enhancese tolorance of tobacco seedlings to salt, drought and freezing. Plant Physiology , 148: 1953-1963.

Yamaguchi, T., and E. Blumwald. 2005. Developing salt tolerant crop plants: Challenges and opportunities. Retrieved March 1, 2010, from www.sciencedirect.com.

Yeo, A. 1998. Moleculer Biology of salt tolerance in the context of whole plant physiology. Journal of Experimental Botany , 49 (323): 915-929.

Yesayan, A. H. 2008. Determination of Salt tolerance in Wild Einkorn wheat (Triticum boeoticum Boiss) under in vitro conditions. Crop Wild Relative. , 6: 10.

Zahra. S. (2010). The salicylic acid effect on the tomatoe Lycopersicum esculentum Mi II.) germination, growth and photosynthesis. Journal of Stress Physiology and Biochemistry , 16 (3): 4-16.

Zhao, M., Q. Tian, and W. Zhang. 2007. Nitric oxide synthesis dependent Nitric oxide production is associated with salt tolerance in Arabidopsis. Plant Physology , 144: 206-217.

Zhu, J. 2001. Plant salt tolerance. TRENDS in Plant Science , 6 (2): 66-71.

139

ANNEX

TRADITIONAL KNOWLEDGE

While giant swamp taro is not famous in poetry, dances and songs it does have its own legend. Strangely enough quite often farmers have found an octopus in their giant swamp taro pits and these octopuses are not just lying around, they are found hugging the taro plant. How these octopuses get to the pits is quite a mystery. Legend has it that the octopus and the giant swamp taro are brothers and sisters, according to the legend this link is formed on the resemblance of the suckers present on the octopus legs and the fruiting flower of the taro. This could mean that it is the sibling link that attracts the octopus to the taro pits.

Figure 2.22 (left) Octopus. Figure 2.23 (right) mature giant swamp taro flowers with berries.

Another relation of the giant swamp taro to the Octopus given by the Tuvaluan farmers where they call the compost soil around giant swamp taro ‘ulu feke’ meaning Octopus head, however the reason for this is unknown (Iese, 2005). Apart from this the majority of the traditional knowledge is mainly about its cultivation and harvest that can be found in the sections 3.3.

140

Ikaraoi Cultivar Group Response

In vitro Ikaraoi

% INITIAL (week 1) FINAL (week 8) Salt NaCl Salt ASW Salt NaCl Salt ASW Salt Solution Solution Solution Solution 0% Salt

0.5%

1.0%

141

1.5%

2.0%

142

Multiple Comparisons Height Tukey HSD

(I) (J) 95% Confidence Interval VAR00002 VAR00002 Mean Difference (I-J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.45000 .28059 .499 -1.2303 .3303 1% .27000 .28059 .871 -.5103 1.0503

dimension3 1.5% -.09000 .28059 .998 -.8703 .6903 2% -.28500 .28059 .848 -1.0653 .4953 0.5% 0% .45000 .28059 .499 -.3303 1.2303 1% .72000 .28059 .085 -.0603 1.5003

dimension3 1.5% .36000 .28059 .702 -.4203 1.1403 2% .16500 .28059 .977 -.6153 .9453 1% 0% -.27000 .28059 .871 -1.0503 .5103 0.5% -.72000 .28059 .085 -1.5003 .0603 dimension2 dimension3 1.5% -.36000 .28059 .702 -1.1403 .4203 2% -.55500 .28059 .285 -1.3353 .2253 1.5% 0% .09000 .28059 .998 -.6903 .8703 0.5% -.36000 .28059 .702 -1.1403 .4203

dimension3 1% .36000 .28059 .702 -.4203 1.1403 2% -.19500 .28059 .957 -.9753 .5853 2% 0% .28500 .28059 .848 -.4953 1.0653 0.5% -.16500 .28059 .977 -.9453 .6153

dimension3 1% .55500 .28059 .285 -.2253 1.3353 1.5% .19500 .28059 .957 -.5853 .9753

