(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Results of the first Joint US-USSR Central Pacific Expedition (BERPAC) : Autumn 1988"

RESULTS OF THE FIRST JOINT 

US— USSR CENTRAL PACIFIC 

EXPEDITION (BERPAC) 



W^r 






** 



Hi 



W-WM^m 



AUTUMN 1988 



UNITED STATES DEPARTMENT OF THE INTERIOR / Fish and Wildlife Service 








OtMCO 






RESULTS OF THE FIRST JOINT US-USSR 

CENTRAL PACIFIC EXPEDITION 
H (BERPAC) 




Oi 
1 = 

s>- 
SI 



a 
iP- 
i un 

! -D 
: eO 
i □ 

I a 

• i-^ 
i a 
: m 
= o 



AUTUMN 1988 



\&iQQQS biCLZ 



Results of the First Joint 
US-USSR Central Pacific Expedition 
(BERPAC) 

Autum 1988 





John F. Turner 
Director, US Fish and 

Wildlife Service, 
Washington, DC 



Yuriy A. Izrael 
Chairman. USSR State 

Committee for Hydrometeorology 
Moscow, USSR 



Harold J. O'Connor 

Project Leader, USA 

US Fish and Wildlife Service 

Patuxent Wildlife Research Center 

Laurel. Maryland 



Alia V. Tsyban 

Project Leader, USSR 

Institute of Global Climate and Ecology 

State Committee for Hydrometeorology 

Academy of Sciences 

Moscow, USSR 



Copies of this publication may be obtained from the Publications Unit. US Fish and Wildlife 
Service, 1 849 C Street, NW, Mail Stop 130— ARLSQ. Washington. DC 20240. 



Suggested Citation: 

Nagel, P. A. (ed. ) ( 1 992). Results of the First Joint US-USSR Central Pacific Expedition 

(BERPAC), Autumn 1988. US Fish and Wildlife Service. Washington. DC. 



Disclaimer: 

The opinions and recommendations expressed in this report are those of the authors and do not necessarily reflect the 
views of the US Fish and Wildlife Service, nor does the mention of trade names constitute endorsement or 
recommendation for use by the Federal Government. 



Foreword 



Included here in its entirety is the BERPAC Program paper, published in 1991 [J.F. Turner, H.J. O'Connor, Yu.A. Izrael, and 
A. V. Tsyban (eds. )( 1991 )]: BERPAC — A Program for Long-term Ecological Research of Ecosystems of the Bering and Chukchi 
Seas and the Pacific Ocean. National Fund for the Patuxent Wildlife Research Center, Bowie, Maryland]. BERPAC refers to 
joint research of the US and USSR in the Bering Sea and central Pacific Ocean since 1 977. The Bering Sea portion of the project 
has included three joint research expeditions (1977, 1984, and 1988) between the two countries. The central Pacific portion of 
the project was established in 1 988 when the first joint expedition took place. The central Pacific segment, therefore, has not had 
the opportunity to go through the same maturing process as that of the Bering Sea. As time goes on, and research continues, goals 
and objectives to be accomplished will be further developed. The paper is included here as a Foreword to show the correlation 
with this volume 's sister monograph. Results of the Third Joint US-USSR Bering & Chukchi Seas Expedition (BERPAC), Summer 
1988. 

BERPAC 

A Program for Long-term Ecological Research of Ecosystems of the Bering and Chukchi Seas and the Pacific Ocean 



Introduction 

Deterioration of ecosystems on a large scale threatens 
many functional equilibria in the biosphere. This problem is 
particularly urgent for the World Ocean, which is the sink for 
many different pollutants that can produce significant ecological 
impacts. 

The ocean is able to assimilate a certain amount of 
anthropogenic compounds due to "self-purification" without 
visible deterioration of the ecosystem. However, continuous 
increase in the flux of pollutants to the ocean creates the need 
for study of the resistance of marine ecosystems to anthropogenic 
impacts. Investigations of ecological consequences and 
elucidation of causal relationships between the impact levels 
and adverse biological effects are only poorly understood for 
the marine environment. The study of these interactions and 
responses is interdisciplinary in character and covers different 
fields of biology, ecology, chemistry, and physics of the sea. 

The dynamics of marine ecosystems, including biological 
and physical processes and biogeochemical cycles, are closely 
related to changes in the climate of the Earth. The predicted 
global warming may have a pronounced effect on certain vital 
processes in the World Ocean, especially the resistance of its 
ecosystems to anthropogenic contamination. This is because 
the living ocean determines, to a great degree, the normal 
functions of the Earth's climatic system. 

Long-term observations of physical, geochemical, and 
hydrobiological processes are necessary for the assessment of 
ecological consequences of contamination in the ocean 
environment and isolation of local anthropogenic effects 
compared to the effect of climatic variability. 

The Bering Sea is located between the coasts of the Soviet 
Far East (USSR) and Alaska (USA), and naturally an interest 
in the study of its ecosystems has been shown by Soviet and 
American scientists (Izrael & Tsyban, 1983a, 1977, 1990; 
Roscigno, 1990). 

In spite of comprehensive studies carried out in the Bering 
Sea in the last few years (Izrael etai, 1988b; Izrael & Tsyban, 
1989, 1990; Coachman, 1990; Roscigno, 1990), a number of 
the oceanographic, hydrochemical. and biological parameters 



determining its ecosystem functions are as yet poorly known 
when compared with, for instance, the Baltic, Mediterranean, 
and Black Seas. For example, the joint bilateral program of 
Bering/Chukchi investigations have been carried out for more 
than 13 years with the production of three monographs of 
cruise results. However, the as yet inadequate data on the 
characteristics and processes occurring in the ecosystems of 
the Bering Sea and North Pacific waters has led to the 
organization and implementation of an international program: 
Long-term Ecological Research of Ecosystems of the Bering 
and Chukchi Seas and the Pacific Ocean (BERPAC Program). 

Goals, Objectives, and Scientific Basis of the BERPAC 
Program 

Goals 

The goal of the BERPAC Program is to examine the status 
of marine ecosystems of the Pacific Ocean, Bering Sea, and 
Chukchi Sea and to assess their role in determining global 
climate. BERPAC will study the dynamics of these ecosystems 
related to conditions of global climate change and anthropogenic 
contamination. 

Objectives and Scientific Basis of the BERPAC Program 

Objectives of the BERPAC Program consist of the study 
of the biogeochemical cycles of contaminants, related 
oceanographic processes, and food-web interactions in the 
North Pacific waters that flow through the Bering/Chukchi 
Seas, including study of the behavior of organic pollutants at 
the water/sediment interface, since sediments are sources of 
the secondary pollution of ecosystems. Important topics of 
study are the control and the accumulation of pollutants in 
bottom deposits, and the study of their migration within the 
sediments and exchange with overlying waters are important 
topics of study. 

1. Assessment of Ecological Consequences of Contamination 

Progressively severe changes in chemical contamination 

of the ocean biosphere are on the increase. Anthropogenic 



in 



impacts influence not only the biotic component of the marine 
environment but different abiotic components as well. Such 
impacts lead to even more significant changes in the World 
Ocean and in the biosphere as a whole. 

Specific features of the Bering Sea and other ecosystems 
w ith "background" levels of contamination are such that they 
are especially vulnerable because of the continual input of 
small doses of pollution. This leads to a gradual accumulation 
of pollutants, and may ultimately cause the degradation of the 
ecosystems. Therefore, ecological investigations and 
monitoring of the background regions of the ocean, especially 
in such highly bioproductive zones as the Bering Sea, are of 
g reat importance. In order to assess the ecological consequences 
of the pollution and isolate anthropogenic effects from the 
background of natural variability, it is necessary to make long- 
term observations of fundamental physical, chemical and 
biological processes in selected areas of the above regions. 
These regions differ in their geographical location, as well as 
in the subsystems of their ecosystems, and are subjected to 
different anthropogenic impacts. 

2. Study of the Processes Determining the Assimilative 
Capacity for Contaminants in Marine Ecosystems 

In the marine environment, various physical, chemical, 
and biological processes occur through which contaminants 
can be eliminated from the ecosystem without serious 
disturbances of the biogeochemical cycles of the elements or 
changes in the biota. Diverse oceanological investigations 
carried out in the last few years have shown that the biotic 
component is important in the fluxes of pollutants. 

The ability of an ecosystem to protect itself against a 
foreign interference at the expense of many biological, physical, 
and chemical processes is its natural ■■immunity"' and the 
measure of this immunity is its assimilative capacity. 

According to the contemporary interpretation (Izrael & 
Tsyban. 1983b, 1989; Izrael et a/., 1988b.c), the assimilative 
capacity of a marine ecosystem is an integral function of its 
ex ist ing environmental status that reflects the ability of physical, 
chemical, and biological processes forelimination of pollutants 
and their impacts on the biota. 

When using the concept of assimilative capacity in practice, 
it is necessary to bear in mind that a marine ecosystem occupies 
a finite volume that may be isolated on the basis of the spatial 
distribution of organisms of various trophic levels, groups of 
ecologically similar species, and production/destruction 
processes, as well as physical and chemical characteristics. 
Hence, the assimilative capacity of each specific ecosystem 
also has a value that objectively characterizes existing properties 
of the marine em ironment. This value could be determined in 
practice on the basis of integrated investigations and monitoring 
of the marine environment, carried out in accordance with 
existing methodological recommendations (Izrael & Tsyban. 
1983b, 1985, 1987, 1989; Izrael etal., 1988b). 

The use of this concept in the BERPAC studies will 
include investigations of the following basic problems: 
/. quantitative assessment ol the balance of chemical elements 
in the ecosystem and possible changes in residence times due 
to disturbances: 2. assessment of adverse biological effects at 



the level of population and communities: and.v determination 
ol the critical concentrations at whichcontaminants adversely 
impact the marine organisms and communities. 

Thus, a conceptual model of the assimilative capacity, 
based on a better understanding of the laws of marine ecosy stem 
functions, can serve as a theoretical basis for the dev elopment 
of forecasts of both the immediate and long-range 
consequences of anthropogenic and climatic impacts on the 
ocean ecosystems. 

3. Study of the Elements of the Biogeochemical Carbon Cycle 
and Its Role in Global Climatic Processes 

Global warming predicted in connection with the 
developing greenhouse effect depends directly upon the 
biogeochemical cycle of carbon — the most important process 
forming the Earth's climate. The basic elements of this cycle 
are carbon-dioxide and other "greenhouse gases" exchanged 
within the ocean-atmosphere system, the function of the 
carbonate system, and the turnover of organic forms of carbon 
in the ocean. 

The most intensive uptake of atmospheric CO, occurs at 
high latitudes, as a result of favorable thermal and hydrological 
conditions (low sea surface temperature and permanent 
downwelling) in the region. These peculiarities explain the 
important role of the Bering Sea, a subarctic body of water 
having a large area, in the global cycle of carbon dioxide. 

The relationship between the rates and directions of CO, 
flow within the ocean-atmosphere system directly affects the 
functioning of the carbonate system. So. in the conditions 
where global warming is induced by an increase in the 
concentration of atmospheric CO : . a shift of the equilibrium 
between carbonate forms of carbon in seawater might occur, 
which will be accompanied by a decrease of pH and. 
consequently, elevation of the lysocline. 

Investigations of these processes, directly affecting the 
sedimentation of organic carbon and the vital functions of 
marine organisms, are only possible with direct determination 
of all components of the carbonate system (i.e.. HCO„ CO,, 
H : CO,. andCO,). 

To fully understand all of the characteristics of the oceanic 
portion of the global carbon cycle, it is necessary to stud) the 
processes of the circulation of its organic forms in the 
composition of dissolved and particulate matter and in the cells 
of living organisms (Zaitsev. 1970. 1980, 1985). 

The dynamic equilibrium of dissolved and particulate 
organic matter, living matter, and the content of organic carbon 
within water masses depends on the relations between 
production/destruction processes established in the ecosystem. 
In this connection, the predicted effects of global warming on 
the bioproductivity of the Bering Sea ecosystem will influence 
the organic carbon cycle. In order to studv possible changes, 
long-term observations of the concentrations of all organic 
forms of carbon are necessary. 

Thus, to establish the carbon balance in the Bering Sea 
ecosystem, comprehensive long-term observations of all carbon 
constituents in the aquatic interface and the study of quantitative 
and qualitative composition of both the carbonate system and 
organic forms of carbon are required. 



IV 



4. In\ estimation ot'the Physical Mechanisms Related toClimate 
Variations 

Existing global physical models ot'the ocean-atmosphere 
system do not make it possible to predict possible climate 
changes on a regional scale because of the extreme complexity 
of the modeled systems. Additional investigations of the 
physical development of regional models, in particular of a 
model for the Bering Sea. are an important need for long-term 
climate forecasting at the present time. 

This problem could be solved on the basis of long-term 
oceanological observations in different regions of the Bering 
Sea. which are aimed at the acquisition of systematic information 
on the vertical distribution of temperature, heat content of the 
active layer and its variability with time, the structure and 
variability of ocean circulation, heat transfer by the basic sea 
currents, and heat and moisture fluxes across the sea surface. 

To develop the above models it is necessary to know the 
regularity of water mass formation in the deep basins of the 
Bering Sea. The following issues are not yet clear: North 
Pacific water must be involved in bottom water formation, but 
given the topographic isolation of Bowers and the central 
basins, how and where docs this lake place? Are sources the 
same for the different basins' What are the flushing rates (e.g.. 
residence times)? 

There are three hypothetical mechanisms by which bottom 
water might possibly be formed: /. modification of surface 
(upper layer) water within the confines of the sea by cooling 
and brine enhancement through ice formation, creating water 
sufficiently dense to sink to the bottom: 2. subsurface mixings 
of North Pacific water with appropriate Bering Sea waters as it 
crosses the sills in the Aleutian-Komandorskiy Island arc 
passages; and 3. direct advection of deep North Pacific water 
in through Kamchatka Strait and then sequentially through the 
gaps into the other basins. 

The BERPAC Program will investigate the mechanism of 
deep water formation, renewal rates, and Hushing of the basins. 

Area of Investigations 

While selecting the study areas and location of stations in 
the Bering Sea. the diversity and contrast of ecological conditions 
in different regions of the sea were taken into account. 

In order to reflect a variety of ecological conditions in the 
Bering Sea more completely, it seems appropriate that integrated 
expeditions include work on polygons located in different 
areas of the sea (with the purpose of obtaining representative 
data on the structure and functions of the basic marine 
ecosystems) and work across transects (with the purpose of 
determining the space and time variations of the key ecological 
parameters). 

Investigations within the framework of BERPAC will be 
conducted on four polygons where investigations were carried 
out in 1 98 1 (during the integrated ecological expedition aboard 
the research vessel ( R/V ) Akademik Shirshov) and in 1 984 and 



1988 (during the second and third Soviet-American ecological 
expeditions aboard the R/V Akademik Korolev) (Izrael & 
Tsyban. 1987, 1990; Izrael etcd.. 1988a; Roscigno. 1990). 

Deep stations will be repeated at four centered polygons in 
the four deep basins. The center station of each polygon will 
also be a location for a mooring containing sediment traps and 
current meters, funding permitting. Four other mooring 
locations will cover the entrance from the North Pacific (in the 
deep channel northwest of Komandorskiy Island), the main 
gaps in the ridges north of Attu, and a location on the east side 
of the Central Basin under the Bering Slope Current. The 
mooring locations are also deep oceanographic stations, and 
1 1 additional stations will provide continuity among the deep 
waters. 

In addition to polygons, observations are planned at stations 
along the transects located in areas that are not yet completely 
understood, such as the Gulf of Anadyr, the Chirikov Basin, the 
Gulf of Alaska, the northern portion of the Pacific Ocean, and 
the deep-water central and southwestern areas of the sea. 
Larger scale studies in the Chukchi Sea and central Pacific 
ecosystems are also planned. The program for individual 
expeditions will be discussed specifically during joint symposia. 

Proposed Observations 

Complex observations during the ecological expeditions 
include meteorological (including aerological and geophysical 
studies), oceanographical. and ecological observations. 
Specifically, the following observations will be made: 

A. Meteorological observations will include routine 
observations of meteorological parameters, such as studies of 
direct solar radiation intensity and ultraviolet irradiation, cloud 
and cloud type studies, and collection of samples of atmospheric 
precipitation for chemical analyses. Aerological and 
geophysical observations will include temperature and wind 
sounding with the aid of radiosondes. Air samples will be 
collected lor determination of sulfates and nitrogen oxides. 
Visual observations of oil and oil product contamination on the 
sea surface will be recorded. 

B. Oceanographic observations at designated sampling 
depths in the water column will include temperature, salinity, 
nutrients, oxygen content, water color and transparency, 
biogenic elements, alkalinity, and petroleum hydrocarbons. 
Tracers for water mass types will include stable isotope content 
of seawater (oxygen, deuterium, tritium, freons, silica, and 
carbon 14). In addition, current velocity and direction will be 
determined, and sediment trap collections will be made. 

C. Ecological observations will include studies of the 
atmosphere, sea surface microlayer, water column, and bottom 
deposits in the environment. 

/. Atmosphere 

In rainfall. pH and the content of organic contaminants 
will be determined. In dust particles, the content of organic 
contaminants and metals will be determined. In the air at the 



sea surface, the content of "greenhouse" gases (C0 2 , nitrogen 
oxides), oxygen, and chlorinated hydrocarbons will be 
determined. 

2. Sea surface microlayer. water column, and bottom 
deposits 

Water samples will be collected in the surface microlayer 
and at standard hydrological depths and at selected experimental 
depths (e.g., thermocline, pycnoline, phyto- and zooplankton 
maxima and sediment-water interface) (Zaitsev, 1980). 

a. In the surface microlayer, the following elements and 
parameters will be determined: 

— organic carbon; 

- contaminants (toxic metals, and aliphatic aromatic 
and chlorinated hydrocarbons), the state of neustonic 
communities; determination of the structural 
characteristics of bacterioplankton; total numbers, 
biomass of microorganisms, most probable numbers 
(MPN) of indicator groups of bacteria (e.g., paraffin- 
oxidizers, PCB-transforming and neurotrophic 
saprophyte groups), and indices of phyto- and 
zooneuston (numbers, biomass, species, size 
composition, species mass and indicator forms), 
mutation (teratogenesis) of zooneuston organisms. 

b. In the water column, the following parameters will be 
determined: 

— water optical indices; 

— contaminants (toxic metals, and aromatic, aliphatic 
and chlorinated hydrocarbons); 

• the total concentrations of organic carbon and its 
composition; 

— elements of the carbonate system (CO,, HCO„ CO : ); 

— characteristics of bacterioplankton (total numbers, 
biomass, MPN, and distribution of indicator groups); 
and their biochemical and genetic capacities; 
structural characteristics of phyto-, microzoo-, and 
mesozooplankton ( numbers, biomass, size, and species 
composition, species mass, and indicator forms); 
functional characteristics of planktonic communities 
(heterotrophic CO, assimilation by bacteria, bacterial 
production, phytoplankton productivity); and 
biosedimentation rate of particulate matter. 

c. In the biota, the following parameters will be 
determined: 

contaminants (toxic metals, and aromatic, chlorinated 
and aliphatic hydrocarbons; and 
organic carbon content, stable carbon, and nitrogen 
isotope content. 

d. In bottom sediments, the following elements will be 
determined: 

determinants (toxic metals, and aromatic, chlorinated 
and aliphatic hydrocarbons). 
total organic carbon and nitrogen; 
stable carbon and nitrogen isotopes; and 



— structural characteristics of zoobenthos (numbers, 
biomass, species composition, and species mass). 

3. Higher trophic levels 

During the expedition, zoological observations will be 
carried out: numbers, distribution, and migrator)' patterns of 
fish, birds, and marine mammals. 

4. Model experiments 

Model experiments will be performed under conditions 
similar to natural situations. During these experiments, the 
following parameters will be studied: 

— photochemical oxidation of organic contaminants; 

— biodegradation potential of bacterioplankton with 
respect to organic contaminants (benzo(a)pyrene. 
PCB, etc.); 

— combined influence of contaminants on biological 
"targets" and establishment of "critical" concentrations 
of the impact on plankton communities in the 
conditions of controlled ecosystems (Izrael el a!.. 
1988a); and 

— sediment respiration and nutrient flux experiments. 

Connection with Other International Programs 

The BERPAC Program has much in common with other 
international programs, but at the same time it has its own 
particular features mentioned earlier. Wide cooperation with 
other similar international projects is built within the framework 
of this program — in particular, in the preparation of joint 
marine expeditions. Wide data exchange is also planned. 

Schedule of Activities and Applications of Results 

Since 1977, successful joint investigations of Soviet and 
American scientists have been carried out in the Bering Sea 
within the framework of the specific theme of the bilateral 
cooperation "Bering Sea" (Project "Comprehensive 
Environmental Analysis;" Subproject "Comprehensive 
Analysis of Marine Ecosystem State and Ecological Problems 
of the World Ocean"). Important stages of this cooperation 
were three joint ecological Soviet-American expeditions in the 
Bering Sea on the R/V Volna (Summer 1977). R/V Akademik 
Korolev ( Summer 1 984 and 1 988 ). and several symposia on the 
preparation of scientific programs and analyses of the results of 
these expeditions, as well as three monographs describing the 
results of long-term Soviet-American investigations in the 
Bering Sea (Izrael & Tsyban, 1990; Roscigno. 1990). H is 
expected that these expeditions will be every four years and 
followed by international symposia and joint publications. 

Monographs on the results of future expeditions will be 
published. It is expected that seminars and symposia within the 
framework of the BERPAC Program will be conducted. Also 
included in the plans are special intercalibrations, a wide 
exchange of specialists, and joint experimental work. 



VI 



References 



Coachman, L. K. (1990). Bering Sea ecosystem: basic 
characteristics and prospects for long-term research. In 
Research on the Bering Sea Ecosystem. Results of the 
Second Soviet-American Expedition. The 37th Cruise of 
the Research Vessel Akademik Korolev. June-September 
1984. Gidrometeoizdat Publishers, Leningrad. 2. (in Russian) 

Izrael, Yu. A. & Tsyban, A. V. (eds. ) (1983a). Research on the 
Bering Sea Ecosystem. Gidrometeoizdat Publishers. 
Leningrad. 157 pp. (in Russian) 

Izrael. Yu. A. & Tsyban. A. V. (1983b). On the assimilative 
capacity of the World Ocean. Reports of the USSR Academy 
of Sciences 272(3), 702-705. (in Russian) 

Izrael. Yu. A. & Tsyban. A. V. (1985). The ecology and 
problems of global ocean monitoring. In Comprehensive 
Global Ocean Monitoring 1. 19-48. Gidrometeoizdat 
Publishers, Leningrad, (in Russian) 

Izrael, Yu. A. & Tsyban, A. V. (eds.) ( 1987). Comprehensive 
Analysis of the Bering Sea Ecosystem. Gidrometeoizdat 
Publishers, Leningrad, 264 pp. (in Russian) 

Izrael, Yu. A. & Tsyban, A. V. (1989). Anthropogenic Ecology 
of the Ocean. Gidrometeoizdat. Leningrad. 528 pp. (in 
Russian) 

Izrael, Yu. A. & Tsyban, A. V. (eds.) (1990). Research on the 
Bering Sea Ecosystem. In Results of the Soviet-American 
Expedition. The 37th Cruise of the Research Vessel Akademik 
Korolev. June-September. 1984. Gidrometeoizdat 
Publishers. Leningrad, 344 pp. (in Russian) 



Izrael, Yu. A., Tsyban. A. V., Panov, G. V., Korsak, M. N. 
et al. ( 1988a). Comprehensive analysis of the Bering Sea 
ecosystem. In Comprehensive Analysis of the Environment. 
Proceedings from the Fifth USSR-US Symposium. 
Gidrometeoizdat Publishers. Leningrad, 528 pp. (in Russian) 

Izrael, Yu. A.. Tsyban, A. V., Ventzel, M. V. & Shigaev, V. V. 
( 1988b). Generalized model of the assimilative capacity of 
a marine ecosystem. Reports of the USSR Academy of 
Sciences 380(2). (in Russian) 

Izrael. Yu. A.. Tsyban, A. V., Ventzel, M. V. & Shigaev. V. V. 
(1988c). Scientific basis for ecological standardization of 
the anthropogenic impact on marine ecosystems (using the 
example of the Baltic Sea ecosystem. Oceanology 28(2). 
(in Russian) 

Roscigno, P. F. (ed. ) (1990). Results of the Second Joint US- 
USSR Bering Sea Expedition. Summer 1984. US Fish and 
Wildlife Service Biological Report 90( 13). 347 pp. 

Zaitsev. Yu. P. (1970). Marine Neustonology. Naukova 
Dumka Publishers. Kiev, 264 pp. (in Russian) 

Zaitsev, Yu. P. ( 1980). Zooneuston and methods for its study. 
In Methods for Biological Analysis of Sea Water and Bottom 
Sediments, pp. 134-139. Gidrometeoizdat Publishers, 
Leningrad,, (in Russian) 

Zaitsev, Yu. P. ( 1985). Biotic contours in ocean monitoring. In 
Comprehensive Global Ocean Monitoring. Proceedings 
from the First International Symposium 2, 76-83. 
Gidrometeoizdat Publishers. Leningrad, (in Russian I 



John F. Turner 
Director. US Fish and 

Wildlife Service 
Washington, DC 



Yuriy A. Izrael 
Chairman. USSR State 
Committee for Hydrometeorology 
Moscow, USSR 



Harold J. O'Connor 

Project Leader. USA 

US Fish and Wildlife Service 

Patuxent Wildlife Research Center 

Laurel, Maryland 



Alia V. Tsyban 

Project Leader, USSR 

Institute of Global Climate and Ecology & 

State Committee for Hydrometeorology 
Academy of Sciences 
Moscow, USSR 



Protocol of the First Joint US-USSR 

Central Pacific Ocean Expedition 

on the R/V Akademik Korolev 



In accordance with the memorandum of the 1 1th meeting 
of the US-USSR Joint Committee on the Environment 
Protection (Moscow, USSR, February 1988) and the 
recommendation of the "Soviet-American Conference on the 
Ecology of the Bering Sea" (Batumi, USSR. March 1988) and 
the plan of the joint bilateral activity of 02.07-2101. 
"Comprehensive Analysis of Marine Ecosystems and Ecological 
Problems of the World Ocean." the Third Joint US-USSR 
Bering & Chukchi Seas Expedition was held on 26 July 1988 
on board the Soviet research vessel Akademik Korolev. The 
second leg of this expedition was conducted from 9 September 
to 3 1 October in the central Pacific Ocean and South China Sea. 
The delegation was headed by Prof. Alia V. Tsyban and 
Dr. Gregory J. Smith. Nearly 10,000 nautical miles of ocean 
were covered during this leg of the expedition. 

The Soviet delegates were represented by participants in 
the cruise from the USSR State Committee for 
Hydrometeorology and Control of Natural Environment; the 
Academy of Sciences from the USSR; and the Academies of 
Sciences from Ukraine, Belyorussia, and Estonia. A list of 
participants is given as Appendix A. 

Six American delegates joined the expedition during the 
port of call in Hilo, Hawaii, on 9 September 1988 and were 
represented by participants from the US Fish and Wildlife 
Service (Department of the Interior), and the University of 
Washington. A list of participants is given as Appendix A. 

The principal objective of the second leg of the Third Joint 
US-USSR Expedition was to conduct comprehensive studies 
of the ecology of the central Pacific Ocean with an emphasis on 
ecosystem processes and the effects of anthropogenic pollutants 
on those systems. Included in this research was a complete 
zoological, chemical, and microbiological assessment of an 
isolated tropical coral atoll. Caroline Atoll, Kiribati. 
Hydrological stations were included along the route from 
Hawaii to Christinas Island, where a port of call was made to 
bring aboard a wildlife biologist from the government of 
Kiribati for studies of Caroline Atoll. During the period of 
22-29 September, extensive studies of this remote atoll's 
ecology were conducted and an assessment of the atoll's status 
with respect to environmental contamination was made. A 
complete survey of the atoll's plant and bird communities was 
also done. A full ecological station was sampled offshore from 
the atoll and series of four more stations were included enroute 
to Ratawa where the Kiribati representative disembarked. 
After Tarawa, the research vessel proceeded to I IN latitude. 
Along (he 1 I transect, six ecological stations were studied 
in the Marianas section. The first station was m more than 
6.00(1 meters of water, the last station along this transect was 
near the eastern Philippines, and the Akademik Korolev 



proceeded through the Straits of Mindanao into the Mindanao. 
Sulu. and South China Seas. This route avoided Typhoon 
Ruby, which was located in the north Philippines and caused a 
severe loss of life and property in that area. On 25 October 
1 988, a three-day anchor station began in the South China Sea 
at 6°N latitude, three miles west of the 107th parallel. This 
long-term ecological station was the first in a series of five as 
the expedition proceeded into Singapore. Arrival at Singapore 
was on 31 October 1988, completing the second leg of the joint 
American-Soviet expedition of 1988. The route of the 
expedition is shown in the Frontispiece. 

The main scientific tasks: 

/. Biological, chemical, and physical fundamental data 
were collected to provide a comprehensive ecological and 
oceanographic profile of the central Pacific Ocean and South 
China Sea. These data provided a basis for comparison with the 
Bering and Chukchi Seas. 

2. Cooperative studies of the methodology of collection, 
quantification, and chemical analysis of plastic debris were 
conducted to assess the overall hazard of plastics to marine life. 

3. The ecological health of a remote coral atoll in the 
central Pacific was determined. Complete studies of the bird 
and plant communities were done. Studies of chlorinated 
hydrocarbons, their concentrations, distribution, and 
degradation by photolytic and microbiological processes were 
determined, thereby providing an assessment of the effects of 
anthropogenic activities. 

In accordance with specialties of the expedition's 
participants, working groups were organized. At these meetings, 
work schedules, joint studies, and model experiments were 
planned. During the expedition, meetings of the Scientific 
Council Board were also held. Discussed at these meetings 
were the important scientific findings of the expedition, 
comparing the Pacific Ocean and the Bering and Chukchi Seas. 
Also discussed were new areas of investigation that were 
introduced into the research effort with the new delegation of 
American scientists on the second leg. These areas 
included studies of marine plastic pollution, bird ecology, and 
the collection of biota for radionuclide analysis. 
Cooperative studies of the degradation and photolytic products 
of benzo(a)pyrene were expanding during this leg of the 
expedition. 

Complex ecological studies were undertaken for the first 
time in the equatorial part of the Pacific Ocean. Unique joint 
studies thai w ere undertaken to study the ecosystem Of Caroline 
Atoll, an extremely remote area removed from the main sources 
of anthropogenic influences, are especially interesting. During 
the course of the cruise, results of preliminary investigations 
were obtained. 



VIII 



The concentration of dissolved oxygen in the lagoon 
waters of Caroline Atoll reached 1 15 to 126%, and the vertical 
distribution of hydrochemical parameters were relatively 
homogeneous. Salinity in lagoon waters was slightly higher 
than in water sampled outside of the lagoon (ocean water); the 
average values for lagoon and adjacent ocean water were 
36.05 and 35.85 parts per thousand, respectively. Nutrient 
concentrations were lower in lagoon water than in coastal 
water: silicon and phosphorus were two times lower, nitrates 
were one and one-half times lower. According to studies of 
hydrooptical characteristics the lagoon waters do not differ 
significantly from the ocean waters. 

The mean rate of biosedimentation of particulate organic 
matter (POM) in the water column from to 100 m in the area 
investigated in the tropical zone of the Pacific Ocean was found 
to be 14.2 to 72.5 mg of dry matter/m'/day. In coastal waters 
of the atoll this rate was 64.5 mg dry matter/mVday. The rate 
of biosedimentation in the surface water was two times higher 
than the rate at a depth of 100 m. 

The most probable number (MPN) of heterotrophic 
saprophytic bacteria in the water inside and outside of the 
Caroline Atoll lagoon was 1,000 to 10,000 cells/ml. The MPN 
ofparaffin-oxidizing,benzo(a)pyrene-transforming,andPCB- 
transforming bacteria in the water near the atoll was 10 to 
1,000 cells/ml. In the water inside of the lagoon the MPN of 
microorganisms of these groups was less: 3 to 10 cells/ml. This 
information gives us the ability to characterize the atoll's water 
as clean, without anthropogenic pollutants. 

The dominant pollutant in the atoll's ecosystem, in the 
widely investigated range of chlorinated hydrocarbons, was 
DDT and its metabolites (approaching 1 ng/1). Other substances 
observed in the pollutant benzo(a)pyrene were more intensive 
at Caroline Atoll than in the tropical zone of the ocean. In one 
hour, the amount of benzo(a)pyrene destroyed approached 
85%, and after three hours more than 95% of experimentally 
added benzo(a)pyrene was destroyed. Ecotoxicity experiments 
conducted in situ dealing with the influence of benzo(a)pyrene, 
PCB's, copper, and cadmium on planktonic communities of 
the atoll's waters showed higher vulnerability of planktonic 
organisms to toxic metals and were relatively more resistant to 
benzo(a)pyrene and PCB's in comparison with the northern 
seas. 

The distinctive feature of the zooneuston of the Caroline 
lagoon in comparison with the adjoining oceanic waters is the 
presence of a neuritic complex of organisms (larval stages of 
benthic animals and the early stages of Copepoda). The mean 
number of organisms was 64 ind/m\ 26 less than in the coastal 
waters of the atoll. Nevertheless, the zooneuston of the lagoon 
play an important role in the formation of the coastal water 
fauna. 

A unique Acropora-Tridacna reef, dividing the lagoon, 
was discovered in the southern part of the atoll. Its length was 
more than one kilometer and its width was 15 to 20 m. At 
certain places along the reef, the Tridacna formed a dense 
aggregation with numbers approaching 40 ind/m : . 

More than 42,000 m 2 of Pacific Ocean surface water were 
sampled for plastic debris and macroscopic spherules. Plastics 
were recovered at six of the 29 different stations sampled. 



At two of these stations, tar balls were also recovered. The 
density of surface plastics in areas that had positive samples 
ranged from 0.00782 to 0.19481 mg plastic/nr. Because 
marine growth on the surface of plastics may alter the specific 
gravity of floating debris, sampling was also done at the 
thermocline to determine if plastic could be contaminating this 
important subsurface stratum. None of the 5 1 ,600 m 3 of water 
sampled at the thermocline at nine different stations contained 
any plastics or anthropogenic materials. The surface water of 
the Caroline Atoll lagoon was sampled at two locations using 
a 102 neuston net. More than 21,500 nr of water sampled 
showed no evidence of plastics or other debris. Extensive 
sampling of surface and subsurface water was also done in the 
South China Sea. Although a wide variety of anthropogenic 
materials were recovered in surface tows, few plastic cylinders 
(raw material) were found. One sample from the thermocline 
contained plastic line, the most common form of plastic 
recovered in this area. Another surface sample from the South 
China Sea contained more than 1 64 tar balls greater than 4 mm 
in diameter. 

Detailed data on the plants, seabirds, landbirds, mammals, 
reptiles, and human disturbance were obtained from 39 islets at 
Caroline Atoll. The flora consists of 19 species (one new to the 
island) organized into 4 natural and 4 anthropogenic plant 
communities; 92% of the islets are pristine. Eleven species of 
seabirds breed (red-tailed tropicbird is a new record): the sooty 
tern, with 189,000 breeding pairs, was the most abundant 
species. The long-tailed cuckoo was recorded for the first time 
in the Line Islands. 

At-sea observations of marine birds indicated high 
densities near uninhabited islands and in known areas of high 
productivity. Low densities were observed in areas of 
low ocean productivity and in areas with high human 
disturbance (Gilbert Islands, and the Bohol, Sulu, and South 
China Seas). 

Beach surveys for anthropogenic debris were done on nine 
islands of Caroline Atoll. Plastic and styrofoam objects 
accounted for 75 to 80% of the total number of items observed; 
however, glass bottles and fishing gear were significantly more 
important with respect to volume and biomass of debris. 
Observations were also made at sea to determine the amount of 
floating debris. 

Samples were collected at Caroline Atoll and selected 
ocean sampling stations for radiological analyses. These 
analyses will compare natural versus anthropogenic 
radioisotopes. Radioisotope concentrations at Caroline Atoll 
will be compared to similar samples obtained in the Marshall 
Islands that contain contamination from atmospheric nuclear 
weapons testing 30 years ago. 

At the end of the joint expedition on board the Akademik 
Korolev, there was an exchange of preliminary data. Future 
exchanges of data and results of analyses will occur in a series 
of three exchanges: 1. 1 March 1989; 2. 1 June 1989; and 
3. 1 October 1989. 

Both sides note with satisfaction the friendly and 
constructive atmosphere of the expedition's work and the 
effectiveness of joint observations allowing for a variety of 
oceanographic and ecological studies. 



The American delegation w ishes to extend their sincerest 
gratitude to the Soviet participants of the expedition. 
The science staff and ship's crew provided an atmosphere 
conducive not only to research but also the main friendships 
that resulted from our meeting. The American delegation 
would especially like to thank Professor A. V. Tsyban and 



Captain O. A. Rostov tse\ for the wealth of scientific and 
maritime knowledge provided the foundation for this highly 
successful scientific endeavor. 

This protocol was written in English and Russian and was 
signed on board the research vessel Akademik Korolev, 
1 November 1988. Both texts are equally authentic. 



For the American side: 

Leader of the American 
Scientific Delegation. 
Patuxent Wildlife Research Center. 
US Fish and Wildlife Service. 
US Department of the Interior 



For the Soviet side: 
Head of Expedition, 

Leader of Project for 

the USSR Side. 
Deputy Director ol Laboratory lor 

Environmental and Climate Monitoring 

Laboratory . 
Goskomgidromet and Academv of Sciences 



Dr. (1. J. Smith Professor A. V. Tsyban 

(This text is a reproduction of the protocol vv ritten on board the R/V Akademik Korolev m I C )S8. The original w as signed bv both project leader.* 



Acknowledgments 



We gratefully acknowledge and thank the many individuals without whose participation this monograph 
may not have been published with the same quality, accuracy, and clarity. 

We thank the US Fish and Wildlife Service and the USSR State Committee for Hydrometeorology 

for their continued support. 

Steven Kohl and Stephanie Miller, the Coordinator and Associate Coordinator of US-USSR Programs, 
US Fish and Wildlife Service (Office of International Affairs), have provided invaluable assistance 
throughout every phase of this project. Their enthusiasm and energy given to this project, and the people 
involved with this project, are outstanding. 

Without each participant of the expedition, and each author of research results, there would be no need 
for a monograph. There are far too many to name here; however, their names are listed with each subchapter 
and in Appendix A in this volume. It is their interest and excitement for the research presented here, and their 
spirit of cooperation so necessary for an international project, that provide the essence of the scientific 
accomplishments. 

We are indebted to each of the US and USSR chapter editors for their help and their patience with the 
seemingly endless questions and tasks assigned to them, and last but certainly not least, for their sense of 
humor which is often the only saving grace inputting together a volume of this magnitude. Their names are 
listed alphabetically below: 

Sergei M. Chernyak Gregory J. Smith 

Cameron B. Kepler Alia V. Tsyban 

Mikhael N. Korsak Terry E. Whitledge 

Clifford P. Rice 

The "Production Team" at Patuxent Wildlife Research Center — Kinard Boone, Patricia A. Holt, 
Susan A. Liga, Robert E. Munro, Patricia A. Nagel, and John C. Sauer — deserves recognition for their 
dedication to meeting the challenge of producing a quality volume in time for it to be distributed to 
participants on the 1992 Expedition. 



Harold J. O'Connor Alia V. Tsyban 



XI 



Table of Contents 



Page 

Foreword iii 

Protocol of the First Joint US-USSR Central Pacific Ocean Expedition on the R/V Akademik Korolev viii 

Acknowledgments xi 

Frontispiece xiv 

Chapter 1 : ECOLOGICAL INVESTIGATION OF A CORAL ATOLL IN THE CENTRAL PACIFIC .. 1 

Chapter 1 Frontispiece 3 

1.1 Ecological Studies of Caroline Atoll. Republic of Kiribati. South-central Pacific Ocean : 

Part 1. History, Physiography, Botany, and Islet Descriptions 5 

Introduction 5 

Geography 6 

History of Caroline Atoll 6 

Pre-European History: Tuamotuan Period 6 

Post-European History: 17th to 19th Centuries 6 

The Late 19th and 20th Centuries 8 

Methods 9 

Naming Caroline's Motus 9 

Structure and Topography 10 

General Account 10 

Reef Flats 10 

Beaches 11 

Lagoon 12 

Substrata 13 

Hydrology 13 

Climate 14 

Vegetation: Vascular Plants and Floristics 15 

Botanical History 15 

Vascular Plants of Caroline Atoll 15 

Floristics and Ecology of the Motus 21 

Ecological Succession 23 

Basic Serai Stages 23 

Ecological Succession on Motus of Different Size Classes 23 

Species-Area Relationships 25 

Plant Communities 26 

General Account 26 

Natural Herb Mat 26 

Beach Scrub with Suriana 27 

Pandanus Forest 27 

Tournefortia Scrub and Forest 27 

Cordia Forest 29 

Pisonia Forest 29 

Coconut Woodlands 32 

Absent Plant Communities 34 

Description and Ecology of the Motus 34 

Nake Island 35 

Long Island 36 

Windward Islets 37 

South Nake Islets 42 

Central Leeward Islets 44 

Southern Leeward Islets 47 

Conclusion 44 

fables 51 

Figures 63 



Plates 93 

Appendix 1 135 

Appendix 2 136 

1.2 Ecological Studies on Caroline Atoll, Republic of Kiribati, South-central Pacific Ocean: 

Part 2. Seabirds, Other Terresterial Anmials and Conservation 139 

Introduction 139 

History of Ornithological Studies 139 

Methods 140 

Seabird Species Accounts 14! 

Other Birds on Caroline Atoll 147 

Other Vertebrates 149 

Coconut Crabs 150 

Conservation: Attributes of International Significance 151 

Tables 153 

Figures 157 

Plates 163 

1 .3 First Records of the Long-tailed Cuckoo (Eudynamis taitensis) on Caroline Atoll, Southern 

Line Islands, Republic of Kiribati 165 

1.4 A Study of the Benthic Communities of Caroline Atoll (Line Islands, Pacific Ocean) 166 

Chapter 1 References 171 

Chapter 2: INVESTIGATIONS AND ANTHROPOGENIC ECOLOGY 177 

2. 1 Distribution of Chlorinated Hydrocarbons in Ecosystems of the Equatorial Pacific 1 79 

2.2 Distribution of Polycyclic Aromatic Hydrocarbons 183 

2.3 The Occurrence and Microbial Transformation of Benzo(a)pyrene in the Waters of the 

Tropical Pacific (Caroline Atoll, Line Islands, Phoenix Islands Transect. South China Sea) 186 

2.4 Cesium-137 in the Surface Waters of the Central Equatorial Pacific 191 

2.5 Quantity and Distribution of Plastics: An Analysis of Chemical Hazards to Marine Life 193 

2.6 The Role of Solar Irradiation in the Oxidative Transformation of Benzo(a)pyrene 197 

Chapter 2 References 203 

Chapter 3: BIOLOGICAL INVESTIGATIONS IN THE CENTRAL PACIFIC 205 

3.1 A Description of Bacterioplankton 207 

3.2 A Study of Primary Phytoplankton Production 212 

3.3 Mesozooplankton 213 

3.4 Zooneuston of the Tropical Pacific 223 

3.5 Observations of Seabirds along a 14.892-km Cruise Track in the Tropical Pacific Ocean and 

the Bohol, Sulu, and South China Seas 225 

3.6 Evidence for a Major Fall Land Bird Migration Corridor Across the South China Sea from 
Indo-China to the Greater Sunda Islands 247 

Chapter 3 References 253 

General Conclusions 257 

Appendix A 259 



o 

m - 



o 

CD 



O 



o 

CD ' 



O 



o 

CD . 



o 

in . 



ca 
ca 

■x. 






-^-0^^ 





y. 






C 






O 






— 












'-> 






O 






_1 


t , 




C 


'-) 
















_- 


c 


(D 


ca 


KJ 




t/3 


f- 


_) 










C 


p 




® 


c 


) 



z 

< 

u 

o 

u 

u 
< 



C 



.c 
U 




ca 

ca 
ca 
H. 



© 



® 



\> 







o 



® 



. 



^ 



<?' 



® 



/ 




■8 



O 
c/5 



3C 
30 



< 

0. 

04 

UJ 

aa 



■a 
u 
n. 
x 

UJ 



act 

D 

l 

t/3 

D 



o 
■a- 



o 

CD 



O 
CM 



O 
CM 



\l\ 



Chapter 1: I 

ECOLOGICAL INVESTIGATION 

OF A CORAL ATOLL 

IN THE CENTRAL PACIFIC 



Editor: 



CAMERON B. KEPLER 



CAROLINE ATOLL 



1000 
I L 



2000 



METERS 



Chapter 1 Frontispiece. Air-mosaic of Caroline 
Atoll, RNZAF 6569. Reproduced by permission of 
the Lands and Survey Department, New Zealand. 




1.1 Ecological Studies of Caroline Atoll, 

Republic of Kiribati, South-central Pacific 
Ocean 

Part 1. History, Physiography, Botany, and Islet Descriptions 

ANGELA K. KEPLER , CAMERON B. KEPLER 1 , and DAVID H. ELLIS* 

*t/5 Fish & Wildlife Senice, Patuxent Wildlife Researeh Center, Southeast Research Station, Athens, Georgia, USA 
X US Fish & Wildlife Senice, Patuxent Wildlife Research Center, Laurel, Maryland, USA 



Introduction 

On 26 July 1988, the Soviet research vessel Akademik 
Korolev sailed from Vladivostok enroute to Dutch Harbor, 
Alaska. There, Soviet oceanographers joined their American 
colleagues to investigate the Gulf of Alaska and the Chukchi 
Sea in the Third Joint US-USSR Eering & Chukchi Seas 
Expedition. When the arctic research was completed in early 
September, the ship headed toward the central Pacific. A 
rendezvous for a second contingent of Americans took place in 
Hilo, Hawaii, on 9 September. Six Americans joined the ship, 
which set sail on a cruise track of 14,892 km that terminated 

6 weeks later in Singapore. An important part of the expedition 
was research in and around little-known Caroline Atoll, at the 
southeastern edge of the Line Group. On Christmas Island, we 
picked up Katino Teeb'aki, a conservation officer for the 
Republic of Kiribati, who represented his government and 
helped our land-based research efforts. After landing on 
Caroline on 22 September, we camped in 2 locations for 

7 nights, surveying the terrestrial plants and animals on all 
39 islets. Caroline is a remarkably pristine atoll with its native 
plant communities nearly intact on all but three islets, and 
teeming seabird communities that, collectively, are second in 
the Line Group only to Christmas Island ( Kiribati ) in diversity. 
For several historical reasons, the natural values of this 
spectacular blend of marine and terrestrial resources have been 
overlooked. 

Caroline Atoll (Chapter Frontispiece) is situated at 10°00'S 
latitude and I50°13'W longitude in the south-central Pacific 
Ocean. Caroline is the southeasternmost of the Southern Line 
Islands, a group of three islands that also includes Vostok and 
Flint, lying 230 km to its west and southwest, respectively. 
Although anthropologically and geographically within 
Polynesia, Caroline is owned by the Republic of Kiribati 
(formerly Gilbert and Ellice Islands). 

Caroline, 9.7 km long, 2.3 km wide at its widest point, and 
26.9 km in circumference, is a crescentic coral ring with 
39 islets centered on a continuous reef enclosing a relatively 
shallow lagoon. Its total land area above high water is 399 ha, 
with islets ranging in size from 0.02 to 107.5 ha. Islets 
extend along 55% of the reef perimeter. The closed lagoon, 
rich in marine life, contains a maze of patch reefs and remarkably 
clear water. 



The island was "discovered" by de Quiros in 1606. 
Although remnants of an ancient Tuamotuan culture still exist, 
the atoll apparently never supported a long-term permanent 
population and has been less affected by man than most Pacific 
islands. Its European history includes guano export, a 
multinational expedition to observe a solar eclipse, and copra 
production. It has been uninhabited since the early 1930's (a 
factor contributing to its relatively undisturbed ecology), except 
for the presence of one family from 1987 to 1991. The primary 
factors responsible for its lack of permanent settlement are 
remoteness, apparent absence of usable ground water, the 
repeated failure of its coconut plantations, absence of a passage 
into the lagoon, a paucity of safe boat anchorages, and an 
abundance of rats. 

Until the 1988 US-USSR Expedition on the Akademik 
Korolev, only one fairly accurate map was available, which 
named seven islets. We have drafted an accurate map based on 
this field work and recent aerial photographs, also naming 32 
previously unnamed islets, 4 islet groups, and an inlet. During 
8 days of intensive field work, we surveyed all 39 islets, 
walking 33 km in systematic cross-islet transects and around 
islet perimeters. 

Soils, of coral and molluscan origin, are categorized into 
five types, from barren coral rubble to rubble mixed with 
humus and guano. Caroline provides an excellent example of 
soil development through different age and size classes of 
motus. 

Caroline's near-pristine, lush native vegetation supports 
27 species of plants organized into 7 plant communities — 
6 natural and 1 anthropogenic. The atollwide distribution of 
each plant species is mapped. Its insular flora, typical of central 
equatorial islands in their natural state, is 85% indigenous 
(possibly up to 93% ), an extremely high figure for anywhere in 
the world. Though Caroline's islets are covered with 
extensive tracts of native woodland, the Pisonia grandis 
forests, up to 25 m high, are particularly notable as they 
constitute some of the best groves left in the Pacific. Toumefortia 
argentea is abundant, and Cordia subcordata, becoming 
quite rare elsewhere, is present on most islets. Cocos is 
present, but only dominates one islet; 22 islets harbor wholly 
indigenous vegetation. 



Islets of many different age and size classes provide 
excellent examples of soil and vegetation development, 
accompanied by an increasing diversity of bird life. 
On account of its relatively low human disturbance and rapid 
recovery to a more natural state, especially over the last 
70 years, Caroline is one of the few Pacific islands that is truly 
an "outdoor ecological laboratory." Many of its disturbed 
islets have recovered so remarkably they are almost 
indistinguishable from those that have remained pristine. 

An analysis of ecological succession on motus of increasing 
size reveals that by the time a motu reaches 0.8 ha in size, all 
the natural plant communities, most plant species, and most 
species of seabirds are present. This is in striking contrast to 
species-area relationships on inhabited atolls with more 
introduced plant species. 

Foreach motu, the main physiographic features, vegetation 
patterns, seabird colonies, miscellaneous biota (coconut crabs 
and rats), and the effects of human activity (if any), are 
described in detail. Included are vegetation maps for each 
motu, and tables and figures relating to species-area 
relationships. 

Permanent protection of Caroline is currently being sought 
by The Nature Conservancy of Hawaii as it negotiates with the 
government of Kiribati for a Southern Line Islands Wildlife 
Preserve, which includes Caroline, Vostok, and Flint. 

Geography 

Caroline Atoll* (Chapter Frontispiece; Figs. 1,2) is a 
small, low coral atoll situated at LOWS latitude, 150°13'W 
longitude in the south-central Pacific Ocean. It lies 2.800 km 
south of Hawaii. 830 km north of Tahiti, and 1 .030 km west of 
the Marquesas Islands. Its nearest neighbors are Vostok and 
Flint, 230 km to the west and southwest, respectively. 

A recent geographic survey of Caroline by the ICBP 1990 
Line and Phoenix Islands Expedition, using a compact satellite 
navigation computer, indicates that the atoll lies one nautical 
mile east of its previously charted longitude position, 1 50° 1 4" W. 
Its range of coordinates are 09°54' to 10°01'S latitude and 
150°12'to 1 50° 14'W longitude. The actual coordinates given. 
10°00'S and 150°13'W, intersect in the lower lagoon just west 
of the "blind passage." 

Caroline is the southeasternmost of the Line Group 
(Fig. 1 ), a widely scattered group of five atolls, five islands, and 
two submerged reefs lying in the south-central Pacific Ocean 
between 06°N and 12°S latitude and 162 and 1 50°W longitude. 
Scattered across 250.000 km 2 of ocean, the Line Group falls 
naturally into two parts: the Northern Line Islands, four atolls 
and one reef north of the equator, and the Southern Line 
Islands, five islands, one atoll, and one reef to its south. The 
name, Line Group, reflects its equator-straddling location. 



■Note: Caroline Atoll is neither physically, geographically, 
nor politically associated with the Caroline Islands, now part of 
the Federated States of Micronesia, more than 6.000 km to the 
northwest. Because of this confusion, we use the name "Caroline 
Atoll" instead of "Caroline Island." 



Although archaeologically and geographically within 
Polynesia, the Line Group was uninhabited when discovered 
by Europeans; its islands have been variously claimed by the 
United States and England. With the exception of US-owned 
Jarvis, Palmyra, and Kingman Reef, all are now governed by 
the Republic of Kiribati (formerly Gilbert Islands). 

Caroline. 9.7 km long by 2.3 km wide at its widest point, 
is a crescentic coral ring 26.9 km in circumference. It is 
composed of 39 islets (Fig. 2) and a few incipient islets, 
centered on a continuous reef flat, submerged at high tide, that 
encloses a relatively shallow lagoon. The total land area above 
high water is 399 ha. Entirely of coralline origin, its geology, 
soils, climate, and vegetation are typical of low-latitude atolls. 
It is relatively unmodified by man. 

History of Caroline Atoll 

Pre-European History: Tuamotuan Period 

Centuries before Europeans encountered Caroline, this 
lonely atoll was inhabited by Polynesians. No oral traditions of 
this occupation are known, but evidence of former habitation 
was evident when de Quiros found the atoll in 1606. He noted 
"an old canoe, lying on her side." and a small grove of coconuts 
planted on South Island (Bennett. 1840; Markham. 1904). 

No further clues to the atoll" s early history were unearthed 
until Messrs. Brown, Brothers, and Arundel exposed about 50 
ancient Polynesian sites in the 1870's while digging for guano 
(Holden, 1884; Arundel, 1890). Although the largest two were 
marked as "graves" on Arundel's 1883 map (Fig. 4). no bones, 
ashes, or human remains were found. Natives living on 
Caroline called them "marai" (marae). Arundel photographed 
and drew plans of them (Fig. 3): depicted are a platform of coral 
and conglomerate rock, surrounded by 10 smaller slabs 
resembling gravestones, all arranged in a rectangular plan. 
Their findings were later identified as Tuamotuan manic 
( Emory. 1 947). Manic, according to ancient belief, "bound the 
ancestral spirits and gods of the kindred to the land, putting it 
under their eternal guardianship" (Emory. 1947). The largest 
manic was on northwest Nake Island, and a smaller one was 
found near the southern tip of Long Island. Both locations 
conform to such prerequisites for building manic as nearby 
shorelines and birds ( see Description and Ecology of the Motus 
section [Nake Island]), which Tuamotuans believed housed 
divine spirits (Emory, 1947. p. 123). Although members of the 
ICBP 1990 Line and Phoenix Islands Expedition located them 
and took photographs and measurements of the Nake site 
(PI. 36). no field work by archaeologists has been conducted. 

Post-European History: 17th to 19th Centuries 

On 21 February 1606. the Portuguese explorer Pedro 
Fernandez de Quiros. employed by Spain, "discovered" Caroline 
Atoll (Markham. 1904; Stevens & Barwick. 1930). naming it 
San Bernardo. Despite its remote location. Caroline was 
encountered early in Pacific history, long before Tahiti. 
Rarotonga, and Hawaii. This is possibly due to its location, for 
early navigators tended to sail due west from South America 
along lines of latitude, and 10 S was an obvious choice. De 
Quiros. the last ad\ enturer in the Spanish age of discovery, was 



leading his second major trans-Pacific expedition with 3 ships 

and 1 50 men obsessed with finding the fabled "Terra Australis 

Incognito." The descriptions of Caroline by his crew, although 

atvariance with one another.still apply today(Pl. 1). Theirfirst 

at-sea impression was that it was "divided into four or five 

hummocks, and all the rest submerged. Its circumference 

appeared to be ten leagues" (Markham, 1904). After landing, 

they found that 

There was a great number offish inshore, and, owing to the 

water being very shallow, they were killed with swords 

and poles. There were great numbers of lobster and 

crawfish, and other kinds of marine animals. They found 

a great quantity of cocoa-nuts in a heap at the foot of the 

palm trees, many large, and of different sizes. There were 

a great quantity of sea birds of several kinds, and so 

importunate that they seemed to want to attack the men. 

We took plenty of all these things. ..It seemed to the 

Captain that on an island where there are so many trees 

there could not fail to be water. (Markham, 1904) 

Fresh water was crucial to de Quiros and his crew, who 

were suffering from lack of food and water. Despite their 

efforts, however, they failed to obtain water. Disappointed and 

lacking energy, they continued their voyage the following 

morning. Their demoralized state may explain one statement 

that Caroline "consisted of twenty-two islets, uninhabited and 

without water, trees or scrub for wood." 

In 1795, Captain W. R. Broughton, on the British sloop 
Providence, rediscovered the atoll while voyaging from Tahiti 
to Hawaii. He named it in honor of the daughter of the First 
Lord of the British Admiralty (Broughton, 1804): 
The southern extremity was the highest part, covered with 
trees, most probably cocoa-nut from their appearance, as 
they stood in detached clumps along the shore. The 
island.. .appeared to be low, and covered with trees, and if 
I am right in its estimated distance, its length will be about 
five miles in a north and south direction. I named it 
Carolina Island in compliment to the daughter of Sir P. 
Stephens of the Admiralty. 
Because early navigation techniques and communication 
were far less sophisticated than today, especially with regards 
to longitude, Caroline was sighted or "discovered" by several 
more explorers who were unsure of its identity. By 1821 the 
atoll had amassed an impressive collection of coordinates and 
names: San Bernardo, Island of Fish, Thornton, Hurst's, 
Clark's, Independence, and Carolina (which later became 
Caroline). Some navigators equated Caroline with an island 
named "San Bernardo" by the Spanish explorer Mendana in 
1595. Mendana's island has only recently been verified as 
Puka-Puka in the northern Cook Islands (Maude, 1968). 

The best early descriptive account of the atoll's flora and 
fauna comes from an 1835 visit by F. D. Bennett, who was 
reasonably well versed in natural history (Bennett, 1840). He 
noted that the islets then, as now, were "covered with 
verdure. ..surprisingly luxuriant, when compared to the arid 
soil it covers." Although Bennett had visited many atolls, he 
was particularly impressed with the quality of Caroline's coral 
reefs. His party observed "rats of a red-brown color" and 



various birds but no reptiles (Subchapter 1.2). Although he 
discusses "land lobsters (Coenobita species)," no mention is 
made of coconut ciabs (Birgus latro). 

First Occupation : The existence of two small coconut 
groves on Caroline prompted two British entrepreneurs, 
representing the Tahitian firm Collie & Lucett, to establish a 
stock raising venture there in 1846. This first known settlement 
was located adjacent to the main coconut grove on the northwest 
peninsula of South Island; a smaller grove evidently existed 
"on the south-south-west side" of the same island (Lucett, 
1851 ). Tahitian laborers tended pigs, hens, turkeys, and grew 
many food plants, including pumpkins and melons. They dried 
and salted fish, planted coconuts, and extracted coconut oil 
(Maude, 1942a;Garnett, 1983 (and were evidently still therein 
May 1852 (Ellsworth, 1990). 

Political Annexation : Though inhabited in prehistory by 
Tuamotuans, officially "discovered" by the Spanish, and visited 
by British, French, and American ships, it took centuries for 
Caroline to acquire a political identity. It was formally annexed 
to Britain by Captain Nares, R. N., who arrived in the H.M.S. 
Reindeer in 1868, finding 27 residents. 

Caroline was under the control of various merchants in the 
late 19th century: Lionel Brown, Captain Brothers, and later 
John Arundel, a well-known businessman, trader, and guano 
merchant in the Pacific. Arundel's 1883 map (Fig. 4) of 
Caroline is the only reasonably correct chart published until 
this paper. 

The Guano Era : Though bonded under the American 
Guano Act in 1860, no phosphate was dug on Caroline until 
Arundel was granted a 7-year license in 1874. A few months 
earlier, a set of moorings was laid off the lee side of South 
Island, allowing ships of up to 1 ,000 tons to lie safely during 
trade wind weather. Guano was the only successful business 
venture at Caroline: approximately 10,000 tons were shipped 
to California and Australia between 1873 and 1895, when 
supplies became exhausted (Young, ca. 1922). 

Solar Eclipse Expedition : In 1883, Caroline received 
international publicity when astronomers calculated that it lay 
directly under the path of a pending solar eclipse. As a result, 
three parties of astronomers (American, British, and French) 
set up camp on South Island, making detailed observations of 
this celestial event (PI. 2a). At that time Caroline was more 
famous, and housed more people, than before or since: 7 
"natives," scientists, and crewmen totaled 5 1 occupants. 
Legacies from former inhabitants included three houses 
(PI. 2b), two sheds, three huts on smaller motus, nautical 
flotsam and jetsam, and two shallow wells. To this they added 
tents, observatory frames, a marble slab, flagpole, and brick 
"piers" for their telescopes, most of which remained as 
technological litter. 

This expedition (Dixon, 1884; Holden, 1884; Holden & 
Qualtrough, 1884; Trelease, 1884; Young, 1884) also marked 
the first attempt to describe the topography, climate, flora, and 
fauna of the atoll. Drawings included an artist's rendering of 
Caroline and map of the "settlement" (PI. 3) and views along 
South Island's lagoon shore (Pis. 4, 5). Another map (Fig. 5) 
was drafted but is highly inaccurate. 



Observations by the astronomers on vegetation, birds, 
insects, reptiles, marine organisms, et cetera were sketch) 
( Butler &Strecker, 1884; Dixon, 1884). Dixon, the zoologist, 
listed such organisms as •'shrimp." "'hermit crabs," "gnat," 
et cetera. As with Bennett, there was no mention of coconut 
crabs, even though they were evidently abundant on South 
Island in 1910 (Young, ca. 1922). 

The Late lVth and 20th Centuries 

In 1875. C. D. Voy. a naturalist from California visited 
Caroline, collecting mollusks (Pilsbry & Vanatta, 1905a.b) 
and fish (Fowler, 1901). 

As early as 1885, Arundel began to clear land and plant 
coconuts, but his planned copra industry was unsuccessful. In 
I S97 he sold his business to the Pacific Islands Company. Ltd., 
which also failed. The coconut plantations suffered from 
disease and poor vitality, and populations of Polynesian rats 
apparently exploded. By 1904, when the H.M.S. Icarus visited 
Caroline, only six Polynesians lived there. A few months later 
they were repatriated to Niue, and Caroline remained 
uninhabited until 1916. when a new effort was made to develop 
the coconut plantation by Messrs. S. R. Maxwell & Co., Ltd. 
During the uninhabited years. South Island"s vegetation 
and wildlife began to recover from the earlier forest felling 
( Pis. 2-4). When Mr. J. L. Young, then managing director for 
S. R. Maxwell & Co., Ltd. ( Young, ca. 1 922 ), visited the atoll 
in July 1910, he described it as a wilderness, teeming with sooty 
terns, fish and coconut crabs: 

The ground was covered with nests of seabirds which 

latter rose like a cloud when disturbed: the noise of their 

shrieking was so great that one had to shout to enable 

oneself to be heard by his companions. Hundreds of great 

Coconut Crabs were seen: 40 large ones were caught by 

the crew of the schooner in an hour. The reef and the 

lagoon swarmed with fish and small sharks. 

From 1 9 1 6 to 1 929. Caroline was altered more than before 

ui since. All the available land on .South was deforested to 

make room for thousands of palms, and laborers demolished 

huge numbers of coconut crabs and seabirds. which were 

thought to damage the palms (Young, ca. 1922). In addition. 

coconuts were planted on all of the main windward islets. 

southern Nake. and on Mannikiba. Plantation workers in great 

part lived oil the land, feasting on fresh fish, seabirds. seabird 

eggs, turtles, and coconut crabs. 

Copra exports averaged around 1 4 tons per year from 1929 
to 1 934. after which the company ran into debt. Concurrently, 
the French government forbade further recruitment from Tahiti; 
b> l936onl> a lew families were left (N.I.D., 1943). In 1941 
(he atoll earned a price tag of 000 English pounds (Maude, 
personal communication), but was never purchased. 

Occupation leases for Caroline were canceled in 1943. 
after which the British Western Pacific High Commission 
repossessed it (Maude, 1953). However.nev. "queen's leases" 
were granted to M. P. A. Bainbridge of Papeete. Tahiti 
( 1 95 I - 1 964 ) ( Nicholson & Douglas. 1 969), (hen ( 'aplam ( )mer 
Darr ( 1964-1989) of Moorea. French Polynesia. When the 



British granted independence to the Gilbert and Ellice Islands 
in 1979, a new country, the Republic of Kiribati (pronounced 
"Keer-ee-bahss") assumed ownership of Caroline, along with 
most of the Line and Phoenix Islands. 

Apart from occasional parties of Tahitians cutting copra 
and a shipw recked sailor in the early 1 880's. the atoll remained 
uninhabited for over 50 vears. During this time. Caroline's 
vegetation and w ildlife recovered to such an extent that, were 
it not for unpublished manuscripts from Maude (Maude, 
ca. 1938.ca. 1 942a. and no date (.including Young' s(ca. 1922) 
"Memoranda re Tahitian Business"our detailed vegetation 
analysis (including tree diameters), and the 1 990 comparisons 
with Flint and Vostok. we would have been unaware of the 
actual extent of previous human interference or of the rapidity 
of forest recovery (the fact that 60<7r of Caroline's motus 
harbored wholly indigenous vegetation seemed to point to a 
relatively pristine atoll). 

In 1987. the Office de la Recherche Scientifique et 
Technique Outre-Mer ( ORSTOM. a French scientific research 
agency) was requested by the Kiribati government to conduct 
a short study at Caroline on the feasibility of pearl-shell culture 
(G. Monet, personal communication). Their results concluded 
that the atoll would be inappropriate for this type of development. 

Also in 1987, a Scotsman. Ron Falconer, his French wife 
Anne, and two small children settled on Caroline as volunteer 
custodians. From October 1989 to November 1990 a new lease 
was under negotiation by Felix Urima. a French businessman, 
who planned to blast a channel through the reef, construct an 
airstrip, build a small hotel, cut timber, and engage in various 
commercial ventures including fishing, a turtle farm, and pearl- 
shell culture. In April 1 990. Urima' s workers began commercial 
fishing, killing turtles and coconut crabs, and clearing land 
( Kepler. 1 990a). This w as a major new insult to the atoll which, 
in spite of its long history of intermittent human occupation, 
remains to this day "possibly one of the least spoiled o( true 
atolls in the Pacific" (Stoddart. 1976). Reports from our 
expeditions to Caroline (Kepler & Kepler. 1989; Kepler. 
1990a.d) resulted in the short-lived cancellation of Urima' s 
tentative lease in November 1990. As of this writing. Urima 
has returned to unlimited fishing at Caroline's reefs. The 
government of Kiribati appears amenable to negotiations for a 
wildlife preserve (see Subchapter 1.2 [Conservation section], 
this volume ). Conservation efforts are presently underway tor 
Caroline to become part of a triple-island wildlife presen e with 
Vostok and Flint (see Subchapter 1.2 [Conservation section], 
this volume). 

20th Century Scientific Studies : In June 1965, afield party 
from the Smithsonian Institution's Pacific Ocean Biological 
Survey Program ( POBSP) visited Caroline for 2 days ( Clapp & 
Sibley. 1971a). Then vsurvej and specimens added much to the 
pre\ ious botanical and ornithological know ledge of the island. 
Other quick visits were made b\ the Line Islands Expedition on 
9-10 September 1974 and by Roger Perry, then Wildlife 
Warden of the Line and Phoenix Islands, on 12-13 November 
1977. Eleven years later the present authors surveyed the 
terrestrial env ironments of Caroline more thoroughly than had 



been previously attempted. This visit was part of the Third 
Joint US-USSR Bering & Chukehi Seas Expedition, a 
4-month voyage that also ineluded oceanographic studies in 
the Pacific Ocean and South China Sea. 

In 1990, one of the authors (AKK) visited Caroline twice 
with the International Council for Bird Preservation (ICBP) 
1990 Line and Phoenix Islands Expedition. These visits were 
primarily to discuss conservation matters with the Falconers; 
introduce Caroline to Dr. and Mrs. M. Garnett, representatives 
from ICBP; confirm the illegal taking of fish, turtles, and 
coconut crabs by Mr. Urima's men; find the Tuamotuan 
maraes; survey North Arundel Islet; collect invertebrates; and 
fill in missing details from the 1988 expedition. 

Methods 

From 22 to 29 September 1988, Drs. A. K. Kepler, C. B. 
Kepler, D. H. Ellis (USA), and Mr. Katino Teeb'aki (Republic 
of Kiribati) surveyed all 39 motus at Caroline Atoll (Fig. 2), 
gathering detailed information on plants, seabirds, land birds, 
mammals, reptiles, coconut crabs, and human disturbance. 
Some incidental data have been added from the two visits in 
1990 (10-13 March, 18-28 May) by Dr. A.K. Kepler, Captain 
G. Wragg, A. Garnett, M. Linsley. J. Phillips (March), and 
Dr. M. Garnett (May). 

Prior to the first expedition, a series of transects and known 
botanical information were mapped to ensure that 5% of each 
motu was sampled and to maximize our chances of encountering 
all known plant species. Transects on the larger motus (Nake, 
Long, South ) were spaced approximately 400 m apart and, with 
one exception, were perpendicular to the long axes of each islet 
(Tr. 3 on Nake extended first from east to west, then ran south 
parallel to the west coast). If the motus were greater than 
400 m long, we used two transects. Transects on the smaller 
motus passed through their widest points (Fig. 8). Their 
lengths ranged from 77 m (Azure) to 2,000 m (Tr. 3, Nake). 

Considerable modifications were required when we realized 
that the Clapp& Sibley ( 197 la) map (Fig. 7) was incorrect. We 
redrew the transects on Arundel's 1883 map (Fig. 4), secured 
just prior to the expedition. On South Island, due to impenetrable 
draperies of Ipomoea vines, Tr. 3 was omitted, Tr. 5 ran only 
from the lagoon south to the Ipomoea curtain (75 m), and 
Trs. 4 and 6 ran north and south until we reached an impasse 
(PI. 7). 

Compass headings were determined by the configuration 
of each island. Beginning at high water mark, all distances 
(islet dimensions, widths of reef flats and substrates, and plant 
communities) were measured using hip chains with 
biodegradable cotton thread. These parameters were later 
checked against aerial photographs in stereoscopic pairs taken 
in 1 985 by the Royal New Zealand Air Force ( RNZAF), which 
provided 3-dimensional overviews of every islet. Vegetation 
maps for each islet were constructed by drawing the outlines 
and plant communities (based on the aerial photos) on graph 
paper, enlarging, then counting dots. 

Data were collected in a 30-m (15 m to each side of the 
transect) swath along each transect and recorded on field 
forms. Within each plant community we took photographs. 



assessed the relative abundance of each plant species (rare, 
uncommon, common, very common, abundant, and locally 
common; see Vegetation section ) and recorded plant community 
width, plants collected, and substrate type. We also estimated 
the maximum height of the dominant vegetation and percentage 
of ground area covered by each species. 

In addition to the linear transects, an additional 19,300 m 
of perimeter surveys were conducted on 21 islets (Fig. 9). The 
combined distance for linear and perimeter transects was 
32.6 km. Seven tiny islets (Noddy Rock, Skull, Atibu, Bosun 
Bird, Coral, Reef-flat, Fishball) were surveyed completely. 

In 1988. we camped on the atoll for seven nights, 
establishing base camps ( Fig. 8; Pis. 6,8) on the northwest point 
of South (22-24 September) and southwest Long 
(25-28 September). We relocated camps using a Zodiac with 
outboard motor and an inflatable Sevylor canoe (PI. 9). All 
transects were surveyed during daylight hours, beginning at 
dawn. Walking the interislet channels was relatively easy at 
low tide, but became hazardous at incoming or high tide, as the 
numerous black-tipped reef sharks, Carcharhinusmelanopterus 
(PI. 10), regularly charged at our legs and had to be beaten off 
with sticks and coral chunks. 

During the 1990 visits, we stayed on Motu Ana- Ana with 
the Falconers. Work was not intensive, as in 1988. We walked 
or motored an inflatable Lancer, visiting 20 motus: Ana-Ana, 
Kimoa, Pisonia, Eitei, South, North Arundel, Noddy Rock, 
Brothers, North Brothers, Skull, Pig, North Pig, Bo'sun Bird, 
Long, Nake, Mouakena, Shark, Scarlet Crab, Bird, and Fishball. 
Insects were preserved in ethyl alcohol. We used a "Magellan" 
NAV 1000 to obtain accurate geographical coordinates of 
Caroline. 

Naming Caroline 's Motus 

Previous literature has provided vague or incomplete data 
on Caroline's constituent motus (Bennett, 1840; Holden & 
Qualtrough, 1884;Markham, 1904; Stevens &Barwick, 1930; 
Bryan, 1942; Clapp & Sibley. 1971a; Garnett, 1983). This 
confusion resulted because most previous visits had been brief. 
The only charts available were a quite accurate survey by 
Arundel, a guano merchant who mapped the atoll in 1883 
( Fig. 4; Admiralty Chart, No. 979, 1 965 ), and a map. greatly in 
error, drafted by an international Solar Eclipse Party, also in 
1883 (Fig. 5). There are no hydrological navigation charts. 
Unfortunately, the eclipse party map has been used in all 
subsequent scientific, historical, and sociological publications 
(Bryan, 1942 [Fig. 6]; Maude. 1968; Clapp & Sibley, 1971a 
[Fig. 7]; Garnett, 1983). It shows only 25 of the 39 islets and 
many of the shapes are distorted. The 38 islets on Arundel's 
map are similar to those in the RNZAF (1986) aerial 
photographs. Only a few islets appear to have changed in 
minor ways since 1883: major discrepancies in Arundel's map. 
we believe, are due to difficulties involved in the accurate 
rendition of small land areas (i.e., the South Nake Islets). 
Maude (ca. 1938) counted 36 islets but never published his 
information. 

To aid our survey we named 32 islets and 4 islet groups 
(Fig. 2). Our names, whether in English or Gilbertese, reflect 
appropriate aspects of islet biology. Etymology is provided in 



the Description and Ecology of the Motus section. Gilbertese 
names are prefixed with motu. Any name not appearing on 
Arundel's map ( Fig. 4) was given to the islets by us. They have 
been sent to the British Admiralty and US Hydrographic Office 
(along with corrections to the Pacific Pilot) for official 
recognition. 

Structure and Topography 

General Account 

Caroline, one of the oceanic islands contributing to 
Darwin's theory of atoll formation (Darwin. 1842), is a low 
island derived entirely from coral reefs and mollusks, living 
and dead. Although undoubtedly resting upon an ancient 
basaltic base, today's atoll reveals no visible fragments of its 
volcanic heritage. "The atoll consists of a chain of twenty-five 
[sic] little islets, well covered with trees and shrubbery, the 
whole forming a quiet scene of grove and lake, charmingly set 
off by the contrasting ocean" (Holden & Qualtrough, 1884). 
This description could apply equally well today. Caroline is 
actually composed of a ring of 39 "permanent" and 3 incipient 
islets whose total area is 398.94 ha. Most are well wooded, but 
four tiny ones, less than 0. 1 ha in size, are scarcely more than 
coral rubble piled on the reef, supporting sparse patches of 
Tournefortia and Heliotropium. One islet. Noddy Rock 
(PI. 19), is a vestige of a former reef segment. 

Our terminology is based on Tracey et al. (1955) as cited 
by Wiens (1962), to which we have one addition. The name 
motu, Polynesian for "islet" or "small island," is now a technical 
term for detrital reef islands (Danielsson, 1954; Stoddart & 
Gibbs, 1975). In this paper the terms motu and islet are used 
interchangeably; however, as motu is now a bona fide English 
word and not italicized, it may be pluralized by adding an "s" 
(normally Polynesian words are not pluralized with "s"). 

Caroline's overall shape is like a flattened crescent. 9.7 km 
long on its north-south axis. Its perimeter is 26.9 km. measured 
along the outer reef. Its greatest breadth ( including both reefs), 
2.3 km, lies centrally along an east-west transect that includes 
Motu Mannikiba and lower Long Island. The longest islet. 
Long, extends 4.23 km from north to south, while South Island, 
extending 1.2 km from east to west, claims the widest stretch 
of land. 

The motus, lying upon a wide, continuous reef flat which 
encloses an elongated, relatively shallow lagoon, fall naturally 
into groupings of 3 large islands (South, Nake, Long) and 
4 groups of smaller islets! 13 Windwards, 5 Southern Leewards, 
1 I Central Leewards, 7 South Nakes). There are four basic 
motu shapes: 

/. long, linear, and parallel to the reef axis (e.g.. Long 
Island); 

2. small, linear or oval, and perpendicular to the reel axis 
(e g., Southern Leeward Islets i; 

3. triangularorcrescentic.withtheapex lacing the seaward 
reel (e.g.. most of the Windward Islets); and 

4. large and quadrangular, occupying the ends of the atoll 
(e.g.. South. Nake). 

All individual motus are discussed m detail in the 
Description and Ecology of the Motus section. 



Reef Flats 

Caroline's peripheral reefs, which completely surround 
the lagoon and upon which the motus rest, are consistently wide 
(average 562 m, range 396-759 m, N = 100). The windward 
and leeward reefs differ in structure and dimensions. Neither 
are entirely dry, even at the lowest tides. They consist primarily 
of barren calcareous rock, which on other atolls generally 
represents the erosional surface of an older reef. Jagged 
"mushrooms" of newer (but dead) reef dot the leeward reefs 
(PI. 11 ). There are no passes from ocean to lagoon, a typical 
feature of central Pacific atolls (Wiens, 1962). The combined 
area of intertidal and subtidal marine environments that they 
enclose is several times the areaoccupied by terrestrial habitats. 

In the Southern Hemisphere, reef flats tend to be widest in 
the southwest sector and narrowest in the northeast (Wiens, 
1962). Caroline's reefs are quite wide throughout the west 
(PI. 12), and definitely narrowest in the northeast off Nake 
(PI. 13). 

The reef rim. irregularly dentate and 26.9 km in 
circumference, is surmounted by islets for 55% of its length. 
On 72% of all Pacific atolls, less than half the reef circumference 
is occupied by land (Wiens. 1962). Caroline lies within a 28% 
minority in which one-half to two-thirds of the reef rim contains 
land. Corresponding values for 2 Tuamotuan atolls. Rangiroa 
and Raroia. are 33% and 35% (Stoddart & Sachet. 1969). 
Where motus exist, the reef flat is divided into the seaward reef 
flat (PI. 12), islet, and lagoon reef flat (PI. 14). 

At low tide all reef fiats are wadable. Black-tipped reef 
sharks were a threat to our safety in most areas in 1988, but by 
1 990 dozens had been killed. The reef segments that separated 
the Southern and Central Leeward groups and the Leeward and 
South Nake groups were particularly hazardous. Within these 
islet clusters, names such as Blackfin, Shark, and Danger 
reflected this ubiquitous feature of Caroline. 

The only tidal measurements taken were by the Solar 
Eclipse Party (Holden & Qualtrough. 1884). who noted that in 
May 1 883 the greatest daily fluctuation ranged from a maximum 
of 475 mm (1' 7") to a minimum of 125 mm (0' 5"), similar to 
that (around 2') in the Tuamotus (Stoddart & Sachet. 1969); 
Arundel's map (Fig. 4) gives 0.5 m ( 1.5') for Caroline, which 
we have tentatively used in the schematic profiles 
(Figs. 34-36) as the difference between low and high spring 
tidal levels. 

Windward Reef Flats : Constantly pounded by surf 
(PI. 16). the windward reefs are typically narrower than those 
to leeward, averaging 5 19 m ( range 396-759 in), though this is 
less evident from a map than in the field. 

The windward reefs are 13.5 km long, surmounted by 16 
motus that total 63% of its length. This is not surprising, as a 
recurrent pattern on central Pacific and Tuamotuan atolls is that 
motus are more frequent along windward reef rims (Thomas. 
1961; Wiens. 1962). The longest islets are Nake ( 1.980 inland 
Long (4.226 m), both formed from the coalescence of two or 
more smaller islets (Figs. 37-39). The rest vary from 1 S m to 
858 m in length. 

The character of these reef Hats differs, depending on the 
presence or absence of land, interisland distances, lagoon 
depth, and recent weather conditions. In February 1990. part 



10 



of a hurricane hit Caroline, rearranging tons of coral rubble and 
sand on the windward beaches and motus, tearing out 
Tournefortia scrub, obliterating much of the native herb mat, 
exposing beachrock, depositing large chunks of broken reef on 
the seaward reef flats, and changing the sizes and shapes of the 
beach crests. 

Reef Rim w ith Motu : The w idth of the seaward flats is 
quite uniform, averaging 307 m (range 1 93-396 m), occupying 
517c of the rim width. It consists of a slightly raised algal ridge 
bearing the brunt of incessant wave action and a rubbly reef flat, 
partly drying at low water, which sweeps up to the motu's 
beach (PI. 16). Adjacent to Nake, in the northeast, the seaward 
margin forms a shallow moat separating the land and algal 
ridge (PI. 13). 

The motus differ considerably in width, ranging from the 
narrow tip of Long, merely 30 m wide, to Windward, 290 m 
wide. Nake and South islands, forming "caps" to the atoll at its 
upper and lower ends, respectively, exhibit characteristics 
more typical of windward than leeward motus. Because coral 
debris accumulates wherever atoll reefs turn sharply (Thomas. 
1961), these two motus are the widest on the atoll (Pis. 16.17). 
A comparison of maps a century apart ( Figs. 2.4) indicates that 
several layered ridges of coral debris have accumulated on 
northern Nake since 1883. 

Reel Rim without Motu : Zonation within the reef flat is 
less marked where there is no land. Within these interislet reef 
flats, however, certain areas of high water transport have 
carved surge channels and grooves. These are particularly 
e\ ident at the north and south ends of islets ( PI. 18), between 
islets, and within the longer flats. In all these areas, tidal tans 
extend into the lagoon especially at its northern end where 
sedimentation is most active. Caroline has no deep pass or 
navigable channels into the lagoon nor a ship anchorage 
beyond the reef, though small boats may anchor within the 
close lee of South Island during normal trade winds and low 
seas. Landing in an inflatable is best made across the reef 
slightly north of the "boat entrance" (marked by an upright 
anchor I. 

The reef flat between Tridacna and South Island, serrated 
with 6 erosional grooves, one labelled "blind passage." is of 
particular importance to navigators. Its most southerly channel 
is a narrow diverticulum 380 in long within a reef 430 m wide. 
On all previously published maps this passage is drawn as 
though it completely connects ocean and lagoon (Figs. 4-7). 
However, it is a true blind channel (Fig. 50), serving as a 
sheltered anchorage for motored yachts near its lagoon end. hut 
cannot be entered or exited during high winds or moderate-to- 
high surf. 

Leeward Reef Flats : These are wider, flatter, gentler, more 
consolidated, and less filled with rubble than the windward 
reefs(Pl. 1 1 ). Everywhere except within the surge channels, an 
orange-colored alga blankets the coralheads. chunks of upraised 
coral (PI. 11). carbonate rock, and giant clams. This alga is 
found on many atolls — for example Enewetak and Rangiroa 
(Stoddart & Sachet. 1969: US Department of Energy. 1987). 
Living coral is sparse. 



Surge Channels : These occur in a variety of shapes and 
sizes, depending on the distances between motus, the extent 
and buildup of reef flats adjacent to land, and lagoon depth. 
Surge channels and reef grooves (hoa) are deeper on the 
windward side. Aerial photos indicate that the vigorous 
currents washing daily into Caroline's lagoon have created 
largerdebris fans between windward motus than between those 
to leeward (see Chapter Frontispiece). 

Beaches 

Caroline's beaches — the zones lying between low water 
mark and the inland limit of wave-deposited debris — are entirely 
of reef origin. There is, however, considerable variation in the 
sizes of coral rubble, and the proportions of sand and silt with 
which they are mingled. In general, the windward beaches and 
surge channels, subject to winds and swells and in a constant 
state of erosion or deposition, support the greatest variety of 
sediments: well-sorted sands; gravels of coralline, algal, and 
molluscan origin; and a wide variety of sizes of coral rubble. 

Almost all exposed rubble on Caroline is colored from 
pale to dark gray , aconsequence of penetration by cyanobacteria 
(Fosberg, 1953). Typically the oldest rubbles, highest up the 
beach and extending into the interior, are darkest. A marked 
beach crest rises — gently or abruptly — from the windward 
beaches, at the crest of which is deposited an assortment of 
flotsam and jetsam: bottles, plastic, wood, coconuts, et cetera 
(Pis. 16,20). No large chunks ofdisengaged reef were found on 
or near land in 1988. but in 1990 many of these littered the 
windward reefs and shores, the result of cyclonic weather in 
February of that year. Similarly, in 1990, considerably thicker 
deposits of coarse sand had overlain the rubbly windward 
beaches and interislet channels of 1988. 

Alterations to Caroline's beaches provide the major areas 
of change in islet shape. The main areas of aggradation are on 
the northwest and southwest leeward points of the 
windward islets. This is particularly true of the largerones such 
as Brothers (which has joined with a separate islet 
mapped by Arundel ). and Windw ard and Tridacna ( which have 
added more sediment to their southwest points during the last 
century). 

Beachrock : These narrow, elongated strata of eroded 
upraised reef, brown consolidated sands, and reef detritus are 
not abundant on Caroline. Occurring as strips at the low water 
mark, they flank the windward beaches of Nake, Long, and 
South (Figs. 37.38.50; PI. 58) and a few of the leeward islets. 
They become more exposed alter violent storms. Extensive 
areas of a coarser conglomerate (PI. 21 ) hug the inner reef flats 
of western South Island. 

Upraised Reef : In a lew areas, jagged, eroded upraised 
reef {champignon or /<•<<) is evident — for example, the lower 
quarter of Long. A thin soil cover supports a forest of lower 
stature than would otherwise be expected. In this minikarst 
area the rocky substrate is pitted with holes of varying sizes and 
undermined with subterranean tunnels in which at least two 
species of land crabs (Birgus latro, Cardisoma spp. ) shelter 
( PI. 22). Noddv Rock (PI. 1 9). the smallest motu (0.02 ha), and 



11 



many jagged coralline "mushrooms" found on the reel Hats 
(PI. 11 ). are probably remnants of former reef flats formed 
when sea levels were several feet higher than present. 

Lagoon 

A notable attribute of Caroline's lagoon fromaconservation 
perspective is its outstanding clarity and beauty. Throughout, 
but especially near the Central Leeward Islets, the twisted reef 
configurations studded with white sandy channels and deeper 
circular openings generate a spectrum of stunning colors — 
turquoise, apple green, tawny, azure, and royal blue (PI. 25). 

Caroline's lagoon. 8.9 km long, is closed. Though it 
appears to dominate the atoll, its total area is less than that of 
the combined reef flats. The lagoon is relatively shallow, 
tapering in shape and depth at each end, and is crisscrossed with 
living coral. Its bathymetry is unknown. 

In the north the lagoon is more sheltered, as the presence 
of continuous vegetated land buffers the easterly trades. At its 
northern extremity, merging reef flats squeeze the lagoon until 
it disappears east of Pandanus Islet. A filled-in portion of the 
former lagoon penetrates Nake for 300 m as a fishhook-shaped 
mudflat, Sandy Inlet (Fig. 37. PI. 23). before succumbing to 
encircling vegetation. At the lagoon's southern end, where 
winds whip through the "blind channel." it is choppy, having 
more sediment and slightly less visibility. However, w ithin the 
lee of South Island' s north-central curve, the lagoon is frequently 
quiet and reflective (PI. 24). 

Lagoon Hydrology : Although Caroline's hydrology has 
not been studied, it has been closely observed for over two 
years by Ron Falconer, especially the south end of the lagoon 
and "blind passage.'" He has noticed that the lagoon water is 
tv pically "perched" at a level above that of all but the daily high 
tides. High tide water flows rapidly over the reef flats into the 
lagoon but is held back by the reefs as the tide lowers. Ron has 
noted that lagoon water at low tide is about 0.3 m higher than 
water in the "blind passage." Water moves out of the lagoon 
through a few channels that, although deep in places, form 
broad, shallow troughs over the reef flats. A major channel 
with a current flowing west at several knots passes along the 
northwest point of South Island, although water passage is 
impeded hv the reel flats west of South Island. If a channel 
u ere to he blasted through the reef flats, .is has been proposed, 
this delicate hydrolog) would be disrupted. For example, the 
high tide w ater is never more than 20 cm above the coral heads 
and reels m the lagoon. A man-made reef channel for vessels 
could lower water levels 30-40 cm. thereby exposing and 
killing the extensive Acropora Tridacna reefs within the lagoon. 

The "blind passage" northeast ol South Island (Fig. 50) is 
sustained by a powerful northward How of water along the east 
coast of South Island and a strong southward flow of water 
along the seaward reels of Tridacna Islet. The South Island 
flow is apparently augmented bv water draining from a large 
shallow basin on the reef Hats south of the island. Water spills 
into the "blind passage" and drams east at about 4 knots against 
(he pre\ ailing Hade w inds and surf. There is much less current 
at the west (inner) end of the passage, where less water is 
collected, and throughout the passage at low tide i there being 
essential!) no water How out from the lagoon). 



Patch Reefs : A complex series of patch reefs and coral 
knolls ( primarily Acropora spp. i. circular and elongated, flank 
the smaller motus and crisscross most of the lagoon's area (see 
Chapter Frontispiece). They are particularly evident in the 
southern two-thirds of the lagoon, where they approach and/or 
break the surface. Coral limestone bedrock, surmounted by a 
variety of living coral, mollusks. and other invertebrates, 
provides their basic structure (Fig. 10). The atoll's perimeter 
reefs shelter the knolls from storms, surf, and excessive erosion. 

Caroline's lagoon is gradually filling in with ever- 
expanding patch reefs and debris washed in from the fringing 
reefs. Since Arundel's time, the effects of detrital deposition 
can be discerned as changes in the shapes of islets such as Nake. 
Danger, and Arundel (compare Figs. 2 and 4 1. 

Such change is typical of atoll evolution. Geologically. 
Caroline is a few steps behind one of its "neighbor" Line 
Islands, Christmas, where sediments and coral growth have 
converted the original lagoon into a maze of supersaline. 
minilagoons and tiny islets, mostly cut off from the sea. Further 
steps in evolution are exemplified by completely filled-in atolls 
such as Jarvis and Vostok, where not even salty pools remain. 

Tridacna- Acropora Reef: Though the giant clam 
(Tridacna maxima) is an abundant component of Caroline's 
lagoonside reefs, exceptional aggregations flank the most 
southerly windward motus (Brothers through Tridacna. 
Figs. 44 to 48 ). Two especially outstanding reefs extend across 
the lagoon from Tridacna to Ana-Ana (Fig. 10: PI. 26) and 
Tridacna to Kimoa (Fig. 48). where Tridacna are attached to 
Acropora spp. corals, a favored substrate (Braley. 1987). 
Abundant inshore Tridacna on all these islets suggest that their 
density is similar to that on the main reef: up to 20/. 25 nr 
(i.e.. 80/m 2 ), averaging 35/nv for the entire area surveyed 
(Sirenko& Koltun. Subchapter 1. 4. this volume). This exceeds 
the highest densities known: up to60/nv at Reao Atoll. Tuamotu 
Archipelago ( Richard. 1985). Densities of 6-20/m 2 , at Takapoto 
Atoll (Tuamotus) are considered high. Throughout Caroline, 
the clams averaged 18 x lOcminsize. Several species of giant 
clams have suffered greatly in the Indo-Paeific from poaching 
and ov erharv esling: few undisturbed populations exist ( Braley . 
1987). Caroline is thus a special refuge for T. maxima. 

Lagoon Reel Flats : These varv considerably but are 
narrower and more gently sloping than the seaward reef flats. 
They are typically covered with fine coral gravel and coarse 
sand In sheltered areas dower Long. Windward. Crescent. 
South, upper end of lagoon), lush shrubbery — Cordia, 
Toiirnefortia, Pisonia, Cocos — overhangs the lagoon. Here 
fine silt. sand, and/or an algal slime are common (PI. 28). In 
l l )88. narrow, sandy beaches occurred only on the north shore 
of South (PI. 24) and east side of Shark (PI. 29). but in 1990, 
sand was more common throughout Caroline. 

Where the lagoon shorelines arc less sheltered and 
vegetation does not overhang the lagoon, unvegetated rubble 
and sparse herb mats are typical. Here, lagoonside rubble 
av erages less than 2 m wide ( PI. 3 1 ). This contrasts with their 
seaw aid reef Hats, w Inch av erage 2 1 m w ide (PI. 12). 

Lagoon Reef Fauna: A Brief Summary : Caroline's 
marine environment is rich vet essentially undocumented; 
know ledge of its ecosystems is limited to small, outdated lists 



12 



offish and invertebrates (Dixon, 1884; Fowler, 1901; Pilsbry 
&Vanatta. 1905a.b). However, preliminary studies on mollusks, 
benthic invertebrates, and lagoon plankton were begun by 
Soviet scientists in 1988 (Tsyban & Smith, 1988; Sirenko & 
Koltun. Subchapter 1.4. this volume). The lagoon and reefs are 
remarkably pristine, having changed little since their first 
discovery; all early travelers remarked on their beauty, 
abundance, and variety (Bennett, 1840; Markham, 1904). 

The usual assemblage of tropical invertebrates — 
echinoderms. mollusks, crustaceans, porifera, corals, tunicates, 
et cetera — are present. Large numbers of black sea cucumbers 
(PI. 10). about 20 cm long, are particularly abundant on the 
lagoon reef flats of the southern windward islands. They have 
been tentatively identified as Ludwigothuria sp. (B. Sirenko, 
personal communication). Conspicuous fish families include 
paiTot fish (Scaridae), butterfly fish (Chaetodontidae), 
surgeonfish (Acanthuridae), damselfish (Pomacentridae), 
pufferfish (Tetraodontidae). and wrasses (Labridae). 

Substrata 

Throughout the atoll, the substrata reflect a coralline 
origin. There is little "soil" in the accepted sense. Various 
grades of jagged, eroded coral and mollusean rubble (from fist- 
sized to tiny pebbles), together with sand, coralline algae, and 
small proportions of organic litter, humus, and guano, are 
present. Such accumulations of reef and terrestrial debris are 
similar to those of other low, coral atolls ( Fosberg, 1953; 
Stone. 1953: Wiens, 1962; Niering. 1963; Stoddart & Sachet. 
1969; Reese, 1987; Garnett, 1983). 

Generally speaking, atoll soils are calcareous and extremely 
immature, a consequence of their limited age and frequent 
disturbance by storms. Barely modified beyond the reef that 
spawned their presence, they are composed primarily of calcium 
and magnesium carbonates. Water retention, if any, is due to 
accumulated organic matter and its associated chemical changes. 
This is particularly important with respect to guano, which 
reacts with coral sand and humus to form phosphatic hardpan 
and nitrogen-rich "soils" (Fosberg. 1953). 

Reese (1987) categorizes atoll "soils" into five types, all of 
which occur, in different proportions, at Caroline. The degree 
of organic matter, decomposition, amount of humus, and the 
depth of the "soil" strata are directly correlated with age and 
size of the motus. 

/. Accumulations of coral rubble, mainly of stone size. 
These youngest of "soils" are most evident around the edges of 
the motus. acting as a substrate for pioneer herb mats. Often 
extending well inland, they can support surprisingly lush 
Tournefortia scrub. 

2. Unaltered coral sand and gravel. Although exposed 
sand was uncommon at Caroline in 1 988, this substrate occurred 
intertidally where the lagoon was filling in and on actively 
growing sandbars, primarily in the upper lagoon (Pis. 23,28), 
northeast and northwest South (Pis. 24,32), and the lagoon ward 
edge of Shark (PI. 29). In 1990, sand was more ubiquitous on 
Caroline as a result of the deposition of tons of sand during the 
severe February storm. 

3. Soils with a weakly developed A-horizon, with color 
i mly slightly darker than the unaltered sand below, but with no 



evidence of structural development. Especially evident in 
1988 within the ancient interislet channels that compose Long 
Island (PI. 33), much of this substrate is now storm-eroded and 
overlain by fresh sand. 

4. Soils with a more developed A-horizon. deeper and 
darker than above, with some structural development. This 
stage defines areas where the rubbly/sandy substrate approaches 
a true, but poor, "soil." As such, it represents older, more stable 
parts of each island. It is common within the islet interiors 
where Pisonia is (or was) present. Humus and guano fill the 
gaps within the irregular shapes of eroded coral. Its composition 
may be likened to a coarse mixture of gravel, sand, bones, and 
shells, all mixed with sparse amounts of partly-decomposed 
litter. Land crabs are particularly numerous, helping break 
down organic matter into finer particles. 

5. Soils with an accumulation of raw humus on the surface 
and with a relatively deep A-horizon. During this stage 
phosphatic hardpan may develop. These true soils, though 
somewhat depleted by guano diggers, cover significant areas 
on South and Nake. Cocos and/or Pisonia debris adds greatly 
to their dark color and moisture content. This earthy substrate 
is composed primarily of rotting Cocos fronds and fibers that 
have been shredded by coconut crabs. Patches of blackish 
muck on South Island support local patches of Polynesian 
arrowroot (Tacca leontopetaloides). 

On Caroline, hardpan (PI. 76) was present in several areas 
(primarily South, Nake, Long, Emerald, and Mannikiba), 
supporting herb mats and Tournefortia scrub. This substrate 
may be likened to an old asphalt road. 

Caroline provides an excellent example of the progression 
of soil development through islets of different age and size 
classes (see Ecological Succession section). From a wave- 
washed mound of coral rubble, barely above sea level (Fig. 5), 
the substrate gradually improves in texture and fertility as the 
emerging islet ages and organic matter accumulates. Pioneer 
plants are hardy, salt-tolerant, low-lying mats consisting 
primarily of Heliotropium, and later, Tournefortia. Increasing 
numbers of shrubs provide shade and branches for nesting 
seabirds. Larger trees (Pisonia. Cordia, Morinda) add more 
shade and thereby increase humidity, as well as provide 
opportunities for additional organic "fallout": leaves and bird 
remains (nests, eggs, chicks, droppings, regurgitated food, 
dead adults). 

Each stage of substrate development accelerates the 
accumulation of organic material and helps to define an 
emerging, deeper A-horizon. Soil maturity is indicated by 
more organic matter, improved soil texture, and a lowered 
volume of coralline and mollusean debris. Caroline's soils 
barely exceed several centimeters in depth and are always 
intermingled with coral fragments. As a result, they are 
unsuitable for burrowing seabirds, such as petrels and 
shearwaters. 

Hydrology 

Hydrological information is essentially lacking. No 
standing fresh water exists. The quality, extent, and salinity of 
the freshwater lenses, as well as their variability according to 
tide, season, and rainfall, are unknown. At the time of Caroline's 



13 



discovery" ( 1606), de Quiros and his party were desperate for 
fresh water. After noting how lush and green Caroline was, 
they expected to find good water supplies, but there was 
"nothing but salt water in the holes they dug" (Markham, 
1904). Maude ( 1968) suggested, in retrospect, that had they 
waited longer the salt water in their shallow wells might have 
run fresh, as has been his experience on some other atolls. 
During the 19th century, three wells were used — one on Nake 
and two on South (Holden & Qualtrough, 1884). One South 
Island well contained fresh water at 1.5 m depth in 1974 
(Garnett. 1983). We saw no wells but located concrete cisterns, 
one built near the northwest point of South in 1937 and rebuilt 
by the Falconers in 1989. and another uncovered one (dating 
from 1938) within a Cocos—Pisonia grove along Tr. 2, about 
200 in east of the southwest corner of Nake. 

Caroline's paucity of fresh water may be partly responsible 
for the lack of a permanent population. The annual rainfall in 
1989 (Appendix 2) was 1,242 mm (48.9"). However, like the 
similarly lush Nikumaroro and Orona (Phoenix Islands), 
Caroline's rainfall may vary greatly from yearto year, resulting 
in undependable water supplies. In the past, residents relied on 
rainfall catchment for fresh water (Maude, ca. 1938; R. Falconer, 
personal communication). 

Shallow sources of fresh or brackish water may be present 
on most islets, as Pisonia forests occur on 14% of them, but 
very little is known of freshwater lenses supporting Pisonia 
forests (Wiens, 1962). If we assume that Pisonia is not salt- 
tolerant, small water lenses may be present on motus as small 
as 0.2 ha. This is further discussed in the Ecological Succession 
section. 

Climate 

Meteorological records for Caroline were sparse until 
1989, when Ron Falconer began daily records of rainfall and 
wind direction (Appendix 2). Some data is available from the 
plantation years 1916-1920 (Young, ca. 1922) and during the 
1883 Solar Eclipse Expedition (20 April-8 May) (Upton. 
1884), when 203.2 mm (8") fell. The best generalizations on 
weather conditions in this area are found in the Geographical 
Handbook Series: Pacific Islands, Volume 2 (N.I.D.. 1943). 
Wiens (1962), Seelye (1950), Taylor (1973), and various 
papers on the Tuamotus (Stoddart & Sachet, 1969; Sachet, 
1983). Islands in the Line Group experience a wide range of 
climates. In general, those near the equator are dry, with 
rainfall increasing with increasing latitude north or south. 

Caroline experiences a tropical oceanic climate that varies 
little during the year. Temperatures are uniformly warm to hot, 
normally tempered by trade winds from the southeast to 
northeast. Falconer (personal communication) has recorded an 
annual average of 30°C (86°F) (range 26-3PC [78-88°F|). 
Mean annual temperatures for the Central Equatorial Islands 
lie between 24 and 29 C (75-85°F). Surface temperatures 
increase rapidly in early morning and remain hot throughout 
the day. forest interiors are humid. The daily range of 
temperatures exceeds the annual fluctuation in the daily mean. 

Atmospheric pressure, sunshine, and cloud cover are 
probably similar lo the northern Tuamotus — uniform except 



during storms. During our 1 988 surveys the sky and air were 
extremely clear, but we experienced much heavy cloud and 
rain in March 1 990, the aftermath of cyclones "Peni" (centered 
near Vostok) and "Ofu" (centered further west). 

Wind and Rainfall : Caroline is dominated by trade winds. 
As on all low atolls, land topography has no appreciable effect 
on weather. Although it lies within an area primarily influenced 
by southeast trades, there is a small annual oscillatory movement 
northward and southward, so that in reality winds blow from 
the east, northeast, and southeast. This accounts for the east 
and northeast winds that puzzled the eclipse party, who were 
expecting winds from the southeast (Upton. 1884). Data from 
Falconer (Appendix 2) indicates that, at least for 1989 and 
1990, winds blow primarily from the north and northeast, and 
rarely from the southeast (April-August). 

The atoll lies within a belt of variable rainfall, along with 
Vostok, Flint, and the northern Tuamotus. Young (ca. 1922, 
p. 13) notes that Caroline's rainfall is "certainly less than that 
of Flint," giving exact figures for 1919 (2,172 mm) and 1920 
( 1.854 mm) and estimates that there was "probably not more 
than 50" (1270 mm)" during 1916. 1917. and 1919. These 
estimates were based on exact figures from Flint ( 1 .600. 1 .346, 
1,295 mm, respectively). Falconer measured 1.242.1 mm 
(48.9") in 1989 and 2,209.8 mm (87") in 1990. An unusually 
stormy February in 1990 brought 640 mm (25.2") of rain. 
Rainfall distribution isohyets (Taylor. 1973) assign Caroline 
an annual precipitation of approximately 1,500 mm (60"), a 
perfect average ofthe above 6 years (x= 1,513mm). Ingeneral, 
"winter" (May-October) corresponds roughly to a dry season 
and "summer" (November-April) to a wet season. 

Hurricanes and Tsunamis : Atoll islets are active structures, 
undergoing repeated death and rebirth. Violent storms contribute 
to ongoing erosional and rebuilding processes. Storms deposit 
debris not only along the shores of the windward islets 
(Pis. 17,20) and across reef flats into the lagoon but sweeps it 
far inland. 

Although the south-central Pacific is relatively free of 
cyclonic storms (cyclones, typhoons, hurricanes), they do 
occur with enough frequency and devastating force that any 
discussion on climate should include them. Although detailed 
records of hurricanes and tropical storms exist for the inhabited 
Tuamotus since European discovery, many of these may not 
have affected Caroline. However, the following evidence 
suggests that Caroline experienced two major hurricanes last 
century and that periodic violent storms can modify the atoll 
substantially: 

/. Between IH22 and 1825. When de Quiros visited 
Caroline in 1 606. the northwesterly Cocos plantation on South 
Island was healthy. When Bennett arrived in 1834. he noted 
that all the palms were "of dwarf stature." and that "amidst the 
original groves, the number of vigorous seedlings fully 
confirmed Captain Stavers' statement [who had visited the 
atoll in 1828] that these palms had increased greatlv since his 
last visit to the spot" (Bennett. 1840). 

A few years before 1 828. therefore, something had affected 
the palms. By 1834 they were all of an even height and quite 
short, vet bore nuts. French records indicate that two devastating 



14 



storms whipped through the Tuamotus during this time — in 
1 822 and 1 825 (Sachet, 1 983 ). At least one of these could have 
affected Caroline. 

2. The 1878 cyclone. The first unambiguous record of 
major devastation at Caroline comes from the letter of a certain 
J. M. Salmon, dated 1883 and reproduced in Holden (1884). 
Speaking of the time when Messrs. Brown and Brothers took 
possession of Caroline (somewhere between 1865 and 1872). 
he stated that "it seemed as if there had been a storm or 
hurricane at some short period previous, which had desolated 
the place." Arundel ( 1 890) attributed this to a tidal wave that 
swept across the Pacific from South America to New Zealand 
and Australia in 1868 (Arundel, 1890). but atolls do not 
generally suffer greatly from tsunamis because they lack 
focusing relief. Hydrographer of the Navy (1931. Vol. Ill, 
p. 154), however, referring to Caroline, clearly states that in 
" 1 878 a cyclone passed over the islands, destroying most of the 
coconut trees." 

The Great Britain Naval Intelligence Division (N.I.D.. 
1 943. p. 490). in reference to Caroline, also states that "in 1 878 
a hurricane wrought g-eat destruction." This was possibly the 
storm of 6-7 February 1878. an extremely violent one which 
killed I 17 persons on Kaukura Atoll. 750 km southeast of 
Caroline in the Tuamotus (Sachet. 1983). 

3. The 1990 storms. We know from Arundel's chart 
( Fig. 4) that no major islet-altering storm has hit Caroline Atoll 
since 1883. However. oursecondvisittoCarolinewas2 weeks 
after cyclone "Peni," centered near Vostok (February 1990). 
affected the atoll. Violent winds, torrential rain, and high seas 
had uprooted vegetation in some windward areas and greatly 
altered Caroline's shorelines, interislet channels, tidal fans. 
and incipient islets from our 1988 visit. Sand and rubble had 
been rearranged on both windward and leeward islets, Motu 
Atibu virtually disappeared, and the main interislet channel 
that divides Long Island had lost its herb mats and many 
Tournefortia shrubs, becoming smothered with fresh sand. 

Because islets on coral atolls rarely exceed 5 m in elevation, 
the tidal surges associated with Class IV orClass V hurricanes. 
often exceeding 5 m in depth, can overwhelm them, not only 
altering or destroying the vegetation, but in extreme cases 
completely removing them from the coral rim (Frisbie, 1944). 
It is essential to consider the ephemeral nature of Caroline's 
islets in the discussions that follow. 

Sea Conditions : Because the most extensive coral rubble 
deposits occur around northern Nake and southern South 
Island, and because the Cocos plantation of northwest South 
was so badly hit by storms last century, the following Tuamotuan 
generalities (Newell. 1956) probably also apply to Caroline: 
/. prevailing trade winds from the east give heavy seas on 
the northeast or windward side; 

2. southern ocean swells generated in the sub-Antarctic 
break heavily on the south or seaward side; and 

3. occasional hurricanes or tropical storms strike in the 
northwest or stormward quarter. 



Vegetation: Vascular Plants and Floristics 

Botanical History 

All early visitors to Caroline described a well wooded atoll 
with numerous islets whose vegetation extended to the shoreline. 
It has changed little in the 384 years since its Western discovery. 
The first botanical collection and notes were those of Bennett 
in 1835 (Bennett. 1840), who recorded 10 flowering plants and 
a fern and planted Tahitian chestnut, sweet potato, and 
Polynesian arrowroot. The location of his plant collection, if 
it still exists, is unknown (Clapp & Sibley. 1971a). Evidently 
only a single nonnative species (Cocos), surviving as two small 
groves, persisted until the late 19th century. Beginning in 
1885. coconuts were planted extensively on South Island and 
south Nake. but the copra industry failed twice, and from 1929 
to 1987 the atoll was essentially uninhabited. 

Dixon made the first true botanical collection in 1883 
during the Solar Eclipse Expedition (in Trelease, 1884). All 
specimens were from South Island except Laponea ruderalis. 
His collection included several ornamentals and vegetables 
that have not been reported since, an important point as these 
temporary introductions have since been cited in the literature 
as part of Caroline's 35 plant species. Many were not found by 
the POBSP party, yet because no scientific investigations had 
been conducted for 80 years, they were counted as part of the 
atoll flora (Clapp & Sibley. 1971a). Three more visits to 
Caroline, plus periodic searching by the Falconers, have also 
failed to uncover most of these ornamentals. Since Caroline's 
occasional occupants tended gardens (Lucett, 1851), it is 
evident that many introduced plants have died out, lacking 
constant care. 

Vascular Plants of Caroline Atoll 

Plant Collections : To avoid duplicating Long's plant 
collection (Clapp & Sibley. 1971a). we collected only 
5 specimens in 1988 and 33 in 1990. Dr. D. Herbst assisted 
with identification, prepared and deposited the specimens with 
Long's in the Bernice P. Bishop Museum. Honolulu. Hawaii, 
with duplicates in the U.S. National Museum. Washington. 
D.C. Collection numbers preceded by 'K' were collected by 
A. Kay Kepler; those preceded by 'L' are those of the late 
C. R. Long. Earlier collections of Bennett in 1835 (Bennett. 
1 840) and Dixon in 1 883 ( in Trelease. 1 884) are noted by date 
only. 

Working with Long's location records for some species 
has proven difficult. He was working with an incorrect map 
(Fig. 7). which showed only 25 islets instead of 39. Much of 
his work was done at night, which in some places would have 
made it hard for him to determine his exact location. His 
references to South, Long, and Nake are undoubtedly correct, 
and presumably the following: "second islet south of Long" = 
Crescent; "islet northeast of South Island" = Tridacna: and 
"fourth islet north of Bird Islet" = Emerald. Long records 
Pandanus on the "second islet south of Nake Island." which 



15 



lacked Pandanus when we surveyed the island. Moreover, the 
first islet south of Nake supports an extensive grove of large 
Pandanus trees on its eastern (lagoon) shore, and we feel 
confident in ascribing Long's specimen to this island, which 
we had named "Pandanus" because of this grove. To be 
consistent, we have ascribed all his other "second islet" 
specimens to Pandanus Islet as well and assume he made no 
collections on the actual second islet (Danger). 

Species Lists. Annotated Checklist, and Maps of Terrestrial 
Vascular Plants : Following recent authors Sachet & Fosberg, 
19S3;Lamberson, 1987), we do not considerCaroline's transient 
or extinct vascular flora (Table 1) or the vegetables and 
ornamentals in the Falconer's garden as part of Caroline's 
viable flora. Table 2 summarizes the current flora, detailing the 
relative abundance of each species within each plant community . 
These tables are based on sight records supplemented by all 
collections, past and present. No beach drift seeds are known 
from Caroline apart from those species already represented. 
English and Gilbertese names in Tables 1 and 2 are from 
Thaman ( 1987), St. John (1973). and Perry & Garnett (n.d.). 
If no common name is available, the Hawaiian name, familiar 
to many students of Pacific botany, is used. 

One new species record for the atoll, as yet unidentified, is 
called Species A (K-90-23. 24): a single sterile shrub 2.5 in 
high, collected from southwest Motu Mannikiba in coarse 
strand rubble, is similarto Clerodendrum inerme, with leathery 
leaves and arching stems 4-5 m long. It was found by John 
Phillips. 

Table 3 lists the distribution and abundance of plant 
species ( with subdivisions into tree, shrub, and herb components) 
onallmotus. Figures 1 1-25 map the entire atoll distribution of 
each species according to data from transects and aerial maps. 
Families are arranged phylogenetically. according to 
losherg & Sachet ( 1 987 ). with species arranged alphabetically 
within each family. The taxonomy of vascular plants follows 
W. Wagner etal. ( 1990). and ferns follow H. Wagner (personal 
communication). "'< cover" means the percentage of the 
ground area covered by a particular plant species. In all text and 
tables, the following symbols apply: 
New record for Caroline 

Indigenous — plants native to Caroline but also occurring 
elsewhere (I) 

Aboriginal introduction — useful plants brought by 

Polynesians in pre-historical times (AI) 

# Recent introduction — plants of accidental or deliberate 

introduction alter Western discovery of the atoll (Rl) 

A Abundant — generally the major or dominant species in 

a given area 
VC Very common — often seen but not quite as abundantly 

as above 
C Common — generally distributed throughout a given 

area in large numbers 
UC Uncommon — observed uncommonly but more than 10 

times in a given area 
Occasional — here and there, often widely scattered but 

not forming a major component of the vegetation 
R Rare — observed 2 10 times in a given area 
S Single — only one specimen observed 



L Local — found only or principally in one or more restricted 

areas 
D Drift seedling — plant derived from a water-borne seed 
+ Not seen 1988-1990 but probably still present 

PSILOTACEAE 

* Psilotum nudum (L.) Beauv (Fig. 1 1 ) 

Formerly Known Distribution : L-3233 from Nake 
Present Distribution : Cosmopolitan, common on remote 
islands, rare on Caroline. K-90-15 from South. In 1965. 
common on wet base of Cocos only on Nake Island. In 1988 
and 1990. a few clumps found similarly on South Island in 
shady, damp locations, close to lagoon, northwest sector. 
Cocos canopy was 18 m. 

POLYPODIACEAE 

* Phymatosorus scolopendria (Burm. /) Pichi-Sermolli 
(Fig. 12: PI. 34) 

Phymatodes scolopendria (Burm./) Ching 

Polypodium phymatodes L. 

Polypodium scolopendrium Burm./. 

Microsorium scolopendria (Burm.) Pichi-Sermolli 
Formerly Known Distribution : Recorded 1840. collected 
1884; L-3244, L-3250, L-3287 from Nake, Long, and South 
Islands. 

Present Distribution : Range extension from 3 to 1 1 motus. 
Rarely a continuous ground cover, usually locally rare to 
abundant. Commonest on Nake. with cover 10-80%. Well 
represented on South, especially in open areas of the interior, 
where soils are moister. On other motus local distribution 
varied from less than 1 to 80%; accurate mapping is difficult. 
Absent from motus less than 0.6 ha in size, where habitats 
cannot provide cover, moisture, and substrate for both 
sporophyte and gametophyte generations. 

Ecology : Hardy. Leaves burn in sun but can withstand 
very dry conditions. Primarily in Toumefortia scrub, mixed 
forests with Pisonia and Pandanus. or Cocos plantations. 
Associated with Cordia, Morinda, Suriana. In open clearings 
within dying Cocos forests, occurs in dense mats intermingled 
with Boerhavia, Ipomoea, and Portulaca. Sometimes gathers 
in thick bands at the interface of Toumefortia and Pisonia 
forests. Prefers shelter, high humidity, "soil." and relative lack 
of wind, but absent from deeply shaded forests. Rhizomes 
never exposed on ground surface or epiphytic on trunks, as in 
wetter islands such as Hawaii or Samoa (personal observation) 
or m the moister Line Islands (Wester. 1985). indicating that 
Caroline's habitats are suboptimal. Although most ferns are 
not halophytic, tins species grew (rather stunted) in l l >NN 
amongst sparse herb mats ( 1% cover) on older beach sands of 
an ancient reef channel on Long Island (Tr. C). where rainfall 
pi ov ides the sole fresh water, but was ( temporarily?) obliterated 
in February 1990. Rare to uncommon in outer beach strand, 
and beach scrub w ith Suriana on South, Arundel, and Shark. 
Substrata: Dry coral rubble, sand and gravel, rubble with 
sparse humus, lagoon mud, relatively fertile humus, older 
beach sands. 



PANDANACEAE 

* **? Pandanus tectorius Park. (Fig. 11; Pis. 35-38,50) 

Formerly Known Distribution : Recorded 1840, 
unidentified Pandanus; L-3227, Pandanus Islet, seen on Nake 
by Long. 

Present Distribution : A minor plant community ( Plant and 
Communities section), Pandanus is primarily associated with 
Tournefortia or Pisonia on the leeward motus. Most common 
on Nake. with Cocos and Ipomoea. Range extension from two 
to seven motus. 

Phenology: Flowers and fruit in October, March, and 
May. 

Substrata : Variable. Prefers lagoon mud, pure sand, and 
rubble-humus, but survives in almost pure rubble. 

GRAMINAE 

* +? Digitaria species 

Collected 1883 and recorded as IPanicum (Digitaria) 
marginata. Examined by Long, who believes it a Digitaria 
identical to his L-3235. Not found by the authors. 

* Lepturus repens (Forst. f.) R. Br. (Fig. 13; PI. 2a) 

Formerly Known Distribution : Collected 1883; L-3211, 
322 1 . 3236, 3238, 3247, 3259, 3286 from Windward, Tridacna. 
Nake, Long, Emerald, Crab, and South Islands, respectively. 

Present Distribution : On most dry Pacific atolls. K-88-4, 
5 ; K-90- 1 , 2, 1 9 to 2 1, 25 from South, Tridacna, and Ana- Ana. 
On Caroline, range extension from 6 to 26 motus. 

Ecology : Patchy, rare to locally common. Usually in 
exposed herb mats with Heliotropium, Laportea, Portulaca, 
and low Tournefortia scrub. Abundance 1-5% cover where 
not in thick patches. Occasionally inland under Tournefortia, 
Cordia, or Cocos, fitting the generalization that Lepturus, 
though a pioneer, will often persist as undergrowth in forests. 
Tufts tiny (few centimeters), dry and scrappy in exposed areas, 
but to 3 dm where shaded. Never in tall, upright clumps, or with 
the same abundance as on the drier, filled-in equatorial atolls or 
islands with sandier habitats (Christopherson. 1927; Fosberg, 
1953, personal observation). Never forms a turf. 

Substrata : Able to survive in coral rubble of varying 
coarseness, down to high water mark, but preferred habitat is 
part sand. L-3286 was from "numerous clumps under Suriana 
scrub on South Island," perhaps the low, sandy portion of the 
northwest point ( PI. 45 ). our best Lepturus site. Comparison of 
Arundel's chart (1883), recent aerial surveys, and earlier 
photographs indicate that several motus have altered shape 
since 1883. The amount of open area on South Island has also 
decreased markedly since 1883. The distribution of Lepturus 
parallels these changes; there is clearly much less on South 
Island, and more in newly-created islet fringes. 

Since 1965 the lagoon shore of South Island has become 
overgrown by Cocos, so much that both Suriana and Lepturus 
are much less common than previously (Pis. 39,40). However, 
sand and debris will always be shifting, so that Lepturus will 
move from place to place, establishing wherever conditions 
permit. In the second situation, a comparison of Pis. 2a and 24 



from 1 883 and 1988, respectively, shows that a century ago the 
lagoon-facing shores of South Island were far more open than 
the dense Cocos plantations of today. The clumped grass in the 
foreground of PI. 2a is undoubtedly Lepturus, probably mixed 
with introduced grasses not seen since that time (Eleusine 
indica, Eragrostis plumosa) and the dubious Digitaria sp., 
above. 

PALMAE 

**# Cocos nuciferaL. (Figs. 14,36,51; Pis. 2,6,18,24,34,37,44) 

Formerly Known Distribution : Recorded 1840, 1884; 
L-3285 from South Island, extensive groves on South and 
Nake, scattered on north portion of Long. 

Present Distribution ; Range extension from 3 to 1 5 motus. 
Planted groves on South, Nake, and Long; the rest derived from 
drift. 

Phenology: Flowers and fruit year round. 

Ecology : Forms a major vegetation type (Plant 
Communities section). Primarily South and Nake, where 
closed canopy forests average 2 1 m high. 

TACCACEAE 

# **? Tacca leontopetaloides (L.) O. Kuntze (Fig. 1 1 ) 

Tacca pinnatifida Forster 

Formerly Known Distribution : Normally an aboriginal 
introduction on Pacific islands, but on Caroline is first mentioned 
as planted in 1834 (Bennett. 1840); L-3213 and 3219, and 
K-90-7 and 90-16 from moist muck, South Island. L-3234, 
common under Cocos and numerous patches found in muck, 
south end, Nake. 

Present Distribution : Common in northwest South. None 
in flower; each plant had two to three leaves, possibly dying 
back. None found on Nake, despite searching the south end. 
Has large underground tubers, dies back, and though cultivated, 
still occurs spontaneously in Cocos groves on many atolls. 
Currently harvested by the Falconers. 

Ecology : Needs fine, moist soil and shade. Though its 
seeds float for months (Guppy, 1906), it will probably not 
become established on any other motu, due to the prevalence of 
rubbly substrates. 

Phenology : Flowers and fruit in March and May, dies 
back in October. 

URTICACEAE 

* Laportea ruderalis (Forst. Chew (Fig. 15) 

Fleurya ruderalis (Forst. f.) Gaud, ex Wedd 
Formerly Known Distribution : Reported 1840, collected 

1884. L-3215 common in shady areas South Island; 

L-3229 scattered on exposed coral and sand, west side Crescent 

Islet. L-3253 under shade of Cocos and Pisonia on north side 

of Long Island. 

Distribution and Abundance : K-88-3 South Island, Tr. 1, 

elevation 0.3 m, under old Cocos plantation, in humus and 

rubble. Range extension from 3 to 32 islets (Table 3). 

Commonest and most widespread ground cover, 



17 



patchily distributed. Rare to locally abundant, percentage 
cover from less than 1% in herb mats of tiny motus to 60% in 
tall Pisonia forest. Best represented on Nake, Long, Brothers, 
South, Pisonia, Eitei, and Mannikiba, where coverage exceeded 
50% in appropriate habitats. To 1.1 m tall on Kimoa. 

Ecology : Largest specimens found under Tournefortia. 
Pisonia, Cocos, or Pandanus. Tiny (1-2 cm) and tougher in 
sunny, exposed sites. Halophytic, pioneering in herb mats on 
islets less than 0.75 ha in size (e.g., Fishball). Optimum habitat 
is Tournefortia scrub, in sunny clearings, or belts behind beach 
scrub. Uncommon in Pisonia forest. Occurs in both windward 
and leeward sites, but in greater density leeward. Will persist 
through several stages of plant succession if given adequate 
shade. 

Phenology : Flowers and fruit in October. March, and 
May. 

Substrata : Primarily beach gravel or coarse rubble. Also 
rubble-sand mixtures; not lagoon silt. 

OLACACEAE 

* (#?) Ximenia americana L. (Fig. 1 1 ) 

Never previously collected. K-90-170 South Island. 
50-100 m north of cistern, elevation 0.3 m. 10-20 m from 
coastal Tournefortia fringe, within Cocos plantation. Collected 
by crew of the yacht Amanita and posted by Anne Falconer to 
AKK. 

Distribution and Abundance : Locally abundant in one 
location, about 50 bushes ( 3—4 m high, 2-3 m wide ) spread over 
about 100 m. Adjacent to indigenous scrub, on edge of Cocos 
plantation near old settlement. 

Phenology : Flowering in July 1990. 

AMARANTHACEAE 

* Achyranthes canescens R. Br. (Fig. 16, PI. 41 ) 

Never previously collected. K-88-1 South Island, 
Tr. 5, to 0.7 m, elevation 0.3 m, in Tournefortia fringe, coral 
rubble. 

Distribution and Abundance : Quite widespread, primarily 
in interior scrub and forest of 19 motus (Table 3), from tiny, 
barely vegetated Fishball (0.73 ha) to the largest. South 
(1 06 ha). Density variable: from less than 1 % in Tournefortia 
scrub to 50% local ground cover in mixed Pandanus forest. 
Primarily associated with Tournefortia. May be locally abundant 
in clearings in Pisonia forests, pure or mixed. Often in a zone 
dividing Tournefortia and Pisonia trees, especially on Pig, 
Brothers, and Nake. 

Ecology : Never in natural herb mats. Needs shade but 
requires some direct sun: rare in pure stands of Cocos and 
Pisonia. Prefers small, sunny openings in forest or scrub. 
Drought-resistent and probably partly halophytic. Dies back 
annually in the dry season and reappears with winter rains 
(Anne Falconer, personal communication). To 1.5 m tall. 
Little or no capacity for dispersal by sea. On other islands, 
seeds carried by birds, especially fruit pigeons (Guppy, 1906), 



but pigeons are absent from the Line Islands. Perhaps dispersed 
by the long-tailed cuckoo (Ellis et al., 1990). 

Phenology : Flowers and fruit present in October, March, 
and May. 

Substrata : Lushest growth in humus soils of forest interiors. 
Often grows in pure rubble. 

NYCTAGINACEAE 

* Boerhavia repens L. (Fig. 17, PI. 34) 

Boerhavia diffusa L. 

Boerhavia hirsuta: Sensu Bennett, 1840 

Boerhavia species: Dixon. 1884 

Formerly Known Distribution : Reported 1 840, collected 
1 884; L-32 1 0. 3324, 3239, 3225, 3252, 3262, 3289, 329 1 from 
Windward, Tridacna, Nake. Long, Emerald, and South, 
respectively. 

Present Distribution : Cosmopolitan, widespread in the 
Pacific. K-90-164and 165 from Ana- Ana. Range extension on 
Caroline from 6 to 33 motus (Table 3). 

Abundance : Present in every habitat, leeward and 
windward, ranging from less than 1 to 80% cover. Often in 
unpredictable patches. Best locations (>50% cover) on Nake, 
Long, Windward. Pig, Brothers, Arundel. Tridacna. South. 
Ana-Ana, Pisonia. and Pandanus Islets. 

Ecology : Mostly found beneath Tournefortia. either in 
pure scrub or mixed with Pisonia. Cordia. Morinda. Suriana, 
or Cocos. Not in deep Pisonia shade; rarely in herb mats. Thick 
ground cover in indigenous scrub ( Shark ) or within clearings in 
old Cocos-Ipomoea forest (South), where it mingles with 
Phymatosorus, reaching a high density (PI. 34) and large size 
(rooting at nodes, vines exceeded 1 mlong). BIRDS: Bristle- 
thighed curlews fed within the Boerhavia mat in old Cocos 
forests. South. Sticky seeds (.32 cm [one-eighth inch] long) 
found entangled in preened down and adhering to contour 
feathers of a juvenile great frigatebird (PI. 42). Species is 
customarily dispersed around large oceanic areas and within 
atolls by seabirds such as red-footed boobies (Guppy. 1906; 
Ridley. 1930). 

Phenology : Small mauve flowers and seeds present in 
October, March, and May. 

Substrata : Coral rubble with sand or humus, rarely pure 
beach rubble. Lushest growth in humus-and-guano-laden 
rubble clearings where Pisonia forest once grew. 

* Pisonia grandis R. Br. (Fig. 18; PI. 43) 

Formerly Known Distribution : Collected 1884; L-3280 
4 m tree, north shore. South. Small grove, north end. Long. 

Present Distribution : Cosmopolitan, pan-Pacific. Caroline 
range extension from 2 to 29 motus (Table 3). 

Abundance : A major plant community (see Plant 
Communities section). Caroline^ Pisonia forests, some of the 
last remaining groves in the Pacific, are of special conservation 
value. 

Substrata : Occupies, and contributes to. best soils on atoll: 
mixture of rubble, humus, and guano. 



18 



PORTULACACEAE 

* Portulaca lutea Solander ex Forster F. (Fig. 19; Pis. 34,38) 

Formerly Known Distribution : Reported 1840 and 1884; 
L-3233 and 3292, 3231 . 3237, 3255, 3257, from South, 
Pandanus. Nake. Long, and Emerald, respectively, in open 
coral, rubble, gravel, and exposed areas, to 1.5 dm high. 

Present Distribution : Range extension from 5 to 33 islets 
(Table 3). 

Abundance : Along with Heliotropium anomalum is a 
component of the plant community. Natural Herb Mat (see 
Plant Communities section). Widespread, predictable on coast 
and former reef channels but local inland. Covered from one 
to 60% of land area on almost every transect, windward and 
leeward, especially facing lagoon. Best areas are Long, Tr. 4 
(36-m wide meadow); South, north end of Tr. 6 (50 m wide): 
Brothers, lee, almost pure mat covering 20% ground (6 m 
wide); Kimoa. north side (8 m wide), 10 cm high; Eitei, north 
side, 5 cm high. 

Ecology : Primarily occurs along edges of motus in rubble 
mat and open Tournefortia scrub, averages 12 cm high. 
Prominent in sparsely vegetated areas, extending seaward to 
high tide level. Halophytic; highly tolerant of sun. A Hat mat 
in exposed areas but lusher inland, rising to 2 dm tall. Generally 
found with Heliotropium, Lepturus, Boerhavia, or Laportea 
but may form pure mats. Uncommon in Tournefortia scrub 
patchy in clearings within Pisonia forests up to 13 m high 
Exceptionally common in old Cocos groves with Boerhavia 
et cetera (PI. 34); otherwise rare or absent from closed canopy 
Cocos plantations. Pinker stems found in sunny sites. BIRDS 
Provides nesting cushion for masked booby, sooty tern, brown 
noddy. On noddy rock, brown noddies nest on a thick mat of 
pure Portulaca. Feeding location for shorebirds. 

Phenology: Flowers and fruit October, March, and May. 

Substrata : Coral rubble and gravel, fine to very coarse. 
Healthier on older sands and coral-humus. 

ZYGOPHYLLACEAE 

* Tribulus cistoides L. (Fig. 1 1 ) 

Formerly Known Distribution : Collected 1884. L-3245in 
open sandy area among Tournefortia shrubs. Long Island. Not 
seen elsewhere on atoll. 

Present Distribution : Not seen on our surveys, but present 
in 2 sites on west-central Long Island. K-90-161 (collected by 
Anne Falconer), probably from one of same sites as 1965 
collection. Flowers in March. 

SURIANACEAE 

* Suriana maritima L. (Fig. 20; Pis. 6,21,39,40,44) 

Formerly Known Distribution : Collected 1884. L-3220, 
shrub to 1.8 m, east edge of Tridacna Islet. 

Present Distribution : K-90-5, 6 from South Island. Range 
expansion from one to 9 motus (Table 3). 

Abundance : Occasional on Caroline. Forms a vegetation 
unit. Beach Scrub with Suriana ( see Plant Communities section). 



Phenology : Flowers in March and May. 

Substrata : Best sites in sand but also on coral rubble. 

EUPHORBIACEAE 

# Phyllanthus amarus Schum. and Thonn. (Fig. 1 1 ) 

Formerly Known Distribution : Collected 1884. L-3283, 
herb Phyllanthus niruri L. (Trelease, 1884) to 4 dm, common 
on north side of South Island. 

Present Distribution : K-90-10-13, herb, 2 small patches, 
South Island. Limited to a few square meters in the atoll's only 
weedy area, less than 10 m : in two small clearings by the 
recently-renovated cistern. South. A fairly common weed in 
the Society and Tuamotu Islands, therefore probably arrived 
with Tahitian copra-cutters and perhaps again within the last 
2 years. Caroline's only established "weed" (excluding 
Polynesian introductions such as Cocos). 

MALVACEAE 

o * ( ** ?) (#?) Hibiscus tiliaceus L. (Fig. 1 1 ) 

Never previously collected. K-90-8, 90-9 from South 
Island, northwest peninsula, in Cocos plantation near old 
settlement and "landing," in coral rubble and humus, 0.6 m in 
elevation. 

Present Distribution : Two or three large spreading trees in 
heavy Cocos shade, 1 m tall, with recumbent branches forming 
an impenetrable thicket. This species, culturally important to 
Polynesians, is either indigenous, an early Polynesian 
introduction, or an ornamental brought by 1 9th century settlers. 

° * (**'?) (#?) Thespesia populnea (L.) Soland. ex Correa 
(Fig. 1 1 ) 

Never previously collected. K-90-22, 154. 155 from 
South Island, in Cocos plantation and in lagoon strand, northwest 
peninsula, near "landing." 

Present Distribution : Two trees ( 10 m tall), one near the 
cistern, the other in a fringe of native vegetation bordering the 
lagoon. The history of this species is probably the same as 
Hibiscus tiliaceus (above). 

* SidafallaxWalp. (Fig. 1 1) 

Formerly Known Distribution : Collected by Dixon, 1884, 
who found one specimen. 

Present Distribution : Not seen for 106 years. K-90-156, 
157, 158 from South Island, at edge of cistern, north side. One 
clump located in a sunny clearing, recently enlarged by the 
Falconers. 

CONVULVULACEAE 

* Ipomoea macrantha R & S (Fig. 21; Pis. 34,37) 

Ipomoea tuba (Schlecht.) G. Don 
Formerly Known Distribution : L-3228 and 3293, 3242, 
325 1 on South, Nake, and Long, respectively. Trailing vines, 
white flowers, stems to 25 m long climbing over Tournefortia. 
Morinda. and Cocos. 



19 



History : Not collected last century, though plantation 
records indicate that it was a major reason for the abandonments 
of the coconut plantations: "The Pohue Vine |misidentifiedas 
Tuumfetta(=Triumfetta)procumbens], which is the worst pest 
on the island, was reported in 1921 to be under control" 
(Young, ca. 1922). Today it still causes severe damage to 
Cocos on South Island, strangling about two-thirds of the 
plantation (54 ha). 

Present Distribution : Range extension from three to seven 
motus, five Windward and two Southern Leeward Islets 
(Table 3). 

Abundance : Forms part of a vegetation subunit. Dying 
Cocos-lpomoea Forest (PI. 34, Plant Communities section). 
An indigenous, nonparasitic vine, becoming abundant and 
strangling in disturbed areas. Rampant growth over most of the 
interior of South Island, where it forms dense tangles up to 
25 m high. Less dense thickets on southern Nake drape 
Pandanus, Tournefortia, Morinda, and Cocos to 10 m. Our 
transects on Nake were not rerouted or abandoned, as on South. 
Coverage scant elsewhere, generally 2-5%, except in two 
Pisonia sites, where its coverage was 20% (Long Island. Tr. B: 
Windward Islet, Tr. 1 ). 

Ecology : Lush in dying Cocos forests and mixed forest 
with Pandanus, because of relatively fertile soils, moisture, 
humidity, and partly sunny clearings. Strangles all but the 
tallest Pisonia and Cordia. Typically sea-dispersed to atolls 
(seeds germinate after floating up to 1 year in seawater), crawls 
inland, progressively dropping seeds, to attain full size in 
interior forests (Guppy, 1906; Ridley, 1930). Seeds of /. pes- 
caprae are known to be ingested by white terns in the Marshall 
Islands, perhaps as gizzard stones (Fosberg, 1953). Possibly 
these same terns, abundant at Caroline, once aided the seed 
dispersal of /. macrantha. Also characteristic of Cocos 
plantations elsewhere in the Pacific (Fosberg. 1965: Stoddart 
& Sachet. 1969: Lamberson. 1987). 

Substrata : Prefers humus-laden rubble but can grow in 
coarse rubble and sand, especially in leeward areas. 

BORAGINACEAE 

* Cordia subcordata Lam. (Fig. 22: PI. 27) 

Formerly Known Distribution : Collected in 1X84. L-3213 
and 326 1 a, 3228, 3246. and 326 1 b on South. Pandanus. Long, 
anil Emerald, respectively; flowering trees to 4.5 m high in 
leeward coral rubble or along lagoon. 

Present Distribution : Africa to Polynesia. K-90-3 from 
South Island, lagoon edge. Range expansion on Caroline from 
5 to 23 motus (Table 3). 

Phenology : Peak flowering November through April, 
fruits collected in March and Max 

Abundance : A separate, though minor, plant community 
(Plant Communities section). Caroline's Cordia forests, 
typically small and mixed \\ ith other emergents. are some of 
the last remaining gun es in the Pacific and are thus of particular 
importance to conservation. 



* Heliotropiumanomalum H. & A. (Fig. 23: Pis. 17.33.45-47) 

Formerly Known Distribution : Recorded (as 
H. curassavicum) in 1840, collected in 1884. L-3222 and 3288. 
3240, 3248, 3256, 3288 on South, Danger. Long, and Emerald, 
respectively. 

Present Distribution : Pantropical. K-90-17 from Ana- 
Ana. In coral gravel, leeward, and windward shores. Range 
extension on Caroline from 4 to 34 motus (Table 3). 

Abundance : Forms part of a major vegetation unit. Natural 
Herb Mat (Plant Communities section), often associated with 
Laportea, Lepturus, or Boerhavia. Area coverage ranges from 
less than 1% to 50%. Widespread, predictable on wind- and 
salt-blown, low flats where vegetation does not overhang edge 
of motu. Also in ancient reef channels and newly evolving land 
connecting islets. Covers major areas of islets — that is. those 
less than 1.0 ha (e.g., Fishball. Skull, and Bo'sun Bird. Best 
developed on Skull. Tridacna. South. Emerald, and Mannikiba 
(50% coverage, western seaward rim). 

Ecology : Halophytic pioneer. Heights to 22 cm, averaging 
7 em. Thrives in heat and exposure. 

Phenology : Flowers and fruits year round. 

Substrata : Primarily coral rubble and rubbly sand. Marginal 
habitats extend down to high tide line in areas of coarse coral 
chunks, where it is tiny and leathery. 

* Tournefortia argentea L. (Fig. 24; Pis. 8.37.47.48) 

Messerschmidia argentea (L.f.) Johnston 

Formerly Known Distribution : Collected 1884. L-3216. 
3226, 324 1 , 3249, 3258 from South, Tridacna, Nake, Long, and 
Emerald Isle; shrub to 3 m high, edge of lagoon and above high 
tide, with white flowers, 

Present Distribution : Range extension from 5 to 38 motus 
(Table 3). Widespread in the Pacific, especially on small islets. 
Caroline's large tracts are excellent examples of relatively 
undisturbed, pure Tournefortia scrub and forest. 

Abundance : Dominates the atoll woodlands, forming the 
major \ egetation type ( Plant Communities section). On almost 
every motu ranging from a spattering of exposed shrubs within 
herb mats, through scrublands and taller forests to 14 m high. 

Ecology : Supports seven species of breeding seabirds; 
provides feeding habitats for reef herons (Egretta sacra). 
shorebirds, land crabs, and rats. 

Phenology : Flowers and fruits year round. 

Substrata : Pure coral clinker; mixtures of rubble, gravel, 
sand, and humus. 

BRASSICACEAE 

* Lepidium Indentation Montin (Fig. 1 1 ) 

Formerly Known Distribution : Reported in 1S25: "a boat 
load of pepper-grass and pursley" (Paulding. 1X31) and in 
1835. "a Lepidium of luxuriant growth" (Bennett. 1840). 
Collected by Dixon as /.. piscidium Forst in 1883. 

Present Distribution : Widely distributed throughout the 
North and South Pacific. K-90-169 and 171 (collected by 



20 



Alexandre Falconer), on Tridacna and Pisonia, most probably 
in coastal Tournefortia scrub. 

RUBIACEAE 

* Morinda citrifolia L. (Fig. 25; PI. 48) 

Formerly Known Distribution : Reported 1840, collected 
1884. L-32 14. 32 17 and 3282: 3232; 3254 on South, Nake, and 
Long, respectively. 

Present Distribution : K-90-4, 18 from South Island, 
lagoon edge, and Ana-Ana. respectively. Range extension on 
Caroline from 3 to 30 motus (Table 3). 

Abundance : Coverage from 2% to 50%. Basically an 
inland species, widespread and predictable in scrub and forest 
understory. Rarely a component of the canopy, except on 
Raurau. where Morinda grows 12 m tall in a 13 m Pisonia 
forest. Essentially associated with established Tournefortia 
woodlands on motus greater than one hectare in size. Quite 
common on South Island despite major disturbance, occurring 
within beach strand, Cocos plantations, and Cocos-Ipomoea 
interior. Best locations (40-60% coverage): Nake. Tr. 3; 
Tridacna, both transects; Long, Tr. 8; Raurau and Ana-Ana. 

Ecology : Appears early in plant succession: inToumefortia 
scrub as an early pioneer (Stage I), then from Stages II to IV, 
progressively becoming more common and robust. Not in pure 
Pisonia forest (Stage V). Much less common in Pandanus 
stands. Although it thrives best in light to heavy shade, 
preferably growing in moist "soil," one leathery seedling (7 cm 
high) had gained a foothold in exposed rubbly Heliotropium 
flats on Fishball Islet. 

Biogeographical Note : Generally considered a naturalized 
aboriginal introduction on most Pacific islands. Morinda could 
be native to Caroline, as theorized for the northern Line Islands 
(Wester, 1985). Although possibly introduced by early 
Tuamotuan settlers, its present distribution strongly suggests 
that it is indigenous. Throughout the atoll Morinda occurs in 
the greatest densities on motus with no anthropogenic forests 
or in areas distant from historical settlements (Fig. 26). On 
Nake, Morinda occurs frequentl — in places abundantly — within 
the interior Pisonia forests, yet its coverage is only 5-10% in 
mixed Pandanus— Cocos forests in the southern sector. It also 
appears to be part of natural biological succession (Table 6). 
Further support for this theory comes from nearby Flint. 
Though there is no direct archaeological evidence that Flint 
was settled in prehistoric times (Garnett, 1983), our 1990 
surveys found Morinda in all habitats (mixed woodland, native 
coastal scrub. Cocos plantations, and abandoned settlement). 

Originating in Polynesia, Morinda has been widely 
dispersed by man but has apparently also spread, unaided by 
man. "widely by sea in the Malayan and Polynesian Islands" 
(Ridley, 1930). Its air-filled, buoyant pyrenes can float for at 
least 53 days and "its seeds are almost certainly disseminated 
by birds and bats" (Guppy, 1906). It could also be disseminated 
by Coenobita crabs and rats within and between motus, as has 
been found elsewhere by Ridley. 

Phenology : Flowers and fruits year round. 



Substrata : Coral rubble, gravel, sand, and humus. Rarely 
found in coarse clinker. On larger motus. prefers moist soils 
under tall forests. 

GOODENIACEAE 

c * Scaevola sericea var. sericea Vahl (Fig. 1 1 ) 

Scaevola taccada var. sericea (Vahl) St. John 

Never previously collected. K-88-2, Windward Islet, 
central-windward side, elevation 0.3 m. 

Distribution and Abundance : One wind- and salt-sheared 
"hedge," found by K. Teeb'aki on Windward Islet, was growing 
on a coarse rubble beach. "The saltbush..., being recorded for 
the first time too from the island. ..covered approximately 3% 
of the islet's land area [this probably can be translated as "3% 
of the area covered at that location on the transect." as we 
understood from Mr. Teeb'aki's description that it was quite 
small], occupying the mid-windward side. The patch grew 
very low — only up to 2' high with its foliage forming an 
extended raised mat canopy all along the area it occupied" 
(Teeb'aki, 1988). We have been unable to return to this spot to 
observe and photograph it directly. 

Because Scaevola is hardy, halophytic, and widespread in 
the Pacific, it is surprising that it is so rare on Caroline. 
However, none occur on Vostok, and only one clump is known 
from Flint (Clapp & Sibley, 1971b; Garnett, 1983). Fosberg 
(1953) noted that Scaevola seeds are transported by bristle- 
thighed curlews (Numenius taitensis) in the Marshall Islands: 
curlews are common on Caroline ( Subchapter 1 .2, this volume) 
and could have brought seeds from elsewhere. 

Substrata : Coarse rubble, windward beach. 

° * Scaevola sericea var. tuamotensis (Si. John)Fosb. (Fig. 1 1 ) 
Scaevola taccada tuamotensis St. John 

Never previously collected. K-90-168 (collected by 
Alexandre Falconer), northeast peninsula. South Island, in 
coral rubble. 

Present Distribution : One individual, of unknown size. 
with Suriana and Heliotropium, northeast peninsula, South 
Island, facing the inner side of the "blind passage." 

Floristics and Ecology of the Motus 

Size of the Flora : Atoll floras characteristically lack 
diversity. Numbers of species range from 3 to around 150 in the 
Pacific and 284 in the Indian Ocean. The flora of the Southern 
Line Islands is particularly impoverished because of /. their 
easterly location ( far from the major source areas of Australasia); 

2. low profiles (most only rise a few meters above sea level); 

3. lack of topographic diversity (most have a very limited range 
of habitats); 4. low to medium rainfall (approximately 
1,500 mm p. a.); and 5. edaphic factors such as salinity, highly 
calcareous soils, et cetera. Long-distance dispersal and hardiness 
are important factors in establishing a flora, especially since the 
closest high island, Tahiti, is 830 km away, and the ultimate 
source of its flora, the Malayan-Melanesian region, is over 
8.000 km away. South America, the closest continent, is 



21 



approximately 9,000 km distant. The motus of Aitutaki, for 
example, at a similar latitude but further west and wetter, are 
considered depleted with 45 species. Fanning, at a similar 
longitude but wetter, has 123 species. Tarawa, 3,900 km to the 
northwest, receives a similar rainfall but supports 109 species. 

Where an atoll's potential flora is larger, the increased 
shade and greater protection from wind, salt spray, and storms 
result in a greater number of natural plant species on its larger 
motus. However, such atolls are generally inhabited and 
alterations by both aboriginal and modern man have modified 
their original flora. Caroline's isolation, variety of islet areas, 
and minimal human disturbance all contribute to its excellence 
for the study of atoll evolution. 

The number of species presently established on Caroline's 
39 motus is 27 (Tables 2, 3). The previous expedition in 
1 965 (Clapp & Sibley, 1971a) collected 20 species, of which 4 
were new to the atoll. Their total of 35 species, however, 
incorporating reports and collections from the 1800's, is 
misleading. Our total, 6 of which were new records, would 
have brought the atoll total to 44 (plus about 15 more 
unestablished, mostly garden, plants). However, following 
recent custom (see Vegetation section ), we have listed transient 
orextinct members of the flora separately (Table 1). To include 
them would obscure the near-pristine nature of the atoll and 
bias our analyses of species-area relationships. 

The 1 883 drawings of the South Island settlement, inhabited 
when most of Caroline's species were catalogued, shows that 
the island was vastly different (compare Pis. 2 and 24). A 
century ago homes were set amidst large grassy clearings; now 
the site is completely obliterated beneath shady 21-m-tall 
coconut palms. Nine exotic plant species have not been seen 
for over a century (Table 1 ). Evidently most ornamentals and 
domestic vegetables perished during uninhabited periods. The 
present residents struggle to keep garden plants alive because 
of poor soils, irregular rainfall, and foraging land crabs. A few 
native species might also have been eliminated during the 
guano and copra-harvesting years. 

Numbers of Indigenous Plants : A comparison of the 
percentage of indigenous species between different island 
groups (Table 4) shows that Caroline, with 85%* (N = 23) 
indigenous, is unusually high. Only 11 of 44 Pacific atolls 
reviewed have more than 75% of their species indigenous. Of 
these, nine ( including Caroline) are remote and lack permanent 
human occupation. 

The Tuamotu Islands ( 149° to 134°W) lie east and south 
of Caroline, yet they harbor considerably larger floras. Rainfall 
is similar. Three of them average 121 species (Table 4). 
averaging 42 indigenous species. When the variables rainfall 
and distance from a colonization source to the west are 
considered, the proximity of the Tuamotus to the diverse high 



* Note: Perhaps as high as 93% ; the Digitaria sp., if still extant, 
is of unknown identity and origin, and Species A has yet to be 
determined. 



islands of the Societies seems to play a major part in 
determiningtheir indigenous flora. A similar situation exists in 
the southern Cook Islands. Caroline and other remote Line and 
Phoenix Islands are sufficiently isolated from high volcanic 
islands that they exhibit a much simpler flora. Tahiti, the 
closest high island (830 km south), is in the wrong direction for 
direct currents, winds, or vagrant birds to bring seeds to 
Caroline. 

Composition of the Flora (Tables 2. 3 ): Caroline's botanical 
affinities lie with other southern Line Islands and the Tuamotus. 
Although the strand and inland floras consist of pan-Pacific or 
pantropical species, there are several widespread species and 
communities that are notably absent. Those that survive have 
withstood the atoll tests of time — poor soils, scarcity of fresh 
water, periodic inundation by salt water, intermittent cyclonic 
storms and hurricanes, harsh climate, and high seedling 
mortality. Caroline provides an excellent ecological laboratory 
in which floristic correlations with variations in habitat, motu 
size, and leeward/windward aspect may be studied. Fosberg 
(1985) and Sachet (1967) have noted the importance of such 
details in understanding the biogeography and taxonomy of 
Pacific plants. 

Caroline's present established flora includes only one 
weed species (Phyllanthus amarus), represented by a tiny 
patch less than 2 m : in size. There are two Polynesian 
introductions (Cocos, Tacca). Pandanus tectorius, Morinda 
citrifolia, Thespesia populnea, Ximenia americana, and 
Hibiscus tiliaceus, though indigenous, may have been 
introduced by Polynesians or 1 9th-century settlers. (See section 
on Vegetation.) 

Trees: Seven species present. Only three — Pacificwide 
natives — are widespread: Pisoniagrandis. Morinda citrifolia. 
and Cordia subcordata. Two are locally abundant: Cocos 
nucifera and Pandanus tectorius, while the rest, Thespesia 
populnea and Hibiscus tiliaceus. are rare and limited to the old 
settlement site. The absence of typical Pacific species such as 
Calophyllum inophyllum and Guettarda speciosa is notable, as 
they occur naturally on atolls such as Rangiroa. further east 
(Stoddart & Sachet. 1969). 

Shrubs: Five species present, at least four indigenous. 
Only Tournefortia argentea is abundant: its most abundant size 
class is under4 m. Scaevola and Suriana. tough and widespread 
elsewhere, are poorly represented on Caroline. It is noteworthy 
that two varieties of Scaevola sericea are present. Species A is 
represented by a single individual, Ximenia americana. by a 
single, large patch. Pemphis acidula. though common on atolls 
of similar latitude and climate, is absent from most of the Line 
and Phoenix groups (Stoddart & Gibbs, 1975; Fosberg & 
Sachet, no date). This may be due to the paucity of its preferred 
habitats: low rocky substrates (reef rock, conglomerate rock) 
and sand-gravel ridges. 

Herbs: Fifteen species present, at least 12 indigenous. Of 
these only seven are common: Heliotropium anomalum. 
Boerhavia re pens. Portulaca lutea, Laportea ruder alls, 
Achyranthes canescens, Lepturus repens, and Phymatosorus 



22 



scolopendria. Ipomoea macrantha and Tacca leontopetaloides 
are locally abundant, while Phyllanthus amarus, Tribulus 
cistoides, Lepidium bidentatum, and Psilotum nudum are rare 
and localized. Digitaria sp. may be extinct. The fact that Sida 
fallax has only been recorded twice in 106 years is curious. 

Ecological Succession 

We have attempted to trace the development of Caroline' s 
flora from the smallest to largest motus, using field data and 
aerial photos, which reveal past geological processes. Three 
tables provide this analysis of ecological succession: Table 5 
presents Caroline's motus in order of ascending size, together 
with the numbers of plant species and major plant communities. 
Since the atoll's total land area is small, our data provide 
relatively complete floristic lists for each islet and detailed 
maps of their plant communities (Figs. 37-57). The number of 
species varied from 3 growing on 4 tiny islets (0.02 ha each) to 
23 on South ( 104.41 ha). Because the total number of species 
for the entire atoll (27) is also small, the addition of one or two 
rare species contributes significantly to the total flora. Such 
additions must be kept in perspective when evaluating plant 
succession. 

Table 3 provides a summary of plant species distribution 
by islet in decreasing order of abundance, and Table 6 is a 
summary of plant species distribution and relative abundance 
with respect to islet area and the primary mode of seed dispersal. 

Basic Serai Stages 

Islets appear, grow, mature ecologically, or vanish in 
violent storms. Many interacting factors, including geographical 
(islet area, atoll shape, distance from high islands and continents), 
geological (changes in sea level), chemical (nitrates from bird 
droppings, leaf fall, et cetera), climatological (wind, droughts, 
storms, microclimates), and biological (seabirds, rats, land 
crabs, and man, both aboriginal and modern ), constantly interact 
to change conditions. The relative influence of some of these 
factors is evident when comparing the floras on motus of 
different sizes. 

Seed-dispersal mechanisms (Table 6) and the presence of 
underground fresh water are also vital. Unfortunately, the 
relationships between groundwater salinity, species distribution, 
and vegetation patterns on atolls are poorly understood ( Fosberg. 
1985). The presence and relative salinity of permanent water 
depends on Ghyben-Herzberg lenses of varying thickness on 
different islets, and this in turn depends upon island dimensions 
(especially width), soil porosity, rainfall, tidal fluctuation, and 
other hydrological factors. Though groundwater supplies have 
been studied on many atolls (Wiens, 1962; Maude, 1953), each 
island group is so unique that it is unwise to extrapolate 
information from one to the other. 

Caroline's 39 motus fall naturally into 4 size classes: 
motus with areas of a) <0.2 ha, b) 0.2 to 0.7 ha, c) 0.8 to 
25.0 ha, and d) >25.0 ha. These size groupings harbor all 5 of 
the serai stages identified on Enewetak Atoll (Lamberson, 
1987), tailored to reflect Caroline's particular geography, 
geology, and impoverished flora. Each stage may be the sole 
example of ecological succession on an islet or may occur as 



one of several stages. Typically the early stages cover the 
peripheral rubble and scrubby outer zones, while the later ones 
appear as a series of roughly concentric bands progressing 
inland. 

Stage 1 : Early pioneers on sandbars, spits, or small rubbly 
islets subject to storm damage and washover. Harsh conditions, 
intense sun, drying winds, salt spray. High salt concentration 
in the substrate. Lack of fresh water and nutrients. Plant genera 
present include Heliotropium, Portulaca. Lepturus, Boerhavia, 
and seedling oropenTo!/r/ie/o/-f/a scrub. No Cocos. This stage 
is found on many small motus (Noddy Rock, Fishball) and 
former interislet channels (e.g., Long Island). 

Stage II : Thick scrub of mixed genera, often impenetrable. 
Its protective barrier allows for the development of vegetation 
on the larger islets. Seabirds begin to contribute to the soil 
(guano, eggs, regurgitated fish, decaying nesting material). 
Plant genera include Toumefortia, Suriana, Cordia, and 
Laportea. If Cocos present, accompanied by coconut crabs. 
Very common around the periphery of most motus just inland 
of the native herb mats or flanking sheltered shores adjacent to 
the lagoon (South. Kota). 

Stage III: Trees larger, seabirds add further to soil fertility. 
Open grassland may develop in sunny clearings (Tridacna). 
Added plant communities and Cordia-Tournefortia, 
Tournefortia-Morinda, and Pisonia— Toumefortia forests. 
Occurs in the next inner concentric zone of vegetation to 
Stage II on larger motus (Nake, Long) or, more commonly, the 
entire interior of smaller ones (Pandanus, Southern Leeward 
Islets). 

Stage IV : Pisonia dominates the older mixed forest. 
Morinda and Toumefortia reach for the sun. Forests are more 
open. Undergrowth mostly a ground cover of Laportea, 
Boerhavia, Lepturus, and Portulaca. Covers the main portion 
of larger islets. If Cocos and Pandanus present, forms a mixed 
forest with vines (southern Nake, Shark). Coconut crabs 
common. A widespread stage in the center of most motus 
(Central Leeward, Windward Islets). 

Stage V : Pisonia takes over. Other trees are confined to 
the forest edges. Always in the deep interior of the larger islets. 
Little or no ground cover. Abundant nesting black noddies 
(Anous minutus). A more restricted stage (Brothers, Raurau, 
central Nake, Pig). 

Ecological Succession on Motus of Different Size Classes 

To assist discussions of succession on Caroline's motus, 
refer to the individual vegetation maps and graphs 
(Figs. 27-57) and photographs (Pis. 13-80). particularly 
Figs. 27-30. The latter figures summarize the amounts of each 
islet's surface covered by each major plant community, as well 
as providing the numbers and percentages of indigenous species 
for each islet. 

a) Motus with Areas <0.2 ha (Figs. 27,31; PI. 49; 
Tables 5.6): Caroline has four motus in this category, three 
windward and one leeward, whose combined area totals 
0. 15 ha. There are also three incipient islets which, because of 
their temporary character, have not been counted in Caroline's 
overall total (Fig. 2; PI. 15). With the exception of Noddy 
Rock — a jagged, upraised limestone plateau — all consist 



23 



predominantly of coarse coral rubble (75-98% coverage). 
These liny motus are the simplest ecosystems on the atoll, 
representing early Stage I in plant succession. The number of 
plant species per motu averages three, all hardy, sea-dispersed, 
and salt-tolerant pioneers (Heliotropium, Portulaca, Lepturus, 
Tournefurtia). The sole plant community is a natural herb mat 
of varying thickness and extent. Tournefortia, though stunted 
and scattered, is not sufficiently common to form a separate 
scrub habitat. Indigenous vegetation covers 2 to 22% of the 
islet areas. Seabirds. especially brown noddies and red-tailed 
tropicbirds, may attempt to nest. 

b) Motus with Areas 0.2 to 0.7 ha ( Figs. 28,3 1 : Tables 5.6): 
There are five leeward motus in this category whose combined 
areas total 2.21 ha. Their vegetative cover is more extensive 
and diverse than in size class a, with herb mats and Tournefortia 
scrub and forest, but open rubble is still abundant (30-55% 
cover). Plant succession corresponds to late Stage I and 
Stage II. The average number of species is 8.2 (range 6-11), 
one-third of Caroline's total. All vegetation on these motus is 
indigenous except for a few Cocos palms. Seeds are dispersed 
by sea, wind, and birds. 

With the appearance of shrubs, the number of species 
increases markedly, and woodlands, primarily of Tournefortia, 
form and expand to create dense thickets averaging 5 m tall and 
covering 25% of the land area. Canopies of 10 m occur on 
motus Nautonga and Kota. Seabird colonies of up to six 
species (brown and red-footed boobies, great frigatebirds, 
black and brown noddies, white terns) are present. 

A low herb mat, dominated by Heliotropium, Portulaca, 
Boerhavia, and, more rarely, Lepturus, develops first, after 
which Tournefortia quickly becomes established. Shade, 
producing locally humid conditions, and better "soils" derived 
from guano, decomposing leaves, and the activities of land 
crabs and rats, provide appropriate habitat for Laportea and 
occasional Phymatosorus and Achyranthes. The major tree 
species — Pisonia, Morinda. Cordia, and Cocos — subsequently 
appear but are relatively rare. Pisonia, typically an inland 
species assumed to need companion trees and underground 
water (Wiens, 1962; Spicer & Newbery. 1979), could well be 
salt tolerant as it occurs on motus as small as 0.2 ha (Tables 5,6 ). 
In this size class Pisonia occupies only 2-6% of the total islet 
areas. 

c) Motus with Areas from 0.8 to 25.0 ha (Figs. 29,31; 
Tables 5,6): All 27 motus in this category share a similar 
complement of species and plant communities (Tables 5, 6). 
Their combined area totals 1 24.35 ha. They are well wooded 
(Fig. 29), although the leeward motus have a higher proportion 
of rubble and herb mats, and forests are higher to windward. 
Unvegetated rubble covers less land area (219? ) than in size 
classes a and b (87% and 39%, respectively). Within the 
woodlands of these motus, substrates mature from basic rubble 
to primitive "soils" with small, but significant, structural 
development. Then flora shows increasing diversity with size, 
and almost the lull complement of seabirds may nest. 

All natural ecosystems are firmly established; canopy 
heights range from 4 to 2 1 m. On Booby Islet (0.84 ha). Pisonia 
suddenly becomes very common, and the Pisonia forests on 
North Brothers (1.71 ha) and Pig (7.25 ha), at 21 m. are the 



tallest on Caroline. As rich guano and dead foliage accumulate, 
a layer of phosphate-rich humus enables those species already 
present but poorly represented on the small motus (Pisonia. 
Morinda, Boerhavia, Laportea, Achyranthes) to increase in 
abundance and stature (Table 6). Additional species are Suriana, 
Pandanus, Scaevola, Ipomoea, Lepidium, and Species A. 

Plant succession, ranging from Stage III to Stage V in the 
interior, primarily involves forest maturity rather than the 
addition of large numbers of species. On the larger islets, the 
numberof plant species increases by relatively small increments, 
filling out the subcanopy layers and, in the cases of Cordia and 
Pandanus, adding variety to the canopy. 

The average number of plant species is 1 1 .0, ranging from 
4 to 15. If we divide the motus into smaller size classes, we find 
that their species numbers increase slightly with increasing 
size: 8.0 species for areas 0.8-1.0 ha, 9.8 species for areas 
1.1-2.0 ha, 1 1.5 species for areas 2.1-4.0 ha. 1 1.3 species for 
areas 4.1-10.0 ha, and 12.0 for areas 10. 1-22 ha. An increase 
in herbs (range 3-9) is primarily responsible for these higher 
averages (Table 5). 

Despite the large range of motu sizes in this category, plant 
communities are essentially natural (Table 5). Their overall 
species composition is 96% indigenous. Seventeen of the 
motus lack Cocos, the only introduced species in this area 
category, which is represented by small, isolated clumps or 
individual palms. 

On the larger motus, and within the taller forests, more 
species of birds, especially red-footed boobies, great frigatebirds. 
white terns, and black noddies, nest in increasingly large 
colonies, furnishing more minerals to the developing soils, 
especially where Pisonia covers large areas. 

In summary, by the time a motu on Caroline has reached 
0.8 ha in size, all the natural plant communities, most species 
of trees, shrubs, and herbs, and most species of seabirds are 
present. In Caroline's depauperate flora there are few species 
left to increase floral diversity on the larger islets, regardless of 
their size. This is very different from the inhabited atolls such 
as Kapingamarangi. 

Although we do not know when true freshwater lenses 
develop, they may occur in motus of this size class. If we 
assume that Pisonia is not salt-tolerant, limited fresh water 
must be available on motus as small as 0.2 ha. and actual 
freshwater lenses may begin forming at around 0.7 ha, as 
indicated by the sudden proliferation of Pisonia forest 
(Tables 5,6). However, the Falconers have not been successful 
in finding any underground fresh water on Motu Ana-Ana 
(2. 16 ha), which suggests that Pisonia may be salt-tolerant. 

d) Motus with Areas >25.0 ha (Figs. 30,31; Tables 5.6): 
On Caroline, no motus fall between 22 and 75 ha in size. Thus 
the three motus in this category (Nake. South, Long) cover a 
limited range of sizes: 75.98 to 107.50 ha. They average 18.0 
plant species. The floral components and forest heights of 
these larger motus (Figs. 32,33; Table 5) are essentially the 
same as for class c. There are no additional ecosystems 
( mangrove swamps, salt Hats, grasslands, et cetera ) or understory 
layers. Ten species, all rare or uncommon, are present only on 
the larger motus (Table 3): Scaevola. Tribulus, Hibiscus, 
Thespesia, Ximenia, Psilotum, Tacca (introduced in 1834), 



24 



Phyllanthus, Sida, and the dubious Digitaria. Four, possibly as 
many as eight, are indigenous. In 1965, one vine of the 
indigenous Ipomoea pes-caprae was also found, but three 
subsequent surveys failed to locate it. 

Species-Area Relationships 

The relationship between the numbers of plant species and 
island size has long fascinated biologists (Fosberg, 1949; 
Wiens, 1 962; MacArthur& Wilson, 1967; Whitehead & Jones, 
1969), yet data from uninhabited islands is scant. The studies 
from Kapingamarangi (Niering, 1956; Wiens, 1956) and 
Aitutaki ( Stoddart & Gibbs, 1975 ) treat atolls with long histories 
of human occupancy. SomeofthevillagesonKapingamarangi's 
23 motus date to 1200 A.D. Aitutaki's 16 uninhabited motus 
lie adjacent to a westernized volcanic island in an "almost- 
atoll." People on both these atolls have profoundly influenced 
their flora. 

Caroline provides an opportunity to compare the numbers 
of species on motus of different sizes in an uninhabited atoll, 
then to compare the results with Kapingamarangi, Aitutaki, 
and uninhabited islands in the Line and Phoenix groups that 
have no introduced species and have experienced minimal 
human contact. 

Comparisons of Species-Area Relationships with Other 
Atolls : Studies of Kapingamarangi (Niering, 1956 (contributed 
greatly to theories of island biogeography (Mac Arthur & 
Wilson, 1967). Because its motus coverthe same range of sizes 
as Caroline, the two atolls might be expected to exhibit similar 
patterns. However, their species-area relationships are 
completely different. On Kapingamarangi, islets less than 
1 .4 ha showed a constant, small number of species, after which 
islets up to 100 ha showed a direct correlation of area with 
numbers of species. On Caroline, a motu of 1 .4 ha supports 
almost two-thirds of the total number of species, and plant 
diversity on islets up to 107 ha shows only a slight, but not 
necessarily steady, increase (Table 5). 

Species-area relationships on the motus of Aitutaki 
(Stoddart & Gibbs, 1975, Figs. 33 and 34 of that paper) 
conformed to the Caroline model; the number of species 
increased only slightly on motus from 4 to 71 ha. Unfortunately, 
Aitutaki had only one motu less than 1 .4 ha, so comparisons for 
smaller islets cannot be made. The floras of all three atolls have 
been impacted by man, but Caroline far less so than the others. 
Much of the floral diversity on larger islets at Kapingamarangi 
is derived from plants introduced by man and cannot be 
considered normal. Caroline and Aitutaki provide much better 
samples of natural plant species-area relations on atolls. 

Six islands in the Line and Phoenix groups (Maiden, 
Starbuck, McKean, Phoenix, Vostok, Birnie) are uninhabited. 
Their flora is entirely native. All are Caroline's "neighbors" in 
an oceanic sense, and all except Vostok are dry , receiving about 
750 mm (30") of rain yearly. They are old, essentially filled- 
in atolls, containing hypersaline central lagoons or no lagoon 
at all. Although the largest island (Maiden) has the greatest 
diversity, there is only a very small linear increase in plant 
species with increasing area (Table 8). Plant diversity is more 
a function of climate (hot and dry) and distance from source 
areas, than size, similar to the situation on Caroline. 



The Question of Fresh Water : The Kapingamarangi data 
were analyzed with availability of fresh water in mind (Wiens, 
1962;Whitehead&Jones, 1969). These authors suggested that 
1 .4 ha is the threshold at which a freshwater lens can develop. 
Below this size only halophytes can survive. They argue that, 
as there are only a limited number of salt-tolerant species, the 
floral composition on islets below 1 .4 ha is relatively constant. 
On larger islets, species numbers increase in direct proportion 
to land area, because permanent groundwater promotes the 
survival of an increasing variety of nonhalophytic plants. 

The groundwater versus plant model does not apply to 
depauperate Caroline for a number of reasons: first, the number 
of plant species is not constant on islets below 1.4 ha: in fact, 
species are added faster on motus from 0.02 to 1 .4 ha than any 
other size range. 

Second, on Kapingamarangi, the number of species 
increased in direct relation to islet size from 1.4 ha to 100 ha. 
On Caroline, species numbers increased only slightly from 
1.4 to 22 ha and exhibited another minor increase from 70 to 
108 ha (see Fig. 31; Tables 5,6; and Ecological Succession 
section). Thus, Caroline's data do not support the area- 
diversity theory. 

Third, Whitehead & Jones (1969) argued that the flora on 
"small" motus lacking a freshwater lens (i.e., < 1 .4 ha) consists 
only of salt-tolerant strand species. This is not true on Caroline 
(Table 6). In addition to harboring the usual strand species 
(Tournefortia, Portulaca, Laportea, Heliotropium, Boerhavia, 
Lepturus), Caroline' s "small" motus also support inland species 
that are generally considered nonhalophytic (Pison ia, Morinda, 
Achyranthes, Cordia, Phymatosorus). Either these latter five 
species are moderately salt tolerant, or on Caroline the minimum 
islet size with a freshwater lens is much less than 1 .4 ha, or both. 

Fourth, Whitehead & Jones (1969) postulated that the 
nonhalophytic species are those that control overall species- 
area associations. This may be a good generalization for less 
remote islands but does not hold up for atolls with depauperate 
floras (Table 6). For example, on Caroline the halophytic 
Ipomoea macrantha, I. pes-caprae, Scaevola sericea, Sida 
fallax, Lepidium bidentatum, Hibiscus tiliaceus, Thespesia 
populnea. and Tribidus cistoides, which theoretically should 
only occur as strand species on the smaller islets, occur only on 
larger islets. In addition, several nonhalophytes (e.g., Morinda) 
were found at Caroline on small motus where one might only 
expect to find strand species. 

Fifth, the above authors do not mention bird-dispersal of 
seeds, which is probably a factor that needs to be taken into 
account on remote islands: at Caroline, Pisonia and Boerhavia 
contribute to the floral diversity of islets from 0.2 ha to 108 ha. 

Sixth, Caroline does not have an assemblage of nonstrand 
plants that only occur on larger motus; the only naturally 
occurring, nonstrand plant is Psilotum. 

Seventh, the greatest factor complicating our understanding 
of Kapingamarangi's natural evolutionary processes is the 
presence of numerous exotics: of its 98 vascular plants, only 
38 (39%) are indigenous. Its exotics include numerous weedy 
herbs and food plants, which occupy gardens, abandoned 
house sites, taro patches, and plantations (Cocos, Pandanus, 
Artocarpus). These man-made habitats are particularly 



25 



prevalent on larger islands. Such an abundance of exotics, both 
in species and area covered, renders a discussion of natural 
processes on Kapingamarangi almost impossible. Undisturbed 
habitats such as those on most of Caroline's motus, and on 
other uninhabited Pacific islands whose quota of indigenous 
plants exceeds 75%, provide far better data on species-area 
relationships. 

Motu Size in Relation to the Distribution of Trees, Shrubs 
and Herbs : As one progresses from small to large islets 
(Table 5), the number of tree species rises from to 7, the 
number of shrubs from 1 to 4, and the number of herbs from 
2 to 12. Caroline's trends are similar to those at Aitutaki 
(Stoddart & Gibbs, 1975), where the numbers of trees and 
shrubs are relatively constant over a wide range of motu sizes 
(3.8-7 1 ha), while the number of herbs shows a slight increase. 
There are too many recent exotics on Kapingamarangi for 
comparisons to be valid. We believe that if Niering's data were 
reanalyzed, using only indigenous species, similar 
generalizations would be found, viz.: most species on atolls 
establish rapidly on small motus, after which a few additions 
occur on motus of increasing size until the maximum number 
of potentially available species is reached. Cursory examination 
of Niering's Fig. 31, detailing the breakdown of total species 
numbers into indigenous and nonindigenous components, bears 
out this hypothesis. 

Plant Communities 

General Account 

The total area covered by vegetation on Caroline is 
357.55 ha, fully 90% of the combined areas of all the motus. Of 
this, two-thirds (289.82 ha) is woodland. Substantial areas of 
Caroline's native woodlands and herb mats are pristine, and 
85% (possibly as high as 93%) of its plant species are indigenous. 
Twenty-three (60%) of its 39 motus harbor wholly indigenous 
vegetation (Figs. 27-30). Atolls that support substantial areas 
of native forest are typically remote and uninhabited. Where 
people are present, native vegetation is usually confined to the 
smallest motus or the extremities of larger ones — areas with 
marginal human usefulness. 

Typical of atolls, but unusual for the tropics, are monotypic 
stands of shrubs and trees. Caroline is rich in such woodlands 
(Figs. 18,22,24.27-30; Table 9). The present vascular flora of 
Caroline, 27 species, is organized into 7 plant communities 
( 1 1 subcommunities) defined principally by dominant species 
(Fosberg, 1953. 1977a). Eight subcommunities are natural, 
three are anthropogenic (Table 5 ). The subcommunities include 
a mix of dominant species, which are discussed in the major 
Community sections below. 

NATURAL COMMUNITIES: 
Natural Herb Mat 
Toumefortla Scrub and Forest 
Beach Scrub with Suriana 
( 'ordia Forest 
Paiulamis forest 
Pisonia Forest 



ANTHROPOGENIC COMMUNITY: 
Coconut Woodlands 

Natural Herb Mat (67.73 ha) (Figs. 19.23: Pis. 20.33, 
34.45,46,47) 

Widespread and predictable on wind- and salt-blown 
coastal coral rubble and incipient motus, these mats are 
composed primarily of Heliotropium and Portulaca. They are 
pioneers on newly emergent motus, cover most of the ground 
area of small motus, extend inland along ancient reef channels, 
and typify newly evolving land that connects or augments 
established islets. Natural herb mats may persist through all 
five stages of plant succession as long as sunny openings occur. 
Caroline's motus illustrate two general principles: /. the 
smaller the area of an islet, the more extreme is the strand 
character of its vegetation, and its corollary; 2. as islet areas 
enlarge, strand flora becomes less important (Fosberg, 1949). 

The following species are present (see Table 2 for 
abundance indices): 

Trees: Morinda citrifolia (one drift seedling on one motu); 
Shrubs: Tournefortia argentea, Suriana maritima, and 
Scaevola sericea; and 

Herbs: Heliotropium anomalum, Portulaca lutea, Boerhavia 
repens, Lepturus repens, Laportea ruder alls, Lepidium 
bidentatum, and Ipomoea macrantha. 

Near the high water mark, the herb mats are recumbent, 
leathery, and somewhat desiccated. As environmental 
conditions improve further inland, they spread more laterally 
and average up to 7 cm in height. Their rubbly habitat, often 
sprinkled with Tournefortia, resembles a low savannah. 
Although these prostrate herbs can tolerate dazzling sunshine, 
they grow most vigorously when slight shade, and hence a 
higher relative humidity, is present. Under these conditions 
they may attain a height of 22 cm and form a fairly thick mat. 
With too much shade the mats disappear or their species 
proportions and abundance changes according to the presence 
or absence of sunny clearings. Thus, natural herb mats may be 
found in patchy clearings within forests up to 1 3 m tall. They 
are common in the abandoned Cocos plantations of South 
Island, where Boerhavia tends to proliferate into thick mats 
that completely cover the substrate, vying with Phymatosorus 
and Ipomoea for"lebensraum" (PI. 34). A thick, exposed mat 
of succulent herbs is found on Noddy Rock, where Portulaca 
is the primary component. 

Herb mats occurred on almost every transect, windward 
and leeward, ranging from 1% to 60% coverage (Figs. 19,23). 
The most extensive areas (coverage 35-50%) were on Skull. 
Tridacna, South (Trs. 1.4,6). Emerald, and Mannikiba. Mats 
predominated in sparsely vegetated areas. Their widths varied 
according to the age, shape, exposure, and geographic position 
of the motu but were widest on seaward-facing shores 
(Table 7). 

Wide bands of herb mats may encircle an entire motu: to 
windward they average 36 m (Table 7), while, bordering the 
relatively placid and intermittently shaded lagoon, they shrink 
to a mere 0.9 m. On leeward motus, the corresponding figures 
are 18.5 m and 4.2 m. 



26 



Although reef flats are typically wider wherever an islet 
turns sharply, it is not unexpected that these perimeter bands 
are the most extensive on the extremely exposed shores of 
northern Nake( PI. 17)andsouthernSouthIsland. On the latter, 
they are up to 59 m wide. Similarly, on small exposed motus 
(e.g.. Skull. Noddy Rock), they carpet most of the area 
(Fig. 27). Under such conditions, Portulaca and Boerhavia 
develop much redder stems, possibly due to the presence of a 
chemical "sunscreen." 

Associations with Birds : Whether bordering the edges of 
established islands or composing the entire ground cover of 
tiny motus and ancient reef channels, herb mats are nesting 
sites for red-tailed tropicbirds, masked and brown boobies, 
sooty terns, and brown noddies. Herb mats are often used as 
foraging grounds for shorebirds. 

Beach Scrub with Sunana ( 1 .49 ha) (Fig. 20; Pis. 6,24,39,44) 
Uncommon on Caroline, beach scrub with Suriana is 
typically foundon sandorsandy rubble bordering Tournefortia 
or Cocos. On Caroline, it is evidently limited by the paucity of 
low-lying sand and gravel sheets, with which it is normally 
associated elsewhere (Fosberg, 1953; Wiens, 1962;Stoddart& 
Gibbs, 1975). 

The following species are present (see Table 2 for 
abundance indices): 

Shrubs: Tournefortia argentea, Suriana maritima; and 
Herbs: Heliotropium anomalum, Boerhavia repens, 
Portulaca lutea, Laportca ruderalis, Phymatosorus 
scolopendria, and Lepturus repens. 

This plant community was found on 10 motus (Fig. 20), 
either in thick bands or as scattered shrubs. Suriana is most 
robust on sandy substrates, especially fringing the lower lagoon 
on South Island (Fig. 36; Pis. 6,24,39) and on windward 
Tridacna. The fringe, repeatedly interrupted by other species, 
grows to 1 2 m wide and 1.8 m high. Here the shrubs are closely 
appressed and slightly entangled, forming dense shade, which 
supports a sparse understory. On South, where its roots are 
submerged at high tide, it is being shaded out by overhanging 
Cocos (compare Pis. 39 and 40), having retreated since 1965. 
Suriana also occurs as scattered individuals or in open bands in 
coarse rubble. Beach strand up to 60 m wide, containing herb 
mats, Tournefortia, and scattered Suriana, were found on 
South (Tr. 1, PI. 21 ). Long (Tr. C), Brothers. Matawa, Long, 
and the Southern Leeward Islets. 

Pandanus Forest (3.38 ha [this figure is pure Pandanus forest. 
Mixed forests containing Pandanus account for a further 
14.96 ha]) (Fig. 11; Pis. 18.35-38) 

Although several species of Pandanus are native to the 
Line Islands, and their seeds are common components of 
Pacific sea-drift (Ridley, 1930; Stone, 1968), it is probable that 
the groves of P. tectohus on Caroline represent both naturally 
established forests and cultivars transported by early 
Polynesians. Its largest acreages are on two islands that 
contained historical settlements (Nake, South). However, its 
presence within the interior forests of a few motus lead us to 
conclude that it may have experienced a dual introduction. On 
Emerald Isle, 3.20 ha (38% of the islet) supports a mixed forest 



of Tournefortia, Pisonia, and Pandanus. Similarly, Shark 
Islet's interior woodlands of Tournefortia, Pisonia, and Cordia 
(5.52 ha, 70% of the islet's area) also contain a substantial 
amount of Pandanus. The only record of habitation for these 
motus was a possible hut on Shark. The occurrence of Pandanus 
groves or lone trees on other islets (Fig. 1 1 ) is easily attributable 
to drift seedlings. Dried Pandanus seedpods are the most 
conspicuous litter along Caroline's lagoon beaches (PL 38); its 
seeds last for months in seawater (Guppy, 1906) and are 
probably distributed locally by rats and land crabs, as noted 
elsewhere (Ridley, 1930). Carpels from Nake's southern 
mixed woodlands undoubtedly established the grove on 
Pandanus Islet. 

The mixed forest with Pandanus on south Nake (with 
Cocos, Cordia, Pisonia, and Tournefortia) contains up to 50% 
Pandanus attaining heights of 1 2 m (Fig. 37). It does, however, 
look disturbed. 

Many Pandanus trees were felled on South Island during 
the coconut planting era (ca. 1 873- 1 925 ), as we know that they 
were "somewhat numerous" in 1834(Bennett, 1840), but only 
"one or more of the screw pines were found growing in various 
parts of the island" in 1883 (Trelease. 1884). A drawing in this 
latter paper (PI. 50) depicts a grove from South Island denser 
than any remaining today, where Pandanus is uncommon in 
the beach scrub bordering the Cocos plantation. 

Trees were fruiting abundantly in September 1988, 
especially on Nake. The green phalanges, 17.5-20 cm in 
diameter, ripen to yellow and orange when they fall to the 
ground. They are eaten by hermit crabs (Coenobita perlatus 
[PI. 381). 

Tournefortia Scrub and Forest (125.25 ha) (Figs. 24.34; 
Pis. 5.8,20,30,47.51) 

General Distribution : Characteristic of many Pacific 
islands, Tournefortia, a broadleafed evergreen, dominates the 
wooded motus of Caroline, forming 40% of its total vegetative 
cover (Fig. 24). Its pale foliage and hemispherical canopies (to 
14 m tall) typically surround the taller, darker canopies of 
Pisonia and Cordia. 

A hardy halophyte. Tournefortia occurs on every motu 
and in every habitat except pure Pisonia forest. It is tallest, 
widest, and lushest on the windward motus, particularly on 
those where Pisonia is also best developed. Without direct sun 
though, as under dense Pisonia or Cocos, it withers (PI. 24). 

On other atolls Tournefortia forms a narrow or interrupted 
belt inland of the beach or is a component of mixed scrub 
(Fosberg, 1953). However, given the floristic poverty on 
Caroline, especially of shrubs and trees, Tournefortia not only 
has expanded into niches that might elsewhere be occupied by 
combinations of Scaevola. Pemphis, Suriana. Terminalia, 
Hernandia, Thespesia, Hibiscus, et cetera, but frequently occurs 
in pure stands ( 1 13.03 ha) that extend well inland. It thus 
occupies a much higher percentage of the islet areas on Caroline 
than on atolls with greater biodiversity. For example, Nake, the 
largest islet, has the greatest amount of Tournefortia (79.68 ha) 
of any islet: 28.9 ha of pure scrub and forest, 18.28 ha of 
"savannah," 17.48 ha with Cordia, 8.99 ha with Pisonia, and 
6.03 ha mixed with Cocos, Pandanus, and Pisonia. 



27 



Overall, we classify Tournefortia as a shrub (Stoddart & 
Gibbs, 1975). However, following Mueller-Dombois et al. 
( 1 98 1 , p. 58), we also distinguish between its shrub (scrub) and 
tree communities. Because they intergrade. sometimes we 
lump them together ( vegetation maps and schematic profiles of 
the motus) and at other times treat them separately 
(Tables 2.5, 1 and ecological discussions |: 
/. Tournefortia Scrub: <5 m high (x = 2 m). <60% canopy 
coverage (Pis. 20,30,33,47 (.This open scrub growth is typically 
confined to islet perimeters or emergent reef channels and 
covers much of the vegetated rubble on smaller islets. Its 
species composition is similar to that of the taller forest, except 
that herbs are more prominent. 

2. Tournefortia Forest: >5 m high ("x = 8 m), >6C7f canopy 
coverage (PI. 48). This taller, closed forest, with maximum 
height 1 5 m. develops as a second belt of woody vegetation 
approaching the interior of the larger islets. Figure 34 depicts 
a schematic profile through pure Tournefortia scrub and forest, 
while Fig. 35 diagrams a profile of a larger islet where 
Tournefortia is represented only on its periphery. 

Species Diversity in Tournefortia Woodlands : The 
following species occur in both scrub and forest. Those marked 
"*" occur primarily in the scrubland (Table 2). 
Trees: Pisonia grandis, Morinda citrifolia, Pandanus 
tectorius. Cocas nucifera, Cordia subcordata; 
Shrubs: Suriana maritima, Tournefortia argentea, Scaevola 
sericea. Species A; 

Herbs: *Heliotropium anomalum, *Boerhavia repens, 
*Portulaca lutea, *Lepturus repens, *Laportea ruderalis, 
*Achyranthes canescens, Phymatosorus scolopendria, 
Ipomoea macrantha. 

Caroline's tallest Tournefortia stands (12-15 m) occur on 
Nake. On all other windward motus, the Tournefortia canopies 
vary between 6 and 9 m tall, shorter than expected if their 
forests were virgin. This has historical significance: we do not 
know the extent of forest felling (if any ) on the Windward Islets 
(Crescent through Tridacna) during the guano era, but we do 
know that 4.587 coconut palms were planted during 1 9 1 9-20, 
and that "misses" (dead seedlings) were fastidiously replaced 
over the following 2 years (Young, ca. 1922). Thus, their 
forests, though weed-free today, comprise secondary growth 
around 60 years old. It is not surprising that Achyranthei 
canescens and Lepturus repens, both weedy (though 
indigenous), are particularly common inland on some windward 
motus (Figs. 13.16). Tournefortia' s rapid recovery illustrates 
that ecosystems in the pioneer stage generally recover their 
original condition rapidly when left alone (Fosberg. 19S3). 

Stature and Area Coverage : Forming an umbrellalike 
canopy, a typical Tournefortia forest is very simple. Its twisted 
branches ami gnarled trunks stretch about untidily over an open 
understory. The lower branches die off as the trees increase in 
stature. Sometimes a scant coverofherbs develops in restricted 
pockets of better soil, such as a clearing where a dead tree fell, 
a semishaded spot beneath a colony of seabirds, or a site where 
a storm deposited a leu dead fish. 

Tournefortia is abundant throughout the atoll. Areas with 
90-10095 canopy cover were found on Nake (Tr. 4). Long 
(Trs. B.C.4.6. 1 0. 1 2 ). North Pig, Pig. North Brothers, Brothers. 



Crescent, Arundel (Fig. 34). Tridacna (Trs. 1.2). South 
(Trs. 1,4), all 5 Southern Leeward Islets, all Central Leewards 
over 0.5 ha, and Pandanus Islet. Tournefortia is present across 
the entire width of some small motus — for example. Fishball 
(144 in wide). Even on larger motus such as Mannikiba 
(280 m wide). Tournefortia blankets nearly all the land ( PI. 70). 
Long (75.98 ha) is a composite motu: long, narrow, and 
derived from the coalescence of at least five former islets. 
Because 7barn€/brria encircled the perimeters of these ancient 
islets, it is now present in five sets of concentric circles. 
connected by herb mats, down the length of the island ( Fig. 39). 

In the herb mats, Tournefortia is small (x = 1.4 m) and 
widely scattered (Table 10). It may be of typical hemispherical 
shape or irregularly windshorn (Pis. 13.45). On windward 
coasts they typically form a tight wind barrier one or two trees 
thick. Moving inshore from the seaward fringe, the trees 
become progressively taller ( x = 6 m) with a more open 
understory. Cordia often mixes with Tournefortia, either as 
scattered individuals in the understory or canopy, or as small 
groves. On the Southern Leeward Islets, such belts border the 
seaward scrublands. 

Though still widespread in the Pacific. Tournefortia is far 
less abundant than formerly. On inhabited islands it exists 
primarily in relict patches or as edging around anthropogenic 
forests. It rarely covers most of the land area of islets: two 
exceptions are TaongK Marshall Islands) and GaferuK Caroline 
Islands), both in Micronesia (Fosberg. 1956: Wiens. 1962). 
The finest quality Tournefortia forests on Caroline Atoll (15m, 
80% cover) occupy central and northern Nake (Fig. 37); given 
Caroline's history of occupation, these could well be virgin. 
These 15-m Tournefortia compare favorably with 18-m 
specimens found at Jemo Island by Fosberg ( 1956). Perhaps 
Jemo's trees are at the upper size limit for the species, as 
Tournefortia is generally recorded as 3 to 6 m tall (Wiens. 
1962). 

Ecology : Tournefortia is an integral part of the atoll's 
evolution and ecology. Bearing seeds capable of floating for at 
least 4 months in the sea (Guppy, 1906), it is the first woody 
plant to establish on tiny motus (<0. 1 ha), appearing immediately 
after the native herbs have begun to germinate in the coarse 
coral rubble. Requiring little or no soil and adequate rainfall, 
it can grow up to 2 m a year (Fosberg. 1959). Tournefortia' s 
leaves contribute to soil development, paving the way for plant 
succession from Stages I through IV. for it only persists in soils 
that are conducive to the growth of its mesophytic competitors 
(Fosberg. 1953). The most mature trees (x = 9.5 m) occur at the 
Tournefortia— Pisonia interface, but die off as Pisonia expands. 
When Tournefortia has reached its maximum height, most of 
its lower branches have fallen, leafage is reduced, and flowers 
and fruits are few. Tournefortia usually drops out after one 
generation. Seedlings are rarely seen in heavy shade, and fallen 
trees are fairly common on the edge of the interior forests where 
Pisonia replaces it. 

An example of complete replacement of Tournefortia by 
Pisonia is illustrated by nearby Vostok. It has heretofore been 
assumed that Vostok's sole tree species was Pisonia grandis 
(Fosberg. 1936: Bryan. 1942: Clapp& Sibley, 1971b; Garnett, 
1983). However, Young (ca. 1922) stated that when Captain. I. 



28 



Larsen. of the schooner Papeete, planted 1 00 coconuts there on 
3 1 May 1922. he found "Pukatea and Tauhinu trees, et cetera 
60 to 80 feet high;" that is, Pisonia grandis and Toumefortia 
argentea, but no "Tou" trees (Cordia subcordata). By 1935 
only Pisonia remained (Fosberg, 1936); thus, the last natural 
Toumefortia must have been eliminated by Pisonia. 

Along some coasts (Long, Nake, South). Toumefortia 
overhangs the water, its roots immersed at high tide. We found 
floating debris up to 20 m inland within dense Toumefortia 
forest, indicating that this hardy shrub can withstand periodic 
storms and high tides. If a rosette of Toumefortia leaves is 
placed in fresh water, it droops within an hour, indicating that 
its tissues require a high salt concentration in order to maintain 
turgidity (personal observation). Perhaps decreased salinity in 
the ground water, coupled with reduced light intensity in 
advanced serai stages, contribute to the eventual disappearance 
of Toumefortia in the center of coral islands. 

Associations with Birds : Toumefortia is a favored roosting 
and breeding site for most of Caroline's seabirds. The taller the 
trees, the greater the bird diversity they harbor: scrub contained 
four species (36%) and forest, nine (82%). Sooty terns nest in 
tight colonies in its shade, its canopies support large populations 
of red-footed boobies (PI. 51) and great frigatebirds 
(Subchapter 1.2, this volume), and its branches are favored by 
white terns (Figs. 34-36). Toumefortia leaves provide nesting 
material for noddies. 

Cordia Forest (1.39 ha) (Fig. 22; PI. 27) 

General Distribution : Cordia does not form "the main 
native woodland" on Caroline Atoll, as implied by Clapp & 
Sibley ( 197 la) and stated by Stoddart & Gibbs (1975, p. 104). 
It occupies far less area than Toumefortia or Pisonia (Table 9). 
Cordia is generally mixed with other emergents: monotypic 
Cordia forest covers only 1.39 ha, while Toumefortia or 
Pisonia containing substantial amounts of Cordia total 
25 .89 ha. In toto. this is less than 1 0% of Caroline' s woodlands, 
and Cordia is usually subdominant. We treat Cordia forest as 
a separate plant community because of its increasing rarity on 
Pacific atolls, which makes Caroline's groves an increasingly 
important resource in need of conservation. Cordia forest 
occurs primarily on Nake, Windward. Crescent, North Pig, 
Pig, Danger, Shark, and the Southern Leeward Islets. 

History : Bennett (1840) recorded "two species of 
Toumefortia" on Caroline, possibly referring to Toumefortia 
and Cordia. There are no other 19th-century records. From 
Cordia 's present distribution we can infer that it was formerly 
more extensive on South and Nake. Scattered trees within and 
bordering the Coeos plantations suggest that its history is 
similar to the species on Flint: both Flint and Caroline were 
worked simultaneously by the same companies for guano 
( 1872-1890) and copra (into the 193()'s). Pisonia and Cordia 
forests were felled to make room for coconuts. Several 
hundred Cordia logs were exported from Flint to San Francisco 
to be used for furniture. The last logs were exported in 1896, 
6 years after the guano supplies were depleted, but coconuts 
were still being planted (Young, ca. 1922). Today, the belt of 
indigenous vegetation bordering Flint's coconut plantation 
still has many large Cordia trees (Kepler, 1990b), unlike 



Caroline, where today Cordia is rare on South Island. However, 
some of Flint's Cordia trees today may well be those "few tiny, 
struggling. ..trees. ..recently planted" (St. John & Fosberg, 1937). 

Abundance and Distribution : Cordia seeds are dispersed 
by ocean currents and can germinate after 40 days in seawater 
(Guppy. 1906). Requiring the presence of other species 
(Fosberg, 1953), on Caroline it develops both as an understory 
shrub and forest emergent (to 15 m high). It typically occupies 
the woodland periphery, occurring in small circular or linear 
groves, or mixing with Toumefortia and/or Pisonia (Table 5). 
On many other Pacific atolls Cordia forms a mixed scrub with 
Scaevola (Fosberg, 1949). Cordia may form tall, straight- 
trunked trees (PI. 27) or sprawl like Hibiscus tiliaceus. In dry 
rubble sites it may become chlorotic ( PI. 79 ) or semideciduous. 
The tallest groves are on Pig (PI. 27), where six trees averaged 
12.6 m tall, 1 16 cm circumference at 1.5 m high, and 99.8 cm 
around the trunk base. Lush Cordia groves sheltered parts of 
the upper lagoon on Long Island (Tr. 10). 

Flowering times are unpredictable: In November 1989, 
flowers were abundant, extending through March, yet in 
November 1990 not one flower was observed (Anne Falconer 
and AK.K, personal observation). Two flowers were seen in 
September 1988 (personal observation). 

Associations with Birds : Black and brown noddies, 
frigatebirds, and white terns nest in Cordia wherever it is a 
forest component. Great frigatebirds and red-footed boobies 
favor roosting in the lush, lagoonside forest of Cordia and 
Pisonia near the south end of Long Island. 

Pisonia Forest (62.17 ha) (Figs. 18,32,33,35,39,41; 
Pis. 43,52,53; Tables 11-14) 

General Distribution : Although Pisonia grandis is recorded 
as "present" in the two previous scientific accounts of Caroline 
(Trelease, 1884; Clapp & Sibley, 197 la), the quality and extent 
of its forests has not been recognized. Some stands on this atoll 
are outstanding representatives of a major ecosystem that was 
formerly far more widespread in the Pacific. 

Common throughout the atoll, Pisonia occurs on 29 motus. 
Well developed groves, 10-21 m tall and up to 359 cm 
circumference at 1.5 m, are present on 23 of these (Table 1 1 ). 
Although present on motus less than one hectare in size 
(Table 5). it typically occupies interior forests (schematic 
profile. Fig. 35), with individual trees or groves contributing 
from 5% to 100% of the canopy. In general, Caroline's 
windward motus support the lushest forests: the maximum 
height of windward Pisonia forests is 2 1 m; of leeward forests, 
15 m. 

Mature Pisonia forests are monocultures of grandeur. The 
trees bear one to several stout boles of irregular shape, whose 
rotting cavities often harbor large coconut crabs or miniponds 
alive with mosquito larvae. Their scraggly branches 
occasionally bend over and reroot. It is dark and humid after 
the glare of the beach. Walking is easy because the forest floor 
is open except for exposed roots and a scattering of broken 
branches. Few seedlings occur. Overhead, aconstant cacophony 
of bird calls overwhelms the sound of the trade winds, and 
guano spills everywhere. Polynesian rats scurry underfoot. It 
is a curious habitat for a tropical island. 



29 



In September 1988 we saw no flowers or fruit. Anne 
Falconer reported flowers on Motu Ana- Ana in August 1990. 
Pisonia was beginning to bloom on Vostok in March 1990 
(A. Kepler, in prep.). 

A Historical Perspective : Some of Caroline's most mature 
Pisonia groves (to 21 m tall, 660 cm circumference at 1.5 m. 
multiple trunks) appear to be virgin. Overall dimensions of the 
trees, low species diversity, and general character of the plant 
community are similar to the known virgin groves on Vostok 
(personal observation, Table 12). We do not have dimensional 
data (other than height) of these particular areas on Nake, but 
their level of maturity can be seen in PI. 43. 

Despite the advanced stage of ecological succession of 
many groves, especially to windward, planting records from 
1916-1922 indicate that Cocos was planted throughout, not 
only South, but also on Nake. Long, and all the major Windward 
Islets (Young, ca. 1922). Given the standard planting density 
of one tree every 8.5 nr (28 x 28 ft) ( Young, ca. 1922), we have 
calculated the approximate area on each islet given over to 
Cocos plantations, based on the number of coconuts planted 
(Table 13) times the area required for each tree (73 nr). We 
have then compared this to the usable areas based on today's 
forest cover. On all nine Windward Islets, Cocos covered 
79 to 100% of usable ground; in several cases the amount 
calculated for Cocos exceeds the amount of potentially usable 
ground. Thus, Cocos was so intensively cultivated on the 
Windward Islets that essentially all the Pisonia and most 
Toumefortia forests must have been felled. 

Two remarkable points emerge from Table 13: 1. Scarcely 
any Cocos remains today on the nine Windward Islets; seven 
of them bear no trace of the former plantations 
(Figs. 43,44,47,48); and 2. recovery of Caroline's natural 
ecosystems. Stages I through V ( Ecological Succession section ). 
on the windward side has been rapid and, at least on Brothers 
Islet (Fig. 46), reasonably complete with regard to ecological 
succession and species diversity. Today the Windward Islets 
have the lushest and tallest plant communities, with higher 
species diversity than the leeward islets (Table 3). which have 
evidently experienced far less human disturbance. 

This differential disturbance on the w indward and leeward 
sides of the atoll explains enigmas such as 20-m-tall Pisonia 
forest on the leeward Booby Islet (0.84 ha), taller than most of 
the windward forests; the absence of Pisonia on windward 
Tridacna Islet (9.08 ha), which, being close to South Island, 
probably supported Cocos, which was managed longer than the 
more distant windward islets; and the patchy distribution of 
Pisonia in the interiorof several islets (e.g.. Windward, Arundel). 
This last point also applies to Mannikiba ( 2 1 .49 ha), the largest 
leeward islet. According to Young (ca. 1922), 6,000 seed sets 
were brought from Flint to Caroline in 1920 and kept on 
Mannikiba. This "nursery stock" was used to replant "misses" 
on other islets, due mostly to destruction by coconut crabs and 
poor planting. Today. Mannikiba's total acreage of Pisonia 
(Fig. 53) is very small and fragmented relative to the islet's 
si/e: 1.13 ha. 5'< of the total land area. Compare this with Bird 
Islet (Fig. 55). which, as far as we know, has never been 
disturbed: 1.70 ha Pisonia, 42' i of the islet's land area. 



On both Caroline and Flint there is much variation in the 
quality of the regenerated Pisonia forests (Table 12). Some 
trees bear enormous, partly rotting boles, black algae smothering 
the bark, multiple trunks, and few or no understory herbs. 
Other trees are tall and straight-trunked, with characteristic 
whitish bark, and bear no rotting holes in their bases. Such 
observations suggest that when their indigenous forests were 
felled, only minimal cutting was done, and many Pisonias were 
able to regenerate quickly by sprouting from rooted stumps and 
fallen branches. This speculation is supported by the fact that 
some of Vostok's Pisonia trees regenerated similarly. Maude 
(1953, p. 96) stated that "there is room for 8,000 palms on 
Vostok, but only 100 have been planted and most of these have 
been choked in the luxuriant 'buka' (Pisonia grandis) forest: 
no attempt having been made to exploit the island since the 
initial planting." 

Pisonia, a soft, pulpy wood, has a well-known ability to 
sprout or send up suckers from dismembered branches or fallen 
trunks ( Fosberg, 1 953), and it has been noted that older trees are 
virtually indestructible, fire being the only effective means of 
clearing forests (Wiens, 1962, p. 397). The senior author has 
photographed leaf sprouts from partly burned twigs as small as 
1 m long and 5 to 6 cm in diameter. 

Since the existing Cocos plantations on South Island and 
southwest Nake contain few Pisonias, it seems that forest 
clearing was more thorough on the atoll's larger islets than on 
the smaller ones, which today manifest scant traces of their 
former history. Fortunately for Caroline, its coconut plantations 
were plagued by a number of problems, which resulted in their 
being abandoned twice: coconut crabs, seabirds, rats, 
Ipomoea vines, and an unknown disease (see under Coconut 
Woodlands, this section). 

A footnote in Young (ca. 1922, p. 15) stated that "the larger 
portion of the 30,000 trees planted were either badly planted or 
smitten with some disease as in 1927 it was reported by 
Mr. Bunckley that most of them had perished." In 1929 only 
1 3,2 1 5 trees were left and more were being planted. Considering 
the distribution of both palms and natural forests today, it 
appears that plantations continued on South and Nake and were 
abandoned on the smaller islets, allowing for a better recovery 
than might be expected had the Cocos grown to maturity. 

Once a Cocos plantation has been well established and 
subsequently abandoned, Pisonia regrowth is more difficult. 
This is characteristic of mans islands in the tropics. For 
example, on Cousin Island (an ICBP wildlife preserve since 
1968, Seychelles Islands. Indian Ocean), where Pisonia is 
currently reestablishing within adeterioratingCf)cc.v plantation, 
Phillips & Phillips ( 1990. p. 37) envisioned "centuries rather 
than decades before something like a natural ecosystem 
develops." We predict a similar time frame for areas on 
Caroline and Flint where Cocos canopy is over 7095 cover. 

Annual Growth Rates : Data on Pisonia grandis growth 
rales are evidently lacking. On Cousin Island, vegetation 
changes, including Pisonia and Cocos, have been monitored 
since 1 474 (Phillips. 1984; Phillips & Phillips. 1990). However. 
as in) measurements of any tree dimensions are included, 
growth rates cannot be deduced. 



30 



Because of this paucity of data on Pisonia, and because its 
forests have diminished significantly on coral islands this 
century, we are presenting our data from Caroline, with 
comparisons with Vostok and Flint, in the hopes that it might 
inspire more research. 

One point is clear: on all three of the Southern Line 
Islands, Pisonia grandis has recovered fast from disturbance 
(except for total forest elimination), reaching close to its 
maximum height and ecological maturity in 70 years or less. 
Mature Pisonia, under optimal conditions of soil, temperature, 
and rainfall, may attain 35 m, as on Fanning and Washington 
(Garnett, 1983 and personal communication). However, in the 
Southern Line Islands, canopies of similarly virgin Pisonia on 
Vostok rarely exceed 25-30 m tall (Kepler, 1990c). 

Caroline's prime grove — 21 m tall, with circumferences 
(at 1.5 m) to 660 cm, and bearing multiple trunks and root 
suckers — we now know date back only to the 1920's. They 
have thus averaged a growth rate of 0.32 m per year since, say, 
1925 (65 years). Growth was undoubtedly fastest during the 
first few years. 

Further evidence of fast growth rates is provided from 
Flint Island. In 1934 only one small Pisonia was recorded 
(St. John&Fosberg, 1937). Fosberg( personal communication) 
recalled that only scant traces of native vegetation existed at the 
time, virtually the entire island (324 ha) being planted with 
Cocos. In 1990, transect surveys (Kepler, 1990b,d), coupled 
with an analysis of aerial photographs, revealed that single to 
multi-trunked Pisonia trees, now quite common on the windward 
side of Flint, attained maximum heights of 30 m, having 
circumferences at the base and at 1 .5 m of 1 ,000 cm and 200 cm, 
respectively (Table 12). These compare favorably with one 
large Pisonia, presumably virgin, measured on Atafu Island 
(Tokelau) by the US Exploring Expedition in 1 840, which was 
more than 600 cm in circumference at its base and about 12 m 
tall (Wilkes, 1845, Vol. V, p. 9). Furthermore, indigenous 
forests (Pisonia, Cordia, Guettarda), with canopies of 
4 to >20 m, covered 57 ha, 28% of Flint' s vegetated area. Thus, 
numerous Pisonias have not only established themselves since 
the plantation was abandoned in the late 1930's but have 
averaged approximately 0.5 m growth per year. This faster 
growth rate than on Caroline may be due to Flint's higher 
rainfall and greater relative humidity due to the presence of a 
more successful coconut plantation inland: Caroline's annual 
output of copra was 15 tons, compared to 230 tons for Flint 
(Young, ca. 1922; Maude, 1953). 

Species Diversity in Pisonia Forests : Caroline's motus 
harbor every stage in the development of a Pisonia forest, from 
stately monoty pic groves to a single tree. The plant communities 
between these extremes harbor the greatest species diversity 
and most luxuriant growth on the atoll. The following species 
are present (Table 2): 

Trees: Morinda citrifolia, Cordia subcordata, Cocos 
nucifera, Pandamis tectorius, Pisonia grandis; 
Shrubs: Tournefortia argentea; and 

Herbs: Boerhavia repens, Portulaca lutea, Laportea 
ruderalis, Lepturus repens, Achyranthes canescens, 
Phymatosorus scolopendria, and Ipomoea macrantha. 



The number of species within Pisonia forests ranges from 
1 to 14 (Table 14). As Pisonia becomes more dominant, their 
trees are taller (21 m), and species diversity is less (Table 14). 
Here, the average number of species is 3.4. Species diversity 
is also very low at the other extreme of Pisonia development: 
in one young motu (Azure), only a single 6-m-tall Pisonia tree 
is present (x = 4.0 m). The smallest islet on which we found 
Pisonia, Azure is only 0.20 ha in area and 77 m wide (Fig. 55, 
PI. 53); more than half of it is rubble. The width of its scrub is 
only 38 m. Along a transect within the majestic Pisonia grove 
( 100% canopy cover) on Brothers (Fig. 46), we found no other 
plant species, an extreme case of the barrenness of Pisonia 
understory. This grove, 13mtallandextending42mfromeast 
to west, was sharply delineated from the 6-m-high Tournefortia 
forests on both sides and provides a striking example of 
complete ecological succession since its Cocos plantation days 
of the 1920's. 

The highest species diversity occurred with mixed co- 
dominants (Tournefortia, Cordia), and Pisonia coverage 
25-50% (Table 14). Here, the average number of species was 
6.2 (range 3-10). Regardless of the area or width of the motu 
on which they occurred, these mixed stands (x = 7 m tall) were 
always shorter than pure Pisonia forest. 

Ecology : On Caroline, most plant species are established 
early in the evolution of individual motus, increasing 
in abundance and stature while the motus are quite small. 
Pisonia typifies this pattern: single trees occur on 2 motus 
whose areas are only 0.2 ha (Table 6). This suggests that 
Pisonia is partly salt-tolerant, at least in its early growth stages. 
In general, however, motus less than 0.7 ha on Caroline have 
little Pisonia (Table 6). It is difficult to imagine a freshwater 
lens on Motu Nautonga (1 ha), where an 11-m-tall Pisonia 
forest is found (Table 1 1). Further evidence for the salt-tolerant 
nature of Pisonia comes from Vostok, where a Pisonia 
forest, the sole woodland, extends to the edge of the shoreline 
rubble and herb mat. The trees, tightly pruned by wind and salt, 
have no buffer of coastal scrub. During storms, seawater 
reaches Vostok' s interior forest, yet this 24-ha island supports 
one of the largest and tallest (25 m high) groves in the Pacific 
(Clapp & Sibley, 1971b; Fosberg, 1977b and personal 
observation). 

Many Pisonia trees were heavily infested with scale 
insects (Coccidae) and Neuropteran larvae (Chrysopa sp.), 
identified by Dr. Scott Miller (Bishop Museum, Honolulu, 
Hawaii). This appears to be a natural phenomenon, as they 
were also abundant on the virgin Pisonia forests on Vostok and 
also on secondary Pisonias at Flint. 

Relationships Between Pisonia Forest Height and Motu 
Dimensions : Contrary to expectations, the tallest, most mature 
forests did not all occur on the largest motus (Table 1 1 ). The 
three prime forests (90-100% canopy cover) are on Nake 
(107.46 ha), Pig (7.21 ha), and Booby (0.84 ha). Trees on 
Booby are smaller in girth than those on Nake and Pig, but their 
height (20 m) is impressive; as far as we know, Booby was 
never cleared. Fine forests occur on other small, undisturbed 
motus; for example, Pisonia grew to 14 m on Raurau (3.48 ha) 
and to 1 1 m on Kimoa ( 1 .80 ha). 



31 



A positive correlation exists between Pisonia height and 
island width (Fig. 32). On Caroline, motus were 90 m wide 
before closed canopies of 13 m developed (Fig. 32), and tree 
height increased to 21 m with islet width up to 200 m (Pig, 
topmost star in Fig. 32 ). Further increases in islet width did not 
result in taller trees. However, even on motus with sufficient 
w idth, Pisonia did not develop unless other environmental 
conditions were suitable. For example, on Long, Pisonia only 
occurred in the centers of its former islets, not in the scrubby 
areas where coalescence is more recent. Tridacna. seemingly 
excellent for Pisonia, has not yet recovered from its Cocos 
plantations. 

Pisonia-Seabud Relationships : Seabirds are an integral 
part of Pisonia ecology. Its sticky seed capsules adhere to the 
feathers of, and are thus dispersed by, seabirds such as terns, 
boobies, and frigatebirds; thus, its early appearance on small 
motus is not surprising. 

On Caroline, six species of seabirds nest in its branches, 
dropping considerable guano to the ground below. Black 
noddies, amassing in dense colonies, nest almost exclusively in 
Pisonia, along with brown noddies, white terns, great and 
lesser frigatebirds, and red-footed boobies. Pig Islet, with 
7.25 ha of excellent Pisonia forest, supported a dense colony of 
nearly 2,000 pairs of black noddies (Subchapter 1.2. this 
volume). Bristle-thighed curlews feed on the ground beneath 
its open understory. and the long-tailed cuckoo forages within 
ils canopy. 

Seabirds may be so much a part of Pisonia ecology that a 
debate exists as to whether Pisonia actually requires guano for 
successful germination and establishment of seedlings (Shaw, 
1952; Fosberg. 1953;Wiens. 1962). Very high phosphate and 
nitrogen levels arc associated with mature Pisonia, and 
concurrently the development of Pisonia forest results in 
great!) modified soils that perpetuate its existence (Wiens, 
1962; Spicer & Newbery. 1979). The formation of a highly 
acid raw humus on the surface of the ground, sometimes in 
association with phosphatic hardpan. has also been documented 
on several atolls by Fosberg ( 1953. 1956). including Vostok 
(AKK and John Phillips, personal observation). Perhaps 
Pisonia's present distribution, primarily restricted to uninhabited 
islands ( Shaw, 1 952; Wiens. 1 962 l.isin part due to the fact that 
its primary seed carriers, seabirds. rarely coexist for long with 
num. 