143

Multiple Comparisons leaves Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0SALT 0.5SALT .40000 .25854 .535 -.3190 1.1190 1.0SALT .20000 .25854 .938 -.5190 .9190 1.5SALT .15000 .25854 .978 -.5690 .8690 2SALT -.05000 .25854 1.000 -.7690 .6690 0.5SALT 0SALT -.40000 .25854 .535 -1.1190 .3190 1.0SALT -.20000 .25854 .938 -.9190 .5190 1.5SALT -.25000 .25854 .869 -.9690 .4690 2SALT -.45000 .25854 .414 -1.1690 .2690 1.0SALT 0SALT -.20000 .25854 .938 -.9190 .5190 0.5SALT .20000 .25854 .938 -.5190 .9190 1.5SALT -.05000 .25854 1.000 -.7690 .6690 2SALT -.25000 .25854 .869 -.9690 .4690 1.5SALT 0SALT -.15000 .25854 .978 -.8690 .5690 0.5SALT .25000 .25854 .869 -.4690 .9690 1.0SALT .05000 .25854 1.000 -.6690 .7690 2SALT -.20000 .25854 .938 -.9190 .5190 2SALT 0SALT .05000 .25854 1.000 -.6690 .7690 0.5SALT .45000 .25854 .414 -.2690 1.1690 1.0SALT .25000 .25854 .869 -.4690 .9690 1.5SALT .20000 .25854 .938 -.5190 .9190

144

Multiple Comparisons suckers Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.650 .468 .636 -1.95 .65 1% -.500 .468 .822 -1.80 .80 1.5% -1.000 .468 .213 -2.30 .30 2% -.950 .468 .260 -2.25 .35 0.5% 0% .650 .468 .636 -.65 1.95 1% .150 .468 .998 -1.15 1.45 1.5% -.350 .468 .945 -1.65 .95 2% -.300 .468 .968 -1.60 1.00 1% 0% .500 .468 .822 -.80 1.80 0.5% -.150 .468 .998 -1.45 1.15 1.5% -.500 .468 .822 -1.80 .80 2% -.450 .468 .872 -1.75 .85 1.5% 0% 1.000 .468 .213 -.30 2.30 0.5% .350 .468 .945 -.95 1.65 1% .500 .468 .822 -.80 1.80 2% .050 .468 1.000 -1.25 1.35 2% 0% .950 .468 .260 -.35 2.25 0.5% .300 .468 .968 -1.00 1.60 1% .450 .468 .872 -.85 1.75 1.5% -.050 .468 1.000 -1.35 1.25

145

Multiple Comparisons corm Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.1894 .1425 .674 -.586 .208 1% -.3424 .1425 .124 -.739 .055 1.5% -.6156* .1403 .000 -1.007 -.225 2% -.4653* .1384 .010 -.851 -.080 0.5% 0% .1894 .1425 .674 -.208 .586 1% -.1529 .1481 .840 -.566 .260 1.5% -.4261* .1461 .035 -.833 -.019 2% -.2759 .1442 .318 -.678 .126 1% 0% .3424 .1425 .124 -.055 .739 0.5% .1529 .1481 .840 -.260 .566 1.5% -.2732 .1461 .341 -.680 .134 2% -.1229 .1442 .913 -.525 .279 1.5% 0% .6156* .1403 .000 .225 1.007 0.5% .4261* .1461 .035 .019 .833 1% .2732 .1461 .341 -.134 .680 2% .1503 .1421 .827 -.246 .546 2% 0% .4653* .1384 .010 .080 .851 0.5% .2759 .1442 .318 -.126 .678 1% .1229 .1442 .913 -.279 .525 1.5% -.1503 .1421 .827 -.546 .246 *. The mean difference is significant at the 0.05 level.

146

Multiple Comparisons root Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% .9494 .4151 .159 -.207 2.106 1% .5729 .4151 .642 -.584 1.730 1.5% .6922 .4089 .444 -.447 1.831 2% -.0221 .4032 1.000 -1.145 1.101 0.5% 0% -.9494 .4151 .159 -2.106 .207 1% -.3765 .4316 .906 -1.579 .826 1.5% -.2572 .4256 .974 -1.443 .929 2% -.9715 .4201 .151 -2.142 .199 1% 0% -.5729 .4151 .642 -1.730 .584 0.5% .3765 .4316 .906 -.826 1.579 1.5% .1193 .4256 .999 -1.067 1.305 2% -.5950 .4201 .619 -1.766 .576 1.5% 0% -.6922 .4089 .444 -1.831 .447 0.5% .2572 .4256 .974 -.929 1.443 1% -.1193 .4256 .999 -1.305 1.067 2% -.7143 .4139 .424 -1.868 .439 2% 0% .0221 .4032 1.000 -1.101 1.145 0.5% .9715 .4201 .151 -.199 2.142 1% .5950 .4201 .619 -.576 1.766 1.5% .7143 .4139 .424 -.439 1.868