Remnant Pisonia Forests in the Pacific : Though naturally 
and widely distributed from the western Indian Ocean to the 
eastern Pacific (excluding Hawaii), Pisonia grandis has become 
increasingly rare this century (Fosberg. 1953 and personal 
i ommunication). Occupying the interior of most atolls, it may 
have formerly covered the greatest area of any tree species in 
the Pacific (Wiens, 1962). Shaw (1952), summarizing its 
distribution, stated that it only occurs on remote, generally 
uninhabited islands ranging from the western Indian Ocean to 
the eastern Pacific, including Malaysia. However, more recent 
studies, particularly bv Fosberg, indicate that because its habitat 
occupies, and is in part responsible for. the most fertile areas of 
inhabited islands, its formerly extensive forests have been 
largely replaced by coconuts. Though Pisonia's soli wood is 



of little use to either atoll inhabitants or to the timber industry, 
its soils were rich sources of phosphate fertilizer and w ere thus 
greatly disturbed during the guano mining era. 

One of the most extensive Pisonia stands in the Pacific 
( 1 3.5 ha on Vostok) was partly burned in 1977 by members of 
a "scientific expedition" (Fosberg. 1977b). The Royal New 
Zealand Air Force found it smoldering 3 months later (Fosberg, 
1977b, personal communication). In a March 1990 visit to 
Vostok, we found that approximately 1.5 ha were completely 
cleared ( Kepler. 1 990c ), and a further unknown amount of land 
was affected. Other excellent groves exist on Palmyra and 
Washington (northern Line Group land on Nikumarorol Phoenix 
Group). Flint (Southern Line Group); Christmas (northern 
Line Group); Bikar, Jemo, and Ujae (Marshall Islands): and 
Aitutaki, Penrhyn. Suwarrow, and Manihiki (Cook Islands) 
have relatively small stands. Caroline, with 62.73 ha in Pisonia 
forest (36.94 ha in monotypic groves) holds some of the finest 
representatives of this ecosystem in the Pacific, even though 
much of it is not virgin. 

Coconut Woodlands (96.14 ha) (Figs. 14.36: Pis. 18.23.24. 
28-30.33.34.37.39.40.44) 

General Distribution : Cocos. although present on 15 
motus and known historically from another 4. covers significant 
areas only on Caroline's 2 largest islets. South and Nake 
(Table 13). Individual trees and small groves elsewhere are 
drift-derived or remnants of plantings made from 1916-1 920. 

The following species occur in habitats containing Cocos 
(Table 2): 

Trees: Pisonia grandis, Morinda citrifolia, Pandanus 
tectorius, Cordia subcordata, Cocos nucifera, Thespesia 
populnea, Hibiscus tiliaceus; 

Shrubs: Tournefortia argentea, Ximenia americana; and 
Herbs: Boerhavia repens, Portulaca lutea, Laportea 
ruderalis.Achyranthescanescens, Phymatosorusscolopendria, 
Ipomoea macrantha, Lepturus repens, Taccaleontopetaloides, 
Psilotum nudum, Phyllanthus amarus, and Sida fallax. 

The distribution of Cocos (Fig. 14). in order of decreasing 
abundance is as follows: South: Forests old and neglected. 
Healthiest palms line the lagoon, currently shading out strip of 
native scrub. Nake: Southern forests (50-80% Cocos) healthier, 
younger, with more native trees and Pandanus than on South: 
grove of about 50 palms on northeast. Long: Range from < I ' i 
cover (Tr. C) to dense fringe adjacent to lagoon. Emerald; 
Northeast and center-west patches. Mannikiba: Main grove. 
northeast: 40 palms. 20 m high, another patch in south centei 
Ana- Ana: House site, northeast point. Bird. Blackfin, Brothers. 
Nautonga. North Brothers. Pig. Pisonia. Raurau. Shark: lew 
trees each, primarily in Tournefortia. Lone Palm: One tree, 
central forests. 

History : A relatively small coconut grove was planted on 
South Island prior to the 16th century by Tuamotuan settlers 
(Emory, 1947; Maude, 1968). In 1 606. deQuiros noted -plenty 
of palms" and "many cocoa-nuts" (Markham, 1904). Since 
then, every visitor has recorded them as they grew, and still 
grow . adjacent to the boat "landing." A smaller grove evidently 
also existed in the south-southwest portion of South Island 



32 



(Lucett, 1851). According to Maude (ca. 1938). palms were 
also periodically planted — and destroyed — "by whalers and 
other chance visitors to the island." 

Until Arundel's arrival in 1885, Cocos was basically 
confined to this single grove in the northwest sector of South 
Island (Maude, ca. 1942a). In 1885, land clearing began, and 
from then till 1929, nearly 38,000 palms were planted, 29,480 
between 1916 and 1920 and another 7.000 young trees after 
1927 to replace thousands that had perished (Young, ca. 1922). 
Arundel's initial license gave him the exclusive rights to 
occupy Caroline and Flint, planting coconuts and other trees 
for 2 1 years, in return for an annual rental of 50 pounds ( Maude, 
ca. 1942a). In 1929. 13,215 trees remained, after which no one 
has counted them. Our field work and scrutiny of aerial 
photographs indicate that far fewer exist today. 

Caroline's plantations produced copra periodically from 
1 873 to 1934, but never profitably. They suffered greatly from 
the atoll's abandonment from 1901-1916. Dying and poorly 
planted palms presented continual setbacks ( Young, ca. 1 922), 
and in 1878 a hurricane wrought great destruction (N.I.D., 
1943). In addition, plantation managers lamented their poor 
productivity due to choking "by undergrowth and Pohue Vine 
[said to be Tuumfetta (- Triumfetta) procumbens, most likely 
a misidentification of lpomoea mac rant ha], destruction of 
inflorescences by great numbers of seabirds which roosted in 
the tops and broke off the flowers as they appeared," disease, 
and ruination of nuts by Polynesian rats and coconut crabs. As 
a result of this, the resident laborers slaughtered many crabs, 
and "greatly reduced the numbers of sea birds, who migrated to 
unoccupied islets." The rat problem was never resolved and 
appears to be the major reason for repeated failure of the 
plantations on both Caroline and Flint. Their enormous numbers 
and voracious eating habits greatly reduced both the crops of 
potentially healthy nuts as well as the volume of dried copra. In 
1920, 4,600 were trapped on South Island, and hundreds more 
were killed by small terriers introduced specifically to control 
them (Young, ca. 1922). Maude (personal communication) 
recalls that one terrier still survived in the 1940's. Rats still 
abound, especially within coconut groves and Pisonia forests. 
Another serious problem was due to coconut crabs digging up 
recently planted nuts and also their habit of pinching off young 
developing shoots. Evidently after the palms had attained one 
year's growth this was no longer a problem (Young, ca. 1922). 

Before abandonment (1902 to 1916, and after 1934), 
Caroline's plantations were owned by several companies whose 
average annual copra output was approximately 14 tons. From 
1934 to the 1970's, copra was harvested sporadically by small 
parties from Tahiti (Garnett. 1983). but within the last 2 
decades it stopped altogether. 

Despite the relatively fertile soils of South Island, the 
problems in the plantations hampered the establishment of 
permanent settlements on Caroline. In the 1930's, Maude 
estimated that the atoll could support 400 Gilbertese. increasing 
to over 1 .000 "when the island has been fully planted" ( Maude, 
ca. 1938). However, colonists were never established, leaving 
Caroline "one of the least spoiled islands in the Pacific" 



(Stoddart, 1976). As Young's (ca. 1922) unpublished 
"Memoranda" indicate, Caroline is not as pristine as it appears; 
however, the rapid comeback of many of its natural forests on 
the windward side is remarkable (see discussion under Pisonia 
Forests, this section). 

Distribution and Abundance : We recognize four 
subdivisions of the coconut woodlands: Cocos Plantations, 
Dying Cocos-lpomoea Plantation, Scattered Groves on Small 
Motus, and Mixed Forest with Cocos. 

1. Cocos Plantations (34.07 ha) 

Palm forests now dominate South Island and southwestern 
Nake. Although the planting of Cocos on South altered most 
of its original habitats, Nake escaped with less damage: whereas 
Cocos covers 77% of the area on South, it takes up only 6% of 
Nake (11% including mixed forests). The 60 to 100-year-old 
trees form tall, closed canopy woodlands (PI. 24) 21-25 m 
high, the customary maximum height recorded for old 
plantations (Fosberg, 1953). Figure 51 shows the distribution 
and abundance of plant species along a transect running centrally 
through the island, while Fig. 36 depicts a schematic profile of 
the same swath. 

Pure coconut plantations (like all habitats on Caroline) 
harbor relatively few species: up to 7 trees, zero to 2 shrubs, and 
5-11 herbs. The ground vegetation and shrub layers are 
composed almost exclusively of indigenous species, an unusual 
feature. However, skirting the edge of South Island's lagoon, 
tall palms overhang the water and crowd out native plants; 
there were considerably fewer Suriana and Tournefortia in 
1988 (PI. 29) than in 1965 (PL 40). 

2. Dving Cocos-lpomoea Plantation (53.92 ha) 

Mature plantations characteristically become overgrown 
with shrubs and vines (Fosberg, 1953. 1956). lpomoea 
macrantha. the sole vine on Caroline, forms tangled, 
impenetrable thickets. Indigenous, nonparasitic, and widely 
dispersed by ocean currents, it occurs naturally in small numbers 
in natural habitats on Caroline, but grows rampantly in disturbed 
areas. These vine-covered coconut woodlands cover two- 
thirds of South Island's interior (Fig. 50). The Dying Cocos- 
lpomoea forest is moribund. It is bordered by a belt of living 
palms, which in turn are sheltered by a narrow rim of indigenous 
vegetation (Figs. 36.51). 

While surveying the South Island transects, the authors 
stomped over intertwining thickets up to 3 m high (PI. 7) and 
crawled through tightly-knit masses of vines descending from 
the crowns of old palms, Pisonia, and Morinda bushes, until 
this too, proved impenetrable. In sunny clearings dotted with 
dead or dying palms, lpomoea, Boerhavia, and Phymatosorus 
proliferated luxuriantly, lpomoea, one of the prime reasons for 
the double abandonment of copra production, will continue to 
destroy the coconuts, encouraging natural ecological succession 
to begin anew. 

3. Scattered Groves on Small Motus (0.82 ha) 
Drift-derived palms were observed as long ago as 1834 

(Bennett, 1840). In 1916, when planting operations were 
commenced after a break of 14 years, about 40 trees grew 
beyond the plantations (Maude, ca. 1942a). Today, small 



33 



Cocos groves, up to 50 palms, drift-derived and plantation 
remnants, generally close to the shoreline (Pis. 29, 30), occur 
on 1 1 motus. 
4. Mixed Forest with Cocos (6.24 ha) 

This forest type is a simplified version of more complex 
and varied mixed forests that occur on most inhabited atolls. 
Composed of both anthropogenic and native elements, it 
contains a high proportion of Cocos (50-80%) mingled with 
variable proportions of Tournefortia. Pisonia and Pandanus. 
This forest type occurs primarily in southern Nake (Fig. 14), 
but also on Emerald, Shark, and southwest Long, where it 
mixes with Cordia and Tournefortia. 

House Site: A single clearing on Motu Ana-Ana, 
approximately 40 m x 70 m, contains a few Cocos adjacent to 
a vegetable garden and thatched living quarters (PI. 53). 

Associations with Seabirds : Coros-dominated habitats 
were the most depauperate on Caroline: only brown noddies 
and white terns breed there (PI. 54). The noddies nest high 
within the frond and inflorescence bases, whereas the white 
terns preferred lower sites. The absence of other species 
suggests that the anthropogenic Cocos forests seriously inhibit 
seabird use and may continue to do so for decades until they are 
replaced by native vegetation. 

Absent Plant Communities 

Caroline's impoverished flora and simple geology has 
resulted in a limited variety of ecosystems. The atoll is thus 
notable not only for its grand Pisonia forests, extensive 
monotypic stands of Tournefortia, and presence of Cordia 
groves, but also for the absence of several ecosystems that are 
generally considered typical of Pacific atolls: 
/. Sesuvium flats; 

2. Pemphis, Scaevola, and Sida scrub (two Scaevola plants 
are present, and the only two Sida records are from 1884 and 
1990); 

3. Barringtonia, Calophyllum, Guettarda, Hernandia, and 
Ochrosia forests; 

4. Plant associations (except Cocos) typical of native cultures 
on atolls: breadfruit groves (Artocarpus alt His), taro pits 
( Cyrtosperma chamissonis, Colocasia esculenta, Xanthosoma 
sagittifolia), cultivated ornamentals (Hibiscus rosa-sinensis. 
Plumeria spp., et cetera), or weedy grasslands/wastelands 
(Paspalum, Sporobolus. Wedelia, Vigna, et cetera). Even 
widespread introduced strand species such as Terminalia 
catappa and Casuarina equisetifolia are absent. 

In addition, there are no mangroves, peat bogs, marshes, 
ponds, salt flats, or other habitats associated with fresh or 
brackish water. Poorly represented are: 
/. Lepturus grassland. Although Lepturus is present in 
coastal herb mats, and occasionally in patches w ithin the forest 
understory, it does not form a separate plant community. 
However, it may once have covered the extensive clearings on 
South Island (PI. 2). 

2. Mixed forest. Though 6.24 ha of mixed forest (with 
Cocos) occurs (primarily on Nake), it is of such minor 
importance to Caroline's overall vegetation that it is treated as 
a subsection of coconut woodlands. 



Description and Ecology of the Motus 

These islet accounts synthesize the history, physiography, 
vegetation patterns, ecology, seabird colonies, miscellaneous 
biota, and the effects of human activity (if any) on Caroline's 
39 motus (Fig. 2). Mapping is based on the coast-to-coast 
transects, perimeter surveys, complete surveys (smaller motus), 
color transparencies, and aerial photographs. 

All motus are detrital reef islets representing many 
evolutionary stages from barely emerged coral rubble to large 
islets with relatively fertile "soils" supporting lush vegetation. 
There is one tiny old reef platform in its final stages of erosion. 

We discuss and map them in geographic order beginning 
in the north with Nake and progressing down the windward reef 
through Long and the 13 Windward Islets to South Island. 
Beginning anew in the north, we move south through 7 South 
Nake Islets, 1 1 Central Leeward Islets, and finally the 
5 Southern Leeward Islets. 

Because of the variety of islet shapes, "long" or "length" 
refers to the longest dimension lying parallel to the outer reef 
edge (normally north-south) and "wide" or "width" to the 
longest dimension perpendicular to the outer reef edge (normally 
east-west). South Island, the only exception, is considered to 
lie adjacent to the southern reef edge, so its "length" is measured 
east-west. Seabird numbers are from Subchapter 1 .2. Table 1, 
this volume. For convenience in locating particular islets, the 
order is as follows: 

/. Nake Island (Fig. 37) 
2. Long Island (Fig. 38) 



indw; 


ird Islets 


3. 


Bo'sun Bird (Fig. 42) 


4. 


Windward (Fig. 43) 


5. 


Crescent (Fig. 43) 


6. 


Atibu (Fig. 43) 


7. 


North Pig (Fig. 44) 


8. 


Pig (Fig. 44) 


9. 


Skull (Fig. 441 


10. 


North Brothers (Fig. 44) 


II. 


Brothers (Fig. 44) 


12. 


Noddy Rock (Fig. 47) 


13. 


North Arundel (Fig. 47) 


14. 


Arundel (Fig. 47) 


15. 


Tridacna (Fig. 48) 


16. 


South Island (Fig. 50) 



South Nake Islets 



17. 
18. 
19. 

20. 
21 
22 
23 



Pandanus (Fig. 52) 
Danger (Fig. 52) 
Booby (Fig. 52) 
Coral (Fig. 52) 
Lone Palm (Fig. 52) 
Kota(Fig. 52) 
Mouakena (Fig. 52) 



34 



Central Leeward Islets 

24. Mannikiba (Fig. 53) 

25. Blackfin (Fig. 54) 

26. Matawa (Fig. 54) 

27. Emerald (Fig. 54) 

28. Shark (Fig. 55) 

29. Scarlet Crab (Fig. 55) 

30. Nautonga (Fig. 55) 

31. Azure (Fig. 55) 

32. Reef-flat (Fig. 55) 

33. Bird (Fig. 55) 

34. Fishball (Fig. 55) 

Southern Leeward Islets 

35. Raurau(Fig. 57) 

36. Eitei (Fig. 57) 

37. Pisonia(Fig. 57) 

38. Kimoa(Fig. 57) 

39. Ana- Ana (Fig. 57) 

/. NAKE ISLAND (91.72 ha) (Figs. 30,37; Pis. 
35-37,43) 



18,23. 



History : Nake's large size and underground water lens, 
coupled with topography and soils more varied than elsewhere 
on Caroline, attracted early Polynesian settlers. Because early 
European visitors stayed primarily on South Island, there is 
only a single reference to Cocos prior to the late 19th century 
(one tree seen in 1825 by Paulding [1831]). 

The far northwest of Nake (also called North Island in 
Young, ca. 1922) houses the most important archaeological 
site on Caroline — a large marae (Figs. 3,37; PI. 36). Discovered 
during the guano era, the site is marked as "graves" on Arundel' s 
map. Arundel, who was living on the atoll when the marae was 
discovered, describes it thus: "On the north-west end of 
Caroline are some curious old native remains, whether places 
of burial or of sacrifice I cannot determine. I opened one of 
these, but could find no indication whatever to guide me in a 
decision" (Arundel, 1890). The senior author, R. Falconer, and 
G. Wragg located, measured, and photographed this marae in 
1990. The entire courtyard was approximately 18 m long by 
14 m wide. All 10 peripheral stones and the central one were 
easily identifiable from the 1883 plan (Fig. 3), although a few 
had fallen over or broken due to encroaching vegetation. The 
lower wall, partly destroyed by Arundel, had not been 
reconstructed. It is probable that this marae had not been seen 
since the 1 880' s; though discussed by Emory (1947), he never 
visited Caroline personally. 

Northwest Nake is particularly suitable for a place of 
worship and sacrifice: it fits most of the environmental criteria 
indispensable to ancient Tuamotuan religious ritual (Emory, 
1947). First, flat ground was necessary, preferably lying at 
right angles to, or parallel to, the lagoon. Second, it was 
important to have the wind blowing across the marae to waft 
away the smells of sacrificed animals. Third, ceremonial items 
included branches of the Pisonia tree, leaves of Cocos (for leaf 



charms/" rosaries"), and the aerial roots oiPandanus. Fourth, 
feathers from "black terns" (black noddy), frigatebirds, and 
red-tailed tropicbirds were also necessary for rituals. Rather 
than a smooth substrate, the early Polynesians would have had 
to be content with leveled coral rubble and distance from the 
lagoon. The only organism not living near the marae today is 
the tropicbird; however, their elongated tail feathers could 
have been plucked from adults nesting on nearby motus. 

Since marae are sacred places, there is possibly a meaning 
to the location of the main "courtyard" close to the atoll's 
northern tip. Generally the northern extremities of islands were 
auspicious places for Polynesians; it is here, they believed, that 
disembodied spirits were whisked to the netherworld. 

Physiography : Largest in area, Nake is the northernmost 
motu, separated from Long by a 40-m channel (PI. 18). With 
maximum dimensions 2,000 m long and 685 m wide, it is 
basically rectangular with rounded corners and a peninsula- 
like extension in the southeast. 

Nake lies north of the lagoon, having a southern "bay" 
(which we named Sandy Inlet), in which silt, sand, and fine 
coral debris are being actively deposited (PL 23). Sandy Inlet, 
, a hard, flat expanse of fine lagoon mud and sand, is 145 m wide 
at its mouth and extends 200 m north into the main islet. Its 
3.50 ha provide a favorite feeding location for shorebirds, 
especially bristle-thighed curlews. If Arundel's chart (Fig. 4) 
is correct, Sandy Inlet has increased its land area during the last 
century. 

On the reef flats off the west side are extensive remnants 
of jagged upraised reef (PI. 1 1) and occasional beachrock. The 
exposed beaches and reef flats at Nake's north point are 
especially broad, characteristic of reef flats at the exposed 
corners of islands. Comparisons of the northern sweep of 
rubble on recent aerial photos with Arundel' s map indicate that 
much coral debris, in the form of raised ridges (PI. 17), has been 
added since 1883. In the deep, fine coral rubble mixed with 
sand east of the marae, the 1990 expedition found three old 
turtle nests. Overall, unvegetated coral rubble, mud, and sand 
account for only 6% of the land area. In addition, some sparsely 
vegetated expanses of hardpan were noted in the south-central 
sector, just inland of the coast within a belt of Tournefortia 
forest. 

Nake's windward coast, complete with a peaked beach 
crest and discontinuous beachrock, is 30 m wide in the north, 
narrowing to 3 m in the south. Offshore, submerged reef flats 
form a sandy moat bordered by a barrier reef upon which waves 
pound incessantly. 

In the distant past, what we now call Nake consisted of two 
separate motus. Aerial photos (Chapter Frontispiece) reveal an 
oblique, ancient channel about two-thirds of the way down the 
islet. It is now well vegetated in the center but scrubby near the 
coastlines. 

Vegetation : There are 16 plant species (5 trees, 1 shrub, 
10 herbs), 59% of Caroline's flora. Nake is the lushest motu. 
Its woodlands (82.39 ha) are about 80% native and 20% with 
Cocos (PI. 37). Although in 1916 there were about 260 palms, 
and the entire island was evidently planted with 10,544 palms 
in 1918-1919(Young,ca. 1922, Table 13), substantial tracts of 
each major vegetation type occur today. Its interior is rich in 



35 



Pisonia, with the largest acreage (20.79 ha) and some of the 
tallest trees (20 m high) on the atoll (PI. 43. Table 1 1 ). In 
addition, Cordia is well represented: two major groves of 
Cordia-Toumefortia forest occupy 11.8 ha, 2% of Nake"s 
area. Extensive pioneer herb mats. Hanked on their inner sides 
by Tournefortia scrub, occur in the north and east. The 
remaining Cocos, essentially in the southern quarter, comprise 
Caroline's second largest coconut grove. 

Birds : Nake, with 80% of Caroline's breeding seabird 
species, shows a direct correlation between islet size and bird 
species diversity. Nine species of seabirds breed, all with 
larger populations (pairs) than previously reported (Clapp & 
Sibley, 1971a): masked booby (105), brown booby ( 1). red- 
footed booby (496), great frigatebird (522), lesser frigatebird 
1 56), brown noddy ( 390), black noddy (8 14), sooty tern (nesting 
in 1989; Anne Falconer, personal communication), and white 
tern (1,094). 

2. LONG ISLAND (75.98 ha) (Figs. 30,35,38-41; 
Pis. 8,13,18,20,28,33,47,58) 

Third largest in area, this longest of motus covers nearly 
one-third of the atoll's windward side. In the north it is 
separated from Nake by a narrow channel; from its southern tip 
a chain of smaller motus extends south along the windward 
reef. 

Physiography and History : Long — 4.226 m long and 
330 m wide — is somewhat snake-shaped, with an enlarged 
northern "head" and attenuated "tail." From a distance its 
vegetation appears as a series of humps. Long has experienced 
a fairly complex geological history, noted by the Solar Eclipse 
Party: "On some of the islands there are spaces void of 
\ egetation, extending from lagoon to beach, which indicate the 
existence at a former time of a water separation" (Holden & 
Qualtrough. 1884). 

At present. Long is composed of five distinct former islets 
separated by sparsely vegetated channels of coarse sand and 
coral gravel. Aerial photographs also reveal further, older 
subdivisions (discussed below). Coalescence and fracturing of 
the original motus have probably occurred repeatedly. Since 
erosion proceeds faster on an atoll's windward reefs, providing 
coral fragments, coralline algae, and pulverized mollusks, il is 
no surprise that the first series of Caroline's motus to fuse were 
those facing this rich source of parent material. 

Long's coarse rubble beaches (Pis. 13.20) are a mirror 
image of those on Nake: southward, the) widen progressively. 
The swath of unvegetated rubble above high tide line in the 
upper two-thirds of Long averages S m wide, while in the lower 
third il is 40 m wide. Unvegetated coral debris accounts for 
10% of the island's area (Fig. 30). Beachrock, Hanking the 
windward shoreline for most of its length, is more abundant 
than elsewhere on the atoll (PI. 58). 

I ong's lagoon Hank is edged w ith submerged sand and silt 
and is line of the most sheltered parts of Caroline. Sand and 
rubble deposition off the south point has formed an islet in the 
lagoon ( Bo'sun Bird |, w Inch could, in the future, coalesce with 
Long's south point to form a hook. 



An uncommon substrate on Caroline, upraised reef 
( makatea > forms a low rampart ( generally < 1 m high ) paralleling 
the ridge crest inside the vegetation for much of the lower 
quarter of Long. 

In 1 990 Graham Wragg found some scattered large stones, 
similar to those of the marae on Nake. located centrally 100 m 
north of the southern tip of Long, confirming the report of the 
remains of a smaller marae on Long Island (Holden & 
Qualtrough, 1884). Wragg noted that the marae was indeed 
smaller than that on Nake. w ith dimensions approximately 3 m 
wide by 8- 1 m long. Its orientation appeared to be northeast- 
southwest. The wall on one end was evidently smashed by 
storm waves. Only two of the peripheral upright stones were 
still standing; they were of similar size to those on Nake. The 
platform was in reasonable condition, with a huge Pisonia tree 
growing through it. Some rock slabs were large (2 x 2 m). The 
entire marae was situated within a Pisonia grove, with some 
Cocos but no Pandanus nearby. We do not know if the nearby 
coconut grove ( 1 .6 ha) was present before 1 .343 palms ( 20% of 
the islet's area) were planted in 1918-19 (Young, ca. 1922). 
The sheltered location and a Pisonia— Cocos forest, which 
suggests an old clearing, further indicate prior occupation. 

In 1990. G. Wragg also uncovered an RNZAF survey 
marker just inland of Long's southernmost tip. 

Vegetation : There are 15 plant species (4 trees. 2 shrubs. 
9 herbs) on Long. 56% of the total flora. Long's variety of 
habitats, vegetation heights, substrata, and birds make it the 
most diverse islet on Caroline. Only 3% of its area remains in 
Cocos. All the atoll's seabirds have bred here. Its ecology is 
best understood with reference to Figs. 35 and 39—1 1 . 

Within the basic pattern of five coalesced motus. it may be 
seen that: 

/. From north to south (measured from the midpoint of 
each former channel) the motus, of divergent size and shape, 
are approximately 320. 620. 700. 1.840. and 100 m long. 

2. Each former islet, crowned by a Pisonia forest, contains 
concentric rings of decreasing fertility around its core and is 
morphologically similar to islets surrounded by water, except 
that the coarse coral gravel along the former perimeter is less 
marked. More specifically, beach sands and gravel extend for 
200-300 m north and south o( the old channels, after which 
they increasingly accumulate coral rubble, humus, and guano. 

3. The dominant vegetation is Tournefortia, interspersed 
with 4 patches of taller Pisonia forest and scattered clumps of 
Cocos md Cordia. Interrupted herb mats parallel the windward 
coasl and often extend across the island along former channels 
(PI. 33). Vegetation height varies from 2 cm to 15 m. 

4. Plant species diversity is highest in Tournefortia- 
Pisonia and lowest in Pisonia forests. 

5. Long's tallest, most mature Pisonia groves (up to L009S 
Pisonia) occur on the largest of the former islets. The Pisonia 
forest near the south end (Tr. 10). although healthy, is only 
12 m tall. This may be due to its impoverished makatea 
substrate of pitted reef rock barely covered with "soil." Since 
it lies adjacent to Long's most luxuriant Cocos grove, its land 
could well have been cleared in 1918-19. with the Pisonia 
forest taking longer than elsewhere to recuperate. Because tern 



36 



guano increases soil fertility and is important for Pisonia 
growth (Fosberg. 1953). it is of interest that neither black nor 
brown noddies nested here. 

ft. Deep dips in Fig. 39 (lower graph) correspond to east- 
west corridors formed from old channels. Vegetation in these 
relatively infertile, sandy flats is low. similar to that on small 
developing motus (i.e.. native herbs with scattered Tournefortia 
<2 in high). One sandy channel (Tr. C; PI. 33) supported a 
sparse population of Suriana. During the February 1990 
cyclone, all vegetation was either uprooted, washed away, or 
smothered with fresh sand and coral gravel along Trs. A and C 
(personal observation, March 1990). Storm erosion was 
particularly marked within the channel that almost bisects the 
island (Tr. A). 

7. Secondary dips mark even older interislet channels 
("ancient channels"), visible on aerial photographs (Chapter 
Frontispiece) but barely recognizable in the field. They are 
overgrown with Tournefortia and/or Pisonia. 

8. Sharp dips within established forests or herb mats 
denote relatively recent channels gouged out by storms ("recent 
storm cuts"). These were also altered during the winter 1990 
storm. 

Figures 40 and 41 illustrate some differences between the 
windward and leeward coasts. Transect C (Fig. 40) crosses the 
north end of Long through an old interislet channel now filled 
with sand and rubble. Its low profile reflects the simple habitat 
harboring halophytic herbs and Tournefortia shrubs less than 
2 m high. Although the shrubs are scattered, the lagoon half of 
the transect passes through slightly higher ground, which 
encourages denser Tournefortia. This transverse section is 
similar to that of a formative motu such as Fishball (Fig. 56). 
This exposed, scrubby swath, 300 m wide, harbors red-footed 
boobies, great frigatebirds, and a discrete population of masked 
boobies. Approximately 127,000 pairs of sooty terns nested in 
a similar sandy channel 740 m to the south (Tr. A, PI. 59) in 
1988. 

Transect 8 (Fig. 4 1 ) crossed the islet nearer the southern tip 
(Fig. 8). This profile departs significantly from the usual 
parabolic cross-section seen on most of the small motus and 
which exists further north on Long Island. From east ( windward ) 
to west, there is first a wide expanse of coarse, unvegetated 
rubble, followed by rubble dotted with herbs, then Tournefortia 
scrub increasing to 9 m high. Further inland, a forest of 
10-m-high Tournefortia. Pisonia. and Cordia continues 
westward to the lagoon. This leeward margin of Long, extending 
southward nearly to its tip, is the only location on Caroline 
where tall, indigenous vegetation overhangs and shelters the 
lagoon. No herb mat is present. 

In summary. Long contains examples of all major plant 
communities, as well as two minor ecosystems, Pisonia- 
Cordia (3.2 ha) and Cocos-Cordia (0.82 ha). Its woodlands 
total 49.60 ha. Coconut crabs inhabit all areas containing 
Cocos and Pisonia; our rough estimate of their population is 
200 crabs. 

Birds : In 1988, Long supported 9 ( 10 in 1965) species of 
breeding seabirds. as follows (pairs): red-tailed tropicbird (5). 
masked booby (69), brown booby (12), red-footed booby 
(659), great frigatebird (808), sooty tern (179,800). brown 



noddy (207), black noddy (986), and white tern (751 ). From 
1988 through 1990, sooty terns occupied 19 large colony sites 
(Fig. 11, Subchapter 1.2). 

Comments : Polynesian rats were abundant, especially in 
Cocos and Pisonia habitats. It was often possible to see 3 or 4 
simultaneously while conducting daily surveys and 20 or more 
around camp. At night, their numbers increased substantially. 
Azure-tailed skinks (Emoia cyanura) were noted. 

Windward Islets 

This chain of 13 islets occupies the southern half of 
Caroline's east coast. All rest on the same reef flat, separated 
by surge channels varying in width and depth. They can be 
waded with care at low tides, but most harbor black-tipped reef 
sharks: up to four were visible in the shallows within 50 m of 
an observer. 

The motus range in size from Noddy Rock (0.02 ha) to 
Windward (1 1.42 ha). They support every major vegetation 
type, from simple herb mats to magnificent Pisonia forests, 
21 m tall. Because of their constant exposure to trade winds, 
the seaward vegetation is wind- and salt-shorn. Though 
appearing completely natural, all of the Windward Islets were 
intensively planted with Cocos (Table 13) from 1916-1920 
(Young, ca. 1922). However, these incipient plantations 
experienced difficulty and appear to have been abandoned 
within a few years (see Plant Communities section). 

Flanking the lagoon of the southern motus (Brothers 
through Tridacna) and extending westward are coral reefs 
densely studded with giant clams, whose iridescent, 
multicolored mantles add to Caroline's outstanding natural 
assets (PI. 26; Subchapter 1.2, Conservation section). 

3. BO'SUN BIRD ISLET (0.86 ha) (Figs. 29,42; PI. 9) 

We named this motu for its red-tailed tropicbirds, 
commonly called bo' sun birds. The sizeable population is the 
largest on Caroline. In addition, our 1988 records constituted 
the first known breeding of this species on the atoll. 

Physiography : Bo'sun Bird Islet, 165 m west of Long's 
southern tip. is the only motu lying "within" Caroline' s lagoon. 
It shares the same reef as Long, however, and is not a true 
"lagoon motu." 

Amoeboid in shape, Bo'sun Bird is greatly affected by the 
tidal waters that spread across the shallow reef flats and gush 
through the surge channels that separate Long and Windward. 
Because it sits near the inner edge of a wide windward reef flat, 
the layering of sediments around it is complex and transitory; 
our observations indicate that more rubble was deposited on 
the islet's western edge since the aerial photos were taken in 
1985. Its western shoreline rises gradually to a high water 
mark, and slight changes in water level greatly change its 
overall size and shape. At high tide its perimeter resembles the 
shape of Pinocchio' s head — ovoid with a long, expanded nose. 
The "head" is approximately 70 m wide and 115 m long, while 
the "nose" is 45 m long and 15 m wide. 

Vegetation and Birds : Bo'sun Bird Islet, composed of 
coral rubble and sand, supports only natural herb mats 
(Heliotr opium, Portulaca. Lepturus) and Tournefortia scrub 



37 



(to 4 m tall). These two simple plant communities cover 35% 
and 55% of the land area, respectively. For its size, the motu 
is sparsely vegetated, with only four plant species (one shrub, 
three herbs), 15% of Caroline's total flora. There are no 
introductions. 

Bo'sun Bird's most notable attributes are its 4 species of 
breeding seabirds: red-tailed tropicbird (47 pairs in 1988, but 
130 pairs seen in 1990), sooty tern (8,400 pairs), brown noddy 
( 10 pairs), and white tern (6 pairs). 

4. WINDWARD ISLET ( 1 1.42 ha) (Figs. 29,43) 

We named this "Windward" because it is the first major, 
and largest. Windward Islet. 

Physiography : Windward is broadly crescentic in shape. 
508 m long by 287 m wide. It parallels the reef's longitudinal 
axis and is set close to the lagoon. Its seaward beach is quite 
narrow (3 m wide); there is no lagoon beach. 

Vegetation: Windward has 1 1 species of plants (3 trees, 
1 shrub, 7 herbs), 41% of the total flora. A windward crescent 
of halophytic herbs borders a zone of Tournefortia scrub, 
which mixes quite densely with Pisonia and Cordia over most 
of the interior in a bilobed pattern. These latter forests, 
reaching 14 m high in the south and 9 m in the north, total 
8.67 ha. This unusual distribution of central forests undoubtedly 
reflects Pisonia' s recovery from 1 00% land clearing for Cocos — 
1,299 palms— in 1920 (Young, ca. 1922, Table 13). It is 
remarkable that not one Cocos remains as a legacy of this 
disturbance. 

The east-west profile of Windward, similar to that of 
Tr. 8, Long Island (Fig. 41 ), is typical of most motus, except 
that lagoon-facing herb flats are almost nonexistent. Scaevola 
sericea sericea, a new plant record for the atoll, is unique to this 
motu, although S. s. tuamotensis was found on South Island in 
1990. 

Birds : Five species of breeding birds were present, all in 
appreciable numbers (pairs): red-footed booby (163), great 
frigatebird (207). brown noddy (20), black noddy (28), and 
white tern (134). 

Comments : In May 1990, AKK noted a possible motu 
midway between Windward and Crescent Islets during midtide. 
It appeared to be upraised reef like Noddy Rock, but because 
of extensive shallow reefs in this area, it has not yet been 
confirmed. 

5. CRESCENT ISLET (3.10 ha) (Figs. 29.43) 

We named this islet for its cupped shape. 

Physiography : Crescent Islet is 1 90 m long by 225 m wide. 
It is almost entirely composed of coral rubble, with a little 
humus in the interior. The seaward beach is variable (up to 
50 m wide), the lagoon beach, insignificant. 

Vegetation : There are 10 species (3 trees. 1 shrub. 
6 herbs), 37% of Caroline's flora. No introduced plants occur. 
Plant diversity is poorer than on Windward, a reflection of 
small size, poor soils, and scant herb mats. However, woodlands 
cover two-thirds of its area, and the centra] stand of Pisonia and 



Cordia is 87 m wide and up to 1 3 m high. Crescent was heavily 
planted (80% of total area, 228 palms) in Cocos in 1920. but 
today none remain. 

Birds : Crescent Islet was used by the following numbers 
of breeding pairs: red-footed booby (28), great frigatebird (5). 
brown noddy (36), black noddy (60), and white tern (8). 

6. MOTU ATIBU "Coral Rubble Islet" (0.02 ha) 
(Figs. 27,43) 

Motu Atibu was Caroline's smallest and least vegetated 
islet. Third in the windward chain, it measured 13 m x 18 m. 
We named it for its basic rubble character. Vegetation covered 
only 2% of the land surface and consisted of a few Tournefortia 
shrubs (<1 m high) encircled by narrow swaths of low herbs 
and rubble. Its three plant species (one shrub, two herbs) — 
1 1% of Caroline's flora — were among the most meager on the 
atoll. Atibu' s profile was similar to that of Fishball (Fig. 56). 
There were no breeding birds. 

Comments : Since a February 1990 storm. Atibu has 
apparently disappeared, having been reduced to a thin strip of 
coral gravel below high tide level. 

7. NORTH PIG ISLET (5.44 ha) (Figs. 29,44; Pis. 60.61 ) 

We named the fourth windward islet "North Pig" for its 
location immediately north of Pig Islet. 

Physiography : Classically crescentic. North Pig is 350 m 
long and 230 m wide. Though approximately half Pig's area 
and less wooded overall. North Pig has a similar distribution of 
substrates (including sand on the lee side), vegetation, and 
breeding birds. Profiles of the two motus are nearly identical 
(Fig. 45). 

Vegetation : There are 1 1 plant species (3 trees, 1 shrub. 
7 herbs), 41% of Caroline's flora. No introduced plants are 
present. Proceeding south along the windward islets, lagoon- 
side herb mats develop and islet cross-sections assume a more 
perfect symmetry — low at the edges and forming a hump in the 
middle. 

North Pig's three vegetation zones are predictably 
symmetrical: a peripheral band of herbs (more extensive on the 
"horns"), curved belts of Tournefortia, and a spacious central 
forest of mixed Pisonia, Cordia, and Tournefortia. The latter 
(to 20 m tall) covers more than one-half the islet's width and 
one-third its area and includes fine Cordia groves (Fig. 44). 
This excellent forest is surprising because 402 Cocos palms 
were planted on 93% of North Pig's usable land in 1420 
(Young, ca. 1922, Table 13). Measurements from 25 Pisonia 
trees (main trunks) averaged 19 m in height, 221 cm in 
circumference (at 1.5 m), and 261 cm in base circumference 
(Table 12). 

Birds : Five species of seabirds bred: red-footed booby 
( 3 1 pairs), great frigatebird ( 1 7 pairs), brown noddy (76 pairs). 
black noddy (3,199 pairs), and white tern (1 10 pairs). The 
largest colony of black noddies on Caroline nested in the tall 
Pisonias. 

Comments: Rats and coconut crabs were common. 



38 



8. PIG ISLET (7.21 ha) (Figs. 29,44; Pis. 27,41,52,60,61) 

Number 5 down the chain. Pig was named prior to 1883. 
Domestic pigs were introduced to Caroline in 1 828 by Captain 
Stavers but evidently died out before 1834. Reintroduced in 
1848 with the first recorded settlers, it is not known how long 
they lasted. One would expect that they were only on South 
Island, but the statement that "about one-third the distance up 
the lagoon a canvas hut exists on one of the smaller islets on the 
eastern side of the lagoon" (Holden & Qualtrough, 1884) 
suggests that perhaps domestic animals also inhabited Pig. 
Though this is weak evidence, there must have been some 
reason for this curious name. Today, fortunately, no pig 
devastation is evident here or elsewhere on the atoll. 

Physiography : Pig, shaped like a fat kidney bean, is 
330 m long and 255 m across. It is separated from North Pig 
by a channel 60 m wide. 

Vegetation : The islet has 1 1 plant species (4 trees, 1 shrub, 
6 herbs), 4 1 % of Caroline' s flora. Cocos, the only introduction, 
is rare (0.03 ha). In 1920, 538 palms were planted (Young, 
ca. 1922), which covered approximately 79% of Pig's usable 
area (Table 13). 

Pig's vegetation profile (Fig. 45) is classic: a wide, 
windward herb mat, bordered by Tournefortia and Cordia, 
which, in turn, grades rapidly into an outstanding Pisonia 
forest (to 21 m tall, 3.36 ha), one of Caroline's best groves. 
Measurements from five trees, mostly multiple-trunked, 
averaged 16 m in height, 338 cm in circumference (at 1.5 m), 
and 282 cm in base circumference (Table 12). This Pisonia 
also occupies the largest proportion (46%) of any islet area. It 
is striking that such quality forests could regenerate in about 
65 years (see section on Plant Communities, Pisonia Forests). 
In the Cordia forest (PI. 27), also the finest on Caroline, six 
trees averaged 12.6 m in height, 1 16 cm in circumference (at 
1.5 m), and 99.8 cm around their bases. On the lee side of Pig, 
Tournefortia extends directly to the lagoon shore. 

Birds : Five species of seabirds bred: red-footed booby 
(14 pairs), great frigatebird (118 pairs), brown noddy 
(82 pairs), black noddy (1,928 pairs), and white tern (164 
pairs). 

Comments : Rats and coconut crabs were common. In 
1 990 agrayish gecko (possibly mourning gecko, Lepidodactylus 
lugubris) was seen by A. Garnett. 

9. SKULL ISLET (0.02 ha) (Figs. 27,44; Pis. 46,49) 

Sixth in the windward chain, we named Skull Islet after 
finding the skull, tail feather, and eggshell of a red-tailed 
tropicbird, the first evidence that this species bred on the atoll. 
A low shelf of coral rubble and sand, barely above high tide 
mark, this motu is barren except for a small herb mat under five 
Tournefortia bushes ( 1 m high) on the lagoon side. Only 2% 
of the surface area is vegetated. There are three plant species 
(one shrub, two herbs), 1 1% of the atoll's flora. Although 
appearing young, the islet must be more than 100 years old, as 
it is marked on Arundel' s chart (Fig. 4). After February 1990, 
several large reef fragments had washed into the channel close 
to Skull Islet. 



In 1988 there were no birds. However, in March 1990, a 
colony of 150 brown noddies was in a prelaying phase, 
accompanied by 6 sooty terns, a brown booby and a wandering 
tattler. 

70. NORTH BROTHERS ISLET (1.71 ha) (Figs. 29.44: PI. 60) 

The seventh windward motu, we named this islet North 
Brothers because of its location directly north of the named 
motu, Brothers. 

Physiography : North Brothers is shaped like an oval that 
curves lagoonward toward Brothers, 40 m away. The concave 
shorelines and lack of herb mats on the opposite shorelines of 
these 2 islets suggest that they might have been 
formerly connected. Composed primarily of rubbly substrates, 
with slightly better soils centrally, it is 95 m long and 250 m 
wide. 

Vegetation : Plant species number 10 (3 trees, 1 shrub, 
6 herbs), 37% of Caroline's flora. A few Cocos trees are 
present, remains of the 180 planted in 1920 (Young, ca. 1922), 
which covered 100% of all available land on the islet 
(Table 13). Plant communities on North Brothers are simple: 
Tournefortia (more open in the west) rises to an excellent 
Pisonia forest, 80 m wide and 18 m tall, on the east end. 
Average measurements from three Pisonia trees were height 
18 m, base circumference 314 cm, and number of trunks, 2.3 
(Table 12). 

Birds : Five species of seabirds bred on the islet in 1988 
(pairs): red-footed booby (25), great frigatebird (9), brown 
noddy (23), black noddy (40, plus hundreds of old nests), and 
white tern (69). In September 1989, sooty terns nested on the 
windward beach (Anne Falconer, personal communication), 
and in May 1990, a prebreeding swirl of thousands of sooty 
terns swarmed above Brothers and North Brothers. 

Comments : Gecko eggs were seen on Pisonia trunks in 
1990. 

77. BROTHERS ISLET (4.31 ha) (Figs. 29.44.46: Pis. 30.60) 

The eighth windward motu. Brothers Islet was named last 
century after Captain Brothers, who managed a stock-raising 
venture on Caroline. In 1 873, his rights to the atoll passed into 
the hands of John Arundel. 

Physiography : Shaped like a molar tooth, with roots 
extending toward the lagoon, Brothers Islet lies about two- 
thirds of the way down Caroline's windward reef. It is 198 m 
long x 178 m wide through the center. A Tridacna reef extends 
westward from it almost completely across the lagoon. 

An interesting aspect of Brothers' structure is that Arundel's 
chart (Fig. 4) indicates a tiny, separate motu off the southwest 
point. Our survey and the 1985 aerial photos show that this 
motu is now joined to Brothers Islet. Its former identity is 
marked by a small patch of Tournefortia, around which the 
recently deposited sand and rubble is sparsely dotted with 
native herbs. 

Vegetation : There are 1 1 plant species (4 trees, 2 shrubs, 
5 herbs), 41% of Caroline's flora. Cocos, along the leeward 
shore, is the only introduced plant. Three distinct plant 



39 



communities are present: peripheral herb mats (including 
leeward Portulaca with Suriana), Toumefortia scrub and 
forest ( to 6 m high ) bordered with Cordia, and a central Pisonia 
forest. Larger trees had up to 15 trunks and multiple root 
suckers. Measurements of 10 trees (main trunks) averaged 
15 m in height, 140 cm in circumference (at 1 .5 m), and 243 cm 
base circumference. Distances to nearest neighbor for 10 trees 
averaged 4.2 meters. As on its neighbor islets, the Pisonia 
forest on Brothers is striking, especially since it has matured to 
a closed-canopy monotypic stand devoid of any subcanopy 
species (Fig. 40), evidently in about 65 years. In 1920, Brothers 
Islet was planted with 315 Cocos palms, which covered 
approximately 97% of the usable land area (Table 13). 

Birds : Four species of seabirds bred: red-footed booby 
(25 pairs), brown noddy (8 pairs), black noddy (15 pairs), and 
white tern (50 pairs). In May 1990, large numbers of sooty 
terns swirled over Brothers and North Brothers Islets. 

Comments : Many of the mature Pisonia trees contained 
capacious cavities in their boles that housed large coconut 
crabs. In March 1990, several of these holes had feathered 
skeleta of sooty terns (and possibly also brown noddies) 
outside their entrances, along with freshly-snipped Pisonia 
branches (see Subchapter 1.2, Coconut Crabs section). 

12. NODDY ROCK (0.02 ha) (Figs. 27.47; PI. 19) 

We named this ninth motu in the Windward Islets for its 
only known breeding seabird, the brown noddy. In September 
1988. at least 80 pairs were incubating their eggs on the 
Portulaca mat that covers its central lee section. 

Noddy Rock, an eroded limestone plateau (feo), is 26 in 
wide by 9 m long. It is windswept and salty, with waves 
splashing over its eastern edge on most days (it rises only 
0.5 m above high water). During storms it is completely awash 
( Anne Falconer, personal communication ). Only three species 
of plants (11% of Caroline"s flora) grow here, thinly covering 
the western (leeward) third of the island in the following 
proportion: 75% Portulaca, 20% Lepturus, and 5% 
Toumefortia). 

13. NORTH ARUNDEL (0.91 ha) discussed below 
(Figs. 29,34.47; Pis. 14.62) 

14. ARUNDEL (7.34 ha) 

Arundel Islet was named last century in honor of John T. 
Arundel. A British trader and guano digger. Arundel was one 
of the leading figures in the Pacific phosphate industry . directing 
guano and coconut planting operations on Caroline and other 
islands from 1873 to 1897. His most valuable contributions, 
however, were his excellent surveys and maps of several 
central Pacific islands, including Caroline (Fig. 4). The islet 
immediately to its north. Arundel's "cap." we named North 
Arundel. 

Physiography : Arundel's shape is a I at crescent, with 
wedge-shaped North Arundel lying across a short channel 
immediately to Us north. North Arundel is 80 m long x 1 30 m 



wide, while Arundel is 375 m long x 210 m wide. They are 
composed almost exclusively of coral rubble and Hanked on 
their inneredges by Acropora reefs heavily laden with Tridacna 
clams. Arundel's inner "horns" have evidently added more 
sand and rubble since 1883 (Fig. 4). 

Vegetation : There are 1 1 plant species (3 trees. 1 shrub. 
7 herbs) on Arundel, 41% of Caroline's total. There are no 
introductions. North Arundel has 1 1 (4 trees, 1 shrub. 6 herbs), 
41% of Caroline's flora, including one introduction, Cocos. 

The vegetation on this pair of motus, along with Tridacna 
to the south, consists of extensive herb mats, low scrub and 
small interior forests (Fig. 34), slightly less lush than the more 
northerly windward motus. Their woodlands are primarily 
Toumefortia, with a thin belt of Cordia and central Pisonia 
groves (a bilobed pattern on Arundel). Morinda is unusually 
common in Arundel's central forests, and Achyranthes is 
especially abundant on North Arundel. Pisonia occupies only 
13% of the land area on Arundel, compared to 46% on Pig. 
Soils are extremely rubbly . with scant organic matter, a possible 
legacy of the guano era. 

Both North Arundel and Arundel were heavily planted 
with Cocos in 1919-20 (69 and 646 palms, respectively). All 
usable land was cleared (Table 13). Despite the extreme 
paucity of Cocos today, the relatively scant Pisonia present 
today on these islets, compared to those further north, suggests 
that the Cocos plantations were more successful here. Their 
proximity to South probably also guaranteed better maintenance. 

Birds : Five species of seabirds bred on Arundel: red- 
footed booby (37 pairs), great frigatebird (on territory, 
September 1988; breeding confirmed, early 1989 by Anne 
Falconer), brown noddy (1 1 pairs), black noddy (249 pairs). 
and white tern (227 pairs). In May 1990, thousands of sooty 
terns swarmed above Arundel and North Arundel. 

15. TRIDACNA ISLET (9.08 ha) (Figs. 29.48,49; 
Pis. 1,26.48.62.63) 

The 13th and southernmost motu in the windward chain 
was named by the present authors and Boris Sirenko for its 
outstanding coral reef studded with giant clams {Tridacna 
maxima). 

Physiography : Somewhat crescentic. measuring 446 m 
long and 250 m wide, Tridacna is one of the largest motus on 
Caroline. Its ground surface is heavily littered with coral 
rubble, having a sandy strip above the beach crest on the 
windward edge. 

Vegetation : There are 1 3 plant species (2 trees, 2 shrubs, 
9 herbs), 48% of the atoll's flora. For its size, Tridacna's 
\ egetation is surprisingly lacking in tall forests, a legacy of the 
910 Cocos palms planted on 82% of its available land area 
(Table 13). Vegetation patterns follow the usual concentric 
zonation: peripheral herb mats border a discontinuous belt of 
Suriana (windward side), while the large central mass is 
dominated by scrubby Toumefortia— Morinda woodlands, 
which cover 88% of the islet's area, yet only attain 7 m in 
height. In cross-section (Fig. 49). the short woodlands are 
particularly noticeable. Compare the present lack of Cordia. 



40 



paucity of Pisonia, and richness of herbs, both in species 
numbers and abundance, with Pig (Fig. 45) and Brothers 
(Fig. 46). Although there are no introduced plants, thick 
patches of Lepturus also reflect past forest clearing. 

Birds : Fourspeciesof seabirds were nesting in 1988: red- 
footed booby (111 pairs), brown noddy ( 1 1 pairs), black noddy 
(249 pairs), and white tern (227 pairs). Tridacna is periodically 
a major breeding area for sooty terns. Clapp& Sibley (1971a) 
found 4 main colonies totaling 250,000 birds, and large numbers 
nested along the windward beach in August 1 989 (A. Falconer, 
personal communication). Nests were located under shrubs, or 
in open areas bordering them, and were evidently preyed upon 
by coconut crabs. 

16. SOUTH ISLAND (104.41 ha) (Figs. 30,36,50; 
Pis. 1-7,12,16,24,34,39,44,45,50,62) 

History : The history of South Island (called Rimapoto in 
Young, ca. 1 922) is essentially the history of Caroline, for most 
information about the atoll prior to 1965 is from here. It is the 
second largest islet, and the staging area for trips up-lagoon as 
it lies adjacent to both the "boat landing" and "blind passage." 

South Island was inhabited in prehistory by Tuamotuans, 
who planted the first small coconut grove on its northwest 
point. The first Europeans to land, in 1606, found coconuts, 
fish, lobsters, and seabirds in abundance. They dug for fresh 
water in vain. Two hundred years later, in the decade after a 
cyclone in 1825, pigs, sweet potatoes, arrowroot, and South 
Sea chestnut were introduced. However, "the unfriendly 
character of the soil, and the number of land crabs that infest it, 
gave us but little hope of the experiment succeeding" ( Bennett, 
1840). The pigs expired within a few years. The arrowroot, 
tenacious and adapted to island environments, still exists today 
(unless later immigrants brought it). Of the others — plus many 
other later food plants and ornamentals — no trace exists 
(Table 1). (In 1990, we found a few Hibiscus tiliaceus, 
Thespesia populnea, and Ximenia americana. All could be 
indigenous. The first 2 species were often planted by Polynesians 
in copra-cutting settlements [e.g., Flint Island]). Tropical heat, 
droughts, storms, excessive shade from Cocos, poor 
germination, poor soils, terrestrial crabs, and lack of care all 
undoubtedly contributed to their demise. 

The first recorded settlement on Caroline, and first for the 
Line Islands, was in 1846, on the northwest point. These 
settlers, as well as subsequent ones, eked out a spartan living by 
raising stock, drying fish and copra, and digging for guano. 
Their managers built "proper" dwellings, so when US. British, 
and French astronomers arrived to observe the solar eclipse in 
May 1 883. South Island was quite "civilized," far more than it 
is today. Three houses and two sheds "were in good repair," 
and a variety of "anchors, chains, spars, and pieces of the 
woodwork of vessels" littered its reefs ( Holden & Qualtrough, 
1884). Large grassy clearings adjacent to the lagoon 
accommodated several European-style houses (Pis. 2—4). The 
astronomers' account of South Island, illustrated with pen- 
and-ink drawings (Pis. 2-5,50), is the only record of buildings 
on Caroline, apart from mention of perhaps the same dwelling, 
the manager's house, reported in 1936 by the "H.M.S." 



Wellington to be "in excellent condition and spotlessly clean" 
(Maude, ca. 1938), and a copra shed seen by Clapp & Sibley 
(1971a). Arundel also took photographs, including some of the 
marae on Nake, which we have not examined (Arundel's 
memorabilia [photos, letters, diaries, a microfilm, et cetera] are 
deposited in the Rare Book Collection, National Library and 
Pacific Manuscripts Bureau, Research School of Pacific Studies, 
Australian National University, both in Canberra. A.C.T., 
Australia). 

Today, the houses, sheds, brick piers (constructed in 1883 
for telescopes and observatory frames), signboard, flagpole, 
marble slab with inscription "U.S. Eclipse Party, 1883, 
May 6," and all but one of the introduced plants have 
disappeared. In three trips we found no traces of the copra shed, 
nor have the Falconers, after repeated visits over 2 years. All 
that remains of the formerly large clearings are two small palm- 
shaded "flats." in 1988 used by the US and Soviet scientists for 
a base camp and work area. In 1987, the Falconers cleared one 
of these for living quarters, and in 1 990 fishermen expanded the 
other by burning an area 35 x 22 m, then erecting a tin shack, 
cookhouse, and fishtrap, which were destroyed in a summer 
1990 storm. 

Our "civilization list" probably covered all that could be 
seen on South Island without digging: a 26-foot wrecked sloop 
( AK 669 1 J. ), complete with trail to a "Robinson Crusoe-type" 
campsite strewn with remnants of radio and navigational 
equipment, sail, cans, clothing, et cetera (southeast coast); 
assorted flotsam and jetsam ( whisky bottles, Japanese fishballs, 
plastic debris, et cetera); a large rubber ship fender; a bench 
mark from the 1 985 RNZAF survey team; a recently renovated 
concrete cistern (by the landing); and an old wooden canoe 
lying on its side just like de Quiros found in 1606! 

We assume that all the Polynesians, ancient and recent 
(Tuamotuans, Tahitians, Niueans as far as is known), lived in 
native thatched huts similar to the ones on Ana-Ana today. 
Fashioned from coconut palms and pandanus trees, they 
disappear quickly when abandoned. The largest number of 
inhabitants recorded for Caroline ( probably all on South Island ) 
was "two managers and 52 laborers" in 1873 (The Commercial 
Advertiser, 1873). 

The history of South Island's coconut plantations from 
1885 to 1901 and from 1916 to 1929 is discussed under 
Coconut Woodlands (Plant Communities section). 

Physiography : South Island forms the base of the thinly 
crescentic isosceles triangle whose limits define the atoll. Its 
own shape is that of an irregular parallelogram 
858 m wide x 1 ,254 m long at its longest points (Fig. 50). The 
north coast, a curved bay, forms the lower boundary of the 
lagoon. This shore, along with the adjacent northwest 
peninsula, has been the most trodden by man, but the scars 
have healed, leaving few traces beyond the presence of 
coconut palms. 

The reef flats surrounding the outer three sides of South are 
the widest on the atoll, averaging 23 1 m, 578 m, and 363 m on 
the east, south, and west, respectively. The windward and 
leeward reefs immediately to its north are 530 m wide. To 
leeward is the small boat "landing," and to windward, the 
"blind passage." 



41 



Aesthetically, the lagoon fringe is one of the most 
picturesque spots on the atoll. Lush palms overhang a narrow 
beach of blinding white sand and coral gravel, affording idyllic 
views of the azure lagoon and its encircling wooded motus 
(Pis. 7,24,39). 

Vegetation : There are 23 plant species (7 trees, 3 shrubs, 
13 herbs), 85% of the atoll's flora. Cocos dominates South 
Island, occupying 77% of its area. The healthy, but old, closed- 
canopy plantations (21 m tall ) that border its coastlines give no 
indication of the vast extent of the overgrown, dying groves 
that occupy two-thirds of its interior ( Fig. 50; Pis. 7, 34). Here, 
three species of herbs (Boerhavia repens, Portulaca lutea, 
Phymatosorus scolopendha) have proliferated unnaturally to 
form dense carpets, and the vine lpomoea macrantha climbs, 
in tangled, strangling masses, to the tops of the highest palms. 

The natural communities that prevail on other motus are 
only minor components on South (Fig. 50): herb mats (13% of 
the total area), coastal scrub with Suriana ( 1 % ), and Tournefortia 
scrub (4%). Conspicuously absent are prime scrublands and 
forests of Tournefortia, Pisonia, and Cordia, which undoubtedly 
once swept in a lush expanse from shore to shore, stratified and 
zoned as on other motus. Canopy heights of the plantations are 
uniform (21 m). and the outer fringe of indigenous scrub 
(Tournefortia, Cordia, Suriana) and herbs (fleliotropium, 
Portulaca) occupy a small proportion of the island's width 
(Figs. 36,51). Note the abrupt transition of canopy heights as 
they drop to the level of coastal scrub on both sides of the 
plantation (Fig. 51). Pandanus, too, is less extensive than 
formerly: Bennett (1840) called Pandanus "somewhat 
numerous" and PI. 50 reproduces an 1883 painting of a grove, 
denser than any existing today on South. During our survey , we 
observed only one small Pandanus grove and a few scattered 
trees. Bennett also noted that the island was "covered with 
verdure," and there were "trees attaining the height of twenty 
feet." However, it is well to recall that 9 or 10 years previous 
to Bennett's visit a violent storm had whipped over the atoll. 
Drawings from 1883 (Pis. 4,5) depict remnant Tournefortia 
and Pisonia trees larger than this. 

Apart from the coastal buffer zone, little native forest 
remains. Other sizable trees (Pisonia. Cordia), up to 17 m tall, 
are rare, but Morinda, tolerant to both sun and shade, is still 
quite common. Though we have not been able to trace any 
records to Caroline, it is possible that shiploads of Cordia logs 
were exported to San Francisco on guano ships, as was the case 
on Flint, worked simultaneously by Arundel's company (Young, 
ca. 1922). 

A final noteworthy aspect of South Island is that, despite 
its history of sporadic occupation and extensive forest felling 
for coconut plantations, only one "weed," a tiny patch of 
Phyllanthus amarus, and no vegetable or garden ornamentals 
(excluding Polynesian introductions) have survived. (We are 
unsure of the status of Hibiscus tiliaceus, Thespesia populnea, 
or Ximenia americana). The 19th century gardens, once 
drenched in sunshine, have long been buried beneath the deep 
shade of palm groves (compare Pis. 2 and 24). In addition, 
periodic storms, droughts, irregular rainfall, nutrient-poor soils, 
rats, land crabs, and the harsh salty environment must have 
contributed to the eradication of all exotics except traditional 



native food and medicine plants, which are specifically adapted 
for atoll environments. Studies on other atolls, even those near 
high islands (Stoddart & Fosberg, 1972; Stoddart & Gibbs, 
1 975 ), have demonstrated also that exotics survive, despite the 
proximity to source areas containing garden ornamentals and 
weed plants. We learned on our last two visits ( March and May 
1990), however, that a small sunny clearing around the cistern 
has attracted one clump of a weed not previously reported 
(Kyllinga brevifolia), also the location of Phyllanthus. This 
area is now used as an extension of the Falconers' vegetable 
garden on Motu Ana-Ana. Kyllinga is listed as a temporary 
species (Table 1 ). 

Birds : Only 2 species of birds bred on South in September 
1988, a reflection of its paucity of natural habitats: brown 
noddy (163 pairs) and white tern (381 pairs). Bristle-thighed 
curlews are very common, gathering in small flocks on the 
rubbly shores (Subchapter 1 .2, this volume). They also forage 
in the open Ipomoea-Cocos forest, perching on dead coconut 
stumps 6-10 m high, then flying down to feed in the thick herb 
mats. 

Terrestrial Crabs : Caroline's highest population of coconut 
crabs, having many huge individuals decades old. occupies the 
open Cocos forests (Pis. 22,56,57 ). A crude minimum estimate 
for South Island is 500 mature individuals (since March 1990. 
these have become much reduced due to killing and preserving 
in formalin for curios). We also found a fist-sized blue hermit 
crab within a Turbo shell, possibly Coenobita brevimanus 
(Yaldwyn & Wodzicki, 1979; E. Reese, personal 
communication). As elsewhere on the atoll, land crabs such as 
the reddish-purple Cardisoma sp. and scarlet hermit crabs, 
Coenobita perlatus (in Turbo argyrostomus shells), were 
abundant ( PI. 38). A Geograpsus sp., closer to the shore, was 
less common. 

Rats : Polynesian rats were abundant on South, active both 
day and night. The rats were constantly afoot in broad daylight, 
and at night a small flashlight beam often revealed a half dozen 
at a time. 

South Nake Islets (Fig. 52) 

This chain of seven islets extends 1,500 m south from 
Nake on the west side. They range in size from 0.64 ha ( Kota) 
to 7.36 ha (Pandanus). All are well wooded and support every 
natural plant community. Proceeding south, the overall plant 
cover thins somewhat, but not to the dryness and openness of 
the Central Leeward Islets. The herb mats are more extensive 
than on the windward islets, especially to seaward. Aboriginal 
introductions (Cocos, Pandanus) are sparse. We have found 
no historical records indicating human disturbance to these 
islets, thus their vegetation, with the possible exception of 
Pandanus Islet, is evidently natural. The two scrawny Cocos 
are probably drift-derived. 

On the Solar Eclipse Party's map of Caroline (Fig. 5), only 
the top two islets of this group are drawn. The South Nake Islets 
constitute the only cluster of motus that show appreciable 
differences between Arundel's chart (Fig. 4) and the 1985 
aerial photos: most were shown as smaller, and with slightly 
different shapes, by Arundel. The interior vegetation on these 
motus includes mature forests of Tournefortia. Pisonia, and 



42 



Pandanus, so it is unlikely that these differences reflect changes 
to the center of the motus. However, since the islets now appear 
larger, accretions of coral rubble and sand that may have 
occurred in the past 105 years, and are now barren or covered 
only with herb mats, could account for most of the differences 
(see Coral Islet discussion). 

Although we have no actual records of sooty tern colonies 
on this chain of islets, in May 1990 AKK observed pre- 
breeding swirls of this species over Lone Palm, Kota, and 
Mouakena (Subchapter 1.2, Fig. 1 1 ). 

17. PANDANUS ISLET (7.36 ha) (Figs. 29,52; PI. 64) 

This motu was named by the present authors for its coastal 
Pandanus grove, probably a drift-derived offshoot from a 
parent colony on Nake. 

Physiography : Pandanus Islet, first in the chain, is 
irregularly oval, 400 m long and 258 m across. It is nearly twice 
the size shown on Arundel's map (ca. 3.4 ha). It occupies a 
sheltered spot at the apex of the lagoon. Sand, actively filling 
in the adjacent lagoon, is an important component of the 
substrate on Pandanus, extending one-third of the way across 
the islet. Although tidal reef flats are absent on the lagoon edge, 
they average 75 m wide on the seaward side, producing a fairly 
high proportion of rubble compared to the total land surface 
(32%). 

Vegetation : Plant species total 10 (3 trees, 1 shrub, 
6 herbs), 37% of Caroline's flora. Cocus, surprisingly, is 
absent, despite the close proximity to Nake. Pandanus Islet has 
four basic vegetation zones: natural herb mats, Toumefortia 
scrub (with Pandanus), Tournefortia-Pisonia forest, and pure 
Pisonia. Woodlands cover 62% of its area. The widest pioneer 
mats ( 13 m) of any leeward motu occupy its east edge and 
though sparsely vegetated (20% Heliotropium, 5% Lepturus, 
5% Portulaca) reflect active growth toward the lagoon. 
Proceeding west across the island, Toumefortia scrub (2 m 
high), with pockets of pure Pandanus (10m high), merges into 
Tournefortia-Pisonia forest (to 14 m high), whose bimodal 
distribution suggests that the islet was once divided. The 
seaward coast supports open Toumefortia (5 m high), beneath 
which herbs eventually thin out onto the extensive reef flats. 

Birds : Five species of seabirds breed: masked booby 
(2 pairs), red-footed booby (32 pairs), great frigatebird 
(26 pairs), brown noddy (26 pairs), and white tern (52 pairs). 

Comments : Skinks and rats were observed, along with the 
ubiquitous Coenobita and Cardisoma crabs. 

18. DANGER ISLET (2.71 ha) (Figs. 29,52; PI. 65,68) 

We named Danger Islet to commemorate the deep, shark- 
infested channel to its north, a barrier that aborted our first 
(dusk) attempt to survey the South Nake Islets. 

Physiography : Danger, shaped like a thickened comma, is 
approximately 150 m long and 215 m wide. It is composed 
almost entirely of coral rubble; interior humus is scant. Its reef- 
channel flats are 21 m (north) and 14 m (south) wide. The east 



and west beaches, narrow and wide respectively, are typical of 
all the leeward motus. 

Vegetation : Danger has 10 plant species (3 trees, 1 shrub, 
6 herbs), 37% of the total flora. There are no introductions. The 
usual concentric vegetation is clearly zoned: herb mats, 
Toumefortia scrub and forest, central Pisonia, and Cordia in 
the southwest. The herb mats are wide, extending 22 m and 
15 m on the north and south shores, respectively. 

Birds : Four species of nesting seabirds were present in 
1988: red-footed booby (139 pairs), great 
frigatebird (26 pairs), brown noddy (26 pairs), and white tern 
(52 pairs). 

19. BOOBY ISLET (0.84 ha) (Figs. 29,52; PI. 66) 

We named this motu, third in the chain, for its two species 
of boobies, the common red-footed and rarer masked booby. 

Physiography : Booby, shaped like a teardrop, is 70 m long 
and 1 25 m wide. Its coral rubble flats extend 1 m and 30 m on 
the north and south sides, respectively. 

Vegetation : Despite its small size, the most notable 
feature of Booby is its Pisonia forest. 20 m tall and undoubtedly 
virgin. It occupies the exact center of the islet in a circle about 
40 m in diameter. Surrounding this is Toumefortia scrub (to 
8 m tall), thinning out to peripheral bands of coral rubble. 
Although less than one hectare in size. Booby Islet's woodlands 
occupy two-thirds of this area. Booby Islet has nine species of 
plants (two trees, one shrub, six herbs), 33% of Caroline's 
flora, and no introductions. 

Birds : Five species of seabirds breed: masked booby 
(7 pairs), red-footed booby (52 pairs), brown noddy (2 pairs), 
black noddy ( 1 pair), and white tern (6 pairs). 

20. CORAL ISLET (1.70 ha) (Figs. 29,52; PI. 66) 

Fourth from the north. Coral Islet was named for its reef- 
derived coralline substrate. 

Physiography : Shaped like an arrowhead, Coral is 
approximately 130 m long by 200 m wide, more than three 
times the size mapped by Arundel (Fig. 4). Most of its area is 
barely higher than the surrounding interislet channels. The 
shallow reef flats between Coral and its two southern motus are 
only several centimeters deep at low tide; all three may be 
destined to unite. Unless closely inspected, they appear to have 
already merged, a fact which, together with Bryan's incorrect 
map (Fig. 6), helps account for the widely differing number of 
motus attributed to Caroline. 

Vegetation : There are nine species of plants (two trees, 
one shrub, six herbs), 33% of Caroline's flora, and no 
introductions are present. Plant communities comprise a small 
Pisonia forest (0. 1 3 ha), which is surrounded by the predominant 
Toumefortia, which in turn is fringed with a narrow band of 
native herbs. "Soils" are extremely coarse. 

Birds : Five species of seabirds bred in 1988: masked 
booby (1 pair), red-footed booby (28 pairs), great frigatebird 
(2 pairs), brown noddy (6 pairs), and white tern (15 pairs). 



43 



21. LONE PALM ISLET (1.99 ha) (Figs. 29,52; Pis. 66-68) 

We named Lone Palm, fifth in the chain, for its single 
coconut palm, which towers, Hag-like, above a dense mound of 
Towmefortia. 

Physiography : Similar to Kota (to its south), Lone Palm 
is sausage-shaped, 97 m long and 240 m wide, and four times 
the size mapped by Arundel. Although composed almost 
entirely of coral rubble, some sand borders the lagoon. Following 
a pattern prevalent on all the leeward motus, its lagoon beach 
is 2 m wide, while the seaward beach is 17 m. 

Vegetation : Eleven species of plants are present (three 
trees, one shrub, seven herbs). 46% of Caroline's flora. Lone 
Palm's plant communities are simple: a wide band of herb mats 
and open Toumefortia flanks an oval of Towmefortia forest (to 
1 m tall ). A line of Pisonia trees, with a lone Cocos surmounting 
the scrub, easily identifies this islet from lagoon or ocean. 

Birds : Three species of seabirds bred in 1988: masked 
booby (2 pairs), red-footed booby (48 pairs), and white tern 
(9 pairs). In May 1990, we saw a large prebreeding swarm of 
sooty terns. 

22. MOTU KOTA -Red-footed Booby Islet" (0.64 ha) 
(Figs. 28,52: Pis. 66.68) 

We named this motu for its high density of red-footed 
boobies (kota in Gilbertese). 

Physiography : Sixth in line south of Nake, sausage- 
shaped Motu Kota is 50 m long and 175 m wide. At low tide 
it is almost connected to Motu Mouakena. Both surveys 
indicate that coral rubble, the islet' s predominant substrate, had 
further accumulated on its south side since the 1985 aerial 
photos and also since 1988. 

Vegetation : Though barely wooded. Kota has 1 1 species 
of plants (3 trees, 1 shrub, 7 herbs). 41% of Caroline's flora. 
One introduction is present, a single, tattered Cocos, partly 
hidden by vegetation. Two plant communities are present: 
peripheral herb mats and central Toumefortia scrub (to 10 m 
tall |. with a few Pisonia. 

Birds : Three species of seabirds bred in 1988: brown 
booby (1 pair), red-footed booby (12 pairs), and white tern 
(3 pairs). In May 1990, a single masked booby was on territory, 
and sooty terns swirled overhead. 

23. MOTU MOUAKENA "Masked Booby Islet" ( 1 .00 ha) 
(Figs. 29,52; Pis. 15.69) 

This islet was named for its nesting masked boobies, a 
relatively uncommon seabird on Caroline. 

Physiography : Somewhat U-shaped, Motu Mouakena is 
seventh, and southernmost, in the South Nake chain of islets. 
Both sides of the "U" were, in the recent past, separate islets. 
By joining on the west, a narrow, V-shaped inlet was created on 
the lagoon side. Motu Mouakena. 100 m long and 1 60 m wide. 
is extremely nubbly ami infertile; much rubble was reorganized 
during the February 1990 storm. Seventeen meters to its south 
lies a newly emerging shoal of sand and gravel (PI. 15), perhaps 
destined to be Caroline's fortieth motu or perhaps part of 



Mouakena' s southern shore. Since the above storm, rubble has 
further accumulated on this shoal, its adjacent reef Oats, and the 
channel separating it from Mouakena. It already supports one 
Toumefortia shrub, two dozen Heliotropium plants, and very 
scattered Lepturus and Portulaca. 

Vegetation : The number of species is eight ( one tree, one 
shrub, six herbs), 30% of Caroline's flora, with no introductions. 
Mouakena is thinly vegetated with open Toumefortia scrub (to 
9 m tall, 26% cover), a few small Pisonia. and very sparse herb 
mats. 

Birds : Mouakena has less vegetation and fewer birds than 
might be expected from a consideration of its area because 
much of it is unshaded, coarse coral rubble. Though 
unproductive botanically, this provides ideal nesting grounds 
for masked boobies, one of the two species of breeding seabirds 
on the islet in 1988: masked booby (3 pairs) and red-footed 
booby (8 pairs). In May 1990, we saw one great frigatebird nest 
with eggs and a swirl of sooty terns. 

Central Leeward Islets 

This chain of 1 1 motus occupies the central west side of 
Caroline. All are separated by channels, wadable only at low 
tide but prowled by belligerent sharks. Approximately 
1.600 m south of Motu Mouakena lies a sandy shoal (0.5 m 
high. 7 m wide. 4 m long), close to the lagoon edge of the reef 
flats and connected only by a thin thread of rubble to Motu 
Mannikiba to its south. 

The islets range in size from Mannikiba (28.50 ha), the 
most northerly, to Fishball (0.46 ha), the most southerly. All 
support good seabird populations and. though quite well wooded. 
are nonetheless the least lush motus on Caroline. Historical 
records of the Central Leewards are very scant: much of 
Mannikiba's forest was felled to make room for a Cocos 
seedling "nursery" (Young, ca. 1922). The bulk of "40 trees on 
other islets." in Young's totals, were most likely from Shark 
and Emerald. The rest of this group is evidently pristine; the 
natural communities on Bird Islet, in particular, are in excellent 
condition. 

Shark Islet boasts the best sandy beach on the atoll. 
In common with all the motus on Caroline's west rim. 
their lagoonside beaches are narrow and leeward reef 
Hats wide. The leeward flats are composed of a greater 
variety of substrata than the former, including coral rubble 
of several grades (always gray), upraised reef, and 
beachrock. Periodically, thousands of nesting sooty 
terns occupy their open spaces (Clapp & Sibley. 1971a: 
AKK, personal observation; Anne Falconer, personal 
communication). 

24. MOTU MANNIKIBA "Seabird Islet" (21.49 ha) 
(Fig. 29; Pis. 70-72) 

We named this motu for its teeming seabirds, mannikiba 
in Gilbertese. 

Physiography : Largest and most northerly of the Central 
Leeward Islets. Mannikiba is somewhat rectangular with 
rounded corners. Its reel Hats, containing an incipient islet, 
stretch 2.0 km north to the South Nake Islets. 



44 



Mannikiha's maximum dimensions are 700 m long and 
375 m wide. On the lagoon side, the serub skirts high water, but 
when the tide drops, a strip of blinding white sandy coral lines 
the lagoon. To seaward, upraised reef, beachrock, and successive 
layers of gray coral rubble stretch in a wide swath (40 m) 
toward the outer reef, 130 m distant. Throughout the islet the 
substratum is gray coral rubble, with some exposed reef flat 
hardpan in the northeast. Having numerous seabirds, this motu 
might have also contained productive guano deposits. 

Vegetation : Mannikiba, the fourth largest motu, harbors 
1 3 plant species: (4 trees, 2 shrubs, 7 herbs), 48% of Caroline's 
flora. One of its shrubs (Species A), a new record for Caroline, 
has yet to be identified. The only introduction is Cocos, 
occupying 0.1% of the land area. 

Mannikiba' s vegetation, denser toward the north end, is 
clearly zoned: herb mats, Tournefortia scrub and forest, and 
scattered Pisonia groves. The few clumps of peripheral Cocos 
are probably not drift-derived but the remnants of 6,000 "seed 
sets" brought from Flint Island in June 1920. These were stored 
on Mannikiba and "used to replant misses on other islets" 
(Young, ca. 1922). 

Pisonia, though present, occupies only 5% of the land 
area, a small percentage for such a large islet. This suggests that 
a large portion of the interior forests were felled to accommodate 
the coconut "sets." This is also confirmed by the presence of 
several old cut stumps in the interior. A century ago, Holden 
& Qualtrough ( 1 884) noted that "About one-third the distance 
up the lagoon a canvas hut exists on one of the smaller islets on 
the eastern side of the lagoon, and two wooden huts stand on 
one of the western islets, some distance further up the lagoon." 
Mannikiba, the largest western islet, situated about halfway up 
the lagoon, was most likely the site of the wooden huts, erected 
around 1 920 and used for the following few years when the new 
company, S. R. Maxwell & Co., Ltd., was anxious for the 
success of Caroline's plantations. Although nothing more is 
known of Mannikiba's history, collection of guano from its 
numerous seabirds, including large populations of frigatebirds 
and sooty terns, may account for further past disturbance. 

Transect 1 ( north-central sector, PI . 7 1 ) passed through the 
heart of a fine interior forest, while Tr. 2 (south-central sector) 
passed through scrub and herb mats, which may represent part 
of the former Cocos "nursery." Profiles through these two 
cross-island transects resemble those from Brothers (Fig. 46) 
and an old interisland channel on Long (Fig. 40), respectively. 

The low, peripheral herb mats (absent from the lagoon 
side) are composed of 30% Heliotropium, 20% Boerhavia, 
1 5% Tournefortia, and less than 1 % of Portulaca and Laportea. 
They are best represented in the southern sector. The 
Tournefortia forest, 6 m high on both sides, is thick, having 
95% canopy coverage. The Pisonia forests, though fragmented 
(12m high, 1 00% canopy cover), contain Morinda, Boerhavia, 
Achyranthes, Laportea, and Phymatosorus, but none cover 
more than 1 0% of the ground area. 

Seabirds : Six species of seabirds are known to breed: red- 
footed booby ( 184 pairs), great frigatebird (287 pairs), brown 
noddy (161 pairs), black noddy (176 pairs), and white tern 



(195 pairs). No sooty terns nested on this islet in 1988, but 
Clapp & Sibley ( 1 97 1 a) estimated 2,500 pairs in 1 965, and the 
Falconers reported large colonies on Mannikiba, Blackfin and 
Matawa in July-August 1990. 

Comments : Coconut crabs live in the Cocos grove. 
Azure-tailed and snake-eyed skinks (Cryptoblepharus 
poecilopleurus), as well as a gecko, were noted in 1990 (DHE, 
G. Wragg, personal observation). 

25. BLACKFIN ISLET (2.62 ha) (Figs. 29,54; Pis. 3 1 ,73 ) 

We named this motu, second in the Central Leeward chain, 
for two exhilarating shark attacks (near misses) within its 
northern surge channel. 

Physiography : Blackfin, shaped like conjoined ovals, is 
140 m long and 190 m across. Coral rubble covers 30% of its 
surface; all beaches and upper reef flats are of variable widths, 
due in part to the fact that it has. in the recent geological past, 
incorporated a smaller, circular motu into its northern confines. 

Vegetation : Blackfin Islet has nine species of plants (three 
trees, one shrub, five herbs), 33% of Caroline's flora. The only 
introduction, Cocos, is rare. Four plant communities were 
identified. Herb mats are well represented, especially around 
the newly incorporated islet. The Tournefortia scrub, 21 m 
v/ide in the east, is short (to 2 m), but approaches the stature of 
a forest (to 6 m) in the west. The central forests of Cordia and 
Pisonia (0.41 ha) are 9 m high. 

Birds : Three species of seabirds bred in 1988: great 
frigatebird (4 pairs), brown noddy (37 pairs), and white tern 
( 1 1 pairs). In May 1990, one red-footed booby sat tight on a 
nest, while two months later large numbers of sooty terns began 
laying. 

26. MOTU MATAWA "White Tern Islet" (1.71 ha) 
(Figs. 29.54; PI. 55: Subchapter 1.2, PI. 3) 

On arriving at this islet, the authors were swarmed by 
1 5 white terns, all hovering within arm's reach and exhibiting 
the ethereal grace that inspired their former common name, 
fairy tern. I-Kiribati (Gilbertese) call them matawa. 

Physiography : Of oval shape, Motu Matawa is third from 
the north in the Central Leeward chain. It is 105 m long and 
190 m wide. The entire motu, like all of Caroline's small to 
medium islets, is built of coral rubble of varying grades, whose 
unvegetated portion comprises one-fourth or more of the land 
area. Its lagoon beach is 2.5 m wide, while the seaward beach 
(sparsely vegetated) is 6 m wide. 

Vegetation : Matawa has 10 species of plants (4 trees, 
2 shrubs, 4 herbs), 37% of Caroline's flora. There are no 
introductions. The usual plant communities were present, 
dominated by Tournefortia (to 7 m), which covers half the islet. 
Vegetation is less lush and more open as one progresses south 
on the leeward side. Coral rubble, flanking the beaches and 
extending further inland, also becomes more evident. The east- 
central Pisonia-Cordia forest (to 8 m) rises barely higher than 
its surrounding Tournefortia. 



45 



Birds : Four species of seabirds bred: red-footed booby 
(5 pairs), great frigatebird ( I pair), brown noddy (3 pairs), and 
white tern ( 1 3 pairs). Most conspicuous were white terns, with 
9 pairs breeding on the 30-m-wide transect swath. One dark 
morph reef heron fished in the shallows. In summer 1990, 
sooty terns bred. 

27. EMERALD ISLE (8.34 ha) (Figs. 29,54; Pis. 25,74-76) 

Fifth down the chain, we named Emerald for the richly 
colored, translucent lagoon waters that fringe its shorelines. 

Physiography : Of thickened crescentic shape, Emerald is 
330 m long and 240 m wide. Its lagoonside reefs, patch reefs, 
and coral knolls are irregularly patterned with sandy channels. 
It is here that the verdure of the lagoon is most intense. 

Vegetation : Emerald Isle has 12 species of plants (5 trees, 
1 shrub, 6 herbs), 44% ofCaroline's flora. The only introduction 
is Cocos. Four plant communities, with a fairly high species 
diversity, are present: the herb mats, covering one-fourth of its 
land area, are composed almost exclusively of Heliotropium 
( 35% cover) with scattered low Tournefortia (30% cover). The 
Tournefortia scrub and forest attains a maximum height of 8 m 
and. for a little variety, is mixed about equally with Pandanus 
over most of its width ( 144 m) on the seaward side. 

The interior forest (to 1 1 m tall) is also mixed, with 
Pandanus, Tournefortia, Pisonia, and a little Cordia (PI. 75). 
This condition is similar to the mixed forest on Nake, but 
because Cocos is absent, it appears more natural. The existence 
of this 3.20 ha mixed forest, as well as a similar one on Shark, 
prompted us to suggest that Pandanus may be both native and 
Polynesian-introduced. Cocos is present as two small groves, 
complete with coconut crab sign (mounds of shredded fibers, 
PL 57), beside the east and midwest shores. 

We have been unable to trace the history of Emerald's 
forests: the presence of Cocos on the west side and fragmented 
Pisonia suggest past disturbance. 

Birds : Six species of breeding seabirds were present: red- 
tailed Tropicbird (1 pair), red-footed booby (3 pairs), great 
frigatebird (230 pairs), brown noddy (7 pairs), black noddy 
( 150 pairs), and white tern (83 pairs). 

Although we did not locate any red-tailed tropicbird nests, 
two adults circled steadily overhead. Two reef herons (one 
dark morph, one light ) also foraged in the inshore reef shallows. 

28. SHARK ISLET (7.98 ha) (Figs. 29,55; Pis. 29,77) 

We named this islet to commemorate a particularly 
pugnacious shark who was so anxious to procure a human foot 
that it charged shoreward and leaped up onto the beach. 

Physiography : Stoutly crescentic. Shark Islet is 280 m 
long and 310 m wide in the center. The sandy lagoon beach and 
rubbly seaward beach are each 3 m wide. Beyond high water 
the seaward reel Hats extend for 280 m. Like Emerald. Shark's 
reefs and surrounding lagoon waters reflect particularly stunning 
colors, perfect complements to the sparkling pink coral sand of 
Caroline's prime beach. 



Vegetation : There are 12 species of plants (5 trees, 
1 shrub, 6 herbs), 44% of the atoll's flora. Shark has one 
introduction, Cocos, forming 3 clumps along the lagoon beach 
( 1 % of the islet's area). Shark's rings of vegetation approximate 
the islet's outline. Herb mats dot the fine sand lagoonward. 
while to seaward they emerge from coarse rubble. The 
Tournefortia (to 7 m tall) eventually gives way to a 12-m-high 
Pisonia forest studded with Cordia and Pandanus. Centrally 
this mixed forest is unnaturally open, suggesting possible past 
disturbance. 

Birds : Four species of seabirds bred in 1988: great 
frigatebird (118 pairs), brown noddy (37 pairs), black noddy 
(125 pairs), and white tern (44 pairs), red-footed boobies were 
nesting in 1990. The notable colonies of great frigatebirds and 
black noddies are due in part to the extensive Pisonia forest, 
covering half of the islet. 

29. SCARLET CRAB ISLET (0.46 ha) (Figs. 28,55) 

This motu was named by the authors in honor of Coenobita 
perlatus, the scarlet, fist-sized hermit crab that is abundant both 
here and on the entire atoll. 

Physiography : Scarlet Crab, sixth in the chain and only 
40 m long by 1 25 m wide, is a young motu shaped like a closed 
pair of lips. It skirts the southern shore of Shark, separated from 
it by a channel 1 6 m wide. Because its eastern end points into 
the lagoon, there is no true lagoon beach. Together with the 
next three islets. Scarlet Crab's seaward reef flats (480 m) are 
the most extensive on Caroline's lee side. 

Vegetation : Vegetative cover is slight: less than 1% 
area coverage of Heliotropium and Laportea, interspersed 
with 10 small Tournefortia (to 1.5 m). Its species count is 6 
(1 shrub, 5 herbs), 22% of Caroline's flora. There are no 
introductions. 

Birds : Although during storms this motu is undoubted!) 
awash, two species of seabirds were breeding in 1988: brown 
noddy (one pair, on ground) and white tern (two pairs, in low 
scrub). 

30. MOTU NAUTONGA "Sea Cucumber Islet- (0.34 ha) 
(Figs. 28.55) 

We named this motu for the Gilbertese word for the black 
sea cucumbers or "beche-de-mer" (Ludwigothuria sp. ) that are 
densely strewn over all ofCaroline's reef shallows (PI. 10). 

Physiography : Semicircular in shape. Nautonga is seventh 
in the Central Leeward chain, measuring 70 m long and 80 m 
wide. Situated close to the lagoon, it is one of three small islets 
that barely protrude above the reef flats. Nautonga' s perimeter 
beaches are all narrow (2m), and its seaward reef flats are wide 
(495 mi. 

Vegetation : There are nine indigenous species (three 
trees, one shrub, five herbs). 33% of the atoll's flora. Though 
small. Nautonga's vegetation is concentrically zoned, 
comprising herb mats (10-14 m wide), and a central forest of 
Tournefortia and Pisonia (84 m wide) up to 10 m high. 



46 



Birds : Five species of seabirds bred in 1988: red-footed 
booby ( 1 1 pairs), great frigatebird (2 pairs), brown noddy 
(7 pairs), black noddy (32 pairs), and white tern (10 pairs). 
Lesser frigatebirds appeared to be preparing to nest in May 
1 990. One pair of blue-gray noddies, flying toward Azure Isle, 
was seen by the 1990 expedition in May. 

31. AZURE ISLE (0.20 ha) (Figs. 28,55; PI. 53) 

We named this small, wedge-shaped motu for the striking 
blue-green of its nearby lagoon. 

Physiography : Eighth from the north, this small, elongated 
triangle of land is 30 m long and 66 m wide. Its seaward reef 
flats are wide (512 m). Its surge channels are narrow and 
shallow. 

Vegetation : Azure has only seven species (one tree, one 
shrub, five herbs), 26% of Caroline's flora. The Pisonia tree is 
6 m tall. A young motu. Azure is a superb example of an early 
stage of biological succession. Its plant cover consists of a 
single mound of Toumefortia scrub crowned by a single 
Pisonia tree, growing from rubble only one meter above sea 
level. Only 45% of its surface is vegetated; the rest, primarily 
on the ocean side, is coarse rubble. Azure Isle presents what 
may be the minimum width of vegetation (38 m) in which 
Pisonia can develop on Caroline. 

Birds : This motu illustrates the speed at which seabirds 
will utilize newly available habitats. Within its dozen or so 
Toumefortia shrubs (to 4 m tall), three species of seabirds nest: 
red-footed booby ( 7 pairs ), great frigatebird ( 2 pairs ), and white 
tern (2 pairs). A pair of blue-gray noddies were seen in May 
1990. 

32. REEF-FLAT ISLET (0.09 ha) (Figs. 27,55) 

We named this young motu for its primary characteristic — 
its reef flats. Ninth in the Central Leeward chain, this curved 
strip of coarse rubble lies parallel to the surge channels that 
surround it. It measures about 20 m long and 60 m wide. Only 
three species of plants are present (one shrub, two herbs), 1 1% 
of Caroline's flora. They cover less than one-fourth of its area 
and are distributed so sparsely that not one bird was present. 

33. BIRD ISLET (4.05 ha) (Figs. 29,55) 

This is one of the motus named on Arundel ' s chart ( Fig. 4), 
probably reflecting the presence of black noddies and/or sooty 
terns. 

Physiography : Bird is ovoid, measuring 230 m long by 
200 m wide. It sits close to the inner edge of the lagoon reef, 
whereas 400 m of seaward reef flats stretch westward. 

Vegetation : There are 12 species of plants (4 trees, 
2 shrubs, 6 herbs), 44% of Caroline's flora. A small Cocos 
grove is the only introduction. It is well wooded, with very 
narrow herb mats (6% of total area). Toumefortia (to 8 m) and 
Pisonia (to 14 m) each cover 42% of its surface; the rest is 
rubble. The Pisonia forest is of good quality (90-95% canopy 
cover), having scattered Mo rinda. Boerhavia. andAchyranthes 



as an understory. One large clump of Suriana (14 x 14 m, 
2.5 m high) occurred in the islet center (A. Garnett, personal 
observation). Bird Islet shows very few signs of past disturbance, 
having prime plant communities, rich in breeding seabirds. 

Birds : Five species of seabirds nested in 1988: red-footed 
booby (29 pairs), great frigatebird (6 pairs), brown noddy 
(42 pairs), black noddy (329 pairs), and white tern (48 pairs). 
In June 1990, many thousands of sooty terns laid on Bird and 
adjacent Fishball. 

34. FISHBALL ISLET (0.57 ha) (Figs. 28,55,56) 

Eleventh and southernmost in the Central Leeward chain, 
we named Fishball after discovering a large glass fishing float 
with a broken bottom, decorously placed in the islet's center 
within a square of coral slabs. 

Physiography : Paramecium-shaped, Fishball lies close to 
the lagoon and is separated from Bird by a shallow, rubble- 
strewn channel 100 m wide. The motu is 45 m long by 144 m 
wide, with seaward reef flats 595 m in extent. South of the islet, 
the reef flats — wadable at very low tide — stretch 1 .4 km to the 
Southern Leeward Islets. 

Vegetation : The number of plant species is eight (one 
seedling "tree," one shrub, six herbs), 30% of Caroline's flora. 
Figure 56 depicts an east-west cross-section of Fishball, showing 
a vertical profile and the relative abundance and distribution of 
each species. Fishball exemplifies an emerging motu. All 
plants are low and halophytic; most are herbs. The motu is half- 
covered with a sparse herb mat of Heliotropium (10% cover), 
with scattered Laportea. Lepturus , and Portulaca (less than 1% 
cover each). Small Toumefortia shrubs (to 2 m tall) are 
scattered in the central sector, while a tiny drift seedling of 
Morinda, 7 cm high, struggled to gain a foothold in the 
exposed, salty rubble. 

This motu is a fine example of the initial stages of islet 
formation and colonization. It demonstrates that sea-dispersed, 
halophytic herbs first germinate on the coarse rubble, later 
becoming shaded out by Toumefortia, enabling a greater plant 
species diversity to establish. It is very unlikely that a water 
lens is present. 

Birds : Two species of seabirds bred in 1988: red-tailed 
tropicbird (three pairs) and brown noddy (five pairs). In May 
1990, many thousands of sooty terns covered the ground and 
swirled in the air, day and night. On 23 May, no eggs were 
found, but laying occurred on Fishball and adjacent Bird Islet 
in June (A. Falconer, personal communication). 

Southern Leeward Islets (Pis. 14, 78) 

This chain of five small motus lies along the southwestern 
edge of the lagoon. All are built upon piles of rubble about 
3 m high, oriented in an east-west direction, and are separated 
by shallow, narrow channels. They range in size from 1 .5 1 to 
3.67 ha, and their topography, vegetation, and breeding seabirds 
are similar. Although situated on the leeward side of the atoll, 
the Southern Leeward Islets exhibit some windward 
characteristics: they lie opposite and slightly north of a wide 
break in the windward reef, which allows trade winds to sweep. 



47 



uninterrupted, across the lagoon. This promotes their 60-80% 
cover of scrub or forest. Ana-Ana, the southernmost, was 
periodically occupied from 1987-1991 by the Falconer family. 

Of particular botanical interest are the interior forests, 
composed of Pisonia mixed with more Cordia than elsewhere 
on the atoll. Pure Cordia groves (mostly too small to map 
accurately) typically occupy the forest periphery. 

Their history (apart from the last 3 years) is unknown; all 
appear to harbor virgin plant communities with occasional 
drift-derived Cocos or Pandanus. 

35. MOTU RAURAU "Blue- ? ra\ Nodd\ Islet" (3.48 ha) 
(Figs. 29,57; Pis. 14,78,79) 

Northernmost of the Southern Leeward Islets, we named 
this motu for the blue-gray noddies {raurau in Gilbertese) 
observed there. A highly territorial blue-gray noddy was acting 
as though a nest was nearby. 

Physiography : Raurau is ovoid, with a small lagoonside 
bay. and maximum dimensions of 1 80 m long and 23 1 m wide. 
It has the most expansive rubble of all the Southern Leeward 
Islets. This coarse coral clinker extends, apronlike, around the 
islet, widest (40 m) closest to the lagoon and narrowest (10 m) 
to seaward. The seaward reef flats extend 446 m to the ocean. 

Vegetation : The number of plant species is 10(5 trees. 
1 shrub. 4 herbs), 37% of the atoll's flora. Raurau's two plant 
communities are simple: a very scant herb mat is sprinkled 
with Tournefortia, which rises to 6-m-high scrub all around the 
islet. Laportea forms a narrow band at the interface between 
coral rubble and scrub. Centrally a Pisonia forest (to 13 m). 
dotted with Cordia on the periphery, harbors much Morinda in 
the understory. including the tallest Morinda ( 1 3 m) seen on the 
atoll. A handful of drift-derived Cocos and Pandanus. the only 
introductions, dot the scrub. 

Birds : No seabirds were found on transect, but a perimeter 
walk in 1 988 revealed that tour species bred on the west side in 
the Tournefortia scrub: red-footed booby (10 pairs), great 
frigatebird (31 pairs), brown noddy (1 pair), and white tern 
( 2 pairs). This islet, for its size, is particularly rich in frigatebirds. 

Comments : Polynesian rats are present. 

36. MOTU EITEI "Frigatebird Islet " (1.41 ha) (Figs. 29.57; 
Pis. 14.78) 

Second in line from the north, we named this motu for its 
nesting great frigatebirds, eitei in Gilbertese. 

Physiography : Motu Eitei is fat-elliptical. 105 m long and 
2X0 m wide. Lying perpendicular to the reef axis, it touches the 
lagoon edge on its inner side. To seaward, the reel' Hats are 
644 m wide. 

Vegetation : There are eight species of plants (three trees, 
one shrub, lour herbs), $09i of the atoll's flora, with no 
introductions. Eitei is carpeted with three plant communities 
in the usual concentric arrangement. However, there is a slight 
difference in the species composition of the herb mats: on 
transect, the southern mat (2 m wide) consisted solely of 
Portulaca, while the north side contained a 3-m swath of 
Heliotropium, Laportea. and scattered Suriana. Inside the mat 



is a ring of Tournefortia scrub (to 5 m) and a central Pisonia- 
Cordia forest (to 11m). Laportea is particularly abundant, 
while Portulaca, normally confined to the edges, abounds in 
small openings within the interior woodlands. 

Birds : Four species of seabirds bred on Motu Eitei in 1988: 
red-footed booby ( 1 7 pairs), great frigatebird ( 1 4 pairs ). brown 
noddy (6 pairs), and white tern ( 18 pairs). The first blue-gray 
noddy nest for Caroline was found in summer. 1990 
(Subchapter 1.2. this volume). 

37. PISONIA ISLET (2.45 ha) (Figs. 29.57: Pis. 14.78) 

We named this motu for its fine Pisonia forest. 

Physiography : Pisonia, third in the chain from the north, 
is almost circular and lies closely appressed to its neighbor 
islets. Its maximum dimensions are 140 m long and 220 m 
wide. Like Raurau. it possesses a wide "apron" of coral rubble 
and sparse herbs on the lagoon side. Its seaward reef flats are 
300 m wide. 

Vegetation : The number of plant species is 15 (5 trees. 
2 shrubs, 8 herbs), 56% of the atoll's flora. The only introduction 
is Cocos { few, scattered, north and south shores). Well wooded. 
Pisonia harbors the customary three plant communities: the 
herb mat is almost pure Heliotropium, dotted with Suriana. 
One specimen of Lepidium bidentatum w as found in 1 990. The 
Tournefortia scrub and forest, covering half of the motu's 
length and width, grows to 9 m. while the Pisonia— Cordia 
forest, covering 0.86 ha ( 35% of the islet's area), reached 1 m. 

Birds : Despite the beautiful Pisonia forest, no black or 
brown noddies nested. Only three species of seabirds bred in 
1988: red-footed booby (26 pairs), great frigatebird! 14 pairs). 
and white tern ( 10 pairs). Best represented were red-footed 
boobies: a perimeter count yielded 18 tended nests, all in 
Tournefortia scrub. A long-tailed cuckoo was heard in the 
interior. 

Comments : Rats were common: six were noted on a mid- 
morning transect survey. 

38. MOTU KIMOA "Rat Islet" (1.80 ha) (Figs. 29.57: 
Pis. 14.78,80) 

Fourth from the north, we named this motu for its single 
mammalian inhabitant, the Polynesian rat, kimoa in Gilbertese. 

Physiography : Kimoa, smallest of the Southern Leeward 
Islets and shaped like a flared teardrop, is squeezed between its 
neighbor motus. Its maximum dimensions are 92 m long and 
218 m wide, almost four times the size mapped by Arundel 
(Fig. 4). The southeast rubble and herb mats are wide. The 
distance to the outer reef edge is 307 m. Of special note is the 
emergent Tridacna—Acropora reef, which stretches completely 
across the lagoon to Tridacna Islet. This reef is 1 5-20 m wide 
(Fig. 48; Pis. 26. 63 land 1.023 m long, which, together with an 
equal length in blind diverticulae. totals over 2 km. The 
Tridacna clams aggregate in densities up to 80/ in' ( Sirenko & 
Koltun, Subchapter 1.4). 

Vegetation : Kimoa has II species of plants (3 trees, 
2 shrubs. 6 herbs). 419? of Caroline's flora There are no 
introductions. Though small and narrow. Kimoa is well 



48 



vegetated. Its herb mats are composed of Heliotropium on the 
south side and Portulaca (plus Suriana) on the north. The 
interior Tournefortia-Pisonia-Cordia forests (to 1 1 m) cover 
nearly one-half of the islet's area. 

Birds : Four species of seabirds bred in 1988: red-footed 
booby (21 pairs), great frigatebird (3 pairs), black noddy 
(2 pairs), and white tern (7 pairs). Red-footed booby nests were 
located along the perimeter. 

39. MOTU ANA- ANA "Anne 's Islet" ( 2. 1 6 ha) (Figs. 29,57; 
Pis. 5a. 14,54,78,81) 

This motu includes a small settlement with three thatched 
huts (cooking, eating, sleeping), a water tank, chicken coop, 
and garden. It was occupied from 1 989- 1 99 1 by Anne and Ron 
Falconer, 2 small children, chickens, Muscovy ducks, and a 
dog. When we discovered a wooden sign marked "Ana-Ana" 
and adorned with a shell lei, we knew the islet had been named. 

It is interesting to compare Pis. 5a and 8 1 , identical views 
of Ana-Ana 105 years apart. The profiles are indistinguishable, 
showing how little this motu has changed over the years. 

Physiography : Ana-Ana is the southernmost motu in the 
Southern Leeward Islets, 120 m long by 222 m wide at its 
widest point. Approximately 3 m high, it is roughly oval, with 
a hooked point and curved bay facing the lagoon. This point is 
actively growing as more and more rubble is deposited by the 
large flow of water passing through the channel (430 m wide) 
that separates Ana- Ana and South Island. This channel contains 
abundant giant clams that amass into an extensive Acropora- 
Tridacna reef stretching approximately 900 m across the 
lagoon to Tridacna Islet. The outer reef flats measured 
281 m. 

Vegetation : Ana- Ana has 15 species of plants (5 trees, 
2 shrubs, 8 herbs), 56% of Caroline's flora. Introductions 
include Cocos, vegetables, a few ornamentals and, as yet. no 
weeds. Ana-Ana's vegetation is typical of the other Southern 
Leeward Islets, except for the settlement. Narrow trails from 
the southern channel lead to a neat clearing, approximately 
40 m x 70 m, the only inhabited portion of the atoll. We have 
advised the Falconers against introducing exotic plants with 
spreading seeds and have requested them to destroy all 
introductions when vacating the island permanently. 

Ana-Ana has sparse strand vegetation: Suriana, 
Heliotropium, Portulaca, Laportea, and Lepturus. The 
Toumefortia scrub includes Cocos, Cordia, and Pandanus. A 
quality Pisonia forest, 1 5 m high, covers 43% of the islet's area. 

Birds : No breeding seabirds were found on any of the three 
visits to Caroline. However, the Falconers have found a few 
white terns and one great frigatebird nesting in the perimeter 
scrub, as well as groups of brown noddies sitting on the beach. 
Long-tailed cuckoos were seen around the huts in March, 
April, and May 1990. 

Comments : Rats are abundant. Although the house site 
was clean and tidy, 1 2 rats were seen in a pile of coconut debris, 
and others scurried amongst the forest litter. The Falconers 



have trapped over 1 ,300 rats in less than 2 years. Several pale 
geckos with a few spots and largish heads were seen in and 
around the thatched huts (probably mourning geckos). 

Conclusion 

Lushly wooded Caroline Atoll, with the majority of its 
39 islets (399 ha of land) either in near-pristine condition or 
having recovered remarkably from past disturbance, is one of 
the least spoiled atolls in the Pacific. Uninhabited except for 
one family, it harbors plant ecosystems and breeding seabirds 
(Subchapter 1.2, this volume) of national and international 
importance. Its marine and terrestrial ecosystems are prime 
outdoor ecological laboratories for research on geological 
processes including groundwater, fish poisoning, and numerous 
facets of ecology (especially plant succession). Caroline 
boasts outstanding coral reefs thickly studded with giant clams, 
substantial numbers of coconut crabs, breeding sites for green 
turtles, wintering grounds for shorebirds including the rare 
bristle-thighed curlew, ancient Tuamotuan marae, and a 
crystalline lagoon. The variety, abundance, and quality of its 
flora and fauna qualify it for status as an officially recognized 
international preserve (Subchapter 1 .2. Conservation section). 

An expedition of this magnitude entailed the help of many 
people, and it gives us great pleasure to thank them. We are indebted 
to Hal O'Connor and Randy Perry. Patuxent Wildlife Research 
Center, for making possible our participation. Steve Kohl, FWS 
Office of Internationa] Affairs, and Terry Whitledge aided immensely 
by handling innumerable details with their Soviet colleagues. Members 
of the Fish & Wildlife Service Mauna Loa Field Station, especially 
Jim Jacobi. Julia Williams. Jack Jeffrey, and Martha Moore, provided 
welcome logistical support in Hilo. and Paul Sykes willingly shouldered 
additional responsibility that freed CBK to join the expedition. 

On the Soviet side, we thank Professor Alia V. Tsyban, chief 
scientist of the expedition, for extensive help and friendship during the 
voyage. Captain Oleg A. Rostovtsev. Yevgeniy N. Nelepov. Yuri L. 
Volodkovich, and the ship's crew for handling many ship-board 
details and transporting us to and from Christmas Island and Caroline. 
Contacts with our Soviet colleagues would have been far less 
memorable without the translation skills of Svetlana V. Petrovskaya 
and. especially, Valeriya M. Vronskaya. 

We thank Greg Smith and Chuck Stafford for transporting us 
within Caroline's lagoon in the inflatable Zodiac "Tigris." Katino 
Teeb'aki shared in the hard work on transects, and his skills at 
climbing palms and opening coconuts often energized us during the 
wilting midday heat. Abureti Takaio. former Minister for the Line and 
Phoenix Groups, permitted us to work on Caroline and. with the 
residents of Christmas Island, arranged a memorable evening of 
dancing and food, despite the fact that their last supply ship was 
10 months previous: for this and their many kindnesses we are most 
grateful. 

Financial assistance for the 1988 expedition and for writing the 
manuscript was provided by the US Fish & Wildlife Service. Patuxent 
Wildlife Research Center, and the Natural Environment & Climate 
Monitoring Laboratory. Goskomgidromet USSR. 

Grateful thanks are extended to Derral Herbst for identifying and 
preparing plant specimens (deposited in the B. P. Bishop Museum. 
Honolulu. Hawaii ) and George Zug for identifying lizards (deposited 



49 



in the US National Museum. Washington. DC). The Royal New 

Zealand Air Force supplied the aerial photos. Roger Clapp, Ray 

rg, Gene Hclfman, Harry Maude. Ernst Reese, and David 

Stoddart shared unpublished manuscripts and other information. We 

are most grateful to Lynda Garrett and Wanda Manning of the 
Patuxent Wildlife Research Center I ibrary . Laurel, Maryland, for 
digging out obscure historical references. Harry Maude of the 
Australian National University introduced us to plantation records 
and similar "gray literature." which proved indispensable in 
understanding Caroline's past and present ecology. The libraries and 
herbarium at the University of Georgia were also useful. We 
especialK thank Bonnie Fancher for her efficiency, enthusiasm, and 
hard work, often late at night and on w eekends. on the computer and 
in other clerical matters. The manuscript has benefitted from reviews 
by Ron and Anne Falconer. Ray Fosberg. Pat Roscigno, Betty Ann 
Schreiber, Fred Sibley. Terry Whitledge. and Stephen Zeeman. 

AKK. as coleader of the ICBP 1 990 expedition, expresses much 
gratitude to Christoff Imboden (International Council for Bird 
Preservation) and coleader/expedition initiator Martin Garnett, for 
sharing finances. Thanks also to Martin and Annie Garnett. John 
Phillips, and Mark Linsley for help with field work, and to Alve 
Henricson for his sailing skills. The expedition would not have been 
successful without the dedication of Captain Graham Wragg. 



skipper/owner of TeManu, w ho transported us 7.400 km in the central 
Pacific (including two \ isils to Caroline), helped with field work, 
and whose competence and consideration in many areas eased the 
varied hardships associated with 3 months at sea in a 10.5-m 
ketch. Thanks also to Scott Miller for providing insect \ials and for 
preparing and depositing insect specimens in the Bishop Museum. 
Hawaii. 

On remote Caroline, the Falconers were exceptionally hospitable 
hosts, developing a special interest in its wildlife and helping us with 
field work during and after the expedition. Special thanks go to 
7-year-old Alexandre, who discovered the first blue-gray noddy nest 
for the island and 3 new plant records. 

French Polynesian residents who assisted in various ways 
include Jacques Florence. Les and Gloria Whiteley. Rick Steger. 
Michael Poole. Jean Roudeix. and friends who supplied us with 
fruit and \ egetables for the trip. We particularly thank those who have 
aided us in follow-up conservation efforts: Kelvin Taketa. 
Jim Maragos, and staff of The Nature Conservancy-Hawaii, 
Christoff Imboden and staff at ICBP. Alex du Prel. Jean-Michael 
Chazine, Philippe Siu. George Monet. Graham Wragg. the 
Falconers. Customs authorities in Papeete. George Ariyoshi. 
Secretary and Minister to the Line and Phoenix Islands, and 
the Hon. Secretary to the Cabinet. Republic of Kiribati. 



50 



TABLE 1 

Plants reported from Caroline Atoll, but considered to be 
transient or extinct members of the flora.' 



Scientific Name 



English and 
Gilbertese Names 



Date Last 
Reported 



Comments 



CLASS ANGIOSPERMAE 
Family Graminae 

Eleusine indica (L. ) Gaertn. 

Eragrostis amabilis 

(L.)H.&A. 
Family Cyperaceae 
°Kyllinga brevifolia Rottb. 

Family Bromeliaceae 
Ananas comosa L. 
Family Liliaceae 



goosegrass. te uteute 1 884 

lovegrass, te uteute 1884 



kyllinga 



pineapple, te bainaboro 1884 



Crinum sp. 


lily, te kiebu 


1884 

Presently 

cultivated 


Family Moraceae 






°Artocarpus altilis (Park.) Fosb. 


breadfruit, te mai 


Presently 
cultivated 


Fiats carica L. 


common fig, te biku 


1884 


Family Basellaceae 






Boussingaultia gracilis Miers 


Madeira vine 


1884 


Family Leguminosae 






Inocarpus fagifer 


Tahitian or Pacific 


1840 


(Parkins, ex Z) Fosb. 


chestnut, mape (Tahiti), 




(= 1 . fagifents) 


te ibi 




Family Euphorbiaceae 






Euphorbia hirla 


garden spurge, sleeping 


1884 


(= E. pilulifera) 


plant, te kaimatu 




Family Guttiferae 






Calophyllum inophyllum L. 


Alexandrian laurel. 


1884 



Family Caricaceae 
Carica papaya 



Family Cucurbitaceae 
Cucurbita pepo L. 

lpomoea batatas L. 



I.pes-caprae brasiliensis 
(L.) v. Ooststr. 



Family Scrophulariaceae 
Russelia equisetiformis 
Schlecht. 



tamanu (Tahiti), te itai 



Papaya, pawpaw, 

te babaia, te mwemwera 



Pumpkin, te baukin. 

te bamakin 

Sweet potato, te kumara 



beach morning glory, 
pohuehue (Hawaii), 
te ruku 



Coral plant, te kaihaun 
("golden plant") 



Presently 
cultivated 



Presently 
cultivated 
Presently 
cultivated 



1884 



Introduced weed 
Introduced weed 



One clump by cistern. South Is., 
on recently disturbed ground 

Introduced for cultivated fruit 

Introduced ornamental. One 
small specimen found on South 
Is. by Anne Falconer, 1990. 
Collection no. K-90-14 

Not yet established, 2 trees on 
South and Ana-Ana 
Introduced for cultivated fruit 

Introduced "vine climbing over 
portico" (Trelease 1 884) 

Unsuccessful introduction in 
1834. Food plant 



Introduced weed, unsuccessful 



In the 1940's. a "few taller 
Calophyllum and Pisonia" 
(N.I.D. 1943). No other 
reference; did observer confuse 
Calophyllum with Cordial 

Cultivated for fruit in 1884. 
not seen in 1965. In garden on 
Ana-Ana, one on South Is. by 
cistern 

Cultivated in 1884, not found 

in 1965 

Introduced in 1840, not reported 

again until this expedition (tubers 

brought in 1988). Collection 

nos. K-159, 160 

Found in 1965 by copra shed; 

extensive searching on 3 

expeditions in 1988 and 1990 

failed to find it 

Unsuccessful introduction in 19th 
century 



'Since 1988, the Falconers have added more vegetables and ornamentals to their ever-expanding garden: green 
beans, lemon grass, peppermint, okra, banana, Tahitian gardenia (tiare), tomato, breadfruit, red hibiscus, etc. 
°Not previously reported from Caroline Atoll. 



51 



TAB1.K 2 



Vascular flora ol Caroline Atoll: relative abundance of each species within the major ecosystems, with data on seabird utilization 



Scientific Name 



(.minion & Gilbertese 
Name 



Seabird 
Utilization 



o 
o 

2- 



Nalural Ecosystems 



Coastal 



Natural Beach Town. 
Herb Scrub Scrub 
Mat With 
Suriana 



Inland 



Toum. 

Forest 



A nt h ropogen ic Ecosystems 

Coconut Woodlands 

Pisonia Cocos Dying Mixed 

Forest Plantation Cocos/ Forest 

Ipomoea With 

Plantation Cocos 



TREES 
Pisonia grandis 

Morinda citrifolia 
Cocos nucifera 
Cordia subcordata 

Pandanus tectorius 

'Hibiscus tiliai ens 

'Thespesia populnea 

SHRUBS 

Tourneforlia argentea 
Suriana maritima 

'Ximenia americana 
'Scaevola s< rit ea 

Species A 

HERBS 

Heliotropium anomalum 

Boerhavia repens 
Ponulaca lulea 

Laportea ruderalis 

'At hvranthi s i anesi ens 
Phymatosorus scolopendria 



I 1 punas repens 
Ipomoea nun ranlha 

Tacca leontopelaloides 



Lepidium bidentatum 
Psilotum million 

Phyllanlbus amarus 
Sida falkn 

nana sp. 
Iiihiilus i istoides 



pisonia, puka tree, 

te buka 

Indian mulberry, te mm 

coconut, te ni 

sea trumpet, kou 

(Hawaii I. ic kanawa 

pandanus, screwpine, 

te aroka, te kaina 

beach hibiscus, hau 

(Hawaii). ic nui 

milo (Hawaii ). te bingibing 



tree heliotrope, te ren 
bay cedar, te aroa, te 
mount 

monkeyplum 
scaevola, saltbush, 
half-flower, te mao 



"sand rose", hnitiliinn 
( Hau an I 
pigvine, te wao 
yellow portulaca, 
seaside purslane. 
te boimari, te hot 
"nettle", te ukeuke, 
te nekeneke 

maile-scented fern, 
Itittti'c or lawai fern, 
te raukota, te keang 
bunchgrass, te uteute 
morning glory, wild 
moon-flower, te ruku 
Polynesian or island 
arrowroot, pia t Hawaii 
& I ahiti), n- makemake 
peppergrass 
upright psilotum. "reed 
fern", te kimarawa 

'ilimti (Hawaii), te kaura 
crabgrass, te uteute 
puncture vine, te kiebu 



R 


UC 


C 


C 


R-C 


R-VC 




R-O 


UC 




UC 


UC-C 


O 


C 


c 



X X o-uc vc 

O VC-A 
R 



X A UC 
R-UC R-C 



A A 
O 



O-A R-UC 
R-VC VC 

R 
R-UC UC-A 



PR 



VC 

R-A R-A 



UC-C O 

C-A C 
R-C C 



LC LC 

L.R-A 

R R 



LR 



A 


L.UC 


L.UC 


R-C 


R-A 


R-C 


C 


O 


O 


A 


A 


C-A 


R-VC 




L.UC 


R 





O 

1..R 
L.R 




A 



UC-A 



UC 



LA 



R-A 



L.O 

LC 
C 



C-A.L 



R-UC 



UC-A UC C-A 



R-C R 

LC R-UC UC-C 
O A 



R-A R-UC C-A 
R 

UC A 



I \ 



LR 
PR 
PR 
(S) 



Excludes transient and extinct species i | able 1 1 Species arranged according to then overall abundance on the atoll. 

New records foi Caroline. 

Mot seen on this expedition, lasi seen 1965 K'lapp & Sibley, 1971a). 



TABLE 3 

Distribution and abundance of plant species on Caroline Atoll. 1 Motus are arranged geographically from 
north to south (windward), then similarly on the leeward side. 



Windward Motus 



South Nake Motus 





N 


L 


B 


W 


c 


A 


N 


P 


S 


N 


B 


N 


N 


A 


T 


S 


P D 


B 


C 


L 


K 


M 




a 








i 


r 


t 





i 


k 





r 





o 


r 


r 





a a 











o 







k 


n 




ii 


e 


i 


r 


g 


u 


r 





d 


r 


u 


i 


u 


n n 





r 


n 


t 


ii 




e 


IT 


s 


d 


s 


b 


t 




1 


t 


t 


d 


t 


n 


d 


t 


d g 


b 


a 


e 


a 


a 








u 


w 


c 


u 


h 




1 


h 


h 


y 


h 


d 


a 


h 


a e 


y 


1 






k 








n 


a 
r 


e 
n 




P 






B 


e 
r 


R 


A 


e 
1 


c 
n 




n r 
u 






P 

a 




e 
n 








B 

i 


d 


t 




i 
g 






r 




s 




c 


r 
u 




a 




s 






1 
m 




a 








r 














t 




k 


n 


























d 














h 
e 
r 

s 






d 
e 

1 




















TREES(7spp.) 














































Pisonia grandis 


A 


vc 




A 


A 




A 


A 




A 


A 




VC 


A 


R 


R 


VCA 


VC UC 


R 


R 


R 


Morinda citrifolia 


UC 


UC 




UC 


O 




R 


UC 




R 


O 




o 


UC VC 


C 


R R 


R 


R 


R 


R 




Cocos nucifera 


LA 


LA 












LR 






LO 




R 






A 








LR 


LR 




Cordia subcordata 


UC 


UC 




c 


LA 




C 


VC 




C 


VC 




C 


UC 




R 


LA 












Pandanus tectorius 


LA 






























LO 


LA 












"Hibiscus tiliaceus 
































LR 














°Thespesia populnea 
































LR 














SHRUBS (5 spp.) 














































Toumefortia argentea 


A 


A 


VC 


A 


A 


O 


A 


A 





A 


A 


O 


A 


A 


A 


C 


A A 


A 


A 


A 


C 


A 


Suriana maritima 




LR 


















LR 








LC 


UC. 
LC 















SkicvoIci sericea 








S 


































Ximenia americana 






























LA 












°Species A 










































HERBS (15 spp.) 










































Heliotropium anomalum 


C 


VC 


UC 


C 


c 


UC 


UC 


UC 


R 


UC 


C 




UC VC 


C 


c 





UC UC 


UC UC 


R 


Boerhavia repens 


C 


LC 




C 


c 




c 


c 




UC 


C 




A UC 


c 


VC 


c 


O 


O R 


R R 


R 


Portulaca hitea 


C 


C 


c 


C 


c 


UC 


UC 


c 


c 


A 


A 


A 


A C 


c 


c 


cue 


O O 


O O 


R 


Laportea ruderalis 


C 


VC 




c 


UC 




UC 








C 




C 


UC 


UC 


c 


o 


O R 


R R 


R 


°Achyranthes canescens 


LC 


O 






o 




c 


o 




LC 






VCVC 


LC 


R 


o 





R R 


R R 


R 


Lepturus repens 


R 


R 


R 


R 






R 






R 


R 


C 


c 


C 


O 


R 


R 


R R 


R R 


R 


Phymatosorus 


VC 


LA. 






UC 




UC 


UC 










o 


C 


LA. 








R R 




scolopendria 




UC 


























UC 












Ipomoea macrantha 


LC 







UC 








o 














A 












Tacca leontopetaloides 


*LC 




























LR 












Psilotum nudum 


*LR 




























LR 












Phyllanthus amarus 






























LC 












Tribulus cistoides 




s 






































Sidafallax 






























LR 












Lepidium Indentation 






























S 












*Digitaria sp. 






























s 













53 



TABLE 3 - continued 













Central 












Southern 




















Leeward Motus 








Leeward Motus 


T 


% O 






M 


B 


M 


E 


S 


S 


N 


A 


R 


B 


F 


R 


E 


P 


K 


A 


O 




u 


1 


a 


m 


h 


c 


a 


z 


e 


i 


i 


a 


i 


i 


i 


n 


o 


n 


r 




n 


a 


t 


e 


a 


a 


u 


u 


e 


r 


s 


u 


t 


s 


m 


a 


t 


F 


i 




n 


c 


a 


r 


r 


r 


1 


r 


f 


d 


h 


r 


e 





o 


- 


a 


r E 


g 




i 


k 


w 


a 


k 


1 





e 


- 




b 


a 


i 


n 


a 


A 


1 


e a 


i 




k 

i 


1 
i 


a 


1 

d 




e 

t 


n 

CT 




f 

1 




a 
1 


u 




i 
a 




n 

a 




q c 
u h 


n 




b 


i) 










a 




a 




1 














e 






a 










C 

r 
a 

b 






t 


















n M 
C o 

y i 
u : 




TREES(7spp.) 








































Pisonia grandis 


C 


VC 


UC 


C 


A 




O 


R 




A 




A 


C 


A 


VC 


A 


32 


82 


1 


Morinda citrifolia 


R 




R 


R 


R 




R 






R 


D 


UC 


R 


R 


R 


C 


30 


77 


LAI? 


< .'. os nucifera 


LR 


LR 




LO 


LR 










LR 




LR 




LO 




LC 


17 


44 


AI.RI 


Cordia subcordata 


O 


UC 


C 


O 


LO 




O 






UC 




C 


C 


C 


C 


C 


23 


59 


I 


Pandanus tectorius 






R 


C 


C 














LR 




LR 




LR 


9 


23 


LAI? 


"Hibiscus tiliaceus 


































1 


3 


I(AI,RI?) 


"Thespesia populnea 


































1 


3 


1 i AI.RI?) 


SHRUBS (5 spp.) 








































Tournefortia argentea 


A 


A 


A 


A 


A 


R 


C 


C 


R 


A 


UC 


A 


A 


A 


A 


A 


39 


100 


1 


Suriana maritima 






R 














UC 






LR 


LR 


LR 


O 


10 


19 


1 


Si ,h \ ola serii ea 


































1 


3 


1 


'Ximenia mucin una 


































1 


3 




Species A 


S 
































1 


3 


1 


HERBS (15 spp.) 








































Heliotropium anomalum 


C 


C 


C 


C 


C 


R 


C 


C 


R 


R 


UC 


UC 


C 


C 


C 


UC 


38 


98 


I 


Boerhavia repens 


C 


UC 


UC 


C 


UC 


R 


R 


R 




UC 


R 


UC UC UC 


O 


c 


34 


88 


1 


Portulaca lutea 


c 






c 


uc 


R 


R 


R 


R 


UC 


R 


UC 


C 


UC 


C 


LC 


37 


95 


1 


Laportea ruderalis 


uc 


uc 


o 


c 


uc 


R 


R 


R 




R 


R 


UC VC L1C 


O 


c 


32 


82 


1 


\i In ranthes i anescens 


uc 


o 


R 


uc 


LA 




R 


R 




UC 


R 






o 


C 


o 


29 


72 


I 


Lepturus repens 


R 


R 




R 




R 








UC 


R 






R 


R 


VC 


27 


69 


1 


Phymatosorus scolopendr 


uO 








R 






















uc 


14 


3d 


1 


Ipomoea macrantha 




























O 




R 


7 


18 


I 


Idcca leontopetaloides 




































3 


RI.AI * 


PsilotUtn milium 




































3 


I 


Phyllanthus animus 




































3 


X 


Tributes cistoides 




































3 


1 


Sidafallax 




































3 


I 


I epidium bidentatum 




























S 






2 


5 


I 


'Digitaria sp. 


































1 


3 


•> 



Species arranged according to frequency of occurrence. List excludes transient and extinct members of the flora i [able 1 1 
No. motus ha\ ing a particular species. di\ ided in total no. motus \ 100' I 
New records foi Caroline. 
■ Not seen on three visits, bul possiblj still present. 



54 



TABLE 4 

Sizes of Pacific atoll floras, with emphasis on the 
percentages of indigenous plants. 1 







Total 2 


No. 










No. 


Species 


% 




Island Group 


Atoll 


Species 


Indigenous 


Indigenous 


Source 


Caroline Is. 


Kapingamarangi 


98 


38 


39 


Niering, 1962 


(Fed. States 












of Micronesia) 












Cook Is. 


Aitutaki 


45 




50 


Stoddart & 


(New 


(motus) 








Gibbs. 1975 


Zealand) 


Rarotonga 
(motus) 


49 




ca. 60 


Stoddart & 
Fosberg, 1972 


Gilbert Is. 


Onotoa 


60 


50 


83 


Moul, 1957 


(Rep. of 


Tarawa 


109 


28 


26 


Catala, 1957 


Kiribati) 












Northwest 


Kure 


42 


23 


55 


Lamoureux, 1961; 


Hawaiian Is. 










Clay, 1961 


(USA) 


Laysan 


38 


27 


71 


Ely & Clapp, 1973 


Line Is. 


Caroline 


27 


23(25?) 


85(93?) 


This paper 


(Kiribati) 


Christmas 
(Kiritimati) 


69 


19 


28 


Garnett, 1983 




Fanning 


123 


23 


19 


Wester. 1985 




Flint 


37 


14 


38 


St. John & 
Fosberg, 1937 




Maiden 


9 


9 


100 


Garnett, 1983 




Palmyra 


58 


21 


36 


Wester, 1985 




Starbuck 


7 


4 


57 


Garnett, 1983 
(incomplete, 
little known) 




Vostok 


3 


3 


100 


Clapp & Sibley, 
1971b; Kepler, 
1990c 




Washington 


91 


25 


27 


Wester. 1985 


Marshall Is. 


Ailuk 


56 


26 


46 


Fosberg. 1955 


(Fed. 


Arno 


125 


40 


32 


Hatheway, 1953 


States of 


Enewetak 


128 


55 


43 


Lamberson, 1987 


Micronesia) 


Jaluit 


288 


55 


21 


Fosberg & 
Sachet, n.d. 




Jemo 


34 


17 


50 


Fosberg, 1955 




Kwajalein 


89 


25 


28 


Fosberg, 1955, 
1959 




Lae 


61 


35 


57 


" 




Likiep 


91 


31 


34 


" 




Taka 


23 


18 


78 


Fosberg, 1955 




Ujae 


61 


32 


52 


Fosberg, 1955, 
1959 




Ujelang 


50 


29 


58 


" 




Utirik 


55 


26 


47 


'* 




Wotho 


40 


28 


70 


" 


Phoenix Is. 


Kanton 


164 


14 


9 


Degener & 


(Kiribati) 


(Abariringa) 








Gillaspy, 1955 






129 


18 


14 


Garnett, 1983 




Birnie 


3 


3 


100 


Fosberg & 
Sachet, (n.d.) 




Enderbury 


23 


18 


78 


" 




Nikumaroro 


35 


17 


49 


" 




Orona 


ca. 29 


19 


ca. 66 


" 




McKean 


7 


7 


100 


" 




Phoenix 


6 


6 


100 


" 




Manra 


ca. 18 


14 


ca. 77 


" 



55 







TABLE 4 


continued 










TotaF 


No. 










No. 


Species 


', 




Island Group 


Atoll 


Species 


Indigenous 


Indigenous 


Source 


Society Is. 


Tetiaroa 


95 


47 


49 


Sachet & 
Fosberg. 1983 


Solomon Is. 


Ontong Java 


1 50 


58 


39 


Bayliss-Smith. 
1973 


Tokelau Is. 


Nukunono 


55 


35 


64 


Parham. 1971 


(N.Z.) 












Tuamotu Is. 


Raroia 


121 


39 


32 


Stoddart & 


(France) 










Sachet, 1969 




Rarioa 


135 


54 


40 


Doty. 1954 




Takapoto 


106 


33 


31 


Sachet. 1983 


Outlyers 


Clipperton 
(U.K.) 


31 


14 


45 


Sachet. 1962 




Oeno 


17 


14 


82 


St. John & 




(U.K.) 








Philipson. 1960 




Wake 


94 


20 


21 


Fosberg & 
Sachet. 1969 



An updated version of Table 1 I, p. 105. Stoddart and Gibbs ( 1975). 

: Number of species of those indigenous are not always comparable. Ferns are usually included. bu( 
certain ornamentals may not be. Artocarpus, Morinda, and Pandanus may be indigenous, aboriginal 
introductions, or both. Without its full scientific name, a species has an unknown biogeographical 

status. 



56 



E g 
#1 



s s 



i 3 

Is 






I I 

« .a 



1 " 
— P 

I 5 
1 . 

a j 



.a .8 



5 a. i 
1 I 

o a. 



It 3 
» 1 B 






It* 



ss 



2 < 



I I* 



X X X X X 



m ci f*% 



o o o o 



s y S s 

© © o c> 



X X X X X 



i/-i w-i vi >© t— 



- 


X 


,3 


^ 


< 


,3 


<•? 


3 

8 

5 


I 



8 S * K 3 

© © © d © 



X x X XXX XXXXXXXXXXXXXX XX 



X XXXX XXXXXXXXXXXXXXXX XX 



X X XXX XXXXXX X X XX X X X 



X XXXXXXXXXXXXXXXXXXXXXXXXX 



XX XXXX xxxxxxxxxxxxxxxxxxx 



XXXX X XX 



XX XX X 



xx=xxxxSxxxxxxxxxxxxxxxxxxx 



•OooO-cS 10 ™™^^^ 



© © © r* © — O — — M M m — m 



>©(■-]_ v©vi>©"T_,'©r-oooo'/ir->0'»«v>p-'0*or- , ©'0<>r-r- * 2 2 



^---r)nM-NM----NN------n-H 



n © * — ei 



ioei<Nmwi*f«*t(*»ci*rc>w^ , 'ir'i*'> , r 



R? 8 



""> -* "t P. ""( ^ 
O f*i ^ ^1 (N f*i 



II 



| | 

8 .2 






t.N 



< £ ill J3 



3 3-33 

o | « § « o ^ 

oazSSuSzS 



i 1 § -2 1 s 

« s § .a a s. 
s « * s. i « 5 

s 8 * I s 

SSps 



3-!J 



I ill 



i" 1 J 5 i 1 1 

1 u 2 m 03 Z a. 



I Z "a 1 

jlll 
I a £ s 



v h vi 



5 * 

3 S 



"S 1 1 

3 j 3 

in 



s 

i 

I 

1 



57 



u 

1 g 

1 £• 



il 



3 |S 

3 "I 



2 3 

.s I 



5 _ 

il 






a 



6 o 

§ I £ 

- 9 

O O — 



= = 

«S ^ — 



6 o 



Ov — w 



■_-'_- x C at 



HS< o 



< U U (J O a: U 



uoS 



< O U O 3 a! O 



O U O < _J 

uuu-r9o i 

< 3 3 3 O OC 3 -JO 



o o o o 

< 3 3 3 <J OS 3 



<a<KDa:>o 



O O -1 o 

V = = U as oi 

<a:a:OO0£3O_}-J 



<OOa;aiacOa: 



< a: a: a: x x 



> 3 O O X X 



CJ 3 O X X X 



U O X X X 



V U X X 



o 

X X 



_ - 



>■§ 



Ml 



§ fi 



•S. S- « P i 

Ml 8. > 



J <3 



1 1 



il 



4 - 



u < 






o% 



j£i 



1 
& 



£1 



5 3 



s 



a! 

m _ •? 

8> a 3 



II 

ill 



■C « 






•S s 
b s 

.1 

& g 

II 
1 5 



I -g I 

Ji| 



u a 



■S*i 



S 8 



■a* i 

II. 



.5 i ! .s S ) 
_! . £ Q z z 



58 



TABLE 7 

Widths of pioneer herb mats on seaward- and 
lagoon-facing shores, Caroline Atoll. 

Average Width of Pioneer Herb Mat (m) 
Bordering Sea Bordering Lagoon 

Leeward Motus 18.5 4.2 

(3-81) (0-28) 
Windward Motus 36.0 0.9 

(24-69) (0-3) 



TABLE 8 

Species-area relationships of six Pacific islands 
with entirely indigenous flora. 



Island 


Area 


No. Species 


Maiden 


39.3 sq km 


9 


Starbuck 


16.2 sq km 


6 


McKean 


57 ha 


7 


Phoenix 


49 ha 


6 


Vostok 


24 ha 


3 


Birnie 


20 ha 


4 



Islands are arranged according to decreasing area. 
Data is from Garnett ( 1983), Fosberg and Sachet 
(n.d.). Clapp and Sibley ( 1971b), and pers. obs. 



TABLE 9 

Area of plant communities on the islets of Caroline Atoll. 







Total 




Area (ha) 


Area (ha) 


Unvegetated Habitats 




41.39 


Coral Rubble and Sand 


41.39 




Natural Plant Communities 




261.41 


Natural Herb Mats 


67.73 




Beach Scrub with Suriana 


1.49 




Pandanus Forest 1 


3.38 




Tournefortia Scrub and Forest 


125.25 




Cordia Forest : 


1.39 




Pisonia Forest 


62.17 




Anthropogenic Community 




96.14 



Coconut Woodlands 96. 1 4 

Total Area Above High Water 398.94 

1 Pure Pandanus only. Also mixed with Pisonia. Tournefortia. and Cocos. 

2 Cordia. where mixed with Pisonia and Tournefortia, is included in totals 
for those forest communities. 



59 



TABLE 10 

Stature and extent of Toumefortia in the major habitats 
of Caroline Atoll. 



Av. Hgt. (m) Av. Width (m) Toumefortia No. No. 

(in) (m) Cover Motus Transects 



Natural Herb Mat 

Toumefortia 
Scrub & Forest 
Toumefortia - 
Pisonia Forest 



1.4 

(0.3-1.8) 

6 

(0.3-15) 

9.5 

(5-15) 



49 
(3-198) 

55 
(2-287) 

98 
(8-284) 



25 


14 


20 


(5-95) 






81 


38 


71 


(5-100) 






47 


18 


27 


(5-90) 







TABLE 1 1 

Distribution of well-developed (>I0 m height) Pisonia forests on the motus 

of Caroline Atoll. 1 Motus and transects are arranged according to the 

decreasing height of their Pisonia groves. Capitals indicate those motus 

whose forests were felled for Cocos plantations from 1916-20. 







Area of 






Pisonia 


Pisonia 


Motu Area 


Motu & Transect 


Height (m) 


(ha) 


(ha) 


PIG 


21* 


3.36 


7.21 


NAKE, Transect 4 


20* 


20.79 


107.46 


Booby 


20* 


0.12 


0.84 


NORTH PIG 


20* 


1.83 


5.44 


NORTH BROTHERS 


18* 


0.43 


1.71 


NAKE, Transect 3 (central) 


15* 


20.79 


107.46 


LONG, Transect O 


15* 


15.00 


75.98 


BROTHERS 


15* 


0.37 


4.31 


Ana-Ana 


15* 


0.93 


2.16 


Danger 


15* 


0.39 


2.71 


NAKE, Transect 2 


14* 


20.79 


107.46 


Bird 


14* 


1.70 


4.05 


WINDWARD, Transect 2 


14- 


2.97 


11.42 


Raurau 


14* 


1.07 


3.48 


CRESCENT 


13- 


0.51 


3.10 


Mannikiba, Transect 1 


12* 


1.13 


21.49 


Shark 


12- 


2.60 


7.98 


NAKE, Transect 3 (west) 


12- 


20.79 


107.46 


LONG, Transect 1 2 


12- 


15.00 


75.98 


Pisonia 


11* 


0.86 


2.45 


Matawa 


1 1 


0.07 


1.71 


Nautonga 


11 


0.02 


0.34 


NAKE, Transect 3 (southwest) 


II- 


20.79 


107.46 


Kimoa 


11 


0.59 


1.80 


Emerald 


1 1 


3.20 


8.34 


Eitei 


1 1 


0.38 


1.42 


LONG. Transect B 


10- 


15.00 


75.98 


LONG. Transect 8 


10 


15.00 


75.98 


NAKH. Transect 1 


10- 


20.79 


107.46 


WINDWARD. Transect 1 


10 


2.97 


11.42 


Blackfin 


10- 


0.41 


2.62 


NORTH ARUNDEL 


1 It- 


0.18 


0.91 



Toumefortia or Cordia mas be present, but sub-dominant to Pisonia. 
* 90-100% canopy cover. 
50 SO', canop) cover. 



60 



TABLE 12 



Area and Dimensions of Pisonia grandis on Vostok, Flint, and five islets of Caroline Atoll. 



Island/Islet 


Area of 


No. trees 


Mean 


Range of 


Mean 


Range 


Mean 


Range of base 




Pi si >n ui 


or main 


Heigth 


Heights 


cbh 1 


of cbh 


base 2 


circumferences 




ha 


trunks 


(m) 


(m) 


(m) 


(m) 


(m) 


(m) 


CAROLINE 


62.17 
















North Pig 


1.83 


25 


19 


11-21 


221 


110-359 


261 


205-470 


Brothers 


0.37 


10 


15 


15 


140 


50-219 


243 


154-340 


Pig 


3.36 


5 


16 


12-17 


338 


290-660 


282 


230-333 


North Brothers 


0.43 


3 


18 


18 






314 


293-332 


Long 


15.00 


3 


15 


15 






414 


282-500 


Total for above 


















islets at Caroline 


20.99 


46 


18 


11-21 


213 


50-660 


293 


154-500 


VOSTOK 


1 3.5 


58 


18 


10-25 


218 


67-510 






FLINT 


approx. 4 


















(fragmented) 20 


17 


8-30 


160 


60-200 


598 


100-1000 



cbh = 'ircumference at 1 .5 m. 
base = base circumference at 0.3 m. 



TABLE 13 

Number of trees and areas planted in Cocos on Caroline's islets 
during the major planting era ( 1916-1920). also showing remnant Cocos data for 1990. 























Approx. 
















1916- 




% 1990 






Approx. 


Forest 




Islet 




Area Forest of Scrub Usable 




No. 


Area 


% 


Planted 


Islet 


Area 

(ha) 




for Cocos in 1990 1 (ha) 




Cocos 
Planted 


Cocos 2 
(ha) 


Cocos 
1990 


in Cocos 




Town. 


Pis. 


Other 


Cocos 


Total 


70 Yrs Ago 


South 


104.41 


4.20 





1.10 


80.00 


86.10 


13.006 


94.90 


477 


100% 


Nake 


107.46 


30.65 


20.8 


9.41 


5.75 


66.61 


10,544 


76.97 


6 


100% 


Long 


75.98 


32.20 


15.00 


- 


2.40 


49.60 


1.343 


9.80 


3 


20% 


Tridacna(Al) 1 


9.08 


7.97 





0.18 




8.15 


910 


6.64 





82% 


Arundel 


7.34 


4.36 


0.95 


- 





5.31 


646 


4.71 





89% 


N. Arundel (A2) 


0.91 


0.33 


0.19 


- 


few trees 


0.52 


69 


0.50 





100% 


Brothers 


4.31 


2.00 


0.37 


- 


0.0 1 


2.38 


315 


2.30 


0.2 


97% 


N. Brothers (A3) 


1.71 


0.68 


0.43 


- 


few trees 


1.11 


180 


1.31 





100% 


Pig 


7.25 


1.61 


3.36 


- 


0.03 


5.00 


538 


3.93 


0.4 


79% 


N. Pig(A4) 


5.44 


1.31 


1.84 


- 





3.15 


402 


2.93 





93% 


Crescent (A5) 


3.10 


1.56 


0.51 


- 





2.07 


228 


1.66 





80% 


Windward (A6) 


11.42 


5.70 


2.97 


- 





8.67 


1,299 


9.48 
215.10 


(1 


100% 



' "Usable area" does not include unvegetaied rubble or natural herb mats. 
- Based on Caroline's planting densities of 28 x 28 sq ft (Young ca. 1922). 
1 The "A" series of islet names are from Youna (ca. 1922). 



TABLE 14 

Species diversity in Pisonia forests of decreasing maturity. Caroline Atoll. Motus are 
arranged according to the degree of coverage of their constituent Pisonia trees; within 

these groupings.the motu order is dependent upon the total number of plant species 

within this same habitat. Note that there is an inverse relationship between the purilv <<\ 

the true Pisonia forest and species diversity. 



Canopy 


Canopy 


Av. 


Total Ni 


jmbers of Species 


No. 


Cover 


Hgt. 


No. 


Trees 


Shrubs 


Herbs 


Transects 




(mi 


Spp. 


(inch Pisonia) 










100% (Pisonia only) 


13 


1 


1 












1 


100% (co-dominant present) 


15 


3.4 


2 









5 


9 


90-95% 


10 


5.2 


5 




1 




7 


9 


50-909 f 


10 


6.2 


5 




2 




7 


15 


25-50% 


7 


6.2 


4 




1 




6 


5 


<25%' 


9 


5.63 


3 




2 




9 


6 


Motus with one 


6 


4.0 


3 




1 




3 


2 


Pisonia tree only 



















South Island not included, as its Pisonia is too rare and fragmented. 



62 



t ropic of CancT 



* . 



HAWAII 



20 



equator 



KINGMAN REEF 
■ PALMYRA ATOLL 

^ • WASHINGTON IS. 

/ • FANNING IS. 



KIRITIMATI 

[CHRISTMAS IS. I 



10 



• JARVIS IS. 



PHOENIX 

group 

TOKELAU 



AMERICAN 
SAMOA 



KIRIBATI 



Q 



MALDEN 
IS. 



FILIPPO REEF 



STARBUCK IS 



o 

-0 
VOSTOK IS. 

FLINT IS. 



CAROLINE 
ATOLL 



^"AS , s 



10 



FRENCH 
POLYNESIA 



WESTERN 
SAMOA 

NIUE 



COOK ISLANDS 



=>*::».>, 



'K 



«.<,"*, 



tf 



.TAHITI * 



y is 



u^ 



<© 



r <? 20 
o 



tropic of Capricofn 



Fig. I. Line Islands: geographic location in the Pacific Ocean. 



63 







WINDWARD 
ISLETS 



SOUTHERN 
LEEWARD 
ISLETS 



CAROLINE ATOLL 



Fig. 2. Caroline Atoll. Republic of Kiribati, with newly-named islets. Based 
on photos by the Royal New Zealand Air Force, Aerial Plan No. 1036 
9d-h ( 1986). Photogrammetric Branch, Dept. of Lands and Survey, 
RNZAF. New Zealand. 




Fig. 3. Main marae on Nake Island, Caroline Atoll, based on a plan published 
by the Solar Eclipse Expedition ( Holden & Qualtrough, 1 884 ). Figures 
in the margins show side views of the peripheral blocks shown in the 
plan. The two end walls are represented in ground plan. 





Fig. 4. Caroline Atoll, as surveyed by John Arundel, 1883 This is still the 
standard hydrological chart for the atoll (Admiralty Chart No. 9791. 
Thoughquiteaccurate.it has never been used in scientific publications 



Fig. 5. Caroline Atoll, as charted bj the Solar Eclipse Party, also in 1883 
(Holden & Qualtrough. IS84). 



M 



ISO'lS' 150 I4W 



I50°I3' 



CAROLINE 
ISLAND 

From RO. S28. 




Boat Entrance 
Settlement/' j' 



2 nautical miles 



Fig. 6. Caroline Atoll, a modified version of the Solar Eclipse Party's map 
(1883) as portrayed by Bryan (1942). Though highly inaccurate, 
modifications of this map have been used in all publications since 
Bryan (1942). 

N 

t 



SURVEY 
TRANSECTS 

campsites 




I 



CAROLINE ATOLL 

SCALE 1 y* 000 
$00 900 IOO0 2000 

I — I I I I 1 

matari 



Caroline Atoll 




_kJ 



Fig. 7. This map, by Clapp & Sibley (1971a), was based on Fig. 6. 



PERIMETER 
SURVEYS 




CAROLINE ATOLL 

SCALE 134 000 
500 900 I0O0 2000 

inn I 1 1 1 



Fig. 8. Caroline Atoll: survey transects. The distance covered was 13.3 km. Fig. 9. Caroline Atoll: perimetersurveys. The distance covered was 19.3km. 



• 



- coral 
(Acropora spp.) 

- mollusks 
(Tridacna maxima) 

- coral 

(Fungia granulata) 



m* - Holothurioidea 
(Ludwigothuria sp. 

(ww - algae 

(Porolithon sp.) 

# - algae 

[Halimeda sp.) 




iMiiafei^*"^^"* 






edge lateral 



central 




Coral limestone 
Sand 



lateral I edge 



ZONES 

Fig 10. 1 )iagrammatic representation of a portion of the outstanding Acropora-Tridacna reels connecting the islets Ana-Ana, Kimoa and Tridacna. 
(Tridacna maxima), thickly studding the reel', attain densities of 80 per square meter (Sirenko & Koltun. Subehapter 1.4). 



Giant clams 



RARE AND 
LOCALIZED PLANTS 

|^< | PANDANUS 

| m | SCAEVOLA 

| • | PSILOTUM 

["">"] HIBISCUS. THE 

j | TACCA 

fT~| TRIBULUS 

[~T~] SPECIES A 

| a | LEPIDIUM 

| I PHYLLANTHU 




PHYMATOSOR 
SCOLOPENDR 



Pig I i i nnii' distribution map "l rare and/or localized plants on Caroline 

Hibisi us liliaceus, I epidium bidenlalum, Pandanus tectorius. 

Phyllanihus amarus, Psilolum nudum, 8 m vola terit ea, Sida fallax, 

leontopelaloides, Thespesia populnea, Tribulus cistoides, 

Ximenia americana, and Species V Psilotum ma\ still exist on Nake 




Fig. 1 2. Transect distribution map of the fern Phymatosorus scolopendria on 

Caroline Atoll 



66 



LEPTURUS REPENS 




COCOS NUCIFERA 

tj££3 Healthy Cocos 

■'.'■] Dying Cocoa- Ipomoea Fore 

"....] Mixed Forest 

■ Isolated groves 



CAROLINE ATOLL 




CAROLINE ATOLL 

SCALE i J* 00'.' 
900 O SOO 1040 1000 

'■■■■I I I I I 



Fig. 13. Transect distribution map of the grass Lepturus repens on Caroline Fig. 14. Entire distribution map of the coconut Cocos nucifera on Caroline 
Atoll. Arrows indicate areas of highest density. Atoll. 



LAPORTEA RUDERALIS 




ACHYRANTHES CANESCENS 



CAROLINE ATOLL 



Fig. 15. Transect distribution map of Laporlea ruderalis on Caroline Atoll. Fig. 16. Transect distribution map of Achyranthescanescens. Arrows indicate 
Arrows indicate areas of highest density. areas of highest density. 




67 



BOEflHAVIA REPENS 




J PISONIA 
I 1 PISONIACOBOIA 

~2 PI SONIA • TOURNEFORTI A 
[^H Ml XEO FOREST 



PISONIA GRANDIS 




Fig. 17. Transect distribution map of pigvine, Boerhavia repens, Caroline Fig. 18. Entire distribution map of the buka tree, Pisonia grandis, Caroline 
Atoll. Arrows indicate areas of highest density. Atoll. Arrows indicate forests from 10 to 21 m tall. 



POBTULACA LUTEA 




SURIANA MARITIMA 



CAROLINE ATOLL 

•CALI i J« ooo 

900 SOO 1MO 1040 

'■■■■' I I I I 



Fig 19 Transect distribution map of the succulent herb Portulaca lutea, Fig. 20. Transect and perimeter survey distribution map of Suriana maritma. 
Caroline Atoll. Arrows indicate pure Portulaca Hats. Arrows indicate areas of highest densit] 




68 



IPOMOEA MACRANTHA 





Fig. 21. Transect distribution map of Ipomoeamacrantha. Entire distribution 

is shown for South and Nake Islands. Arrows indicate areas having Fig. 22. Entire distribution map of CorJiu subcordata. Arrows indicate small 
significant amounts of this vine. but monolypic. stands. 



m TOURNEFORTIA 

] SAVANNA"- HERB MAT 

'"■•1 MIXED FOREST 

HELIOTROPIUM \ ^ ~.% \ ^s; TOURNEFORT I A ■ MOR IN DA 

ANOMALUM 



TOURNEFORTIA ARGEN 




CAROLINE ATOLL 



Fig. 24. Entire distribution map of Tournefortia argentea. Because this shrub 
Fig. 23. Entire distribution map of Heliotropium anomalum. Arrows indicate dominates Caroline^ woodlands, there are no individual arrows to 

areas of highest density. indicate areas of high density. 




69 



MORINDA CITRIFOLIA 




CAROLINE ATOLL 

SCALE 1:34000 
S00 I0O0 2000 

'■■■■' 1 I I I 



100 



in 



80 



C3 

z 



< 



50 



30 



20 - 



10 






TOURNEFORTIA PISONIA 
NATURAL FORESTS 



MIXED COCOS HOUSE SITE 
ANTHROPOGENIC FORESTS 



Fig. 25. Transect distribution map of Morinda citrifolia. The outlined area on 
rridacna Islet (northeast of South Island) encloses Toumefortia- 
\4orinda forest. Arrows indicate areas of highest density. 



Fig. 26. Evidence for the indigenous status of Morinda citrifolia on Caroline 
Atoll : percentage cover on transects within natural and anthropogenic 
forests. Morinda occurs on 30 ( 7796 ) motus. never in a "planted" 
situation. 



70 



LEFT CIRCLE 

MAJOR PLANT COMMUNITIES 
UNVEGETATED 
HERB MAT 



RIGHT CIRCLE 
PERCENT INDIGENOUS 



^1 



TOURNEFORTIA 
PISONIA 

cocos 

SURIANA 
PANDANUS 



INDIGENOUS 

ABORIGINAL INTRODUCTION 

RECENT INTRODUCTION 



Wll HOUSE SITE 



MIXED FOREST WITH COCOS 



NODDY ROCK 
0.02 ha 



0.005 



0.015 
(75%) 




SKULL ISLET 0.005 

0.02 ha (5%) 

^s^,^ 0.019 

40l§&i0 ? (95%) 





MOTU ATIBU 
0.02 ha 



0.004 
(2%) 



r 



0.02 

I* (98%) 





REEF-FLAT ISLET 
0.09 ha 



fc 



0.02 
(22%) 





Fig. 27. Plant communities and amount of indigenous vegetation on motus less than 0.I ha, Caroline Atoll. The left "pie" depicts the relative amount of amotu's 
total surface area covered by each plant community: numbers indicate actual area in hectares. The right "pie" depicts the numbers and percentages of 
indigenous and anthropogenic species per motu. Data is based on the vegetation maps for each motu (Figs. 37-57) and Tables 2 and 9. 



71 



1.01 (5% 



AZURE ISLE 
0.20 ha 



o.ii m 

(55%H 



0.04 
20%) 





MOTU NAUTONGA 
0.34 ha 



0.13(38%) 



0.02 (6%) 




(11%) 



0.08 (24%) 




SCARLET CRAB ISLET 
0.46 ha 

0.11 (24%) 



0.18(39%) 





FISHBALL ISLET 
0.57 ha 

0.16(28%) 




0.17(30%)^^^ 




0.01 



0.03 



MOTU KOTA 
0.64 ha 




0.29 
(45%) 




Fig. 28. Plant communities and amount of indigenous \ egetation on motus 0.2 to 0.7 ha. Caroline Atoll. See Fig. 27 for explanation of the figure. 



72 



BOOBY ISLET 
0.84 ha 0.12 
(14%) 



o.27 m 

(32%) \ 




0.41 
(49%) 




BOSUN BIRD ISLET 
0.86 ha 




0.25 

(30%) 



^0.20 

(23%) 




NORTH ARUNDEL ISLET 
0.91 ha 

0.33 

(36%) 



0.27 
(30%) 





10 

(100%) 



Fig. 29a. Plant communities and amount of indigenous vegetation on motus 0.8 to 25.0 ha, Caroline Atoll. See Fig. 27 for explanation of the figure. 



73 



MOTU MOUAKENA 
1.00 ha 

0.26 

(26%) 




0.38 
(38%) 



0.36 
(36%) 




MOTU EITEI 
1.41 ha 

0.38 
(27%) 




0.38 

(27%) 



0.57 
(40%) 




CORAL ISLET 
1.07 ha 




0.13(8%) 0.07 
(4%) 



MOTU MATAWA 
1.71 ha 



0.68 
(40%) 



0.78 

(46%) 




0.52 
(30%) 



0.07 (4%) 



(26%) 





Fig. 29b. Plain communities and amount of indigenous vegetation on motus 0.8 to 25.0 ha, Caroline Atoll Sec Fig. 27 for explanation of the figur 



74 



NORTH BROTHERS ISLET 
1.71 ha 



0.43 

(25%) 




0.48 
(28%) 



0.68 
(40%) 




MOTU KIMOA 

1.80 ha 

0.59 

(33%) 



0.50 
(28%) 





12 
(100%) 



LONE PALM ISLET 
1.99 ha 



0.95 

(48%) 



0.39 
(19%) 




0.61 
W (31%) 




MOTU ANA-ANA 
2.16 ha 



0.93 

(43%) 




47 
(22%) 



(7%) 




0.21 (10%) 



Fig. 29c. Plant communities and amount of indigenous vegetation on motus 0.8 to 25.0 ha, Caroline Atoll. See Fig. 27 for explanation of the figure. 

75 




0.69 
(28%) 



1 (7%) 




BLACKFIN ISLET 

2.62 ha 

0.78 
(30%) 



0.63 




0.80 
(30%) 



1 (11%) 




DANGER ISLET 
2.71 ha 



0.77 
(28%) 





10 
(100%) 



CRESCENT ISLET 
3.10 ha 



1.56 
(50%) 



0.61 
(20%) 





10 
(100%) 



Fig. 24d. Plant communities and amount of indigenous vegetation on tnotus 0.8 to 25.0 ha, Caroline Atoll. See Fig. 27 lor explanation of the figure. 

76 



MOTU RAURAU 1.71 

3.48 ha /""" ~~"~~\ (49%) 

(31%)%iixll^v/ 

0.04 0.66 

(1%) (28%) 



(10%) 




BIRD ISLET 
4.05 ha 




0.22 (6%) 

0.41 (10%) 
0.01 (0.03%) 




BROTHERS ISLET 
4.31 ha 



0.58 

(13%) 




0.37 

(8%) 



0.01 
(0.2%) 




NORTH PIG ISLET 


0.97 


5.44 ha 


^ 1 8%) 


1.31 / 




(24%)/ 




l_^-rC\\\v 


Ipttjll 
x\ 1 1 ?? 


**X*X"XvX 


•X'XvffiasS©' \£-*t /o) 




1.84 
(34%) 

Fig. 29e. Plant communities and amount of indigenous vegetation on motus 0.8 to 25.0 ha. Caroline Atoll. See Fig. 27 for explanation of the figure. 

77 



PANDANUS ISLET 
7.20 ha 



2.30 
(33%) 




2.28 
(32%) 




1.61 
PIG ISLET (23%) 

7.25 ha 

0.90 

k(12%) 

'" : lpf 
?*.'■&/ 1 .35 

(46%) ^-j-ijg*: 

0.03 (0.4%) 





ARUNDEL ISLET 
7.34 ha 



4.36 
(59%) 





SHARK ISLET 
7.98 ha 

2.92 

(37%) 





94(12%) 

0.12(1%) 
2.60(33%) " 

Fig. 291". Plant communities and amount of indigenous vegetation on motus 0.8 to 25.0 ha, Caroline Atoll. See Fig. 27 for explanation of the figure. 

78 



EMERALD ISLE 
8.34 ha 



3.20 
(38%) 



1.5(18%) 




1 (8%) 




TRIDACNA ISLET 
9.08 ha 



7.97 
(88%) 




0.18(2%) 
10.21 (2%) 

f 0.72 (8%) 




11 
(100%) 



WINDWARD ISLET 
1 1 .42 ha 
5.7 
(50%) 




2.97 
(26%) 



1.6 
(14%) 




11 
(100%) 



MOTU MANNIKIBA 
21.49 ha 

15.69 
(73%) 



1 (8%) 




2.60 

(12%) 

f 2.04(10%) 

0.03(0.1%) 
1.13(5%) 




Fig. 29g. Plant communities and amount of indigenous vegetation on motus 0.8 to 25.0 ha, Caroline Atoll. See Fig. 27 for explanation of the figure. 

79 



LONG ISLAND 
75.90 ha 



32.2 

(42%) 



18.6(25%) 





SOUTH ISLAND 
104.60 ha 



80.8 ' 
(77%) 




4.2 (4%) 
1.1 (1%) 

13.6(13%) 
gllf 4.9(5%) 



1 (5%) 




NAKE ISLAND 




91.71 ha 


^^^ 18.3 


30.65 / 


J'.W(20°o) 


(33%) / 




KjIvXvX 


||W6.8(7°o) 


\jx : x : x 


;x3^ 3.38 (4%) 


20.8^ 


£3=^6.03 (7%) 


(23%) 


5.75 




(6%) 



Fig Kl. Plant communities and amount of indigenous vegetation on motus over 25.0 ha. See Fig. 27 for explanation ol the figure. 

80 



_l 2 

a 

O 



=> 20 - 



O 

2 



O 





I 




1 1 


11 1 1 


he 


II 


^ 1 


_ 1 






~\ r^ 


1 1 








- 














■ 
■ 


" 










■ ■ 
■ ■ ■ ■ * 




■ 


" 








■ ■ 


■ ■ ■ • X ■ 

■ ■■ ■ 


■ 




" 








■ 
■ ■ 


■ ■ ■ 






- 






■ 


■ ■ ■ 








* 




■ ■ 












- 










• • 




• 


- 






• 


• • 

• • 

• • 


• 
*• 
• • • • 
* •• • • • 
• • • 


• 




* 




• 

1 




• 
1 


J. 




' 





01 



AREA OF MOTU ( HA ) 



Fig. 3 1 . Total numbers of plant communities ( upper graph ) and species ( lower graph ) in relation to islet area, demonstrating plant succession on the differentsized 
islets encircling Caroline's lagoon. Roman numerals refer to size classes of the motus: I = < 0.2 ha, II = 0.2 to 0.7 ha, III = 0.8 to 25.0 ha, 
IV = > 25.0 ha. The break between II and III marks a substantial increase in the diversity, area coverage, and height of the forest ecosystems. Data is 
based on Table 5. 



22 



20 . 



18 



o 



16 



12 



10 



a 8 

O 

z 

< 6 

O 



50 100 



200 



300 



400 



500 



600 



700 



800 



WIDTH OF MOTU (m) 

Fig. 32. Maximum heights of Pisonia forests in relation to width of the motus. Stars represent forests with 90-100% canopy cover; dots represent forest or scrub 
with less than 90% cover. 



81 



20 . 



16 . 



I 12 - 

O 



I 10 



< 
u 



2 . 



AREA OF MOTU (HA) 

Fig. 33. Maximum canopy heights of Pisonia forests in relation to islet area. Stars represent forests with 90-100<7r canopy cover; dots represent forest or scrub 
with less than 909c cover. 



LAGOON 



HERB 
l_MAT I 

I I TOURNEFORT 



IA SCRUB 



llo3m) 



|TOURNEFORTIA 
SCRUB 
Itournefortia FORESTI |, o5ml 

to9m 



Itou 
est! 



HERB 

MAT 

I 



WIND 



OCEAN 
REEF FLATS 






^ i is^^^^i jfisg l a^^Saia M feaga gals gfesge SassS 



^ 



fJjTJj COBAL RUBBLE 
COARSE SAND 



-* y* S RED FOOTED 



GREAT FRlGATEBIRO 



WHITE TERN 



REEF HERON 



% 



LACK NODDY 



BROWN NQDOr 



BRISTLE -TMIGHED CURLEW 



POL TNE SIAN HAT 



i i i S. hematic profile through Arundel Kiel, recovering from disturbance over 60 years ago, showing natural herb mats, Tournefortia scrub and forest, and 
five species of breeding seabirds. Vertical height is exaggerated, 



WIND 



TOURNEFORTIA 

forestt — 



PISONIA FORESTliolSml 



Itournefortia HERB 

"-| SCRUB | MAT | 



OCEAN 
REEF FLATS 




OL VNESIAN RAT 



Fig. 35. Schematic profile through Long Island. Tr. O. Although Long Island has been formed in the recent past by a merger of five smaller islets, this section 
of the islet is very mature, containing natural herb mats, Toumefortia scrub and forest, and tall Pisonia forest. Seven species of seabirds breed. Vertical 
height is exaggerated. 



oc 

REE 



iCEAN |_HERb[tOURNEFORTIA 
F FLATS I MAT! FOREST 



WIND 



SOUTH ISLAND. 
TRANSECT 2 



DYING COCO S-IPOMOEA FOREST 
(to 18m) 



BEACH SCRUB 
COCOS |WITH SURJANAItoSm 

— Lag6on 



plantation! f 

( to 21m) 




I n j — l^s Sg&k .»■">& 



J — i 



E gJsSfl CORAL RUBBLE 
FINE SAND 
HUMUS. GUANO. RUBBLE 



BROWN " ■;■!■" 



H „ 






SHOREBIRDS 



REEF HERON 



1 BRISTLE -THlGHED CURLEW ^^S LON 



! «t^^H POLVN 



ESIAN R«T 



COCONUT CRAB 



G ■ TAILED CUCKOO 



Fig. 36. Schematic profile through South Island, where 77% of the land surface is covered with Cocos forests, primarily in a dying state. Vertical height is 



exaggerated. 



83 



Pg^j RUBBLE 
fm HERB MAT 
i 1 COROIA 

H TOURNEFORTtA 
W3 PISONIA 

I 1 COCOS 

|..»| PANDANUS 
IV | MIXED 

BEACHROCK 



i 




Fig. 37. Nake Island: vegetation and physiography. 



PLANT COMMUNITY 



PLAHT SPEC 


ES 






Bo«i*i»via ••p«»* 




C.co. » ur ,..-. 




Coiltil luttnidlll 






.■v* 


i»nw>n«» nt«e'««l« 


- 



'.:,. »r.. *,»,, 



SUBSTRATES 

ma— udo • •-*• - -.1 

M u m ul )„)■« rubblf 



g^3 RUBBLE 

j^P HERB MAT 

f "| TOURNEFORTIA 

m PISONIA 

[^j COROIA 

F*Tj COCOS 

^s^— BEACHROCK 





LONG ISLAND 



Fig. 38. Long Island: vegetation and physiography. 



AHCIBM1 r.H» NN( i 


o 


Oifi CHANMII 


1 


«, , „, ■ -, BW ,, • 


* 


II»r«tt 0< miMI b U 


1. 1 * 




01 ■ I AMC I l»| 



I ig (9. Long Island: north-south transect showing division into former islets, floristic composition, relative abundance of plant species, degreeofspeciesoverlap, 
and canopy heiehts. Vertical heighl is exaggerated. The exact locations oi the formerly more extensive Cocos plantations are unknown 

84 



Tournefortia- Piaonla 



Tournaforlla rubbli 



Boerhavla repens 

Cocos nuclf era 

Cordia subcordata 

H eliot ropiuni anomalum 

Lepturus repens 

Morlnda citrifolla 

Phymatosorus scolopendrli 

Pisonla grandis 
Portulaca lutea 
Tournefortia argantea 



I 20 



O 10 

z 
< 
o 

I 

LAGOON 

[laal 




100 200 

TRANSECT WIDTH Iml 



OCEAN 

I windward I 



Fig. 40. Long Island: east-west cross-section through Tr. C, a former inter-islet channel, showing floristic composition, relative abundance of plant species, degree 
of species overlap, and canopy heights. Vertical height is exaggerated. 



Opan Tournafortla Scrub 



Flats sand.rubble rubble 



Boerhavla repens 
Cocos nucifera 
Hehotropium anomalum 
Laportea ruderalis 
Lepturus repens 
Phymatosorus scolopendria 
Portulaca lutea 
Suriana maritima 
Tournefortia argentea 



I 

a 



oio . 



LAGOON 

laa ■ 



100 200 

TRANSECT WIDTH |m| 





1 


_ 






1 






1 


— 


1 


1 E= 


1 

I 

1 

1 

1 



300 



OCEAN 
Iwlndward I 



Fig. 41. Long Island: east-west cross-section through Tr. 8. which passes through mature interior Pisonia forest of largest of Long's coalesced motus. Data 
includes floristic composition, relative abundance of plant species, degree of species overlap, and canopy heights. Vertical height is exaggerated. Note 
the absence of low vegetation on the leeward shore. 

85 




Fig. 42. Vegetation and physiography of Windward Islet no. 1: Bo'sun Bird Islet. Scale is larger than on the vegetation maps of other islets. 



RUBBLE 
HERB MAT 



FTT^ RUBBLE 

^M HERB MAT 
fl CORD. A 

"1 TOURNEFORTIA 
Hi "'SONIA 

~1 TOURNEFORTIA -CORDIA 



WINDWARD ISLET 



CRESCENT ISLET 



MOTU ATIBU 




'J ) 



L 



l ig l ! Vegetation and physiography of Windward Islets nos. 2, ' and 4: 
Windward and Crescent Islets, and Mom Atihu ("Coral Rubble 
Kiel"). Atibu appears to have been severel) damaged during the 
February 1990 storm. 



\ CH con °" 

Y | | TOURNEFORTIA 

P^l PISONIA 

| . | COCOS 

I • I SURIANA 



NORTH PIG ISLET 



SKULL ISLET 



NORTH BROTHERS 
ISLET 



BROTHERS ISLET 





Fig 44. Vegetation and physiography of Windward Islets nos. 5 through 9: 
North Pig, Pig. Skull. North Brothers, and Brothers Islets, Note the 
reels extending westward into the lagoon. 



86 



Tournetortia Pisonia Tournelortla Herb Mat 



Achyranthes canescens 
Boerhavia repens 
Heliotropium anomalum 

Ipomoea macrantha 
Morinda citrifolia 

Pisonia grandis 

Phymatosorus scolopendria 

Portulaca lutea 
Tournetortia argentea 



x 
a 



z 
< 
u 



30 _ 



20 - 



10 - 




LAGOON 
Heel 



100 

TRANSECT WIDTH Iml 



200 

OCEAN 
I windward 



Fig. 45. Pig Islet: east- west cross-section through center of islet. Dataincludes floristic composition, relative abundance of plant species, degree of species overlap 
and canopy heights. Vertical height is exaggerated. Pig's profile is especially symmetrical. It is remarkable that this islet was totally felled for coconuts 
in 1920. 



Boerhavia repens 
Cocos nuci tera 
Heliotropium anomalum 
Laportea ruderalis 
Lepturus repens 
Morinda citrifolia 
Pisonia grandis 
Portulaca lutea 
Sun ana maritima 
Tournetortia argentea 




LAGOON 
Heel 



100 

TRANSECT WIDTH Iml 



200 

OCEAN 

I windward I 



Fig. 46. Brothers Islet: east-west cross-section through center of islet. Data includes floristic composition, relative abundance of plant species, degree of species 
overlap and canopy heights. Vertical height is exaggerated. Note the central monotypic stand of Pisonia forest. This islet's forests were totally felled 
in 1920. o 7 



f£23 rubble 

f^^ HERB MAT 

| ) TOURNEFORTIA 

|^9 P'SONIA 

n~l cocos 



Acropora - Tndacna 




H«*f ^/ 



gy^] RUBBLE 

^| HERB MAT 

| ) TOURNEFORTIA MORINOA 

|» «) SURIANA 



Acropora - Tndacna 
Raaft 



<V 




TRIOACNA ISLET 

I I I U 



Fig. 47. Vegetation and physiography of Windward Islets nos. 10 through 12: Fig. 48. Vegetation and physiography of Windward Islet no. 13: Tridacna 

Noddy Rock. North Arundel, and Arundel. See text for explanation of Islet. The best quality Acropora-Tridacna reels extend clear across 

the relatively small amount of Pisonia cover (Description and Ecology the lagoon from this motu. See Desc. and Ecol. of the Motus Section 

of the Motus Section). for explanation of unusual forest cover. 



Tour net or t la 



Tournefortia 

Morinda Tournefortia Herb mat 



Boerhavia repens 
Heliotropium anomalum 
Laportea ruderalis 
Lepturus repens 
Morinda ci t ri folia 
Phymatosor us scolopendria 
Pisonia grandls 
Portulaca lutea 
Tournefortia argentea 



O 

x 



10 - 



o 
z 
< 
o 



LAGOON 



100 

TRANSECT WIDTH Iml 



1 1 


1 














1 


1= 














1 
1 

1 


■ ' 



200 

OCEAN 
I wl ndward I 



1 i: l'i 1 ridacna Islet: cist west cross-section through lower center of motu. Data includes Holistic composition, relative abundance of plain species, degree 
ol species overlap and canopy heights. Vertical height is exaggerated Note the absence of well-developed interior forests, unusual for a motu of this 
si/e (Description and Kcologv of the Motus Section) 

88 



g£3 RUBBLE 

^B KERB MAT 

| | TOURNEFORTIA 

\TT^ COCOS 

P 7J DYING COCOS IPOMOEA 

| * «J SURIANA 

1"»^ PANDANUS 

~Z^~ BEACHROCK 




Fig. 50. South Island: vegetation and physiography. Note the accepted landing route across its leeward reef flats. 



E • 

OuUr B«iCh Scrub X * 

r 



Boerhavirt repens 
Cocos nucltera 
He 1 1 ol ' opium anomalum 
Ipomoea macrantha 
Lapor tea ruder alls 

Lepturus repens 
Monnda citrtfolia 

Phymalosorus scolopendria 

Pisoma grandls 

portulaca lutea 

Sur lana man t ima 

Tournefortia argentea 




LAGOON 
( lee ) 



240 360 

TRANSECT WIDTH (m) 



720 

OCEAN 
(pari windiwcrd ) 



Fig. 5 1 . South Island: distribution and abundance of plant species along Tr. 2, which runs at an angle of 60° from the lagoon to the south shore through the western 
center of the islet. Data includes floristic composition, relative abundance of plant species, degree of species overlap and canopy heights. Vertical height 
is exaggerated. Horizontal scale is half that of the profiles from smaller motus. 

89 



fTTj BUBBLE 

|^ HERB MAT 

"2 TOUHNEFOHT 

J COHOIA 

[*"3 PISONIA 

|»1^ PANDANUS 

f~r*1 cocos 

PANDANUS ISLET 




t£ J . : 'I RUBBLE 

|B MERB MAT 

I 1 TOURNEFORTIA 

rT*^ PISONIA 

!• • -1 cocos 

^ BEACMROCK 




Fig. 52. Vegetation and physiography of the 7 South Nake Islets: Pandanus, 

Danger, Booby, Coral, and Lone Palm Islets, MotuKota( "Red-footed Fig. 53. Vegetation and physiography of Central Leeward Islet no. 1: Motu 



Booby Islet"), and Motu Mouakena ("Masked Booby Islet"). 



Mannikiba ("Seabird Islet"). 



)"3 RUBBLE 

| HERB MAT 

[ ) TOURNEFORTIA 

1 .J PISONIA 

^■■^ PANDANUS TOURNEFORTIA 

\: :'■: I coroia 

|" • -1 cocos 





C3 RUBBLE 

■ HERB MAT 

| 1 TOURNEFORTIA 

003 CORDIA 

[■ • «1 COCOS 

!■»■! PANDANUS 

[~3 PISONIA 

r~7~| SURIANA 




• Urt 400 



i i I Vegetation and physiography ol the Central 1 eeward Islets nos 2 Fig. 55 Vegetation and physiography of the Central Leeward Islets nos. 5 

through 4: Blackfm Met. Mom Matawa ("Fair} Tern Met"), and through 11: Shark and Scarlet Crab Islets, Motu Nautonga ("Sea 

Emerald Isle Cucumber Islet"), Azure Isle, Reef-flat, Bud and Fishball Islets 

90 



Achyranthas canascana 

Bonhivia rep«ns 
He I iot r opium anomalum 
L aportea rudaralis 
Lepturus repans 
Morinda citrif olia 
Portulaca lutaa 
Tournafortia argantaa 





Hi 













■ 

_ i 


E 

h- 

a 

X 

ho 

z 
< 
a 


■ 
■ 

J i — — —i— A 1 





OCEAN 



50 



100 



TRANSECT WIDTH ( m ) 



LAGOON 
Iwtftdward I 



Fig. 56. Fishball Islet (no. 1 1. Centra] Leewards): east-west cross-section through the center of this young motu, which exhibits early stages of geological and 
biological evolution. Data includes floristic composition, relative abundance of plant species, degree of species merlap and canopy heights. Vertical 
height is exaggerated. 



f-j,;^ BUBBLE 

Q^| HERB MAT 

~] TOURNEFORTIA 

r~~i cordia 

"2 PISONIA 

[. . .| cocos 

|~j~| HOUSE SITE 

[ ■ | PANDANUS 

, | • | SURIANA 



MOTU RAURAU 



MOTU E1TEI 



PISONIA ISLET 



MOTU KIMOA 



MOTU ANA ANA 



N 

i 





SOUTHERN LEEWARD ISLETS 



Fig. 57. Vegetation and physiography of the 5 Southern Leeward Islets: Motus Raurau ("Blue-gray Noddy Islet"). Eitei ( "Frigaiebird Islet"). Pisonia Islet. Kimoa 
("Rat Islet"), and Ana-Ana ("Anne's Islet"). 




_ lb 

PI. 1 . A dawn view of Caroline as seen from the ocean in 1 988 but virtually identical with that seen by the atoll's Western "discoverer.' 
de Quiros, in 1606. 








PI. 2a. A clearing on South Island from which the Solar Eclipse Party made their observations in 1 KX3. Today the area is covered with 
dense Cocos forest (from Holden & Qualtrough, 1884). 




93 




PI. 2b. One of the three European-style houses that have ever been built on Caroline, drawn in 1883 (ibid.). 







■■■■ " 







PI. 3a. An artists very lice rendering of Caroline In 1883 (from Holden & Qualtrouith. 18841. 



94 



I.IIUIIl' MHI.l .'I llll SOI'TBRR* Wll <>» < AKfU.IM. iKLisrm 




a 






..<•<" 



V 









7^. 



,% *: 



j «* ** 









* * 



tif ;.■■■!.. r *tt'i. 




PI. 3b. Map of the "settlement" on South Island, as drawn by the Solar 
Eclipse Party (ibid.). 



■B3* 




. 4. Two lagoon views a century ago along the north coast of South 
Island (from Holden & Qualtrough. 1884). Compare these 
drawings with Plate 24. 



95 




PI. 5a.b. Motu Ana-Ana, unnamed in 1883 (Holden & Qualirough, 
1884) but appearing virtually identical then and today 
( Plate 8 1 ). Below is a substantial Tournefortia tree along South 
Island's lagoon edge. 




PI. 6. Base camp, northwest peninsula. South Island. 




46 



* 






PI. 7. The junior author beside an entangling thicket of Ipomoea 
macrantha, dying Cocos-Ipomoeaforest, interior South Island. 







97 




PI. 9. Relocating camps with inflatable canoe. Cameron Kepler and Katino Teeb'aki, Bo'sun Bird Islet. 




PI. 10. BI.Kk tippped reel shark [Carcharhinus melanopterus), a numerous and aggressive denizen "l Caroline's lagoon. Note the 
abundant set cucumbers i / udwigolhuria sp. ). 



98 




PI. 11. Jagged, upraised reef ifeo or champignon), leeward reef, southwest Nake Island. Note the wide reef Hats. 




PI. 12. Floating a small boat across the soutwest reef Hats in calm weather from the "boat entrance" to the "landi 
the wide reef flats. The Akademik Korolev drifts offshore. 



' on South Island. Note 



99 



4fefe 







PI, 1 3. Beach crest, sand) rubble, seaward moat, and narrow reef flats off northeast Long Island. 




PI. 14 I Ktensive lagoon reef flats south of Arundel Island on the windward side. Note the Five southern leeward islets in the distance. 



100 





PI. 15. An incipient motu. barely connected to Motu Mouakena's southern shore. See also Fig. 69. 




PI. 16. Windward reel and peaked beach crest on South Island, with the recently wrecked remains of a 26-foot sloop. 




I 



;.-^> 










PI. 17. S 



Successive ridges of coral rubble forming extensive gravel flats, northeast Nake Island. 






■ 






PI. 18. Channel between the two northern islets. Long and Nake. Note the mixed forest with Coi os and Pandanus. 



102 




PI. 19. Noddy Rock (0.02 ha), an emergent reef platform along the windward reef flats. A northward view. 






PI. 20. Windward beach. Long Island, showing wide rubble flats inland of the beach crest, rimmed by oceanic flotsam and jetsam. 



103 




Beachrock, wesi shore. South Island 
blinding coral beach. 



Russian vessel Akademik Korolev drifts offshore 




PI. 22. A large coconut crab (Birgus (afro) shelters in a subterranean cavit) in the ft 



104 




.. 




PI 23. Sand> Inlet, a filled-in portion of the lagoon, extends its fishhook-shaped mudflat 300 m northward into Nake's landmass. Here 
grow the healthiest and most productive Cocos on Caroline. Note the bristle-thighed curlews in the foreground. 




PI. 24. South Island's pure ( '.>< OS plantation, looking west along the lagoon. This extensive grove has now obliterated all traces of the 
former "settlement" (Plates 2-5). 



105 




PI. 25. Brilliantly-colored, crystalline lagoon waters adjacent to Emerald Isle (Central Leeward Islets) are studded with twisted reef 
configurations and sandy channels. 




-.« • -. 



- 






■ 














PI. 26. An impressive cross lagoon reel ol Acropora sp. coral and Tridacna maxima clam shells joins Tridacna Islet with Motu Kimoa. 




106 




PI. 27. Cordia forest (to 1 2.6 m tall ), Pig Islet. 



PI. 28. Sand, silt, rubble, and hardpan mingle on the upper 
reaches of Long Island adjacent to the lagoon. 








PI. 29. Caroline's best sandy beach lines the lagoon shore of Shark Islet. The fine sand is overlain with sparkling pink granuli 





J29[tee^4fe^i 




I 'I JO Sheltered baj Brothers Islel Raurau Islet lies across the lagoon. Note the sparse herb mat and silt] shallow waters 



108 







PI. 31. Narrow lagoon beach lined with Tournefortia scrub. Blackfin Islet (Central Leewards). 






PI. 32. Recent sand additions to South Island's northeast point, which is in part covered with excellent natural herb mats and healthy 
Suriana scrub (right). 



109 




PI. 33. An old interislet channel (Transect C, Long Island), now filling in with herbs, Toumefortia scrub, and Cocas. Note the nesting 
masked boobies in middle right. 



1 





&*' 



^ 




T$ 



PI. 34. A large clearing within dying ( o< 05 /pom 
Phymatosorus and Ipomoea 







st, interior South Island. Note the 



» v vi «%Bk... 



prolific mats of Boerhavia interlaced wuh 



110 



PI. 35. Pandanus forest, south Nake. 



if A * 'j- 





PI. 37. Mixed fores! \uth Cocos, southwest Nake Island. 






7 .-» - 




i'l *s Orange, scarlet, and green phalanges of Paw/anus rest on a clump ol Portulat a. The ubiquitous Coenobitaperlatus forage on their 
siriui;\ flesh. 




PI. 39. Inner edge of lagoon. South Island, 1988. Cocos is progressively shading out the beach scrub with Suriana maritima. 




PI. 40. Inner edge of lagoon. South Island, 1965, taken from approximately the same location as Plate 39. Note the greater extent of sand 
and Suriana coverage above high water than today, due to less encroachment and shading by the palms. 



113 




PI. 41. Heavy understory of Achyranthes canescens, Boerhavia repens, and Phymatosorus scolopendria in a clearing adjacent to Pisonia 
forest. Pic Islet. 




I'l 42. Boerhavia Fruits on feathers and hill of a groat frigatebird. 



114 



PI. 43. Inside a mature Pisonia grandis forest, interior Nake Island. Note 
the barren, dark aspect, virtually devoid of undergrowth except 
root suckers. 




*7 PI. 44. Fringe of Suriana, northeast point. South Island. 



L^MflA* 



_ 




PI. 45. Well developed natural herb mat. primarily Heliotropium anomalum and Lepturus repens. Seattered Tournefortia form 
"savannah." Here the sandy soils are conducive to the growth of lush Heliotropium, northeast point. South Island. 




PI. 4ft. Detail ol Heliotropium anomalum, with remains of the first evidence of tropicbirds on Caroline. Skull Islet. 



116 






V 






! 



PI. 47. Toumefortia scrub, fringed by a natural herb mat, and occupied 
by a colony of sooty terns. An old interislet channel, northern 
Long Island. Note the nesting red-footed boobies. 




PI. 48. Toumefortia- Morinda forest, with nestingbrown noddies, interior 
Tridacna Islet. 



117 



PI. 44. Skull Islet (0.02 ha), with brown nodd) tems, looking east to the 

windward reef. 




PI 50. Anartistic impression of&Paihldiiin prove. South Island, in 1 SS3 
(from Holder & Qualtrough, ISS4). 



118 




PI. 51. Red-footed booby in Tournefortia scrub. Motu Raurau. 




PI. 52. Mature Pisonia grandis canopy with incubating black noddies and a white tern. Pig Islet. With a canopy height of 2 1 m. this was 
the most majestic interior forest on the atoll, although it cannot be older than 65 years. 



19 




Azure Isle (Central Leewards) — an example of a motu containing a single Pisonia tree. Note the narro 
interislet channel. View east from Motu Nautonga. with Brothers Islet in the distance. 



hut Mill shark-patrolled. 




PI. 54. Caroline's sole clearing, with Tahitian-style huts. Motu Ana-Ana. 



120 




PI. 55. While tern "nest" balances precariously on the upper midrib of a coconut frond. 




PI. 56. Average-sized coconut crab iBirgus lalro) on Caroline. Compare its size with the coconut. 



121 







PI. 57. Piles of fibrous shavings — coconut crab sign. 




I 'I ss Beachrock al the lower end of Long Island, typicallj found al low water. 



122 




"• A 




PI. 59. Sooty terns (Colony A) — approximate^ 128,000 pairs occupy a scrubby swath of vegetation in the north-central portion of Long 
Island, the location of an old interislet channel. Note Cocos on the right and a mound of Pisonia at rear. 







PI. 60. Windward Islets numbers 5-9 (left to right): North Pig. Pig. Skull (not visible). North Brothers, and Brothers. 



123 




PI. 61. North Pig (left) and Pig (right) Islets: a southeaster!} mow across the coral-studded lagoon from Shark Islet. Note the exp 
central Pisonia forests. 




PI, 62. Arundel Islet (foreground), looking south southwest across IYidacna Islet toSouth Island. Distant Mom Ana- Ana lies on the right. 



124 







■■ M matter ■=^i»' ■ * *>■* r- nls^%. « . r * TV 

PI. 63. Detail, Tridacna maxima reefs, lagoonside of Tridacna Islet. This dense aggregation of giant clams amassed up to 80 per square 
meter. 




PI. 64. View of Pandanus Islet (center) west down the channel separating Nake (right) and Long (left) Islands. 



125 




. 65. Danger Islet (South Nake number 2), looking due west across the shallow upper lagoon from Long Island. 




PI. 66. South Nake Islets numbers 3 6 1 right to left): Booby. Coral. Lone Palm, and Kota. Westerly view across the shallow upper lagoon 
from Long Island 



126 




PI. 67. Lone Palm Islet (South Nake number 5): a southerly view from the shallow tidal flats of Coral Islet. Plate 68 provides a more 
detailed view of the lower part of the chain. 




PI. 68. South Nake Islets numbers 2-6 (right to left): Danger. Booby, Coral, Lone Palm, and Kota, from a boat. This view shows better 
detail of the lower part of the chain than Plate 67. 



127 




PI. 69. Motu Mouakena (South Nake number 7), with its sandy cay. A westerly view from the upper lagoon. Compare with Plate 15. 




PI. 70. North end, Motu Mannikiba "Seabird Islet" (Central I ee wards number I ), showing mounds ofPisonia and a closer Cocos grove 
Note coral "mushroom" in laj n 



12S 



j**4MMfl&- V* 




PI. 71. 



of Pisonia. 



, looking west along Transect 



Tournefortia, then a distant patch 



r ?>*''• 




■ 







y-VV^ 





s£ 



s:ip; : x£-v 







PI. 72. Motu Mannikiba, looking west along Transect 2. Low Tournefortia scrub covers a coarse, rubbly substrate which was probably 
once an inter-slet channel. 



129 




PI. 73. Blackfin Islet (Central Leewards number 2) lies to the left of the ship. View west-southwest from the tip of Long Island. 




PI. 74. A stunning "blue hole" within the lagoon ofl the southwest point ol Emerald Isle (Central Leewards number 4) 



130 



PI. 75. Mixed Pandanus-Tournefortia fores!, interior Emerald Isle. 




PI. 76. Emerald Isle, looking west across hardpan and open lagoonside 
scrub to a densely vegetated interior. 






•>i.' 





131 




PI. 77. Shark Islet (Central Leewards number 5): view across crystalline shallow waters to Caroline's sole sandy beach. 





PI. 78. 



Southern I eeward Islets numbers I 5 (right to left): motus Raurau .mil Eitei, 
northwest from 1 ridacna Islet. 



Pisonia Islet, motus kimoa and Ana Ana View 



32 




PI. 79. View of Motu Raurau ("Blue-gray Noddy Islet") fromMotu Eitei ("Frigatebird Islel" >. Note the two nutrient-starved, chlorotic 
Cordia trees growing in almost pure rubble (center left). 





PI. 80. 



View of Motu KimoaC'Ra 
and Cordia. 



Its central forest, typical of the Southern I.< 



133 




Motu Ana-Ana ("Anne's Islet"): a view with giant ray, from the shallows of the lagoon's southern end adjacent to South Island. 
Note the similarity to Plate 5, dating from 1883. 



L34 



Appendix 1 

Reef Information for Navigators 



We include this section because no accurate hydrological 
chart exists, and the Pacific Islands Pilot ( Hy drographer of the 
Navy, 1982) section for Caroline is incomplete. Arundel's 
1883 map (Admiralty Chart No. 979, Fig. 4) is still used today. 

Caroline has neither a deep pass, nor navigable channels 
into the lagoon, nor a ship anchorage beyond the reef. In 1 873, 
a set of moorings was placed off the west coast of South Island 
for the convenience of guano ships, approximately "a mile 
north of the south-west point, in about 60 fathoms of water and 
some distance from the shore" (Maude, ca. 1942). These are 
long gone, although small boats can still anchor within the 
close lee of South Island during normal trade winds. Today's 
ships, however, must drift well offshore after approaching the 
atoll from the west (PI. 12). 

Of special note is a possible extension of the perimeter reef 
south and southwest of Caroline. Arundel's map notes: "Reef 
reported to extend four cables from southeast point." This 
information probably originated in Findlay's South Pacific 
Directory of 1884. quoted by Holden (1884). Evidently the 
windward reef of South Island extends approximately 1.7 km 
from its southeast point. From here "this reef sends out two 
branches to a distance of 2.5 km. one toward the southeast, the 
other toward the southwest and is consequently dangerous to 
approach at night." Arundel's map does not include this 
bifurcation which, according to Findlay ( 1 884 ). extends at least 
across the width of South Island. He also states that "a landing 
(not always safe) may be effected on the north side of the 
southwest bifurcation, described above." No trace of these 
submerged reefs is evident on the RNZAF aerial photos. 



The "boat entrance" (Fig. 4). a narrow nick in the outer 
leeward reef, marked by the stock and ring of an anchor 
and immediately to the west of South Island's northwest point, 
is not necessarily the easiest route to the lagoon. Landing 
is possible across the steep-to reef at many locations along 
the leeward reef; opposite the southern end of Ana-Ana is 
good. 

Landing is fairly straightforward by the anchor when the 
seas are calm, especially when one becomes familiar with the 
crooked notch that narrowly pierces the outer reef. After 
negotiating a powerful backwash, one's boat is swept onto the 
shallow reef flats — liberally laced with chunks of jagged 
reef — which is exposed at low tide and barely covered at high 
tide. A swift current passes west out of the lagoon between 
South Island and Ana- Ana, sweeping over the reef at the notch. 
Only small craft with virtually no draft can effect the 500-m 
journey to South Island. Because the shallows are unchanneled 
and not navigable even at high tide, one's small boat must be 
carefully hauled through the water to a sheltered landing spot 
adjacent to South Island's northwest point (PI. 12). 

An alternative landing method used by yachts in calm 
weather is via the "blind passage" (Structure and Topography 
section. Fig. 50). adjacent to the northeast corner of South 
Island. Despite the fact that the inner one-third of this narrow 
diverticulum is calm ( and used for the residents' yacht mooring), 
the outer two-thirds are rough and dangerous most of the time. 
Its channel leading to and from the open sea is particularly 
turbulent and should not be attempted without assistance from 
the residents, and only at first light. 



135 



Appendix 2 

Weather Data. Caroline Atoll. 1989- 1 WO 

A. Wind Direction and Speed (mph), 1989 



Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 

T~ NNE 15 NW 18 E25 E12 E 20 E 15 E 25 E 20 NE 10 N 20 NE 8 

2 NE 10 W5 E 15 E 16 E 20 NE 18 SE 20 E 12 NE 12 N 15-17 NE 10 

Variable 

3 NE 10 NE20 E 12 SE 14 S 15-20 E 18 E 20 SE 25 NE 15 NE 15 NW 20 

4 NE 15 NE 15 E 12 SE 15 E 15 E 15 E 20 SE 20 NE 15 NE 12 NW 20 

5 NE20 NE 15 NE 15 E 12 NE 8 E 15 E 18 E 18 NE 12 E 12 big NW 18 

swells 



6 


NE20 


NE20 




SE 10 


NE8 


E 15 


E 10 


E 18 


- 


NE 12 


E 10 


N 10 


7 


NE 12 


SE 18 


E 12 


E 14 


NE 10 
E 10 


NE25 


- 


E 12 


- 


NE 15 


NE 12 


NW 18 


8 


NE 12 


E 15 


calm 


SE 1(1 


E 10 


NE20 


SE 12 
squalls 


NE 18 


- 


NE 15 


calm 


NW 18 


9 


NE 12 


E 18 


SE 10 


E20 


E 15 


NE 18 


SE20 


E 12 


- 


NE 12 


N 10 


NE 10 


10 


NE 15 


E 16 


NW 10 


E20 


E 15 


E 15 


SE 16 


E 12 


- 


NE 10 


NE 12 


NE 10 


11 


NE 10 


NE 16 


calm 


SE 25- 

E 10 
squalls 


E 15 


NE 15- 
E 15 


E 15 


calm 


- 


NE 10 


E 18 


NE5 


12 


NE20 


SE 14 


E 10 


W 14 


E 15 


E 15 


E 12 


E 12 


- 


NE 16 


E 17 


NE 12 


13 


NE15 


E 12 


E 12 


E 16 


NE 16 


NE 16 


E 12 


- 


- 


NE 11 


E 14 


NE 12 


14 


NE 15 


E 10 


NE 10-18 


E20 


NE 16 


NE8 


E 12 


calm 


- 


ENE 10 


E 10 


NE 14 


15 


NE15 


NE8 


E8 


E 18 


E 15 


E 10 


E 10 


- 


- 


NE 10 


calm 


NE 14 


16 


NE 15 


SE8 


E 12 


E 18 


E22 
squalls 


E 10 


NE 10 


- 


- 


NE 1.2 


calm 


NE 15 


17 


NE 18 


E20 


NE 15 


E 18 


E25 


E 15 


E 10 


- 


- 


NE 14 


calm 


NE 12 


18 


NE 12 


E 12 


E8 


E 18 


NE 18 


E 15 big 
swells 


E 15 


- 


- 


NE 15 


N 


NE 14 


19 


NE 12 


E 12 


E 10 


E 18 


NE 14 


E 10 


E 16 


- 


- 


NE 13 


NW 10 


E 12 


20 


NE 18 


E 12 


E 10 


E 18 


NE 14 


E 15 


E 18 


- 


- 


E 15 


NW 15 


E 10 


21 


NE 15 


NE 10 


SI X 


E 14 


E 12 


NE20 


E20 
squalls 


- 


- 


E 16 


N 16 


E 10 big 
swells 


22 


NE20 


E 15 


calm 


E 14 


E 12 


E 10 


E15 


- 


- 


NE 14 


NE 14 


E 12-NE 10 


23 


NE 18 


thundei 

squalls 


N 18 
thunder 


SF. 14 


SE 10 


F 10 


E 15 


- 


NE 20-30 


E20 


NE 16 


calm 


24 


NE 15 


E25 | 
E 15V 


N 12 


SE 14 


SE 10-25 


F 10 


E 12 


- 


E 20-30 


NF 15 


M i: 


NE 15 


25 


NE 12 


N 12 


E 12 


E 18 


SE 17 


E 10 


- 


E20 


NE 15 


NE 12 


NE 16 


26 


NE 12 


E 15 


N 10 


E 10 


\l I 1 


E 10 


calm 


- 


E20 


NE 15 


N 10 


\l I-- 


27 


NE 12 


1 25 


SE 10 


NE 10 


E 18 


E 10 


calm 


- 


NE 17 


E9 


E9 


NF. L5 


28 


NE 12 


E 14 


i: 12 


E 1 i 


E 18 


E 10 


E 15-35 


- 


E 1 5 


NE 10 


F8 


NE 15 


>9 


NE 15 




E 14 


1. 18 big 
swells 


E 15 


E 18 


E 15 


- 


N 10 


NE8 


E7 


NF 12 


JO 


M 12 




1 I ! 


1 I ! 


- 


E 18 


F 15 


- 


calm 


calm 


- 


F 10 


51 


calm 




E 18 








SE 18 


- 




N 20 




E 16 



136 



B. Rainfall, 1989-1990 



Month/Year Mean Monthly Mean Number of 
Rainfall (mm) Rain Days 



Jan "89 


71.1 


7 


•90 


177.8 


22 


Feb '89 


160.0 


14 


•90 


640.1 


10 


Mar '89 


259.1 


20 


'90 


215.9 


14 


Apr '89 


190.5 


16 


'90 


48.3 


6 


May '89 


66.0 


10 


'90 


325.1 


8 


Jun '89 


48.3 


11 


'90 


78.7 


11 


Jul '89 


45.7 


12 


'90 


68.6 


8 


Aug '89 


35.6 


12" 


'90 


109.2 


14 


Sep '89 


50.8 


3" 


'90 


81.3 


6 


Oct '89 


73.7 


11 


'90 


175.3 


9' 


Nov '89 


78.7 


7 


'90 


134.6 


14 


Dec '89 


162.6 


11 


'90 


154.9 


9 


Annual '89 


1.242.1 


134 


'90 


2.209.8 


131 



Source: Ron Falconer, Caroline Atoll (personal communication). 
■"Based on 16 days' data. 
h Based on 9 days' data. 
'Based on 18 days' data. 



137 



1.2 Ecological Studies on Caroline Atoll, 

Republic of Kiribati, South-central Pacific 
Ocean 

Part 2. Seabirds, Other Terrestrial Animals, and Conservation 

CAMERON B. KEPLER . ANGELA K. KEPLER and DAVID H. ELLIS* 

' US Fish & Wildlife Service, Patuxent Wildife Research Center, Southeast Research Station. Athens, Georgia, USA 
I S Fish & Wildlife Service. Patuxent Wildlife Research Center, Laurel, Maryland, USA 



Introduction 

On 26 July 1988. the Soviet research vessel Akademik 
Korolev sailed from Vladivostok enroute to Dutch Harbor, 
Alaska. There, Soviet oceanographers joined their American 
colleagues to invest : gate the Gulf of Alaska and the Chukchi 
Sea in the Third Joint US-USSR Bering & Chukchi Seas 
Expedition. When the arctic research was completed in early 
September, the ship headed toward the central Pacific. A 
rendezvous for a second contingent of Americans took place in 
Hilo. Hawaii, on 9 September. Six Americans joined the ship, 
which set sail on a cruise track of 14.N42 km that terminated 

6 weeks later in Singapore. An important part of the expedition 
was research in and around little-known Caroline Atoll, at the 
southeastern edge of the Line Group. On Christmas Island, we 
picked up Katino Teeb'aki, a conservation officer for the 
Republic of Kiribati, who represented his government and 
helped our land-based research efforts. After landing on 
Caroline on 22 September, we camped in 2 locations for 

7 nights, surveying the terrestrial plants and animals on all 
39 islets. Caroline is a remarkably pristine atoll with its native 
plant communities nearly intact on all but three islets, and 
teeming seabird communities that, collectively, are second in 
the Line Group only to Christmas Island (Kiribati) in diversity. 
For several historical reasons, the natural values of this 
spectacular blend of marine and terrestrial resources have been 
overlooked. 

Approximately 1,000. 000 seabirds of I I species bred mi 
Caroline Atoll in November 1988. The most abundant species, 
with over 900.000 birds in 1988. was the sooty tern (Sterna 
fuscata). Two species (red-tailed tropicbird [Phaethon 
rubricauda], blue-gray noddy [Procelsterna cerulea]) are 
reported breeding for the first time. The known seabird fauna 
now includes one tropicbird. three boobies, two frigatebirds, 
and five terns. 

Seabird distribution on Caroline is determined by the 
distribution of plant communities, rats, coconut crabs (Birgus 
latro). and the prevailing trade winds. Red-tailed tropicbirds 
and ground-nesting brown noddies (Anous stolidus) nested on 
small islets relatively free of rats and coconut crabs, masked 
and brown boobies (Sula leucogaster) preferred exposed 
windward beaches, primarily on Long and Nake. The tree- 
nesting red-footed booby (Sula sula) and the frigatebirds 



attained their highest nest densities in areas with reduced wind 
speed. The black noddy (Anous minutus) was found in dense 
colonies, generally high in Pisonia trees in the center of small 
islets, while the uncommon blue-gray noddy (Procelsterna 
cerulea) nested solitarily on open coral rubble. Sooty terns 
nested in large colonies, generally near or under relatively open 
Tournefortia scrub but also in open areas under Tournefortia 
and closed-canopy P/.sy»i/« forests. Tree-nesting brown noddies 
and white terns (Gygis alba) were found throughout the native 
forests and were the only species that nested in anthropogenic 
forests. Disturbed forests on South and Nake held the lowest 
seabird population densities, and no birds nested on inhabited 
Motu Ana-Ana. 

About 300 bristle-thighed curlews {Numenius tahitiensis). 
a rare shorebird, overwinter on Caroline, foraging in all terrestrial 
habitats, including Pisonia and Cocos-Ipomoea forests. We 
extended the known winter range of the long-tailed cuckoo by 
discovering a small population on the atoll, the first record for 
the Southern Line Islands. 

The known lizard fauna was increased from three to six 
species. Approximately 2.200 coconut crabs inhabited 
12 islets on Caroline. Although primarily associated with 
coconut plantations, we also found them in Pisonia and 
Tournefortia. 

We now know that the populations of seabirds and coconut 
crabs on Caroline Atoll are of national and international 
importance. The black noddy (17,000 birds) and white tern 
(8,000 birds) populations are the largest in the Republic of 
Kiribati, while the red-footed booby population (7.000 birds) 
is the fifth largest in the world. 

History of Ornithological Studies 

"There were a great quantity of seabirds of several kinds, 
and so importunate that they seemed to want to attack the men" 
(Markham. 1904). So wrote the Portuguese explorer deQuiros 
on 21 February 1601, the first European to see Caroline Atoll. 

Precisely what seabirds were present remained a mystery 
until the island was surveyed 364 years later by the Pacific 
Ocean Biological Survey Program (POBSP) (Clapp & Sibley, 
1971a). Prior to this expedition, accounts of the avifauna had 
been incomplete and somewhat confused. Bennett (1840) 
described red-footed boobies, a frigatebird (species '.'). white 



139 



terns, bristle-thighed curlews, tattlers (Heteroscelus incanum 
and H. brevipes), and "a great number of small pigeons" with 
white heads (certainly noddy terns, perhaps both A. minimis 
and A. stolidus). The "shoal birds" that greeted him were 
probably sooty terns. His most unusual contribution was 
mention of a possible flightless rail: "The other birds of the 
coast were a kind resembling a cool..." (p. 372). 

The 1883 Solar Eclipse Party (Subchapter 1.1. History of 
Caroline Atoll section) published a few sketchy notes, adding 
lesser golden-plover (Pluvialis dominicd), reef heron (Egretta 
sacra), and masked booby (Sula dactylatra) ("gannet") to the 
bird list. Of dubious identity were two species of "seagull" and 
a "snipe" (Dixon. 1884). Holden. one of the astronomers, 
heard "the notes of a singing bird." which prompted us to add 
mist nets to our equipment in the hopes of capturing an 
Acrocephalus warbler. This resulted in our discovery of the 
long-tailed cuckoo (Eudynamistaitensis) (Ellis et al., 1990) and 
piqued our curiosity about what Holden might really have 
heard. 

The POBSP expedition spent 3 days on Caroline in June 
1965. They found 10 species of seabirds (9 breeders). 
4 migrant shorebirds, and a reef heron (Clapp& Sibley. 1971a), 
providing rough population estimates for each species. This 
w ork laid the foundation for later expeditions. Brief visits to 
Caroline by the Kiribati government in 1974 and Roger Perry 
in 1977 (Garnett, 1983) added no further information. 

The 1988 expedition to Caroline was longer and more 
extensive than all former visits. We found three new island 
records: a breeding seabird ( red-tailed tropicbird ). a shorebird 
(Sanderling [Crocethiaalba]), and a migratory land bird (long- 
tailed cuckoo), and determined islet-by-islet distributions for 
each species. Our population estimates, calculated from field 
work, aerial photographs, and detailed vegetation analysis. 
indicate that Caroline's avifauna is far more important than had 
previously been suspected (King, 1973; Garnett. 1983). In 
March and May 1 990. the ICBP 1990 Line and Phoenix Islands 
Expedition (Subchapter 1.1. Methods section) filled in minor 
gaps in our knowledge. Caroline's residents added another 
breeding seabird, the blue-gray noddy, in summer 1990. 

Methods 



Distribution and Habitat Preference 

We described se\ en major plant communities on Caroline 
Atoll (Subchapter 1 . 1 i. With the use of aerial photos and the 
transect data, we mapped the communities found on each islet. 
Bird distribution was determined and plotted using these islet 
vegetation maps. If a species nested within a particular plant 
community, it was plotted on the distribution maps as occurring 
throughout that community unless determined otherwise. 

Population Sizes and Breeding Phenology 

We measured transect distances for each islet using a hip- 
chain and biodegradable cotton thread. We recorded all birds 
seen within the 30-m-wide strips: transect width was estimated 
visually. We assigned birds to one of several mutual I v exclusive 
categories: adults present, adults on territory, adults on nests 
(contents unknown ). eggs, naked chicks, dow ny chicks, chicks 
w ith remiges erupting, chicks with scapular feathers, or chicks 
in juvenile plumage. We created a range of possible laying 
dates for each egg and chick using known growth parameters 
for each species (Kepler. 1978; Kepler & Kepler. 1978). This 
enabled us not only to estimate seabird populations, but also to 
determine and plot a rough breeding phenology for each 
species (Figs. 3.5.7.9.10.12). In these figures, the height of the 
bar for each category ("downy," "scapulars." etc.) represents 
the number of nests found or estimated with that development 
stage in September 1988. The bar width represents the 
approximate time span over which eggs could have been laid 
to produce that stage, while the "no. day s" is a count back from 
the survey dates to accommodate growth and development that 
had occurred. Thus, while each figure shows what breeding 
stages we found, we extend those nests back in time to show- 
roughly when they would have begun. The number of clutches 
begun per day is determined by dividing the number of nests 
per stage by the time span in days over which those eggs were 
laid. 

Sooty terns nested in dense colonies. Each colon) was 
mapped, and its total si/e (m') w as calculated. A minimum of 
10 plots (3x3 mor.3 x 6 m). within which all eggs and chicks 
were counted, were randomly located along a compass line in 
each colony. The population size of each colon) w as estimated 
from these plot densities. 



From 22 2 l > September I9S8, C. B. Kepler. A. K. Kepler, 
D. H. Ellis, and K. Teeb'aki surveyed all of Caroline's 39 islets 
except North Arundel Islet, naming most of them (Fig. 1 ; see 
Subchapter 1.1, Methods section). We established 50 linear 
transects, extending 13,300 x 30 m. laid out to ensure that at 
least y< of each islet was sampled lor birds and plants (see 
Subchapter I.I. Methods section and Fig. 8). Sampling was 
increased with 19.300 m of perimeter surveys along the 
windward and leeward coasts of 21 islets (Subchapter 1.1. 
Fig. 9). On Noddv Rock. Skull. Atibu. Bo'sun Bird. Coral. 
Reef-flat, and Fishball (Fig. 1). we made total counts of the 
breeding seabirds. All sur\ cv s w ere conducted during daylight 
hours. Some incidental data have been added from the 1990 
ICBP expedition. 



Mist Nets 

We operated 4 ATX 4-shelf 36-mm mesh mist nets 



(2.6 x 12 m) 



43.5 net hours, according to the following 



schedule: 14.5 net hours (daylight) beneath a 10-15 m Cocos 
canopj on South. 27.5 net hours (day and night i in Pisonia- 
Cocos interface 1 12 m tall) near Tr. 10 on Long, and 1.5 net 
hours m Pisonia-Tournefortia within a 4-6 m canopj onTr.4, 

Long. One cuckoo was collected (USNM 607 19 1 1. 

Collecting Other Vertebrates 

Lizards that were active and conspicuous were collected at 
base camps on South and Long, either bv hand or with a 
blowgun firing steel darts. No attempt was made to search lor 
reptiles under coral, litter, or in other concealed locations 



140 



Rats were collected with a blowgun or snap traps baited with 
coconut, the former proving far more effective because most 
traps were sprung by hermit crabs. We preserved all specimens 
in formalin and sent them to the US National Museum. 

Seabird Species Accounts 

Eleven species of seabirds occur at Caroline, most of 
which breed in large numbers. They include one tropicbird, 
three boobies, two frigatebirds. and five terns. 

Red-tailed Tropicbird (Phaethon rubricauda) (Figs. 2.3; PI. 1 ) 

Red-tailed tropicbirds breed at widely scattered locations 
throughout the tropical Pacific and Indian Oceans. In the Line 
Group, they nest from Palmyra south to Starbuck ( Perry, 1980), 
with a large population (8,500 birds) on Christmas Island 
(Clapp, 1967). Prior to our expedition it was unknown from 
Caroline, Vostok, or Flint. 

Distribution and Habitat Preference : Our first indication 
that red-tailed tropicbirds nested on Caroline was the discovery 
of the skull, tail feather, and broken egg ( Subchapter 1 . 1 , PI. 46) 
under a small Tournefortia bush on a previously unnamed islet 
between Pig and North Brothers Islets. We named this sparsely 
vegetated collection of rubble "Skull Islet" (Subchapter 1.1. 
pis. 46,49). We later found 47 nests on another islet, naming 
it Bo' sun Bird (Fig. 1 ) after the species' common name. 

All nests were located under relatively open Tournefortia 
scrub less than 3 m tall in open, windy locations, with the 
majority (91%) on small islets (0.24-0.86 ha). All nests were 
under shrubs with few stems within a 0.5 m 2 nest space, and 
most had peripheral cover on the sides of the shrubs, both 
important factors in nest-site selection (Clark etal., 1983). All 
nests were in areas relatively free of Polynesian rats (Rattus 
exulans) and coconut crabs (Birgus Ultra): 5 nests on Long 
were within 50 m of the island's south point. 

There are large populations of Polynesian rats and coconut 
crabs on Caroline's bigger, more wooded islets. This rat, 
though basically vegetarian, is an effective seabird predator 
(Kepler, 1967; Norman, 1975) that in some years has taken 
65% of the red-tailed tropicbird eggs and 1 00% of the chicks on 
Kure Atoll (Fleet, 1972). Coconut crabs are also known bird 
predators (Clapp & Sibley, 1971a; Helfman, 1979; Reese. 
1987). It may be no accident that tropicbirds on Caroline 
are restricted to small, relatively open islets that harbor few, if 
any. rats and crabs or occur only on the tip of Long Island, 
where predator densities are low. We saw no rats on Bo'sun 
Bird Islet. Although rats could swim the 1 65 m to the islet, the 
nearly continuous presence of black-tipped reef sharks 
(Subchapter 1.1. PI. 10) in the channels surrounding the islet 
provides protection to its nesting tropicbirds. 

Numbers : In September 1988. we found a total of 
56 active nests on 5 islets (Fig. 2, Table 1) and estimated a 
minimum population of 60 pairs. The May 1990 expedition 
found 130 nests on Bo'sun Bird: our revised estimate for 
Caroline is approximately 300 birds. Bo'sun Bird Islet was 
surveyed by POBSP in June 1965, and no tropicbirds were 
located on the ground or in the air (F. Sibley, personal 



communication ). It is unlikely that red-tailed tropicbirds were 
present but overlooked at that time, suggesting that they have 
colonized the atoll only recently. The Caroline population is 
now the second largest colony known from the Line Group, and 
Caroline is only one of five islands in the archipelago where 
red-tailed tropicbirds are known to breed. 

Phenology : Of the 56 nests found in 1988, 54 contained 
eggs or chicks (Table 2). The 33 chicks were divided into 4 age 
classes (Fleet, 1974; Diamond, 1975a). which, together with 
the 21 eggs, provided an indication of laying phenology for 
140 days prior to our arrival (Fig. 3). Eggs in surviving nests 
had been laid at a fairly even rate from early May (possibly 
starting earlier) through September. The finding of only two 
additional pairs on territory, and only one courtship flight, 
indicated that laying was ending. On 24 May 1 990. many nests 
contained eggs and downy chicks (75% nests with chicks), and 
pairs were still courting. 

On Christmas Island, peak laying generally occurs from 
June to October ( Schreiber & Ashmole. 1 970), later than those 
parts of the 1988 and 1990 breeding seasons we observed on 
Caroline. 

Masked Booby (Sula dactylatra) (Figs. 4,5; PI. 2) 

The masked booby is widely distributed in the Atlantic, 
Indian, and Pacific Oceans. Clapp (1967) estimated that 
19,100 masked boobies bred in the Line and Phoenix Islands, 
with about 13,000 of them in the Line Islands, mostly 
(ca. 9,000) on Jarvis. 