147

Multiple Comparisons Dying Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% .9000 .3336 .062 -.028 1.828 1% .6000 .3336 .381 -.328 1.528 1.5% .2500 .3336 .944 -.678 1.178 2% -.1000 .3336 .998 -1.028 .828 0.5% 0% -.9000 .3336 .062 -1.828 .028 1% -.3000 .3336 .897 -1.228 .628 1.5% -.6500 .3336 .300 -1.578 .278 2% -1.0000* .3336 .028 -1.928 -.072 1% 0% -.6000 .3336 .381 -1.528 .328 0.5% .3000 .3336 .897 -.628 1.228 1.5% -.3500 .3336 .832 -1.278 .578 2% -.7000 .3336 .229 -1.628 .228 1.5% 0% -.2500 .3336 .944 -1.178 .678 0.5% .6500 .3336 .300 -.278 1.578 1% .3500 .3336 .832 -.578 1.278 2% -.3500 .3336 .832 -1.278 .578 2% 0% .1000 .3336 .998 -.828 1.028 0.5% 1.0000* .3336 .028 .072 1.928 1% .7000 .3336 .229 -.228 1.628 1.5% .3500 .3336 .832 -.578 1.278 *. The mean difference is significant at the 0.05 level.

148

Multiple Comparisons weight Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.2290 .2155 .825 -.828 .370 1% .0575 .2155 .999 -.542 .657 1.5% -.1525 .2155 .954 -.752 .447 2% .0385 .2155 1.000 -.561 .638 0.5% 0% .2290 .2155 .825 -.370 .828 1% .2865 .2155 .674 -.313 .886 1.5% .0765 .2155 .997 -.523 .676 2% .2675 .2155 .727 -.332 .867 1% 0% -.0575 .2155 .999 -.657 .542 0.5% -.2865 .2155 .674 -.886 .313 1.5% -.2100 .2155 .866 -.809 .389 2% -.0190 .2155 1.000 -.618 .580 1.5% 0% .1525 .2155 .954 -.447 .752 0.5% -.0765 .2155 .997 -.676 .523 1% .2100 .2155 .866 -.389 .809 2% .1910 .2155 .901 -.408 .790 2% 0% -.0385 .2155 1.000 -.638 .561 0.5% -.2675 .2155 .727 -.867 .332 1% .0190 .2155 1.000 -.580 .618 1.5% -.1910 .2155 .901 -.790 .408

149

Multiple Comparisons Chlorophyll Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.0851 1.6748 1.000 -4.743 4.572 1% .3699 1.6748 .999 -4.288 5.027 1.5% .2893 1.6748 1.000 -4.368 4.947 2% .7089 1.6748 .993 -3.949 5.366 0.5% 0% .0851 1.6748 1.000 -4.572 4.743 1% .4550 1.6748 .999 -4.203 5.112 1.5% .3744 1.6748 .999 -4.283 5.032 2% .7940 1.6748 .990 -3.863 5.452 1% 0% -.3699 1.6748 .999 -5.027 4.288 0.5% -.4550 1.6748 .999 -5.112 4.203 1.5% -.0806 1.6748 1.000 -4.738 4.577 2% .3390 1.6748 1.000 -4.318 4.997 1.5% 0% -.2893 1.6748 1.000 -4.947 4.368 0.5% -.3744 1.6748 .999 -5.032 4.283 1% .0806 1.6748 1.000 -4.577 4.738 2% .4196 1.6748 .999 -4.238 5.077 2% 0% -.7089 1.6748 .993 -5.366 3.949 0.5% -.7940 1.6748 .990 -5.452 3.863 1% -.3390 1.6748 1.000 -4.997 4.318 1.5% -.4196 1.6748 .999 -5.077 4.238

150

Katutu Cultivar Group Response

In vitro Katutu

% INITIAL (week 1) FINAL (week 8) Salt NaCl Salt ASW Salt NaCl Salt ASW Salt Solution Solution Solution Solution 0% Salt