Distribution and Habitat Preference : Eighty-four percent 
of masked booby nests (159) were on the windward, rubbly 
shores of Long and Nake Islands, extending to the north end of 
the atoll. Fifteen additional nests were scattered along the 
lagoon edges of five South Nake Islets (Table 1 ). Nests 
consisted of bare scrapes with exposed sand, usually within a 
sparse ground cover of Portidaca and Heliotropium (PI. 2). 
Over half the nests were amassed in one open colony on Nake 
that extended nearly 1,000 m, beginning approximately 150 m 
south of Tr. 2 and extending about 50 m north of Tr. 4 
(Subchapter 1.1, Fig. 8). Here a nearly unbroken Heliotropium 
mat 30-80 m wide, with patches of Tournefortia, occupied the 
area between the leading edge of the Tournefortia scrub and the 
beach crest. Nests were 20-30 m apart in the densest section 
(nearTr. 3). All nests were exposed to the sun, unlike those of 
the brown boobies. Some adults and juveniles roosted under 
the scrub; guano deposits indicated regular occupancy. 

A loose group of 7 breeding pairs was scattered on a broad 
plain of low herbs along a partially filled old interislet channel 
370 m south of the north end of Long Island (Tr. C. 
Subchapter 1.1, Figs. 8,40). Four more pairs nested in coral 
rubble along the channel separating Nake and Long, one pair 
with a downy young only 2-3 cm above high-tide flow on an 
"islet" between fingers of the channel, a precarious location 
where nesting surely must fail in stormy periods. No birds were 
seen there in March and May 1 990, following a severe storm in 
February 1990. Four pairs nested singly along a leeward 
1 ,000-m stretch of lagoon shore on the northern end of Long 
(Fig. 4); hardpan was the primary substrate. 



141 



Numbers : In September 1988 we found 189 masked 
booby pairs (Table 3), including those on territory (with or 
without nest scrapes) and juveniles (with or without attending 
adults). We found no "clubs" of nonbreeding birds. We 
covered most of the habitat favored by this species except the 
northern 300 m of Nake Island; in 1990 a few scattered pairs 
nested there. Our population estimate, including pairs we 
might have missed, was approximately 200 breeding pairs. 
The only other population estimate was of "ca. 10" birds 
(Clapp&Sibley, 1971a);in 1965 POBSP biologists (F.Sibley, 
personal communication) surveyed all locations where we 
found breeding pairs. Thus, 200 pairs represents a major 
increase in the population on Caroline Atoll. 

Phenology : In June 1965 only four masked booby nests 
containing eggs were found ( Clapp & Sibley, 1971a). indicating 
that nesting began in May or June. We found nests in all stages 
in September 1988 (Table 3. Fig. 5 1. The large age class in 
April may include some juveniles that could fly ( i.e.. were older 
than 1 80 days). We may have undercounted naked chicks, not 
wishing to expose them to the sun by frightening the brooding 
adult. Laying began in April or earlier, peaked in June and July 
(Fig. 5). and continued until our survey in late September. The 
34 pairs on territory, many with nest scrapes (Table 3 ). indicated 
that laying was still in progress and would continue into 
October. 

In March 1 990. 3 1 pairs were on territory or were attending 
nests, eggs, or older chicks, indicating that a new breeding 
season was under way as the previous season was ending. By 
May 1 990 there were 63 nests, mostly with eggs, and there were 
no older chicks. Thus, the 1990 season augments the 1988 data 
and suggests an annual cycle with egg laying beginning slowly 
in lebruary and March, peaking in June and July , and declining 
to a low ebb from December to February. 

The large number of fledgedjuveniles and nests with older 
chicks, in both September 1988 and March 1990. indicated that 
the 1988 and 1989 breeding seasons were very successful. It 
also suggested that potential predators (rats and coconut crabs ) 
posed little hazard to this hardy species. 

Brown Benin (Sula leucogaster) (Fig. 2) 

This widely-distributed pantropical species has an 
estimated population in the Line and Phoenix Islands of about 
$,200 (Clapp. 1967; Perry. 1980). with over half of them 
(2.000) recently found on Maiden Island, in the Southern Line 
Group. 

Distribution and Habitat Preference : Breeding brown 
boobies on Caroline were restricted to the w mdward edges of 
Tournefortia scrub and forest, generally within 15-20 m of 
high water. In 19SS we found nests on four islets (Fig. 2, 
Table I). Long, with 12 pairs, was the onlv islet supporting 
more than a single pair. They w ere located on the northern two- 
thirds of the island: four pairs formed a loose colony near the 
head of Tr. A ( Subchapter I.I. Fig. Si. All nests were under 
Tournefortia bushes approximately 3 m tall. In March 1990. 
we found 20 pairs of brown boobies, all on windward Nake as 
far as (he islet's northern extremity. There was no evidence of 
nesting on Long Island. On May 22. 1990. only three nests, all 
with engs. were found on Nake. 



On 22 September 1988. we saw 2 birds plunge-div ing w ith 
masked and red-footed boobies approximately 500 in west oi 
South Island. On the atoll, flying brown boobies were observed 
soaring only along the windward beaches. Two birds roosted 
on the south-central beach of South, and another was found 
roosting on Kota. 

Numbers : We counted 1 5 pairs during perimeter surveys 
in 1988, yet found none on the transects. Since we covered 
virtually all the w indward beaches ( Subchapter 1.1. Fig. 9 1. we 
are confident that fewer than 20 pairs nested on the atoll. Our 
population estimate for 1990 was 25 pairs. 

The POBSP (Clapp & Sibley, 1 97 1 a) found three nests on 
Nake in June 1965, estimating a population of 15 birds. Even 
though our surveys triple the known population, the brown 
booby remains a rare seabird on Caroline. 

Phenology : With the exception of one recently 
fledged juvenile, all nests contained eggs in September 1988 
(Table 3). Clapp & Sibley (1971a) found eggs in June 1965. 
In March 1990, the 20 pairs were all on nests whose contents 
ranged from eggs to an older juvenile. However, only 
2 months later, only three nests, containing eggs, could be 
found. These data from 3 years suggest that the species may 
have trouble rearing young on the atoll. More juveniles should 
have been encountered, especially in May 1990. Predationby 
Polynesian rats or coconut crabs could limit reproduction on 
the atoll. 

Red-footed Booby (Sula sula) (Figs. 6.7: Subchapter 1.1. 
PI. 51) 

This pantropical booby numbers over 55.000 individuals 
in the Line Group (Clapp. 1967; Perry, 1980), making it one of 
the most important regions in the world for this species. 
Caroline holds the fifth largest known red-footed booby colony 
( see Nelson. 1978). The largest known colony ( 140.000 pairs) 
is found on Tower Island (Galapagos): 3 of the 5 biggest 
colonies occur in the Line Group. 

Distribution and Habitat Preference : In 1988, the red- 
footed booby bred on 28 islets, ranging in size from Nautonga 
(0.34 ha) to Nake (107.46 ha) (Fig. 6). On the Windward 
Islands, red-foots occurred from Nake to Tridacna. absent only 
from the smallest islets ( Noddy Rock. Skull Islet. Motu Atibu ). 
The species was also widespread on the leeward islets, extending 
from Pandanus to Eitei. The tiny islets ( Fishball, Azure, Reef- 
flat) were not occupied. 

Red-foots are tree nesters w hose distribution on Caroline 
closely matched that of Tournefortia scrub and forest 
(Subchapter 1.1. PI. 51). Thev sometimes utilized smaller 
Pisonia or Cordia trees where thev intermingled with 
Tournefortia and occasionally built nests in the tallest (>15 mi 
Pisonia. They nested in smaller Tournefortia patches within 
the peripheral scrublands, especially those not directly exposed 
(o the trade winds. They clearly av oided smaller islets because 
of the lack of suitable Tournefortia in which to breed. Thev 
nested inward from (he vegetated edges of the islets, generally 
at 3-6 m in height, and were distributed in broken rings around 
the smaller motus in areas of moderate winds. A higher 
percentage of the population occurred on perimeter surveys 
than on cross-island transects. 



142 



Red-foots were absent from South Island, which was 
primarily covered with Cocos (Subchapter 1.1, Figs. 50,51). 
Even though Tournefortia occurred on all its coastlines, no 
boobies nested in them. Ana-Ana was also unoccupied: the 
presence of a family of four people, a cat (removed in 1990), 
and a dog undoubtedly discouraged nesting attempts. Red- 
foots also avoided the mixed forests of south Nake, which 
contained much Cocos and Pandanus (Subchapter 1.1, 
Fig. 37 ). Red-footed boobies were thus found only in Caroline' s 
indigenous woodlands, primarily in Tournefortia >2 m tall; 
they avoided anthropogenic plant communities and man. 

Red-foots used a wider range of habitats for roosting. 
Nonbreeding birds were found throughout the taller indigenous 
trees, even in leeward situations where Pisonia and Cordia 
overhung the lagoon (as on Long Island). 

Numbers : The POBSP(Clapp& Sibley, 197 la) estimated 
5,000 ± 25% Red-foots on Caroline in June 1965, with about 
2,000 + 25% nesting pairs. In 1988, we sampled systematically 
more than 7% of the available habitat on all motus except 
Crescent (4.6% sampled) and North Arundel, and estimated 
that 2,22 1 pairs of red-footed boobies nested on 27 of Caroline' s 
islets (Table 1). We found an additional 1,234 roosting, 
nonbreeding birds. We know (Kepler, 1969; Nelson. 1978) 
that fewer boobies remain in their colonies during the day than 
at night. Thus, an unknown fraction of the population was at 
sea when we conducted our counts. Impressive flights of red- 
footed boobies returned each evening: 3^4 birds arrived for 
each one that had remained behind, many undoubtedly mates 
of incubating birds. To approximate the number of returning 
nonbreeding birds, we doubled the number of roosting adults 
to allow for an additional 1,234 adults and juveniles. Thus, our 
conservative estimate was at least 7,000 individuals. 

Because red-footed boobies were so dependent upon 
Tournefortia, we determined the nesting population on each 
islet by multiplying the number of nests found on transects by 
the ratio of sampled to total Tournefortia area. Perimeter 
counts (Subchapter 1.1, Fig. 9) were used if the number of red- 
foots observed exceeded the number calculated from the cross- 
island transects. 

Long Island held the greatest number of nests (659), 
mostly in the leeward Tournefortia and Tournefortia— Pisonia 
edge. Bird densities were typically highest on the largest islets: 
Windward and Tridacna, the largest Windward Islets, held 163 
and 1 1 1 nests, respectively; and Mannikiba. the biggest leeward 
islet, harbored the largest population ( 1 84) of the entire leeward 
side. There were exceptions, however: Pandanus, with four 
times the area of Tournefortia of any of the South Nake Islets, 
held fewer birds than three much smaller islets (Table 1 ). 

Tournefortia scrub and forest covered approximately 
125.25 ha (Subchapter l.l,Table9). Overall, there were 1.75 
red-footed booby nests/ 1 .000 nr of Tournefortia forest. Nest 
densities for occupied islets by island groups (Table 4) showed 
that red-foots favored areas less exposed to the trade winds: 
most nests on the windward motus were protected by well- 
developed Pisonia forests. The exposed Central Leeward 
Islets held the lowest nest densities ( 1 . 2 nests/ 1 .000 m 2 ), far less 
than on the South Nake Islets (5.3/1,000 nr), which are 



protected by the northern edge of Long. The greatest densities 
(7.8 nests/ 1 ,000 nr) occurred on the South Nake Islets south of 
Pandanus. 

Broadly speaking, red-foots breed in well-dispersed 
colonies. A record density of 600 nests/ 1 ,000 m 2 on Tromelin 
Island (Indian Ocean) is exceptional. Elsewhere, 
53 pairs/1,000 nr on Tower Island (Galapagos), 40/1,000 nr 
on Moku Manu (Oahu. Hawaii), and 27/1,000 m 2 on Half 
Moon Cay (Honduras ) are more consistent high-density colonies 
(Nelson, 1978). Only on tiny Motu Kota (Subchapter 1.1, 
Fig. 52), with 1 2 nests in 303 nr of Tournefortia (40/1 ,000 nr ). 
did we find such density, and for this reason we named the islet 
"Kota" (Gilbertese for red-footed booby). 

Phenology : In September 1988, we located 339 nests. Of 
the 152 whose contents could be determined, 87 were empty, 
63 contained eggs, and 2 held downy chicks. We saw dozens 
of flying juveniles along the windward coasts. Most pairs were 
building or guarding their nests during a pre-laying stage that 
lasts from 1 1-35 days (Nelson, 1969). Of the pairs with nests, 
57.2% had yet to lay and 41.4% had laid their eggs between 
mid-August and late September (Fig. 7). Applied to the total 
breeding population, approximately 1,270 nests were in the 
prelaying stage and would be expected to produce eggs 
throughout October. An additional 9 1 9 nests had a mean laying 
date in early September (Fig. 7). Red-footed boobies were 
synchronous with brown boobies but delayed relative to masked 
boobies. 

In June 1965. nests containing prelaying adults, eggs, and 
young in all stages indicated that the birds were in the midst of 
a protracted breeding season extending from January to June. 
Our data reveal that no successful nesting occurred in May- 
June 1988. Data from March and May 1990 indicate that nest- 
building began in January (or earlier), with eggs laid from 
January to May. Red-footed boobies in other tropical locations 
have variable, opportunistic breeding seasons that depend 
upon food availability (Nelson. 1978); our data suggest that 
similar pressures could be operating at Caroline. 

Color Morphs : Red-footed boobies are polymorphic 
(Nelson, 1978). The basic plumages are brown or white, with 
brown morphs having many combinations of tail, back, scapular, 
foot, and bill colors. A variety of brown forms and white forms 
occurred on Caroline, with a ratio of 9:1 (337 brown to 
35 white), which contrasted sharply with Nelson's (1978) 
statement that "in the Line and Phoenix Islands all birds are 
white morphs." Most of the dark morphs were the "white- 
tailed" form (see Nelson, 1978. pp. 660-661 ). The variations 
and proportions of plumage types show clinal change in the 
Line and Phoenix Islands ( F. Sibley, personal communication), 
and the question of plumage morphology needs much more 
study in the central Pacific. 

Great Frigatebird (Fregata minor) (Subchapter 1.1. Figs. 7,8 
and PL 42) 

The great frigatebird breeds at widely scattered locations 
throughout tropical waters in the Atlantic, Pacific, and Indian 
Oceans. It is known to breed on all of the Line Islands except 
Starbuck (Perry. 1980). 



143 



Distribution and Habitat Preference : Great frigatebirds 
nested on 25 islets, including Nake, Long, and most of the 
larger islets (Fig. 8. Table 1 1, ranging in size from Azure 
(0.20 ha) to Nake ( 107.46 ha). Every occupied islet had some 
Pisonia forest, even if only a single tree (Azure). The larger 
islets lacking Pisonia forest (Arundel. 7.34 ha; Tridacna. 
9.08 ha) lacked frigatebirds in 1988. although frigatebird 
chicks were present on Arundel in early 1989 (Anne Falconer, 
personal communication). 

Although great frigatebirds were similar in nest 
requirements to red-footed boobies, there were significant 
differences: the frigates tended to nest higher in, and closer to, 
the outer edge of the canopy (although nests were found as low 
as 1.3 m). Nest sites were more sheltered from the wind than 
those of red-foots and in locations where the birds could take 
flight easily. Such site preferences may explain the association 
with Pisonia. Pisonia reaches 25 m on Caroline, taller than 
other tree species, providing a windbreak on most islets. The 
largest colonies (Nake. Long, Pig. Mannikiba) were found 
leeward of these stands. We found nests in Tournefortia, 
Pisonia. and Cordia. They were often in the Tournefortia- 
Pisonia interface, generally in the taller Tournefortia. One 
colony on south Long overhung the lagoon in a dense Pisonia 
stand. Frigates were not found in any anthropogenic forests 
and were absent from then-inhabited Ana-Ana. 

Numbers : The previous population estimate for great 
frigatebirds on Caroline was 10. 000 birds (Clapp & Sibley, 
1971a: Perry, 1980). We calculated that 2.427 pairs bred or 
attended territories. An additional 617 birds roosted, thus the 
entire population was approximately 5,471 individuals. A 
large but undetermined number of birds soared over the atoll 
throughout the day, and an uncountable number of birds, 
including Hedged juveniles that would ultimately return to the 
island to nest (Diamond. 1971). were undoubtedly at sea. 
Because this species is difficult to count accurately, it is unclear 
if the population has changed since 1965. 

Phenology : In frigatebirds. the scapulars, which first 
appear at 81 days in Fregata magnificent (Diamond. 1973). 
erupt before the primaries. Because we lack chick stage data 
for F. minor and F. arid, we have modified ages from Diamond 
( 1 973 ) for F. magnificens, using the hatching times for F. arid 
and F. minor from Nelson (1976) and fledgling ages from 
Diamond! 1 975b) to construct approximate development stages 
for the species on Caroline. Since (hey Hedged at an earlier age 
than /•'. magnificens, we have reduced the ages for chicks with 
erupting primaries for F. arid and F. minor, kept the duration 
of the earlier stages approximately the same, and reduced the 
period in juvenile plumage. 

We found 214 nests in 1988. Of the 144 in which we 
determined contents, 49 contained eggs or young chicks. 
27 held chicks with developing scapular feathers, and 
68 contained older chicks (fable 5 1. The additional 70 adults 
occupied nests oi unknown contents; mans probably held eggs 
or young chicks or were empty. We saw fewer than 
10 displaying males, so the breeding season was winding 
dow n. This was also indicated by the high proportion 1 87' i |oi 
nests with chicks, mam ol them old. A major laving effort had 



begun in March-April (Fig. 9) and continued into September. 
In March 1 990. an abundance of fly ing juveniles and occasional 
larger chicks down to the downy stage indicated that the 
previous year's breeding season was ending. A small number 
of males were beginning another courtship cycle. By May 
1990, courtship and egg-laying were still under way, and nests 
contained eggs or small chicks up to the "remiges" stage. Peak 
laying on Christmas Island ( Pacific Ocean) occurs from March 
to May (Schreiber & Ashmole. 1970), the same laving cycle 
observed on Caroline in 1988 and 1990. 

Lesser Frigatebird (Fregata ariel) (Figs. 8.10) 

The lesser frigatebird is. along with F. minor, one of the 
true pantropical species. It breeds and disperses widely within 
the tropical Pacific (Sibley & Clapp. 1967). One of the largest 
populations in the world (30.000-85.000) breeds on McKean 
Island, in the Phoenix Group (Garnett. 1983). Lesser frigatebirds 
breed on four of the Line Islands, with the population on 
Maiden (7.000) the largest in the archipelago (Perry. 1980). 

Distribution and Habitat Preference : In June 1965. lesser 
frigatebirds were found nesting in one compact colony on the 
leeward north end of Long (Clapp & Sibley. 1971a). We found 
a single colony in leeward Pisonia forest on western Nake 
(Fig. 8). both in September 1988 and May 1990. The birds 
nested high (to 18 m) in the Pisonia and Pisonia-Conlia edge 
facing an open Tournefortia savannah. Although primarily 
composed of F. ariel. a few F. minor were scattered along all 
but the eastern edge of the colony. West of the birds, across the 
open forest. F. minor and Sula sula nested in a mixed colony in 
a denser stand of Tournefortia. Birds were seen soaring over 
Nake. Long, and the leeward islets but were not found roosting 
or nesting away from the colony on Nake. How ever, in March 
1990. approximately 650 lesser frigatebirds were swarming 
above, and roosting on. Motu Nautonga in a light cluster. 
possibly preparing for nesting. 

Numbers and Phenology : Pacific Ocean Biological Sun ev 
Program biologists estimated a population of 1.000 lesser 
frigatebirds on Caroline in June 1965. with 400 ± \0% breeding: 
only eggs were found (Clapp & Sibley. 197 la). On Christmas 
Island. F. ariel laid in May and June in 1959. 1963. 1964. and 
1 967 ( Schreiber & Ashmole. 1970). Of 46 nests found in 1988. 
we could inspect the contents of only 26: all contained 
feathered chicks (Table 5). Laying dates ranged from March 
through July (Fig. 10). with a peak from April to June. Caroline's 
lesser frigates, therefore, appeared to be synchronous with 
those on Christmas. 

Because w e did not determine the colony limits, we cannot 
provide a population estimate. There was a minimum of 
20( ) birds in 1 988 ( 46 nests, plus roosting and flying indiv iduals) 
and 500 pairs in 1990. 

Sooty Tern (Sterna fuscata) (Figs. 11.12: Subchapter 1.1. 
PI. 59) 

This tern is the most widespread and abundant tropical 
seabird in the world. Under favorable conditions it forms 
immense colonies numbering into the millions. It is known to 
breed on 7 of the Line Islands: the largest population in the 



144 



Pacific is found on Christmas Island (15,000,000 at highest 
count), and 3,000,000 have been recorded on Starbuck (Perry, 
1980). 

Distribution and Habitat Preference : To date, 19 colonies 
from 10 islets are known for the years 1965, 1988, 1989, and 
1990 (Fig. 11). In September 1988, we found three colonies, 
two on the northern half of Long and one on Bo' sun Bird Islet; 
all fit the general habitat description in Clapp & Sibley ( 1 97 1 a). 
Colony A, nearly square, was 210 m on a side. Eggs were 
placed under a savannah-type Tournefortia scrub, from 1-4 m 
tall with approximately 60% canopy cover (Subchapter 1.1, 
PI. 59). The substrate was coral rubble mixed with sand, 
covered by Heliotropium (5%), Portulaca (1%), Laportea 
(< 1 % ), and Lepturus (< 1 %), typical of old interisland channels. 
Colony 1 was located in a broad sandy corridor with two large 
"groves" of Tournefortia. The northern subpopulation extended 
116 m along the windward beach, but 248 m along the lagoon. 
The southern subpopulation began 28 m further south along the 
beach, fronted the seaward reef for 86 m, and was shaped like 
a blunt triangle, its apex pointing toward the lagoon. Most 
chicks were under Tournefortia, which consisted of shrubs 
2-4 m high with 80% canopy cover. The substrate was also 
older beach sands mixed with coral rubble and covered with 
Portulaca (40% cover), Lepturus (<5%), and Heliotropium 
(<5%). The Bo'sun Bird colony, a rough oval approximately 
55 m wide by 70 m long, was under 2-3 m high Tournefortia 
with 75% cover, on coral rubble/sand sparsely carpeted with 
Portulaca and Heliotropium. 

Numbers : Populations were determined by measuring 
colony dimensions, then counting eggs and/or chicks in 9-m 2 
sample plots located at random points along a compass line. 
Because juveniles moved as we approached, they were counted 
6 m ahead of us in estimated 3 x 6 m plots. The Colony I 
subcolonies (North, South) were treated separately. 

Colony size (rounded) in 1988 ranged from 
1 27,000 ± 30,000 "nests" (Colony A) to 1,500 ±750 new eggs 
on Bo'sun Bird Islet (Table 6). There were an additional 
6,900 + 1 ,600 nearly-fledged chicks in the Bo'sun Bird colony, 
resulting from eggs laid three months earlier. 

The total number of eggs and chicks was 188,000 ±21%. 
Actual numbers of adults are difficult to estimate but in other 
studies have exceeded the number of eggs and young by factors 
of more than two because innumerable eggs and chicks were 
lost, colonies often overlapped, and many nonbreeding adults 
joined the prebreeding swarms or associated with breeding 
birds. Schreiber & Ashmole (1970), relying on POBSP data 
from Johnston Atoll (north-central Pacific), estimated that four 
adults were present for each egg laid. Pacific Ocean Biological 
Survey Program data from Johnston (Amerson & Shelton, 
1976) indicated that about 600,000 adults were present in a 
colony with 105,000 eggs, or approximately 5.7 adults/egg. If 
we assume that real numbers of terns in our colonies lay 
midway between 4 and 5.7 times the number of eggs and 
chicks, then the number of sooty terns using Caroline Atoll 
would have ranged between 720,000 and 1,100.000 birds 
(91 1.800±21%). This is twice the estimate provided by Clapp 
& Sibley (1971a), even though we found fewer colonies. 



In March 1990, laying was just beginning in two colonies 
on Long Island, (625 x 1 50-3 1 5 m wide and 1 80 x 1 60 m wide). 
Enormous numbers of birds, both on the ground in densities up 
to 9 or 1 pairs/nr and in the air, made it impossible to calculate 
a reasonable population figure. According to Anne Falconer, 
these two colonies were very successful. Similarly, counting 
was difficult in May 1990 when six large prebreeding swirls 
hovered like huge clouds of gnats over discrete islets and islet 
groups (Fig. 1 1). Our 1988 estimate of approximately one 
million birds is probably a conservative count for the atoll as a 
whole on an annual basis. 

Phenology : The incubation period in sooty terns is about 
4 weeks (Dinsmore, 1972). Young fledged 7-8 weeks after 
hatching, although fledgling ages, dependent upon food supply 
(Schreiber & Ashmole, 1970), are highly variable. 

Four separate sooty tern colonies had been started over the 
12-week period prior to our study in 1988 (Table 6). On Bo'sun 
Bird Islet a new wave of laying was just beginning in an open 
area immediately southwest of most of the colony, while nearly 
fledged chicks scurried about beneath the Tournefortia. 
Undoubtedly many young had already fledged, so many more 
eggs would have been laid in early July by this colony than 
indicated (Fig. 12). The two colonies on Long were established 
at different times: the short-tailed juveniles in Colony 1 
preceded the large number of eggs, hatching eggs, and downy 
chicks of Colony A by 3-4 weeks. 

The July-September laying period on Caroline in 1988 is 
very different from the bimodal breeding (May-June, 
December-January) reported from Christmas Island, Pacific 
Ocean (Schreiber & Ashmole, 1 970), and the May laying dates 
noted for Caroline by the POBSP in 1965. Additional data 
(Anne Falconer, personal communication) indicate that sooty 
terns may lay any time (Fig. 11), certainly January through 
September (1988 to 1990). Severe storms, which destroyed 
large Long Island colonies in February 1990, were perhaps 
responsible for reinitiating breeding activities on the leeward 
side of the atoll within the next few months. A great deal more 
research will be needed on Caroline before the breeding seasons 
for this species are fully understood. 

Brown Noddy (Anous stolidus) (Fig. 13) 

This tern, primarily a tree nester, is widely distributed 
throughout the warm oceans of the world. It is abundant in the 
Line and Phoenix Groups, with an estimated total population 
exceeding 40,000 birds. Brown noddies are most abundant on 
Palmyra Island (10,000 birds). 

Distribution and Habitat Preference : The brown noddy is 
second only to the white tern in the number of motus (28) upon 
which it is known to breed (Fig. 13). It utilized the smallest 
(Noddy Rock, 0.02 ha) and largest (South, 104.41 ha) motus, 
nesting upon coral rubble and in plant communities ranging 
from the simplest herb mats to Tournefortia. Pisonia, Cordia, 
Cocos, and the mixed anthropogenic forests of South and 
Nake. Most pairs were well dispersed, nesting from the outer 
edges of Tournefortia to the central, inner branches of Pisonia, 
and from the ground to the crowns of 25-m Cocos. When 
nesting sympatrically with black noddies in Pisonia. the brown 



145 



noddies typically occupied portions of branches closest to the 
trunk. Brown noddies nested almost solitarily in the Cocos 
canopy on South, were found within dense colonies of black 
noddies and white terns in tall Pisonia forests, with red-footed 
boobies and great frigatebirds in Tournefortia, and amidst 
sooty tems and red-tailed tropicbirds (Bo'sun Bird Islet). 
Apart from a few ground nesters on Raurau and Fishball, the 
only ground-nesting colony (80 nests) was located on a 
Portulaca mat on Noddy Rock — a site free of predators, 
although flooded during storms. 

Brown noddies often formed loose roosting "clubs" on the 
atol 1 ' s beaches. Aggregations of 1 5-20 birds were found on the 
west coast of South and on Sandy Inlet, south-central Nake. 

Numbers : Clapp & Sibley ( 1 97 1 a) estimated a population 
of 1,000 birds in June 1965, with about 800 birds breeding 
(with eggs and young). We estimated a total population of 
1,491 breeding pairs (Table 1 ). Because nests high in Cocos 
palms were difficult to detect, we undoubtedly overlooked 
many, and our estimate of approximately 3.000 birds is 
conservative. Although larger than the population estimated 
by POBSP (Clapp & Sibley, 1971a), uncertainties about the 
1965 survey coverage (F. Sibley, personal communication) 
prevent us from knowing if Caroline's population has changed 
over the past 25 years. 

Phenology : On Christmas Island, the timing of egg laying 
varies between colonies. In general, peak laying occurs from 
March to May, and from November to December. On Caroline, 
mating and nest-building were found in March 1990, but by 
May only a few eggs had been laid. Eggs and young were found 
in June 1965 (Clapp & Sibley, 1971a) and in September 1988 
(present study). We found 246 nests in September 1988 and 
determined the contents of 106: 103 held eggs, 3 held downy 
chicks. The incubation period is 35-37 days (Dorward & 
Ashmole, 1963), so all viable eggs had been laid within the 
previous 40 days (mid-August to late September). Because 
many nests were being built, we feel confident that laying 
continued into October. Clearly more research is needed to 
determine whether laying occurs in regular cycles. 

Black Noddy ( Anous minutus) (Fig. 14) 

The black noddy is widely distributed in the tropical 
Atlantic and Pacific. It is abundant in the Line and Phoenix 
Groups, with populations of 16,000 estimated in the Phoenix 
Islands (Clapp, 1967) and over 46.000 in the Line Group. 
Centers of abundance are Palmyra (20,000) and Christmas 
(14.500) (Perry, 1980). 

Distribution and Habitat Preference : The black noddy is 
a tree-nesting species that on Caroline prefers tall stands of 
Pisonia. The largest colonies (61% of the population) were 
found in the grand Pisonia forests (to 25 m) on Pig and North 
Pig, We found breeding birds on 18 motus. with colonies 
exceeding 200 pairs in the Pisonia on Nake, Long, Arundel, 
and Bird (Fig. 14). The only significant colony not primarily 
associated with Pisonia was found on Tridacna, where 
approximately 230 pairs nested in the tallest (ca. 8 m), most 
central Toumefortia-Morinda forest. Black noddies always 
nested in dense colonies near islet centers and were integral 



components of these plant communities: their droppings, 
coating the ground with a film of guano, constantly enriched 
the islet's meager soils. 

Numbers : Clapp & Sibley (1971a) estimated that 
7,000 ± 25% birds were on Caroline. During our visit the 
population was much larger: 5,122 pairs were estimated for Pig 
and North Pig alone (Table 1 ). Basing our numbers primarily 
on the densities of sampled colonies in Pisonia, we estimated 
that nearly 8,400 pairs were nesting during our 1 988 visit. Our 
population estimate approached 17.000 birds, to which an 
unknown number of nonbreeding birds could be added. These 
values place the Caroline population far above that for 
Christmas, making it the largest known population in Kiribati. 

Phenology : Black noddies were just beginning a new 
breeding season. On 27 September we observed hundreds of 
birds gathering Tournefortia leaves floating along the windward 
shore (Long) or flying with fresh leaves to their nests (Pig, 
North Pig). Of the 1,085 pairs counted on transect, 536(49%) 
perched as pairs, were defending nest sites, or were building 
nests. An additional 273 pairs were attending nearly-completed 
nests but were not incubating. The remaining 276 pairs were 
incubating. Thus, 75% of the pairs had not laid eggs. The 
contents of 230 nests were unknown, although we assumed 
they contained eggs because of the incubating positions of the 
adults. Of 46 nests into which we could see, 45 held a single 
egg, and one contained a downy chick less than 5 days old. 

The breeding seasons for black noddies on Christmas 
Island and Johnston Atoll peak in April and May (Schreiber& 
Ashmole, 1970; Amerson & Shelton, 1976), where pairs are 
highly synchronous, laying most of their eggs within a 
2-3- month period. The Caroline colony, also synchronous, 
but beginning egg-production in September, would be expected 
to peak in October/November, six months out of phase with the 
colonies further north. In 1990. however, black noddies were 
just beginning to mate and nest in March, and by May some 
were still sitting tightly on nests, while others had chicks in all 
stages. 

Blue-gray Noddy (Procelsterna cerulea) 

Blue-gray noddies nest widely across the Pacific from the 
Kermadec Islands to Hawaii. They are scattered throughout 
the Line and Phoenix Groups. In the Line Islands, they were 
formerly known to breed only on Christmas and Maiden 
(Perry, 1980). Eggs are placed in nests minimally provided 
with twigs and may be on coral rubble, sheltered under 
vegetation, or under coral slabs to depths of I m (Rauzon 
etai, 1984). 

The blue-gray noddy was recorded as "present" on Caroline 
by Perry ( 1 980). Clapp & Sibley ( 1 97 1 a) noted birds over the 
lagoon hut saw none on land. When we approached Caroline, 
we saw two from the ship and later observed three Hying across 
the lagoon. We also saw three birds perched on the leeward 
islets, one each on the reef flats of Nautonga and Fitei. A third 
bird Hushed repeatedly from a small clearing around a pile of 
bottles on Raurau. but we failed to find a nest. In March and 
May 1 990, we observed blue-gray noddies on all of the Southern 
Leewards, plus Azure and Nautonga in the Central Leewards. 



146 



In summer 1990, Alexandre Falconer found one small 
blue-gray noddy chick, attended by its parents, on an open 
expanse of coral rubble on Motu Eitei, the first breeding record 
for Caroline. Eitei is adjacent to Raurau, which we predicted 
was the most likely breeding location for this species. 

Blue-gray noddies evidently breed in very small numbers 
on Caroline. Nests are hard to find, given their cryptic placement, 
the small number of birds present, and the extent of open 
habitat (67.7 ha of herb mats and 41 .4 ha of consolidated coral 
rubble). 

White Tern (Gygis alba) (Fig. 15, PI. 3; Subchapter 1 . 1 . PI. 55) 

The white tern is a widely-distributed pantropical species 
occurring in moderate numbers throughout the Line and Phoenix 
Groups. Clapp ( 1967) estimated 10,000 birds in the Phoenix 
Group, and Perry (1980) estimated 17,050 birds for the Line 
Islands. 

Distribution and Habitat Preference : White terns, the most 
widely distributed breeding bird on Caroline, nested on 32 of 
the 39 motus (Fig. 15). The only islets not occupied were tiny 
and sparsely vegetated. 

White terns nested from 1 to 15 m above the ground, 
wherever a branch or frond provided a relatively stable platform 
in Tournefortia (PI. 3), Pisonia, Cordia, Pandanus , or Cocos 
(Subchapter 1 . 1 , PI. 55). They did not form dense colonies but 
were scattered from the edge to the center of each islet, even on 
the windward sides, although they normally selected sites not 
directly exposed to the prevailing trade winds. They utilized 
isolated trees, scrub, or forest. An unusual departure from the 
white tern's usual mode of "nesting" was an egg laid in an old 
black noddy nest, 6 m up in an 8-m-tall Tournefortia on 
Tridacna Islet. 

White tern densities varied from islet to islet (Table 7). At 
one extreme, we found only two nests on Raurau 
(0.07/1,000 nr). Densities on other islets ranged from 
0.75/1,000 m 2 (Shark) to 6.67/1,000 nr (Nautonga) with a 
mean density of 1 .38 pairs/1 ,000 nr of woodland. Overall, the 
Windward Islets supported the highest densities. Although 
white terns also nested in anthropogenic forests, their densities 
were low: we believe that the low densities on South Island and 
the Southern Leewards (Table 7) are attributable to man. Of 
South' s 104.47 ha of vegetated land, only 4.2 ha (4.4%) was 
native woodland (Subchapter 1.1, Fig. 50); fully 84% was 
either Cocos (18.3 ha) or Cocos-lpomoea (62.5 ha) forest. 
Although most of the Southern Leewards are covered in 
unmodified natural forest, central Ana-Ana has been partly 
cleared (0.21 ha) to accommodate thatched huts and a garden. 
The activities of a family of four, with a dog and cat (until 
October 1990), have apparently depressed the white tern 
population on Ana-Ana and, perhaps, even on nearby islets. 
We found no white terns on Ana- Ana during our visit, although 
the Falconers, who vacated the atoll in summer 1991, assured 
us that they occasionally nested. 

Numbers : We used the total woodland area of each islet 
to calculate islet populations (Table 1 ) from our transect data. 
More birds were found on the largest islets except South Island. 
We estimated 1,094 pairs for Nake, 751 pairs for Long, and 
nearly 400 pairs for Tridacna; these 3 islets accounted for over 



half the population (and over half the native woodland). We 
estimated that 3,957 pairs bred on Caroline. This doubles the 
numbers of Clapp & S ibley ( 1 97 1 a) and cited by Perry ( 1 980) 
and exceeds by 3,000 the largest population formerly known 
for the Line Islands. 

Phenology : Of 569 pairs of white terns recorded on 
transect, 437 were roosting without obvious signs of eggs or 
chicks, 107 were incubating, and 25 had chicks (often adults 
were not present). Of the 25 chicks recorded, 17 were downy, 
7 retained extensive traces of down with remiges, and 1 was 
almost ready to fly. Incubation takes about 36 days ( Ashmole, 
1963); young may require from 40 to 96 days to fledged 
(Gibson-Hill, 1950; Ashmole, 1968). Nearly all chicks were 
far from Hedging and were less than 4 weeks old. 

On Christmas Island, Schreiber& Ashmole (1970) found 
that peak laying occurred in April-August each year, with 
some laying in each month. On Caroline, Clapp & Sibley 
(1971a) noted that about half of the birds had eggs, half had 
young in June 1965. In March 1990, we found only a few eggs 
and downy chicks, but in May a larger number of pairs were 
breeding, with eggs and chicks in all stages. Although we 
found that white terns on Caroline do lay during the peak period 
on Christmas, it was clear that in 1 988 most eggs were laid after 
mid-August. 

Other Birds on Caroline Atoll 

Seven species other than seabirds have now been recorded 
on Caroline. Six of them are migrants (five shorebirds and a 
long-tailed cuckoo). The reef heron is apparently resident, 
although no nest has yet been found. 

Reef Heron (Egretta sacra) 

We found 15 reef herons scattered on 8 islands: Nake (1), 
Long (2), Pig (1), Brothers (3), South (2), Mannikiba (2), 
Matawa(l), and Emerald (2), as well as on the open reef flats 
( 1 ). Although birds were found on both the seaward and 
lagoonward sides of the islets, most were along the lagoon 
edge, as also found by POBSP in 1965 (Clapp & Sibley, 
197 la). We estimated that approximately 30 birds were using 
the atoll. We found no signs of breeding. Of the 15 individuals 
we observed, 5 were dark, 8 were white, and 2 were of the pied 
morph. 

Lesser Golden-plover (Pluvialis dominica) 

This plover used the beaches and herb mats, generally to 
seaward. In September 1988, we found them on Nake (1), 
Long (4), Tridacna (4), and Mannikiba (1), estimating a total 
population of 20-30 birds, the same number found by POBSP 
(Clapp & Sibley, 1971a). In March 1990, we observed eight, 
and in May, three, all in winter plumage. 

Wandering/Siberian Tattler (Heteroscelus incanum or 
H. brevipes) 

In September 1988, we located 18 tattlers on 6 different 
islets: Nake (3), Long (3), Crescent ( 1 ), Arundel (2), South (7), 
and Emerald (2). All birds were either alone or in pairs and 
generally remained in the intertidal zone, although they often 



147 



foraged on herb mats close to the beach scrub. The total 
population was approximately 40 birds. Those few birds heard 
were all H. incanum. We saw six tattlers in March 1990 and 
several in May of the same year. 

Ruddy Turnstone ( Arenaria interpres) 

One turnstone w as found on the windward beach of Motu 
Mannikiba in September 1988. and five on atoll beaches in 
March 1990. The Caroline population probably does not 
exceed 1 5 birds. 

Bristle-thighed Curlew (Numenius tahitiensis) 
(Subchapter 1.1, PI. 23) 

The bristle-thighed curlew, common in the Line and 
Phoenix Groups, is a widespread migrant to the low atolls of the 
central and South Pacific during the boreal winter (Pratt el ai, 
1987). One of the world's least-studied shorebirds, the species 
is considered rare throughout its range (Johnsgard, 1981; 
Marks et ai, 1990) and is a candidate for the US Fish & 
Wildlife Service Endangered Species List (Gill. 1990). Clapp 
& Sibley! 197 la) estimated 20 birds for Caroline in June 1965. 

We counted 83 birds on 12 of Caroline's islets in 1988. 
including the 3 large islands (Nake. Long, South) and motus in 
the Windwards, Central Leewards, and Southern Leewards. 
In March 1990, we saw 20 curlews on 10 islets during 
incidental observations throughout the atoll, bringing the total 
number of islets on which they have been recorded to 16. On 
our return trip (May 1990) we only saw three curlews (only 
eight islets visited). Undoubtedly, curlews occur on all islets, 
utilizing essentially all plant communities. Although they are 
most conspicuous on the beaches and reef flats, higher numbers 
may actually forage in the forests during the day. The Falconers 
(personal communication) note that small numbers of curlews 
remain all year. They are least common between April and 
August and most abundant after September/October. This 
correlates with preliminary information from Rangiroa Atoll. 
Tuamotu Archipelago (Gill. 1990; Gill & Redmond, in prep.). 

Perimeter Habitats : On a complete perimeter count of 
South Island in 1988, we found 29 curlews. Twenty-one were 
foraging and loitering on the windward east coast, principally 
above the beach crest on coral rubble interspersed with herb 
mat. Similarly. 14 of 20 curlews found on Long and the 
Windward Islets foraged along the windward beach crest, with 
only 6 birds found on the lagoonward shores. Curlews were 
equally common on windward and leeward shores in the 
leeward islets, occupying habitats composed of coral rubble 
and sand. While the numbers indicate that curlews showed a 
preference tor windward shores, they may be biased because 
most birds were there in late evening (19 birds on South). 
Perhaps the\ use the relativeh open areas for roosting and 
foraging at dusk. Certainly the largest concentrations ( 13. 14) 
were found late in the day. We found our largest flock ( 14. 
Sands Inlet, Nake) at 1 600 h, foraging on compacted, silt) sand 
at the lagoonward end of the inlet, while single curlews dotted 
the interislet channels and shallow tidal reef flats. 

Vegetated Habitats : We found bristle-thighed curlew s on 
natural herb mats, in Tournefortia scrub. Pisonia forest, and in 



Cocos habitats, both in the healthy peripheral plantations and 
w ithin the dying Cocos— Ipomoea woodlands ( Subchapter 1.1. 
Fig. 36. PI. 34). One was captured in a mist net under a dense 
Cocos canopy. Disintegrating plantations in the center of 
South (54 ha) held a large population: calculated numbers 
produced an estimate of 154 curlews. They foraged over the 
//wwoefl-strewn ground, frequently using broken-topped 
coconut trunks as lookouts. We also found 5 curlews on 
transects in Pisonia forests up to 20 m tall on Nake (calculated 
population, 41 ). They were foraging on the relatively open, 
although dimly lit, forest floor. 

Numbers : From the 1988 data we estimated a population 
of ±300 curlews: 41 birds in Pisonia, 154 in Cocos-Ipomoea, 
43 on the beaches of South Island (29) and the Sandy Inlet of 
Nake (14), and another 62 scattered over the remainder of the 
atoll. Because 154 of them were calculated from the sighting 
of a flock of 7 curlews on one transect on South, there may be 
a bias in our population estimate. Incidental observations made 
off-transect did show, however, that curlews were common in 
the Cocos-lpomoea woodlands, and we believe that the numbers 
on the 104 ha that compose South Island approximated our 
estimated density (about 1.5 birds/ha). 

Bill Length : Bristle-thighed curlews show great variation 
in bill length immediately after the breeding season. Because 
birds of the year migrate south before their bills reach adult 
length (R.Gill, personal communication ). the ratio of "long" to 
"short" bills provides a rough estimate of juvenile survival. Of 
3 1 curlews seen in September. 20 were clearly adult length. 
7 were conspicuously shorter, and 4 were "intermediate" 
(probably young birds). All March and May birds had long, 
adult-sized bills. 

Some subadults also remain on their Pacific wintering 
grounds for up to 3 years, during which time they pass through 
a flightless phase (Gill. 1 990: Marks el ai. 1990). Noflightless 
birds were seen. 

Foraging : We saw one curlew chase and capture a small 
Polynesian rat at dusk on the south shore of South Island. The 
bird bashed the rat on the coral rubble, then ran rapidly about 
with the rat dangling from its bill. After about 5 minutes, the 
bird swallowed the rat with vigorous gulps. 

Polynesian rats, abundant on Caroline (especially in 
Pisonia- and Cocas-dominated habitats), remain within the 
forest during the day. but many move to the beach crest and tide 
line at dusk. They provide abundant potential prey for curlews, 
which can easily capture them on the open rubble. The 
synchronous appearance of rats and curlews at the beach- 
woodland interface at dusk may be part of the foraging strategy 
of this large shorebird. The presence of curlews beneath the 
forest canopy may also be partly associated w ith this source of 
food. 

Sanderling (Crocethia alba) 

One sanderling in winter plumage was seen at water's 
edge on the windward beach of Long Island on 27 September 
1988. Although Sanderlings arc well-known fall migrants in 
the Line and Phoenix Islands (Clapp & Sibley. 1967, 1968), 
this is the first record for Caroline Atoll. 



148 



Long-tailed Cuckoo (Eudynamis taitensis) (Fig. 16) 

The long-tailed cuckoo breeds in New Zealand and winters 
in the southwest Pacific. The center of its winter range lies in 
central Polynesia, but birds have been recorded as far as Palau 
in the northwest and Pitcaim Island in the southeast. Although 
occurring throughout French Polynesia and the Cook Islands, 
it had not been recorded from the Line Islands prior to our 
expedition (Bogert, 1937: Clapp& Sibley, 1 97 1 a,b; Pratt et al. , 
1987; Ellis etal, 1990). 

We found long-tailed cuckoos on 4 of Caroline's 39 motus 
(Fig. 16). We heard its distinctive monosyllabic and disyllabic 
call notes on South, Long, and Pisonia, identified one on Nake, 
and on 28 September collected a male in a mist net on Tr. 4, 
Long Island ( USNM 607 191). Soon after our return home we 
sent a description and photograph of this species to the Falconers: 
they, and AKK, have since seen them several times on Motu 
Ana- Ana in March, April, and May 1989-90. 

All the cuckoo sightings were at canopy or subcanopy 
level, and three of the four birds were found in Pisonia. 
The South Island cuckoo was located in a Cocos canopy over 
20 m high. The netted male flitted secretively within an 
undisturbed, tangled low-canopy (4—6 m) Pisonia-Tournefortia 
interface. We suspect that this elusive migrant occurs throughout 
the mid-to-upper levels of Caroline's forest canopy. 

These records establish the long-tailed cuckoo as a winter 
visitor to Caroline Atoll. Our observations on four islets, 
including the southernmost, northernmost, windwards, and 
leewards, suggest that many individuals were present. A 
March 1990 first sighting on Vostok (J. Phillips, personal 
communication) further suggests that the species disperses 
regularly to the Southern Line Group. 

Other Vertebrates 

Lizards 

Although "small lizards" were observed on Caroline 
in 1825 (Paulding, 1831), it wasn't until 1965 that the 
first collections were made (Clapp & Sibley, 1971a). We 
collected four additional lizard species, which increased the 
known terrestrial herpetofauna from three to six (Table 8). 
Although all are indigenous, the azure-tailed skink (Emoia 
cyanura) is suspected of being partly dispersed by man 
(Brown, 1956). All but two of the small lizard species known 
from the Line Islands (Crombie, 1 990) have now been found on 
Caroline. 

Turtles 

We found three Pacific green sea turtles ( Chelonia mydas), 
a threatened species (McKeown, 1978), at Caroline in 1988. 
Two were swimming over the lagoon reef flats, one west of 
Arundel, the second east of Ana- Ana. The third was in the open 
sea about 1 00 m west of South Island near the "boat entrance." 
Ron Falconer has seen up to seven turtles in the lagoon in a 
single day. In April and May 1990, AKK saw workers from 
Tahiti capture and kill a minimum of four green turtles in the 
lagoon; two more entered the lagoon during the following 
4 months (R. Falconer, pesonal communication). 



In March 1990, AKK and G. Wragg found three old nests, 
presumably of this species, on the northwest coast of Nake 
within 100 m of the northern tip of the islet. These are the first 
known turtle breeding records for the atoll. Young (ca. 1922) 
notes that the copra plantation laborers ate green turtles from 
September to December each year. The February 1990 storm 
added large amounts of sand to Caroline's shorelines, providing 
potential new habitat for turtle nesting. 

Terrestrial Mammals 

None of the terriers (see Subchapter 1.1) that were 
introduced to control rats on South Island in the early part of 
this century (Young, ca. 1922) have survived (F. Sibley, 
personal communication; R. Falconer, personal 
communication). In May 1990 the Falconers kept a dog and a 
cat on Motu Ana-Ana. Despite the fact that both animals 
generally remained close to the settlement, the dog regularly 
visited the other Southern Leeward Islets and accompanied the 
family on excursions in their sailing canoe throughout the atoll. 
As a result of our recommendations, the cat was removed from 
Caroline in October 1990. The Falconers, with their dog, 
vacated the atoll in mid- 1991. 

Bennett ( 1 840) noted "rats of a red-brown color," the first 
reference to rodents on Caroline. Dixon ( 1 884) found that rats 
were "not numerous" and that they nested "just at the base of 
the fronds" of the coconuts. Two specimens collected by the 
POBSP proved to be Rattus exulans (Clapp & Sibley. 1 97 1 a). 
They reported that rats were uncommon and restricted to South 
Island. 

The 19th and 20th century settlers found rats (presumably 
R. exulans) to be extremely abundant and very destructive to 
the coconut plantations. Maude (ca. 1938) states that rats 
destroyed the nuts, and that they contributed greatly to the 
eventual abandonment of copra enterprises on Caroline and 
Flint. They voraciously devoured both growing and fallen 
nuts, as well as dried copra. Being arboreal, they also lapped 
the juices of the flower stalks, preventing nut development 
(Young, ca. 1922). In a single year ( 1920) over 4,600 were 
trapped on South Island (Maude, ca. 1938). Thousands more 
were killed by terriers introduced to Caroline in a vain attempt 
to control them. 

We found rats on almost every islet; they were especially 
abundant on South, Long, Nake, and in the vicinity of coconut 
palms on smaller islets. We recorded rats during daylight hours 
on most transects, especially within the Pisonia forests. At our 
campsites on Long and South we noticed groups of 1 0-20 each 
night, so tame as to approach within 1 m while we were eating. 
The rats evidently undergo wide population fluctuations, as 
they were less abundant in March and May 1990 than in 
September 1988. 

We suspect that rats periodically reach most motus, and 
that the islets apparently lacking rats ( such as Noddy Rock ) are 
too small and/or depauperate to support a resident population. 
Because/?. e.v;</fl/f5isaknown seabird predator (Kepler, 1967; 
Fleet, 1972; Norman, 1975), the restriction of some species 
(i.e., red-tailed tropicbird) to small islets may be due to rat 
populations on larger islets. 



149 



We found rats throughout the Southern Leeward Islets and 
learned from the Faleoners that they are an abundant nuisance 
on Ana-Ana. They trapped over 1 .300 animals in 2 years and. 
like the pioneers before them, rely upon a dog to help keep them 
at bay. 

Marine Mammals 

On March 14, 1990, members of the Line and Phoenix 
Islands Expedition observed a minimum of 1 Pacific bottlenose 
dolphins (Tursiops gilli) in the open sea about 500 m off the 
southeast corner of South Island. 

Coconut Crabs 

The coconut crab (Birgus latro, Coenobitidae), the largest 
terrestrial invertebrate on earth, ranges throughout the tropical 
Indo-Pacific (Subchapter 1.1, PI. 56). It is highly esteemed as 
a source of food throughout its range, and for this reason is rare 
or absent on or near most inhabited islands. Because it is 
heavily exploited by man, it is under consideration for 
endangered species status (Reese, personal communication). 
Since March 1990, dozens of Caroline's coconut crabs have 
been killed for food and for preservation in formalin as curios 
for the Tahiti tourist market. Because of the increasing numbers 
of visitors to Caroline over the past 2 years, it is important that 
Caroline's coconut crabs receive protection. 

History : Young (ca. 1 922) was the first to mention coconut 
crabs on Caroline. In 1910 he wrote that "hundreds of great 
Coconut Crabs were seen: 40 large ones were caught by the 
crew of the schooner in an hour" on South Island. 

It is hardly credible that these enormous crabs, the dominant 
terrestrial animal of the atoll environment, could have been 
overlooked by all visitors prior to the 20th century. Perhaps 
their populations had been reduced or extirpated by earlier 
inhabitants. It is of interest in this regard that members of the 
1934 Mangarevan Expedition saw no coconut crabs on nearby 
Flint Island ( Fosberg, personal communication ), nor were they 
mentioned in a historical summary paper on Flint by Maude 
(ca. 1942b). Today Flint has perhaps the greatest density of 
coconut crabs in the world (Kepler, 1990b). 

Young ( ca. 1 922 ) noted that coconut crabs were considered 
a great nuisance by plantation laborers, who killed them 
mercilessly. Evidently the crabs dug up newly planted nuts and 
snipped off emerging shoots. On the smaller islets, visited less 
frequently than South, Nake, and Long, these depredations 
were difficult to control. Thus the small motu plantations were 
abandoned within a tew years of initial planting, resulting in a 
remarkably rapid recovery of the original vegetation (see 
Subchapter I.I. Ecological Succession section). 

Distribution and Habitat Preference : In 1988 and 1990, 
coconut crabs were abundant in the Cocos plantations of South 
and Nake, and present, in varying densities, on 1 2 other motus 
(Fig. 18). Although generally associated with ( 'ocos, we found 
them in woodlands of Pisonia, Cordia, and Toumefortia, as 
well as on rubble beaches (especiallj alter dusk). Although 
capable of survi\ ing without coconut palms, these crabs appear 
to seek them out. In the open understory of the tall plantations. 



or in groves of only one or two palms, telltale piles of shredded 
coconut husk fibers (Subchapter 1.1. PI. 57) disclosed the 
crab's presence. 

Because the prevalent coarse rubble substrates on Carol ine 
are hard to burrow into, coconut crabs occupied a variety of 
shelters: mounds of fallen coconuts and rotting palm fronds ( to 
1.5 m high), piles of rubble pushed against tree roots, sand 
burrows, tunnels within the/<?» (Subchapter 1.1, PI. 22), or 
large cavities in the boles of mature Pisonia trees. Coconut 
crabs also use a variety of shelters on the Tokelau Islands 
(Yaldwyn & Wodzicki. 1979) and Flint (AKK. personal 
observation). 

Numbers : Though conspicuous and slow-moving, coconut 
crabs are very difficult to count. Environmental variables such 
as rainfall, tide, lunar cycle, and population size and age classes 
all affect their activity (Reese. 1965: Helfman. 1977a,b). 
Although unable to conduct mark-recapture studies, we did 
make incidental observations on the numbers of individuals 
seen during transect and perimeter surveys. Coconut crabs are 
generally nocturnal, but we often found them during daylight, 
at times exposed on coral gravel beaches close to the waterline. 
Reese (personal communication) suggests that the abundance 
of rats occupying the same habitat may "force" the crabs to be 
more diurnal, as has been reported from the Indian Ocean. Our 
estimate of the population on Caroline is approximately 
2, 200 individuals, based on the number of daytime observations, 
the area covered, and the fact that only one out of every three 
or four individuals may be present on any given night ( Helfman, 
1977b; Reese, 1987). 

Foraging : Since the first detailed description of coconut 
crabs in 1 705, their shy, curious habits have been the subject of 
folklore, speculation, and misinformation (see Reyne. 1939). 
No scientist has yet published a documented account of a 
coconut crab actually opening a coconut (Helfman. 1979), 
which is widely held to be their consummate foraging behavior. 
Helfman is convinced that they do so, as he has found piles of 
coconut fiber and observed crabs walking w ith husked, opened 
nuts in places where he was the only other possible coconut 
husker. We repeat Helfman's ( 1979) assertion that coconut 
crabs do husk fallen coconuts. The piles of finely separated 
fibers (Subchapter 1.1, PI. 57) we encountered are totalis 
different from those produced by stick or machete husking, the 
two methods commonly employed by Pacific peoples. The 
crab tears virtually every fiber off individually, a process so 
painstakingly slow it probably takes days. We did not observe 
this on Caroline, but in March 1 990, AKK. on uninhabited Flint 
Island, observed a large male coconut crab that had just husked 
a coconut and was enlarging a small crack in the center of the 
smooth nut in a manner similar to that described by Gardiner 
(1907) in Reyne (1939, p. 297). 

On Caroline we observed the aftermath of coconut crab/ 
sooty tern predation or scavenging. On Brothers Islet, several 
entrances and pathways leading to coconut crab holes were 
strewn with the feathered skeleta of adult sooty terns (and 
possibly brown noddies ), along w ith numerous, freshly snipped 
branches of Pisonia up to 0.7 m long (Subchapter 1.1. 
Description and Ecology of the Motus section). This was also 



150 



recorded on Tridacna Islet by Clapp & Sibley (1971a) for sooty 
tern eggs and chicks and by Reese (1987) and Helfman ( 1979) 
on Enewetak, Micronesia. 

Size : Living in a rich environment free of predators, 
coconut crabs attain huge sizes on Caroline. The bodies of the 
largest males were as wide as a full-sized, unhusked coconut 
(Subchapter 1.1, PI. 56), giving them weights of at least 4 k 
(Helfman. personal observation). Thorax widths for 10 crabs 
(2 females with eggs. 8 males) averaged 129 mm. The thorax 
of the largest male measured 200 mm across, making it. along 
with many measured on Flint in 1990, one of the largest 
recorded coconut crabs in the world (the previous record was 
178 mm, Helfman. 1977a), with an age estimated to exceed 
40 years (E. Reese, personal communication). 

Conservation: Attributes of International Significance 

When Bennett (1840) stated that "no reefs we had seen 
could compete with those of Caroline for novelty and beauty," 
he was seeing an essentially pristine ecosystem through the 
eyes of a well-traveled naturalist. Caroline is stunning, but its 
value in today's shrinking world goes well beyond its physical 
beauty. Caroline' s exceptional attributes need to be elucidated, 
for the atoll has remained essentially unknown, even to some 
who have evaluated its worth (King, 1973; Garnett, 1983, 
1984). Man's presence anywhere, especially on pristine or 
near-pristine islands, generally brings rapid, often irreversible, 
changes. There are few, if any, islands remaining in the Pacific 
that can claim the impressive array of natural features exhibited 
by Caroline (Nicholson & Douglas, 1969). We believe that it 
is imperative that this atoll, which has managed to escape large- 
scale human disturbance, should remain undeveloped. 

Caroline was inhabited from 1988 to 1991 by a single 
family who lived a spartan, ecologically sound lifestyle. There 
are no roads, vehicles, stores, jetties, or services (water, sewage, 
or food), and no communication. There is no passage into the 
lagoon or safe sea anchorage. 

One of the most important of Caroline's attributes is its 
relative lack of disturbance. Aside from obvious human 
impacts on South. Nake. and Ana-Ana, the majority of its 
motus are dominated by indigenous vegetation and its reefs are 
basically pristine. There is no obvious pollution to alter the 
chemistry of the lagoon, beyond the flotsam and jetsam that 
spatter the windward beaches. It is thus an exceptionally clear 
and clean ecological laboratory that presents a picture of 
lagoon ecosystems "before" extensive disturbance by man. and 
one that provides the marine biologist with an unparalleled 
opportunity to study undisturbed natural communities. The 
atoll is rich in marine vertebrates and invertebrates; the maze 
of reefs and coral heads in the lower half of the lagoon has the 
highest recorded density of living Tridacna (20/. 25 m 2 ) ever 
recorded ( Sirenko & Koltun. Subchapter 1 .4. this volume), one 
of the few undisturbed world populations of this species 
(Subchapter 1.1. PI. 26). 

Caroline's many islets of different sizes provide excellent 
examples of soil and vegetation development, accompanied by 
variations in the diversity of bird life (Fig. 17). Many of its 



disturbed islets have recovered so remarkably they are almost 
indistinguishable from those which have remained pristine. 
The changing shapes of the islets, bearing emerging and mature 
plant communities, graphically portray a natural terrestrial 
atoll ecosystem. Caroline's concentric pattern of plant 
community development and the relationships of these 
communities to islet size, shape, and location on the atoll rim 
will continue to provide insight into evolutionary processes on 
atolls if they are left undisturbed. 

Caroline's insular flora, typical of central equatorial islands 
in their natural state and covering 70% of the atoll's land area, 
is of both national and international importance. The 27 extant 
plant species are 85% indigenous (possibly up to 93%), an 
extremely high figure for anywhere in the world. Six of the 
seven plant communities are natural. Lushly wooded, Caroline 
possesses some of the largest and grandest Pisonia (Pisonia 
grandis) forests known (Subchapter 1.1, PI. 43), occurring on 
29 islets. Although not as majestic as the prime forests on 
Washington and Fanning (Northern Line Group), which enjoy 
a heavier rainfall, those on Caroline are some of the finest 
representatives of this forest community in the entire Pacific. 
The 62 ha of Pisonia forest may well cover a larger area than 
on any other Pacific atoll. 

Caroline possesses significant stands of the hardwood kou 
(Cordia subcordata), a tree that is now rare in the Pacific. 
Caroline's groves (Subchapter 1.1, PI. 27), though small and 
often occurring in mixed native woodlands, total 26 ha. possibly 
the greatest area on any Pacific atoll. Its extensive coverage of 
tree heliotrope (Tournefortia argentea) is also notable: scrub 
and forests of this species form 40% of the atoll woodlands 
(Subchapter 1.1. PI. 47). Caroline's groves are some of the 
most unmodified in the Pacific: elsewhere Tournefortia is 
typically restricted to coastal fringes surrounding anthropogenic 
plantations (R. Fosberg. personal communication). 

Caroline offers many opportunities for ecological research 
under reasonably pristine conditions. Valuable clues as to the 
nature of underground water supplies may lead to a better 
understanding of the regulation of water supplies on inhabited 
islands. Marine biological and biomedical research could 
unearth clues as to the causes and treatment of ciguatoxicity of 
fishes and crabs. Such topics are increasingly important as 
more islands are subjected to disturbance and pollution. For 
example, the abundant red snapper (Lutjanus vaigiensis) and 
red spotted crab (Carpilius maculatus), both of which are 
notorious for their potent poisons, are safe to eat on Caroline. 
Associated with Caroline's plant communities are 
1 1 species of breeding seabirds numbering well in excess of 
1,000,000 individuals. The populations of most of these 
species are of national importance (Table 9). Caroline has the 
fifth largest red-footed booby colony (Subchapter 1.1, PI. 51) 
in the world. Its black noddy and white tern (PI. 3) populations 
are the largest in Kiribati. Under the 1975 Republic of Kiribati 
Wildlife Conservation Ordinance (amended in 1979). all 
known seabirds, migrant shorebirds. and endemic land birds 
are "fully protected throughout the Gilbert Islands" (Garnett, 
1983, p. 128). However, their protected status is in doubt on 
Caroline, due to attempts to lease the island for development. 



151 



Caroline deserves protection similar to five closed areas on 
Christmas Island and seven island sanctuaries in the Line and 
Phoenix Groups (Garnett, 1983). 

Caroline is an important wintering ground for the bristle- 
thighed curlew, a rare shorebird and candidate for the US Fish 
& Wildlife Service Endangered Species list. Some subadults 
remain all year on the atoll. Adult curlews pass through a 
flightless phase on Pacific islands, and Caroline provides a 
predator-free environment for this vulnerable phase of the 
curlew's life history. 

Caroline is exceptional in harboring a robust population of 
coconut crabs (Subchapter 1.1, Pis. 22, 56). These large 
invertebrates are abundant in the Cocos plantations of South 
and Nake and are found in good numbers in the indigenous 
Pisonia forests on most of Caroline's larger motus. 

Although green turtles are not abundant on the atoll, 
worldwide populations of these marine reptiles have suffered 
so greatly from overexploitation that remote, predator-free 
islands such as Caroline provide important, though small, 
sanctuaries. Since 1978 the Pacific green sea turtle has been 
reclassified by the United States Department of the Interior as 
threatened and the Pacific hawksbill sea turtle as endangered. 

From an archaeological point of view, Caroline houses 
one intact Tuamotuan marae ( ancient religious site) and another 
smaller site, partly destroyed by storms. The main site 
(Subchapter 1.1, Fig. 3, PI. 36), basically undisturbed since the 
1 870' s, is a relic of prehistoric occupation worthy of protection, 
being the only one of its kind in the Line and Phoenix Islands. 

Currently Caroline Atoll is owned by the government of 
the Republic of Kiribati and does not enjoy any legal protection 
(Garnett, 1983; Government of Kiribati, personal 
communication). Over the last 50 years it has been leased to 
private individuals who have scarcely altered the atoll. The 
benign management of the past is no guarantee for the future, 
and from October 1989 to the present, pressures to develop the 



atoll have mounted rapidly. Proposed schemes included an 
airstrip, a blasted channel through the reef, a hotel, a casino, 
logging, and commercial harvest offish and lobsters. In March 
1990, commercial harvesting of fish, the taking of coconut 
crabs, and illegal killing of green turtles began, emphasizing 
that no island, however remote, is guaranteed protection through 
isolation. In addition, during the past 2 years Caroline has 
become more visited than ever before, mostly without the 
knowledge or consent of the Kiribati government. 

There are many reasons why Caroline is inappropriate for 
resident tourists or development (remoteness, distance from 
medical aid, no regular water supply, no passage into the 
lagoon, etc.; see Kepler. 1990a). Caroline could support a 
limited number of ship-based ecotourists each year. 

Recommendations for an international preserve began in 
January 1989. During the 1990 ICBP expedition to the Line 
Islands, the team leaders discussed conservation matters with 
Kiribati government officials and key scientists in French 
Polynesia. Fortunately, documentation was obtained of illegal 
land clearing and wildlife disturbance during two visits to 
Caroline (Kepler, 1990a,b,c). The Kiribati government is 
considering altering their plans for the development of Caroline 
in favor of wildlife preservation. During summer 1990, French 
customs officials in Tahiti temporarily banned the exploitation 
of Caroline by French Polynesian nationals. 

As of December 1 990, The Nature Conservancy of Hawaii 
has restated its interest in establishing a triple-island preserve 
on Caroline, Vostok, and Flint and has begun discussions with 
the Kiribati government on Tarawa. The fate of these special 

islands may rest upon the results of these negotiations. 

We have a number of people to thank for their assistance in this 
project; however, our acknowledgments for Parts 1 and 2 of the 
manuscript. Ecological Studies of Caroline Atoll. Republic of 
Kirabati. South-central Pacific Ocean are listed at the conclusion of 
Part 1 . and are not repeated here. Again, it is our sincere pleasure to 
(hank these individuals. 



152 



TABLE 1 



Estimated number of breeding seabird pairs on Caroline Atoll, September 1988. 



Location 

Nake 
Long 
Windward Islets 

Bo'sun Bird 

Windward 

Crescent 

Atibu 

North Pig 

Pig 

Skull 

North Brothers 

Brothers 

Noddy Rock 

North Arundel 

Arundel 

Tridacna 
South 
South Nake Islets 

Pandanus 

Danger 

Booby 

Coral 

Lone Palm 

Kota 

Mouakena 
Central Leeward Islets 

Mannikiba 

Blackfin 

Matawa 

Emerald 

Shark 

Scarlet Crab 

Nautonga 

Azure 

Reef-flat 

Bird 

Fishball 
Southern Leeward Islets 

Raurau 

Eitei 

Pisonia 

Kimoa 

Ana-Ana 



Red-tailed Masked Brown Red-footed Great Lesser Sooty 

Tropicbird Booby Booby Booby Frigatebird Frigatebird Tern 



5 
47 



105 
69 



25 



496 


522 


659 


808 


163 


207 


28 


5 


31 


17 


14 


118 


25 


9 


* 


* 


37 


* 


111 


- 


32 


26 


139 


- 


52 


- 


28 


2 


48 


- 


12 


1 


8 


- 


184 


287 


* 


4 


5 


1 


3 


230 


* 


118 


11 


2 


7 


2 



29 



10 


31 


17 


14 


26 


14 


21 


3 



56 



Brown Black Blue-gray White 
Noddy Noddy Noddy Tern 



1,094 
751 



- 


390 


814 


179,800 


207 


986 


8,400 


10 


_ 


- 


20 


28 


- 


36 


60 


_ 


76 


3.194 


- 


82 


1,928 


_ 


23 


40 


- 


8 


15 


- 


80 


- 


_ 


11 


249 


- 


11 


230 


- 


163 


- 


_ 


26 


_ 


- 


33 


33 


- 


2 


1 


- 


6 


- 



161 


176 


37 


- 


3 


- 


7 


150 


37 


125 


7 


32 


42 


329 


5 


- 



6 
134 



110 
164 

69 

50 



227 
396 

381 

52 
37 
6 
15 
9 
3 



195 
11 
13 
83 
44 
2 

10 

2 

48 



Total Estimated 
Pairs 



56 



189 



2,427 



56 188,200 1,491 8,392 



3,957 



Breeding confirmed in 1989 or 1990. 



153 



TABLE 2 

Stages in the breeding cycle of the red-tailed 

tropicbird, Caroline Atoll, 27-29 September 1988 

(ages after Stonehouse, 1962). 



Approximate Age in 
Nest Stage Days From Laying No. Nests 



Juv. 


90-133 


18 


Remiges 


69-89 


4 


Scapulars 


58-68 


5 


Downy 


44-57 


6 


Egg 


0-43 


21 


Pairs on 






Territory 


- 


2 



TABLE 3 

Stages in the breeding cycle of the boobies of Caroline Atoll, 
21-29 September 1988. 



Nest stage'/Approximate age in days from laying 



Flying Pairs on 

Species Juv. Juv. Scapulars Remiges Downy Naked Eggs Territory 



Masked >164 145-164 115-144 89-114 

Brown > 1 64 144-164 114-144 88-114 

Red-footed - >150 111-150 75-110 



55-88 45-54 0-44 
54-88 44-54 0-44 
54-74 45-53 0-45 



No. nests in each stase 



Masked 

Brown 

Red-footed 



40 



many 



38 


4 


33 


34 


- 


- 


3 


8 


29 


- 


919 


1.270 



For descriptions of nest stage, see C. B. Kepler ( 1978). 



TABLE 4 

Density of red-footed booby nests in occupied Toumefortia habitats on 
islet groups, Caroline Atoll. September 1988. 



Islet Group 



Number Estimated Area of Nests/1.000 nr 

Occupied Number Toumefortia of Available 

Islets Nests (nr) Habitat 



Nake 


1 


496 


300,650 


1.6 


Long 


1 


659 


322.000 


2.0 


Windward Islets 


8 


434 


25 1 ,900 


1.7 


South Nake Islets 


7 


319 


59,801) 


5.3 


Central Leewards 


6 


239 


197,500 


1.2 


Southern Leewards 


4 


74 


39.600 


1.9 


rotal 


27 


2,221 


1.170,550 


1.9 



154 



TABLE 5 

Stages in the breeding cycle of frigatebirds on Caroline Atoll. 
21-29 September 1988. 

Nest stage/Approximate age in days from laying 



Species 


Juv. 


Primaries Scaps. Downy Naked 


Eggs 


Great 
Lesser 


191-220 
181-210 


101-190 81-100 (56-80) a (56-80) 3 
91-180 71-90 56-70 46-55 

No. nests in each stage 


0-55 
0-45 


Great 

Lesser 


22 

4 


46 27 ( 30 J ) 
13 4 5 


19 




Duration of naked and downy chick stages are lumped because it was often 
impossible to see into canopy nests. 



Colony Location 



Long Island, A 



Long Island. 1 N 



Long Island. 1 S 



TABLE 6 

Sooty tern colonies on Caroline Atoll. 27-28 September 1988. 

Calculated 

Population Nest Approx. Weeks 

Area (Mean Pairs ± SE) Stage From Laying 



44,100 m : 127,449 ±30.429 hatching eggs, 4-5 

downy chicks 

24,200m 2 41,382 ± 5,808 chicks with 7-10 

short tails, 
juv. plumage 

6,400 nr 10.9441 1.536 " 7-10 



Bo"sun Biid Islet, 
old 
new 

Total 



3,375 m : 6,883 ± 1,575 fledglings 

3,375 m 2 1,538 ± 758 new eggs 

75,075 nr 188.196 ±40.106 



11-12 
1-9 



TABLE 7 

Density of white terns on occupied islets by islet group, Caroline 
Atoll. September 1988. 









# White 


Density 




# 


Vegetated 


tern 


(pairs/ 


Islet Group 


Islets 


Area (ha) 


Pairs 


1,000 nr) 


Nake 


1 


66.63 


1,094 


1.64 


Long 


1 


49.60 


751 


1.51 


South 


1 


86.10 


381 


0.43 


Windward Islets 


9 


36.09 


1,164 


3.23 


South Nake Islets 


6 


8.50 


122 


1.44 


Central Leeward Islets 


9 


33.56 


408 


1.22 


Southern Leeward Islets 


4 


6.47 


37 


0.57 



All Occupied Islets 



31 



286.88 



3,957 



1.38 



155 



TABLE 8 

Lizards collected on Caroline Atoll, 1965-1988. 
Specimens: 



Species 



Clapp & Sibley 1 97 1 a Present Study 



Mourning gecko 

Lepidodactylus lugubris 
Polynesian gecko 

Gehyra oceanica 
Snake-eyed skink 

Cryptoblepharus poecilopleurus 
Moth skink 

Lipinia noctua' 

Emoia impar 
Azure-tailed skink 
Emoia cyanura 



USNM 158355-57 
USNM 158353-54 

USNM 158358 



USNM 299773 
USNM 299772 

USNM 299768-70 

USNM 299771 



1 USNM 158358 has recently been reidentit'ied by R.I. Crombie as Lipinia 
noctua. not Emoia nigra, as reported in Clapp & Sibley (1971a). 



TABLE 9 



Species 



Comparative abundance of Caroline's breeding seabirds in the Line Group. 

Estimated Population Comparative Abundance in the Line Group 



Red-tailed tropicbird 


300 ' 


Masked booby 


400 


Brow n boobv 


40 


Red-looted boobv 


7.000 


Great frigatebird 


6.1(H) 


lesser frigatebird 


200+ 


Sooty tern 


912.000 


Brovwi noddv 


3.000 


Black noddy 


17.000 


Blue-gray nodd) 


<10 


White tern 


8.000 



Second largest population 
Fourth largest population 
Third largest population 
Third largest population 
Third largest population 

Third largest population 
Third largest population 
Largest population (largest mi Kiribati) 

Largest population (largest in Kiribati) 



' Based upon nest count in 1990. 



156 




BROWN BOOBY 
RED-TAILED TROPICBIRD 



WINDWARO 
ISLETS 




Fig. 1. Caroline Atoll, Republic of Kiribati, with newly-named islets. 
RED TAILED TROPICBIRD 1988 



22 - 
20 - 
18 - 



CAROLINE ATOLL 



Fig. 2. Distribution map of breeding red-tailed tropicbirds and brown 
boobies on Caroline Atoll. September 1988. In this and the following 
distribution maps, arrows indicate concentrations of breeding birds. 



a 

z 



0) 

< 

a. 

o 

z 



16 
14 
12 

10 




uveniles 



primaries 



scap 



downy 



eggs 



1.0 



0.5 



>- 
< 

Q 

a. 

LU 
0. 



CO 
LU 

X 

u 



o 
6 



160 



APRIL 



140 

MAY 



120 



100 80 


60 


40 


JUNE JULY 




AUG. 


NO. DAYS 







20 

SEPT. 



Fig. 3. Approximate laying dates for red-tailed tropicbird nests found on Caroline Atoll in September 1988. In this and the following similar figures the numbers 
of nests begun during a given time period (bars) were determined by tallying each nest into one of several age classes (bar labels): bar widths indicate 
length in days for each class. The dotted line connects mean number of surviving clutches begun per day for each class. For example, a juvenile found 
in September began its egg stage in the previous May or June. The number of days are counted backwards from field observations. 

157 



MASKED BOOBY 




50% of population 



CAROLINE ATOLL 



SCALE i .14 ooo 
100 O too 1000 

'■••■' I I l_ 





MASKED BOOBY 1988 


9) 






40 




~~"\ 




- 




/ 




i 










\ 




- 


/ 




\ 






32 


/ 


24 




/ 








„ 


,' *ggs 
_i 1 


z 

> 
< 
■J 16 

s 


/ 


rer 


iiges 


< 

Q. 

o . 


scapulars 
■ ' 








180 160 

APRIL 



140 120 100 80 60 40 

MAY JUNE JULY AUG. 



Fig. 4. Distribution map of breeding masked boobies on Caroline Atoll, 
September 1988. 



Fig. 5. Approximate laying dates for masked booby nests found on Caroline 
Atoll in September 1988. 



RED- FOOTED BOO 




ROLINE ATOLL 




Fig 'i Distribution map oi breeding red footed boobies on Caroline Atoll, Fig. 7. Approximate laying datesfoi red footed boob) nests foundonCaroline 
September 1988 Atoll in September 1988 



158 



■ GREAT FRIGATEBIR 
► LESSER FRIGATEBIRD 




Fig. 8. Distribution map of breeding great and lesser frigatebirds on Caroline 
Atoll, September 1988. 



50 



40 



30 



-i 20 

</> 

IT 

< 

a. 

6 10 

z 



juveniles 



primaries 



GREAT FRIGATEBIRD 1988 



scaps 



naked/ 
downy 



_l_ 




eggs 



220 200 180 160 

FEB. MARCH APRIL 



140 


120 


100 


MAY 




JUNE 


NO. DAYS 







80 60 

JULY 



40 

AUG. 



20 o 

SEPT. 



Fig. 9. Approximate laying dates for great frigatebird nests found on Caroline Atoll in September 1988. See Figure 3 for explanation. 

159 



15 



10 



o 

z 



< 



LESSER FRIGATEBIRD 1988 



u veni les 



primaries 



scaps 



down 




220- 
FEB 



200 

MARCH 



180 160 

APRIL 



140 

MAY 



120 100 

JUNE 



80 60 

JULY 



40 

AUG. 



20 

SEPT. 



NO DAYS 

Fig. 10. Approximate laying dates for lesser frigatebird nests found on Caroline Atoll in September l c 



9 MAR 89 




OMARS MAY 90 



'SEP 86. • 89 & 

AR 90. -* MAY 90 IN -31 

* r SEP 88. • FEB 89 4 MAY 90 



#MAR89,-^MAY89IN=3I 



• * "JUN 65, O MAY 90 
IN =31 



juveniles 
N number of colonies 



BO' SUN BIRD 

• •* SEP 88 IN = 2I 



O AUG 894 MAY 90 
O AUG 894MAY 90 
-' : JUN 65 ,° AUG 89 IN 4 

CAROLINE ATOLL 



MO l«*0 1NO 

J I I I ' 



SOOTY TERN 1988 



BO SUN 
BIRD 




LONG IS 

"a" 



L2 



Fig. 1 1. Distribution mapofbreeding sooty ternsonCaroline Atoll, September Fig. 12. Approximate lasing dales for sooty tern young found on Caroline 
1988-July 1990.. Atoll in September 1988. See Figure 3 for explanation. 



160 



BROWN NODDY 




BLACK NODDY 



CAROLINE ATOLL 




CAROLINE ATOLL 



Fig. 13. Distribution map of breeding brown noddies on Caroline Atoll 
September 1988. 




Fig. 14. Distribution map of breeding black noddies on Caroline Atoll, 
September 1988. 



LONG-TAILED CUCKOO 




Fig. 15. Distribution mapol breeding white terns on Caroline Atoll. September Fig. 16. Tentative distribution map of the long-tailed cuckoo on Caroline 
1988. Atoll. The species most likely utilizes all well-wooded islets 



161 



SEABIRD BREEDING 
SPECIES DIVERSITY 




COCONUT CRAB 



CAROLINE ATOLL 





5C»l€ 






BO 


900 

1 


1 1 


,000 




CAROLINE ATOLL 



Fig. 17. Seabird breeding species diversity by islet. Caroline Atoll. 



Fig. 18. Distribution map of coconut crabs on Caroline Atoll. 



162 




PI. 1 . Incubating red-tailed tropicbird, Bo'sun Bird Islet, Caroline Atoll, 25 September 1 988. The nest scrape is in fine coral rubble under 
a Toumefortia shrub. 




PI. 2. Masked booby adult with egg on coarse coral rubble substrate with Portulaca mat. Nake Island. Caroline Atoll, 
26 September 1988. 



163 




PI. 3. White lern adult with egg in typical nest site, a dead Ttmrmforiia branch. South Island. Caroline Atoll. 23 September 1988. 



164 



1.3 First Records of the Long-tailed Cuckoo 
(Eudynamis taitensis) on Caroline Atoll, 
Southern Line Islands, Republic of Kiribati 

DAVID H. ELLIS . CAMERON B. KEPLER, ANGELA K. KEPLER*, and KATINO TEEB'AKI 
US Fish & Wildlife Service. Patuxent Wildlife Research Center, Laurel, Maryland, USA 

US Fish & Wildlife Service, Patuxent Wildlife Research Center, Southeast Research Station, Athens, Georgia, USA 
Wildlife Conservation Unit, Christinas Island, Republic of Kiribati 



Introduction 

The long-tailed cuckoo {Eudynamis taitensis) performs 
what is perhaps the most remarkable overwater migration of 
any land bird (Lack. 1959). It breeds in New Zealand and is 
known to winter in the islands of the central Pacific Ocean, with 
stragglers seen as far as Palau to the west and Pitcaim to the east 
(Bogert. 1937). Although the Line Islands are along the 
northeastern perimeter of this range, the cuckoo has never been 
reported for the Line or Phoenix Islands (Pratt etai, 1987). In 
1883. when Dixon (1884) visited Caroline Atoll (10°S. 150°W) 
at the southeastern end of the Line Islands, he reported that a 
colleague had heard "the notes of a singing bird." but no land 
bird was collected. In June 1965, biologists from the Pacific 
Ocean Biological Survey Program visited Caroline Atoll but 
failed to detect land birds (Clapp & Sibley. 1971 ). 

We encountered the long-tailed cuckoo on four islets 
during our 22-29 September 1988 survey of all 39 islets of 
Caroline Atoll as part of a research team from the Soviet 
oceanographic research vessel Akademik Korolev. Harsh 
monosyllabic or disyllabic call notes, presumably of this species, 
were first heard on three islets: South. Pisonia. and Long (see 
Kepler et a/., Subchapter 1.1. this volume, for islet locations). 
Then, on 25 September, a single bird, probably of this species, 
was heard and briefly seen on South Island. On 26 September, 
an individual was positively identified (CBK; ca. 25 m; xlO 
binoculars) on Nake Island. 

In an attempt to capture this species, or any other 
undiscovered land bird, we operated mist nets at three locations 
on the atoll. In 14.5 net hours (daylight hours only; ATX 
4-shelf nets. 2.6 x 12 m; mesh size 36 mm) beneath a 10-15 m 
Cocos canopy on South Island, only a single bristle-thighed 
curlew (Numenius tahitiensis) was captured. On Long Island 
two nets along a Pisonia-Cocos interface (canopy at 6-8 m) 
were unsuccessful in 27.5 net hours. Finally, on 28 September 



we collected a male long-tailed cuckoo on Long (US National 
Museum No. 607191) in 1.5 net hours along a Pisonia- 
TournefortialCordia interface with a short canopy (4-6 m) 
where two of us ( AKK, KT) had heard and followed a cuckoo- 
sized land bird for about 20 minutes the day before. 

The following measurements of the specimen were taken 
immediately after collection: mass 125 g, length 411 mm 
(central rectrices still growing so measurement was to the tip of 
the worn rectrices adjacent to central rectrices). culmen 
25.4 mm, and wing chord 179 mm. Soft part colors within 
30 minutes of death, compared with Smithe's ( 1 975 ) color key, 
were ridge of bill. No. 219 Sepia; lateral margin of bill, No. 86 
Pale Neutral Gray; lower mandible, No. 53 Buff-yellow; iris. 
No. 124 Buff; foot pad, No. 153 Trogon Yellow; and dorsal 
surface of foot and tarsus. No. 150 Bunting Green. 

These records establish the long-tailed cuckoo as a 
winter visitor to the Line Islands. Since our 1988 visit, we 
learned from correspondence and personal discussions with 
the atoll's only human inhabitants and wardens, Ronald and 
Anne Falconer, that cuckoos occasionally occurred in 
April. 1989, near their dwelling on Motu Ana-Ana. the 
southernmost leeward islet. This was confirmed by AKK with 
further sightings on two subsequent trips to Caroline in March 
and May 1990 (Kepler. 1990). Our original observations at 
five widely scattered locations (the most distant were 9 km 
apart) suggested that several individuals were present on the 
atoll during our visit. Subsequent observations suggest that 
this species disperses regularly to Caroline Atoll and perhaps 
some others of the better-vegetated Line and Phoenix Islands 
as well. 

We are pleased to thank Mr. Harold J. O'Connor and Mr. Steve 
Kohl (US Fish & Wildlife Service) and Professor Alia V. Tsyban 
(Goskomgidromet. USSR ) for organizing the Third Joint US-USSR 
Bering & Chukchi Seas Expedition that made these observations 
possible. 



165 



1.4 A Study of the Benthic Communities of 

Caroline Atoll (Line Islands, Pacific Ocean) 



BORIS I. SIRENKO and VLADIMIR M. KOLTUN 

Zoological Institute, USSR Academy of Sciences, Leningrad, USSR 



Introduction 

The study of coral atolls, which constitute one of the most 
highly productive biological systems in oligotrophic tropical 
waters, is of considerable theoretical and practical value. On 
one hand, rapid human population growth over the past few 
decades, with the concomitant shortage of protein-rich foods, 
has driven a never-ending search for new protein sources. The 
study of coral reefs as highly productive biological systems 
would be helpful for establishing marine farming facilities. On 
the other hand, burgeoning industrialization and the increasingly 
intensive use of all natural systems has caused considerable 
damage to coral reefs (Gomez & Yap, 1985). There is, 
therefore, an urgent need for coral reef monitoring. Uninhabited 
Caroline Atoll, situated far from principal sea routes, would be 
considered an excellent monitoring site. With this in mind, we 
spent a week studying the status of the coral reefs and surveying 
the benthic communities around South Island and the Southern 
Leeward Islands ( Kepler e t al. , Subchapter 1 . 1 , this volume ) in 
Caroline Atoll (Fig. 1 ). Regrettably, time constraints did not 
permit a study of sufficient scope and depth. The present paper 
is therefore limited to a general description of benthos 
distribution in the accessible portions of the reef and to an 
account of a uniquely interesting reef situated within the atoll 
lagoon. 

At low tide, the Caroline Atoll Lagoon is linked to the 
ocean by shallow passes of no greater than 0.5 m in depth. 
Several narrow intralagoonal reefs subdivide the central lagoon 
at low tide. The depth of the lagoon does not exceed 10 m. The 
largely sandy bottom includes isolated patches of fragile coral 
colonies, mostly Acropora. 

With few exceptions, the outer side of the reef near the 
southern islands has the classical structure of most coral reefs 
(Preobrazhenskiy, 1986). The narrow shallow-water lagoon 
facing the islands gradually becomes a reef flat cemented by 
encrusted calcareous algae. This is rather extensive, averaging 
562 m: range 396-759, N = 100 sites (Kepler et al.. 
Subchapter 1.1, this volume). The reef Oat is surmounted by 
widely spread limestone coral knolls or "coral heads." Further 
seaward, the reef flat, with the usual channels and overhanging 
ledges, breaks off. At a depth of 5-6 m it becomes a buttress 
zone (i.e., a radically crosscut sloping terrace consisting of 
individual spurs or benches that become narrower in the 
oceanward direction I. Ii is the latter zone that constitutes the 
main portion of the reef, with its abundant growth of corals 
(Acropora, Pocillipora, etc. ). It is this growth that accounts for 
the origin and continued development of the reef and the entire 
coral atoll. It was the status of the coral settlements in the 




£:&) 



3 / ^ / 

4 / rX-* ' Acropora-Tridacna Reef 

5 (QA * 




Fig. 1. Location of the Acropora-Tridacna reef and of hydrobiological 
section I-I of Caroline Atoll. 

buttress zone that afforded an indication of the "health" of the 
reef. Our studies showed that not less than 50^ of the surface 
area of the outer slope of the Caroline Atoll reef was covered 
by living coral colonies. Assessment on this basis, using the 
method of cross sections and areas (Gomez & Yap, 1985), 
showed the condition of the reef to be sound. The notorious 
"crown of thorns" starfish (Acanthaster planet), responsible 
for the devastation of reefs in other parts of the Pacific Ocean, 
was seen only once. 

The littoral zone of the lagoon facing South Island was 
sandy. The sand was filled with holes made by burrowing 
PolychaetaandBalanglossi. Also present, sometimes in clusters, 
were the mollusks Cerithium columna and Ccrithiuni sp., a few 
Calappa sp. crabs, and gastropod mollusksAfe/ar/iape undulata 
and Nerita plicata, the latter settling by water's edge on trees 
whose branches dip all the way down to the water. 



166 



The northeastern side of South Island was washed by 
strong oceanic surf throughout our stay at the atoll. The 
prevailing wind in this area is easterly (see Kepler et <//.. 
Subchapter 1.1. Appendix 2, this volume), an observation that 
is consistent with the presence on the eastern side of South 
Island of extensive coarsely fragmented coral limestone banks 
2-2.5 in high, attributable to windstorms. These storm banks 
contained large numbers of fresh bivalve mollusk (Asaphis cf. 
violascens) shells, providing evidence fora large population of 
these burrowing animals. The shallow islandward lagoon and 
reef flat, with its extensive sand lenses, is inhabited by the 
aforementioned bivalves, large numbers of gastropods of the 
genera Cerithium, Drupa, Cyprae, et cetera, as well as by 
predatory mollusks of the genus Conus, including unusually 
large specimens of Conns ebreus (up to 55 mm long). 

The coral reef on the western side of South Island received 
closer scrutiny. The distribution of organisms over the 
hydrobiological section (I-I) is indicated in Fig. 2. 

The beach of the island showed many large red hermit land 
crabs Coenobitaperlata, as well as Ocypode sp. crabs. Next to 
the beach lay a zone of lifeless coral limestone remarkable for 
its extreme lack of living organisms during the season of the 
year when we visited. This zone became very narrow to the 
south. The fact that the sea had until recently extended this far 
was evidenced only by small numbers of Melarhaphe undulata 
clinging to the underside of sun-baked coral slabs. It appears 
that when the wind direction changes to westerly, powerful surf 
inundates this area during high tide. During our visit, however, 
surf continued to crash against the opposite side of the atoll. 
The dry, lifeless zone ended with a ledge approximately 1 m in 
height. The water reached this ledge during high tides. The 
dominant fauna consisted of small hermit crabs, Grapsus sp. 
crabs, holothuria. and the mollusks Nerita plicata and Thais 



armigera. Further seaward, the stones in the never-drying 
pools of the islandward part of the lagoon harbored the following 
fauna: Diadema sp. urchins; the mollusks Cypraea moneta, 
Cerithium columna, Cerithium sp., Vasum tuhiferum. Conns 
ebreus, and Conus sponsalis; the crabs Eriphia sp. and Actaea 
sp.; three species of sea cucumbers (Holothuroidea); Linckia 
sp. starfishes; and black and grey brittle stars (Ophiuroidea). 
Part of the coral limestone in the same area was encrusted with 
the calcareous algae Porolithon sp. and Lithothamnion sp. To 
seaward, the littoral pools became deeperand somewhat larger, 
interdigitating to form the islandward lagoon. The bottom in 
this area consisted of coral limestone with small patches of 
coral fragments and sand. The bottom was strewn with 
limestone knolls that resulted from the intense buildup of 
calcareous algae and corals of the genera Porites, Pocillopora, 
Aeropora, and Montipora. These limestone knolls were 
cavernous and often caved in under the weight of an adult 
human. The fissures and caverns contained numerous small 
crabs and other crustaceans, as well as sponges. Also prevalent 
were algae bushes (Halimeda sp.). 

The bottom sloped upward towards the tidal strip, where 
it was succeeded by a highly tenacious, firmly attached fauna 
of a very particular kind. Unfortunately, because our 
examination was confined to the upper portion of the reef flat, 
the richly populated interior portion remained unstudied. The 
cavities and fissures of the top portion of the reef flat revealed 
an abundance of corals of the aforementioned genera. Dead 
coral colonies, as well as all of the old coral limestone formations, 
were densely encrusted with calcareous algae that cemented 
the reef-flat surface together. There were many gastropods 
(large Turbo argyrostomus and the smaller Drupa ricina, 
D. morum, D. grossularia, etc. ) as well as large orange-colored 
hermit crabs. The perpetually surf-washed tidal areas at the 




Q«fe f(3 





Corals: 




Crustaceans: 


(put) 


- Aeropora sp. 


& 


- Ocypode sp. 


<st& 


- Pocillopora sp. 


¥% 


- Grapsus sp. 


tMV\ 


- Montipora sp. 


** 


- Eriphia sp. 


C=0 


• Porites sp. 


$k 


- Aklaea sp. 






if 


- Coenobita periata 




Algae: 


A= 


- Paguridae sp. 


<***. 


- P'rolithion sp. 






MBSr< 


- Lithotamnion sp. 







Mollusks: 

- Malarhaphe undulata y^ 

© - Nerita plicata \U/a, 

• Cerithium spp. 

ff - Nassa sp. 

(2) - Turbo argyrostromus 

<Q - Vasum tuhiferum 

- Conus spp. 

(S - Cypraea spp. 

Q - Thais armisera, and Drupa spp. 



Echinodermata: 
■ Heterocentrotus sp. 



Diadema sp. 

"7a[ - Lichia sp. 

< y%, - Ophiuroidea 
^^ - Holothuroidea 




Fig. 2. Distribution of organisms on the reef on the western side of South Island of Caroline Atoll (Section I-I). 



167 



vers edge of the Hat exhibited magnificent specimens of the sea 
urchin (Heterocentrotus sp.), with spines as thick as cigars. 
The limestone caverns sheltered smaller crabs (Eriphia sp. and 
\i taea sp. ). The edge of the reef Hat dropped abruptly some 
5-6 m, in places forming an overhanging ledge that overlooked 
the most abundantly populated and significant portion of the 
reef, namely the buttress zone with its rich variety of coral 
species. 

Caroline's lagoon enclosed a unique natural structure, an 
unusual Acropora-Tridacna reef ( Fig. 1 . Chapter Frontispiece). 
This reef extended from the southernmost islet (Ana-Ana) in 
the Southern Leeward Islands into the interior of a shallow 
lagoon. In the middle of the lagoon, it divided into two 
branches: a northern branch that ended in the center of the 
lagoon, and an eastern branch that extended all the way to the 
opposite shore. The reef, in effect, partitions the lagoon into 
two parts; a wide channel had to be dug to permit a small flat- 
bottomed rubber boat to pass. The width of the reef varied from 
1 5 to 20 m, attaining 30 m at its fork. Most of the reef surface 
stood above sea level, with many of the Tridacna maxima and 
corals partially drying out during very low ebb tides. 

The distribution of organisms over a typical cross section 
of the reef is shown in Fig. 3. Five distinct zones (two edges, 
two lateral strips, and a central strip) were clearly evident. 

The edge strips, populated mostly by fragile colonies of 
Acropora secale, A. palmerae, and Acropora sp., are growth 
zones where vital activity keeps increasing the reefs width. 
The dominant corals of the genus Acropora completely cover 
the steep slopes of the reef as well as the neighboring rudimentary 
reefs that do not reach the surface of the water in the lagoon. 
The lush coral growth resembles a huge domed topiary. 



The thickness of the living coral layer ranges from 20 to 70 cm. 
Parts of the reef slope exhibited scarps that probably formed as 
a result of the collapse of fragile coral colonies unable to carry 
their own weight. It is at these scarps that measurements were 
made of the longest coral branches, some of which were found 
to attain a length of 70 cm. The spaces between the fine coral 
tentacles of the reef slope were filled with beaded ( moniliform ) 
algae (Halimeda sp.). Also evident at the base of the reef slope 
were a few, mostly large, specimens of Tridacna maxima. 
Beginning at a depth of 5 m and extending to the reef base were 
sparse, isolated growths of fungiform coral ( Fungia granulosa). 
The colonies of Acropora were fine and fairly fragile, as is 
typical in the still-water portions of many lagoons. It was risky 
to approach the edge of the reef, since the loose and brittle coral 
colonies tended to crumble underfoot. 

Beyond the edge zones (including the reef slopes, and 
extending several tens of centimeters into the reef flat) lie 
lateral zones where Tridacna maxima clams are especially 
abundant. These lateral zones have a width of 3-5 m on either 
side of the reef (Fig. 3). Particularly striking was the very high 
density of Tridacna. firmly attached to the reef surface by the 
byssus. which formed veritable bunches atypical for these 
large mollusks in other regions. Not uncommon were bunches 
of five or six clams attached to one another by the byssus. The 
average shell length for Tridacna found atop the reef was 
12-13 cm, with a maximum length of 1 9-20 cm for individual 
specimens. No less striking was the variety of coloration 
exhibited in their mantles, probably occasioned by the presence 
of symbiotic algae. We were able to identify as many as 
10 shades of blue, green, and light brown pigmentation. 
The average Tridacna population density in the lateral zone 



^ - coral 

(Acropora spp.) 

| - mollusks 

(Tridacna maxima) 
<X> - coral 

(Fungia granulata) 



«* - Holothurioidea 
(Ludwigothuria sp.) 

mm - algae 

(Porolithon sp.) 

$ - algae 

(Halimeda sp.) 







r 







edge i later; 



Coral limestone 



central 



lateral | edge 






ZONES 



i ross section of the Kcropora rrirfacna reef in the lagoon of Caroline Atoll. 

168 



was estimated at 35 living individuals/m 2 , with some 0.25-m 2 
patches containing as many as 20 of these mollusks. In addition 
to Tridacna clams, thu lateral strips included small colonies of 
Acropora sp. coral and some less abundant lamellate Montipora 
sp. Also present were algae {Halimeda sp. ) occupying spaces 
between Tridacna shells and coral colonies, as well as calcareous 
fouling algae encrusting the mollusk shells and dead portions 
of coral colonies. 

A central strip 7-10 m in width accounts for most of the 
reef flat. This site harbored most of the dead coral colonies and 
empty Tridacna shells, firmly bonded to the reef surface by 
calcareous fouling algae (Porolithon sp.). Nearly 809r of the 
surface area of the central strip was covered by these algae, 
which gave the middle portion of the reef added strength. The 
live Tridacna population density in this strip was one order of 
magnitude lower (an average of 4-5 individuals/m : ) than in the 
lateral strips. The strip also contained widely scattered colonies 
of Acropora sp. and, more commonly . bluish scales of Montipora 
sp. coral colonies, including small clumps of Halimeda sp. 
Another denizen of the central strip is the holothurian 
Ludwigothuria sp. (approximately 1 specimen/nr ). 

The central and lateral strips exhibited a few small holes 
4-8 cm in diameter surrounded by empty shells of smaller, 
thoroughly consumed Tridacna. These were probably the 
remains of meals taken by predators, namely small octopuses 
that are able to open bivalve shells without damaging them. 
The typical division of the reef into the aforementioned /ones 
is disrupted in some places. In these instances, the central strip 
of the reef contained a shallow depression of friable structure 
populated by corals and Tridacna, with small amounts of 
calcareous fouling algae. This is probably an intermediate 
stage in the merging of individual smaller reefs with the larger 
reef traversing the entire lagoon. One such smaller and still- 
growing reef is shown on the left side of the reef cross section 
in Fig. 3. 

It is interesting to note that the Acropora-Tridacna reef is 
a natural farm producing large bivalves of commercial value. 
The efficiency of the "farm" is difficult to assess without data 
on its productivity. However, a count of the Tridacna present 
is possible. According to the most conservative estimates, the 
surface of the Acropora- Tridacna reef, extending 1 km into the 



lagoon of Caroline Atoll, contains approximately 300,000 
Tridacna clams, the raw weight of their flesh equaling not less 
than 30 tons. It should be noted that the Acropora-Tridacna 
reef actually investigated was not the only one in the lagoon 
(Chapter Frontispiece: Kepler cial.. Subchapter 1.1. this volume. 
Figs. 47,48,57). There were, in fact, several such reels, and if 
we assume that they are of similar structure, the above figures 
can be multiplied by an appropriate factor to show the actual 
reserves of valuable food protein available in Caroline"s 
waters. 

The unusually high density of living Tridacna in Caroline's 
lagoon was especially striking, exceeding any previously known 
populations of both Tridacna maxima and T crocea (usually 
more abundant in other parts of the World Ocean). For 
example, the Palau Islands (western Pacific) were reported to 
have just six T. maxima and 153 T. crocea within an area of 
1 . 1 00 m- ( Hardy & Hardy, 1 969), whereas the same area on the 
Acropora-Tridacna reel in Caroline Atoll lagoon contained 
16,500 T. maxima. Richards ( 1 985 ) found that T. maxima in the 
Tuamotus numbered 6-20/nr at Takapoto Atoll, and up to 
60/rrr at Reao Atoll. Although these are the highest densities 
previously reported, they do not equal the numbers found in the 
densest patches on Caroline. 

The very considerable effect of such an enormous mass of 
large mollusks on the entire atoll is also noteworthy. This is 
because in symbiosis with zooxanthellae. which are the principal 
food of the Tridacna clam, the latter experience intensive 
growth and in turn enrich their habitat with proteins (Ricard& 
Salvat. 1977). Dataconcerning the natural Acropora-Tridacna 
reef could be put to use in creating artificial reefs in other parts 
of the ocean to achieve considerably enhanced productivity. 

The sound condition of the coral reef around Caroline 
Atoll, as well as the presence in its lagoon of a uniquely 
interesting natural feature in the formof the Acropora-Tridacna 
reef, may be deemed sufficient grounds for organizing a marine 
reserve in the area. 

The authors are grateful to US colleagues Kay and Cameron 
Kepler and D. Smith, participants in the 47th voyage of the research 
vessel Akademik Korolev, lor their active assistance in field work on 
Caroline Atoll. They also wish to thank Yu.N.Latypovforidentifying 
the corals. 



169 



Chapter 1 References 



Admiralty Chart, No. 979. (1965). Central Pacific Ocean: 

Islands between 160° East and 150° West Longitude 

J. D. Potter. London. 
Amerson, A. B„ Jr. & Shelton. P. S. ( 1 976). The natural history 

of Johnston Atoll, central Pacific Ocean. Atoll Res. Bull. 

192, 1-479. 
Arundel. J. T. ( 1 890). The Phoenix Group and other islands of 

the Pacific. New Zealand Herald, Auckland. July 5 and 12, 

1890. 
Ashmole, N. P. ( 1 963 ). The fairy tern Gygis alba on Ascension 

Island. Ibis 103b. 365-378. 
Ashmole, N. P. ( 1968). Breeding and molt in the white tern 

(Gygis alba) on Christmas Island, Pacific Ocean. Condor 

70. 35-55. 
Bavliss-Smith.T. P. (1973). Ecosystem and Economic System 

of Ontong Java Atoll, Solomon Islands. Cambridge University 

Ph. D. Thesis. 
Bennett. F. D. ( 1 840). Narrative of a whaling voyage round the 

globe from the year 1833-36, Vol. I. BibliothecaAustraliana 

46, 1-402. 
Bogert, C. ( 1937). Birds collected during the Whitney South 

Sea Expedition. XXXIV. The distribution and the migration 

of the long-tailed cuckoo (Urodynamis taitensis Sparrman). 

Am. Mus. Novit. 933, 1-12. 
Braley. R. D. (1987). Spatial distribution and population 

parameters of Tridacna gigas and T. derasa. Micronesica 

20, 225-246. 
Broughton. W. R. (1804). A Voyage of Discovery to the North 

Pacific Ocean. . . Performed in His Majesty's Sloop 

Providence. . . in 1795, 1796. 1797, 1798. T. Cadell and 

W. Davies. Rep.. London. 
Brown. W. C. ( 1956). The distribution of terrestrial reptiles in 

the islands of the Pacific Basin. Proc. 5th Pac. Sci. Conf. 3A. 
1479-1491. 
Bryan, E. H. ( 1942). American Polynesia and the Hawaiian 

Chain. Tongg Publ. Co.. Honolulu. 253 pp. 
Butler. A. G. & Strecker. H. (1884). Memorandum of the 

butterflies, etc. of Caroline Island. In Report of the Eclipse 

Expedition to Caroline Island May 1883. Natl. Acad. Sci. 

Mem. 2, 92-96. 
Catala, R. L. A. ( 1957). Report on the Gilbert Islands: Some 

aspects of human ecology. Atoll Res. Bull. 59, Oct. 3 1 . 
Christopherson, E. ( 1927). Vegetation of Pacific equatorial 

islands. B. P. Bishop Mus. Bull. 44. 79 pp. 
Clapp, R.B. (1967). Line Islands. Unpublished report. 32 pp. 
Clapp. R. B. & Sibley. F. C. ( 1967 ). New records of birds from 

the Phoenix and Line Islands. Ibis 109. 122-125. 
Clapp. R. B. & Sibley, F. C. ( 1968). Additional new records 

of birds from the Phoenix and Line Islands. Ibis 110, 
573-575. 
Clapp. R. B. & Sibley. F. C. ( 1971a). Notes on the vascular 
flora and terrestrial vertebrates of Caroline Atoll, Southern 
Line Islands. Atoll Res. Bull. 145. Washington, D.C.. 
Smithsonian Institution. 20 pp. 