0.5%

1.0 %

151

1.5 %

2.0 %

152

Multiple Comparisons height Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -1.11000* .35088 .017 -2.0858 -.1342 1% -.58500 .35088 .459 -1.5608 .3908 1.5% -.15500 .35088 .992 -1.1308 .8208 2% -.95000 .35088 .060 -1.9258 .0258 0.5% 0% 1.11000* .35088 .017 .1342 2.0858 1% .52500 .35088 .567 -.4508 1.5008 1.5% .95500 .35088 .058 -.0208 1.9308 2% .16000 .35088 .991 -.8158 1.1358 1% 0% .58500 .35088 .459 -.3908 1.5608 0.5% -.52500 .35088 .567 -1.5008 .4508 1.5% .43000 .35088 .737 -.5458 1.4058 2% -.36500 .35088 .836 -1.3408 .6108 1.5% 0% .15500 .35088 .992 -.8208 1.1308 0.5% -.95500 .35088 .058 -1.9308 .0208 1% -.43000 .35088 .737 -1.4058 .5458 2% -.79500 .35088 .165 -1.7708 .1808 2% 0% .95000 .35088 .060 -.0258 1.9258 0.5% -.16000 .35088 .991 -1.1358 .8158 1% .36500 .35088 .836 -.6108 1.3408 1.5% .79500 .35088 .165 -.1808 1.7708 *. The mean difference is significant at the 0.05 level.

153

Multiple Comparisons leaves Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.15000 .29182 .986 -.9615 .6615 1% .25000 .29182 .912 -.5615 1.0615 1.5% .20000 .29182 .959 -.6115 1.0115 2% -.50000 .29182 .431 -1.3115 .3115 0.5% 0% .15000 .29182 .986 -.6615 .9615 1% .40000 .29182 .648 -.4115 1.2115 1.5% .35000 .29182 .752 -.4615 1.1615 2% -.35000 .29182 .752 -1.1615 .4615 1% 0% -.25000 .29182 .912 -1.0615 .5615 0.5% -.40000 .29182 .648 -1.2115 .4115 1.5% -.05000 .29182 1.000 -.8615 .7615 2% -.75000 .29182 .084 -1.5615 .0615 1.5% 0% -.20000 .29182 .959 -1.0115 .6115 0.5% -.35000 .29182 .752 -1.1615 .4615 1% .05000 .29182 1.000 -.7615 .8615 2% -.70000 .29182 .125 -1.5115 .1115 2% 0% .50000 .29182 .431 -.3115 1.3115 0.5% .35000 .29182 .752 -.4615 1.1615 1% .75000 .29182 .084 -.0615 1.5615 1.5% .70000 .29182 .125 -.1115 1.5115

154

Multiple Comparisons suckers Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.650 .519 .721 -2.09 .79 1% -1.900* .519 .004 -3.34 -.46 1.5% -1.200 .519 .151 -2.64 .24 2% -1.200 .519 .151 -2.64 .24 0.5% 0% .650 .519 .721 -.79 2.09 1% -1.250 .519 .123 -2.69 .19 1.5% -.550 .519 .827 -1.99 .89 2% -.550 .519 .827 -1.99 .89 1% 0% 1.900* .519 .004 .46 3.34 0.5% 1.250 .519 .123 -.19 2.69 1.5% .700 .519 .662 -.74 2.14 2% .700 .519 .662 -.74 2.14 1.5% 0% 1.200 .519 .151 -.24 2.64 0.5% .550 .519 .827 -.89 1.99 1% -.700 .519 .662 -2.14 .74 2% .000 .519 1.000 -1.44 1.44 2% 0% 1.200 .519 .151 -.24 2.64 0.5% .550 .519 .827 -.89 1.99 1% -.700 .519 .662 -2.14 .74 1.5% .000 .519 1.000 -1.44 1.44 *. The mean difference is significant at the 0.05 level.

155

Multiple Comparisons corm Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.0537 .1597 .997 -.498 .391 1% -.4578* .1620 .045 -.909 -.007 1.5% -.3011 .1597 .333 -.746 .144 2% -.2850 .1577 .376 -.724 .154 0.5% 0% .0537 .1597 .997 -.391 .498 1% -.4041 .1640 .108 -.861 .052 1.5% -.2474 .1618 .546 -.698 .203 2% -.2313 .1597 .598 -.676 .213 1% 0% .4578* .1620 .045 .007 .909 0.5% .4041 .1640 .108 -.052 .861 1.5% .1567 .1640 .874 -.300 .613 2% .1728 .1620 .823 -.278 .624 1.5% 0% .3011 .1597 .333 -.144 .746 0.5% .2474 .1618 .546 -.203 .698 1% -.1567 .1640 .874 -.613 .300 2% .0161 .1597 1.000 -.429 .461 2% 0% .2850 .1577 .376 -.154 .724 0.5% .2313 .1597 .598 -.213 .676 1% -.1728 .1620 .823 -.624 .278 1.5% -.0161 .1597 1.000 -.461 .429 *. The mean difference is significant at the 0.05 level.

156

Multiple Comparisons root Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% .7863 .4413 .390 -.442 2.014 1% .4489 .4475 .853 -.797 1.694 1.5% .3021 .4413 .959 -.926 1.530 2% .3750 .4356 .910 -.837 1.587 0.5% 0% -.7863 .4413 .390 -2.014 .442 1% -.3374 .4531 .945 -1.598 .924 1.5% -.4842 .4469 .815 -1.728 .760 2% -.4113 .4413 .884 -1.639 .817 1% 0% -.4489 .4475 .853 -1.694 .797 0.5% .3374 .4531 .945 -.924 1.598 1.5% -.1468 .4531 .998 -1.408 1.114 2% -.0739 .4475 1.000 -1.319 1.172 1.5% 0% -.3021 .4413 .959 -1.530 .926 0.5% .4842 .4469 .815 -.760 1.728 1% .1468 .4531 .998 -1.114 1.408 2% .0729 .4413 1.000 -1.155 1.301 2% 0% -.3750 .4356 .910 -1.587 .837 0.5% .4113 .4413 .884 -.817 1.639 1% .0739 .4475 1.000 -1.172 1.319 1.5% -.0729 .4413 1.000 -1.301 1.155

157

Multiple Comparisons dying Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% .1500 .3024 .988 -.691 .991 1% .0000 .3024 1.000 -.841 .841 1.5% -.1500 .3024 .988 -.991 .691 2% -.4500 .3024 .573 -1.291 .391 0.5% 0% -.1500 .3024 .988 -.991 .691 1% -.1500 .3024 .988 -.991 .691 1.5% -.3000 .3024 .858 -1.141 .541 2% -.6000 .3024 .282 -1.441 .241 1% 0% .0000 .3024 1.000 -.841 .841 0.5% .1500 .3024 .988 -.691 .991 1.5% -.1500 .3024 .988 -.991 .691 2% -.4500 .3024 .573 -1.291 .391 1.5% 0% .1500 .3024 .988 -.691 .991 0.5% .3000 .3024 .858 -.541 1.141 1% .1500 .3024 .988 -.691 .991 2% -.3000 .3024 .858 -1.141 .541 2% 0% .4500 .3024 .573 -.391 1.291 0.5% .6000 .3024 .282 -.241 1.441 1% .4500 .3024 .573 -.391 1.291 1.5% .3000 .3024 .858 -.541 1.141

158

Multiple Comparisons weight Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.2580 .2065 .723 -.832 .316 1% -.2225 .2065 .818 -.797 .352 1.5% .3185 .2065 .538 -.256 .893 2% -.0890 .2065 .993 -.663 .485 0.5% 0% .2580 .2065 .723 -.316 .832 1% .0355 .2065 1.000 -.539 .610 1.5% .5765* .2065 .049 .002 1.151 2% .1690 .2065 .924 -.405 .743 1% 0% .2225 .2065 .818 -.352 .797 0.5% -.0355 .2065 1.000 -.610 .539 1.5% .5410 .2065 .075 -.033 1.115 2% .1335 .2065 .967 -.441 .708 1.5% 0% -.3185 .2065 .538 -.893 .256 0.5% -.5765* .2065 .049 -1.151 -.002 1% -.5410 .2065 .075 -1.115 .033 2% -.4075 .2065 .287 -.982 .167 2% 0% .0890 .2065 .993 -.485 .663 0.5% -.1690 .2065 .924 -.743 .405 1% -.1335 .2065 .967 -.708 .441 1.5% .4075 .2065 .287 -.167 .982 *. The mean difference is significant at the 0.05 level.