Clapp, R. B. & Sibley. F.C. (1971b). The vascular flora and 
terrestrial vertebrates of Vostock Island, south-central 

Pacific. Atoll Res. Bull. 144. 1-10. 
Clark. L., Ricklefs, R. E. & Schreiber, R. W. ( 1983). Nest-site 

selection by the red-tailed tropicbird. Auk 100, 953-959. 
Clay, H. (1961). Narrative report of botanical fieldwork on 

Kure Island. 3 October 1959 to 9 October 1959. Atoll Res. 

Bull. 78. 1-4. 
Commercial Advertiser, The. ( 1873). From Caroline Island. 

Honolulu. Two-line note in newspaper, March 5, 1873. 
Crombie.R. I. (in prep. ). Herpetofauna of Oceania: Island lists 

and species distributions. 
Danielsson. B. ( 1954). Native topographical terms in Raroia, 

Tuamotus. Atoll Res. Bull. 32, 92-96. 
Darwin. C. R. ( 1 842 ). The Structure and Distribution of Coral 

Reefs. Smith. Elder & Co.. London, xii + 214 pp. 
Degener, O. & Gillaspy. E. (1955). Canton Island. South 

Pacific. Atoll Res. Bull. 41. Aug. 15. 
Diamond. A. W. (1971). The ecology of the sea birds of 

Aldabra. Philos. Trans. R. Soc. hand. B260. 561-571. 
Diamond. A. W. ( 1973). Notes on the breeding biology and 

behavior of the magnificent frigatebird. Condor 75. 

200-209. 
Diamond, A. W. (1975a). The biology of tropicbirds at 

Aldabra Atoll. Indian Ocean. Auk 92, 16-39. 
Diamond. A. W.( 1975b). Biology and behavior of frigatebirds 

Fregata spp. on Aldabra Atoll. Ibis 117, 302-323. 
Dinsmore, J. J. ( 1972). Sooty tern behavior. Bull. Fla. State 

Mus. Biol. Sci. 16(3). 129-179. 
Dixon. W. S. ( 1884). Notes on the zoology of Caroline Island. 

In Report of the Operations of the American Expedition 

to. . . Caroline Island May 1883. Natl. Acad. Sci. Mem. 2, 

90-92. 
Dorward, D. F. & Ashmole. N. P. ( 1963). Notes on the biology 

of the brown noddy (Anous stolidits) on Ascension Island. 

Ibis 103b. 447-457. 
Doty. M. S. & Morrison. J. P. E. ( 1954). Interrelationships of 

the organisms on Raroia aside from man. Atoll Res. Bull. 35. 

Nov. 30. 
Ellis, D. H., Kepler. C. B., Kepler. A. K. & Teeb'aki, K. (1990). 

Occurrence of the long-tailed cuckoo {Eudynamis taitensis) 

on Caroline Atoll. Kiribati. Emu 90. 202. 
Ellsworth. S. G. (ed. ) ( 1990). The Journals of Addison Pratt: 

Being a Narrative of Yankee Whaling in the Eighteen 

Twenties, A Mormon Mission to the Society Islands. . .. in 

the Eighteen Forties and Fifties. University of Utah Press, 

Salt Lake City, Utah. 606 pp. 
Ely.C. A. & Clapp. R. B. ( 1 973 ). The natural history of Laysan 

Island, northwestern Hawaiian Islands. Atoll Res. Bull. 171. 

xi + 361. 
Emory, K. P. (1947). Tuamotuan religious structures and 

ceremonies. B. P. Bishop Mus. Bull. 191, 102 pp. 
Findlay, A. G. (1884). Directory for the Navigation of the 

South Pacific Ocean. 5th Edition. London. 



171 



Fleet, R. R.( 1972). Nesting success of the red-tailed tropicbird 

on Kure Atoll. Auk 89. 651-659. 
Fleet. R. R. ( 1974). The red-tailed tropicbird on Kure Atoll. 

A. O. U. Monogr. 16, 1-64. 
losberg, F. R. ( 1936). Vegetation of Vostok Island, Central 

Pacific. B. P. Bishop Mas. Spec. Publ. 30, 19. 
Fosberg. F. R. ( 1949). Atoll vegetation and salinity. Pac. Sci. 

3. 89-92. 
Fosberg. F. R. (1953). Vegetation of Central Pacific atolls. 

Atoll Res. Bull. 23. 1-26. 
Fosberg, F. R. ( 1955). Northern Marshalls Expedition. 1951- 

1952. Land biota: Vascular plants. Atoll Res. Bull. 39, 1-22. 
Fosberg, F. R. (1956). Military Geography of the Northern 

Marshalls, prepared under the direction of the Chief 

of Engineers. US Army, and the US Geological Survey. 

320 pp. 
Fosberg, F. R. ( 1959). Vegetation and flora of Wake Island. 

Atoll Res. Bull. 67, 1-20. 
Fosberg, F. R. (ed. ) (1965). Man's Place in the Island 

Ecosystem: A Symposium. 10th Pacific Science Congress, 

Honolulu, Hawaii, 1961. Bishop Museum Press. 264 pp. 
Fosberg, F. R. (1968). Critical notes on Pacific island plants 2. 

Micronesica 4(2), 255-259. 
Fosberg, F. R. ( 1977a). Coral island vegetation. In Biology 

ami Geology of Coral Reefs, Vol. III. (O. A. Jones & 

R. Endean, eds.), pp. 256-277. Academic Press. 
Fosberg, F. R. ( 1977b). An irresponsible scientific expedition. 

Atoll Res. Bull. 219, 4-5. 
Fosberg. F. R. ( 1983). The human factor in the biogeography 

of oceanic islands. C. R. Soc. Biogeogr. 59(2), 147-190. 
Fosberg, F. R. ( 1985). Present state of knowledge of the floras 

and vegetation of emergent reef surfaces. Vol. 5. Proc. 5th 

Int. Coral ReefCongr., Tahiti. 1985. 
Fosberg. F.R.& Sachet, M.-H.l 1969). Wake Island vegetation 

and flora. 1961-1963. Atoll Res. Bull. 123, 1-15. 
Fosberg, F. R. & Sachet, M. -H. ( 1987). Flora of the Gilbert 

Islands, Kiribati, checklist. Atoll Res. Bull. 295, 1-33. 
Fosberg. F. R. & Sachet, M. -H. (MS). Flora of the Phoenix 

Islands. Unpublished manuscripts. Pt. 1, 85 pp. ; Pt. II. 52 pp. 
Fowler, H. W. (1901). Fishes from Caroline Island. Proc. 

Acad Nat. Sci. Phila. 53, 324-326. 
Frisbie, R. D. (1944). The Island of Desire. Garden City. New 

York. 234 pp. 
Garnett, M. C. (1983). A Management Plan for Nature 

Conservation in the Line and Phoenix Islands. Pt. I: 

Description, 318 pp.: Pts. II, III: Policy and 

Recommendations, 131 pp. Unpublished reports. 
Garnett, M. C. ( 1984). Conservation of seabirds in the South 

Pacific region: A review. ICBP Tech. Pub. 2. 547-557. 
Gibson Hill. C. A. ( 1950). Notes on the birds of the Cocos- 

Keeling Islands. Bull. Raffles Mus. 22. 212-270. 
Gill, R. F. (1990). Ecology of curlews in French Polynesia 

during spring. In Summary of the Proceedings from the 

Workshop on Bristle thighed Curlews. Anchorage, Alaska 

(R. F. Gill & C. M. Handel, eds). 22 pp. 
Gill.R. E& Redmond, R. L. ( in prep.). Distribution and habitat 

associations of bristle thighed curlews on Rangiroa Atoll. 

15 pp. 



Gomez, E. D. & Yap. H. T. ( 1 985 ). Setting up an international 

coral reef monitoring system. In Comprehensive Global 

Monitoring of the World Ocean (A. V. Tsyban & 

A. S. Sarkisyan, eds.). pp. 16-31. Proceedings of the First 

International Symposium. Vol. 2. Gidrometeoizdat 

Publishers. Leningrad, (in Russian) 
Guppy. H. B. (1906). Observations of a Naturalist in the 

Pacific Between 1896 and 1899. Vol. II. Plant-Dispersal. 

Macmillan & Co., New York. 627 pp. 
Hardy. J. T. & Hardy, S. A. ( 1969). Ecology of Tridacna m 

Palau. Pac. Sci. 23, 467-472. 
Hatheway, W. H. (1953). The land vegetation of Arno Atoll, 

Marshall Islands. Atoll Res. Bull. 16. April 30. 
Helfman. G. S. (No date). Ecological considerations for the 

management of the coconut crab. Birgus latro (L). 

Unpublished manuscript. 8 pp. 
Helfman. G. S. ( 1977a). Ecology and Behavior of the Coconut 

Crab, Birgus latro (L. >. M. S. Thesis. University of Hawaii, 

Parti. 109 pp. 
Helfman, G. S. (1977b). Temporal Distribution of Crabs. 

M. S. Thesis, University of Hawaii. Part II. 158 pp. 
Helfman. G. S. (1979). Coconut crabs and cannibalism. Nat. 

Hist. 88(9), 76-83. 
Holden, E. S. (1884). History, in Report of the Eclipse 

Expedition to Caroline Island May 1883. Natl. Acad. Sci. 

Mem. 2, 20-22. 
Holden, E. S. & Qualtrough, E. F. ( 1 884). Description of the 

island. In Report of the Eclipse Expedition to Caroline 

Island Max 1883. Natl. Acad. Sci. Mem. 2. 22-26. 
Hydrographer of the Navy. (1931. 1982). Pacific Islands 

Pilot. Vol. III. Ministry of Defense. Taunton. Somerset, 

England. 
Johnsgard. P. A. ( 1981 ). The Plovers. Sandpipers, and Snipes 

of the World. University of Nebraska Press. Lincoln. 

Nebraska. 
Kepler. A. K. ( 1990a). Caroline Atoll: Fact sheet, including 

conservation attributes and recommendations. Unpublished 

manuscript. 12 pp. 
Kepler, A. K.( 1990b). Flint Island: Fact sheet. Unpublished 

manuscript. 6 pp. 
Kepler, A. K.( 1990c). Vostok Island: Fact sheet. Unpublished 

manuscript. 6 pp. 
Kepler. A. K. (1990d). Report: ICBP 1990 Line and Phoenix 

Islands Expedition, Februarj 20-May31, 1990. Unpublished 

manuscript. 14 pp. 
Kepler, A. K. (in prep. ). Ecological studies of Vostok Island. 

Republic of Kiribati. 
Kepler. A. K.. Kepler. C. B. & Ellis. D. II. (1991). Ecological 

studies of Caroline Atoll. Southern Line Islands. Republic ol 

Kiribati. Part I. History, physiography, botany, and islet 

descriptions (Subchapter I. I , this volume). 
Kepler. C. B. (1967). Polynesian rat predation on Laysan 

albatrosses and other Pacific seabirds. Auk 84. 426-429. 
Kepler. C. B. (1969). Breeding biology of the blue-faced 

booby on Green Island. Kure Atoll. Publ. Nuttall Ornithol. 

ClubS, 1-97. 
Kepler. C. B. (1978). Ihc breeding ecology of sea birds on 

Monito Island. Puerto Rico. Condor80, 72-87. 



172 



Kepler, C. B. & Kepler, A. K. (1978). The sea birds of Culebra 

and its adjacent islands, Puerto Rico. The Living Bird 16, 

21-50. 
Kepler. C. B. & Kepler, A. K. (1989). Trip report, Soviet- 
American oceanographic expedition to the central Pacific: 

Hilo. Hawaii to Singapore, 9 Sept.-31 Oct. 1988. Unpublished 

manuscript. 9 pp. 
King. W. B. ( 1973). Conservation status of birds of central 

Pacific islands. Wilson Bull. 85. 89-103. 
Lack, D. (1959). Migration across the sea. Ibis 101, 374-399. 
Lamberson, J. O. ( 1 987). Natural history of terrestrial vascular 

plants of Enewetak Atoll. In Vol. II. The Natural History 

ofEnewetak Atoll. Chapter 3, pp. 17-35. US Department 

of Energy. 348 pp. 
Lamoureaux, C. ( 1961 ). Botanical observations on leeward 

Hawaiian atolls. Atoll Res. Bull. 79. 1-10. 
Lucett. E. ( 185 1 ). Rovings in the Pacific from 1837 to 1849. 

with a Glance at California, Vol. II. By a merchant long 

resident at Tahiti. 2 Vols., Longman. Brown and Green. 

London, xi + 371 pp. 
Mac Arthur, R. H. & Wilson, E. O. ( 1967). The Theory of Island 

Biogeography. Princeton University Press. Princeton. New 

Jersey. 203 pp. 
Markham. Sir Clements. (1904). The Voyages of Pedro 

Fernandez de Quiros. 1595-1606, 2 Vols.. Hakluyt 

Society. Series II, Nos. 14-15. The Hakluyt Society. 

London. 
Marks, J. S.. Redmond, R. L., Hendricks, P., Clapp, R. B. & 

Gill. R. E. ( 1990). Notes on longevity and flightlessness in 

bristle-thighed curlews. Auk 107(4), 779-781. 
Maude, H. E. (ca. 1938). Caroline Island. Unpublished 

manuscript. 8 pp. 
Maude, H. E. (ca. 1942a). Caroline Island: Part L. Ms. from 

author, of unknown origin. 6 pp. (printed but not published). 
Maude, H. E. (ca. 1942b). Flint Island: Part N. Author's 

manuscript., of unknown origin. 4 pp. (printed but not 

published). 
Maude, H. E. (1953). The British Central Pacific Islands: A 

report on land classification and utilization, Proc. 7th Pac. 

Sci. Congr. 60, 89-97. 
Maude, H. E. (1968). Of Islands and Men. Oxford University 

Press, London and New York. 397 pp. 
Maude, H. E. (No date). Bibliography of Books. Pamphlets. 

and Articles on the Phoenix and Line Islands. Ms., printed 

but not published, in author's possession. 4 pp. 
McKeown. S. (1978). Hawaiian Reptiles and Amphibians. 

The Oriental Publishing Company, Honolulu, Hawaii. 

80 pp. 
Moul, E. T. (1957). Preliminary report on the flora of Onotoa 

Atoll. Gilbert Islands. Atoll Res. Bull. 57. 1-48. 
Mueller-Dombois, D.. Bridges, K. W. & Carson. H. L. ( 198 1 ). 

Island Ecosystems: Biological Organization in Selected 

Hawaiian Communities. Hutchinson Ross Publishing 

Company, Stroudsburg. Pennsylvania. 583 pp. 
Nelson, J. B. ( 1969). The breeding ecology of the red-footed 

booby in the Galapagos. J. Anim. Ecol. 38, 181-198. 
Nelson, J. B. (1976). The breeding biology of frigatebirds — a 

comparative review. The Living Bird 14, 1 13-155. 



Nelson, J. B. (1978). The Sulidae: Gannets and Boobies. 

Oxford University Press. Oxford, England. 1,012 pp. 
Newell, N. D. (1956). Geological reconnaissance of Raroia 

( Kon Tiki ) Atoll, Tuamotu Archipelago. Bull. Am. Mus. Nat. 

Hist. 109.311-372. 
Nicholson, E. M. & Douglas, G. (1969). Draft Check List of 

Pacific Oceanic Islands. Micronesica 5(2), 327^63. 
N. I. D. (Great Britain Naval Intelligence Division). (1943). 

Pacific Islands, Vol. II, Eastern Pacific. H. M. Stationery 

Office. London, xvi + 739 pp. 
Niering. W. A. (1956). Bioecology of Kapingamarangi Atoll, 

Caroline Islands: Terrestrial aspects. Atoll Res. Bull. 49. 

1-32. 
Niering, W. A. ( 1 963). Terrestrial ecology of Kapingamarangi 

Atoll, Caroline Islands. Ecol. Monogr. 33, 131-160. 
Norman, F. I. (1975). The Murine rodents Rattits rattus, 

exulans, and non'egicus as avian predators. Atoll Res. Bull. 

182, 1-13. 
Parham, B. E. V. ( 197 1 ). The vegetation of the Tokelau Islands 

with special reference to the plants of Nukunono Atoll. N. Z. 

J. Bot. 9. 576-609. 
Paulding, H. ( 1 83 1 ). Journal of a Cruise of the United States 

Schooner Dolphin. ... G. & C. & H. Carvill, New York. 

iv + 258 pp. 
Perry, R. (1980). Wildlife conservation in the Line Islands, 

Republic of Kiribati (formerly Gilbert Islands). Environ. 

Consen: 7,311-318. 
Perry, R. & Garnett, M. (No date). The Natural History and 

Birdlife of Christmas Island. ICBP, Cambridge, England. 

36 pp. 
Phillips, N.J. ( 1984). Spatial relationships of Pisonia grandis 

and Cocos nucifera. Cousin Island Research Station (ICBP. 

Seychelles Islands. Indian Ocean) Technical Report. 9 pp. 
Phillips, N. J. & Phillips, V. E. (1990). Monitoring vegetation 

changes on Cousin — A comparative summary of changes 

since 1976. Cousin Island Research Station (ICBP. Seychelles 

Islands, Indian Ocean) Technical Report. 43 pp. 
Pilsbry, H. A. & Vanatta, E. G. (1905a). Mollusca of Flint and 

Caroline Islands, in the central Pacific. Proc. Acad. Nat. Sci. 

Phi la.. 57,291-293. 
Pilsbry, H. A. & Vanatta, E. G. (1905b). On some Pacific 

Cerithiidae. Proc. Acad. Nat. Sci.. Phila., 57. 787-789. 
Pratt, H. D., Bruner, P. L. & Berrett, D. G. (1987). A Field 

Guide to the Birds of Hawaii and the Tropical Pacific. 

Princeton University Press, Princeton, New Jersey. 
Preobrazhenskiy. B. V. (1986). Modern Reefs. Nauka Press, 

243 pp. (in Russian) 
Rauzon, M. J., Harrison, C. S. & Clapp, R. B. (1984). Breeding 

biology of the blue-gray noddy. J. Field Ornithol. 55. 

309-321. 
Reese, E. S. ( 1965). The ecology of the coconut crab. Birgus 

latro(L. ). Bull. Ecol. Soc.America46(4). 191-192. Abstract. 
Reese, E. S. ( 1987). Terrestrial environments and ecology of 

Enewetak Atoll. In The Natural Histoiy of Enewetak Atoll, 

Vol.1., pp. 187-202. 
Reyne. A. ( 1939). On the food habits of the coconut crab 

(Birgus latro L. ), with notes on its distribution. Arch. Neer. 

I. Zool. 3. 283-320. (in French) 



173 



Ricard, M. & Salvat, B. (1977). Faeces of Tridacna maxima 

( Mollusca-Bi val via), composition and coral reef importance. 

In Proceedings: Third International Coral Reef Symposium, 

Rosenstiel School of Marine and Atmospheric Science. 

University of Miami. Florida. 
Richard. G. (1985). Richness of the Great Sessile Bivalves in 

Takapoto Lagoon. 5th Int. Coral Reef Congress, French 

Polynesia, pp. 368-37 1 . 
Ridley. H. N. ( 19301. The Dispersal of Plants Throughout the 

World. L. Reeve & Co.. Ltd.. Ashford, United Kingdom. 

744 pp. 
RNZAF. (1986). Aerial photographs of Caroline Island 

(Gilberts), No. 6569. taken by the Royal New Zealand Air 

Force Plan No. 1036 9d-h, Department of Survey and Land 

Information, Wellington, New Zealand. 
Sachet, M. -H. ( 1 967 ). Coral Islands as ecological laboratories. 

Micronesica 3. 45—49. 
Sachet, M. -H. ( 1983 ). Takapoto Atoll. Tuamotu Archipelago: 

Terrestrial vegetation and flora. Atoll Res. Bull. 277, 1^46. 
Sachet. M. -H. & Fosberg. F. R. (1983). An ecological 

reconnaissance of Tetiaroa Atoll, Society Islands. Atoll Res. 

Bull. 275. 1-28. 
St. John, H. ( 1973). List and summary of the flowering plants 

in the Hawaiian islands. Pac. Trap. Bot. Garden Mem. 1 
519 pp. 
St. John, H. & Fosberg, F. R. ( 1 937 ). Vegetation of Flint Island, 

Central Pacific. Bernice P. Bishop Museum, Occas. Pap.. 

Honolulu. Hawaii. 12(24). 3—1. 
St. John. H.&Philipson,W.R.( I960). List of the flora of Oeno 

Atoll. Tuamotu Archipelago, south-central Pacific Ocean. 

Trans. R. Soc. N.Z. 88(3). 401-403. 
Schreiber, R. W. & Ashmole. N. P. ( 1970). Seabird breeding 

seasons on Christmas Island. Pacific Ocean. Ibis 112. 

363-344. 
Seelye. C. J. ( 1950). Rainfall and its variability over the central 

and south-western Pacific. N.Z. J. Sci. Tech. [B] 32(2). 
11-24. 
Shaw. H. K.A.( 1952). On the distribution of Pisonia grandis 

R. Br. (Nyctaginaceae), with special reference to Malaysia. 

Ken Bull., pp. 87-97. 
Sibley. F. C. & Clapp. R. B. (1967). Distribution and dispersal 

of central Pacific lesser frigatebirds, Fregataariel. Ibis 109. 
528-337. 
Smith. G. J. & Volodkovich. V. 1.. ( 1988). Preliminary report 

on marine plastics: Methods of collection and quantification. 
Unpublished manuscript. 31 October 19SS. in Soviet- 



American Joint Expedition III {1988), Appendix 



pp. 



Smithe. F. B. (1975). Naturalist's Color Guide. American 
Museum of Natural History. New York. New York. 

Spicer, R. A. & Newbery. D. McC. (1979). The terrestrial 
vegetation of an Indian Ocean coral island: Wilingili, Addu 
Atoll, Maldive Islands. [.Transect Analysis of the Vegetation. 
Atoll Res. Bull. 231. 1-21. 

Stevens, H. N. & Barwick, (i. F. I 1930). New light on the 
discover) of Australia as revealed In the journal of captain 
Don Diego de Prado y Tovai. Hakluyt Society. Series II. 
No. 64. The Hakluyt Society, London. 



Stoddart.D. R. ( 1976). Scientific Importance andConservation 

of Central Pacific Islands. Unpublished manuscript. 28 pp. 
Stoddart. D. R. & Fosberg. F. R. (1972). Reef islands of 

Rarotonga with list of vascular plants. Atoll Res. Bull. 160. 

1-22. 
Stoddart. D. R. & Gibbs. P. E. (1975). Almost-atoll of 

Aitutaki: Reef studies in the Cook Islands, South Pacific. 

Atoll Res. Bull. 190, 1-158. 
Stoddart. D. R. & Sachet. M. -H. (1969). Reconnaissance 

geomorphology of Rangiroa Atoll. Tuamotu Archipelago. 

with list of vascular flora of Rangiroa. Atoll Res. Bull. 125. 

1-44. 
Stone, B. C. ( 1968). Notes on Pandanus in the Line Islands. 

Micronesica 4( 1 ). 85-93. 
Stone, E. L. ( 1953). Summary of information on atoll soils. 

Atoll Res. Bull. 23. 1-6. 
Stonehouse. B. ( 1962). The tropic birds (genus Phaethori) of 

Ascension Island. Ibis 103b. 124-161. 
Taylor. R. C. (1973). An atlas of Pacific Islands rainfall. 

Hawaii Inst. Geophys., Data Report 25. 
Teeb' aki. K. ( 1 988 ). Report of visit to Caroline Atoll. Southern 

Line Islands. Wildlife Conservation Unit, Christmas Island. 

Republic of Kiribati. Unpublished manuscript. 12 pp. 
Thaman,R.R.(1987). Plantsof Kiribati: A listing and analysis 

of vernacular names. Atoll Res. Bull. 296. 42 pp. 
Thomas, W. L. (1961 ). The variety of physical environments 

among Pacific Islands. In Man's Place in the Island 

Ecosystem (F. R. Fosberg. ed.), 3rd Edition, 1970. 264 pp. 

Bishop Museum Press, 1965. 
Tracey, J. I.. Jr.. Cloud. P. E. & Emery. K. O. (1955). 

Conspicuous features of organic reefs. Atoll Res. Bull. 46. 

National Research Council. Washington. D.C. 
Trelease. W. (1884). Botany of Caroline Island. In Report of 

the Eclipse Expedition to Caroline Island May ISS3. Natl. 

Acad. Sci. Mem. 2. 87-90. 
Tsyban. A. V. & Smith. G. J. (1988). Protocol of the Third 

Soviet-American Joint Expedition. Pan II. The Central 

Pacific Ocean and South China Sea. aboard the R/V Akademik 

Korolev, 9 September - 31 October. 1988. Unpublished 

report. 6 pp. 
Upton. W. ( 1 884). Meteorology of Caroline Island. In Report 

of the Eclipse Expedition to Caroline Island May 1883, 

pp. 41-86. 
US Dept. of Energy. (1987). The Natural History ofEnewetak 

Atoll. Vols. I and II. 228 and 348 pp.. respectively. Office of 

Scientific and Technical Information. US Deartment of 

Energy. 
Wagner. W. L. Herbst. D. R. & Sohmer. S. H. ( 1990). Manual 

of the Flowering Plants qfHawai 'i. Vols. I and II. University 

of Hawaii Press and Bishop Museum Press. 1. 853 pp. 
Wester. L. (1985). Checklist of the vascular plants of the 

northern Line Islands. Atoll Res. Bull. 287. 38 pp. 
Whitehead. I). R. & Jones. C. E. (1969). Small islands and the 

equilibrium theory of insular biogeographv. Evolution 23. 
1 7 1 - 1 79. Lancaster. Pennsylvania. 
Wiens. H.J. (1956). The Geography of Kapingamarangi Atoll 
in the Eastern Carolines. Atoll Res. Bull. 48. 



174 



Wiens, H. J. (1962). Atoll Environment and Ecology. Yale Young, C. A. ( 1884). Introduction. In Report of the Eclipse 

University Press, New Haven and London. 532 pp. Expedition to Caroline Island May Iti83. Natl. Acad. Sri. 

Wilkes. C. USN (1845). Narrative of the United States Mem. 2. 11-20. 

Exploring Expedition during the years 1838, 1839, 1840, Young. J. L. (ca. 1922). Notes on Flint, Caroline and Vostock 

1841. 1842. Phila, Lea & Blanchard. 5 Vols, and an atlas. Islands. In Memoranda re Tahitian Business. S.R.Maxwell 

Yaldwyn, J. C. & Wodzicki. K. (1979). Systematics and & Co.. Ltd., Pacific Manuscripts Bureau, Research School of 

ecology of the land crabs (Decapoda: Coenobitidae, Pacific Studies. Australian National University. Canberra. 

GrapsidaeandGecarcinidaelof the Tokelau Islands, central A. C. T. PMB Manuscript 23. 1-18. 
Pacific. Atoll Res. Bull. 235. 1-53. 



175 



Chapter 2: 

INVESTIGATIONS AND 
ANTHROPOGENIC ECOLOGY 



Editors: 

SERGEI M. CHERNYAK, 

CLIFFORD P. RICE, & 

GREGORY J. SMITH 



2.1 Distribution of Chlorinated Hydrocarbons 
in Ecosystems of the Equatorial Pacific 

SERGEI M. CHERNYAK and VALERIYA M. VRONSKAYA 

Institute of Global Climate and Ecology, State Committee for Hydrometeorology and Academy of Sciences, Moscow, USSR 



Introduction 

It is a well-known fact that chlorinated hydrocarbon 
pesticides (CHP's), which earned their developers a Nobel 
Prize, have been both a great benefit to mankind by making it 
possible to rescue up to one-half of the world's cereal-grain 
harvests from pests and a detriment to the environment by their 
trait of extreme stability causing them to build up in the 
environment. In just a few years, their environmental hazard 
has become apparent. 

The multiplicity of pathways and the high rates of CHP 
transport led to a situation whereby in the early 1 970' s they had 
accumulated on a worldwide scale, being detectable in virtually 
all environments, including mountain-peak glaciers and deep- 
ocean depressions. Later, the world was to see a no less 
extensive buildup of other classes of chlorinated hydrocarbons 
(CH's). namely polychlorinatedbiphenyls(PCB's), chlorinated 
terpenes, dioxins, benzofurans. et cetera, which were being 
used not only in agriculture but also in various industries and 
even in health care fields. 

Despite the fact that most industrialized countries had 
invoked total or partial bans upon the use of CH's in open-cycle 
processes, their production on a global scale remains almost 
undiminished, since no effective alternative for protecting 
agricultural harvests in tropical countries has yet been devised. 

There is now a considerable body of scientific literature 
indicating that the World Ocean is the ultimate repository for 
CH's. According to the most reliable estimates (Tanabe. 
1 985 ), ocean waters are today the repository of as much as 70% 
of all the CH's ever released into the environment. Previous 
studies (Chemyak et al., 1985b) showed that the rates of CH 
buildup and migration in constituents of the oceanic environment 
depend first and foremost on the physicochemical and 
hydrologic-geographic characteristics of various portions of 
the World Ocean. Hence, predictive estimates of increases in 
ocean pollution, which are bound to occur globally, require 
information on the forms and amounts of CH's present in 
various media and especially in little-studied regions such as 
the Pacific Ocean, where only a few expeditions have concerned 
themselves with CH's (Izrael & Tsyban, 1989). Our own 
recent studies (Chernyak&Mikhaleva. 1985a) were the first to 
include nearly in situ studies of microbial and photochemical 
PCB decomposition processes, which are the only processes 
presently at work to rid the oceanic environment of these 
xenobiotic substances. We were particularly interested in 
investigating PCB distribution at Caroline Atoll, a coral island 
remarkable for its diverse flora and fauna, where these CH's 
have accumulated, in markedly altered form and at a distance 



of many thousands of kilometers from their sources. Their 
presence at these locations are probably occurring through 
atmospheric and oceanic transport. 

The present paper sets out the findings and conclusions of 
comprehensive studies conducted in 1988 in the equatorial 
Pacific and at Caroline Atoll, which were obtained during the 
First Joint US-USSR Central Pacific Expedition aboard the 
R/V Akademik Korolev. 

Materials and Methods 

The locations of the sampling stations are shown in Fig. 1 . 
The media sampled in order to establish the specifics of 
hydrochemical processes involving chlorinated hydrocarbons 
in background regions of the Pacific and coral-reef ecosystems 
were seawater, sediment, plankton, neuston, demersal 
organisms, and fishes. Specific studies conducted at Caroline 
Atoll also included corals, flora and fauna, and eggs of local 
bird species. 



100° 


110° 120° 130° 140° 


150° 160° 170° 180° 


170° 160° 150° 






' HJ 






40 




y4& 


Pacific 




30' 




J<$ 


Ocean 


Hawaiian 
Islands ^^ 


20' 




\ South A 
S China V> t 










/Sea ^'jj&iS'^S 8 ™-*- 


V 




10 




%.-'/j 


\ 


Christmas 






*Y /£* *> r-, 


"""■•«,, 


Island v 


0° 




Vjfoic=£2C-\ 


<b 








"V__ G a ^^^\_ 


ip <a_ 






^^s 3 *? Q ^r 




Caroline 
Aloll 


10 






\ ^ 




20 



Fig. 1 . Expedition route and station locations in the Pacific Ocean and South 
China Sea (9 September to 3! October 1988). 

Seawater samples, 100 I in volume, were filtered through 
XAD-2 resin at the rate of 20 1/h. The sorbed CH's were eluted 
with 80 ml of ethanol mixed with an equal volume of 2% 
sodium sulfate solution. The water-alcohol solution was 
doubly extracted with n-hexane ( using 25 ml of hexane in each 
of the two extractions). The extracts were concentrated to a 
volume of 4-5 ml using a rotary evaporator, purified with 
concentrated sulfuric acid, neutralized with a 5% NaHCO, 
solution, rinsed twice, dried over sodium sulfate, and 
concentrated in a stream of pure nitrogen to a volume of 1 ml. 
The concentrate was injected into a Hewlett-Packard gas 
chromatograph with the aid of an autosampler. The 
chromatographic analysis conditions were as follows: 



179 



fused-quartz capillary column, length 30 m, and inside diameter 
.•I 0.32 mm; chromatographic phase DB-1 (0.25 urn). The 
temperature program lor the column was initial temperature. 
1 20 C" ( I mini, rising to 250 C at the rate of 5°C/min. The 
chromatographic analysis time was 40 min. the injector 
temperature. 225°C, and the electron-capture detector 
temperature. 300°C. 

The sediment samples were centrifuged for 20 min at a 
speed of 2,000 rpm to achieve total deposition of the silt. The 
residue was extracted with acetone, then doubly extracted with 
a 3:1 hexane-acetone mixture. The combined extract was 
washed by mixing with an equal volume of 29? sodium sulfate 
solution. The hexane layer was separated off, and the aqueous- 
acetone solution reextracted with additional hexane. The 
combined hexane extract was concentrated, then purified, first 
with sulfuric acid (to remove extracted organic compounds), 
then with tetrabutylammonium sulfate (to remove any sulfur 
compounds). The purified solution was concentrated down to 
I ml in a stream of pure nitrogen and chromatographically 
analyzed. The biological samples were crushed to obtain a 
homogeneous mass, defatted with acetone, then treated, 
following the procedure used, with sediment samples. 

In order to investigate the effect of photochemical processes 
on the behavior of CH's in background ecosystems of the 
Pacific, experiments were conducted on the decomposition of 
a standard Aroclor 1232 solution in waters drawn from the 
equatorial Pacific and around Caroline Atoll under the action 
of sunlight. 

The experiment was run in two 5-1 reactor vessels, one 
exposed to sunlight, the other shielded using light-blocking 
foil. The surface area involved was 400 cm 2 . The sterilized 
seawater in the reactors were spiked with an acetone solution 
of Aroclor 1232 to yield PCB concentrations in the water of 
100 ng/1. The samples were extracted with n-hexane (twice. 
50 ml each time), then concentrated to a volume of 2 ml in a 
rotary evaporator. They were then purified with concentrated 
sulfuric acid and chromatographically analyzed. Microbial 
degradation of the PCB under the same conditions was 
investigated for control purposes. 

Results and Discussion 

Data on CH levels in Pacific Ocean waters are presented 
in Tables I and 2. 

Analysis of these results clearly demonstrated the 
dependence of the distribution of various PCB components on 
their molecular structure. 

The distribution of hcxachlorocvclohexanc (HCII) is 
noteworthy since its total concentration was fairly high, though 
it was still several times lower than in the Bering and Chukchi 
Seas (Chernyak ci al., 1002). which arc just as far removed 
from areas where this compound continues to be used. The 
composition of the IICH mixture (containing as much as 90' - 
of the (/.-isomer) indicates thai the sources of pollution are 
probably equatorial countries that employ \ast amounts of 
technical-grade hexachloran on their crops. The relatively low 



TABLE 1 

Chlorinated hydrocarbon levels in the surface waters 
of the Pacific (ng/1). 

Chlorinated Hydrocarbon Levels 
Station Total 
Number HCH DDT DDD DDE PCB 



Caroline Atol 












114 


0.02 


0.1 


0.1 


(1.1 


(1.05 


Caroline Atol 


. lagoon 










L-l 


0.01 


1.0 


0.5 


0.3 


0.02 


L-2 


0.01 


0.8 


0.3 


0.3 


0.01 


L-3 


0.01 


1.(1 


0.3 


0.5 


0.02 


L-4 


(1.(11 


1.1 


0.3 


0.3 


002 


Phoenix transect 










1 15 


2.7 


0.07 


0.02 


0.01 


0.2 


1 Id 


2.4 


0.05 


0.01 


0.01 


0.2 


1 17 


2.9 


0.07 


0.01 


0.01 


0.3 


1 IS 


2 "i 


O.OS 


0.02 


0.02 


0.2 


1 19 


2.3 


0.02 


0.01 


0.01 


0.4 


12(1 


2.5 


0.05 


0.02 


0.01 


0.3 


Marianas transect 










121 


3.0 


0.09 


0.02 


0.01 


0.5 


122 


2.3 


0.12 


0.02 


0.03 


0.3 


123 


2.4 


0.05 


0.01 


0.01 


0.1 


124 


2.4 


0.02 


0.01 


0.02 


0.3 


125 


2.5 


0.01 


0.01 


0.01 


0.4 


126 


2 2 


O.OS 


0.02 


0.01 


0.4 


South China 5 


ea 










127 


4.2 


0.18 


0.04 


0.02 


0.3 


I2S 


4.5 


OKI 


0.02 


0.03 


0.4 


124 


4.8 


0.10 


0.04 


0.05 


0.3 


130 


3.7 


0.22 


0.10 


0.03 


0.3 


131 


3.3 


0.IS 


DOS 


(Ids 


0.3 



TABLE 2 

Distribution of chlorinated hydrocarbons over the water column in 
Pacific Ocean waters. 







Chlorinated 




Depth 




llv drocarbon Leve 


s 


inn 


Total 








HCH 


DDT 


PCB 




Slain 


m 1 1 5 







2.7 


0.10 


0.2 


III 


2.5 


O.OS 


0.2 


100 


2.0 


0.05 


0.2 


1.000 


1.7 


001 


0.2 




Slain 


in 120 




II 


2.5 


oos 


0.3 


10 


2.3 


(KIS 


0.3 


100 


2.1 


0.04 


0.2 


1.000 


IS 


0.04 


0.2 



ISO 



hexachloran levels in the water here are likely due to their high 
volatility, which appears to be driving most of the HCH present 
in these hot climates into the atmospheric compartment. 

The rather unusual distribution of HCH within the water 
column — that is, the marked drop in its level that occurs at 
considerable depths — is explained by the relatively high 
solubility of this pollutant in seawater and by the fact that it is 
present largely in dissolved form (in contrast to other globally- 
occurring CH's. which are largely sorbed and precipitate 
together with suspended matter). Hexachlorocyclohexane 
isomers are practically the only common pollutants whose 
behaviorin the open ocean can be explained largely on the basis 
of hydrochemical factors. 

Also worthy of note is the fact that there was almost a 
twofold lower concentration level of HCH in the waters of the 
atoll lagoon versus the ocean water surrounding the lagoon. 
The only obvious explanation was the presence of a temperature 
gradient, which entailed differing rates of photochemical and 
microbial transformation of the cyclohexane ring, and the 
accelerated evaporation that was taking place in the lagoon. 

Another interesting finding was that the lagoon water of 
Caroline Atoll contained roughly one-half as much PCB as did 
the surrounding ocean water. Moreover, the composition of 
the CH's was significantly different: the lagoon water contained 
virtually no highly chlorinated PCB congeners, which was 
probably due to the higher rates of photochemical processes in 
the thoroughly heated shallow water of the lagoon. It should be 
noted, however, that as a general rule, PCB levels in the 
equatorial Pacific were only slightly lower than in the northern 
Pacific, even though sample composition turned out to be 
considerably different. 

Whereas most of the PCB' s in the Bering Sea consisted of 
di- and trichlorobiphenyls, the major constituents of PCB's 
in the equatorial waters were tri- and tetrachlorobiphenyls 
as well as heptachlorobiphenyls. Analysis of the CH's sorbed 
by suspended matter revealed a clear dependence of pollutant 
levels on latitude. For example, the content of HCH isomers in 
suspensions from the equatorial Pacific was almost 50 times 
lower than in the circumpolar parts of the ocean (Table 3 ). This 
may have been due to the significant shift in sorption-process 
equilibria associated with a 25°C rise in temperature. There 
was a marked (almost tenfold) change in PCB levels in 
suspended matter, whereas the levels in the water layer remained 
virtually constant. It is a curious fact that the equatorial Pacific 
is a unique region of the World Ocean where the PCB mixture 
appears to be equally apportioned between the suspended 
matter and the dissolved phase. 

Of special interest is the distribution of the extensively 
used pesticide DDT in Pacific Ocean ecosystems. Dichloro- 
dipheny ltrichloroethane levels in water samples from the Bering 
and Chukchi Seas have decreased considerably over the past 
decade due to restrictions on the use of this compound imposed 
by a number of industrialized countries. In fact, in some 
instances these levels come close to analytical zero (Chernyak 
et al.. 1989). In the equatorial Pacific, however, the DDT 



TABLE 3 

Chlorinated hydrocarbon levels (Ug/g dry weight) in suspended 
matter in the Pacific Ocean. 



Station 


Date 


Depth 


Chi 


urinated 


Hydrocai 


•bon 


No. 




(m) 


a-HCH 


Y-HCH 


DDT 


PCB 


113 


09/2 1 





54 


43 


135 


258 


Caroline 


09/23-29 





5 


5 


174 


31 1 


Atoll 














116 


10/02 


20 


5 


4 


164 


101 


116 


10/02 


100 


5 


5 


105 


126 


119 


10/04 


65 


5 


4 


192 


168 


119 


10/04 


100 


3 


3 


127 


171 


119 


10/11 





5 


3 


167 


315 


120 


10/11 





5 


5 


103 


153 


120 


10/11 





4 


4 


143 


93 


120 


10/12 





3 


4 


215 


139 


120 


10/13 





5 


4 


287 


189 


120 


10/14 





5 


4 


274 


63 


120 


10/15 





3 


3 


159 


76 


121 


10/16 





5 


3 


315 


154 


121 


10/17 





4 


3 


252 


127 


122 


10/18 





4 


4 


338 


129 


122 


10/18 


130 


3 


3 


309 


85 


123 







5 


3 


421 


1X1 


124 







5 


4 


357 


108 



hazard remains considerable: its levels average 0. 1 ng/1, which 
is typical of the areas most severely impacted by human 
activities, namely the North Atlantic and the Indian Ocean 
(Tanabe etal.. 1982). 

Results for the microbial and photochemical degradation 
of PCB's in the equatorial Pacific are presented in Figs. 2-4. 
The gradual loss of several of the PCB congeners are plotted 
over time in Fig. 2. On incubation with natural populations of 
microbes from these central Pacific waters, some congeners 
were reduced by more than 50% in 10 days (e.g., BZ#s. 
Ballschmitter and Zell numbering system for PCB congeners: 
Ballschmitter&Zell, 1980)7, 16. 49, 52, and 42. In Fig. 3, the 
degree of microbial degradation of the PCB's is organized by 
the PCB homologue group. From this presentation it is 
observed that the greater the degree of chlorine substitution of 
the bipheny 1 ring, the more resistant it is to microbial breakdown. 
Plotted for comparison are the relative rates of breakdown of 
the homologous groups for the two regions that were studied, 
the Bering Sea. and the central Pacific Ocean. Between these 
two locations, the rates of degradation of especially the mono- 
and dichlorobiphenyl homologues are faster in the warmer 
water of the Pacific. Figure 4 presents the times loss of total 
PCB's for the central Pacific based on a comparison between 
microbial degradation, photochemical degradation, and 
photochemical processes under the influence of added PAH's. 
It is apparent here that microbial degradation accounts for most 
of the breakdown in this area and that PAH's may have the 
capacity to inhibit photochemical breakdown of PCB's. 




Percentage ot initial PCB concentration 



1 2 3 4 5 6 7 



Fig.2 Photochemical breakdown of individual PCB component 1 . 
A 





Fig. 4. Breakdown of PCB's (dichlorobiphenyls) through microbial and 
photochemical action: 1) microbial; 2) photochemical: 3) photo- 
chemical in the presence of PAH's. 



Conclusions 

As a result of human activities, chlorinated hydrocarbons 
occur in all media constituting the Pacific ecosystems 
investigated. Levels of HCH isomers in the equatorial Pacific 
turned out to be lower than in polar areas. This is understandable, 
given the high volatility of this pollutant. 
Dichlorodiphenyltrichloroethane levels in certain parts of the 
Pacific (such as the Caroline Atoll) were found to be similar to 
those observed in impacted basins, although the concentration 
of this compound in the equatorial Pacific taken as a whole 
were not high. Polychlorinated biphenyls occurred in all the 
water samples studied. It is now clear that despite the slowness 
of photochemical and microbial breakdown of PCBs, these 
processes nevertheless play an important role in the self- 
purification of Pacific Ocean ecosystems. 



Fig. 3. Microbial breakdou n of PCB components in the waters of the Bering 
Set (A) and central Pacific Ocean (B). 



2.2 Distribution of Polycyclic Aromatic 
Hydrocarbons 



EHA R. URBAS. NATALYA I. IRHA, and UVE E. KIRSO 

Chemistry Institute of the Estonian Academy of Sciences, Tallinn, ESSR 



Introduction 

Three distinct regions (Caroline Atoll, tropical Pacific and 
South China Sea) during the First Joint US-USSR Central 
Pacific ( BERPAC) Expedition were sampled and analyzed for 
polycyclic aromatic hydrocarbons (PAH's). Methods for 
collection and analyses are described in Distribution of PAH's 
( Irha et al. , 1 992) in Results of the Third Joint US-USSR Bering 
& Chukchi Seas Expedition (BERPAC), Summer 1988 (Nagel, 
1992). The regions that were sampled are identified in the 
Frontispiece to this volume. 

Polychlorinated aromatic hydrocarbons are important 
natural and anthropogenic contaminants of marine ecosystems. 
These compounds are associated with petroleum pollution, 
including natural oil seepage, and with industrial contamination 
and many exhibit carcinogenic and mutagenic properties. In 
this paper, the presence and quantity of PAH's in both biotic 
and abiotic components of tropical Pacific ecosystems is 
reported. 

Results 

Caroline Atoll 

Determinations were made of the composition and 
distribution of PAH's (Table 1 ) present in water and marine- 
organism suspensions sampled in inshore waters off the atoll, 
as well as in the following media: bottom sand, island soils, 
corals, and siphonales algae. The findings (Table 2) indicated 
that the PAH pollution of this ecosystem was negligible. There 
were no PAH's in samples of surface sand from the atoll, and 
the soils taken from a palm tree forest area showed PAH levels 
below background. 

The total PAH content of coral samples did not exceed 
9.6 |ag/kg of dry weight. The following representatives of 
four- and five-ring PAH's were detected: benzo(e)pyrene 
(BeP), 62.5; indeno( l,2,3-cd)pyrene (IPy), 23; benzo(a)pyrene 
(BaP). 14.5 (in percent by weight). 

The benthic sand of the atoll lagoons contained four- and 
five-ring PAH's (Table 2); the major constituents (in percent 
by weight of the total amount of PAH's present) were 
benzo(b)fluoranthene (BbF), from 38.7 to 53.6%; and BaP, 
from 16.8 to 36.6% (Fig. 1). 

The surface waters off the atoll contained six four- and 
five-ring PAH's, whose total amount did not exceed 
1.55 ng/1, the dominant constituents being the five-ring 
carcinogens BbF and BaP (Fig. 2). Suspended matter sampled 







Table 1 






List of specific PAH's identified by displacemenl-eludonal liquid 


chromatog 


raphy 


Name 


Symbol 


Structural formula 


Carcinogenicity 
(Lee e/u/., 1981) 


Pyrene 


Py 


£9 







Chrysene 


Chr 


oS° 




+ 


Benz( a (anthracene 


BaA 


ocx9 




+ 


Benzol etpyrene 


BeP 


& 




0/+ 


Benzol blfluoranthene 


BbF 


^3 




++ 


Benzol k )fluoranthenc 
Benzo(a)pyrene 


BkF 
BaP 


cab 
cxA 




++ 
++ 


Benzo(g.h,i)perylene 


BPer 


cfi? 




+ 


Dibenzla.h (anthracene 


DBA 


€8 




+ 


Indenol 1,2.3-cdlpyrene 


IPy 


go9 




+ 


Note: 
Classification 




Svmbol 


Criterion: % of animals 

that de\clopcJ lesions 


noncarcinogenic 
weakly carcinogenic 
strongly carcinogenic 






+ 
++ 





33 
>31 



at a depth of 90 m included four kinds of four- and five-ring 
PAH's, whose total concentration was 0.68 ng/1, with the four- 
ring IPy and benz(a)anthracene (BaA) predominating (Fig. 2). 

The total PAH level in the gills of Mexican (northern red, 
Pensacola) snapper {Lutjanus campechanus) specimens was 
moderate, amounting to 0.08 |ig/kg; the corresponding value 
for muscle tissue was 1.13 Hg/kg of fresh weight. 

The following five-ring PAH's were identified as present 
in the gills (in percent by weight): BbF. 62.5; BaP, 33.5; 
benzo(k)fluoranfhene (BkF), 3.7; moreover, chrysene (Chr) 
and benzophenanthrene (Bp) were also present. The main 
constituents of the total PAH mix were four-ring compounds: 
BaA, 88.5%, and Chr, 1 1.5%; also present were traces of BbF 
and BkF. 

Polycylic aromatic hydrocarbon levels in various 
components of the Caroline Atoll ecosystems indicated that 
this area was only slightly polluted by these specific compounds. 



183 



TABLE 2 

PAH le\cls in particular media of the Caroline Atoll ecosystem 
(September 1988). 

Medium Units Py BaA Bnf BkF BaP Bep 

Cm- 



Water, surface ng/1 qualitatively tr. 



>5 0.90 0.21 ir. 



layer 




present 










Suspension 


ng/1 


0.6 


0.8 


tr. 


tr. 


- 


(90m) 














Benthic sand. 


u/kg 


- .. 


- 


- 


- 


- 


lagoon 














Lagoon I 




- 


- 


- 


- 


- 


Lagoon III 




- 


tr. 0.44 


0.08 


0.30 


- 


1 agoon V 




- 


0.60 0.60 


0.09 


0.26 


- 


Atoll soil 


u/kg 


- 


- 


- 


0.83 


- 


Aeropora corals 


u/kg 


2 2 


- 


- 


1.4 


6.6 


Siphonales algae 


u/kg 


- 


- 


- 


0.07 


- 


(Caulerpaceae) 














ir. - trace 















06 



2 



(b) 



n 



n 



BaA BbF BkF BaP 



BaA BbF BkF BaP 



Fig. I PAH levels lu/kgi in henlhic sand of lagoons V (a) and 111 (hi "I 
Caroline Atoll (September 1988) 
|1 egend:] A = PAH level, u/kg. 



1.0 



u 0.5 
X 

< 

a. 



1 




(a) 




(b) 



n 




n 



Py BaA BbF BkF BaP BeP 



Py BaA BkF BaP 



i l ontentol PAH's (ng/1) and their distribution (in percent bj weight) 

in the surface (0 0.5 m) layer ol watei (a) and inasuspension(b)drawn 
from a depth <>i 90 m ofl Caroline \toll (September 1988) 
[Legend:! A = I>AII level, ng/1 



Tropical Pacific 

Polycyclic aromaitc hydrocarbon levels in the surface 
w aters were not high. In the open ocean, the total concentrations 
ranged from 0.8 to 43.8 ng/1, with a maximum of 396.6 ng/1 in 
the Marianas sector. The composition and distribution of the 
PAH's identified in the surface waters of the open ocean 
indicated that the main compounds present were the following 
five-ring carcinogens (in percent by weight): BbF. 86.3: BkF, 
1 1 .0; BaA and Chr. 1 .2; BeP. 1 .0; BaP. 0.5: and IPy was present 
qualitatively. 

The total PAH's present in the waters of the Marianas 
sector were 1 to 1 00 times higher in concentration than in the 
open ocean. The dominant compounds were four- and five- 
ring PAH's ( in percent by weight), namely BaA and Chr. 38.7; 
BeP, 35.2; andpyrene(Py), 22.7. TheBbFandBaPconstnuted 
3.2% and 0.2% by weight, respectively. 

In areas where PAH levels in the water were below the 
sensitivity threshold of the liquid chromatography technique 
employed, a more sensitive method for just BaP was employed. 
Benzol a (pyrene is useful as a representative tracer of other 
PAH's (analytical methods described in Irhae/o/.. 1992). The 
following BaP levels were noted: up to 0.58 ng/1 in the open 
ocean and up to 1.36 ng/1 in the Marianas sector. 

Unlike the surface layer of water, suspended matter sampled 
in the tropical Pacific was found to contain as many as 10 lour- 
to six-ring PAH's. All samples taken at depths of 0, 20. 58, 80. 
1 30. and 200 m contained the PAH's IPy, BaA, Chr, BbF. BkF, 
and BaP. the dominant ones being IPy and BaA. which accounted 
for 30-99% and 30-50%. respectively, of the total amount 
present. In certain instances, suspensions from \ arums depths 
also contained IPy and benzo(g,h,i)perylene ( BPen. 

Bottom sediments of the tropical Pacific were found to 
contain only three kinds of four- and five-ring PAH's. namely 
BaA, BbF. and BaP. with the total PAH content not exceeding 
4.5 Ug/kg. The dominant PAH (by weight) was BbF, which 
accounted for 93% of the total amount of PAH's present. 
Although the composition of PAH' s in the bottom sediments of 
the Marianas sector of the Pacific was similar. IPy and BbF 
were the dominant PAH's (50.7% and 47.3' - . respectively, of 
the total weight of PAH's present). 

South China Sea 

Polycyclic aromatic hydrocarbon levels in the waters of 
the South China Sea were much higher than in the open ocean 
and in the Marianas sector (Table 3). The total amount of 
PAH's in the water \ aried w idel) : the PAH mixture identified 
in the surface microlayer consisted largely of four- and five- 
ring PAH's (in percent by weight): BeP. 45.5: BaA and Chr. 
28.7: Py . 24.6; BbF, 0.9; and BaP. 0.2. Their weight ratios and 
composition in the surface layer were somewhat different 
(Table 3 ). moving tow aids increased fractions of BaA and Chr. 
and BeP. Considerably higher levels of PAH's. by weight. 
were delected in the benthic portion of the water column, the 
dominant I> All being BbF (Table 3). 

In those sea areas where the PAH level fell below the 
analytical detection limit, the BaP level did not exceed 
1.58 ng/1. 



184 



TABLE 3 

PAH levels in particular media of the South China Sea ecosystems (Station 127. September 1988). 



Medium 



Units 



Pv 



BaA BeP BbF BkF BaP IPy BPer DBA 



Water, surface ng/1 qualitativel) 

microlayer present 



170 67 



4.5 



Water, surface ng/1 

layer (0 to 0.5 m) 



150 



175 



278 



5.5 



1.6 



Benthic layer ng/1 qualitatively 290 

present 



290 



14,000 



300 



Benthic sediment u/kg 



0.46 0.14 0.15 



Neusum 



u/kg 



1.4 



1.7 



.3 0.6 0.1 



Suspension. 
58 m 



ns/1 



4.7 



0.6 



0.9 0.06 0.01 0.07 tr. 0.07 tr. 



Suspension, 
80 m 



ng/1 



0.6 



0.4 



0.01 



0.04 



tr. 



0.06 



Hydroids 



trace 



u/kg 



6.5 



0.6 



0.7 



The total PAH level in the suspended matter drawn from 
depths of 58 and 80 m were 6.4 and 3.9 ng/1. respectively 
(Table 3). The PAH mix was found to contain ten four- to 
six-ring PAH' s in the following relative amounts (in percent by 
weight): IPy. 65.2; BeP. 17.4; BaA and Chr. 1 1.0; BeP. 1.0; 
BPer. 1.0: BbF. 0.9; BkF.0.2; and IPy anddibenz(a.h (anthracene 
(DBA) were present in trace amounts. 

The sediment samples contained three five-ring 
carcinogenic PAH's (Table 3), namely BbF. BkF. and BaP. 
The total PAH content of the sediment did not exceed 
0.75 (ig/kg of dry weight, the dominant constituent being BbF 
(61.3% of the total content by weight). 

The PAH mix detected in the marine biota indicated that 
PAH's were fairly abundant. The total PAH content of the 
neuston of the South China Sea was 6.6 ug/kg (Table 3). The 
mix consisted of six PAH's. Most of the total amount was made 
up of four- and five-ring compounds (i.e.. BaA. BeP. BbF. and 
BaP). 

It should be noted that the PAH mix detected in the marine 
hydroids was similar to that found in the sediment (Table 3 ) and 
consisted of strongly carcinogenic five-ring compounds, namely 
BbF. BaP. and BkF. Their total content in the hydroids was 
10 times that in the sediment, and the weight ratios reflected a 



larger fraction of BbF. 83.3%. The relative content of BaP and 
BkF was 8.9% and 7.6%, respectively. The same carcinogenic 
PAH's were detected in benthic organisms (Petrosia sponges. 
Table 4). 

TABLE 4 

Comparative PAH levels in sponge tissues (u/kg of dry weight) 

and fish livers (u/kg of fresh weight) from the South China Sea ( 1 ) 

and Bering Sea (2). 



PAH 


1 






2 




Sponge SI 


lark liver* 

n = 2 


Spongt 


: Liver of 
walleye pollack 


Chr 


_ 


6.15 


_ 


34.00 


BbF 


0.70 


0.27 


0.90 


4.50 


BkF 


0.09 


0.22 


0.16 


tr. 


BaP 


0.40 


1.00 


0.60 


13.00 


BaA 


qualitatively 
present 


1.00 


- 


qualitatively 
present 



BPer 



2.80 



Specimens caught in the tropical Pacific. 



185 



It may be noted that the total content and mix of PAH's 
present in the tissues of South China Sea sponges was not much 
different from those established for the same aquatic organisms 
in the Bering Sea (Table 4). The predominant PAH's 
(in percent by weight) for South China Sea and Bering Sea 
sponges were BbF (54 and 59) and BaP (36 and 33.6), 
respectively. 

Concentrations of PAH's in livers of sharks ranged from 
1.2 to 10.2 (ig/kg of fresh weight (Table 4). The predominant 
compounds were the following four- and five-ring PAH's 



(in percent by weight): Chr, 53.7; BPer, 24.7; BaA. 8.7; BaP, 
8.7; BbF, 2.3; and BkF, 1.9. 

For comparison we note that the levels and distribution of 
PAH's in aquatic organisms of the Bering Sea (sponges and 
walleye-pollack liver) also indicate relatively high total PAH's 
in fish livers and the buildup of the strong carcinogens BbF, 
BaP, and BkF as the dominant PAH's in sponges (Table 4). 

Analysis of the findings of the present study points to 
severe petroleum hydrocarbon pollution of the sea and to the 
buildup of the carcinogenic PAH's — BbF and BaP — in some 
aquatic organisms. 



2.3 The Occurrence and Microbial 

Transformation of Benzo(a)pyrene in the 
Waters of the Tropical Pacific (Caroline 
Atoll, Line Islands, Phoenix Islands 
Transect, South China Sea) 

YURIY L. VOLODKOVICH and OLGA L. BELYAEVA 

Institute of Global Climate and Ecology, State Committee for Hydrometeorology and Academy of Sciences, Moscow; USSR 



Introduction 

Among the numerous organic pollutants working to 
produce dangerous and undesirable changes in the chemical 
and biological status of the marine environment, polycyclic 
aromatic hydrocarbons (PAH's) of both natural and human 
origin must be singled out as being particularly noxious. 
Exhibiting considerable molecular stability, as well as 
pronounced carcinogenic and mutagenic properties, compounds 
of this series are a serious hazard to marine life. At present, 
researchers have been choosing ben7.o(a)pyrene (BaP) as an 
indicator of environmental pollution by PAH's. 

Benzo(a)pyrene and other PAH's have been found to 
occur in many marine ecosystem components from the arctic 
latitudes (Volodkovich & Belyaeva, 1987) all the way to the 
Antarctic, affecting mariculture and other activities. 

The proximity of human sources of pollution has a powerful 
impact on the marine environment, increasing BaP levels in 
waters by factors ranging from tens to hundreds (Tsyban et ai, 
1985b; Tsyban ef a/., 1986). 

For example. BaP levels in some parts of Los Angeles 
harbor have reached, and even exceeded. IIS ng/1. while 
concentrations in the top 0.5 m of the water column in the 
equatorial Pacific have been reported in the 1-6 ng/1 range 
(Shilina, 1982). 



In view of the fact that elevated PAH levels in seawater are 
due largely to the proximity of a given sea area to pollution 
sources, the study of BaP circulation and elimination processes 
(including microbial transformation) in as yet unimpacted 
open sea and open ocean areas is of considerable interest and 
value. 

Microorganisms, which are virtually ubiquitous in the 
World Ocean, play a major role in the functioning of marine 
ecosystems and in the biochemical cycling of various 
compounds, including pollutants such as petroleum crude and 
PAH's (Izrael & Tsyban, 1989). The processes involved in the 
microbial transformation of aromatic hydrocarbons and 
heterocyclic compounds have been investigated in sufficient 
detail. On the other hand, the rates of microbial degradation of 
BaP in the marine environment, as well as the local and 
regional scale of these processes, have thus far been neglected 
(Tsyban et ai, 1986). 

Integrated studies of biogeochemical cycling of PAH's. 
using BaP as an example, were performed in 1988 during the 
cruise of the R/V Akademik Korolev as an extension 
of work begun in 1981 in the South China Sea and the 
Marianas sector. The research represented the first time that 
such work has been undertaken in the tropical Pacific (in the 
Caroline Atoll area and the Line Islands/Phoenix 
Islands transect). 



186 



Methods 

Sampling 

The studies were performed on both rainwaterand seawater 
samples taken from a variety of depths ranging from the surface 
down to the benthic portion of the water column. Also 
investigated were natural bacterioplankton communities in 
surface and near-surface layers. 

Seawater samples were taken with Niskin bottles. 
Rainwater samples were collected in vessels of surface area 
0.5 m : mounted near the bow of the vessel. The BaP present in 
the water was extracted from 1-1 samples by triple benzene 
extraction. Final processing and analysis of 1 00 ml volumes of 
benzene extract of BaP were carried out ashore upon completion 
of the cruise. 

Seawater samples for use in the microbiological 
experiments were taken using sterile sampling equipment 
(a 5- 1 plastic Niskin bottle, and a 5-1 glass bottle, in the case of 
Caroline Atoll lagoon). The water samples containing natural 
microorganism communities were decanted into glass bottles 
under sterile conditions. Further processing was carried out in 
the shipboard microbiology lab. 

The In Situ Experiments to Assess the Biodegradation Potential 
of Seawater Microflora in Relation to BaP 

The process involved in the microbial transformation of 
BaP was studied under the conditions similar to the in situ 
experiments. These involved placing 250-ml samples 
containing natural microflora communities in dark glass bottles 
of volume 0.5 1. A weighed amount of BaP dissolved in a 
minimum quantity of acetone (0.05 ml ) was added to the bottles 
immediately after addition of the water ( the acetone evaporated 
within a few minutes). The transformation process was 
simulated in runs using BaP concentrations of 2 and 10|ig/l. In 
order to allow for abiotic factors, each experiment included a 
control in the form of a sample of sterile water drawn from the 
same depth and containing the same concentration of BaP. 
Each complete pair of tests (experiment plus control) was run 
three times. 

In order to make the conditions of the in situ experiments 
as close as possible to natural, the series of glass vessels were 
placed in baths with running water from the surrounding sea, 
with the baths themselves mounted on the deck of the research 
vessel. In the case of the Caroline Atoll, the sample-containing 
vessels were placed in a plastic cassette placed on the bottom 
of the lagoon (at a depth of 8 m). Exposure equaled 5-7 days, 
depending on the seawater temperature. On completion of 
exposure, the microbial activities were terminated by adding 
several milliliters of concentrated hydrochloric acid to the 
sample. The residual BaP in the experimental and control 
samples were extracted using benzene. The final BaP 
concentration values were determined ashore. 

Chemical Analysis 



Quantitative determinations of the residual BaP levels in 
the seawater samples involved concentrating the resulting 
benzene extracts down to a volume of 1 ml and then analyzing 
them on alumina plates using thin layer chromatography with 
heptane-benzene-acetone (100:60:6.7, by volume) as the 
solvent system. The BaP-containing area of the adsorbent was 
eluted with acetone, after which the BaP was transferred to 
n-octane. 

The benzene extracts of BaP that were obtained in the 
microbial experiments were treated without chromatographic 
thin-layer separation. The evaporated portion of the benzene 
extracts were eluted with 2-ml solutions of n-octane containing 
1 x 10 7 g/ml of 1,12-benzopyrelene, which was used as an 
internal standard. 

Quantitative determinations of BaP (in n-octane solution) 
were made by a fluorescence-spectrum analysis method, relying 
on the Shpolsky effect (Shpolsky et al., 1952; Fedoseeva & 
Khesina. 1968). The analyses were performed with a DFS- 12 
spectrograph at a temperature of -196°C using supplementary 
standards (BaP and othercompounds). The minimum sensitivity 
of the method for BaP was 1 x 10 '" g/ml, with error brackets 
of 10%. 

The simulated rates for microbiological degradation were 
determined from the difference between the initial (introduced) 
and the final (remaining) mass of BaP in the separate reactors. 

Results and Discussion 

The first series of studies concerned the equatorial Pacific 
(Fig. 1 ) in the rectangle bounded by 7° and 0°S and 150° and 
1 80°W. Analysis of the findings showed that BaP levels in the 
waters of the Line Islands/Phoenix Islands transect were 
remarkable for their extreme purity. Benzo(a)pyrene levels in 
the waters of the transect down to depths of 2.000 m 
corresponded to the minimal background value of 1 ng/1 and 
lower. Certain samples contained no detectable amount of BaP 
whatever (Table 1 ). Thus, the average BaP level for the entire 
water column at Station 1 17 was just 0.06, 0.018 ng/1. In the 
majority of cases, the BaP levels did not exceed 5 ng/1. As seen 
in Fig. 2, zones with elevated BaP levels were of a local 
character. High levels were recorded for depths of 
1,000-1,500 m (6.5 and 1 1.3 ng/1, respectively), with only one 
instance (10.3 ng/1. Station 1 18) of a high concentration in the 
top 0.5 m of the water column. It must be emphasized that the 
peak levels recorded were a whole order of magnitude lower 
than BaP levels in open ocean and other portions of the Pacific 
as found in earlier studies (Izrael&Tsyban, 1989). The limited 
occurrence and low levels of BaP buildup in seawater along the 
Line/Phoenix transect may be attributed not only to the 
remoteness of this geographic region from the principal human 
sources of pollution (industrialized areas and shipping routes) 
but also to the absence of any pronounced PAH flows in this 
ultraoligotrophic part of the ocean. Nor does economic activity 
in the waters of the Kiribati Republic archipelago appear to be 
having any significant deleterious effect on this marine 
environment. 



187 



100' 110" 120 130 140' 


150 c 160° 170° 


180 : 170 160 150 c 


5 sr^ 






\ South A 
S China \2> t 
("V/Sea %*' 
"Vs. .-jo ^> 


Pacific 
Ocean 


Hawaiian 

Islands t *^ ) 
i 

Christmas 
-..•3 Island n 




if>* 


Caroline 




a 


Atoll 

«0 



Fig. 1 . Expedition route and station locations in the Pacific Ocean and South 
China Sea (9 September to 31 October 1988). 



If this minimal concentration of BaP serves as an indicator 
representing a whole series of carcinogenic PAH"s, then the 
equatorial zone in question can be readily classified as one of 
the "background" portions of the World Ocean. 

The studies at Caroline Atoll (10°N, 150°W) indicated 
that BaP was present in individual water samples both from the 
interior portion of the lagoon and from the inshore waters at 



Stations 
120 119 118 117 116 115 



Depth, m 



i i i 



J L 



0.5 

10 



45 
100 



500 



Kldil 



15(H) 



2000 




BP concentration, ng/1 



□ - n -' 



□ 



->1 -2 



□ ->2-4 



->4-6 



->6-8 



Fig. 2. Occurrence of benzol a Ipyrene in the waters of the Line/Phoenix 
transect between 7°S and 0°S in the equatorial Pacific ( October 1 988 1. 



TABLE 1 

Benzol a )pyrene concentrations (ng/1) in the waters of the central Pacific (September and October 1988). 


Region Station No. 

0.5 


Depth, m 

10 25 45 100 


500 1.000 1.500 2.000 


Caroline Atoll 
Lagoon 0.0 


0.1 0.1 


0.1 


Inshore waters 1 14 0.2 


0.1 0.1 0.0 0.1 





Line/Phoenix 


115 


0.4 


1.0 


3.7 


2.5 


4.7 


4.1 


4.4 


4.7 


- 


transect 


116 


2.5 


1.4 


0.6 


1.3 


1.5 


4.1 


6.5 


0.9 


0.9 




117 


0.1 


0.3 


0.3 


0.7 


0.2 


0.6 


1.1 


1.2 


0.9 




118 


10.3 


0.0 


1.1 


2.1 


5.8 


0.6 


1.6 


5.9 


— 




119 


2.4 


3.7 


2.2 


1.4 


7.2 


1.9 


11.3 


6.5 


0.5 




120 


2.0 


0.6 


0.0 


0.3 


1.0 


2.9 


2.7 


2.8 


0.6 


Marianas 


121 


80.0 


7.0 


5.9 


9.2 


3.6 


5.5 


5.3 


2.5 


2.7 


transect 


122 


104.0 


96.0 


94.0 


101.0 


86.0 


6.6 


9.7 


111.0 


96.0 




123 


57.5 


84.0 


94.0 


62.0 


84.0 


56.0 


39.3 


6.7 


5.0 




124 


38.2 


96.6 


68.4 


5.0 


14.0 


12.5 


6.6 


14.5 


0.8 




125 


2.7 


I.I 


3.5 


11.3 


1.6 


2.3 


2.1 


0.9 


10.8 




126 


12.2 


12.1 


2.3 


1.9 


3.0 


9.0 


14.3 


1.0 


6.2 



South China 


127 


2.0 


3.2 


0.9 


16.0 


173.0* 




Sea 


127 


1.1 


2.3 


0.9 


0.6 


105.4* 






129 


18.5 


8.2 


3.2 


12.4 


9.3 


27.5** 




131 


64. X 


17.1 


1 1.0 


6.0 


10.0 


148.8*** 



Sea-bottom depths: 



* - 6 1 m 

** - 200 m 

*** - 250 m 



INS 



depths of up to 100 in. However, these also corresponded to 
background values not exceeding 0.2 ng/1 and averaging 
0.08 ng/l (Table 1 ). Particularly noteworthy was the absence 
of BaP in the top 0.5 m of the water column inside the lagoon. 

Without excluding the possibility of PAH biosynthesis in 
the highly productive coral ecosystem of Caroline Atoll (such 
as the phenomenon reported for the Clipperton Lagoons; 
Niassat el <//. . 1 968 ). it may be assumed, with some confidence, 
that the high rates of photochemical oxidation and microbial 
transformation of PAH's characteristic of equatorial waters are 
conducive to the elimination of PAH's from the waters of 
Caroline Atoll and especially from the surface layer. 

Quite the opposite situation was observed in the Marianas 
trough region along the Marianas transect (between 142°E and 
128 K. along 1 1°N) (Fig. 3). Whereas the BaP levels at the 
extremities of this sector, with one exception, ranged from 
0.9 to 9.2 ng/l (Table I ). the water column in the central portion 
of the transect (Stations 122 and 123) contained elevated and 
even maximum concentrations (up to and over 100 ng/l). As is 
evident from Fig. 3. high BaP levels from 38 to 104 ng/l were 
noted not only in the surface layer but also in deeper waters. 
Thus, the BaP concentration in the upper 100-m water layer at 
Station 122 averaged 96.2 ng/l. while the average value at 
1 ,500-2.000 m was 103.5 ng/l. The lowest BaP concentrations 
occurred at the western end of the transect, where the BaP 



HI', nii.i - i 









Fig. 3. Distribution of benzo(a)pyrene in the waters of the Marianas transect 
along 1 1°N (between 142°Eand 1 38°E) in the tropical Pacific in 1981 
and 1988. 



levels at Stations 1 26 and 1 25 in waters up to 2,000 m deep were 
4.03 and 6.88 ng/l. respectively. The average value for eastern 
Station 121, at depths from 10 to 2.000 m. was 5.17 ng/l. 

Analysis of rainwater samples taken at 1 1°N along the 
transect had BaP levels of 7.2, 8.8, and 9.2 ng/l; that was one 
order of magnitude lower than the levels in rainwater measured 
along the transect in 1981 (Izrael&Tsyban, 1989). Nevertheless, 
these findings point to the possible contribution of PAH's to 
the Marianas sector of the Pacific via long-distance atmospheric 
transport of organic pollutants. 

High BaP levels (up to over 100 ng/l) were noted in the 
waters of the Marianas sector during the previous study period 
in 1981 (Izrael & Tsyban, 1989). These ranged from 80 to 
1 20 ng/l. However, the pollution in all instances was confined 
to the top 2 m of the water column with peaks in the surface 
microlayer (SML). 

Since most of the waters of the Marianas sector lie in the 
area of influence of the northern branch of the tradewind 
current, it may be supposed that the invasion of deeper layers 
by pollutants (including BaP) occurs through downwelling of 
surface water due to the anticyclonic circulation of water 
masses in this region. 

The BaP levels in the southwestern South China Sea had 
a broad range of values from 0.9 to 173 ng/l (Table 1 ). The 
lowest BaP levels occurred in open waters at Station 127 
(Fig. 1). Two days of work at this station yielded similar 
estimates, the BaP levels over most of the water column 
varying from 0.6 to 2.0 ng/l and averaging 1 .5 ng/l. The bottom 
layer (at a depth of 61 m) exhibited a very high BaP concentration 
and with an average value of 1 39.2 ng/l. indicating severe PAH 
pollution. 

The BaP concentrations in the waters of stations lying 
close to the eastern coast of Singapore (Stations 1 29 and 131) 
hadelevated BaP concentrations in the lOto 100-m layerof the 
water column (averaging 8.2 and 13.2 ng/l, respectively). The 
maximum BaP lev els were detected in the top 0.5 m layeron the 
one hand and at the sea bottom on the other. These were as high 
as 148. 8 ng/l (Table 1). 

This peculiar pattern of BaP level distribution over the 
water column can be attributed to the influx of PAH's with the 
oil pollution that impacts the surface waters at the stations in 
question as well as the deeper waters below. It should be noted 
that the areas of the South China Sea investigated are busy 
maritime thoroughfares for vessels carrying both crude oil and 
refinery products as well as sites of intense offshore drilling 
activity. 

The potential physiological ability of microflora to 
transform BaP in the central Pacific was studied in a series of 
in situ experiments using natural bacterioplankton communities 
from Caroline Atoll and the southwestern portion of the South 
China Sea. 

The results of the process simulations expressed in terms 
of the degree of elimination of the artificially-introduced 
amount of BaP are shown in Fig. 4. The bacterioplankton of the 
marine areas investigated were able to transform the 
polyaromatic hydrocarbon in question. On the whole, the level 
of microbial degradation reached 57-98% of the mass of BaP 
introduced into the samples (Table 2). In order to gain deeper 



189 



1(H) 
BaP tranformation 
as percentage of 
initial mass (%). 80 




ontrol 



Days 



Fig.4. 



Dynamics of BaP transformation in the in situ experiments using 
microflore from waters of Caroline Atoll lagoon (a) and from the 
South China Sea (b) at Station 127 (1988 data). (Initial BaP 
concentration: A- 2 |ig/l; O- 10 ug/l;#- control) 

insight into the natural processes of BaP elimination from the 
marine environment, we also undertook a study of the dynamics 
involved in microbial transformation. The results obtained 
from individual runs of each experiment yielded reasonably 
consistent values. 

As may be seen from Fig. 4, the highest microbial 
transformation rates were noted in experiments using microflora 
taken from Caroline Atoll lagoon water. The series of samples 
placed in natural conditions at a depth of 8 m contained just 
50% of the initially introduced amount of BaP (initial 
concentration 10 ug/1) after only 2 days; after 5 days had 
elapsed, 97% of the initial BaP had been transformed. The rate 
of microbial transformation of BaP in the lagoon waters was 
high, with the curve of the degradation process close to linear. 



The rate of microbial transformation of BaP in the South 
China Sea was much lower than for the atoll. With an initial 
BaP concentration of 2 ng/1, only one-third of its initial mass 
had been transformed after 5 days. However, after 7 days of the 
experiment, the microflora from the top 0.5 m of the water 
column was able to transform as much as 77% of the introduced 
BaP. With a higher concentration (10 ug/1). microbial 
transformation in the experiment produced a consistent value, 
34% of the initial mass over the third through the fifth days of 
the experiment where a period of decreased microfloral activity 
occurred (Fig. 4), the incremental transformation amounting to 
just 0.6 ug of BaP. In these tests, the total mass of the BaP 
degraded after 7 days; at the time which the experiment was 
terminated, the mass did not exceed 5.68 ug (56% of the initial 
concentration). On the whole, the rate of microbial 
transformation of BaP turned out to be not particularly dependent 
on the initial concentration (2 and 10 Ug/1). Assuming the 
process curve to be more or less linear for both the equatorial 
and the tropical Pacific, we were able to estimate the average 
rates of microbial transformation of BaP. The resulting values 
were 1.94 ug/l/day for the waters of Caroline Atoll and 
0.81 ug/l/day for the South China Sea. 

The abilities of marine microflora to degrade BaP are 
therefore relatively high. In the case of the Caroline Atoll, the 
biodegradation rates exceeded those of similar processes 
investigated in impacted areas such as the Baltic Sea (Tsyban 
ctal., 1985a). The results of our 1988 studies demonstrate the 
need to consider PAH, and especially BaP. metabolism as an 
exceedingly important process for detoxification through the 
elimination of the pollutants in question from the highly 
dynamic ecosystem that constitutes the tropical zones of the 
World Ocean. 



TABLE 2 

Dynamics of benzo(a)pyrene transformation by microflora from central Pacific waters in the in situ 

experiments (October 1988). 



Region of 

Operations and 

sampling depth. 

m 



BaP concentration, ug/1 

Exposure, Initial At end of experiment 
days Cj Control Exp. 



Microbial 
transformation of 

BaP, in % of 
concentration C, 



Caroline Atoll 
Lagoon, S m 



South China Sea. 
Station 127,0.5 m 



Same as above 






id 


1 


10 


2 


10 


3 


10 


4 


10 


5 


10 


5 


10 


() 


2.0 


3 


2 


5 


2 


7 


2 


7 


2 





10 


3 


10 


5 


10 


7 


10 


7 


10 



9.9 



9.9 



2.0 



2.0 



10 



9.9 



7.8 
4.9 
3.? 
1.8 
0.3 





1.71 
1.37 
0.46 



6.62 
6.00 

4.32 




22 
51 
65 
82 
97 




14.5 

3 1 .5 

77.0 




33.8 

40.0 

56.8 





190 



2.4 Cesium- 137 in the Surface Waters of the 
Central Equatorial Pacific 



VLADIMIR I. MEDINETS 5 , VLADIMIR G. SOLOVEV , and BORIS V. GLEBOV* 

'State Oceanographic Institute, Odessa Branch, USSR 

* Institute of Global Climate and Ecology, State Committee for Hydrometeorology and Academy of Sciences, Moscow, USSR 



Introduction 

The study of radioactive contamination of ocean waters 
throughout the world is one of the most important tasks facing 
researchers concerned with the effects of human activities on 
the ocean environment. Man-induced radioactivity of the 
marine environment has three origins: 7. nuclear weapons 
testing; 2. nuclear power plant operation; and 3. operation of 
nuclear fuel treatment and reprocessing plants. 

The need for more information about the present status of 
radioactive contaminants of World Ocean waters prompted 
our research efforts in the course of the 47th cruise of the R/V 
Akademik Korolev in the central portion of the central equatorial 
and western Pacific from September to November 1988. 

The indicator of radioactive pollution selected was cesium- 
1 37, a radionuclide with a long half-life. It is this radionuclide, 
together with strontium-90 and plutonium-239, that presents 
the greatest potential threat to the marine environment. 

Methods 

Seawater samples of large volume (0.8-1.1 m 1 ) were 
drawn from various depths (0-200 m) using NIVA and Malysh 
immersion pumps. Selective extraction of the cesium- 1 37 was 
performed with the aid of Milton-T fibrous sorbent impregnated 
with copper ferrocyanide (Vakulovsky, 1986). Following 
extraction, the sorbent was reduced to ash in a flameless muffle 
furnace at a temperature not exceeding 450°C. The ash residue 
was then hermetically sealed in a polyethylene capsule. In 
December 1988. the samples were measured using a gamma- 
spectrometry setup at one of the laboratories of the State 
Oceanographic Institute, Odessa Branch. The error of the 
cesium- 137 determinations did not exceed 10%. The detection 
threshold was 0.01 Bq. 

Results 

Cesium- 137 levels in surface waters were determined in 
the central equatorial portion of the Pacific, as well as in the 
Philippine, South China, East China, and Japan Seas. The 
results are summarized in Table 1. Figure 1 shows the location 
of cesium-137 level measurements in the surface (0-3 m) layer 
of Pacific Basin waters along the expedition route. 



Analysis of the data showed that the spatial distribution 
of cesium-137 levels was very uneven. The absolute 
minimum and maximum concentrations were recorded in the 
equatorial Pacific. The minimum level recorded in the course 
of the entire expedition was situated at 8°N, 1 56°35'W between 
Hawaii and Christmas Island. This was at a time when the 
Akademik Korolev was traversing the northern boundary 
of the intratropical convergence zone. Precipitation was 
abundant, and the resulting dilution was what, in our view, 
lowered the cesium- 1 37 concentration in the top layer down to 
0.7 Bq/m\ 

A maximum concentration of 7.0 Bq/m 3 was recorded in 
the vicinity of Kusaie Island in the Carolines group. 

During this period, the vessel was at its shortest distance 
from the Marshall Islands that include the Bikini and Eniwetok 
Atolls, sites where nuclear weapons were tested in the 1950's. 
The elevated levels observed were attributable either to the 
transport of the cesium-137 enriched water masses from the 
Marshall Islands or to the presence of the local sources on the 
Kusaie Island, lying eight miles north of the cruise track. A 
more detailed survey of the region might allow identification of 
sources. 

Elevated (3.4-3.9 Bq/m 1 ) cesium-137 levels in the waters 
off Tarawa and Caroline Atolls and Christmas Island may have 
been due to the transport of oceanic waters from the Tuamotu 
Archipelago, where nuclear weapon testing is regularly 
conducted by France. The considerable difference between the 
cesium-137 levels in samples taken two miles off Caroline 
Atoll and those measured in the waters of its lagoon was 
especially noteworthy. 

The cesium-137 level in the lagoon water was double the 
value obtained for oceanic water taken off the atoll. This may 
be attributed to cesium-137 enrichment of the inner lagoon 
waters by the flushing of radionuclides deposited on the atoll 
surface by atmospheric fallout. Analysis of the average levels 
of cesium-137 for each region (Table 2) showed that the 
'cleanest' waters of all the ocean areas studied were those of the 
equatorial Pacific. In terms of radioactive pollution, this region 
must be ranked between the Bering Sea (Medinets et al. . 1 992) 
and the rest of the regions of the Pacific. 

The findings of the present study agree with the results 
obtained by Japanese researchers who worked in the central 
Pacific in 1980-82 (Nagaya & Nakamura, 1985). 



191 



TABLE 1 



Cesium- 137 levels in Pacific Ocean waters in the autumn of 1988. 



TABLE 2 

Mean cesium- 137 levels in seawater bv region. 



Date 


Cooi 


dinates 


Sampling 


Mean levels 




Latitude 


Longitude 


depth ( m ) 


Bq/m 3 


09-15 


8°00' N 


156°35' W 


0-3 


0.7 


09-16 


4°40' N 


157°04'W 


0-3 


2.7 


09-18 


1°28'N 


157°43' W 


0-3 


2.6 


09-19 


1°I6'N 


I56°00'W 


0-3 


3.5 


09-20 


5°49' S 


153°05'W 


0-60 


3.6 


09-2 1 


6°53' S 


I53°21'W 


- 7(1 


2.9 


09-25 


9°59' S 


!50°15'W 


(1- 100 


1.9 


09-28* 


9°59' S 


15015' W 


0-3 


3.9 


10-02 


6°37' S 


!61 a 44'W 


0- 100 


3.6 


10-04 


3°48' S 


1 72°03' W 


0- 100 


-i 2 


10-07 


0°16' S 


177°38' 1 


0- 20 


2.0 


10-08 


l°20'N 


172 20' E 


0- 100 


1.9 


10-11 


2°41'N 


168 C 17'E 


0- 3 


3.3 


10-11 


3°03' N 


I61°01'E 


0-3 


2.3 


10-12 


4 20'N 


162°27'E 


0- 3 


7.0 


10-13 


7°21'N 


156°42' E 


0-3 


4.4 


10-14 


10 08' N 


153=45' E 


0-3 


5.9 


10-15 


1016' N 


147°45'E 


0-3 


3.4 


10-16 


i r 12' n 


I42°55'E 


0-3 


2.9 


10-17 


11 1 5' N 


1 39°35' E 


0-3 


3.0 


10-18 


noo'N 


1 36°00' E 


0- 120 


4.4 


10-18 


1 1 °05' N 


134°07'E 


0-3 


4.9 


10-19 


1 1 ()()' N 


I32°21'E 


- 3 


3.7 


10-20 


1 1 °00' N 


130 33' E 


0-3 


3.3 


10-21 


ID 59' N 


128°46'E 


- 200 


5.5 


10-24 


5°I2'N 


I14°05'E 


0-3 


6.2 


10-25 


5 50' N 


I09°27'E 


0-3 


5.3 


10-27 


6°00' N 


I06°54' E 


- 50 


5.2 


10-27 


6°00' N 


106°54'E 


- 50 


5.2 


10-28 


5°14'N 


106°27' E 


- 55 


4.2 


10-28 


5°14'N 


106°27' E 


- 60 


4.5 


10-29 


4°18'N 


)05°54' E 


0-80 


4.8 


10-29 


3 23' N 


105°19'E 


- 60 


4.3 


10 30 


: 1 9' n 


1 1 14 54' E 


0-4(1 


3.4 


11-06 


4 20' N 


I06°42'E 


(1- 70 


5.9 


11-07 


4 4I'N 


II 47' E 


- 70 


5.2 


1 1-08 


8°04' N 


I16°13'E 


0- 3 


4.9 


1 1-09 


ll°35'N 


1 1 8°42' E 


0-3 


6.3 


1 1-10 


I7°26'N 


114 53' E 


0- 120 


4.9 


nil 


21 I6'N 


121°43'E 


- 10 


4.8 


11-12 


24 58' N 


12 V 39' E 


0- 10 


4.3 


11-13 


2') 00' N 


1 27°00' E 


- m 


4.3 


11-13 


S2 17' N 


I28°09'E 


0- 10 


4.4 


11-14 


36 41' N 


130°52'E 


■ lu 


5.3 


11-15 


41 M'N 


I32°12'E 


0- 10 


4.7 


II 15 


43 Ol'N 


131°59'E 


0- Hi 


5.6 


Sample 


draw n Iron 


the lagoon ( 


1 Caroline Ah 


II. 



Region 



Mean level 
Bq/m 



Bering Seal 50-60° N) 
(from Medinets <■/ <//.. 1992 ) 

Pacific Ocean (10-40 N) 
(from Nagaya& Nakamura, l e )S5i 

Equatorial Pacific (5 N- 1 0°S) 

South China Sea 

East China Sea 

Sea of Japan 



2.4 ±0.2 

5.2 ± 0.3 

3.610.3 
5.0 ±0.2 

4 3 ± 0. 1 
5.2 ± 0.3 




Fig. 1. Map of sampling locations to determine cesium- 137 levels (Bq/m 1 ) in 
the surface layer (0-3 m). (^denotes results obtained in the present 
study; denotes results reported by Nagaya & Nakamura. 1985) 



Comparison of the overall radiation picture in the Pacific 
region with our findings on the contamination status of the 
Black and Baltic Seas conducted from 1 986 to 1 OSS show s that 
cesium- 137 concentrations in the Pacific arc. at present. 10 to 
20 times lower than in the seas of the European region impacted 
bv local resources. 



192 



2.5 Quantity and Distribution of Plastics: 
An Analysis of Chemical Hazards to 
Marine Life 

GREGORY J. SMITH' and CHARLES J. STAFFORD ; 

"Wildlife International Ltd., Easton, Maryland, USA 

f US Environmental Protection Agency, Analytical Chemistry Laboratoiy, Beltsville, Mankind. USA 



Introduction 

In recent years, plastics in the marine environment have 
been recognized as important pollutants of marine ecosystems 
(Laist, 1987; Prater, 1987; Wolfe, 1987). Medical waste, 
plastic debris, and other types of refuse have washed ashore on 
Atlantic Coast beaches at alarming rates. These events have 
created new public awareness of the critical nature of the 
plastic waste disposal. Plastics present a unique disposal 
problem because the same attributes that make many types of 
plastics useful also enhance their longevity and buoyancy in 
the world's oceans. 

The distribution of plastics has been studied in various 
estuarine. coastal, and oceanic waters of the world. 
Concentrations of plastics in surface waters have been associated 
with oceanic convergence zones or the proximity to shipping 
lanes (Colton et ai. 1974; Wong et ai. 1974; Wolfe, 1987). 
Although the distribution and abundance of plastics in the 
oceans have been studied almost exclusively in surface water, 
it is well known that many types of plastics do not float and 
must occur in marine sediments (Shaw, 1977). Gradual sinking 
of those plastics that do float is also possible as bacteria, 
diatoms, hydroids. and other marine life grow on the surface of 
floating plastics (Carpenter et al., 1972; Colton et ai. 1974; 
Winston, 1982). Morris (1980) indicated that floating plastics 
gradually sink to the bottom or to a denser horizon where they 
attain neutral buoyancy and remain in suspension. Because the 
stratum of the pycnocline represents a marked change in water 
density, this would appear to be a likely area for subsurface 
accumulation of plastics with the appropriate neutral buoyancy. 

Potential adverse effects of plastics in the marine 
environment include aesthetic, physical, and chemical. Clearly, 
floating plastic debris or litter that occurs on beaches is visually 
unpleasant. Physical impacts of plastics have been well 
documented. Entanglement and ingestion of floating plastics 
by sea turtles have been found in several areas (Balazs, 1985; 
Carr, 1987). Fish have been found to ingest plastic pellets 
(Carpenter et ai. 1972), as have seabirds (Connors & Smith, 
1982;Furness, 1985; Day & Shaw, 1987; Fry etai. 1987). In 
several studies seabirds were found to have consumed massive 
quantities of plastics (Day et al.. 1985: Furness. 1985; Fry 
etai, 1987; Ryan. 1987). The physical ingestion of plastics has 
been detrimental with respect to causing digestive system 
impaction, ulcerative lesions, and reducing meal size 



(Day etai. 1985; Fry etai. 1987; Ryan, 1988a). However, 
there continues to be much speculation regarding the potential 
chemical hazards of plastics to seabirds and other marine 
organisms. 

Raw plastic pellets have generally been considered to be 
biologically inactive, although manufactured plastics often 
contain additives known to be toxic (van Franeker, 1985). 
Moreover, since Carpenter et al. (1972) published the first 
account of polychlorinated biphenyls (PCB's) adsorbing to 
plastic spherules in seawater. there has been concern for the 
potential for PCB exposure to fish and wildlife that ingest 
plastics. Ryan et al. ( 1988b) found evidence to suggest that 
seabirds assimilated PCB's from ingested plastic particles. 
However, Fry etai ( 1 987 ) suggested that plastics were unlikely 
to pose a significant toxic hazard to birds compared to the 
physical impaction effects that may result. Polychlorinated 
biphenyls are commonly found in certain types of plastics and 
the potential for plastics to adsorb PCB's from ambient water 
would suggest two possible sources of PCB contamination 
from marine plastics. One potential source would be intrinsic: 
PCB's incorporated into the plastic during manufacturing. A 
second source, extrinsic, represents PCB' s adsorbed or absorbed 
from contaminated water. To date, there is still little known 
about the association of environmental contaminants occurring 
both in and on plastics in the marine environment and their 
potential toxic hazard to marine organisms. 

In this paper we report the abundance and distribution of 
plastics in areas of the central Pacific Ocean and the South 
China Sea. Both surface water and water at the level of the 
pycnocline were sampled. Chemical extractions of the plastics 
recovered were analyzed for several organic pollutants and 
subsequent field and laboratory experiments were conducted 
to elucidate potential toxic hazards that plastics may pose to 
fish, seabirds, and other marine organisms. 

Methods 

Field Sampling 

Sampling was conducted during the Third Joint 
US-USSR Bering & Chukchi Seas Expedition to the two seas 
and the central Pacific Ocean, along the segment from Hawaii 
to Singapore. Trawls for plastics were made from 15 September 
to 30 October 1988, along the 18,400-km cruise track of the 
Soviet R/V Akademik Korolev. During the cruise, the vessel 



193 



spent 8 days at Caroline Atoll, Kiribati, a remote coral atoll at 
9°59.09"S, 150° 14.04" W. Neuston sampling of the lagoon 
water of Caroline Atoll was done using a stationary 102 u. 
(0.5 x 1 ml neuston net placed in the tidal currents of the 
entrance to the lagoon. The stationary net was left in position, 
with the opening facing toward the lagoon, over a 7-day period, 
during both flood and ebb tides. Also, one additional location 
in the central lagoon was sampled during two 30-min tows 
using two neuston nets ( 102 |i) towed alongside an inflatable 
boat. 

Three different types of sampling methods were used in 
the open ocean. These included /. surface sampling using a 
1 x 2-m tucker trawl (in the open position) equipped with a 
94 u. net; 2. surface sampling with a 102 U. neuston net with a 
0.5 x 1- m opening; and 3. subsurface water sampling at the 
pycnocline horizon by deploying and hauling the tucker trawl 
in the closed position and fishing at depth in the open position. 
The pycnocline was determined using Soviet equipment 
installed aboard the research vessel to determine temperature, 
salinity, and density. Each net apparatus was equipped with a 
flow meter to determine the linear distance sampled that was 
later expressed on a surface area basis for surface sampling and 
a volume basis for sampling at depth. 

Surface water trawls from the ship averages 15 min with 
half of the rectangular opening of each net below the surface. 
The tucker trawl was towed off the bow of the ship at speeds 
between 82.0 and 280.6 cm/s. Neuston trawls were done from 
the stern of the ship while drifting at speeds from 
5.4 to 55.0 cm/s. At the end of each tow the sample was 
removed and passed through a series of acetone/hexane rinsed 
stainless steel sieves with pore sizes of 4 mm, 1 mm, 500 |i, and 
106 u to fractionate the sample. Each fraction was examined 
under magnification for the presence of plastics; marine 
organisms were separated from the sample for storage. Plastics 
recovered from the sample were immediately characterized as 
to size, type (raw pellet, fragments of plastic objects, fishing 
gear), and frozen in chemically-clean glass jars. 

Chemical Analysis 

All plastics (fragments and pellets) were removed from 
I lie fractionated sample, weighed, and extracted by shaking the 
sample three times in separate 5-ml rinses of hexane. Hexane 
extraction was used to remove nonpolar organics from the 
surface of the plastics and to avoid chemically dissolving the 
plastic samples. The extracts were combined and concentrated 
to a volume of I ml using a stream of dry nitrogen. Sample 
extracts were analyzed for chlorinated hydrocarbons using a 
Hewlett-Packard (HP)589()A gas chromatograph equipped 
with a "'Ni electron capture detector and a 30m DB1 column. 
An IIP 5890A interfaced with an HP 5970 mass selective 
detector, in the full scanning mode from 50 to 450 atomic mass 
units, was used to detect the presence of other halogenated 
hydrocarbons and petroleum hydrocarbons. 

Plastic Adsorption and Chemical Release Experiment 

1 o assess the possible chemical hazards of plastic to 
marine life two experiments were conducted. Theseexperiments 
were designed to evaluate the potential adsorption of organic 



compounds from surface water and to determine if the 
gastrointestinal environment of birds could cause the release of 
organic compounds from raw polyethylene pellets. 

To determine the adsorption potential of plastics 
from surface water, 100 g of new polyethylene pellets 
were placed in a 1-m-diameter brass-wire mesh enclosure in 
Baltimore Harbor. Chesapeake Bay. After 24 h. the pellets 
were collected, handled, and analyzed according to 
the procedures described above. At the same time the plastics 
were collected, a sample of the surface microlayer was collected 
by contact and adhesion of surface water to a glass plate. 
The plate was dipped repeatedly and rinsed with methylene 
chloride to obtain a sample volume of approximately 1-1 
of water from the microlayer. Water microlayer samples 
were extracted with methylene chloride and analyzed using 
the same methods as the plastic extracts. The purpose of 
the microlayer sample was to facilitate the comparison 
between compounds found in the surface microlayer and 
what was potentially adsorbed to the plastic pellets. 
This experiment was conducted to assess the possible 
adsorption characteristics of plastics under environmental 
conditions. 

To determine if chlorinated or petroleum hydrocarbons 
could potentially be released from polyethylene pellets in the 
digestive system of birds, a simulation experiment was 
conducted. The experiment was patterned after those of 
Kimball and Munir ( 1 97 1 ) using a 42 C HC1 bath to simulate 
the physicochemical conditions in the digestive system of 
waterfowl. Hydrochloric acid was added to 1 N saline solution 
to yield digestive solutions with pH"s of 1.4 and 2.8. Two 
50-g samples of polyethylene pellets were added to glass jars 
containing either a 1 N saline only (controls). pH 1.4 solutions, 
pH 2.8 solutions, or a hexane (without saline). These solutions 
were agitated in a water bath at 42°C. At the end of 8 days, the 
plastics and the digestive solutions were extracted with 
methylene chloride and these extracts were analyzed for 
chlorinated and petroleum hydrocarbons using the methods 
described above. 

Results and Discussion 

Over 80,000 m : of surface water and 93.000 m 3 of 
subsurface water were sampled during the course of the 
expedition (Table 1 ). The total number of stations sampled 
using either the 947 \i tucker trawl, the 102 |i neuston net. or 
both at a single location was 28 and 8 for the Pacific Ocean and 
South China Sea. respectively. Sampling at Caroline Atoll was 
done at only two locations. Plastics were recovered from 219i 
of the sample collected in the Pacific Ocean; five of six positive 
samples contained opaque polyethylene pellets. The sixth 
sample contained plastic fragments, not raw pellets. The 
concentration of plastics at positive stations varied considerably. 
with a maximum of 0. 1 8566 mg plastic/m : (Table 2 1. Although 
sampling was done along the cruise track from as far north as 
10°N to as tar south as 10°S. plastics were only recovered from 
stations near the ION latitude area (Fig. I ). Tar balls were 
found at only two of the stations sampled in the open waters of 
the Pacific (Table I). 



194 



TABLE 1 



Sampling information and recoveries of plastics from the central Pacific Ocean and South China Sea. Sampling 
was done from 15 September through 30 October 1988. 









Volume(m') 


Total 


Number of Sampling Stations 


with 


Geographic 


Sample 


Net 


or Area (m : ) 


No. 


Plastic 


Plastic 


Tar 


Region 


Type 


Mesh(u) 


Sampled 


Stations 


Fragments 


Pellets 


Balls 


Pacific Ocean 


Surface 


947 


39.214 


23 


5 


5 


2 






102 


3.627 


8 


1 





2 




Pycnocline 


947 


5 1 ,657 


9 











Caroline Atoll 


Surface 


102 


21.571 


fixed 
station 













Surface 


102 


2.508 


2 trawls 











South China Sea 


Surface 


947 


15.213 


8 


7 


2 


2 






102 


394 


1 


1 










Pycnocline 


947 


42,107 


4 


1 









TABLE 2 

Occurrence and concentrations of plastics from the central Pacific 

Ocean and South China Sea. Sampling was done from 

15 September through 30 October 1988. 

Plastic" Concentrations (mg/m : ) 
Stations Sampled at Positive Stations 

Geographic No. of No. w/ 

Range Locations 11 Plastics Mean Range 



Pacific Ocean 28 6 

South China Sea 8 7 



0.07340 0.01309-0.18566 
0.19478 0.00126-0.69632 




■^ 



>m. M ,!,:( 



Plastic includes pellets and fragments. 
' Several stations were represented by more than one sample. 



In the South China Sea, seven of the eight different 
locations sampled had plastics present in the surface water, but 
only two of the seven had plastic pellets (Table 1 ). Most of 
the plastics found in this area consisted of synthetic fishing 
line and secondary, manufactured plastics in the process of 
breaking down. Plastic concentrations in the South China 
Sea samples were considerably greater than those found in 
the Pacific Ocean, with a maximum concentration of 
0.69632 mg plastic/m 2 (Table 2). Although only two stations 
sampled had tar balls present, one of these samples contained 
hundreds of tar balls and 164 of these were greater than 4 mm 
in size. 

Sampling of the lagoon water at Caroline Atoll was done 
at two locations, one in the central lagoon and the other at a 
fixed station at the entrance channel to the lagoon. At the fixed 
station, current flowed through the neuston net at an average 



Fig. 1 . Locations sampled along the cruise track of the Akademik Korolev, 
September-October 1988. Sampling locations are indicated by solid 
dots. Plastic densities at positive stations are given in mg per meter 
square. Major currents for October 1988 are shown by the arrows. 

velocity of 4.08 cm/s. No plastics or other anthropogenic 
materials were recovered from any of the samples at Caroline 
Atoll (Table 1). 

The concentrations of plastics in the open Pacific Ocean 
stations that had plastics present in the surface water were 
below mean concentrations reported for the subtropical North 
Pacific (Wong el al, 1974; Shaw & Mapes, 1979) and higher 
than from other studies in the North Pacific (Day & Shaw, 
1987). Results from this study indicate that a much greater 
concentration of raw plastic pellets occur in the central Pacific 
compared to the North Pacific. There are no other data from the 
central Pacific region that can be used for temporal comparisons. 

The greater frequency of occurrence of raw material 
plastic pellets compared to discarded plastic objects and tar 
balls in the South China Sea was somewhat surprising. The 
occurrence of discarded plastic may be expected from the 



195 



heavy shipping traffic in this region of the world. Moreover, 
there is extensive offshore oil production in the South China 
Sea and a high occurrence ot tar balls would also be expected. 
The greater frequency of occurrence of plastic pellets compared 
to tar balls is noteworthy. The longevity of plastics in the 
marine environment likely contributes to the increase of this 
material in the oceans. 

The distribution of plastics in the Pacific Ocean was 
largely a function of the major currents. All stations sampled 
that contained plastics were along a l()°N latitude convergence 
lying between the North Equatorial Current and the North 
Equatorial Countercurrent (Fig. 1 ). The most likely source of 
the plastics recovered at this convergence would either be from 
ships or as industrial waste from plastic-producing Pacific Rim 
countries. 

All plastic sample extracts were analyzed for the organic 
compounds listed in Table 3. None of the samples contained 
quantifiable concentrations of organochlorine pesticides, PCB's, 
aliphatic hydrocarbons, or polycyclic aromatic hydrocarbons 
included in the analyses. The absence of detectable levels of 
these contaminants suggests that the plastics that were collected 
at sea did not adsorb any of these compounds, either because 
the plastic surface did not facilitate this or because the 
contaminants were not present in sufficient concentrations. 

In the field experiment conducted in Chesapeake Bay's 
Baltimore Harbor, plastic polyethylene pellets left floating in 
the water for 24 h were extracted and analyzed using the same 

TABLE 3 

Organic compounds included in the analysis of extractions of 
plastics 



Chlorinated 
Pesticides 
ami PCB's 



Aliphatic 
Hydrocarbons 



Polycyclic 

Aromatic 
Hvdrocarbons 



Heptachlorepoxide 
Oxychlordane 
Trans - chlordane 
Cis - chlordane 
Trans - nonachlor 
Cis - nonachlor 
Dieldrin 
I .mil i ii 
p.p" - DDT 
p.p" • DDD 

p.p" 1)1)1 

PBC(1254) 



n - dodecane 
n - tridecane 
n - tetradecane 
n - pentadecane 
Nonycyclohexane 
n - hexadecane 
n - heptadecane 
n_- octadecane 
n - nonadecane 
n - eicosane 

Tetrameth) I- 
pentadecane 



Naphthalene 

Fluorene 

Phenanthrene 

Anthracene 

Fliioranthrene 

Pj rene 

1.2. -ben/anthracene 

Chrysene 

Benzo(b)fluoranthene 

9,10- 

diphenylanthrocene 

Benzo(e)pyrene 

Benzo(a)pyrene 
1.2.5.6- dibenz 
anthracene 
Benzo(g,h,i)perylene 
Pen lene 



The lower limit of quantification was 0.05 ug/g for aliphatic 
hydrocarbons, 0.4 ug/g lor polycyclic hydrocarbons, mis ug/g for 
chlorinated pesticides, and 0.25 ug/g foi l'( !B's based on a 0.2 g 

sample. 



methods as those collected at sea. A sample of the microlayer 
was taken the same time these plastics were placed in the water 
and was analyzed for the same contaminants as the plastic 
extracts. Two aromatic compounds, pyrene and fluoranthene, 
were detected in the microlayer sample: however, only 
fluoranthene was present in the extract of the polyethylene 
pellets. 

These data indicate that polyethylene pellets have the 
potential to adsorb at least one organic compound, fluoranthene. 
from ambient seawater. The specific mechanism of adsorption 
could be either a binding of the organic compound to the plastic 
surface directly or to a film of water surrounding the plastic 
pellet. It is possible that other contaminants, especially those 
occurring in high concentrations in the microlayer, could also 
be adsorbed to floating plastic debris. Therefore, a potential for 
the transfer of certain organic pollutants from plastics to 
organisms that ingest them does exist. Plastics may serve as a 
vehicle for pollutant transport that may enhance exposure of 
organisms; however, if plastics also concentrate contaminants 
from the water this would result in an even greater hazard to 
marine organisms. 