159

Multiple Comparisons chloro Tukey HSD (I) salt (J) salt Mean 95% Confidence Interval Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% .9656 1.4986 .967 -3.202 5.133 1% .5020 1.4986 .997 -3.665 4.669 2% .3215 1.4986 1.000 -3.846 4.489 2% .4041 1.4986 .999 -3.763 4.571 0.5% 0% -.9656 1.4986 .967 -5.133 3.202 1% -.4637 1.4986 .998 -4.631 3.704 2% -.6442 1.4986 .993 -4.812 3.523 2% -.5615 1.4986 .996 -4.729 3.606 1% 0% -.5020 1.4986 .997 -4.669 3.665 0.5% .4637 1.4986 .998 -3.704 4.631 2% -.1805 1.4986 1.000 -4.348 3.987 2% -.0979 1.4986 1.000 -4.265 4.069 2% 0% -.3215 1.4986 1.000 -4.489 3.846 0.5% .6442 1.4986 .993 -3.523 4.812 1% .1805 1.4986 1.000 -3.987 4.348 2% .0826 1.4986 1.000 -4.085 4.250 2% 0% -.4041 1.4986 .999 -4.571 3.763 0.5% .5615 1.4986 .996 -3.606 4.729 1% .0979 1.4986 1.000 -4.069 4.265 2% -.0826 1.4986 1.000 -4.250 4.085

160

Total Plant Response

In vivo plants after salt application

% salt Groups

Ikaraoi Katutu 0% control

0.5%

1.0%

161

1.5%

Multiple Comparisons HEIGHT Tukey HSD (I) (J) Mean 95% Confidence Interval SALT SALT Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.78000* .23310 .009 -1.4218 -.1382 1% -.15750 .23310 .961 -.7993 .4843 1.5% -.12250 .23310 .985 -.7643 .5193 2% -.61750 .23310 .066 -1.2593 .0243 0.5% 0% .78000* .23310 .009 .1382 1.4218 1% .62250 .23310 .062 -.0193 1.2643 1.5% .65750* .23310 .042 .0157 1.2993 2% .16250 .23310 .957 -.4793 .8043 1% 0% .15750 .23310 .961 -.4843 .7993 0.5% -.62250 .23310 .062 -1.2643 .0193 1.5% .03500 .23310 1.000 -.6068 .6768 2% -.46000 .23310 .283 -1.1018 .1818 1.5% 0% .12250 .23310 .985 -.5193 .7643 0.5% -.65750* .23310 .042 -1.2993 -.0157 1% -.03500 .23310 1.000 -.6768 .6068 2% -.49500 .23310 .214 -1.1368 .1468 2% 0% .61750 .23310 .066 -.0243 1.2593 0.5% -.16250 .23310 .957 -.8043 .4793 1% .46000 .23310 .283 -.1818 1.1018 1.5% .49500 .23310 .214 -.1468 1.1368 *. The mean difference is significant at the 0.05 level.

162

Multiple Comparisons LEAVES Tukey HSD (I) SALT (J) SALT dimension4 95% Confidence Interval Mean Std. Lower Upper Difference (I-J) Error Sig. Bound Bound dimension2 0% dimension3 0.5% .12500 .19942 .971 -.4241 .6741 1% .22500 .19942 .791 -.3241 .7741 1.5% .17500 .19942 .905 -.3741 .7241 2% -.27500 .19942 .642 -.8241 .2741 0.5% dimension3 0% -.12500 .19942 .971 -.6741 .4241 1% .10000 .19942 .987 -.4491 .6491 1.5% .05000 .19942 .999 -.4991 .5991 2% -.40000 .19942 .267 -.9491 .1491 1% dimension3 0% -.22500 .19942 .791 -.7741 .3241 0.5% -.10000 .19942 .987 -.6491 .4491 1.5% -.05000 .19942 .999 -.5991 .4991 2% -.50000 .19942 .093 -1.0491 .0491 1.5% dimension3 0% -.17500 .19942 .905 -.7241 .3741 0.5% -.05000 .19942 .999 -.5991 .4991 1% .05000 .19942 .999 -.4991 .5991 2% -.45000 .19942 .164 -.9991 .0991 2% dimension3 0% .27500 .19942 .642 -.2741 .8241 0.5% .40000 .19942 .267 -.1491 .9491 1% .50000 .19942 .093 -.0491 1.0491 1.5% .45000 .19942 .164 -.0991 .9991

163

Multiple Comparisons SUCKERS Tukey HSD (I) (J) Mean 95% Confidence Interval SALT SALT Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.650 .356 .362 -1.63 .33 1% -1.200* .356 .008 -2.18 -.22 1.5% -1.100* .356 .019 -2.08 -.12 2% -1.075* .356 .024 -2.06 -.09 0.5% 0% .650 .356 .362 -.33 1.63 1% -.550 .356 .535 -1.53 .43 1.5% -.450 .356 .714 -1.43 .53 2% -.425 .356 .755 -1.41 .56 1% 0% 1.200* .356 .008 .22 2.18 0.5% .550 .356 .535 -.43 1.53 1.5% .100 .356 .999 -.88 1.08 2% .125 .356 .997 -.86 1.11 1.5% 0% 1.100* .356 .019 .12 2.08 0.5% .450 .356 .714 -.53 1.43 1% -.100 .356 .999 -1.08 .88 2% .025 .356 1.000 -.96 1.01 2% 0% 1.075* .356 .024 .09 2.06 0.5% .425 .356 .755 -.56 1.41 1% -.125 .356 .997 -1.11 .86 1.5% -.025 .356 1.000 -1.01 .96 *. The mean difference is significant at the 0.05 level.

164

Multiple Comparisons CORM Tukey HSD (I) (J) Mean 95% Confidence Interval SALT SALT Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% -.1200 .1073 .797 -.416 .176 1% -.4029* .1081 .002 -.701 -.105 1.5% -.4551* .1065 .000 -.749 -.162 2% -.3738* .1051 .004 -.663 -.084 0.5% 0% .1200 .1073 .797 -.176 .416 1% -.2829 .1109 .084 -.588 .023 1.5% -.3351* .1093 .021 -.636 -.034 2% -.2538 .1079 .134 -.551 .044 1% 0% .4029* .1081 .002 .105 .701 0.5% .2829 .1109 .084 -.023 .588 1.5% -.0523 .1101 .990 -.356 .251 2% .0290 .1087 .999 -.271 .329 1.5% 0% .4551* .1065 .000 .162 .749 0.5% .3351* .1093 .021 .034 .636 1% .0523 .1101 .990 -.251 .356 2% .0813 .1072 .942 -.214 .377 2% 0% .3738* .1051 .004 .084 .663 0.5% .2538 .1079 .134 -.044 .551 1% -.0290 .1087 .999 -.329 .271 1.5% -.0813 .1072 .942 -.377 .214 *. The mean difference is significant at the 0.05 level.

165

Multiple Comparisons ROOT Tukey HSD (I) (J) Mean 95% Confidence Interval SALT SALT Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% .8844* .3120 .040 .025 1.744 1% .5200 .3143 .465 -.346 1.386 1.5% .5022 .3098 .486 -.351 1.356 2% .1913 .3056 .971 -.651 1.033 0.5% 0% -.8844* .3120 .040 -1.744 -.025 1% -.3644 .3224 .790 -1.253 .524 1.5% -.3823 .3179 .750 -1.258 .494 2% -.6932 .3139 .181 -1.558 .172 1% 0% -.5200 .3143 .465 -1.386 .346 0.5% .3644 .3224 .790 -.524 1.253 1.5% -.0178 .3202 1.000 -.900 .864 2% -.3287 .3162 .837 -1.200 .543 1.5% 0% -.5022 .3098 .486 -1.356 .351 0.5% .3823 .3179 .750 -.494 1.258 1% .0178 .3202 1.000 -.864 .900 2% -.3109 .3117 .856 -1.170 .548 2% 0% -.1913 .3056 .971 -1.033 .651 0.5% .6932 .3139 .181 -.172 1.558 1% .3287 .3162 .837 -.543 1.200 1.5% .3109 .3117 .856 -.548 1.170 *. The mean difference is significant at the 0.05 level.