The other aspect of chemical-induced toxicosis resulting 
from plastic ingestion is that of the potential release of chemicals 
from the plastic to the organism. In the second experiment, 
none of the extracts of the acidic digestive solutions, or the 
hexane extraction of plastic pellets contained quantifiable 
concentrations of any of the hydrocarbon contaminants. These 
data suggest that even when exposed to mild heat, acids, and 
hexane, polyethylene pellets did not release appreciable 
quantities of chemicals. However, plastics subjected to real 
avian digestive systems would also be exposed to digestive 
enzymes and tor possibly much longer periods of time. Ryan 
( 1988b) found that domestic chickens fed polyethv lene pellets 
retained 98.3% ofthemoveran 18-day period. Moreover, there 
are many types of plastics and additives to plastics that could 
potentially be released during digestive processes in marine 
vertebrates. 

Results of this study show that raw material plastic pellets 
are becoming increasingly more common in areas of the Pacific 
Ocean far removed from industrial sources. Although tar balls 
were once considered more common and widely distributed 
than marine plastics, their occurrence in the present stud) was 
less than that of plastics even in areas where oil development 
and tanker traffic is hea\ \ . Plastic pellets do have the potential 
to adsorb certain organic contaminants from seawater: however, 
the types of compounds that can be adsorbed and possibly 
concentrated is not well understood. Polyethylene pellets 
subjected lo conditions simulating the avian digestive system 
did not release detectable levels of chemicals However, other 
types of plastics, and those that contain additives, were not 
tested. 

Clearly, fish, seabirds, and other marine organisms will 
continue to be exposed to plastics at increasing rates. Regulations 
prohibiting ocean dumping of plastics have already been 
enacted: however, recycling, waste management, degradable 
plastics, and other alternatives must continue to he developed 
and implemented to abate the global problem of plastics in the 
oceans of the world. 



|0d 



This project was part of the First Joint US-USSR Central 
Pacific Expedition aboard the Soviet Research Vessel, Akademik 
Korolev. We express our appreciation to the US Fish and 
Wildlife Service. USA, and the State Committee for 



Hydrometeorology. USSR, who made our participation possible. 
We also thank H. O'Connor, A. Tsyban, O. Rostovsev, Y. Volodkovich, 
B. Aleksandrov, L. Polishchuk, S. Kohl, A. Krynitsky, M Wong, and 
P. Wills. 



2.6 The Role of Solar Irradiation in the 
Oxidative Transformation of 
Benzo(a)pyrene 

NATALYA I. IRHA, EHA R. URBAS. and UVE E. K1RSO 
Institute of Chemistry, Estonian Academy of Sciences, Tallinn, ESSR 



Introduction 

Organic pollutants entering water bodies, including the 
oceans, contain a wide range of substances. Some of these are 
quite easily subjected to degradation under the action of natural 
agents, while many other compounds are stable and persist in 
various compartments of the ecosystem. Among the persistent 
compounds are the carcinogens belonging to the polycyclic 
aromatic hydrocarbon (PAH) group. It has been established 
that PAH' s enter the ocean in several ways, emissions into the 
atmosphere as a result of incomplete combustion of fuels, 
spillage during transportation of oil, and so forth. On the other 
hand, it is known that there exists a natural occurrence of 
PAH's (natural background) in the oceans. It is the view of 
several researchers (Tsyban, 1975; Izrael. 1984; Kirso et ai. 
1988) that the main mechanism of self-purification of the 
hydrosphere from carcinogenic PAH' s is biological ( bacterial ) 
oxidation. However, there are also other mechanisms for 
removal, including photooxidation under solar irradiation ( Mill 
etai, 1981; Bockris, 1982). It may be assumed that this process 
plays a significant role, especially in the surface water layer 
where the content of soluble oxygen is near \00 c /c saturation 
(i.e., 10 '■' mol/1). There are numerous factors that modify this 
process, including the physicochemical characteristics of water 
( turbidity , transparency, and the presence of other compounds ) 
and the air above the sea surface (condition of the atmosphere, 
i.e., environment and weather conditions). Figure 1 illustrates 
the characteristic sunlight spectra at the sea surface (Ranby & 
Rabek. 1978). Photochemical reactions in the marine 
environment are also moderated by the level of atmospheric 
pollution with aerosol particles and smoke, as well as changes 
in ozone and other impurities. Solar irradiation at wavelengths 
less than 285 nm is to a large extent absorbed by ozone in the 
upper layer of the atmosphere. Therefore, the active spectra for 
this process are usually at wavelengths greater than 285 nm. 
Variations in the characteristics of the ozone layer therefore 
influence all the photochemical processes occurring on the 
oceans' surface. 



1000 



E 
o 


200 


3. 




u 


50 


o 




m 


10 


S-4 

n. 





UJ 



0.5 



0.1 




ISO 



320 



360 



400 



X,,nm 

Fig. 1 . Seasonal variations in solar spectra at the sea surface. I = July, 

2 = December. E = Spectral Energy; A. = Wavelength (Ranby & 
Rabek. 1978). 

The extent of these processes and their contribution to self- 
purification of the marine environment from carcinogenic 
PAH's are determined by both the value and distribution of 
solar irradiation energy (Mill et al, 1981; Rabek, 1985)andthe 
level of pollution of seawater, as well as the concentration and 
composition of pollutants. The latter may have an influence 
upon the pattern and intensity of the degradation processes of 
PAH's (Kirso & Gubergrits, 1971; Gubergrits et al.. 1975). 
Although systematic studies have been carried out on the 
photooxidation of individual PAH's in water (Kirso et ai, 
1971; Gubergrits etai, 1975; Paalme etai, 1976. 1983). there 
are few data on these processes under natural conditions. 



197 



Therefore, during the cruise of the R/V Akademik Korolev 
(July-November 1 988 ). a study was undertaken on the kinetics 
of the oxidative photolysis of a typical carcinogenic PAH — 
benzol a tpyrene (BaP) — (Clar, 197 1 ) in the Bering Sea and the 
tropical waters of the Pacific (Table 1 ). 

It is known that the Bering Sea is almost wholly in the 
subarctic zone, excluding its northern parts, which are in the 
arctic temperature zones (Izrael & Tsyban, 1987). The main 
body of its waters is characterized by a subarctic structure 
whose specific feature is the existence of cold and warm 
intermediate layers. The upper layer thickness average 
25-50 m, the salinity being 32.8-33.4%o and the temperature, 
about 5 to 7°C. According to Izrael and Tsyban ( 1 987), PAH" s 
are permanent and typical components of these ecosystems. 

TABLE 1 

Exposure of BaP solution in seawater (the 47th cruise of the 
R/V Akademik Korolev). 



Month 


Exper. 


Coordinates 


Water 


Average dose of 


(1988) 


Number 




Temp. 
t=°C 


solar radiation 
Q MJ/m : during 
the first 3 hours 



Aiiyust 





1 


53°58'N/I76°28'W 


14.1 


1.48 




2 


53°58'N/176°28 , W 


15.7 


1.09 


September 












3 


09°54'S/156°23 , W 


26.4 


3.50 




4 


04°02'S/154°07'W 


27.0 


3.09 




5 


09°59'S/150°15'W 


27.0 


4.90 




6 


()9°()0'S/I50°0()'W 


27.0 


3.67 


October 












7 


09°00 , S/150°00 , W 


27.0 


2.70 




8 


09°0()'S/150°0()'W 


27.0 


3.00 



Taking into account the low influence of this area from 
human activities, the physicochemical parameters of the 
atmosphere above these waters and characteristics of the surface 
water layer, a study of sunlight photolysis of PAH' sin seawater 

at lower temperatures and low intensity of solar irradiation was 
of interest. The same experiment was carried out under tropical 
conditions — that is. at higher temperatures and considerable 
insolation with a salinity of 35.4%o. 

Methods and Materials 

The surface water samples taken at a depth of to 0.5 m 
were sterilized by autoclaving for 2 h. cooled, and filtered 
through main layers of Sterilized cloth into a sterilized dish. 
The BaP solution in seawater was prepared in 5-1 glass cylindrical 
reactors whose sides were covered with black paper (water 
column height 20 cm). Benzol a (pyrene was introduced as an 
ethanol solution (ethanol content up to 0.01295 ) by constantly 
mixing with a magnetic stirrer for 15 minutes. The BaP 
concentration was varied from 0.7 to 6.6 nanomoles mm). 



This solution was exposed to sunlight at the temperatures of the 
surface water for each given region. Simultaneously, the total 
solar irradiation dose, Q (MJ/m : per h), was measured 
pyranometrically (Table I). The experiments were carried out 
with the following solutions: 

(a) BaP solution in seawater; 

(b) BaP solution in sterilized seawater; and 

(c) BaP solution in sterilized seawater protected from 
light by black paper (autoxidation). 

The exposures lasted for approximately 3 h in the first half 
of the day. The solutions were periodically sampled and the 
BaP concentration measured by chromatography techniques. 
Conditions of analysis were as follows: 

A 'Jasko' HPLC chromatograph (Japan), a 
fluorimetric detector, X tx = 295 nm, X cm = 408 
nm, solvent methanol-water (95:5) isocratic 
regime, eluent flow-rate 0.7 ml/min using a 
25 x 0.25 cm ODS column, sample volume was 
100 ul. 
Statistical kinetic data processing was performed using the 
least-squares method. 

Results and Discussion 

From kinetic data (Figs. 2-5. Table 2), it follows that 
during the first hour of exposure, a decrease in the BaP 
concentrations in seawater is described by a formal-kinetic 
equation for the first-order reaction where c and c, are the 
initial BaP concentration at zero time, and that at a certain time 
t, k is the constant of the first-order reaction, the dimensions for 
this constant are per second (s '). 

ln(c,/c,) = kt 

The rate constant values obtained for the Bering Sea areas, 
the tropical part of the Pacific and the Caroline Atoll (Table 2) 
are of the same order as those found (Mill et <//.. 1981) by 
photolysis of BaP in water (X = 366 nm. BaP concentration 
5x 10 s m, t° = 22-28°C, k = 3.86 ±0.71 x 10V). The half 
decay time of BaP under these conditions was 0.69 h. In our 
experiment the average rate constant value per unit Q was 
(1.1 ±0.1) 10V. 

Irrespective of the areas under study and the sterility 
of water in those treatments that were subjected to 
photolysis (Table 2). the initial stage of the process, as 
mentioned above, is well described by an equation for the first 
order kinetics. At the same time, according to experimental 
data (Figs. 2-4), the oxidation of BaP in seawater 
without sunlight (autoxidation) for the Bering Sea areas and 
the tropical pari of the Pacific is described by a formal- 
kinetic of a second-order type (r = 0.90-0.98). The rate 
constant value of BaP autoxidation in the Bering Sea water 
at a concentration of 5 x 10 s M (July, t =21°C, 
Q,, r 0.68 MJ/m" per h) was (0.4 ± 0.01) 10 4 M's ! . but in 
the tropical part at a concentration of 1.7 x 10 s (October 
t°=27°C. Q r ,,~3.00 MJ/nr per hi (9.08 ± 1.90) 1() 4 M 's '. 
respectively. 



198 



TABLE 2 

Kinetic characteristics of photochemical transformation of 
BaP in seawater under solar irradiation in the following areas: 

(a) the Bering Sea 

(b) the tropical part of the Pacific 

(c) the Caroline Atoll 

(47th cruise of the R/V Akademik Korolev, July-November 19881. 



Experiment 


Initial BaP 


Rate 


Number of 


Correlation 


Sterility of 


No. See 


concentration 


constant 


data points 


coefficient 


media 


Table 1 


M 10 K 


10V'(k±n) 




r (first-order) 








(a) 








1. 


1.47 


1.69±0.13 


5 


0.99 


- 


2. 


4.20 


1.60 ±0.08 


7 


0.99 


+ 




4.44 


1.20 + 0.16 
(b) 


6 


0.97 


- 


3. 


6.59 


2.84 ±0.48 


5 


0.96 


+ 




2.85 


2.85 ±0.98 


4 


0.90 


- 


4. 


2.06 


1.60 ±0.49 

(C) 


4 


0.92 


+ 


5. 


1.90 


2.99 ±0.70 


4 


0.95 


- 




2.14 


4.05 ±0.51 


4 


0.98 


- 


6. 


0.70 


4.20 ±0.11 


4 


0.99 


+ 


7. 


11.10 


8.77 ±0.23 


4 


0.99 


- 


8. 


2.00 


3.67 ±0.27 


6 


0.99 


- 



C/Co 



c 
o 



c 

— 

c 

<u 
o 

c 
o 
o 

a. 

CO 



1.01 


s*-.- 


-•- 


'* 


► • 


"c 




6 






0.5 




O 


/ 


3 




v a.a 








b * 



o 







? 



t .hour 



Fig. 2. Kinetics of BaP degradation under sunlight irradiation at the surface 
of the Bering Sea (coordinates: 67°42'N 1 15°43'W): a) in sterilized 
seawater: b) in non-sterilized seawater; and cl by autoxidation (w/o 
light). 

Thus, in the areas of exposure the rate constant values of 
BaP autoxidation in seawater differed within one order of 
magnitude. The shape of the kinetic curves (Figs. 2-4) also 
gives evidence of a negligible decrease in BaP concentration 
over time without sunhght. 

While in laboratory experiments by photo-initiated 
UV-irradiation within a relatively wide wavelength range — 
around 200 nm — and a high oxygen atmosphere in a liquid 
medium, the degradation of BaP did not change until the 



C/Co 
1.0 



c 
o 



•5 0.5 



c 
o 
o 

(X 

a 

CO 



Fig. 3. 



p 
© 




c 


4 

3 

2 


G)"n 




b © 

-n - 






©0 




~© 

OO-- 
a ~ 


1 




©u 









o 







9 



t .hour 



Kinetics of BaP degradation under sunlight irradiation at the surface 
ofthe Bering Sea (coordinates: 53°58'N 175°28'W): a) in sterilized 
seawater: b) in non-sterilized seawater; and c) by autoxidation 
(w/o light). 



disappearance of the initial substance (zero-order reaction) 
(Gubergritse/fl/., 1975: Paalme etal, 1983). Under natural 
conditions in water (i.e., by sunlight photolysis), a decrease in 
the reaction rate is observed after the first hour of exposure 
(Figs. 2-5 ). This is especially noticeable in the Bering Sea area 
( Figs. 2-3 ). From the graphs ( Figs. 2-5 ), it follows that the BaP 
half-decay times (t" : ) in the Bering Sea waters were not less 
than 1 h. but in the tropical part ofthe Pacific, they were 0.6 h. 
The initial rate constant value of sunlight photolysis of BaP in 
the tropical zone water averaged 8.9 x 1 O 12 mole s ', exceeding 



199 




12 3 4 t, hour 

F ; ig. 4. Kinetics of BaP degradation under sunlight irradiation in the tropical 
part of the Pacific (coordinates: 09°54'N I56°23'W): a) in sterilized 
seawater; b) in non-sterilized seawater; and c) by autoxidation (w/o 
light). 




2 t, hour 



Fig- 5. Kinetics of BaP degradation under sunlight irradiation in the lagoon 
of the Caroline Atoll (coordinates: 09°54'N I5fv2.vwi: al j„ 
sterilized seawater; bi in nonsterilized seawater: c) by autoxidation 
lu/o light). 

by x3 the rate tor the Bering Sea area at similar initial 
BaP concentrations (2.44 x 10 i: mole s '). For comparison, 
it may he pointed oui thai at the initial concentrations of 10 " M, 
the initial rate of sunlight photolysis of BaP in the Caroline 
Atoll is much higher than that found for BaP autoxidation 
in water under laboratory condition 1.67 x 10 i: and 
0.3 x 10 - |: mole s '. respectively (Kirso et al., 1983). 

Thus, experimental data, as expected, gave evidence of a 
dependence of BaP degradation on the location of exposure, 
intensity of solar UV-irradiation, and temperature of the 
environment, and data agreed well w ith the results obtained by 
Graupera and associates (Graupera et al., 1988). Obviously, 
during photolysis of BaP by sunlight in seawater, the existence 
ofmicroimpurities, inorganic components, salinity, and general 



water composition all have a major impact on photochemical 
reactions in different areas of the world oceans. A comparison 
of experimental kinetic parameters and literature data (Mill 
et al., 1981 ) suggests that proceeding from the values of k at 
T"- of BaP under similar conditions of its photolysis in water 
(X = 366 nm and concentration of 5 x 10 s M) the value of 
quantum yield evp (the number of molecules subjected to 
transformation as a result of adsorption of one energy quantum ) 
is almost equal to <p - 5.4 x 10" 4 — that is, less than one (Mill 
et al.. 1981). Consequently, the processes under study are 
complex, involving competing chemical reactions. Thus, the 
degradation of BaP is initiated directly or indirectly and proceeds 
under solar irradiation. 

Photochemical reactions were investigated with soluble 
oxygen in the presence of different inorganic and organic- 
components in water bodies. Zafiriou (Zafiriou et al.. 1989) 
presented the following scheme for generation of free radicals 
in the marine environment: 

O, + hv > initiation of radicals R' 

R- + R,NO > RNO\ R : NOH, R : NOR 

seawater + O, + hv - - > initiation > oxygen con- 
taining radicals 

seawater + : + hv > superoxide 

0.+ NO - - >00 * NO* H + - -> * NO, - -> 2 

seawater + *0 ; + hv — > H * O * OH product 
seawater + H * O * OH + hv - - -> *0 2 + H : *0 

According to (Mopper et al.. 1989) the concentration of 
high-energy ( more than 4 kcal/mole ) oxygen-containing radicals 
in seawater is low and makes for hydroxyl groups (OH) — for 
example, in subtropical coastal areas 1 1.9 x 10 ls M and for 
open sea 1.1 x 10 IB M, correspondingly. Consequently, it may 
be assumed that BaP (and other PAH's) is subjected to 
photodegradation due mainly to secondary photochemical 
reactions w ith different reactive radicals formed directly under 
the action of light quanta or indirectly (see the scheme). 
According to our results, the rate of BaP transformation depends 
primarily on the sunlight intensity. Obviously, then, the 
mechanism of photochemical oxidation of organic xenobiotics 
of the PAH type is not different in northern and southern areas 
(i.e., the amount of oxidizing particles sufficient for 
transformation is generated whose excess favors the first-order 
reaction [pseudomonomolecular] relative to BaP). 

A study of the influence of inorganic salts and micro- 
impurities on the photochemical processes in the marine 
ecosystems requires further research. 

To sum up, it should be pointed out that the experiments 
(under natural conditions) were carried out with only one 
reference PAH. ben/o( a)pyrene, and in the marine environment 
many other PAH"s are present (see Fig. 6). To estimate the 
reactivities of other PAHs undergoing photooxidation in 
seawater. the data obtained by Paalme et al. (1988) were 
considered. It appears that the rate of the process for individual 
homologs differs by factors of over 140 (Fig. 6). Anthracene 
and its derivatives are easily oxidized, while the more condensed 
systems — for example, coronene — remain more stable. It 
may be assumed that as a result of photochemical oxidation, the 
quantitative ratio of PAH's m the marine environment shifts 
low, ml the heavier homologs. 



200 



a 
□ 
□ 



20 



10 



EL 



.-•• 



PI 

•■■ 

-■■' 
-•■' 



/ 
/ 
/ 
/ 
/ 
/ 
/ 
/ 
/ 
/ 
/ 
,-■ 

.■■' 

,-■ 
,-■ 
/ 

/ 

/ 



BaP N 



K 

Ac 



JZL 



R 



-■•■ 
.••■■ 



Pn 



Bba BaA TF Chr 



^A 



Y\ B B B ^ 



/ 
/ 
/ 
/ 
/ 
/ 

/ 

/ 
/ 
/ 
.-- 
.-- 
/ 
/ 
/ 
/ 
/ 

J 



PI 



5 

C PAH 



/ 
/ 



/ 
/ 



- 3 



Pv 



Per BacA BajA BahA BeP Cor BkF BbF 



Fig. 6. Relative rates (v) of photoinitialed transformation of PAH's in water: (1 ) (Paalme etal., 1983); and their content (ng/1) in the water of the Bering Sea; 
(2); and the Baltic; and (3) (Kirso el at.. 1989). 



Thus, under conditions similar to natural ones (i.e., under 
sunlight), it has been shown that a certain amount of PAH's 
may be subjected to sunlight photolysis. The degree of BaP 
transformation in the experiments was governed by the intensity 



of total sunlight irradiation in a given region and by the 
conditions of exposure (temperature, turbidity, etc.). 
Autoxidation plays an insignificant role in self-purification of 
the marine environment from BaP. 



201 



Chapter 2 References 



Balazs, G. (1985). Impact of ocean debris on marine turtles: 
entanglement and ingestion. In Proceedings of the Workshop 
on the Fate and Impact of Marine Debris, 27-29 November 
1984, Honolulu, Hawaii (R. S. Shomura & H. O. Yoshida, 
eds.), pp. 387-429, NOAA Technical Memorandum NMFS. 
NOA ATM-NMSF-SWFS-54, US Department of Commerce. 

Ballschmitter, K. & Zell. M. ( 1 980). Analysis of polychlorinated 
biphenyls (PCB's) by capillary gas chromatography. 
Frensenius ' Z. Anal. Chem. 302, 20-3 1 . 

Bockris. J. O. (1982). Environmental Chemistry: Moscow. 
Khimiia Publishers, Moscow, 670 pp. (Russian translation) 

Carpenter, E. J., Anderson, S. J.. Harvey, G. R.. Miklas, H. P. 
& Peck. B. B. (1972). Polystyrene spherules in coastal 
waters. Science 178, 749-750. 

Carr, A. ( 1 987 ). Impact of nondegradable marine debris on the 
ecology and survival outlook of sea turtles. Mar. Pollut. 
Bull. 18, 352-356. 

Chernyak. S. M„ Kolobova, T. P. & Vronskaya, V. M. ( 1985). 
A study of chlorinated hydrocarbons. In Studies of the Baltic 
Sea Ecosystem, No. 3. Gidrometeoizdat Publishers, 
Leningrad, (in Russian) 

Chernyak, S. M. & Mikbaleva. I. M. (1985). Chlorinated 
hydrocarbons in various media constituting the Baltic Sea 
environment. In Studies of the Baltic Sea Ecosystem, No. 2. 
Gidrometeoizdat Publishers, Leningrad, (in Russian) 

Chernyak, S. M., Vronskaya. V. M. & Kolobova, T. P. (1989). 
Chlorinated hydrocarbons. In Studies of the Bering Sea 
Ecosystem, No. 2. Gidrometeoizdat Publishers, Leningrad. 
(in Russian) 

Chernyak, S. M., Vronskaya. V. M. & Kolobova, T. P. (1992). 
Migratory and bioaccumulative peculiarities in the 
biogeochemical cycling of chlorinated hydrocarbon. In 
Results of the Third Joint US-USSR Bering & Chukchi Seas 
Expedition (BERPAC). Summer 1988 (P. A. Nagel, ed.). US 
Fish and Wildlife Service. Washington, D.C. 

Clar, E. (1971). Polycyclic Hydrocarbons, Vol. 2. Academic 
Press, London. 

Colton, J. B. Jr., Knapp. F. D. & Burns, B. R. ( 1974). Plastic 
particles in surface waters of the northwestern Atlantic. 
Science 185,491-197. 

Connors, P. G. & Smith. K. G. ( 1982). Ocean plastic particle 
pollution: Suspected effect on fat deposition in red phalaropes. 
Mar. Pollut. Bull. 13. 1 8-20. 

Day, R. H. & Shaw. D. G. (1987). Patterns in the abundance 
of pelagic plastic and tar in the North Pacific Ocean. Mar. 
Pollut. Bull. 18.311-316. 

Day. R. H.. Wehle, D. H. S. & Coleman, F. C. (1985). Ingestion 
of plastic pollutants by marine birds. In Proceedings of the 
Workshop on the Fate and Impact of Marine Debris, 
pp. 27-29, November 1984, Honolulu, Hawaii (R. S. Shomura 
& H. O. Yoshida, eds.). pp. 387-429, NOAA Technical 
Memorandum NMFS, NOAA-TM-NMFS-SWFS-54, US 
Department of Commerce. 



Fedoseeva, G. E. & Khesina, A. Ya. (1968). The use of 

quasilinear luminescence spectra for the quantitative 

determination of a series of polycyclic hydrocarbons. 

J. Appl. Spectroscopy, Vol. 9 2,282-288. (in Russian) 
Fry, D. M., Fefer, S. I. & Sileo, L. (1987). Ingestion of plastic 

debris by Laysan albatrosses and wedge-tailed shearwaters 

in the Hawaiian Islands. Mar. Pollut. Bull. 18. 339-343. 
Furness, R. W. ( 1985). Plastic particle pollution: Accumulation 

by procellariiform seabirds at Scottish colonies. Mar. Pollut. 

Bull. 16, 103-106. 
Graupera. B. M., Braunera, M. S., Gonzalez, R. C, Conde, 

I. A. P. & Solionova, L. ( 1988). Studies on the degradation 

of benzo(a)pyrene by solar ultraviolet light and temperature. 

Rev. Cubana Hig. Epidemiol. 2691. 106-1 16. 
Gubergrits, M. Ya., Kirso, U. E. & Paalme, L. P. (1975). 

Transformation of Carcinogenic Compounds in the 

Hydrosphere. Znanije Publishers, Moscow, (in Russian) 
Irha, N. I., Urba, E. R. & Kirso. W. V. ( 1992). Distribution of 

PAH's. In Results of the Third Joint US-USSR Bering & 

Chukchi Seas Expedition (BERPAC), Summer 1988 (P. A. 

Nagel, ed.). US Fish and Wildlife Service, Washington. 

D.C. 
Izrael, Yu. A. (1984). Ecology and Control of the State of 

Environment. Gidrometeoizdat Publishers, Moscow, (in 

Russian) 
Izrael, Yu. A. & Tsyban, A. V. (1987). Comprehensive 

Analysis of the Ecosystem of the Bering Sea. Gidrometeoizdat 

Publishers. Leningrad, (in Russian) 
Izrael, Yu. A. & Tsyban. A. V. ( 1989). Ocean Ecology under 

Human Impact. Gidrometeoizdat Publishers, Leningrad, 

527 pp. (in Russian) 
Kimball, W. H. & Munir, Z. A. ( 1971 ). The corrosion of lead 

shot in a simulated waterfowl gizzard. J. Wildl. Manage. 2, 

360-365. 
Kirso, U., Belyhk, L. & Stom, D. (1983). Kinetics of 

benzo(a)pyrene oxidation in aqueous solution. Oxid. Comm. 

4(1-4), 35-44. 
Kirso, U. & Gubergrits, M. ( 197 1 ). Kinetics of oxidation of 

3,4-benzpyrene. phenol and 5 methylresorcinol activated by 

UV-irradiation. Proceedings of Estonian Academy of 

Science, Chemistry, Geology 20(2). 134-139. (in Russian) 
Kirso, U., Paalme, L., Voll. M., Urbas, E. & Irha, N. ( 1990). 

Accumulation of carcinogenic hydrocarbons at the sediment 

water interface. Marine Chemistry 30. 337-341 . 
Kirso. U., Stom, D.. Belykh. L. & Irha, N. (1988). 

Transformation of Carcinogenic and Toxic Substances in 

the Hydrosphere. Valgus Publishers, Tallinn, (in Russian) 
Laist. D. ( 1 987 ). Overview of the biological effects of lost and 

discarded plastic debris in the marine environment. Mar. 

Pollut. Bull. 18.319-326. 
Lee, M. L., Milos. V., Novotny. K. & Bartle. D. (1987). 

Analytical Chemistry of Polycyclic Aromatic Compounds. 

Academic Press, New York. 5. 441—449. 



203 



Mamotov, G. & Wehry, E. (1988). Chemical reactivity of 
polycyclic organic compounds absorbed on coal fly ash and 
related solid surface final report. Energy Res. Abstr. 

Medinets, V. I.. Solov'ev, V. G. & Glebov, B. V. (1992). 
Cesium- 137 in the seawaters. In Results of the Third Joint 
US-USSR Bering & Chukchi Seas Expedition (BERPAC), 
Summer 1988 (P. A. Nagel. ed.). US Fish and Wildlife 
Service, Washington, D.C. 

Mill, T., Mabey, W. K., Lan, B. Y. & Baraze. A. (1981). 
Photolysis of polycyclic aromatic hydrocarbons in water. 
Chemosphere 10(11/12). 1281-1290. 

Mopper. K. & Zhou, X. ( 1989). Photochemical production of 
hydroxy radicals at the sea surface: implications of oceanic 
carbon cycling. 32nd IUPAC Congress, Stockholm. Bookof 
Abstr. p. 2. 

Moms, R. J. (1980). Plastic debris in the surface waters of the 
South Atlantic. Mar. Pollut. Bull. 11, 164-166. 

Nagaya. Y. & Nakamura, K. (1985). A study of certain 
radionuclides deposited in the deep waters of the Pacific 
through radioactive precipitation. In Comprehensive Global 
Monitoring of the World Ocean. Proceedings of the First 
International Symposium. 2-10 October 1983. Vol. 2 
( Yu. A. Izrael, ed. ). pp. 429—44 1 . Gidrometeoizdat Publishers, 
Leningrad, (in Russian) 

Niassat, P., Echardt, I. & Ottenwalder. I. (1968). Presence of 
benzo3, 4pyrene in the waters of the Clipperton Atoll lagoon. 
Comp. Rend. Acad. Sci. Paris. Series D. 267, 1771-1774. 
(in French) 

Paalme. L. & Gubergrits, M. (1976). Kinetics of separate and 
co-photodegradation of benzo(a)pyrene and 3- 
methylchloranthene. Proceedings of Estonian Academy of 
Science. Chemistry, Geology 25(4), 271-275. (in Russian) 

Paalme. L., Uibonuu, H.. Gubergrits, M. & Jacquignon. P. 
(19831. The initiated oxidation of polycyclic arenes in 
aqueous solution. Oxid. Conun. 4( 1-4). 27-34. 

Pruter, A. T. ( 1987). Sources, quantities and distribution of 
persistent plastics in the marine environment. Mar. Pollut. 
Bull. 18.305-310. 

Rabek, J. F. ( 1982). Experimental Methods in Photochemistry 
and Photophysics. Vol. 1, Mir Publishers. Moscow. (Russian 
translation) 

Ranby. B. & Rabek. J. F. (1975). Photode gradation, 
Photooxidation and Photo-stabilization of Polymers. Mir 
Publishers. Moscow. (Russian translation) 

Ryan. P. G. ( 1 987). The effects of ingested plastic on seabirds: 
Correlations between plastic load and body condition. 
Environ. Pollut. 46, 1 19 125. 

Ryan, P. G. ( 1988a). Plastic ingestion and PCB's in seabirds: 

Is there a relationship? Mar. Pollut. Bull. 19. 174-176. 
Ryan. P. G. (1988b). Effects of ingested plastic on seabird 
feeding: Evidence from chickens. Mar. Pollut. Bull. 19. 
1 25- 128. 
Shaw, D. G. ( 1977). Pelagic tar and plastic in the Gulf of 
Alaska and Bering Sea: l l )75. Sci. Total Environ. 8. 13-20. 
Shaw, D. G. & Mapes. G. A. ( 1979). Surface circulation and 
the distribution of pelagic tar and plastic. Mar. Pollut. Bull. 
10. 160-162. 



Shilina. L. I. (1982). In Integrated Global Monitoring of 
Natural Environmental Pollution, pp. 238-242. 
Gidrometeoizdat Publishers, Leningrad, (in Russiani 

Shpolsky. E. V.. Il'in, A. A. etal. ( 1952). Fluorescence spectra 
of coronene in frozen solutions. Dokladny AN SSSR 1 . 
935-939. (in Russian) 

Tanabe, S., Tatsukawa. R.. Kawano. M. & Hidaka. H. ( 1982). 
Global distribution and atmospheric input of chlorinated 
hydrocarbons in the western Pacific, eastern Indiana and 
Antarctic Oceans. J. Oceanogr. Soc. Jpn.. Vol. 38. 

Tanabe, S. (1985). Distribution, behavior and fate of PCB's in 
the marine environment. J. Oceanogr. Soc. Jpn.. Vol. 41. 

Tsyban, A. V. (1975). Bacterioneuston and problems of 
degradation in surface films of organic substances released 
into the sea. Prog. Water Technol. 7(3/4). 792-799. 

Tsyban, A. V., Khesina, A. Ya., Ventzel, M. V., Volodkovich, 
Yu. L. & Korsak. M. N. el al. (1985a). Occurrence and 
microbial degradation of the carcinogenic hydrocarbon. 
BaP. in open waters of the Baltic Sea. In Studies of the Baltic 
Sea Ecosystem, pp. 218-234. Vol. 2. Gidrometeoizdat 
Publishers. Leningrad, (in Russian) 

Tsyban. A. V., Panov. G. V.. Ventzel, M. V.. Volodkovich. 
Yu. L. & Korsak, M. N. ( 1986). Long-term environmental 
studies on impacted and background regions. In Integrated 
Global Monitoring of the Status of the Biosphere. 
pp. 45-60. Gidrometeoizdat Publishers, Leningrad, (in 
Russian) 

Tsyban, A. V.. Volodkovich. Yu. L. & Belyaeva. O. L. (1987). 
A study of the circulation of benzo(a)pyrene in the Bering 
Sea. In Comprehensive Analysis of the Bering Sea 
Ecosystem, pp. 218-232. Gidrometeoizdat Publishers. 
Leningrad, (in Russian) 

Tsyban. A. V., Volodkovich. Yu. L., Panov, G. V.. Khesina. 
A. Ya. & Yemakov. Ye. A. (1985b). Distribution and 
microbial oxidation of carcinogenic hydrocarbons in some 
regions of the World Ocean, with benzol a Ipyrene as the 
model compound. In Environmental Consequences of 'Ocean 
Pollution, pp. 88-112. Gidrometeoizdat Publishers. 
Leningrad, (in Russian) 

Vakulovskiy. S. M. (ed.) (1986). Methodological 
Recommendations for the Determination of Radioactive 
Contamination of Waters. Gidrometeoizdat Publishers. 
Moscow. 78 pp. (in Russian) 
Van Franeker, J. A. (1985). Plastic ingestion m the North 
Atlantic fulmar. Mar. Pollut. Bull. 16. 367-369. 

Winston, J. E. ( 1982). Drift plastic — An expanding niche for 

a marine invertebrate' Mar. Pollut. Bull. 10. 348-351. 
Wolfe. D. A. ( 1 987 ). Persistent plastics and debris in the ocean: 
An international problem of ocean disposal. Mar. Pollut. 
Bull. 18. 303-305. 
Wong. C.S.. Green. D.R.&Cretney. W. J. (1 974 (.Quantitative 
tar and plastic waste distributions in the Pacific Ocean. 
Nature 247. 30-32. 
Zafiriou, O. C, Dister, B.. Moffett. J. M.. Micinski, E. & 
Blough, N. V. (1989). Photochemical production of free 
radicals in the marine environment. 3 2nd IUPAC Congress. 
Stockholm. Book ofAbstr. p. 201. 



204 



Chapter 3: 

BIOLOGICAL INVESTIGATIONS IN 
THE CENTRAL PACIFIC 



Editors: 

CAMERON B. KEPLER & 
MIKHAIL N. KORSAK 



3.1 A Description of Bacterioplankton 

VASSILIY M. KUDRYATSEV, VLADIMIR O. MAMAEV. and NADEZHDA V. SUKHANOVA 

Institute of Global Climate and Ecology, State Committee for Hydrometeorology and Academy of Sciences, Moscow, USSR 



Introduction 

An assessment of the role played by bacteria in 
biodegradation processes occurring in the World Ocean requires 
data concerning bacterial population counts, bacterial population 
distributions, and a number of other functional characteristics. 

Studies carried out in the course of the First Joint 
US-USSR Central Pacific Expedition in the equatorial Pacific 
and the South China Sea produced new findings characterizing 
the present state of microbiocoenoses in this part of the World 
Ocean. 

Materials and Methods 

The microbiological studies described below were 
conducted using methods set out in several handbooks 
(Romanenko& Kuznetsov, 1974;Tsyban, 1 980; Tsyban etal.. 
1988). Analyses of total counts, biomass, bacterioplankton 
production, indicator group distribution, and degradation 
process rates were performed at 1 8 stations. The samples were 
taken from 6 to 19 depths using 5-1 Niskin bottles. To allow for 
overall bacterial counts, samples of 20-50 ml were passed 
through "Synpor 8" membrane filters with pore size 0.23 urn. 
The filters were desalinated onto filter paper moistened with 
distilled water, dried, and dyed with a 5% carbolic erythrosine 
solution. The bacteria deposited on the filters were counted by 
direct oil-immersion microscopy (xl350 magnification. 20 
visual fields). 

The average bacteria volume was assessed by measuring 
bacterial cell size with the aid of an ocular ruler. The mean 
bacterial cell volume was found to be 0.3 urn'. 

The daily bacterial biomass production rate was calculated 
on the basis of CO : assimilation in darkness. The latter was 
ascertained using a radiocarbon technique (Romanenko. 1964; 
Sorokin. 1971a). The determinations were made in 
1 00- 1 20-ml jars. The radioactivity of the working solution of 
carbon-labeled sodium carbonate introduced into the sample- 
containing jars was 18 x 10 6 counts/min. The samples were 
incubated in darkness for 24 h at the temperature of the water 
where the sample was taken. Once the incubation was 
completed, the samples were fixed with a 40% formaldehyde 
solution, then passed through "Synpor-7" filters (pore size 
0.35 (tm). The radioactivity of the bacteria deposited on the 
filters was measured using an '•Intertechnique'" liquid 
scintillation counter. The scintillation cocktail was ZhS-8. 
The CO, assimilation rate in darkness (T ass ) for the bacterial 
plankton was calculated by means of the formula 

T — '* x Qart 

± ass ' 

R xt 



where r the radioactivity of the bacterial cells on 

the filter (counts/min); 
the hydrocarbonate content of seawater 
(mg/1 ) determined by direct titration of 
0. IN HC1 in the presence of methyl 
red; 
— the radioactivity of the isotope Na l4 CO, 
used in the experiment (counts/min); 
t the incubation time (h). 

The bacterial biomass production was obtained by 
calculation, setting P h = T MS x 16.6. Bacterial plankton 
respiration was determined by applying the formula 



C„ 



k 



D = 1J1, 



where D 



— the amount of oxygen expended on decom- 
position (mg 1 ' d '); 
T ass — rate ofCO : assimilation in darkness (|ig of 

C perl" 1 d '); and 

the coefficient of 7 is the ratio of oxygen uptake to 

CO, assimilation. 

The studies were carried out in the tropical and equatorial 

portions of the Pacific Ocean, as well as in the South China Sea. 

The locations of the stations are indicated on the Frontispiece. 

Results 

The equatorial Pacific is characterized by upwelling. The 
latter occurs mostly along the boundaries of west-to-east zonal 
flows and alternates with surface-water downwelling zones. 
The intrusion of deep waters rich in biogenic elements into the 
euphotic layer determines the way in which biocoenoses develop 
(i.e., their spatial and trophic structures, productivity, and other 
functional characteristics) (Vinogradov, 1978). 

Some temperature stratification was evident during the 
period of our studies in the 0-8°N, 1 60- 1 80° E rectangle of the 
equatorial Pacific. The water temperature and dissolved oxygen 
content remained virtually constant down to a depth of 150 m. 
The parameters in question declined rapidly below this level. 
however, which necessarily affected the formation and 
distribution of microbiocoenoses. Results of analyses 
(Figs. 1-3) indicate that bacterioplankton counts and biomass 
in the 0-100-m layer varied within fairly broad limits. The 
highest bacterial population density was noted at Station 1 16, 
where the average total count and biomass were 
591 x 10 3 cells/ml ' and 13.29 ug C V d' 1 , respectively. The 
lowest bacterioplankton concentration was noted at 
Station 118, where the average total bacteria count was 
199 x 10 3 cells/ml 1 and the biomass 4.48 ug C V d '. The 
average values of the total bacteria count and biomass in 



207 



the 0-100-m layer of the equatorial Pacific were 
360 x 10 3 cells/ml 1 and 8.10 ug C 1 ' d ', respectively. Such a 
level of development of bacterioplankton and its biomass is 
typical of the bacterial population density in the euphotic zone 
with background upwelling (Sorokin, 1978). According to 
data reported in the literature, the principal bacterioplankton 
count and biomass maxima in stratified tropical waters are 
usually situated at the upper boundary of the thermocline. The 
formation of the microbiocoenosis in this zone is attributable to 
the arrival of biogenic substances from lower layers, to the 
stable existence of the phytoplankton population in the euphotic 
layer, and to the stable existence of the phytoplankton population 
in the euphotic layer. This in turn ensures that the 
bacterioplankton receives a steady and readily assimilable 
supply of organic matter (Sorokin, 1971b, 1982). 

According to our observations, the vertical distribution of 
microflora (Figs. 1-3) included several regularly observed 
maxima occasioned by the delivery of labile organic matter to 
the zone of maximum phytoplankton synthesis. These lay at 
15-25 m for Stations 115-117, at the upper boundary of the 
thermocline (100-250 m). and at a depth of 1.000-1.500 m 
(Stations 118-120) (i.e., at the boundary of mixing of the 
Antarctic waters). 

The principal bacterioplankton concentration maximum 
lay at a depth of 100 m. The microflora counts and biomass 
were, on the average, 1 .3 times the corresponding values for the 
euphotic zone. The level of microbiocoenosis development in 
deeper waters (at depths of 1,000-2,000 m) was lower. The 
total bacterioplankton counts and biomass value here turned 
out to be lower than at the upper boundary of the thermocline 
by a factor of 1.5. The constraints limiting microfloral 
development in these deeper waters were very probably the 
lower temperature and high pressure. 

There was a clearcut tendency for the bacterial population 
density in the 0.5-250-m layer to decrease in the east-to-west 
direction. Thus, the bacterioplankton counts and biomass at 
Stations 115 and 116, situated in the eastern part of the 
equatorial zone, were, on the average, twice as high as at 
Station 1 20. lying to the west of them. This observation was in 
keeping with the declining intensity of upwelling from east to 
west (Vinogradov. 1978). 

Thus, total bacterioplankton counts and biomass in the 
waters of the equatorial Pacific, which have suffered less of an 
impact from human activities that other parts of the Ocean, are 
fully comparable with those observed for oligotrophic and 
mesotrophic waters. 

As in the quantitative parameters, the functional 
characteristics undergo considerable change as one moves 
from east to west. Data obtained in the course of the present 
study made possible some quantitative assessments of certain 
functional characteristics of microbiocoenoses. such as 
microfloral activity, bacterial biomass production rate, and the 
rate of degradation of organic matter in the equatorial Pacific. 

Analysis of the results (Figs. 1—1) shows that microfloral 
activity in the 0- 1 00-m layer in the equatorial zone varied over 
a broad range and had a tendency to increase in the east-to-west 
direction despite the gradual diminution of bacterial population 







1 T ■ ' ■ ' ' ' 






10 




Li_—— 


."* " '- . 




















\ 


25 














• 3 


♦ 2 


'J 






.. — * — * 


-x— J— « "■ 




£ 45 




* 


♦ 




Q 




"*^i * 






100 












m** 


■ > 


* • 




250 


T— -—"*'"* 






' ? 


• •V"^ 




1000 


■f 






1500 


< , 


• 








~~'~"~i* . 






Station 115 






2 


4 




>: 


e it 





025 




' 




75 1 



0.25 5 075 




X-..' 




,y 



Figs. 1-3. Vertical distribution of: ( 1 ) water temperature in °C; (2) dissolved 
oxygen in mg/l (3) total bacteria count in millions of cells/ml: and 
(4) CO. assimilation by bacteria in darkness. ingCI ' d '. All data 
refer to the equatorial Pacific. 

density. Thus, the values for the amount of CO, assimilated by 
bacteria per day under conditions of darkness averaged 1 .5-2.0 
times higher for Stations 1 19 and 120 than for the more easterly 
Stations 115-116. Both bacterial biomass production and 
organic matter degradation also increased. Whereas the bacterial 
production values in the 0-100-m layer at Stations 1 15 and 
1 Id averaged 15.2 and 16.6 (.tg C 1 ' d"\ the corresponding 
values at Stations I 19 and I 20 were 25.5 and 32.5 ng C 1 ' d '. 
The respiratory uptake of oxygen by microorganisms in the 
western area was virtually double. The rate of organic mailer 
degradation increased from 78.7 HgC l' 1 d' 1 at Station 1 16 to 



208 



in 0-100 m layer 



Station 115 Station 1 16 Siaiton H7 Siaiion 1 18 SlMion 1 19 StMion 120 



.U 



n h _Q 



Suiiiin 1 15 Sluiion I Ift Station 1 17 Suihon I IH Suimn 1 19 SlMion 120 



in 500-2.000 m layer 



D 



Sluiion I IK Suimn ll>) 

I I BjLlcnjI priwluumn 
Klug C/l/duy [v]] Orgunn nulla dcgrjduiii] 



Fig. 4. Bacterial production and organic-matter degradation in the equatorial 
Pacific. 

104.4 jag C 1 ' d ' at Station 1 19. The average rate of bacterial 
organic matter degradation in the0-100-m layer in this part of 
the Pacific equaled 75.5 ug C 1 ' d '. The P/B coefficient for 
microflora in the 0-100-m layer averaged 2.9, which was 
somewhat higher than the value previously obtained by Sorokin 
(1973). The average value of the P/D coefficient was 0.31. 

Maximum microfloral activity in the 0-100-m layer 
occurred at a depth of 25 m. The rate of assimilation in darkness 
here attained a level typical of mesotrophic waters, averaging 
1.94 ugC I'd'. The respiratory oxygen uptake by microflora 
also showed peak values, averaging 0.28 mg/1 — a finding that 
is in good agreement with that of Sorokin ( 1973). The rate of 
bacterial degradation of organic matter in the euphotic zone 
reached maximum levels on the order of 103.9 ug C 1 ' d ' 
(Fig. 4). 

The activity of microflora in the 100-250-m layer was 
somewhat lower compared with the euphotic zone. The rate of 
CO, assimilation in darkness averaged 1.10 |ig C 1' d'. 
Bacterial biomass production was lower than in the top layer by 
an average factor of 1.2. The bacterial degradation rate 
averaged 58.0 Ug C 1 ' d '. The P/B coefficient was 2.2, the 
P/D coefficient, 0.3 1 . 

Bacterial activity diminished sharply at depths exceeding 
250 m. The rate of bacterial biomass production in the 
500-2, 000-m layer was. on the average, almost five 
times lower compared with the upper layers of the water 
column, ranging from 2.5 to 5.5 |lg C 1 ' d ' and averaging 
4.2 ug C 1 ' d "'. The bacterial degradation rate decreased more 
than fivefold compared with that of the euphotic zone. This 
decreased rate of degradation was clearly attributable to lower 
temperatures. The P/B coefficient in the 500-2,000-m layer 
was 0.7. The water column at 500-2,000 m was found to 
include two layers of heightened activity, one at 1,000 m, the 
other at 2,000 m. However, bacterial production and degradation 
rates in these maximum-activity layers did not attain the 
average values characteristic of the top layer. 

The next transect lay along 10°N in the western Pacific 
(Frontispiece). The waters in this area exhibited lower 
productivity as compared with those of the equatorial portion 
of the central Pacific. The level of development of 



microbiocoenoses at most of the stations corresponded to the 
upper limit of productivity for oligotrophic waters. The bacterial 
population density in the 0-100-m layer ranged from 129 to 
545 x 10 1 cells/ml 1 (Figs. 5-7). The highest bacterioplankton 
concentration occurred at Station 123, where the average count 
and biomass reached 387 x 10-' cells/ml ' and 8.7 ug C 1 ' d '. 
The highest microfloral activity was observed at Station 121, 
where the rate of CO, assimilation by bacterioplankton averaged 
1.76 Ug C 1 ' d "', which corresponds to the upper limit for 
oligotrophic waters. The lowest microfloral activity was noted 
at Station 122. The daily rate of CO, assimilation in darkness 
here was 3.4 times less than at the preceding station. 

Bacterial biomass production in the 0- 1 00-m layer ranged 
from 2.1 to 52 ug C 1 ' d ' (Fig. 8). The oxygen uptake due to 
bacterioplankton respiration averaged 0.12 mg/1. The rate 
of organic matter degradation by bacteria lay in the 
27-94-ug C 1' d ' range, with an average value of 
47 ug C 1 ' d '. The P/B coefficient was 2.3. 

The portion of the water column lying below the 
thermocline (150-500 m) exhibited relatively low bacterial 
population levels. The total bacteria counts here ranged from 
109 to 492 x 10' cells/ml', the biomass from 3.9 to 
7.7 ug C 1 ' d "'. The mean bacteria count and biomass for the 
150-500-m layer turned out to be 237 x 10- cells/ml' 
and 5.3 Ug C 1' d ', respectively. The microfloral activity 
was somewhat lower than in the supernatant 100 m of the 
water column. The rate of CO, assimilation by bacteria in 
darkness ranged from 0.12 to 1.36 ug C 1' d\ averaging 
0.62 ug C 1 ' d ' . The most intense microfloral activity occurred 
at Station 125, where the mean CO, assimilation rate reached 
0.78 ug C 1 ' d ' . The lowest rate of 0.47 ug C 1 ' d ' corresponded 
to Station 122. 

The bacterial biomass production rate turned out to be 1 .5 
times lower than in the 0-100-m layer. Oxygen uptake due to 
bacterioplankton respiration in the 150-500-m layer averaged 
0.08 mg/1, indicating a relatively low rate of degradation. The 
degradation rate OB ranged from 25 to 42 ug C 1' d '. with an 
average value of 33 ug C 1 ' d '. 

A trend towards increased bacterial population densities 
was noted in the deeper ( 1 .000-2,000 m) portion of the water 
column. For example, the total bacteria counts at 2,000 m 
for Stations 1 2 1 and 1 22 were much higher than further up 
the water column. Some increase in bacterioplankton 
concentrations was likewise noted at 1,500 m at Stations 123 
and 125. The average total bacterial count and biomass for 
the 1,000-2,000 m layer was 281 x 10' cells/ml' and 
6.3 ug C I ' d '. respectively. These values were quite closely 
comparable with the data for the 0-100-m layer. 

The results showed that microfloral activity in the 
deeper portions of the water column was suppressed by 
low temperatures and high pressures. The rates of CO, 
assimilation by bacteria in darkness for the 1 ,000-2,000-m 
layer ranged from 0.11 to 0.42 ug C 1' d ', averaging 
0.22 ug C 1 ' d '. Bacterial biomass production was, on the 
average. 2.8 times lower than in the 150-500-m layer. 
Oxygen uptake due to bacterioplankton respiration averaged 
0.03 mg/1. The rate of degradation of organic matter was 
minimal, the average value for the layer being 12 ug C 1 ' d '. 
The P/B coefficient was 0.5. 



209 






2 


. 




6 


8 


11 


) 


025 




15 




75 


1( 


p 


10 




ZS 




40 


5 






2 


4 


6 


6 10 





025 




5 


75 t 






Q 


25 




05 


C 


"5 


1 
















' 1 


10 








3 


k 

t 2 


4 « 


xl 


25 






• 




t 




» 


45 

E 






• 




4 




•• 


t 
n 




• 






* 




// 


100 








* 


* 






250 








*' 




^ 




500 

i D00 


» 






• 










• / 








Slanon 12: 




2000 


* I 


. 


i 











24 30 





2 


4 




6 


6 U 





25 




05 




7S 1 



25 5 




1500 
2000 



25 


5 75 1 




J ! 




*"■ i 


• 








3 . 


> 




/ i 








.. . 




' 




/ 


. 


/ 






■ 


^ * f 












,'• ") ) 


















. .■-" 




A '*• 


Slatwn 122 


/ X 




t ' ' 








i • ; 






25 OS 



25 5 




Figs 5-7. Vertical distribution of: 1 1 1 water temperature in C';(2)dissolved 
oxygen in mg/1 (3) total bacteria count in millions of cells/ml; and 
(4) CO assimilation h> bacteria in darkness, in g C 1' d' 1 . All 
daia applj to the western Pacific. 

The bacterioplankton distribution over the water column 
in the western Pacific exhibited some temperature stratification, 
with several bacterial population and acti\ it) maxima present 
(Figs. 5-7). Three bacterioplankton concentration peaks were 
quite clearly in evidence: the first one lay below the /one of 
maximum phytoplanklon synthesis at a depth of 45 m. the 
second above the upper boundary of the thermoeline at a depth 
of 100 ill. and the third at 1.500 m. Elevated microfloral 
acti\ it} was noted in the euphotic /one (10 15 ml. above the 



I nl ml nl 



mil iimi m layei 



Station 121 Station 122 Station 123 Station 124 Station 125 Station 126 



JL nl n I rn I 



Station 121 Station 122 Station 123 Station 124 Station 1 25 



■ to in layei 



in l.000-2.(HMi,n layei 



Station 121 Station 122 Station 123 Station 124 Station 125 

1 I Bacterial production 



D.0, 



Fig. 8 Bacterial production and organic-matter degradation in the western 
Pacific. 

thermoeline. and at a temperature drop boundary. This behax ior 
of bacterioplankton distribution and activity over the water 
column was closely related to water dynamics and was typical 
of the oligotrophic waters of the tropical oceans. 

The next series of studies was conducted in the South 
China Sea. whose waters are more polluted than those of the 
central Pacific. Analyses of total biomass. bacterioplankton 
production, and bacterial degradation were performed at five 
stations. The results of the measurements are presented 
in Figs. 9-11. These show that the bacterioplankton count 
in the euphotic layer (0-50 m) ranged from 176 to 
61 1 x 10' cells/ml '. The highest bacterial population densities 
were noted for Stations 1 30 and 131. situated in coastal waters: 
the lowest values were obtained at pelagically situated 
Station 127. The bacterioplankton count at certain levels of 
the water column attained 611 x 10' cells ml'. The 
average bacterioplankton count in this area equaled 
354 x 10' cells/ml : . 

The bacterioplankton biomass varied over a broad range 
of values from 3.16 to 11.81 (.tgCl'd 1 . Data for the 0-50-m 
layer averaged for individual stations yielded values from 6.67 
to 9. 1 6 jug C I ' d ' . Mean data for bacterioplankton biomass and 
counts for the euphotic /one place the investigated portions of 
the South China Sea in the oligotrophic category. 

Low bacterioplankton counts notwithstanding, microflora] 
activity was relatively high. Thus, the rate of CO. assimilation 
by bacteria in darkness ranged from 0.52 to 4.07 (ig C 1 d '. 
averaging 1.60 p.g C 1' d ' for all the areas studied. 
Bacterioplankton production for individual stations 
ranged from 1 2.8 to 39.7 p.g C 1 ' d ' (Fig. 1 1 ). The value for 
the entirety of the area investigated was 26.6 |ig C 1 ' d '. 

The rates of bacterioplankton respiration in the 0-50-m 
layer calculated as averages for particular stations ranged 
from 0.11 to 0.34 mg O, 1' d '. The range of bacterial 
respiration rates for individual samples ranged from 0.07 to 
0.58 mg O, I'd'. The maximum bacterioplankton respiration 
rate was observed at Station 1 30. the minimum at Station 1 29. 
The average daily oxygen uptake by bacterioplankton in the 
area was 0.23 mg/1 ' . which w as in good agreement with values 
calculated from experimental data on the respiration of the 
plankton community as a whole. 



210 






6 


12 


16 


24 3C 





2 


4 


6 


e ic 





0.25 


05 




75 1. 



E 10 

Q 

Q 25 

45 






1 






6 


12 


'8 Stattor2427 


30 





2 


4 


6 8 


10 





025 


05 


075 


1.0 







2 


4 


6 


8 IC 





25 


05 




075 1. 



"V> 



,/■' 



/ 





Figs. 9-10. 



Vertical distribution of: ( 1 )watertemperaturein°C;(2)dissolved 
oxygen in mg/1: (3) total bacteria count in millions of cells/ml; 
and (4) CO, assimilation by bacteria in darkness, in g C 1 ' d '. 
All data apply to the South China Sea. 



Total bacterial degradation in the water column at individual 
stations ranged from 41 to 127 (ig C 1 ' d '. The average for a 
whole of the area studied was 86 U. C 1 ' d" 1 . 

The vertical distribution of bacterioplankton over the 
water column within the euphotic zone included a clearly 
evident peak at a depth of 1 m (Figs. 9, 1 0). The mean bacterial 
density and biomass in this layer were 450 x 10' cells/ml ' and 
10.1 u.g C 1 ' d "', respectively. Both counts and biomass then 
decreased with depth. 

Microfloral activity varied considerably over the water 
column. Two microfloral activity peaks were observed at 
several stations: one lay at a depth of 10 m, the other at or near 
the sea bottom. Enhanced production-degradation activity 
was likewise noted at the corresponding depths. Bacterial 
biomass production in the peak-activity layers was 2.5 times 
the bacterial production obtained for the entire sea area studied. 
The degradation rate OB in the bottom layer was 3.5 times that 
noted for the surface layer. 

Discussion 

Particular portions of the Pacific Ocean exhibited differing 
levels of microbiocoenosis development. The equatorial Pacific 
and South China Sea areas are closely similar with respect to 
bacterioplankton and biomass parameters in their euphotic 
zones. On the other hand, the level of development of the 
bacterial population in the western Pacific was found to be 
much lower. The South China Sea exhibited relatively high 
microfloral activity. Values characterizing bacterial production 
and organic matter degradation in the sea were similar to those 
reported for the mesotrophic waters of the tropical ocean. The 
level of production-degradation processes in the western Pacific 
was almost half of that in both the equatorial ocean and the 
South China Sea and corresponded to values characteristic of 
oligotrophic waters. 



n 



n 



in 0-55 m layer 



Station 126 Station 127 



Station 128 



Station 129 Station 130 



Station 131 



Scale: 



I I Bacterial production 
10 Ug c/l/day EH Organic matter degradation 



Fig. 1 1 . Bacterial production and organic-matter degradation in the South China Sea. 

211 



3.2 A Study of Primary Phytoplankton 
Production 

MIKHAEL N. KORSAK 

Institute of Global Climate and Ecology, State Committee for Hydrometeorology and Academy of Sciences, Moscow, USSR 



Introduction 

The study of the rate of formation of new organic matter 
by phytoplankton photosynthesis in the tropical Pacific reported 
in the present paper was carried out during the First Joint 
US-USSR Central Pacific Expedition aboard the R/V Akademik 
Korolev in 1988. The ocean areas investigated included 
portions of the central and western Pacific, which had received 
little previous attention. Primary production in the area of the 
first transect, beginning near Caroline Atoll and ending at 
Tarawa Island (Republic of Kiribati), ranges from values 
characteristic of oligotrophic parts of the ocean 
(100 mg C/nr/day and lower) to values corresponding in 
mesotrophic areas of the Pacific (25 mg c/nr/day) (Sorokin, 
1976). 

The central portion of the tropical Pacific (stations of the 
second transect) may be characterized as an oligotrophic 
productivity zone where a major role in primary production 
enhancement is played by synoptic phenomena such as cyclones, 
tornadoes, waterspouts, etcetera (Sorokin, 1976). Situated in 
the open portion of the northern tropical Pacific (i.e., within the 
northern tradewind zone), the second transect studied showed 
a primary rate of organic matter production by phytoplankton 
of about 1 00 mgC/m7day (Sorokin, 1976). The relatively low 
rales of primary production in this part of the Pacific are 
attributable largely to the deficiency of biogenic elements in 
the photosynthetic layer. These low concentrations are in turn 
due to anticyclonic circulation, which produces downwelling 
of nitrogen- and phosphorus-poor surface waters. It is the 
resulting low nitrogen and phosphorus levels in the 
photosynthetic layer that limit photosynthesis rates in the 
phytoplankton community. 

Despite the absence of significant seasonal variations in 
illumination and water temperature in the tropical ocean, 
considerable seasonal changes of photosynthesis rates have 
been reported in spring and autumn (Sorokin, 1976). 

Materials and Methods 

Studies al Stations 1 14-120 in the central tropical Pacific 
were conducted from 27 September to 7 October 1988. The 
work al Stations 121-126 along the Marianas transect was 
performed from L 6 October to 2] October. Primary production 
was determined by means of a radiocarbon version of the "jars 
method" proposed by Sorokin (Sorokin el al., 1983). Work al 
each station included measurements of photosynthesis in a 
surface-water sample (C ps ) as well as determinations of 
photosynthesis in the water layer as a function of phytoplankton 
distribution over (he water column (the coefficients K,). 



The light curves (the coefficients K,) were determined at one 
station in each transect. The sample incubation was usually 8- 
10 h, beginning in the morning. The radioactivity of filters with 
l4 C-labeled phytoplankton and of the working NaH l4 CO, 
solutions was measured using a Nuclear Chicago "Mark 2" 
scintillation counter. Sample radioactivities were counted 
using liquid scintillator cocktails of previously described 
composition (Sorokin, 1976; Sorokin et al., 1983). 

Primary production was calculated using the standard 
formula, with a factor of 1.5 to correct for l4 C loss due to 
phytoplankton sample filtering (Sorokin, 1976). All 
determinations of primary production were carried out in 
triplicate. The extent of the photosynthesis zone was taken to 
equal the white-disk transparency multiplied by three. Samples 
for determining phytoplankton production were taken using 
5-1 Niskin bottles at depths of 0.5: 10; 15; 25; 45; 70, and 
100m. 

Results 

The vertical structure of phytoplankton communities in 
high-transparency tropical ocean waters is characterized by 
several phytoplankton growth peaks or maxima in the euphoric 
zone, whose depth sometimes exceeds 100 m. Two layers with 
elevated phytoplankton concentrations are usually in evidence. 
The first of these occurs at a depth of 1 0-30 m and is associated 
with a photosynthesis-optimal light zone; the second lies at 
depths of 70-90 m and is related to heightened biogenic 
element levels in (he vicinity of the pycnocline (Sorokin. 
1 976). As is evident from Table 1 , the depth of the photosynthesis 
layer at all stations of both transects was usualh slightly in 
excess of 100 m. As a rule, only a single photosynthesis peak 
was observed (the one in the 15-25-m depth range), since the 
pycnocline lay below the photosynthesis layer boundary (see 
Table 1 ). However, Stations 1 22 and 1 26 did exhibit a second 
relatively small primary production peak at 70 m (Table 1). 
Primary production values for the topmost levels of the water 
column were usually markedly lower than at depths of 
10-15 m. which was probably due to photic inhibition of 
photosynthesis by the high-intensity incident light. Stratification 
of water masses o\ er the w ater column had no significant effect 
on primary organic-matter production by phytoplankton. 
inasmuch as the top 100 m of the water column was 
homothermal. Thus, the water temperature in the top 100 mof 
the water column at stations of both transects varied within just 
1-2°C. Salinity in the same laser varied within the same 
narrow limits, so that the primary production level at various 
depths depended largely upon light intensity, amount of 
phytoplankton present, and biogenic-element availability. 



212 



The local primary production peaks occurring at a depth of 
70 m at Stations 1 22 and 1 26 were probably due to the elevated 
levels of biogenic elements present there. 

The values obtained for primary production of organic 
matter by phytoplankton for the whole of the photosynthetic 
zone ranged from 70 to 140 mg C/nr/day (see Table 1 ). The 
maximum phytoplankton production ( 140 mg C/nr/day) was 
observed at Station 116; minimum production 
( 70 mg C/nr/day) occurred at Station 120 (Table 1). The total 
average levels of phytoplankton productivity in this part of the 
Pacific during the period of our studies was in line with the 
higher range of primary phytoplankton production values 
typical of oligotrophic zones of the World Ocean (i.e., about 
101 mg C/nr/day). 

For stations of the second transect, situated to the northwest 
of the first at about 8°N, the range of variation of primary 
production rates (90-338 mg C/nr/day) exceeded the range of 
values for Stations 1 14-120 (Table 1). The mean primary 
production at Stations 121-126 corresponded to the lower 
range of productivity rates for oligomesotrophic zones of the 
World Ocean (i.e.. about 172 mg C/nr7day) (Table 1 ). 

Comparing the primary organic-matter production values 
measured during the 1988 expedition with those obtained 



during the 1984 expedition, which covered roughly the same 
parts of the Pacific, we note that both primary productivity and 
the range of variation of primary production rates were greater 
in 1984. The rates of primary production of organic matter by 
phytoplankton over the Marianas transect in late July-early 
August 1984 ranged from 1 00 mg C/nr/day to 1.16gC/nr7day, 
the mean value being about 400 mg C/nr/day . The highest 
primary production values for phytoplankton occurred at the 
westernmost stations of the transect. The rather high primary 
production rates in this part of the ocean in 1 984 were probably 
due to the seasonal arrival of waters rich in biogenic elements 
that originated in the equatorial divergence area to the south of 
the study area. 

Phytoplankton productivity rates for the Marianas transect 
in 1988 were somewhat lower than those recorded in 1984. 
The differences can probably be attributed to fluctuations in 
the arrival of waters from the equatorial divergence region. 
On the whole, the level of primary production of organic 
matter in the central tropical Pacific was in keeping with 
expectations based on previous studies ( Sorokin, 1 976 ) and on 
the findings of the 1 984 expedition, corresponding more or less 
to the level associated with oligomesotrophic zones of the 
World Ocean. 



TABLE 1 

Primary production (mg C/nr/day) at different stations 
in the central Pacific. 



Depth 








Station Number 












(m) 114 


1 15 


1 16 


119 


120 


121 122 


123 


124 


125 


126 




0.5 2.9 


1.7 


2.3 


1.2 


0.42 


3.5 1 .9 


2.9 


0.40 


1.2 


1.4 




10 1.7 


2.3 


2.5 


1.8 


1.2 


5.2 5.6 


2.7 


0.48 


0.53 


3.0 




15 1.5 


2.6 


5.6 


3.6 


1.6 


7.2 10.5 


6.0 


2.7 


2.7 


4.5 




25 2.8 


1.1 


1.5 


2.1 


1.3 


1.7 7.8 


2.0 


3.4 


3.7 


2.7 




45 0.7 


0.6 


1.0 


1.6 


1.0 


0.74 1.3 


0.97 


0.45 


1.3 


0.6 




70 0.44 


0.32 


0.6 


0.58 


0.16 


0.74 2.3 


0.3 


0.3 


0.46 


0.7 




100 0.06 


0.02 


0.08 


0.14 


0.02 


0.21 0.13 


0.01 


0.00 


0.05 


0.0 




* P,„2 80.5 


84.2 


14(1 


132 


70 


177 338 


143 


89 


125 


16' 1 




* P,„: is the primary produi 


:tion for the 


photosynthesis 1 


ayer down to 100 m ( m 


g C/m 


7day). 









3.3 Mesozooplankton 



TATIANA A. PAVLOVA and AUDREY S. KUL1KOV 

Institute of Global Climate and Ecology, State Committee for Hydrometeorology and Academy of Sciences, Moscow, USSR 



Materials and Methods 

The materials for this work were collected in the eastern 
equatorial region of the Pacific Ocean at seven stations ( Stations 
1 14- 120) whose coordinates ranged from 10°Sto 150°W,and 
0°S to 178°W (Fig. 1). The zooplankton was collected in 
daylight with a large Juday Net. with a mesh of 168 urn and a 



throat diameter of 37 cm. from the 0-50. 50-100, and 
100-200-m levels. The samples were fixed with a 40% 
solution of formaldehyde and were processed by the standard 
methods (Korshenko, 1988: Tsyban et al„ 1988) under an 
MBS-9 binocular microscope in a Bogorov chamber. The 
sample was concentrated and poured into a Petri dish. While 
it was being examined under the binocular microscope, the 



213 



180 W 



150'W 



a) 




Boundary of 
Tradewind 
Current 



180 W 170' 



150' 

5' 

x|x convergence 

o 



b) 




180 W 170 ' 




Fig. I. Character of currents in the region studied: (a) circulation of waters 
at the surface; (h) circulation of waters at a depth of 100 m; and (c) 
circulation of waters at a depth of 2(H) m. 

largest animals ( over 2 mm long ) were washed out and classified 
into systematic groups using a biolam R-7 microscope. The 
residue of the sample was diluted 1 0-20-fold, depending on the 
concentration, and 2-3 portions of 10-20 ml each were taken 
with a 5-ml plunger pipet for counting. On the basis of this data, 
the mesozooplankton numbers in 1 m 1 and under 1 m : were 
calculated. A micrometer eyepiece was used to determine the 
body length of representatives of each species. On the basis of 
these values and Chislenko's nomograms (Chislenko, 1968; 
Vinogradov & Shushkina, 1987), the biomass of each species 
was determined, and the total biomass of mesozooplankton in 
I in' and under 1 m was calculated. Twenty samples were 
analyzed. 

Results and Discussion 

li is well known (Korshenko, 1988) that the characteristics 
of the distribution of zooplankton depend, to a considerable 
degree, on the hydrological structure of the water masses. 
Figure 1 shows a diagram of the currents in the studied region 
on the surface and at depths of 1 00 m and 200 m. The dynamics 
of the water masses at the easternmost station near Caroline 



Atoll and at adjacent Station 1 14 are low; no major currents 
pass through this area. The depths of 100 m and 200 m at 
stations in the central portion of the region are characterized by 
a water mass transport to the west. The w ater masses of the 
western portion of the section (Stations 119 and 120) are 
affected by the surface South Tradew ind Current. In addition, 
at the westernmost point of the region (Station 120). an 
anticyclonic equatorial current is observed (Gorshkov, 1974). 

The heterogeneity of the regional water structures suggests 
a nonuniform horizontal and vertical distribution of the 
mesozooplankton as well as its qualitative diversity. As is 
evident from Table 1 . the composition of the mesozooplankton 
differed qualitatively at different points of the section. A total 
of about 1 80 forms of mesozooplankton were determined as a 
result of the analysis of the samples. The average number of 
species for a station ranged from 75 to 139 and. for an 
individual level, from 39 to 84 (i.e., it differed by a factor of 
two) (Table 1; Fig. 2). The smallest number of species was 
found at the station near Caroline Atoll: at the 0-50-m and 
50-100-m levels, there were 39 and 56 species, respectively. 
The number of species increased from south to north toward the 
equator, reaching a maximum of 1 39 at Station 17. The vertical 
structure of species composition also varied. In the majority of 
the cases, the number of species increased appreciably with 
depth, and it remained practically unchanged only at the 
equator (Fig. 2a). 

The species observed in the waters of the region studied 
were encountered at different frequencies. Four gradations of 
species occurrence were distinguished: rare, under 1 sp/m'; 



Number of species 



2 


3 30 


40 


s|l 


60 70 80 


90 100 


Depih. m. 








-.1 l!" Ins 
V \ 

l\ll5 \ 


HI? 


50- 








!\ \ \ / 










M4^\ 


sj?« b fl 






100 








1 I 














\l 




200 - 










^ 



Number of species 
140 




120 119 118 117 116 115 114 Sun.) 



I ig 2. Number ol meso/ooplankton species: (a) at an individual level; and 
Ibl average for the station. 



low, from I to 10 sp/m 1 ; normal, from k) to 100 sp/m'; and 
massive , > 100 sp/m 3 (Table 1 ). As a rule, determined forms 
of mesozooplankton were found at the same frequency. 
However, it is evident from Table 1 that, in many cases, the 
frequency of species occurrence increased in the western 
portion of the section. The number of massive species at an 
individual level ranged from 4 to 14. The maximum number of 
massive species ( 1 3-14) was recorded in the western portion of 
the section at the equator; the minimum number was recorded 
in the eastern portion. We note that the populations of only 2 1 
species of zooplankton reached numbers above 100 sp/m 1 ; 18 
consisted of the order Copepoda. Among other taxonomic 
groups, massive concentrations were formed by Flassisagitta 
enflata, Oikopleura sp. I, Euphausia similis var. "armata." 

Some massive species reached a high density at all stations 
and levels. They included Clausocalanus porgens, Oithona 
spp., Oncaea venusta, Corycaeus gibbulus, Microsetella rosea, 
and Oikopleura sp. I. Of ihese, the most numerous species was 
the cyclopoid. O. venusta. On the other hand, 60 rare species 
occurring in 5-15% of the samples were counted. On the 
whole, the lists, of planktonic organisms observed in equatorial 
waters of the Pacific Ocean (Geinrikh, 1960; Vinogradov & 
Voronina, 1963; Arashkevich, 1972; Stepan'yants, 1977) are 
similar to those that we obtained (Table 1 ). 

Mesozooplankton species composition in the section from 
Caroline Atoll to the equator was approximately uniform, from 
east to west for species of Cyclopoida, Harpacticoida, 
Appendicularia, and Siphonophora, as well as for the minimal 
content of Euphausiacea and Chaetognatha(Fig. 3, Table 1). In 
the eastern portion of the region, such groups as Ostracoda, 
Mysidae, and Salpidae were practically absent. Of the 13 
determined Polychaeta, only two species were detected in the 
region near Caroline Atoll. Some species of Chaetognatha 
(i.e., Ferosagitta ferox. Pterosagitta draco, Parasagitta 
speticoela, Sagitta sp., Sagitta pulchra) found in the western 
portion of the region were absent from Station 1 14. Although 
species composition at the eastern stations was poor, relative to 
those to the west, some species were found there that were not 
found at any of the other stations (i.e., the Far-neritic 
[Vinogradov & Voronina. 1963] copepod Undinula vulgaris 
and certain siphonophores). 

In the western section of the region, the number of species 
of Cyclopoida, Harpacticoida, Euphausiacea. Amphiphoda, 
Calanoida, and Chaetognatha increased significantly, 
sometimes severalfold, in comparison with the eastern section 
(Fig. 3). As one moved toward the equator, from Station 1 17. 
the number of previously undetected deep-sea species of 
Calanoida (Neocalanus gracilis, N. robustion, Bradycalanus 
sp., Rhincalanus cornutus, Bradyidus armatus, Euchirella 
amoena, Pleuromamma abdominalis, P. gracilis, Haloptilis 
acutifrons, H. longicornis, Candacia longimana, and 
Labidocera detruncata) increased in the samples. This may 
have been due to a more intense mixing of equatorial waters in 
the region of their increase or to more active vertical daily 
migrations of mesozooplankton (Vinogradov & Voronina, 
1963). The western section was also much richer in the species 
composition of Cyclopoida: previously undetected species of 



Numbers, sp./m 
2000 ■ 

1500 - 



a) 



1000 - 



500- 




120 119 118 117 116 115 114 Stations 



b) 



400 

300 

200- 

100- 








— i 1 1 1 i 

120 119 118 117 116 115 114 Stations 



Fig. 3. Horizontal distribution of mesozooplankton in the 200 m surface 
layer: (a) distribution of numbers; and (b) distribution of biomass. 



Corycaeus lautus, C. robustus, and six species of the genus 
Sapphirina, Copilia longistylis, and Pachysoma dentatum 
were present in the samples. The Harpacticoida group in the 
samples near Caroline Atoll was represented by the single 
species, Microsetella rosea. The new species of Clytemnestra 
scutellata and Miracia sp., from deeper waters, appeared in 
samples from the western stations. In contrast to the eastern 
section, where Amphipoda were represented by species in the 
family Hyperiidae, to the west, there were more species of 
Platyscelidae and Pronoidae. Isolated specimens of Ramosia 
sp., Vogtia serrata, and Maresearsia sphaera; representatives 
of the Gerionidae family; larvae of cephalopod mollusks; and 
large forms of Appendicularia ( Oikopleura sp. II) were observed 
only in the western stations. Thus, moving from Caroline Atoll 
to the equator, the diversity and abundance of mesozooplankton 
increased. 

Differences in species composition were characteristic not 
only of the extreme western and eastern points of the region: a 
number of interesting characteristics were also discovered at 
Station 117. located at the center of the region. All the 
taxonomic groups were represented most completely in the 
community of zooplankton found at this station. For example, 
of 30 species of Amphipoda, 22 were found at Station 117. and 
of 13 species of Polychaeta, 1 1 were found. Large forms of 
Tomopteris sp.. Oikoleura sp. II, a single specimen of 
Nematoscelis gracilis, and concentrations of large Euphausia 
similis var. "armata" and tomopterids were also found in the 
waters of Station 1 1 7. 



215 



TABLE 1 

Qualitative composition of mesozooplankton. 

(Notation: *-rare species, numbers below 1 sp/m'; 

**-low-number species, numbers from 1-10 sp/m'; 

**-regular species, numbers from 10-100 sp/m 3 ; and 

: *-massive species, numbers in excess of 100). 



14 



115 



116 



18 



119 



120 



1. Foraminifera 
Radiolaria 

2. Spongotrochus 

3. Collosum sp. 
Hydrosa 

4 Gerionidae gen. sp. 

5. Liriope tetraphylla 
Siphonophora 

6. Agalma sp. 

7. Marrus sp. 

8. Ramosia sp. 

9. Vogtia serrata 

1 0. Maresearsia sphaera 

1 1 . Sulculeolaria quadridentata 

12. 5. quadrivalvis 

13. Galetta australis 

14. Diphyes bojani 

15. D. dispar 

1 6. Lensia achilles baryi 

17. L. campanella 

18. L. multicristata 

1 9. Lensia spp. 

20. Muggiaea atlantica 

21. Eudoxoides mitra 

22. E. spiralis 

23. Chelphyes appendiculata 

24. C. contorta 

25. Ceratocymba leucartii 

26. Abylopsis eschschottsii 

27. A. tatrugona 

28. Bassia bassensis 

29. A/>\7« schmidti 

30. Diphydae gen. sp. 
Polychaeta 

3 1 . Alciopa parasitica 

32. Alciopidae gen. sp. 

33. Krohnia sp. 

34. Rhynchonerella sp. 

35. Maupasia sp. 

36. Lopadorhynchus appendicular 

37. Pelagobia logicirrata 

38. Phillodocidae gen. sp. 
40. Tomopteris elegans 
41 Tomopteris sp. II 

42. Travisiopsis levinseni 

43. T. lobifera 

44. Typhloscolecidae gen. sp. 
Mollusca 

45. Cephalopoda larvae 
Crustacea 

46. Ostracoda: Conchoecia sp. 
Copcpoda 

47. ( 'alanus minor 

48. ( . pauper 

49. Neocalanus gracilis 

50. /V. robustior 

5 I . Neocalanus spp. cop. 

52. Undinula darwinii 

53. £/. vulgaris 

54. Calanus spp. cop. 

55. Bradycalanus sp. 

56. Eucalanus attenuates 

57. £. subcrassus 

58. Rhincalanus cornutus 



* 
** 

* 
* 

* 



** 

* 

* 



** 

* 
* 

* 
* 



* 
** 



** 


** 


* 


** 


** 


*** 


*** 




** 


** 


** 


** 


*** 


*** 




* 


* 


* 




* 






* 


* 


* 


* 


* 






* 


* 


* 


* 






*** 


*** 


** 


*** 


** 


*** 


*** 


* 


* 












** 


**# 


*** 


** 


** 


*** 
* 


*** 

* 


** 


** 


** 


** 


** 


** 


** 




* 


* 


** 
* 




** 

* 


* 
* 



216 



59. Acrocalanus gibber 

60. A. gracilis 

61. A. monachus 

62. Calocalanus pavo 

63. Paracalanus aculeatus 

64. Mecynocera calusi 

65. Clausocalanus arcuicornis 

66. C pergens 

67. Pseudocalanus minimis 

68. Pseudocalanidae gen. sp. 

69. Bradyidus armatus 

70. Euchirella amoena 

71. Aetidae gen. sp. 

72. Euchaeta marina 

73. Scolecitricella orientalis 

74. Scoletrix danae 

75. Phaena spinifera 

76. Centropages calaninus 

77. C. elongatus 

78. C gracilis 

79. C. longicornis 

80. Centropages spp. cop. 

8 1 . Pleuromamma abdominalis 

82. P. gracifc 

83. Lucicutia flavicornis 

84. Z.. oralis 

85. Heterorhabdus papilliger 

86. Haloptilis acutifrons 

87. W. longicornis 

88. Candacia catula 

89. C longimana 

90. C. paehydactyla 

91. C truncata 

92. Labidocera detruncata 

93. Pontcllina plumata 

94. Arcatia negligens 
Cyclopoida 

95. Oithona spp. 

96. Oncaea venusta 

97. (9. notopus 

98. Corycaeus agilis 

99. C asiaticus 

100. C. fO/H.V 

101 . C crassiuscidus 

102. C.flaccus 

103. C.gibbulus 

104. C.japonieus 

105. C. /m/mv 

106. C. longistylis 

107. C. robust us 

108. C. speciosus 

1 09. Corycaeus spp. cop. 

1 10. Sapphirina auronitens 

111. S. gastrica 

112. 5. intestinata 

1 13. 5. metallina 

114. 5. nigromaculata 

115. 5. opalina 

116. S.stellata 

117. Sapphirina spp. cop. 

1 1 8. Copilia longistylis 

119. C mirabilis 

1 20. C. cpiadreta 

121. Pachysoma dentation 
I 22. Nogagus muraji 
Harpacticoida 

123. Microsetella rosea 

124. Clytemnestra scutellata 

125. Miracia sp. 

1 26. Copepoda nauplii 



TABLE 1 - continued 

114 115 116 






** 


** 


** 


** 


** 


*** 


*** 


#**# 




** 



** 
** 
** 



17 



18 



* 


** 


** 


** 


* 




*** 


** 


**# 


** 


** 


** 


** 


** 


** 


** 


*** 


** 


*** 


** 


*** 


*** 


*** 


*** 


*** 


*** 


*** 


*** 


*** 


*** 


*** 


*** 


** 


*** 


*** 


**** 


* 


**** 


**** 


* =t- =fc 


**** 


*** 


**** 


* 


*** 



** 

** 



119 



** 
* 



** 


* 


** 


** 


* 


** 




** 


* 


*#* 


*** 


** 



** 



** 



120 



** 


** 


*** 


*** 


** 


** 


*** 


** 


*** 


**** 


**** 


*** 


** 


*** 


**** 


**** 



*** 


*** 


*** 


** 


*** 


*** 


** 


** 


** 



* 


** 


** 


** 


** 




* 




** 


** 






** 


** 


** 


** 


* 


* 


* 


*** 


** 


* 


* 


** 



** 
* 



* 


* 


* 


* 


* 


* 


*** 


** 


** 


*** 


*** 


*** 


**** 


*** 


#*** 


*** 


**** 


**** 


*##* 


*** 


$; % % ^ 
* 


*** 


**** 


* 



** 


** 


** 


** 


** 


** 


** 


** 


** 


** 




* 


* 


** 






* 


* 







* 



217 



TABLE 1 Continued 

114 115 llfi 



117 



118 



119 



120 



Mysidacea 

127. Mysidae yen. sp. larvae 

128. Gammaridae num. 
Amphipoda 



129. 


Vibilia chuni 


130. 


Paraphronima graciis 


131. 


Lestrigonus shizongeneios 


132. 


L. shoemakeri 


133. 


Hyperia fabrei 


134. 


Hyperietta stephenseni 


135. 


Phronima atlantica 


136. 


P. curvipes 


137. 


Phronimella elongate 


138. 


Phrosina semulunata 


139. 


Lycaeopsis zamboangae 


140. 


Eupronoe armatus 


141. 


E. maculata 


142. 


E. minuta 


143. 


Paralycaea gracilis 


144. 


Parapronoe parva 


145. 


Brachyscelus crusculum 


146. 


Oxycephalic longipes 


147. 


Leptocotis tenuirostris 


148. 


Calamorhynchus pellucidus 


149. 


Platyscelus armatus 


150. 


PI. ovoides 


151. 


Hemityphis tenuimanus 


152. 


Paratyphis parvus 


153. 


Paratyphis sp. 


154. 


Amphithyrus muratus 


155. 


Tetrathyrus forcipatus 


Ruphausiacea 


156. 


Euphausia similis var. armata 


157. 


Euphausia spp. furcilia 


158. 


Nematoscelis gracilis 


159. 


Stylocheiron affine 


160. 


Stylocheiron spp. furcilia 


161. 


Eupnausiacea caliptopici 


Decapoda 


162. 


Decapoda larvae 


163. 


Lin ifer sp. 


164. 


Echinodertnata larvae (onhiopl.) 


Chaetognatha 


165. 


Pterosagitta draco 


166. 


Aidanosagitta sp. 


167. 


Ferosagitta ferox 


168. 


E. robusta 


169. 


Flassisagitta enflata 


170. 


Parasigitta septicoela 


171. 


Sagitta pulchra 


172. 


Sagitta sp. 


173. 


Serratosagitta pacifica 


174. 


Chaetognatha spp. imm. 


Thaliacea 


175. 


Heliosalpa virgula 


176. 


Salpa sp. 


177. 


lasis zonaria 


178. 


Weelia cylindrit a 


179. 


Salpidae yon. sp. 


Doliolidae 


ISO. 


Thalia demot ratica 


181. 


Dolioloides rarum 


1S2. 


Doliolidae yon. sp 


App 


endicularia 


183. 


Oikopleura sp 1 


184. 


Oikopleura sp II 


IS5. 


Pisces Ian ae 



* 
** 



* 
** 



* 
** 



** 



** 
** 



* 

** 



** 



* 
** 



** 

** 



* 
*** 



** 
** 



* 
*** 



** 
** 



^ * * * * * 



$£$:): *** $$$4 



2 IS 



As noted previously, the distribution of mesozooplankton 
within the region is mainly due to the heterogeneity of 
hydrological conditions found there (Korshenko, 1988). Asa 
result, two bands of increased zooplankton biomass are formed 
in the surface layer along the equator (Vinogradov & Voronina, 
1963). In the Southern Hemisphere west of 140°W, the zone 
of abundant plankton narrows to 5°S^°S. At the same time, 
zooplankton productivity in the equatorial region increases 
from east to west. We found that zooplankton biomass ranged 
from 5 1 to 200 mg/m 3 . The western portion of the region 
(Stations 1 19 and 120) was located in the zone of abundant 
plankton (Fig. 1 ). The total biomass of mesozooplankton 
ranged from 89 to 340 mg/m 3 and averaged 201 mg/m 3 
(Table 2). The total biomass of zooplankton in the eastern 
portion of the section in the upper 200 m layer did not exceed 
163 mg/m 1 (Fig. 3). At the westernmost point, and at Station 
117. the biomass was maximum and amounted to 321 and 
340 mg/m 3 , respectively. The total numbers of mesozooplankton 
at different stations ranged from 547 to 2,170 sp/m 3 and 
averaged 1,044 sp/m 3 (Fig. 3; Table 2), which doubled the 
known quantity of plankton organisms (493 sp/m 3 ) known in 
the northern portion of the equatorial region (Korshenko, 
1 988). In the eastern section, the numbers of mesozooplankton 
ranged from 547 to 765 sp/m 3 . As one moved toward the 
equator, starting at 6°S, the numbers of mesozooplankton 
increased to a maximum of 2,170 sp/m 1 at Station 120 
(Table 2). It was shown that the qualitative composition and 
structure of the community at Stations 1 17 and 120 differed 
appreciably from those recorded at the other stations in the 
region. This is evidently related to the characteristics of the 
structure and dynamics of the water masses. 

The vertical distribution of the mesozooplankton was also 
different at different stations. At six of seven stations (Fig. 4). 
most of the mesozooplankton was concentrated in the upper 
100-m layer, and the maximum was in the 0-50-m layer 
(Table 3; Fig. 4), corresponding to findings cited by Korshenko 
(1988). A significant increase in zooplankton numbers and 
biomass with depth was tracked only at the westernmost point 
(Station 120). Mesozooplankton biomass as a whole increased 
with depth owing to an increase in the fraction of large-sized 
animals. 



In addition to analyzing the vertical distribution of total 
numbers and biomass, we also examined the vertical distribution 
of individual taxonomic groups. We found that the numbers of 
Calanoida and Euphausiasea decreased with depth, yet the 
biomass remained practically unchanged because of an increase 
in the fraction of large-sized forms (Fig. 5; Table 3). The 
numbers and biomass of Siphonophora, Polychaeta, Ostracoda, 
Hapracticoida, Amphipoda, Chaetognatha, and Salpidae 
increased with depth. The numbers and biomass of Cyclopoida, 
Mysidae, Decapoda, and Echinodermata, on the other hand, 
decreased with depth. The numbers of Doliolidae were constant 
at all the levels, and the biomass increased with the relative 
content of large-sized animals. 

In order to analyze the structure of the mesozooplankton 
community, we divided the determined forms into 17 major 
systematic groups. The frequency of occurrence (by numbers 
and biomass) of 14 of these basic taxonomic groups is shown 
in Table 4. The maximum fraction of numbers, as an average 
for the cross section, belonged to Calanoida — 53.8%. In 60% 
of the cases, Calanoida comprised over half of the total numbers. 
According to Korshenko (1988), the numbers of Calanoida in 
the 0-200-m layer amounted to 45% of the total value. The 
highest numerical percentage of Calanoida (66.5%) was found 
on Station 1 14. In the western portion of the section, Calanoida 
numbers decreased to 40%. The biomass amounted to over half 
of the total value only in the region of the Caroline Atoll, owing 
to the small number of other taxonomic groups composing the 
community. To the west, the relative content of Doliolidae was 
comparatively high, and the fraction of groups other than 
Calanoida was insignificant. We noted a tendency for the 
number of taxonomic groups to increase toward the equator. 
On one hand, the fraction of numbers and biomass of the 
dominant group (Calanoida) decreased, while the relative 
content of Cyclopoida, Chaetognatha, and Siphonophora 
increased. For example, the fraction of numbers and biomass 
of Cyclopoida at the equator was 48 and 29%, respectively. In 
addition to the general tendency of the Calanoida fraction to 
decrease, in many cases the typical (Fig. 6) structure of the 
community broke down as a result of concentrations of animals 
of a particular species. For example, the biomass fraction of 
Euphausiacea was maximum (35.5%) at Station 1 17 as a result 



TABLE 2 



Total numbers and biomass of mesozooplankton in the 200 m surface layer. 



Stations 




Numbers 


sp/m" 






Biomass 


sp/m' 




Levels, m 


0-50 


50-100 


100-200 


0-200 


0-50 


50-100 


100-200 


0-200 


1 14 


755 


775 


_ 


765 


158 


131 


_ 


144 


115 


1.354 


826 


428 


759 


148 


202 


150 


163 


116 


791 


806 


246 


547 


87 


158 


55 


89 


117 


2,056 


1,153 


523 


1 .063 


387 


290 


342 


340 


1 IS 


1.028 


704 


699 


783 


124 


122 


146 


134 


119 


1,170 


1.947 


878 


1.218 


227 


220 


211 


217 


120 


1 ,439 


2.137 


2,560 


2.170 


126 


226 


466 


321 


Average values 


1.127 


1,193 


889 


1 .044 


180 


193 


228 


201 



Remark: For Station 1 14. the data and results were obtained for the 0-100 m layer. 



Depth, m 



500 



Numbcrv sp/ltl 

1500 2000 




200 



Biomass. mg/m 
100 200 300 400 500 




Fig. 4. 



Vertical distribution ofmesozooplankton; (a) vertical distribution of 
numbers: and (b) vertical distribution of biomass. 



of massive development of Euphausia similis \ar. "armatu" in 
this area. Korshenkol 1988) notedasimilarca.se for a population 
of E. similis. The same value (35.5% of total biomass) was 
shown by the biomass fraction of Chaetognatha at specific 
levels of the western portion of the section. A typical portrait 
of the structure of the mesozooplankton community of the 
region is shown in Fig. 6. 

The general pattern of variability of the vertical structure 
had the following characteristics: both the numbers and biomass 
of Calanoida, Decapoda. and Appendicularia and the biomass 
fraction of Cyclopoida and Siphonophora decreased with depth. 
The relative content of other groups (Harpacticoida. Amphipoda. 
Euphausiasea, Ostracoda. Chaetognatha. Salpidae. Doliolidae. 
Siphonophora. and Polychaeta) increased with depth. 

Conclusion 

/. The qualitative composition of the mesozooplankton 
was nonuniform in the region studied. As one moved toward 
the equator, species diversity of the zooplankton increased. 
The smallest number of species was recorded in the region of 
Caroline Atoll. In the majority of cases, the number of species 
increased with depth. It was postulated that the heterogeneitv 
of the horizontal and vertical distributions ofmesozooplankton 
resulted from vertical daily migrations and equatorial ascents 
of the water masses. 

Twenty-one species ofmesozooplankton reached numbers 
above 100 sp/m': of these groups, the group of the most diverse 
composition was Calanoida. The number of massive species 
increased from east to west. 











TABLE 3 














Vertical distribution of mesozooplankton 








Levels 





50 


m 






50- 


00 m 


100 


200 m 


Taxons 


N 




B 




N 




B 


N 


B 


I . Syphonophora 


3.2 




17.2 




5.3 




19.4 


7.7 


20.5 


2. Polychaeta 


2.1 




3.2 




2.6 




3.3 


3.8 


6.2 


3. Ostracoda 


0.3 




0.2 




1.2 




1.0 


4.4 


3.5 


4. Calanoida 


687.0 




60.9 




588.0 




65.6 


424.0 


59.0 


5. Cyclopoida 


553.0 




30.0 




429.0 




3.8. 1 


328.0 


23.0 


6. Harpacticoida 


31.0 




0.2 




73.0 




0.5 


47.0 


0.6 


7. Amphipoda 


1.3 




5.2 




2.6 




9.8 


3.6 


9.8 


8. Mysidae 


1.2 




0.1 




2.6 




0.3 


0.7 


0.1 


9. Euphausiasea 


12.7 




19.7 




3.4 




18.3 


13.3 


21.8 


10. Decapoda 


1.7 




3.0 




1.6 




2 2 


1.1 


1.8 


1 1 . Echinodermata 


5.9 




0.0 




4.7 




0.0 


3.2 


0.0 


12. Chaetognatha 


1 3.5 




21.4 




17.8 




24.7 


25.3 


57.6 


13. Salpidae 


0.8 




10.4 




1.5 




3.1 


1.6 


14.2 


14. Doliolidae 


0.9 




14.4 




1.4 




20.4 


1.0 


18.3 


15. Appendicularia 


154.0 




4.9 




69. 1 




2.1 


33.7 


1.0 


16. Pisces 


0.6 




2.1 




1.3 




3.6 


0.8 


2.3 


All Groups 


1.227 




180 




.193 




193 


889 


228 


N - average numbers, sp/m 3 




















B - average values ofbiomass, 


sp/m 



















220 



N, sp./m 





B. mg/m 




N, sp./m 


1 






Siphonophora 


80 




40 






60 




30 






40 


20 


20 






20 


10 


III 









n 












B.mg/m Harpacticoida 



20 119 118 117 116 115 114 C\ 



N. sp./m 




B, 


mg/m 


Calanoida 


101)0 - 


XQ 








soo - 


6Q 








600 - 
400 - 


4Q 

2(1 




\i 


/ • \ / 


200 - 


II 








o- 






i i 


ill i 




120 119 118 117 116 115 114 C 



N, sp./m" 



120 119 118 117 116 115 114 C T 



30 

20 

10 







20. 






A Euphausiacea 




nil 










80. 










60 










411 








- 


;u 








- 


ii 






tt » — -• 


- 




! 


1 1 


— i r 1 1 ■ 



120 119 118 117 116 115 114 C T 



N, sp./m 

1000 
800 
600 
400 
200 




hi I 



B, mg/m ' Cyclopoida 




120 119 | |,s 117 116 115 114 C- 



N, sp./m" 

10 




1. ma/m" 



Amphipoda 







B. mg/m 


Chaetognatha 


, sp./m 


9o| 
70 






40 - 


50 






30- 


30 






20 . 


10 





^\vv 


10 - 






~* \ 







1 1 1 1 1 1 



120 1 19 118 117 116 115 I 14 C- 




20 119 118 117 116 115 114 C T 



1 
N. sp./m 




B, mg/m" 


Appendicularia 


120- 








ioo- 








80- 








60" 


5 






40- 


4 
3 


^\ 


A\/^" 


20- 


2 
1. 


' V 


\_ 


0_ 










120 119 118 117 116 115 114 C T 



Fig. 5. Horizontal distribution of the main taxonomic groups of mesozooplankton. 



221 



TABLE 4 













Structure of mesozooplankton community. 














Stations 


114 


1 


5 


1 16 


117 


118 


119 


1 


20 


Average 
































values 


Taxa 


% N 


9c B 


', N 


', B 


'. \ 


% B 


% N 


9c B 


% N 


%B 


', N 


%B 


', N 


', B 


'. N 


\ B 


1 . Syphonophora 


0.3 


2.5 


0.1 


1.3 


0.3 


7.8 


1.4 


10.5 


0.6 


7.1 


0.5 


17.8 


0.4 


8.4 


0.5 


7.9 


2. Polychaeta 


0.0 


0.0 


0.3 


2.5 


0.4 


2 i 


0.5 


2.3 


1.3 


0.3 


0.4 


3.8 


0.1 


1.5 


0.6 


1.8 


3. Ostracoda 


— 


— 


0.3 


0.8 


0.4 


1.8 


0.3 


0.4 


0.9 


1.1 


0.2 


0.8 


0.1 


0.3 


0.3 


0.7 


4. Calanoida 


66.5 


54.3 


55.9 


42. S 


51.0 


34.0 


57.9 


20.1 


56.2 


31.9 


48.8 


34.8 


40.0 


28.2 


53.8 


33.7 


5. Cyclopoida 


24. 8 


10.2 


30.2 


11.3 


28.2 


12.2 


20.6 


4.5 


24.2 


10.5 


30.0 


9.6 


47.9 


28.9 


29.4 


12.4 


6. Harpacticoida 


1.4 


0.0 


4.2 


0.1 


6.6 


0.2 


4.3 


0.2 


8.4 


0.2 


5.1 


0.2 


2.6 


0.2 


4.7 


0.2 


7. Mysidae 


— 


— 


— 


— 


0.2 


0.2 


0.0 


0.0 


0.1 


0.1 


0.2 


0.1 


— 


— 


0.1 


0.1 


8. Amphipoda 


0.2 


2.0 


0.5 


6.1 


0.2 


4.2 


0.7 


6.4 


0.3 


5.4 


0.2 


2.6 


0.1 


0.8 


0.3 


3.9 


9. Euphausiasea 


0.5 


0.4 


0.4 


0.8 


1.0 


1.0 


2 2 


35.2 


0.6 


1.0 


0.8 


~> 2 


1.4 


3.7 


1.0 


6.3 


10. Decapoda 


0.3 


0.6 


0.1 


0.9 


0.6 


3.2 


0.3 


1.3 


0.1 


2.0 


0.5 


0.3 


0.1 


04 


0.4 


1.2 


1 1 . Chaetoanatha 


1.0 


9.3 


3.1 


20.5 


1.8 


13.7 


2 2 


8.7 


1.9 


18.6 


2.1 


19.6 


1.2 


16.7 


1.9 


15.3 


12. Salpidae 


0.0 


1.0 


0.0 


8.2 


0.3 


7.7 


0.1 


1.2 


— 


— 


0.3 


4.6 


0.1 


2.5 


0.1 


3.6 


13. Doliolidae 


0.2 


18.3 


0.1 


3.6 


0.1 


10.4 


0.1 


3.3 


0.1 


11.3 


0.1 


1.5 


0.0 


6.0 


0.1 


7.7 


14. Appendicularia 


5.3 


0.8 


4.1 


0.9 


3.9 


0.7 


8.4 


1.5 


6.7 


1.3 


10.9 


1.9 


5.7 


1.7 


6.4 


1.3 


% N - fraction of to 


al numbers 






























' 1 B - fraction of total biomass. 


































"3 -D 


— i 


u 


j= 


c 








































rj o 






OJJ 


■* 






j 


-z 








- C 


§■ 


3 


u 








X 


Hi 


u 


a. 

< 


1 \'l 


6 T\p 


ca 


structure 




total 


m 


mt 


ers 


; ai 



'•J 


DO 


a 


-D 


o. 




E 


- 


^T" 









7Z 


— 


s: 


c 


Q. 


E 


a 


y. 


Q 


























X 


u 


a. 
< 




%Ti 














ton c 


tmmunits 


: (a) structure 


ot 



and ih) structure of total biomass. 



2. In numbers and biomass of mesozooplankton. this 
region can be characterized as medium-productive, with the 
western section more productive than the east. The station 
poorest in mesozooplankton was near Caroline Atoll. The 
average numbers of mesozooplankton were maximum in the 
0-50-m layer and decreased w ith depth. On the other hand, the 
average biomass increased with depth owing to the increase in 
the fraction of large-sized forms 

3. The structure o\' the mesozooplankton community as a 
whole was characterized by the dominance of the fraction of 
numbers and biomass of Calanoida and Cyclopoida. In the 
majority of cases, half of the total consisted of Calanoida. In 
the western portion of the section, the relative content of other 
groups (Chaetognatha. Euphausiacea) increased, while the 
fraction of Calanoida decreased. 



3.4 Zooneuston of the Tropical Pacific 

YUZENALY P. ZAITSEV. LEONID N. POLISHCHUK, and BORIS G. ALEXANDROV 

Department of Active Marine Surface Hydrobiology, Institute of Southern Seas Biology, Odessa Branch, USSR 



Introduction 

Seventy-nine zooneuston samples were taken at 23 stations 
in tropical areas of the Pacific Ocean (Table 1 ). The sampling 
equipment consisted of a PNS-2 two-tier plankton-neuston net 
and an MNT fry-neuston trawl (Zaitsev, 1 97 1 ). The PNS-2 net 
(mouth area 0.1 nr; mesh size 1 50 urn) permitted simultaneous 
manual skimming of the upper surface of the upper layers of the 
pelagial (i.e., of neustal [0-5 cm] and subneustal [5-25 cm]). 
The MNT trawl (moutn area 0.39 m 2 ; mesh size 350 p.m) made 
possible high-speed catching of mobile neustons at a vessel 
trawling speed of 3 m/s. The average skim duration was 
10 min. 

To facilitate the study of neuston community formation, 
the whole of the ocean area investigated was subdivided into 
three types of zones, namely 1. atoll lagoons (shallow water; 
complete or partial isolation from the ocean; developed littoral 
communities); 2. semienclosed marginal seas (relatively little 
exchange with the ocean; mostly shallow water; well-developed 
pelagic-shelf communities); and J. open-ocean waters (pelagic 
communities predominant). The sampling effort covered an 
extensive portion of the tropical Pacific from Caroline Atoll in 
the east (154°51'E). The latitudinal limits of station locations 
were 1 1°20'N and 8°40'S. 

TABLE 1 

Neuston sample collection in the tropical Pacific. 



Area 




\ 


umber of 






Static 


ns 


Sampl 


es 






PNS-2 


MNT 


Caroline Atoll 


3 




22 


— 


Caroline Atoll/Phoenix Islands 


4 




8 


4 


Phoenix Islands/Gilbert Islands 


6 




12 


1 


Marianas sector 


5 




10 


4 


South China Sea 


5 
70 




18 
9 


- 


Total 23 





The neuston (surface plankton) of the tropical Pacific has 
been described in publications of a general character (Geinrikh. 
1 964 ). There are also papers on the distribution of its individual 
components (e.g., of copepods of the Pontellaidae family 
[Sherman. 1963; Voronina. 1964] and oceanic water striders of 
the genus Halohates [Herring, 1961]). Results of pleuston 
studies are reported in a major survey paper by Savilov ( 1 969). 



The need for the present study was dictated by a considerable 
hiatus in observations, as well as by the scarcity of quantitative 
data on neuston distribution in conditions of increasingly 
widespread and intense pollution of the World Ocean. 

The zooneuston of the tropical Pacific found during the 
present studies was represented by epineuston consisting of 
oceanic water striders (Halohates) as well as by hyponeuston: 
copepods of the family Pontellidae, marine snails of the genus 
Janthina, larvae of benthic invertebrates, and fishes. In addition, 
allowance was made for the usual neuston components in the 
form of semisubmerged organisms: Physalia (Siphonophora), 
Velella velella, and Porpita pacifica (Chondrophora). 

Analysis of the quantitative distribution of organisms 
leads to some basic conclusions regarding the distinctive 
features of neuston in the types of water areas studied. 

The average counts of animals in the open ocean and in the 
South China Sea were comparable (2,730 and 3,160 ind/m 3 ), 
while the concentration of animals in the lagoon at Caroline 
Atoll was 50 times lower. Larger organisms (medusae, salps, 
fish larvae and fry , euphausids hyperiidea, et cetera) constituted 
aconsiderable fraction of the total neuston biomass, accounting 
for as much as 2.5 g/nV (Station 1 19) and sometimes attaining 
values an entire order of magnitude higher ( Station 128), where 
Porpitae were dominant. 

The relative neuston content was richer in Caroline Atoll' s 
lagoon, the reverse of that for animals overall (Table 2). The 
low numbers of oceanic animal species in the lagoon against a 
background of early developmental stages of benthic 
invertebrates (Decapoda, Gastropoda, Cirripedia, et cetera) 
were due to shallowness of the waters and their relative 
isolation from the open ocean. The distinctiveness of the 
individual sea area categories (atoll, marginal sea. open ocean) 
was clearly evident from the proportion of benthic animal 
larvae in the total neuston count. Their percentage content 
decreased steadily from its maximum value of 99% in the 
lagoon to 75% in the South China Sea and 46% in the open 
ocean. The epineustonic Halohates, which breed along coasts, 
were similarly distributed, while the Pontellidae distribution 
behaved inversely (0.24% and 54%, respectively). 

In addition to benthic-invertebrate larvae, the neustal also 
contained large numbers of pelagic animals at early stages of 
ontogenesis, which validates calling neuston "the ocean's 
breeding ground." 

Comparison of the numbers of organisms in the microlayers 
investigated confirmed the existence of conditions conducive 
to neuston growth in all of the water areas studied. The sole 
exception to this rule was the waters of the Marianas sector 
(Stations 121-126) (Table 3). 