166

Multiple Comparisons DYING Tukey HSD (I) SALT (J) SALT dimension4 95% Confidence Mean Interval Difference Std. Lower Upper (I-J) Error Sig. Bound Bound dimension2 0% dimension3 0.5% .5250 .2255 .140 -.096 1.146 1% .3000 .2255 .672 -.321 .921 1.5% .0500 .2255 .999 -.571 .671 2% -.2750 .2255 .740 -.896 .346 0.5% dimension3 0% -.5250 .2255 .140 -1.146 .096 1% -.2250 .2255 .856 -.846 .396 1.5% -.4750 .2255 .221 -1.096 .146 2% -.8000* .2255 .004 -1.421 -.179 1% dimension3 0% -.3000 .2255 .672 -.921 .321 0.5% .2250 .2255 .856 -.396 .846 1.5% -.2500 .2255 .802 -.871 .371 2% -.5750 .2255 .084 -1.196 .046 1.5% dimension3 0% -.0500 .2255 .999 -.671 .571 0.5% .4750 .2255 .221 -.146 1.096 1% .2500 .2255 .802 -.371 .871 2% -.3250 .2255 .602 -.946 .296 2% dimension3 0% .2750 .2255 .740 -.346 .896 0.5% .8000* .2255 .004 .179 1.421 1% .5750 .2255 .084 -.046 1.196 1.5% .3250 .2255 .602 -.296 .946 *. The mean difference is significant at the 0.05 level.

167

Multiple Comparisons WEIGHT Tukey HSD (I) SALT (J) SALT dimension4 95% Confidence Mean Interval Difference Std. Lower Upper (I-J) Error Sig. Bound Bound dimension2 0% dimension3 0.5% -.2435 .1520 .498 -.662 .175 1% -.0825 .1520 .983 -.501 .336 1.5% .0830 .1520 .982 -.335 .501 2% -.0252 .1520 1.000 -.444 .393 0.5% dimension3 0% .2435 .1520 .498 -.175 .662 1% .1610 .1520 .827 -.257 .579 1.5% .3265 .1520 .204 -.092 .745 2% .2183 .1520 .605 -.200 .637 1% dimension3 0% .0825 .1520 .983 -.336 .501 0.5% -.1610 .1520 .827 -.579 .257 1.5% .1655 .1520 .812 -.253 .584 2% .0573 .1520 .996 -.361 .476 1.5% dimension3 0% -.0830 .1520 .982 -.501 .335 0.5% -.3265 .1520 .204 -.745 .092 1% -.1655 .1520 .812 -.584 .253 2% -.1082 .1520 .953 -.527 .310 2% dimension3 0% .0252 .1520 1.000 -.393 .444 0.5% -.2183 .1520 .605 -.637 .200 1% -.0573 .1520 .996 -.476 .361 1.5% .1082 .1520 .953 -.310 .527

168

Multiple Comparisons CHLORO Tukey HSD (I) (J) Mean 95% Confidence Interval SALT SALT Difference (I- J) Std. Error Sig. Lower Bound Upper Bound 0% 0.5% .4402 1.1106 .995 -2.618 3.498 1% .4359 1.1106 .995 -2.622 3.494 1.5% .3054 1.1106 .999 -2.753 3.363 2% .5565 1.1106 .987 -2.502 3.615 0.5% 0% -.4402 1.1106 .995 -3.498 2.618 1% -.0043 1.1106 1.000 -3.062 3.054 1.5% -.1349 1.1106 1.000 -3.193 2.923 2% .1163 1.1106 1.000 -2.942 3.174 1% 0% -.4359 1.1106 .995 -3.494 2.622 0.5% .0043 1.1106 1.000 -3.054 3.062 1.5% -.1305 1.1106 1.000 -3.189 2.928 2% .1206 1.1106 1.000 -2.937 3.179 1.5% 0% -.3054 1.1106 .999 -3.363 2.753 0.5% .1349 1.1106 1.000 -2.923 3.193 1% .1305 1.1106 1.000 -2.928 3.189 2% .2511 1.1106 .999 -2.807 3.309 2% 0% -.5565 1.1106 .987 -3.615 2.502 0.5% -.1163 1.1106 1.000 -3.174 2.942 1% -.1206 1.1106 1.000 -3.179 2.937 1.5% -.2511 1.1106 .999 -3.309 2.807

169