223 



TABLE 2 

Components (counts expressed as individuals/m') of neuston 

biocoenoses in the basic water area types of the tropical 

Pacific. 



B, in 
mg/m * 



Component 



Basic water area type 
Caroline South Open waters 
Atoll China of the 
lagoon Sea Pacific 



1. 


Pleuston 





1.27 


0.20 


2. 


Epineuston 


0.31 


0.13 


0.05 


3. 


Hyponeuston 

a) benthic-invertebrate 


37.36 


188.35 


164.13 




and fish larvae 


37.36 


143.15 


75.71 




b) Pontellidae and Janthia 





45.20 


88.42 


4. 


Others* 
Relative neuston 


26.63 


2,968.9 


2,566.6 




content, % 


59 


6 


6 



* This category groups together mesozooplankton components 
found in the surface layer, but distinct from the enumerated 
categories of organisms in being more uniformly distributed 
over the water column. 

TABLE 3 

Dominance of organisms in neustal expressed as a 
percentage. 



Sea area 



', 



Caroline Atoll lagoon 
South China Sea 
Open ocean 
Marianas sector 



1.3 ±0.1 
2.5 ±0.6 
1.5 ±0.3 
0.9 ± 0.4 



A degree of symmetry of neuston distribution with respect 
to the equator was noted (see Fig. 1 ). This was evidence of the 
effect of the northern and southern tradewind currents and of 
the anticyclonic current between them, 

The observed pattern of organism distribution agrees with 
the data of Voronina ( 1964), who established that peak counts 
of the most common pontellid Pontella tenuiremis occurred at 
1°N and 1 C S. The meridional components of surface current 



100 
6 
4 
2 


• 
• 


• 


• \ * 

• . 


• 









10 
North 


8 


6 4 2 2 4 

Degrees of latitude 


6 8 

South 



Fig. 1 . 



Latitudinal distribution of neuston biomass (B) in the tropical waters 
of the Pacific Ocean. 



point outwards from the equator, so that species that inhabit the 
topmost water layer are constantly carried away from it. The 
departing water is replaced by deeper-lying water free of 
surface animals. The result is an acute trough in surface species 
along the equator (Sverdrup et al., cited by Voronina, 1964). 

Constituting an agglomeration of hydrobionts. neuston 
attracts both aquatic and aerial predators. We know, for 
example, that decapod crustacean and pontellid larvae are part 
of the diet of tunas, bonitos, and other epipelagic fishes ( Marchal. 
1959). Porpita are consumed by marine turtles, while fish 
larvae and fry are food for marine snakes (Zaitsev, 1971). 
Neuston animals are an important part of the diet of seabirds. 
Flying fish are caught on the fly by frigate birds (Fregatidae) 
and terns (Laridae). Procellariiformes birds (Puffinus) have a 
variety of ways of feeding on neuston; albatrosses ( Diomedea) 
and fulmars (Fulmarus) snatch their prey from the surface, and 
fork-tailed storm-petrels (Oceanodroma) do so while 
performing a "mincing walk" on water (Boaden& Seed, 1985). 
In waters close to shore, neuston is eaten by skimmers 
(Rhynchopidae) with their long cultiform gonys (Zaitsev, 
1971 ). According to the latest count (Day et al., 1984). there 
are at least 50 species of birds that feed on neuston. 

In conformity with what is usually the case with distribution 
in the ocean, neuston is most profuse in areas where currents 
converge. However, recent decades have seen a proliferation 
in the same areas of various kinds of plastic debris ( Day et al. , 
1984). Looking for neuston, marine birds often swallow these 
foreign objects, whose effect is invariably harmful and often 
fatal. With their high adsorption coefficients, oil lumps and 
plastic objects tend to have a buildup of toxic substances on 
their surface (Osipov & Charykov, 1987), thus inhibiting the 
development of invertebrates and fishes in the neustal. This, in 
turn, can only have a negative impact on the bioproductivity of 
the entire ocean. 



224 



3.5 Observations of Seabirds along a 

14,892-km Cruise Track in the Tropical 
Pacific Ocean and the Bohol, Sulu, and 
South China Seas 



ANGELA K. KEPLER 1 , CAMERON B. KEPLER\ DAVID H. ELLIS 5 , and JEFFREY S. HATFIELD } 
US Fish & Wildlife Service, Patuxent Wildlife Research Center, Southeast Research Station, Athens, Georgia, USA 
*US Fish & Wildlife Service, Patuxent Wildlife Research Center, Laurel, Maryland, USA 



Introduction 

This study forms a small part of the research efforts of the 
First Joint US-USSR Central Pacific Expedition (Line and 
Phoenix Groups. Gilbert Islands. Micronesia, inland Philippine 
Seas, and South China Sea). Seabird observations were made 
from the Soviet research vessel (R/V) Akademik Korolev 
(7.000 tons, 1 24 m in length) from Hilo, Hawaii, to Singapore 
(Figs. 1,2) via Christmas Island (02°N, 157°W) and Caroline 
Atoll (10°S, 150°W). 

The primary objectives of the expedition were to 
characterize and contrast the fundamental oceanographic, 
hydrochemical, microbiological, hydrobiological, and 
ecological parameters of arctic and tropical marine ecosystems. 
Emphasis was placed on the primary productivity and ecological 
health of these two major areas, including pollution studies 
involving multidisciplinary experiments conducted jointly by 
scientists of both countries. 

In this paper, we report the marine distribution of seabirds 
and other transoceanic migrants, such as shorebirds and ducks, 
during the tropical portion of the cruise, which covered 
14.892 km from 9 September to 31 October 1988. Weconducted 
a total of 1 6 1 hours of observations on transects representing an 
area of 3,609 km 2 , during which time the ship traveled 
4.5 1 1 km (Table 1 ). Our observations covered 21 degrees of 
latitude and 1 07 degrees of longitude. Because the cruise track 
traversed many island archipelagos, we subdivided it into the 
following 7 regions (Figs. 1,2), all lying between 14°N and 
10°S latitudes: 

Region 1 Line Islands, including waters south of Hawaii 
(150°Wto 160°W); 

Region II Phoenix Islands to the international dateline ( 1 60°W 
to 180°); 

Region III Gilbert Islands (180° to 165°E): 
Region IV Caroline Islands', Micronesia ( 165°E to 136°E); 
Region V Philippine Sea and Basin ( 1 36°E to 1 25°E); 
Region VI Inland Philippine Seas: Bohol (Mindanao) and 
Sulu Seas. Balabac Strait (125°E to 1 17°E); and 



Not to be confused with Caroline Atoll (Island), Southern 
Line Islands, at 10°00S latitude, 1 50° 13'W longitude. 



Region VII South China Sea. Philippines to Singapore ( 1 1 7°E 
to 103°E). 

We interpret our results within the contexts of 1. breeding 
phenology, nonbreeding dispersion, and migration (Fig. 3); 

2. proximity to known breeding colonies and nearest landfalls; 

3. previous at-sea records; 4. anthropogenic factors such as 
population density, environmental alterations to coastal habitats, 
and pollution (see Chapters 2,3); and 5. general areas of 
upwelling, providing locally rich feeding areas. 

Seabirds, more than any other group of living organisms, 
illustrate that the world's oceans are united. For example, 
parasitic jaegers (Stercorarius parasiticus) breed in Siberia 
and Alaska, then migrate south to winter in south temperate 
waters of the Pacific. Indian, and Atlantic Oceans. En route, in 
the Pacific, they skirt all four continents, passing through 150 
degrees of latitude and at least 80 degrees of longitude. 

The 1988 US-USSR expedition provided an opportunity 
to study assemblages of birds, highly visible indicators of the 
health of marine ecosystems, over vast areas of the Pacific 
Ocean. Studies of the marine environments utilized by these 
birds provide the knowledge to encourage practical action 
toward their conservation. 

Previous Studies 

Although the broad distributions of central and western 
Pacific seabirds are well-known (Murphy, 1936; Mayr, 1945; 
Delacour & Mayr, 1946; Baker, 1951; Clapp, 1967; Clapp & 
Sibley, 1967, 1968; Amerson, 1969; King, 1970. 1973, 1974a; 
Nelson, 1975, 1978; Perry, 1980; Engbring, 1983; Garnett. 
1983, 1984; Gould, 1983; Harrison, 1985; Pratt, Bruner & 
Berrett, 1987), much remains to be learned of their detailed 
distribution patterns in the Southern Line Islands and areas 
west of Micronesia. Information on seabirds in southeast Asia, 
at least in English, is incomplete and often outdated (Delacour 
& Mayr, 1946; Delacour, 1947; King & Dickinson, 1975; 
Nelson, 1978; Harrison, 1985). We are unfamiliar with the 
literature in Asian languages and have not pursued the numerous 
reports and publications resulting from marine oriented trips to 
islands and reefs in the Asian region through which we passed 
(UNEP, 1984a,b;IUCN, 1 988 a,b), some of which may contain 



225 



Not thern 

Mariana 

Islands 



Marshall Islands 



^T^io/i* 



.. M/t-M 



Papua NewGuinea -I • ' knmwWT; 



Kiribati * 






PACIFIC OCEAN /./.• 

i I 

! 






Tuvalu 






■ ft/l?- k *IWT_IM*Tl 



SSm # * \» 



10 5 10/4 



■ '• - \ , /2 e 



*«Oi»„ ,s«.»«° s 
Tokelau 



\"" 



< s _$ to'i 



Fig. I. Cruise track of the R/V Akademik Korolev: Regions I (Line Islands), II (Phoenix Islands). Ill (Gilbert Islands), and IV (Caroline Islands. Micronesia), 
15 September-17 October 1988. Solid lines on cruise track represent daily observation hours, dotted lines, hours of darkness. 



PACIFIC 
OCEAN 




AK AOEMIK 



I - ID CRUISE TRACK 

' 1 KOROLEV 

La J CRUISE REGION BOUNDARIES "*«". 

1 D f ff) 



Fig, 2. Cruise track of the WV Akademik Korole\ : Regions V (Philippine Sea and Basin), VI (Bohol and Sulu Seas), and VD (South China Seal. 18-31 October 
l l iss Solid lines on cruise track represenl dail) observation horns, dotted lines, hours ol darkness 



lib 



TABLE 1 

Relative abundance of seubirds in each sector of the R/V Akademik Korolev cruise track. 1 5 September-3 1 October 1 988. 













Region 










I 


I] 


III 


IV 


V 


VI 


VII 


Total 


Observation Hours 


29.95 


21.88 


18.00 


42.78 


13.70 


14.57 


20.45 


161.33 


Kilometers Tra\eled 


















Durinc Observations 


772.91 


656.73 


554.70 


1.210.87 


356.67 


404.14 


555.38 


4.5 11.41 


Area Covered bv 


















Observations (km : ) 


618.33 


525.38 


443.76 


968.70 


285.34 


323.31 


444.31 


3,609.13 


No. Bird Species Seen 


27 


25 


14 


19 


6 


4 


8 


46 


No. Individuals Seen 


457 


1.796 


495 


799 


34 


22 


65 


3.668 


Average No. Birds Seen/hr 


15.3 


82.1 


27.5 


18.7 


2.5 


1.4 


3.2 


22.7 


Average No. Birds Seen/10 km 


















of Observations 


5.9 


27.3 


8.9 


6.6 


1.0 


0.5 


1.2 


8.1 


Average Bird Density 


















Per lOknr 


10.20 


45.72 


14.6 


10.38 


1.76 


1.63 


3.95 


12.61 


No. Families Seen 


7 


7 


5 


9 


4 


4 


4 


11 



Each hour is the combined total of two people simultaneously viewing opposite sides of the ship, and thus equals 2 hours of observation 
in some other papers. 



relevant seabird records. Most of the previous information on 
breeding and at-sea distribution of tropical seabirds was gathered 
from 1963 to 1969 (Humphrey, 1965) by the US National 
Museum's (Smithsonian Institution) Pacific Ocean Biological 
Survey Program (POBSP). 

Methods 

Daily at-sea observations were conducted from the flying 
bridge of the R/V Akademik Korolev, 12 m above the sea 
surface and within a viewing arc of approximately 180°. 
During the first week we honed our observation skills, using 
Harrison ( 1 985 ) as a major reference. Thereafter, we maintained 
almost constant watch during daylight hours (29 days during 
9 September-3 1 October). Methods were based on those of the 
POBSP (King, 1970; Gould. 1974), modified by techniques 
utilized elsewhere (Tasker <?/«/., 1984;Haney, 1985;Gould& 
Forsell. 1989). The watch rotated between three observers 
(AKK, CBK, and DHE), with two observers on watch at all 
times. A change of one observer took place every hour on the 
hour, each person alternating 2-hour watches with a 1-hour 
break to reduce fatigue. Observations began 10-15 min before 
sunrise and terminated 1 0-1 5 min after sunset. Because we did 
not have dedicated ship time, observations were interrupted by 
periodic oceanographic sampling stations, during which time 
no seabird counts were made. Counts were not conducted 
within 10 km of oceanic islands where we landed (Christmas 
Island, Caroline Atoll, and Tarawa), but were made close to 
land in the Bohol, Sulu. and South China Seas. 

Our ship speed averaged 15 knots during observations, 
higher than the 10 knots that Gould & Forsell ( 1989) consider 
ideal. At this time of year in the tropical Pacific, our higher 



speed did not cause identification problems because few species 
were found in large numbers. When approaching or leaving a 
sampling station, we interrupted observations if the ship was 
moving less than 5 knots. 

Bird counts were contained within rectangular strips 
extending 400 m to each side of the ship. Observers stood left 
and right of the midline of the ship, counting all birds seen 
400 m or less ahead of the ship on their side, to a line 
perpendicular to the ship's direction at their position. Thus 
each hour of observation represents the pooled records of two 
observers watching a combined strip 800 m wide. Data were 
recorded on standardized field forms using local time. The 
following information was included: identification (to species 
or subspecies when possible), number of birds per sighting, 
feeding flocks and other associations, flight direction, plumage 
(adult, juvenile, sex), and weather. Due to poor lighting or 
weather conditions, some birds were identified only to genus or 
family. All birds sitting, flying, or flushed within the transect 
were counted, and their different behaviors noted. No birds 
were collected. The ship's position (latitude/longitude) and 
speed were recorded at the beginning and end of each hourly 
observation period. Ship-following species were noted during 
position checks and intermittently during the watch and were 
recorded when first seen. 

Because the visibility of different species at sea varies 
greatly, we subdivided the maximum 400 m transect width into 
3 bands corresponding to the approximate detection distances 
of each species: /. 1 00-m band — small species, ordinarily seen 
relatively close to the ship, includes shorebirds, storm-petrels, 
and Bulwer's petrel (Bulweria bulweri); 2. 30-m band — gadfly 
petrels, shearwaters, most larids, and anatids, and 3. 400 m 
band, large or conspicuous species such as boobies, frigatebirds. 



227 




Fig. 3. Status ol bud species observed along the cruise track of the R/V 
Akademik Korolev, 15 September— 3 1 October 1988. Nonbreeding 
visitors and migrants migrate north or south to breed. 



tropicbirds, and white terns (Gygis alba). These different 
detection distances were used as a basis for calculating the 
densities and area coverages for each different species (Hunt 
era/., 1 98 I;Briggs <■/<//.. 19871. Forease of comparison within 
regions and with previous studies (King, 1970; Tasker et ai, 
1984) and because the number of birds normally seen in 
tropical seas is lower than that observed in colder waters, bird 
densities were calculated in 3 ways: number/hour (per 
2 observers), number/10 linear km, and number/10 km 2 . In this 
paper, numbers/ 10 km' are used for comparison between 
species and regions; the other 2 calculated values are reported 
in Table 2. 

Prior to the expedition, we examined skins of Pacific 
seabndsat the US National Museum. Washington. DC. focusing 
on variable, polymorphic, and particularly difficult species to 
identify. Terminology used in this paper follows the American 
Ornithologists' Union Check-list (AOU. 1983. 1985). The 
order of species within families, and alternate English names, 
follow Harrison ( 1985). 

Results 

Forty-six species (or field-recognizable subspecies) were 
observed along the 14.892-km cruise track of the R/V Akademik 
Korolev from 15 September to 31 October 1988 (Figs. 1,2). 
These were represented by members ol 1 I families of seabirds 
(including gulls, terns, and skuas), shorebirds (including 
phalaropes), and ducks (Table 3). 

The most abundant species was the sooty tern {Sterna 
fusi ata), accounting for 62.8', of the total number of birds 
(Fig. 3), but restricted to Regions I through IV. Noddy tern 
{Anous sp.. primaril) brown noddy, A vtolidus), wedge-tailed 
shearwatei I Puffinus pacificus), and white tern followed as the 
next most frequently observed species We observed birds 
belonging to the following three broad groupings (Fig. 3. 
fable 4 ); 



a) Resident breeders . Seabirds may or may not have 
annual breeding cycles, which affects the temporal patterns of 
their dispersion at-sea. Within an archipelago, a species may 
exhibit asynchronous egg-laying periods on different islands 
or in separate colonies on the same island. For example, in the 
Line Islands, red-footed boobies (Sahi sulci) laid primaril) 

TABLE 2 

Seabtrd densities along the cruise track of the R/V Akademik 
Korolev, 15 September-31 October 1988. 



Region 



Species 



Number Birds 
Per Hour Peril) Per 

Linear Km 10 Km 3 



Phoenix petrel 
Tahiti/Phoenix petrel 
Herald petrel 
White-necked petrel 
Cook's petrel 
Stejneger's petrel 
Bulwer's petrel 
Unidentified petrel 
Flesh-footed shearwater 
Wedge-tailed shearwater 
Sooty shearwatei 
Christmas shearwater 
Audubon's shearw ater 
Unidentified shearwater 
Unidentified shearwater/ 

petrel 
Wilson's/Madeiran 

storm-petrel 
Leach's storm-petrel 
Unidentified storm-petrel 
Red-tailed tropicbird 
Masked booby 
Red-footed booby 
Brown hoohy 
Great frigatebird 
Bristle-thighed curlew 
Gray-backed tern 
Soot) tern 
Brown noddv 
Black noddv 

White tern 
Unidentified tern 

Phoenix petrel 
Mottled petrel 
Herald petrel 
While necked petrel 
Cook's petrel 
Bulwer's petrel 
Unidentified petrel 
Flesh-footed shearwater 
Short i.uled shearwatei 
Wedge tailed shearwater 
Soot) shearwater 
Christmas shearwatei 
Little shearwatei 



0.17 
0.03 
0.07 
0.10 
0.20 
0.06 
0.10 
0. 1 7 
0.03 
1.37 
0.50 
0. 1 3 
0.17 
0.10 
0.70 

0.37 



0.07 
0.01 
0.03 
0.04 
0.08 
0.02 
0.04 
0.07 
0.01 
0.53 
0.19 
0.05 
0.07 
0.04 
0.03 

0.14 



0.1 1 

0.03 

0.04 
0.07 
0.13 
0.06 
0.20 
0.11 
0.03 
0.88 
0.32 
0.09 
0.11 
0.07 
0.04 

0.72 



0.03 


0.01 


DOS 


0.17 


0.07 


0.32 


0.13 


0.05 


0.07 


0.80 


0.31 


0.49 


0. 1 3 


0.05 


0.07 


0.07 


0.03 


0.03 


0.17 


0.07 


0.01 


0.13 


0.05 


0.09 


0.07 


0.03 


0.04 


sss 


3.44 


5.73 


0.10 


0.44 


0.07 


0.17 


0.07 


1 1 


0.10 


0.04 


0.05 


0.03 


001 


0.03 


0.09 


0.03 


0.05 


Oils 


0.02 


0.03 


0.05 


0.02 


0.03 


52 


0.11 


0.17 


0.14 


0.05 


DOS 


0.09 


0.03 


0.16 


000 


0.03 


DOS 


0.05 


0.02 


0.03 


0.05 


0.02 


0.03 


52 


0.1 1 


0.17 


0.09 


0.03 


0.05 


0.05 


0.02 


0.03 


0.05 


0.02 


0.03 



228 



TABLE 2 -continued 



TABLE 2 - continued 



Region 



Species 



Number Birds 
Per Hour Peril) Per 

Linear Km 10 Km 



Region Species 



Audubon's shearwater 


0.05 


0.02 


0.03 


Unidentified shearw ater 


0.09 


0.03 


0.05 


Unidentified shearwater/ 


0.05 


0.02 


0.03 


petrel 








Leach's storm-petrel 


0.09 


0.03 


0. 1 6 


White-throated storm-petrel 


0.09 


0.03 


0.16 


Red-tailed tropicbird 


0.05 


0.02 


0.02 


Masked boob) 


0.32 


0.11 


0.13 


Red-footed boob) 


0.05 


0.02 


0.02 


Great frigatebird 


0.73 


0.24 


0.30 


Lesser golden-plover 


0.05 


0.02 


0.08 


Gray-hacked tern 


3.66 


0.12 


0.20 


Soot) tern 


77.91 


25.96 


43.27 


Brow n noddy 


0.09 


0.03 


0.05 


White tern 


0.73 


0.24 


0.31 


III Herald petrel 


0.56 


0.02 


0.03 


Cook's petrel 


0.06 


0.02 


0.03 


Bulwer's petrel 


0.1 1 


0.04 


0.20 


Unidentified petrel 


0.34 


0.11 


0. 1 8 


Wedge-tai led shearwater 


0.50 


0.16 


0.27 


Unidentified shearwater/ 


0.17 


0.05 


0.09 


petrel 








Leach's storm-petrel 


0.22 


0.07 


0.36 


Masked boob) 


0.28 


0.09 


0.11 


Ruddy turnstone 


0.06 


0.02 


DOS 


South polar skua 


0.06 


0.02 


0.03 


Sooty tern 


7.39 


2.40 


4.00 


Brown noddy 


13.11 


4.26 


7.09 


Black noddy 


0.06 


0.02 


0.03 


Unidentified noddy 


0.17 


0.05 


0.09 


White tern 


4 94 


1.60 


2.01 


IV Kermadee petrel 


0.05 


0.02 


0.03 


Bulwer's petrel 


0.47 


0.02 


0.80 


Flesh-footed shearwater 


0.05 


0.02 


0.03 


Wedge-tailed shearwater 


4.98 


1.76 


2.93 


Unidentified shearwater/ 


0.14 


0.05 


0.08 



petrel 
Wilson's/Madeiran 

storm-petrel 
White-tailed tropicbird 
Red-footed boob) 
Brown booby 
Great frigatebird 
Northern shoveler 
Lesser golden-plover 
Sharp-tat led/pectoral 

sandpiper 
South polar skua 
Parasitic jaeger 
Unidentified skua 
Black-naped tern 
Sooty tern 
Brown noddy 
Unidentified noddy 
White tern 



0.02 



0.01 



0.04 



0.91 


0.32 


0.40 


0.84 


0.30 


0.37 


0.05 


0.02 


0.02 


0.02 


0.01 


0.01 


0.02 


0.01 


0.01 


0.05 


0.02 


0.08 


0.12 


0.04 


0.20 


0.09 


0.03 


0.05 


0.05 


0.02 


0.03 


0.14 


0.05 


0.08 


0.02 


0.01 


0.01 


4.70 


1.66 


2.77 


4.21 


0.15 


0.03 


4.68 


1.65 


2.75 


0.87 


0.3 1 


0.38 



V Streaked shearwater 
Wedge-tailed shearwater 
Unidentified shearwater 
White-tailed tropicbird 
Lesser golden-plover 
Black-naped tern 
White tern 

VI Wedge-tailed shearwater 
Great frigatebird 
Red-necked phalarope 
Pomarine jaeger 
Unidentified tern 
Unidentified gull/tern 

VII Masked booby 
Brown booby 
Unidentified frigatebird 
Red-necked phalarope 
Pomarine jaeger 
Parasitic jaeger 
Unidentified skua 
Caspian tern 

Bridled tern 
Crested tern 
Unidentified tern 





Number Birds 




Per Hour 


Per 10 


Per 




Linear Km 


10 Km 


0.44 


0.17 


0.28 


1.53 


0.59 


0.99 


0.07 


0.03 


0.05 


0.15 


0.06 


0.07 


0. 1 5 


0.06 


0.28 


0.07 


0.03 


0.05 


0.07 


0.03 


0.04 


0.14 


0.03 


0.08 


0.07 


0.03 


0.03 


0.76 


0.27 


1.36 


0.07 


0.03 


0.04 


0.07 


0.03 


0.04 


0.14 


0.05 


0.08 


0.05 


0.02 


0.02 


0.25 


0.09 


0.11 


0.98 


0.04 


0.05 


1.61 


0.06 


2.96 


0.10 


0.04 


0.07 


0.15 


0.05 


0.09 


0.29 


0.11 


0.19 


0.25 


0.09 


0. 1 5 


0. 1 5 


0.05 


0.09 


0.10 


0.04 


0.07 


0.25 


0.09 


0. 1 5 



from May to June on Christmas Island in 1965 (Clapp, 1967) 
but from August to November on Caroline Atoll ( Kepler el ai. 
Subchapter 1 .2. this volume) data indicated an egg peak from 
August to November. Other species disperse variable distances 
from their colonies during the breeding cycle. The case of 
Audubon's shearwater {P. Iherminierl) was particularly 
interesting. Generally seen within ISO km of its breeding 
grounds, we found a bird 1,100 km away from its nearest 
known colony. Christmas Island, but approximately 400 km 
from Maiden, a little-known island that has unexplored potential 
habitat for Audubon's shearwaters. Our data also included 
range extensions of resident breeders. Forexample, the wedge- 
tailed shearwater, the most abundant shearwater in the Pacific. 
has rarely been recorded west of 180° in the Caroline Islands 
(Micronesia). Philippine Sea. and southeast Asia, yet 849! of 
our sightings (N = 245) occurred in this area. 

b) Nonbreeding visitors . These seabirds tend to be highly 
seasonal in their postbreeding movements. We observed birds 
that breed from the arctic to the antarctic. Forexample. Cook's 
petrels {Pterodroma cooki) breed on islands in temperate 
waters of the South Pacific, then undergo long transequatorial 
migrations into tropical and north temperate waters to spend 
their austral w inter. Because of the enormous distances of such 
migrations, and because the distributions of birds are still 
poorly known in some regions, our data added range extensions 
for Kermadee petrel [P. neglecta) and little shearwater 
(P. assimilii I. 



229 



TABLE 3 



Seabird abundance l\\ familj along the cruise track of the R/V Akademik Korolev, 15 September-31 Octobei 1988 

Regions 



Famik 


1 


II 


III 


IV 


V 


VI 


VII 


Total 






Total No. Buds 


457 


1.796 


495 


799 


34 


2(1 


67 


3,668 






No. Species 


26 


24 


13 


18 


6 


3 


s 








Procellariidae 


117 


35 


22 


243 


28 


2 


- 


447 


I12.1 1 *', i 




Hydrobatidae 


17 


4 


4 


1 


- 


- 


- 


26 


(0.71', ) 




Phaethontidae 


4 


1 


- 


39 


2 


- 


- 


46 


(1.2591 I 




Sulidae 


30 


8 


5 


38 


- 


- 


6 


87 


(2.3791 




Fregatidae 


5 


16 


- 


1 


- 


1 


2 


25 


(0.6891 i 




Anatidae 


- 


- 


- 


1 


- 


- 


- 


1 


10.03', i 




Charadriidae 


- 


1 


- 


2 


2 


- 


- 


5 


iO .14', I 




Scolopacidae 


4 


- 


1 


5 


- 


- 


- 


1(1 


(0.279! ) 




Phalaropodidae 


- 


- 


- 


- 


- 


11 


33 


44 


(1.2091 ' 




Stercorariidae 


- 


- 


1 


12 


- 


1 


11 


25 


(0.6891 i 




Laridae 


280 


1 .73 1 


462 


457 


2 


5 


15 


2.952 


(80 489: 




* For geographic 


limits of each 


region, see 


Figs. 1 and 


2. 






















TABLE 4 















Seabird species abundance along the cruise track of the R/V Akademik Korolev, 
15 September-31 October 1988. 



Species 



No. Birds Seen 



Total Status 



Region 



III 



IV 



VI 



VII 



FAMILY PROCELLARIIDAE 
Phoenix petrel {Pterodroma alba) 
["ahiti/phoenix petrel (Pi. rostrata/alba) 
Mottled petrel {Pi inexpe< tata) 
Kermadec petrel (Pi. neglecta) 
Herald petrel (Pi. arminjoniana) 
W hite necked petrel (Pi. externa) 
Cook's petrel (Pi. cooki) 
Stejneger's petrel (Pi. longirostris) 
Bulwer's petrel {Bulweria bulweri) 
I indent 1 1 led petrel 

Sue. iked shearwater {Calonectris leucomelas) 

Flesh footed shearwater {Puffinus carneipes) 

Wedge tailed shearwater (P. pacificus) 

Soots shearwater (/'. griseus) 

Short tailed shearwater (P. tenuirostris) 

Christmas shearwater (P. nativitatis) 

I .title shearwater ( /'. assimilis I 

Audubon's shearwater (/'. Ihenninieri I 

1 Inidentified shearwater 

i nidentified shearwater/petrel 

I Will 'i HYDROBATIDAE 

Wilson s/\ laden an storm-petrel 

(0i eanites oceanicus/Oceanodroma castro) 

White throated storm-petrel {Nesofregetta fuliginosa) 

1 each's si Dim petrel 1 0< eanodroma leucorhoa) 

Unidentified storm-petrel 



2 
3 


1 
7 


1 


6 


3 


1 


2 






3 


2 


2 


5 


2 


6 


1 
41 


1 
7 


9 


15 


2 




4 


1 




5 


1 




3 


2 




21 


1 


3 



113 



3 6 



447 




7 


RB 


1 


RB 


1 


MS 


-i 


WMS 


4 


WMS 


10 


WMS 


10 


WMS 


5 


MS 


27 


RB 


13 




6 


WMN 


4 


MS 


293 


RB 


17 


MS 


1 


MS 


5 


RB 


1 


MS 


6 


RB 


6 




31 





26 



1 


WMS 


2 


RB 


7 


WMN 


5 





230 



TABLE 4 - continued 



Spe 



No. Birds Seen 



Total Status 



Region 



IV 



VI 



FAMILY PHAETHONTIDAE 

Red-tailed tropicbird {Phaethon rubricauda) 

White-tailed tropicbird [P. lepturus) 

FAMILY SULIDAE 
Masked boob) (Sula dactylatra) 
Red-looted booby (S. sula) 
Brown booby (5. leucogaster) 

FAMILY FREGATIDAE 

Great frigatebird (Fregata minor) 

Unidentified frigatebird 



24 
4 
2 



16 



39 



36 

2 





46 






5 


RB 




41 


RB 




87 




1 


37 


RB 




41 


RB 


5 


9 


RB 




25 






23 


RB 


T 


■} 





FAMILY ANATIDAE 

Northern shoveler (Anas rlypeata) 



1 

1 WMN 



FAMILY CHARADRIIDAE 

Lesser golden-plover (Pluvialh dominica) 

FAMILY SCOLOPACIDAE 
Bristle-thighed curlew (Numenius tahitiensis) 
Ruddy turnstone (Arenaria interpres) 
Sharp-tailed pectoral sandpiper 
[Calidris acuminata/C. melanotos) 



5 WMN 

10 

4 WMN 

I WMN 

5 WMN 



FAMILY PHALAROPODIDAE 
Red-necked phalarope (Phalaropus lobatus) 

FAMILY STERCORARIIDAE 
South polar skua (Catharacta maccormicki) 
Pomarine jaeger (Stercorarius pomarinus ) 
Parasitic jaeger (S. parasiticus) 
Unidentified skua 

FAMILY LARIDAE 
Caspian tern {.Sterna caspia) 
Black-naped tern (S. sumatrana) 
Gray-backed tern (5. lunata) 
Bridled tern (5. anaethetus) 
Sooty tern (.V. fuscata) 
Crested tern (S. bergii) 
Brown noddy (Anous stolidus) 
Black noddy (A. minimis) 
Unidentified noddy 
White tern (Gygis allui) 
Unidentified tern 
Unidentified gull 
Unidentified lurid 



44 



2 


8 






266 


1.705 


133 


201 


3 


2 


236 


18 


5 




1 








3 


200 


3 

1 


16 


89 


37 



31 


44 
25 


WMN 




5 


WMS 


-> 


3 


MN 


3 


5 


MN 


6 


12 
2,952 




5 


5 


MN 




2 


RB 




10 


RB 


3 


3 


RB 




2.305 


RB 


-i 


2 


RB 




254 


RB 




6 


RB 




203 






146 


RB 


5 


7 
2 
2 




65 


3,668 





Totals 



457 1,796 



495 



799 



34 



2(1 



RB = resident breeder in the tropical Pacific. 

W = nonbreeding visitor 

MS = migrant: south to temperate/antarctic breeding colonies. 
MN = migrant: north " " /arctic 



231 



c) Direct migrants . This grouping includes shorebirds 
( plovers, sandpipers, phalaropes ) and ducks as well as seabirds. 
They nunc quickly through tropical waters from wintering 
grounds further north or south en route to breeding areas in the 
opposite hemisphere. Although some follow general routes 
and can be predicted at certain times of year, the overall dearth 
of studies in certain portions of the Pacific leaves much to be 
learned of their at-sea distribution. For example, Stejneger's 
petrel (P. longirostris) migrates from Chile to Japan, yet 
sightings had been scant in between before this expedition. 

Species Accounts 

Family Procellariidae 

Shearwaters and petrels provided the greatest species 
diversity (17) of any family. Seven breed and disperse within 
the tropical Pacific, 1 1 breed in the temperate South Pacific and 
migrate to wintering grounds in the North Pacific, and 1 breeds 
in the temperate North Pacific and migrates south to winter 
(Table 4). 

Although we did not encounter large migrating flocks, this 
family ranked second in total numbers seen (447; Table 3) and 
was particularly abundant in Regions I and II (Line and 
Phoenix Islands), after which species richness declined markedly 
to the west (Table 4). The western limit of procellariids 
(wedge-tailed shearwaters) was the Sulu Sea (08°47'N, 
121 28'E). 

Densities of individual species ranged from the rarer 
migrants at 0.03 birds/10 km 2 (Table 2) to the widespread 
resilient breeder, wedge-tailed shearwater, whose numbers 
peaked in the Line Islands at 0.88 birds/ 1 km 2 . Several species 
contributed to feeding flocks (Table 5 ). accounting for 10% of 
their participants. 

Phoenix Petrel (Pterodroma alba) : All seven of this 
rather uncommon species were sighted in Regions II and III. 
within its relatively small range in the central Pacific. Phoenix 
petrels sighted in the Line Islands were either within 1 .200 km 
of their breeding grounds at Christmas Island or flying west 
toward colonies in the Phoenix Islands. Their highest densities 
were in the Line Islands (0.1 1/10 km 2 ). None associated with 
other birds or participated in feeding flocks. 

Tahiti/Phoenix Petrel (Pterodroma roM rata or alba) : One 
Pterodroma found slightly east of Maiden Island (central Line 
Islands i was either a Tahiti or a Phoenix petrel, look-alikes 
difficull to distinguish in the field. Tahiti petrels breed in the 
Society and Marquesas Islands. 800 km and 1.670 km. 
respectively, from its observed position. 

Mottled Petrel (Pterodroma inexpcciata) : A single bird 
was seen flying southeast near Birnie Island (Phoenix Group) 
on 4 October. This species breeds in New Zealand during the 
austral summer, so this individual was likely migrating south 
from its winter quarters in the North Pacific. Its density in 
Region II was 0.03/10 km 

Kermadec Petrel (Plerodroma ne v lee la) : The Kermadec 
petrel breeds in several island groups just south of the Tropic 
ol Capricorn from Lord Howe Island to the coast of Chile. 
Formerly considered sedentary, recent records indicate thai it 
ranges widely into the North Pacific (Gould iV' King. 1967: 



Amerson, 1969; Harrison. 1985; Bailey et al., 1989). Several 
records exist for the area extending from just south of Hawaii 
to the Marshall and Phoenix Groups. An old record from Duke- 
of-York Island (Bismarck Archipelago) indicates that this 
species may also occasionally straggle almost to New Guinea 
(King, 1970). 

Kermadec petrels are not listed for the Gilbert Islands 
(Amerson, 1969) nor for any of the Caroline Islands (Pratt 
et al.. 1987): hence, the following observations extend the 
known range for this species: 2 Kermadec petrels ( 1 light and 
1 dark phase) were seen on the morning of 1 2 October at the far 
eastern edge of Micronesia (04°03'N. 163°30'E). approximately 
157 km south of Kosrae. The dark-phase petrel was flying 
directly south, while the light-phase bird had joined a feeding 
(lock of over 200 sooty terns mingled with small numbers of 
wedge-tailed shearwaters. Bulwer's petrels anil south polar 
skuas (Catharacta maccormicki). Kermadec petrels were seen 
only in Region IV with a density of 0.03/10 km 2 . 

Herald Petrel (Plerodroma arminioniana) : The Pacific- 
breeding range of this medium-sized gadfly petrel includes 
Easter Island, the Pitcairn Islands, Tuamotus. Marquesas. 
Gambiers, and, further west. Tonga and Chesterfield (Coral 
Sea). In the nonbreeding season, herald petrels remain primarily 
in the Southern Hemisphere, occasionally wandering north of 
the equator (Harrison. 1985; Bailey etal.. 1989 (where they are 
typically observed more than 150 km from land (King. 1970). 
We observed four adults (three dark phase, one light) in the 
Line, Phoenix, and Gilbert Groups. Two occurred just outside, 
the others within, the known pelagic range of this species 
(Harrison. 1985). The two dark-phase birds, beyond the 
eastern borders of their known range (05°24'N. 156 o 60'W), 
were flying southwest on 2 October (ca. 380 km east of 
Washington Island. Line Group). Densities were highest in the 
Line Islands (0.04/10 km 2 ). Three of the birds were Hying 
directly southwest. 

White-necked Petrel ( Pterodroma externa) ( includes both 
P. e. externa and P. e. cervicalis): We observed 10 of these 
gadfly petrels: 3 occurred in September and October within 
their main wintering grounds in the central Pacific between the 
equator and Hawaii (0.07/10 km 2 ), and 7 spanned the Phoenix 
Group, where densities were highest (0. 1 7/10 km 2 ). They are 
known to be abundanl in (he former location from May to 
November ( King. 1 967 ). 

There are two subspecies that breed on opposite sides of 
the temperate Pacific; the white-necked {externa) in the 
Kermadec Islands northeast of New Zealand, and the Juan 
Fernandez (cervicalis) in the Juan Fernandez Islands of Chile. 
Seven birds were flv ing south or southwest I four Hew together), 
presumably returning to breed on southern temperate islands 
during the austral summer. 

Cook's Petrel (Pterodroma cooki) : Cook's petrel breeds 
m the austral summer in New Zealand and the Juan Fernandez 
Islands. Although this iransequatonal migrant ranges widelj 
between the south and north temperate Pacific as tar as the 
Aleutian Islands, its pelagic movements are not fully understood, 
lew records exist from the central Pacific. There are two 
records from the Phoenix Islands and sightings near the Hawaiian 
Islands in (he northern spring (King, 1967; Harrison. 1985: 



232 



TABLE 5 

Geographic Distribution of Flocks. "No. Birds" refers to the number of each species present in all 

feeding (locks of each region. Species are arranged according to their overall relative abundance in all 

feedinc flocks. No flocks were seen in Regions V-VI1. 



Species 



Region 



111 



IV 



Flocks 



No. 
Birds 



Flocks 



No. 
Buds 



Flocks 



No. 
Birds 



Flocks 



No. 
Birds 



Sooty tern 
Noddy sp. 

Wedge-tailed shearwater 
Brown noddy 
White tern 
Red-footed booby 
Great frigatebird 
White-tailed tropicbird 
Masked booby 
South polar skua 
Stercorariid sp. 
Audubon's shearwater 
Bulwer's petrel 
Kermadec petrel 
"Shearwater/petrel" 



85 



6 


1.074 


2 


108 


1 
1 


200 
200 


1 


1 






3 


189 


1 


2 


3 


III 






3 


6 


4 


20 


1 


18 


1 


1 






1 


30 


2 


7 






1 


4 


1 


3 






1 


2 


1 


1 






1 
1 

1 


1 

1 
5 



Totals 



85 



1 .095 



239 



651 



Pratt el al., 1987). They have not been previously recorded 
from the Gilbert Islands (Amerson, 1969). Our records are as 
follows: 

Line Islands . Single birds were seen on 15 September at 
10°38'N. 156°16'W; 10°13'N. 156°19'W (Hying north): 
09°29'N. 156°26"W (with 2 Bulwer's petrels); 16Septemberat 
06°16'N. 156°50'W (flying north): 04°55'N. 157°00'W (flying 
south): and on 19 September at 00°22'S. 156°37'W. 

Phoenix Islands . All single birds flying south, as follows: 
2 October at 06°29'S, 162°19'W; 3 October at 05°16'S. 
166 o 53'Wand05 o 12'S. 167°06'W. 

Gilbert Islands . One bird flying southeast on 7 October at 
00°49'S. 179°10'E. 

We observed 12 adults (10 on transect) between 
15 September and 7 October 1988. Two had particularly pale 
plumage. Sightings occurred just north of and within the 
Northern Line Group, where densities were highest 
(0.13/10 km 2 ), in the Phoenix Group, and east of the Gilbert 
Islands. Of seven birds flying in a direct compass direction, 
four from the Line and Phoenix Islands were flying south, and 
one from the Gilbert Islands was flying southeast. Thus 71% 
of individuals of this species were flying in the approximate 
direction of their New Zealand breeding grounds, when such 
movements are expected. Only one Cook" s petrel was associated 
with other birds (two Bulwer's petrels). 

Stejneger's Petrel {PterodromalonvirostrisY . We observed 
two adult Stejneger's petrels in the Northern Line Group 



(05°56'N, 156°53'W; 06°24'N. 156°50'W) on 16 September 
1988, about 170 km east of Palmyra Atoll. There are few 
central Pacific records; POBSP personnel saw some near the 
Phoenix Islands (King. 1967). although this is not mapped in 
Harrison ( 1985). There appear to be no records of Stejneger's 
petrel between Hawaii and the Phoenix Islands. 

Additional sightings were made during the ICBP 1990 
Line and Phoenix Islands Expedition (ICBP. 1990; Kepler, 
1990). We observed no Stejneger's petrels in the Line Islands 
during March and April but saw several birds heading north in 
the same area during May ( 1 1° to 15°S, 149° to 15 1°W). 

Bulwer's Petrel (Bulweria hulueri) : This species was 
relatively common ( N = 27 ) throughout the four regions of the 
Pacific covered by the cruise. Because of its small size, it was 
only counted within 1 00 m of the ship ( see Methods). Since this 
species breeds and ranges at widely-scattered locations in the 
Central and western Pacific, all our observations fell within its 
expected range. Regions I through IV. Three-quarters of our 
sightings were in Micronesia, primarily south of Kosrae and 
north of Ulithi Atoll. None were seen west of Yap (139°E). 
The major Bulwer's petrel breeding grounds in the northwest 
Pacific (Bonin, Volcano, islands of Taiwan and China) are well 
north of Micronesia, and Bulwer's petrel is regarded as a 
species whose numbers decrease gradually with increasing 
distance from land ( King, 1970). Breeding occurs from April 
to September (King, 1967). Of our early October birds, equal 
numbers were heading either north or south. 



233 



The overall density of Bulwer's petrel in Micronesia was 
0.80/10 km 2 , lour times the density encountered in Regions I 
through III (Line. Phoenix and Gilbert Islandsi. 

Streaked Shearwater (CalonectrisleucomelusY . This large 
shearwater breeds on coastal islands off China, Japan and 
Korea and is known to travel southward during October and 
November toward its main wintering area in the New Guinea- 
northern Australia area. We observed six individuals, all from 
18-20 October in the far western Pacific close to the Philippines 
(Region V). Two were recorded as having darker plumage. 
Sightings were restricted to a very narrow band of ocean 
between 130°15'and 1 35° 14'E longitude. These were probably 
postbreeding migrants: three ( 50% ) were flying due south, and 
all were solitary. 

Flesh-footed Shearwater (Puft'iniis carncipes) : Flesh- 
footed shearwaters breed during the austral summer on islands 
off Australia and New Zealand, and winter in the North Pacific 
north of the subtropical convergence. We observed four flesh- 
footed shearwaters in Regions I. II and IV. all migrating birds 
returning to the south temperate Pacific to breed. Two were in 
the central Pacific south of Hawaii and in the Phoenix Islands, 
where the species has been reported in very small but regular 
numbers during the migration months. October to April ( King, 
1967). The rest occurred in a more expected sector of the 
western Pacific, although in a relatively narrow swath (05° 1 2'S 
to 03°59'N, 164°08'E to 163°46'E), viz. between the western 
Gilbert and far eastern Caroline Islands. The densities of these 
transequatorial migrants were equal throughout (0.03/ 10 knr). 

Wedge-tailed Shearwater ( Pitffimts pucitlcus ) : The wedge- 
tailed shearwater has long been considered the most common 
widespread shearwater of the southwest Pacific (Mayr, 1945; 
Jenkins, 1979). This species accounted for 8% of our total 
sightings (Fig. 4). It breeds on numerous islands throughout 
most of the tropical and subtropical Pacific from eastern 
Australia to Mexico south to the Pitcairn Islands. We observed 
293 individuals, 213 (73%) of which occurred in Micronesia. 
The two largest concentrations, of 150 and 22 birds, both 
occurred on 13 October near Pohnpei. Our sightings ranged 
from immediately south of Hawaii through the Line. Phoenix, 
and Gilbert Islands, Micronesia, and the Philippine Islands to 
the far western edge of its range at 121°28'E longitude. 

Although the pelagic distribution of this species is well 
known for the central and eastern Pacific regions and Marshall 
Islands (King, 1967. 1970. 1974b; Amerson, 1969), observations 
are surprisingly scarce in the Carolines-Belau-Philippine region 
( King, I974b;63. 93). For this reason, we discuss our records 
in waters west of 180' longitude (Table 6). where 84' i 
(N = 245) of our wedge-tail sightings occurred: 

a) Gilbert Islands (Region III. IS0-165°E). Although 
wedge-tailed shearwaters are known from this area ( I larrison, 
19X5). we have been unable to find records of specific sightings 
in a detailed summary of published data on wedge-tailed 
shearwater distribution (King, 1974b). They are not listed, 
even as visitors, in the Gilbert Islands by Pratt el id. ( 1987) or 
Amerson ( 1969). 




Fig. 4. Relative abundance of species or species groups on the cruise track of 
the R/V Akademik Korolev. Total number of birds seen was 3.6hS. 
belonging to 1 1 different families. 

Although a resident breeder on several of the northern 
Marshall Islands (Amerson, 1969), the closest recorded sightings 
to the Gilbert Islands are 10 dark-phase birds seen on 
2 November 1960 at Jaluit Atoll, southern Marshall Islands 
(Morzer Bruyns, 1965:58) and between 04°00' and 02°30'S, 
169-155°E in October 1951 southwest of the Gilberts between 
Niutao Atoll and the Admiralty Islands ( Mac Donald & Lawford, 
1954). These latter birds were reported as possibly P. came i pes. 
which has a totally different flight pattern. Our nine sightings 
were close to the equator at 179°E longitude, approximately 
300 km north-northeast of Arorae Island. All birds were flying 
south. 

b) Micronesia (Region IV, 165-136°E). Although known 
to be resident breeders in the northern Marshall Islands 
( Amerson, 1 969:295 ) and central Carolines ( Murphy. 1951:9), 
pelagic records of wedge-tails are sparse (King. 1974b:93). 
Breeding colonies are also difficult to locate. Prattefa/. ( 1987) 
state that this species is rare in western Micronesia: the 
westernmost record at these latitudes i s evidently 1 50°E ( District 
ofTruk). 

We observed wedge-tailed shearwaters (N = 213) 
throughout the Caroline chain of islands east to I32°59'E 
longitude: 73' ; of wedge-tails observed on our cruise were in 
this region. 

c) Philippine Sea and Basin (Region V. 136-125 Fl. Of 
21 wedge-tailed shearwaters seen west of 150°E (east of the 
Philippines and ca. 450 km north of Belau). all were dark phase 
(Table 6). Those in direct flight were heading either south or 
west. This species mas visit Belau more frequently than 
records indicate. ( )w en ( 197 1 ) does not mention its occurrence, 
although Pratt ei id. listed it as a visitor. The overall density of 



234 



wedge-tails in this area, generally depauperate in seabirds, was 
0.99 birds/10 km 2 , far higher than for any of the other five 
species seen there. 

d) Philippine Islands (Region VI. 125-1 17°E longitude). 
Our October observations confirm the presence of this species 
in the Sulu Sea (Delacour & Mayr, 1946; MacDonald & 
Law ford. 1954). where we saw two dark-phase adults flying 
south (0.08/10 knr). 

e) Color Morphs. Flight Directions, and Densities . Pacific 
Ocean Biological Survey Program studies revealed much about 
the migratory movements of the wedge-tailed shearwater in the 
central and eastern Pacific (King. 1967. 1970. 1974b). In 
general, the dark-phase population predominates south of 
10°N latitude, the approximate area covered by the east- 
flowing Equatorial Countercurrent and west-flowing South 



Equatorial Current. Dark-phase birds move northward 
following the currents during the northern summer, reversing 
their movements in the fall. 

Dark-phase birds observed during October would 
be expected to be heading south: of 39 individuals of known 
flight direction. 31 flew south and 4 southwest, toward 
the nutrient-rich waters of the Equatorial Countercurrent 
upwelling between 4°N and 9°N latitude (King, 1974a,b). 
Our largest concentration of wedge-tails was near Pohnpei 
(06°N, 157°E), contributing to the highest daily density 
of this species (10.95 birds/10 km : ). At this time the sea 
was "boiling" with fish, which attracted a great variety of 
seabirds. Such densities of Wedge-tails are comparable to 
those within the countercurrent latitudes in the central Pacific 
(King, 1974a). 



TABLE 6 

Distribution, abundance, and behavior of wedge-tailed shearwaters west of 180° 

7 October-23 October 1988. 



Pacific Ocean and inland Philippine Seas, 

















Overall 


















Nearest 




Density 










Location 






Island 


No. 


(birds per 




Flight 




Reg 


ion Lat. 


Long. 




Date 


or Group 


Birds 


10 knr) 


Phase 


Behavior Direction 


III 


01°00'S-02°55'N 


180°00'-165°00'E 


10/7-10/11 


Gilberts 


9 


0.27 










00°49'S 


179°10'E 


10/7 




Arorae 


8 




- 


Directed flight 


S 




00°49'S 


179°10'E 


" 






1 




- 




S 


IV 


02°55'N-11°10'N 


165°00'-136°00E 


10/11- 


10/17 


Micronesia 


213 


2.93 










03°53'N 


I64°08E 


10/12 




Kosrae 


1 
1 




Dark 


Directed flight 


S 




04°03'N 


163°31E 


" 




" 


1 

17 

1 




■• 


Feeding flock 
Directed flight 


N 




04°30'N 


161°57'E 


.. 




.. 


7 

1 




- 


» 


S 

NE 




05°24'N 


159°04'E 


10/13 




Pohnpei 


2 




Dark 


Solo feeding 


- 




05°32'N 


158°37'E 


" 




" 


1 




- 


Flushed by ship 


- 




05°43'N 


158°25E 








22 




- 


Feeding flock with 
I South Polar Skua 


- 




06°02'N 


158°06E 


" 




" 


1 




- 


Directed flight 


NE 




06 I7'N 


157°51'E 






(in sight) 


150 




- 


Feeding flock: 7 spp 
408 birds 


, - 




06°28'N 


157°41E 


" 




" 


1 

2 
2 




Dark 


Directed flight 


S 

sw 
s 




06°38'N 


157°31E 


" 




" 


2 




" 


" 


sw 




08°59'N 


155° HE 


10/14 




Murilo 
(Hall Is.) 


1 








s 


V 


ll°10'-10°00'N 


136°00'-125 o 00'E 






Yap, Belau 


21 


0.99 










11°02'N 


135°19'E 


10/18 




120 km NW of Yap, 


2 




Dark 


Directed flight 


s 










400 














1 1 00'N 


1 34°36'E 


- 




Belau 


1 

2 

1 
1 




" 


" 


s 

w 
s 




1 1 : '0()'N 


132°59'E 


10/19 






3 
10 






Circling 


w 




1 l°00'N 


132'59'E 


10/19 




Belau 


1 


0.99 


Dark 


Directed flight 


s 


VI 


10 OO'N 


I25 C 00'-117°00E 






Philippines 
(Sulu Sea) 


2 


0.08 










08°49' 


1 2 1 °44' 


10/22 




Negros. Mindanao 


1 




Dark 


Directed flight 


s 




08°47' 


121°28' 








1 








s 



235 



Dark-phase birds accounted for 979? of sightings where 
color phase was noted (57). Only two light-phase individuals 
were seen (Region 1. 10°46'N and ()6°45'N), both south of 
Hawaii, where the majority of the population is light phased. 

Short-tailed Shearwater {Puffinus tenuimstris) : A wide- 
ranging transequatorial migrant, with movements and breeding 
phenology similar to those of the flesh-footed and sooty 
shearwaters (P. griseus), this species breeds off southeastern 
Australia and winters off the west coast of North America and 
the Bering Sea. We observed one adult (2 October) heading 
south through the eastern extremity of its known migratory 
pathway, the eastern Phoenix Islands (06°29'S, 162°19'W). 

Sooty Shearwater (Puffinus griseus) : Sooty shearwaters 
arc wide-ranging, transequatorial migrants that breed from 
October to May in the Southern Hemisphere on islands off 
Australia. New Zealand, and Chile, migrating to equivalent 
high latitudes in the northern Pacific during the nonbreeding 
season. We observed very small numbers of birds on their 
annual southward migration. All 17 observations were in the 
south-central Pacific. We saw them on only 2 days: 
16 September (N = 15). east of Washington Island (Line 
Group) and 2 October (N = 2), between Sydney and Starbuck 
Islands (east of the Phoenix Group). All were adults 
Hying south, and none were feeding or associated with other 
species. Densities were 0.32 birds/ 10 km : in Region I and 
0.05 birds/10 km 2 in Region II. Enormous flocks of migrants 
have been encountered in this area ( King, 1 967: CBK. personal 
observation). 

Christmas Shearwater (Puffinus ncitivitatis) : The Christmas 
shearwater, a year-round resident in the tropical Pacific, tends 
to remain fairly close to its breeding grounds all year, although 
it is generally seen more than 180 km from land (King. 1970). 
We observed four individuals in the Line Islands (05°N, 
157°W) that were Hying west, and a single bird in the Phoenix 
Group, flying north. Densities were low: 0.09 birds/10 km 2 
(Line Islands) and 0.03 birds/10 km 2 (Phoenix Islands). 

Little Shearwater (Puffinus assiniilis) : This small, 
distinctive "aukish" shearwater breeds and disperses within 
south temperate waters, generally only occurring at-sea north 
to 25°S latitude. However, its pelagic dispersal is not well 
known and wanderers have been recorded near the Marquesas, 
Marshall, and Hawaiian Islands (King. 1967). Clapp ( 1967) 
listed a doubtful record from Christmas Island in the late 
I950's. We add a single straggler, seen on 4 October in the 
Phoenix Islands (03 59'S, 171°31'W) close to Birnie Island 
and Hying north. Within 20 minutes, we also observed the 
similar Audubon's shearwater: both flew close to the ship, 
enabling us to compare si/e, bill length, and the degree of white 
present on the underw ings. 

Audubon's Shearwater (Puffinus Ihcnninicri) : This small 
shearwater is resident in the central and western Pacific: our 
five sightings ( Regions I and II ) were within its known range. 

Our sightings of Audubon's shearwater at 02°N. 157 W 
were evidently from adjacent colonies on Christmas Island. 
However, we saw two birds flying northwest in the Southern 
I ine Islands (06 32'S, I52°35'W) on 21 September 1988, 
approximately 1. 1 00 km from Christmas Island and more than 
1,800 km from Phoenix Island, the nearest known colonics 



(Clapp. 1967; Stoddart. 1976: Garnett. 1983). Since this 
species usually ranges at sea within 180 km of its breeding 
islands (King. 1967). we speculate that an unknown colony lies 
w ithin the Southern Line Islands. Maiden Island, approximately 
400 km distant from the Audubon's shearwaters in question, is 
a likely possibility. Little visited by biologists, approximately 
one-third of its interior is covered by a landlocked, supersaline 
lagoon with subterranean connections to the sea, containing a 
maze of interconnected islets and salt flats (Garnett. 1983; 
RNZAF. 1986). These islets have never been surveyed by 
ornithologists (R.B. Clapp. personal communication). They 
provide potential habitat for Audubon's shearwaters, since 
they resemble sites occupied by this species on Christmas 
Island. Bloxham ( 1925) recorded two species of shearwaters 
of unknown identity on Maiden: Garnett ( 1983) proposes that 
one of them may have been Audubon's shearwater. Starbuck. 
another stark, arid guano island with a similar interior, is 
another probable source. 

On 5 May 1990. a single Audubon's shearwater was 
observed flying north-northwest around 10°S, 155°W, 
approximately 640 km due south of Maiden (ICBP. 1990). 
further suggesting the presence of a colony in the Southern 
Line Islands. 

In Region II, we saw one Audubon's shearwater in a 
feeding flock at 03 D 57'S. 171°31'W. near a large breeding 
colony (ca. 12.000 birds) on Phoenix Island (Garnett. 1983). 

Family Hydrobatidae 

The storm-petrels were represented by three (possibly 
four) species (Table 4) and five unidentified individuals. 
Small numbers (N = 26) were present at sea from the Line 
Islands west to Micronesia. Wilson' s/Madeiran storm-petrel 
in the Line Islands accounted for the greatest densities 
(0.72 birds/ 10km : ). Overall, storm-petrels accounted for().7'< 
of the total number of birds on the cruise (Fig. 4). Typically 
solitary feeders, they never participated in feeding flocks. Our 
westernmost observation was of a Wilson' s/Madeiran storm- 
petrel at 03°59'N, 163°45'E (south of Kosrae). 

Wilson' sSiorm-PelvcliOcecinitesoceiiniats) and Madeiran 
(Harcourt's) Storm-Petrel (Oceanodronui custro) : Wilson's 
storm-petrel breeds on subantarctic islands off South America 
and in Antarctica and ranges widely throughout all the world's 
oceans (Murphy. 1936: Murphy & Snyder. 1952). In the 
Pacific it migrates northward to wintering grounds within 
tropical and north temperate waters (Harrison. 1985). It is 
rarely seen in the Pacific except in the far east (King. 1967; 
Huber. 1971;Crossin, 1974). Records and sightings exist from 
the Marshall Islands. Solomons. New Hebrides. New Caledonia, 
waters close to Hawaii, the Phoenix Islands, and Christmas 
Island. 

The Madeiran storm-petrel, although not congeneric with 
Wilson's, appears remarkably similar in the field. It is a 
resident breeder in the tropical Pacific: therefore, its dispersal 
range overlaps with the wintering areas of Wilson's, primarily 
west of the international date line. 

Eleven Wilson/Madeiran storm-petrels were observed at 
00°28'S, 156°32'W, east of Jarvis Island. Hying southeast. A 
Wilson's, flying south, was located southeast of Kosrae 



236 



(04°03'N, 163°31'E). Huber ( 1971 ) has shown that Wilson's 
storm-petrels move through the Marshall Islands from April 
through September, sometimes in considerable numbers. 

White-throated or Polynesian Storm-Petrel (Ncsotrcuetta 
titliaiiiosa) : The white-throated storm-petrel is an uncommon 
central Pacific resident. We saw two (0. 1 6 birds/ 1 knr ) flying 
north on 5 October in the Phoenix Islands, 200 km northwest 
of the nearest land. McKean Island, where the world's largest 
population! 1.000 birds) breeds (King. 1973). Approximately 
500 birds also breed on nearby Phoenix Island, remaining in 
adjacent waters throughout the year, with limited dispersal 
eastward along the South Equatorial Current ( Harrison, 1 985 ). 

Leach's Storm-Petrel (Oceanodroma leucorhoa) : This 
species breeds at subarctic and temperate latitudes in the North 
Pacific, wintering primarily north of the equator. Its at-sea 
range is centered in the central Pacific, with greatest winter 
densities in a broad belt along the equator (Crossin, 1974). Our 
seven observations, all between and 3°S latitude, fell within 
the known range of the species (i.e., the Line, Phoenix, and 
Gilbert Groups). Five were flying north, and two were flying 
southeast. 

Family Phaethontidae 

Tropicbirds were represented by two species observed in 
small numbers (N = 46) from the Line Islands west to the 
Carolines, including the Gilbert Islands. Most sightings were 
of solitary birds during the morning hours, up to 300 km 
from the nearest landfall. The highest density was that of the 
white-tailed tropicbird {Phaethon leptums) in Micronesia 
( 0.40 birds/ 1 km : ), six times higher than elsewhere. Tropicbirds 
accounted for 1.3% of the total birds seen (Fig. 3, Table 3). 

Red-tailed Tropicbird (Phaethon rubricatuhi) : The red- 
tailed tropicbird ranges widely in the tropical and subtropical 
Pacific, breeding on many islands. Typically solitary, it is 
highly pelagic and is often observed many hundreds of 
kilometers from the nearest landfall (Harrison, 1985). Long- 
term studies have found that it is observed in roughly the same 
density regardless of distance from land (King, 1970). 

We observed five individuals in Regions I and II between 
1 7 September and 5 October. In the Line Islands, two occurred 
close to islands (Christmas. Maiden). The remainder were 
approximately 300 km equidistant from Caroline, Maiden, and 
Starbuck. all of which harbor small breeding colonies (Clapp, 
1967; Gould el al., 1974). although that on Caroline is the 
largest ( Kepler et al. . Subchapter 1 .2, this vol.). All birds were 
characteristically solitary; one was resting on the water. This 
was also found by POBSP, whose number of sightings exceeded 
one thousand: 87% of sightings were of lone birds, and 14% 
were sitting on the water (Gould et al.. 1974). Pacific Ocean 
Biological Survey Program found that birds were most 
commonly observed during morning hours: 80% of our birds 
were seen before 0820 h. All flying birds were adults flying 
south. One bird was observed in the Phoenix Islands 
approximately 300 km northwest of McKean Island, where 
some 500 breed (Gould et al.. 1974). 



Densities were highest in the Line Islands 
(0.07 birds/10 knr). Such at-sea abundance is low, probably 
because the species was breeding. On Caroline Atoll, for 
example, eggs and chicks were present in late September 
(Kepler et al., Subchapter 1 .2. this vol.). 

White-tailed Tropicbird (Phaethon leptums) : Like 
P. rubricauda, the white-tailed tropicbird is a resident breeder 
throughout the tropical Pacific. It is. however, less pelagic. Of 
the 41 individuals we observed, 38 were seen on 13 October, 
with Pohnpei in sight most of the day: most were alone or in 
groups of up to four or fewer. Maximum numbers were seen 
in the morning. Those seen with directed flight were generally 
traveling toward or away from Pohnpei. Ten (26%) were 
sitting on the water, and four had joined a large feeding flock. 
White-tailed tropicbirds favor high islands for nesting and are 
known to nest in trees and rocky cliffs on Pohnpei (Baker, 
1951). 

A single white-tailed tropicbird was sighted 400 km north 
of Belau. where it breeds on several islands (Baker, 1951). Our 
last sighting was of a lone bird on 21 October in Region V. on 
the far northwestern edge of the species' range (Harrison. 
1985). It was flying east approximately 200 km east of Samar 
Island (Philippines) at 10°49'N. 127°53'E. 

Family Sulidae 

Boobies were seen throughout the study area except in 
Region V, the Philippine Sea and Basin. All three pantropical 
species were seen in the Line Islands only. Although we saw 
only 87 boobies, this family ranked third in total numbers, 
contributing 1.3% of all individuals observed on the cruise 
(Fig. 4, Table 3). It was most common in the Line Islands and 
Micronesia (Figs. 5,6; Table 3). Many sightings were within 
80 km of their nearest breeding islands, but there were several 
exceptions, one being a possible new pelagic record of a 
masked booby (S. dactylatra) in the central South China Sea 
(see below). 

All three species of boobies commonly participate in 
feeding flocks. This tendency is greatest with the red-footed 
booby (5. sula)but is much less overall than for more gregarious 
species such as sooty terns or wedge-tailed shearwaters (King, 
1970). Our data ( 14 flocks) indicate that boobies participated 
in 21.4% of the flocks (Table 7), accounting for 1.6% of the 
total number of flocking birds (Fig. 7). Of 893 feeding flocks 
in a large study area centered on the Hawaiian Islands, boobies 
participated in 12.3%. although their numbers only accounted 
for 1.8% of the total (King. 1970). 

Masked Booby (Sula dactylatra) : Masked boobies, resident 
breeders in the tropical Pacific, were seen sporadically across 
the entire cruise track (N = 37). They were most common in 
the Line Islands, where densities reached 0.49 birds/10 knr. 
Small numbers were present in the Phoenix and Gilbert Groups 
and South China Sea (Table 2). Although wide-ranging 
throughout the Pacific, this species was not mapped as occurring 
west of around 143°E (i.e., the Marianas chain) (Harrison, 
1985) but is recorded elsewhere from the Philippines, coastal 



237 



IV 
N 799 



V 

N 34 




111 

N 495 




Ml 


PROCELLARIIDAE 


E3 


HYDROBATIDAE 


in 


PHAETHONTIDAE 


^ 


SULIDAE 


■1 


FREGATIDAE 


ED 


ANATIDAE 


EH 


CHARADRIIOAE 


M 


SCOLOPACIDAE 


E n 


PHALAROPODIOAE 


in 


STERCORARIIDAE 


□ 


LARIOAE 



VI 
N 22 



VII 
N 65 





Fig. 5. Relative abundance of birds bj famil} in Regions I-III. 



Fig. 6. Relative abundance ot birds b\ tanuh in Regions [\ \ II See Fig. 5 
for legend. 



Java. Indo-China, Malaya, and Borneo (Delacour & Mayr, 
1946; King & Dickinson. 1975). There appear to be no records 
from the central South China Sea, where we saw a single 
juvenile, flying west, on 2? October at 05°49'N, 107°30'E, 
a location equidistant (ca. 400 km I from Borneo, Vietnam, and 
the Malay Peninsula. 

Nonmigratory. masked boobies generally occur within the 
\ icinity of their breeding islands, hut the} roam far out to sea. 
They characteristically follow ships: one-quarter of the birds 
circled around or followed the R/V Akademik Korolev. 
Juveniles-subadults accounted for two-thirds of our overall 
sightings, especially in the Line Islands, where they breed. 

Red-footed Booby (Sula sulci) : This pantropical booby 
ranges across the entire tropical and subtropical Pacific, breeding 
in many localities. We observed 41 individuals in the Line. 
Phoenix, and Caroline Islands. 

Although nonmigratory, the lack of subadult birds at some 
colonies has given rise to the hypothesis that although adults 
arc relativel) sedentary, many juveniles disperse from their 
natal islands (Schreiber & Ashmole. 1970). Pacific Ocean 
Biological Survey Program personnel have shown that red- 
foots generally remain within the vicinity of breeding or 
roosting areas, and that adults are seldom encountered more 
than 80 km from land ( King. 1967). Our sightings conformed 
to this pattern. For example, we passed 500 km east ol Palmyra, 
which harbors the largest colony of red-loots in the world 
(25.000). yet saw none. The only red-foots seen in the Line 
Islands were three light morph adults, Hying east, within 
I 10 km of Maiden Island, where approximate!) 2.000 birds 



TABLE 7 

Participation b\ species or species groups in 14 feeding flocks 
alona the 14,892km cruise track. 



Species 






Flocks 


Participation 








Number 


Percent 


Sooty tern 


10 


71.4 


White tern 






8 


57.1 


Noddies 






5 


35.7 


Wedge-tailed 


shearwater 


4 


28.6 


Other shearw; 


Hers 


and petrels 


4 


2S.6 


Boobies 






3 


21.4 


Greal frigatet 


iird 




2 


14.3 


Stercoral! uls 






2 


14.3 


White-tailed tropic 


bird 


1 


7.1 



breed (Clapp. 1967). and 1 juvenile, also flying east, around 
460 km northwest of Caroline Atoll. In the Phoenix Group, we 
obsen ed one dark morph adult in a feeding flock about 100 km 
equidistant from several islands. 

In the central Pacific, red-foots lay from February to 
November, depending on the island and food supply (Nelson. 
1978). In the Southern Line Islands, the maun' egg-laying 
period from 1988 to 1989 was September to October (ICBP. 
1990: Kepler era/., Subchapter 1.2, this vol. i. which may have 
accounted for the small number of birds seen on our cruise. 
During March and May 1 990, at-sea observations throughout 
the Southern Line Islands recorded much larger numbers of 



238 



rod-foots than in September 1988. These were primarily 
juveniles dispersing after a successful breeding season. 

On 1 3 October we observed 30 red-foots in the vicinity of 
Pohnpei. Most of these were participants in two feeding flocks, 
one of which was the second largest on the cruise. Both flocks 
developed immediately after a storm. The proportion of light- 
to dark-phase birds at Pohnpei was 22:9. Red-foot density was 
highest on this day ( 1 .62 birds/10 km : ). Overall densities were 
1 times greater in Micronesia than elsewhere, due to the large 
numbers seen feeding near Pohnpei. 

Brown Boobv (Sula leuconaster) : This pantropical species 
ranges widely in the Pacific and breeds on almost every island 
group, although typically in smaller numbers than red-foots. 
The brown booby, less dependent on the tropical blue waters 
preferred by the other two species, is often found in inshore 
waters, including harbors and estuaries, which are more polluted 
and more accessible to man. Because of this. Nelson ( 1978) 
considers that the brown booby has suffered more than its 
congeners, probably accounting for their small numbers in 
populated areas such as Micronesia and the South China Sea. 
Brown boobies are seldom encountered more than 80 km from 
land (King. 1967), but Harrison (1985) noted that there is 
evidence to support small-scale dispersal. 

We observed only nine individuals in the Line Islands, 
Caroline Islands (Micronesia), and South China Sea. We 
attempted to relate these individuals to nearby colonies. In the 
Northern Line Islands (two sightings at 02°02'N, 157°37'W) 
the closest colony was Christmas Island (N = 100), 32 km 
distant. 

In Micronesia, the only brown boobies (two adults) were 
at 10°03'N, 150°02'E, approximately 150 km northwest of 
Magur, Namonuito Atoll, Truk (07°N, 147°E), and 220 km 
from uninhabited East Fayu Island, Truk (08°N. 151°E). 
Although brown boobies nest on uninhabited islands in the 
Marshall Islands (Amerson, 1969;SPREP, 1989),theircolonies 
are rare and little known in the Caroline Islands. Some birds 
were seen at Truk in 1945 (Baker, 1951). and earlier this 
century, "incredible numbers of seabirds." which Nelson ( 1978) 
suspects may have included brown boobies, were reported on 
West Fayu, Gaspar Rico, and Magur. Nelson further notes that 
"East Fayu is also a breeding site," implying a colony of this 
species. Although the ornithology of these islands is virtually 
unstudied (NID, 1945: Nicholson & Douglas, 1969; Owen, 
1 97 1 , 1 977a, b; Ray Fosberg, personal communication), several 
colonies of brown boobies may still exist in the Caroline 
Islands. It is unlikely that the brown boobies we observed were 
from the Marshall Islands, the nearest known colony of which 
is Enewetak. about 1.400 km to the east. 

Seabird information in the more remote islets of the 
Philippines and in the South China Sea is similarly scarce and 
outdated (Delacour& Mayr. 1946: King & Dickinson. 1975; 
Nelson, 1978; Harrison, 1985). Nelson states that "the status 
of the Brown Booby here is little known. There are some, could 
be many and may be few." Their only known nesting colony 
near Balabac Strait is at Tubbataha Reefs (68°N, 1 20°E) in the 
central Sulu Sea (Worcester. 1911, in Nelson, 1978, IUCN. 
1988a). 



In the South China Sea, we observed five brown boobies 
in the Balabac Strait close to numerous islands off the southern 
tip of Palawan (07°44'N, 1 16°58'E) and near northwest Borneo 
(05°92'N, 114°03'E and 04°40'N, 113°20'E). The closest 
known colony is Spratly Island (Nan-Sha Reefs), evidently the 
only remaining colony of brown boobies in Malaysian waters 
that has not been overexploited (Nelson, 1978) that was still 
extant 16 years ago (Haile, 1964, in Nelson, 1978). Both 
locations were within 30 km of coastlines. Our brown boobies, 
approximately 200 km west of this reef and flying east, could 
have originated there (or on several other little-known islands 
in the southwest Sulu Sea). 

Overall densities of brown boobies were low everywhere, 
but were four to six times higher in the Sulu Sea-South China 
Sea than in the Line Islands and Micronesia, respectively 
(Table 2). These distributional patterns are most likely due to 
the proximity of our cruise track to potential breeding islands 
and nearshore waters. We saw no brown boobies in feeding 
flocks. 

Family Fregatidae 

Frigatebirds were seen throughout our cruise track except 
for the Gilbert Islands and the Philippine Sea. The only 
identifiable species was the great frigatebird (Fregata minor): 
two distant, unidentifiable frigatebirds were seen in the central 
South China Sea. Surprisingly, no lesser frigatebirds (F. oriel) 
were seen, even though they winter in the western tropical 
Pacific (Sibley & Clapp, 1967). Densities (Table 2) were 
highest in the Phoenix Islands (0.30 birds/10 km 2 ), where the 
largest populations in the Pacific breed, and lowest in 
Micronesia, where frigates are known to be scarce. Overall, 
frigates accounted for only 0.7% of all birds observed on the 
cruise (Fig. 4). 

Great Frigatebird (Frcvcita minor) : We observed 23 great 
frigatebirds in the Line, Phoenix, and Caroline Islands. Densities 
were highest in the Phoenix Islands (Table 2), where over 
30,000 are known to breed (Stoddart. 1976). 

We saw only five great frigatebirds in the Line Islands, 
where around 13,000 breed (Clapp. 1967; Garnett 1983; 
Kepler et al, Subchapter 1.2, this vol.). These low numbers 
may reflect the fact that birds are concentrated on and near 
colonies during the breeding season (Clapp, 1967; Kepler 
et al.. Subchapter 1.2, this vol.). 

In Micronesia, we saw only one frigatebird, part of a 
large feeding flock near Pohnpei. Evidently frigatebirds are 
infrequent in the Caroline Islands (Baker. 1951; King, 1967; 
Pratt et al., 1987). We located only one reference to breeding 
colonies of this species in this region (Niering, 1961 (.otherwise 
they "are probably resident, especially in the eastern part" 
(Baker, 1951). 

The juvenile frigate in the southwest Sulu Sea (08°02'N. 
1 17°28'W) and the adult in the South China Sea near Vietnam 
(05°48'N, 106°5 l'E) could have come from anywhere, as they 
may wander thousands of kilometers (Sibley & Clapp, 1967; 
Nelson, 1975). Great frigatebirds were minor participants in 
feeding flocks, contributing only 0.3% of the total numbers 
(Table 8). 



239 



Family Anatidae 

Onl) one vagrant duck was observed (below), with an 
overall density in Micronesia of 0.01 birds/10 km 2 . 

Northern Sho\e\er (Anas chpeata) : This Holarctic breeder 
is a rare but regular vagrant to Oceania. Recorded locations in 
Micronesia are the Marianas, Pohnpei east to Wake, and the 
Marshalls (Baker. 195 1 : Pratt et al, 1987). 

We saw an unsexed bird in eclipse plumage three times 
w ithin a half hour on 1 3 October near Pohnpei. Each time it « as 
Hushed by the ship. Its characteristic bill shape and bicolored 
speculum made identification relatively eas) . 

Family Charadriidae 

We observed only one member of this family in small 
numbers during its southward migration from arctic breeding 
grounds. 

Lesser Golden-Plover ( Plmialis dominica) : We saw five 
golden-plovers: Phoenix Islands. Caroline Islands (north of 
the Hall Islands and Ulithi Atoll), and in the Philippine Sea. 
where its density was highest (0.28 birds/10 km 2 ). 

This long-legged plover breeds in Siberia and arctic Alaska, 
migrating annually to islands throughout the tropical Pacific. 
Most abundant from August to April, small numbers are 
present all year ( Pratt et al.. 1 987). It is also a common migrant 
in southeast Asia (King & Dickinson. 1975). 

During an oceanographic station on 19 October, one 
individual, uncounted on the transects, circled the ship several 
times and landed, remaining on board for a week, during which 
time it weakened considerably. Its final attempt at flight 
resulted in drowning. 

Family Scolopac idae 

Ten members and three species of this family were 
observed, all present in small numbers scattered within the 
Line. Gilbert, and Caroline Islands. All were migrants from 
Holarctic breeding grounds. 

Shorebirds are rarely seen in large numbers at sea. even 
during migration. For example. POBSP personnel on the 
Town soul Cromwell, working replicate tracks in a 1 7 1 million- 
ha study area from March 1964 to June 1965. observed only 
four species, each only once (King, 1970). Nonseabird migrants 
(sandpipers, plovers, phalaropes, and a duck) observed from 
the R/V Akademik Korolev accounted for 1.6% of the total 
birds seen (Pig. 3). 

Bristle-thighed Curlew (Numenius taliitiaisi.s) : One of 
the world's least-studied shorebirds. the bristle-thighed curlew- 
is considered rare throughout its range and is a candidate for the 

1 S Fish & Wildlife Service Endangered Species List (Gill, 
1990; Marks et al.. 1990). During the boreal winter, it is 
common in the Line Islands, northwest Hawaiian Islands, and 
southeast Polynesia hut uncommon to rare elsewhere in the 
Pacific (Pratt et al. I9S7). All our at-sea sightings were on 
16 September in the Northern Line Islands (density 0.09 
birds/10 km ) clustered in a small area north-northeast of 
Fanning Island and close to Christmas Island (04 55'N to 

02 02'N, 157 00'Wto 157 37'W). Half the birds were flying 
south, the most expected direction for September, as curlews 
are most abundant in (he Line Islands and Tuamotus between 



October and April (Gill. 1990; Kepler et al.. Subchapter 1.2. 
this vol.). We saw no more until arriving on Caroline Atoll one 
week later, where we estimated a population of >300. 

Ruddy Turnstone (Armaria interpres) : Another arctic 
breeder that winters in the Pacific, the ruddy turnstone is a 
widespread migrant to the Pacific, mostly betw een August and 
May. It is common from the Hawaiian Islands and Micronesia 
south to Samoa and Fiji (Pratt et al.. 1 987). A single turnstone 
circled the ship in late afternoon about 80 km east of Abemama 
Atoll, Gilbert Islands, on 8 October. 

Sharp-tailed (Calidris acuminata )/Pet:(ora\ Sandpiper 
(C. mclanotos) : These similar species are arctic breeders that 
winter in the South Pacific. We observed five birds in fall 
plumage flying south in a tight flock on 160ctoberat U°10'N, 
143°17'E. Both species have been recorded from Micronesia, 
especially C. acuminata in the western portion of the archipelago 
(Pratt etal., 1987). 

Family Phalaropodidae 

One member of this family was found in the South China 
Sea, one of its favored w inter quarters. 

Red-necked Phalarope (Phalaropus lohatus) : A 
circumpolar breeder, this phalarope migrates south to winter at 
sea. Although it can be encountered almost anv w here in the 
Pacific-southeast Asian region between the Aleutians and the 
equator, it favors two areas: coastal Peru and the South China 
Sea (Harrison, 1985). 

We first observed small flocks of this species on 
22 October on the eastern edge of the Sulu Sea (08°59'N, 
123 C 12'E). They continued to appear thereafter, 
characteristically flushing from the water close to the ship's 
sides. Those that were not flushed were observ ed resting on the 
water surface or feeding in areas where currents or wind 
produced debris or foam lines. 

We saw 44 phalaropes. 1 .20^ of the total number of birds 
on the cruise. Their highest density was in the South China Sea 
(2.96 birds/ lOknf). 

Family Stercorariidae 

Three species of skuas/jaegers were seen (N = 13) 
plus 12 unidentified stercorariids (Table 4). The) occurred 
primarily in Micronesia and the South China Sea. where 
densities were 0.16/10 km 2 and 0.35/10 km. respectively. 
None were seen in the Line and Phoenix Islands, or in the 
Philippine Sea anil Basin (Regions I. II. V). Stercorariids 
accounted for 0.795 of all birds observed on the cruise. 

Although ranging w idely. movementsoi skuas and jaegers 
are little known except that some migrator) pathways tend to 
follow coastlines (King. 1967). Some juveniles evidently 
remain on the wintering grounds all year. 

Only the Pomarine jaeger has been recorded lor southeast 
Asia (Tenasserim, Malaya, central Thailand, and Hong Kong) 
(Kingcx; Dickinson. 1987:153). However. King A: Dickinson 
noted that three other species ma) occur there. We report 
possible first sightings of the parasitic jaeger in the South China 
Sea. Since all siereoranid sightings are rare for the western 
Pacific and southeast Asia, our data on I 2 unidentified species 
follows: 



240 



a) 1 2 October, two in feeding flock at 04°03'N, 163°3 1 'E; 

b) 13 October, five flying north, one in feeding Hock 
at 06°16'N, 157°51'E. Heavy squalls and winds from the 
north: 

c) 1? October, one Hying east at 10°19'N, 147°39'E; and 

d) 25 October, six Hushed by ship from resting position on 
water at 05°44'N. !08°01'E. 

South Polar or McCormick's Skua iCaiharacta 
maccormicki) : This thickset, gull-like seabird breeds in 
Antarctica. It is an uncommon migrant in the Pacific, known 
only from the Hawaiian. Line, and Phoenix Islands, possibly 
west to the Gilbert Islands (Pratt etui. 1987). Harrison (1985) 
notes that juveniles disperse much further north than adults, 
possibly following a clockwise path across the Pacific rim past 
Japan (May-July) to British Columbia and California 
(September-October), then back to Antarctica to breed in the 
austral summer. 

We observed five south polar skuas (three light- and one 
dark-phased adults and one juvenile) in Regions III and IV 
(Gilbert and Caroline Islands). All Hew very close to the ship, 
and four participated in feeding flocks. 

Our sightings fall within the known range of this species 
( Harrison, 1 985 ), although other authorities indicate that south 
polar skuas do not range this far west (Mayr. 1945: Baker. 
1951; Pratt. 1987). Since published records are sparse (they 
were not listed in Baker. 195 1 ). our exact sightings follow: 

a) one juvenile, 8 October, flying south at OTOO'N, 
172°40'E, just south of Tarawa, Gilbert Islands. 

b) two adults, light phase, 12 October, in a feeding flock 
at 04°03'N, 1 63°3 1 'E, south of Kosrae. The feeding flock was 
composed of six species: two south polar skuas, two Pomarine/ 
parasitic jaegers. >200 sooty terns. 17 wedge-tailed shearwaters 
(dark phase). 1 Kermadec petrel. 2Bulwer'spetrels. Itdeveloped 
in midmorning immediately after a series of rain squalls had 
moved through the area from the south. 

c) two adults (one very dark, one light), 13 October at 
05°43'N, l5°25'E,just south of Pohnpei. The dark morph was 
flying southeast, the light one had joined a small feeding flock 
of 22 wedge-tailed shearwaters. Four hours later heavy winds 
and squalls hit us from the north, after which a massive 
concentration of fish attracted a large feeding flock which 
included an unidentified skua. 

The highest densities of south polar skuas were in 
Micronesia (0.05 birds/10 km : ). 

Pomarine Jaeger or Pomarine Skua {Stercoruriits 
poinariimsY . This Holarctic-breeding jaeger migrates south to 
winter throughout the tropical, subtropical, and temperate 
areas of the Pacific. Evidently in good breeding years adults 
move south in August-September, followed by juveniles in 
September-October (Harrison, 1985). Our entire cruise track 
was within its known range. 

From 22 to 25 October we observed three Pomarine 
jaegers in the Sulu Sea and South China Sea ( Table 4 ). One was 
a light-phase adult, the rest, juveniles. One juvenile was 
flushed from a resting position on the water, while the others 
flew east or south. The highest densities of this species were 
in the South China Sea (0.07 birds/ 10 km 2 ). 



Parasitic Jaeger or Arctic Skud(Stercorarius parasiticus) : 
The Holarctic breeding and Pacific wintering ranges overlap 
those of the Pomarine Jaeger (Harrison. 1985). Since this 
species has not been recorded from the central Pacific and is 
uncommon in the western Pacific (for example Bismarck 
Archipelago and Bonin Islands), its migration routes are thought 
to remain close to the continental coasts (King, 1967). King& 
Dickinson ( 1975) mentioned that it might occur in southeast 
Asia. 

Our observations came from the Caroline Islands and 
South China Sea. One juvenile was sighted on 12 October, 
south of Kosrae (04°22'N, 162°2 1 'E) flying east; and one adult 
was seen on 14 October north of the Hall Islands (09°46'N, 
154°43E), flying southeast. We also saw parasitic jaegers on 
24 October off the north coast of Borneo (04°40'N, 1 13°20'E. 
one juvenile, flying southeast) and on 25 October in the central 
South China Sea (05°49'N, 107°30E, one adult dark morph. 
flying southwest). 

Three of our five birds on transect were headed in a 
southerly direction, which is appropriate for October, when 
they dispersed south from their breeding grounds. 
Greatest densities (Table 2) were in the South China Sea 
(0.09 birds/ 10 knr). which fits the hypothesis that migratory 
routes tend to follow continental coasts. 

An additional parasitic jaeger was encountered during our 
3-day layover near Mui Bai Bung Cape, Vietnam at 05°47'N, 
107°45'E): On 27 October at 1605 ha juvenile, dark morph 
parasitic jaeger pursued an adult male shikra (Accipiterbadius) 
with prey, a barn swallow (Hirundo rustica), for about 
4 minutes (Ellis et ai, 1990). Finally the jaeger bound to the 
shikra and/or its prey. All three birds whirled about three 
revolutions and fell into the sea. After about 7 seconds in the 
water, the shikra flapped away. The jaeger remained in the 
water with its prize, drifting near the ship's stern. 
Oceanographers later found the swallow's head in a plankton 
net. 

Family Laridae 

This family was represented by 9 species of terns, plus 
4 unidentified larids. totaling 2.953 individuals. Larids 
accounted for 81% of all birds sighted on the cruise. They 
dominated the pelagic waters of Regions I to IV, after which 
their numbers diminished considerably. Sooty terns (Sterna 
fuscata) were the most abundant species overall, accounting 
for 62.8% of all birds seen on the cruise (Fig. 3), with a 
maximum density of 43.27/10 knr in the Phoenix Islands. 

Larids. which characteristically breed on oceanic islands 
(sooty tern, black and brown noddies, white tern), were present 
in far greater abundance than species that typically inhabit 
areas closer to continents (Caspian tern, S. caspia; black-naped 
tern, S. sumatrana; and bridled tern, 5. anaethetus). We 
expected to see far more coastal birds in southeast Asia. 
Human population pressures and habitat alteration evidently 
weigh heavily on the natural resources in this area. 

Sooty, white, and noddy terns were major participants in 
feeding flocks, accounting for 88.129! of all Hocking birds 
(Table 5). 



241 



Caspian Tern (Sterna caspia) : This large, cosmopolitan 
tern primarily breeds inland in Europe (including the Baltic Sea 
and interior USSR) and North America, along sea coasts and 
inland along rivers and lakes, migrating and dispersing 
southward during the nonbreeding season, remaining very 
close to continental coastlines. Apart from its winter status in 
Japan as a visitor, the Caspian tern is rare or uncommon in 
southeast Asia (Harrison. 1985: King & Dickinson. 1987): 
records are from Cambodia. Laos, extreme south Vietnam, and 
Thai land, but not from Singapore (Hails, 1987). We observed 
five Caspian terns. Hying east and south, on the last day of the 
cruise. 31 October, in Singapore Roadstead. Absent elsewhere, 
this species had a density of 0.49 birds/10 knr here. 

Black-napcd Tern [Sterna sinnatrana) : A small, mostly 
white, rather delicate tern, the black-naped is mainly sedentary, 
breeding in the warm watersof the western Pacific and southeast 
Asia. It is primarily a resident of seacoasts and offshore islets 
(Harrison. 1985; King & Dickinson, 1987), extending eastward 
no further than 180° longitude (King, 1967). 

We recorded two, one adult 80 km north of Ulithi Atoll 
(District of Yap, Micronesia) and a juvenile approximately 
550 km east of the Philippines. 

Gray-backed or Spectacled Tern (Sterna lunata) : The 
gray-backed tern breeds in the Hawaiian, Phoenix, Line, and 
Tuamotu Islands, plus several other island groups to the 
southwest. At sea, it occurs as a vagrant or migrant w ithin these 
approximate boundaries (King, 1967; Harrison, 1985). Their 
range is poorly known, probably because they are similar to. 
and often associated in small numbers with, the abundant sooty 
tern. 

We recorded eight gray-backed terns in the Line and 
Phoenix Islands. Two adults w ere seen on 20 September within 
30 km of Maiden (04°03'S. I55°01'W). Within the Line 
Islands, gray-backed terns are known to breed on Christmas. 
Jarvis. and Maiden (Clapp. 1967). thus we were probably 
observing birds from the Maiden colony (500-1.000 birds), 
which would have been breeding at that time (Clapp. 1967; 
Grossman & Grossman. 1974; Garnett. 1983). 

In the Phoenix Islands we saw two birds centrally in the 
group and six ( including one juvenile ) north of McKean. Gray- 
backed terns breed on Enderbury (N = 10.000 birds). Phoenix 
( N = 1 .800 birds), and McKean ( N = 23.000 birds) < King. 1 973; 
Garnett, 1983). 

Gray-backed tern densities were greatest in the Phoenix 
Islands (0.20 birds/10 km>. close to the largest breeding 
colonies in the world. Because breeding was in progress, most 
tems were likel) to be on or near land. 

Bridled or Brown-winged Tern (Sterna anaethetus) : 
Frequenting offshore waters and open ocean in the tropics and 
subtropics, tins species is the ecological counterpart of 
S. lunata in the western Pacific. We identified three bridled 
terns in Singapore Harbor; they are known to be offshore 
residents of the Malay Peninsula (Hails. 1987; King & 
Dickinson. 1987). 

Sooty Tern (Sterna luscata) : The most abundant seabird 
in the tropical Pacific, the soots tern breeds colonial!} in large 
numbers on almost every island group. Ii ranges widely 
between the tropics of Capricorn and Cancer. 



We observed this species in the Line, Phoenix, and 
Gilbert Islands and Micronesia. It was the most abundant bird 
observed on the cruise (N = 2.305), nearly eight times more 
numerous than the secondmost common species, the wedge- 
tailed shearwater (Table 4). Sooty terns comprised 62.8% 
of the total number of birds observed (Fig. 4). 

Sooty terns ranged throughout the Line Islands, with 
heavier concentrations near Christmas and Jarvis. Major 
colonies, of approximately two million each, exist on these two 
islands. The Christmas populations have declined drastically 
in recent years (from an estimated 15 million 10 years ago) due 
to predation by rats and cats, direct human exploitation for 
food, and reproductive failure due to the 1982-1983 El Nino 
Southern Oscillation (King. 1973: Gould. 1974b; Garnett. 
1984; Schreiber & Schreiber. 1984; E. Schreiber, personal 
communication). 

In Region II. sooty tern populations are estimated at 
around four million (Gould, 1974b). Here we observed the 
highest density of all species (43.27 birds/ 1 knr), eight times 
greater than sooties in the Line Islands (Table 2). Sooty terns 
are abundant year-round in the Phoenix Islands, with highest 
densities recorded from October to December (Gould. 1974b). 

Their numbers at sea depend on their breeding cycles, 
which are often complex and unpredictable, even within a 
single island. In 1965, and from 1988 to 1990. Caroline Atoll 
had 19 colonies, each on a different cycle, giving rise to egg- 
laying in virtually every month (Clapp & Sibley. 1971a: 
Anne Falconer, personal communication; Kepler et <;/.. 
Subchapter 1.2, this vol.). 

In Region III (Gilbert Islands) smaller numbers of sooty 
terns were observed (N = 1 13): we found moderate densities 
(4.00/10 knr). although the species is reported as uncommon 
at sea (Amerson. 1969). Our cruise track passed close to 
Aranuka. which is thought to harbor breeding colonies 
(Amerson. 1969). 

In Region IV (Micronesia), we only observed sooty terns 
south of Kosrae and Pohnpei (N = 201): 200 of these were 
members of a single feeding flock. This species is known to 
breed in small numbers in Micronesia on Pohnpei. on several 
uninhabited atolls in the northern Marshalls (Baker. 1951; 
Amerson. 1969). and on Helen Reef (Engbring. 1983). 

Although ranging entirely across the Pacific Ocean through 
southeast Asia into the Indian Ocean (Harrison. 1985). sooty 
terns have been seen only infrequently in the far western 
Pacific (Baker. 1951; Gould. 1974b). The POBSP found that 
the perimeter of their peripheral breeding localities is also the 
normal limit of their pelagic dispersal, and records beyond this 
have historically been attributable to climatic disturbances 
such as typhoons (Gould. 1974b). 

Sooty terns, highly gregarious feeders and breeders, 
participated in 7 1 .49! of all 1 4 feeding flocks and contributed 
70.8'f of all birds within the flocks (Tables 5.7; Fig. 7). Of 
2,305 sooty terns observed. 1.467 (64' r) were in feeding 
Hocks, as was also found (65'i ) by Gould ( 1974b). Over our 
entire cruise track, only 4'i were juveniles, suggesting that few 
soot) colonies had bred recently. By contrast, during March to 
May 1990 in the Line Islands, a much larger proportion of 
juveniles were observed (ICBP. 1990). 



242 



Crested Tern ( Thalasseus hcr^ii) : We observed only two 
of these large, coastal, yellow-billed terns. Although a resident 
breeder on several island groups in the central and western 
Pacific and southeast Asia, it is everywhere uncommon ( King, 
1967; King & Dickinson. 1967; Amerson. 1969). Our birds 
appeared on the last day of the cruise as we entered Singapore 
Roadstead. 

Brow nor Common Noddy (Anous stolidus) : This abundant 
tern breeds on almost all island groups in the Pacific, extending 
westward into southeast Asia (King, 1967; Pratt el ai. 1987). 
Although widespread, it generally occurs within 80 km of its 
breeding colonies. Much remains to be learned of the precise 
details of its dispersal patterns (Harrison, 1985). 

The brown noddy was the third most abundant bird seen 
from the R/V Akademik Korolev (N = 259). An additional 
203 noddies, seen in Micronesia, could not be identified to 
species. Since we saw no black noddies (A. minutus) in 
Micronesia, it is likely that they were also A. stolidus. 

Although we sighted brown noddies in the Line, Phoenix. 
Gilbert, and Caroline Islands, we counted only five birds total. 
All well out to sea, some were 900 km from the nearest known 
breeding colony (ca. 7.700 birds at Phoenix Island. POBSP 
estimate, in Garnett, 1983). 

Despite the fact that the Gilbert Islands are heavily 
populated, brown noddy densities were highest there 
(7.09 birds/10 km 2 ): they were the most abundant bird in the 
region (Table 2). This species, which can nest in trees, is much 
more tolerant of humans than most seabirds. Pacific Ocean 
Biological Survey Program personnel also found that/4, stolidus 
was common throughout the Gilberts up to 80 km from land 
(November to December. 1962 and 1964). It breeds on six 
islands in the southern Gilberts, including Abemama, near 
which we observed 235 brown noddies on 8 October. 

In Micronesia, 18 brown noddies and 200 unidentified 
noddies (probably A. stolidus) close to Pohnpei probably 
originated on that island, where they are known to breed 
(Baker. 1951). 

Brown noddies were regular members of feeding flocks, 
contributing 5% of their numbers (15% if the unidentified 
noddies are included) and participating in 36% of flocks 
(Tables 5.7). Corresponding figures from King (1970) are 
7 and \\9c. 

Black or White-capped Noddy (Anous minutus) : This 
species, smaller and darker than the brown noddy, breeds and 
disperses over the same range. It is more sedentary, however, 
and feeds closer inshore. Water temperature plays an important 
part in the pelagic distribution of both noddy species (Murphy, 
1936): both are largely absent from cold-water upwellings. 

Despite the black noddy's widespread distribution in the 
Pacific, we only observed six. Five were in the Line Islands 
within a few kilometers of Christmas Island, where 
approximately 10,000 breed (Garnett, 1983). The remaining 
bird was in the Gilbert Islands, where A. minutus may still 
breed. Twenty years ago there were small breeding colonies on 
eight islands (Amerson, 1969), including Maiana, which lay 
within 20 km of our cruise track. 

Greatest densities (0.11 birds/10 km 2 ) were within the 
Line Islands, where breeding populations exceed 40,000 
(Stoddart, 1976). Colonies are smaller in the Phoenix Islands, 



but it is surprising that we did not see one, as we passed within 
80 km of Orona ( Hull ). where 1 0.000 birds have been reported 
(King, 1973). None participated in feeding flocks. 

White or Fairy Tern (Gve/.v alba) : Tolerant of man, the 
white tern breeds throughout the tropical Pacific. As with 
noddies, it is most common within 80 km of its breeding 
colonies, but may wander great distances out to sea (King, 
1967; Pratt et ai, 1987). 

The white tern was the fourth most common bird (Fig. 4) 
seen on the cruise (N = 146), occurring from the Line Islands 
to the Philippines. Many sightings were of ones and twos; the 
largest group (18) had been attracted to a large feeding flock 
south of Pohnpei, Caroline Islands. 

At-sea densities of white terns, as for many other tropical 
species, reflect in part their breeding phenology, especially for 
the more remote islands. Line Islands densities were low 
(0.05 birds/10 km 2 ) in September 1988 but much higher in 
March to May 1990 (ICBP, 1990). Perry (1980) estimated 
17,050 white terns in the Line Islands and noted that they 
ranged widely and bred all year. Densities in the Phoenix 
Islands were higher (0.31/10 km 2 ), a surprise considering that 
the overall population (10,000 birds; Clapp, 1967) is smaller. 

White tern densities were greatest (2.01 birds/10 km 2 ) in 
the Gilbert Islands, due in large part to 89 birds seen the day 
before anchoring at Tarawa. In Micronesia, 36 G. alba were 
observed south of Kosrae and Pohnpei, and one was found 
north of Ulithi Atoll. Baker ( 1 95 1 ) implies that it is common 
in Micronesia, especially on low islands: Ulithi is listed as a 
breeding site. 

White terns are not listed for the Philippines (Delacour & 
Mayr, 1946; King & Dickinson, 1975). The species evidently 
rapidly decreases in density west of Micronesia. Our most 
westerly sighting, an adult flying southwest at 1 1°00'N, 132°59'E 
was approximately 480 km north-northwest of Belau, the 
closest landfall, and on the extreme western Pacific limit of the 
white tern's known range. This is the maximum distance that 
we observed this species from land. 

Feeding Flocks 

A seabird feeding flock is regarded as an association of 
five or more individuals acting as a unit (Gould, 1974b). In this 
study we encountered both monospecific and mixed-species 
flocks. We did not see any large associations of seabirds that 
were not feeding flocks (i.e., flocks of direct migrants). The 
following analysis deals with the size, abundance, composition 
(species and family), and geographic distribution of the 
14 feeding flocks we encountered. 

Flock Size and Abundance 

Flock sizes ranged from 8 to 420 birds, averaging 
148 birds per flock. The total number of birds in feeding flocks 
(N = 2,070) represented 56.5% of our bird sightings, a figure 
comparable to the 69.5% found in 893 flocks by POBSP ( King, 
1970). 

The smallest flock (N = 8), of seven white terns and one 
brown noddy, occurred in the Gilbert Islands on 8 October. 
The largest flock (3 October in the Phoenix Islands) was 
monospecific (420 sooty terns). The second largest (N = 408) 



243 



contained 7 species and was predominantly composed of 
wedge-tailed shearwaters (8695 ) and unidentified noddies. It 
was encountered on 13 October adjacent to Pohnpei. Three 
environmental factors undoubtedly contributed to the size and 
diversity of this flock: 

/. The flock was near a high island at 07°N. within the 
Equatorial Countercurrent boundaries (ca. 4°N and 9°N ). where 
particularly rich upwellings provide feeding grounds for both 
fish and birds (Ashmole & Ashmole, 1967: Gould, 1974a); 

2. Flying fish, considerably more abundant compared to 
the previous 3 days, indicated increased productivity; and 

3. The flock was attracted to masses of predatory fish 
(primarily tunas) vigorously leaping from the water, causing 
the water to "boil." 

Flock Composition 

Five families (Laridae. Procellariidae. Fregatidae, Sulidae, 
Phaethontidae ), 1 2 species, and 3 unidentifiable groupcategories 
were represented in feeding flocks. Sooty terns dominated 
the feeding flocks, followed numerically by noddies and wedge- 
tailed shearwaters (Fig. 7; Tables 5,7). These three species 
groups accounted for 95% of all birds seen in flocks. 

Sooty terns participated in 71.4% of all flocks (Table 7). a 
similar proportion to that found by POBSP (76.0%) over a 
longer time frame (King, 1970). White terns were found in 
57% of the flocks, noddies (primarily brown noddy) in 38%, 
and wedge-tailed shearwaters in 29%. 

White terns were also regular, though minor, members of 
pelagic feeding flocks, found in 57% of all flocks and 
representing 2% of total numbers (Table 7, Fig. 7). King 
( 1970) found that white terns participated in 9% of flocks and 
noted that the relatively high flocking tendency of white terns 




lie. 7. Relative abundance >>i spe*. us ..i -,|x\ io groups loniul in .ill feeding 
flocks (N = 14). Total numbei ol birds was -.(172 (\ = I4S). 



was of interest since this species had been thought to be a 
solitary feeder. Our data also suggest that the white tern 
commonly joins mixed-species feeding flocks. 

White terns are generally more solitary on land and at sea 
and have not been found to exhibit such a high rate of flock 
participation (57.1%) as we observed (King, 1970). This 
species, however, joined flocks only in small numbers. Overall. 
five white terns on our cruise joined over half the flocks, but 
their total numbers within flocks accounted for only 2.1% 
(Fig. 7; Table 7). 

Geographic Distribution 

We encountered feeding flocks in Regions I through IV 
(Table 5). viz.. in the Pacific Ocean from 150°W to 158°E 
longitude. We saw no flocks from Pohnpei westward to 
Singapore. 

We found only one flock, composed of 85 sooty terns, in 
the Line Islands. More than half the flocking birds were within 
the Phoenix Islands (Table 5). where the highest seabird 
populations and greatest species diversity were found. Great 
frigatebirds joined flocks only in the Phoenix Islands. 

In Region III (Gilbert Islands), sooty terns, brown noddies, 
and white terns were the only species present in feeding flocks. 
Sooty terns and brown noddies accounted for 92% of flocking 
birds in this area (Table 5). 

Region IV (Micronesia) was relatively rich in feeding 
flocks that contained several species not encountered elsewhere: 
stercorariids, Kermadec and Bulwer"s petrel. Audubon's 
shearwaters, and white-tailed tropicbirds. 

The feeding flocks we observed were found in areas 
known to be nutrient-rich. In broad terms, latitudes north of 
10°N and south of 10°S are poor in nutrients. The zone in- 
between, especially from 04 or 05°N to 09 or 1 0°N ( Equatorial 
Countercurrent) and from 00 to 09 or 10°S. is considerably 
richer in plankton and schools of small fish and tuna (Ashmole 
& Ashmole. 1967). Superimposed on this general pattern are 
local upwellings of plankton or "fronts" that occur close to 
islands and arc particularly evident between 1 °30'S and 05°N 
(King & Hida. 1957). 

The geographic distribution of our feeding flocks fits these 
general patterns (Fig. 8). Their latitudinal limits were07°N and 
07°S. At 02°S. within the plankton bloom on either side of the 
equator, we encountered a feeding flock of 153 birds. Rocks 
were more frequent around island clusters. We saw four flocks 
each close to the Phoenix and Gilbert Groups, and the day our 
ship arrived at 05°N (13 October, south of Pohnpei! we 
immediately observed a substantial increase in flying fish. 
From 04°N to 07°N, we encountered 3 feeding flocks of 223. 
22. and 408 birds, respectively. The dearth of fish. and. hence, 
seabirds. north of 10°N was particularly evident. Not only 
were there no feeding flocks, there was a substantial decrease 
m (he number of birds compared to all other Pacific areas 
(Table 1 ). 

We saw no flocks in southeast Asia, where total numbers 
of birds were low. Here we-encountered complicating factors 
due to heavy pressures from commercial and subsistence 
fisheries and other human population factors, which override 
changes in oceanography that are associated with the continental 



244 



shelf. Southeast Asia has long been recognized as an outstanding 
area for fish and. until the last few decades, was similarly rich 
in seabird colonies (Nelson. 1978). 

Regional Discussions 

Region I (457 birds. 26 species. 7 families) 

Region 1 (Fig. 1 ) begins with the waters between the Line 
Islands and Hawaii north to 14°N. The Line Islands are a 
scattered group of five atolls, five islands, and two submerged 
reefs straddling the equator between 06°N and 12°S latitude, 
and 162° and 150°W longitude. They are all low islands with 
extremely varied ecology ranging from barren, tropical deserts 
with scarcely any vascular plants to lush forests of coconut 
palms and/or indigenous vegetation. 

All except three are uninhabited and thereby serve as 
suitable habitats for tropical seabirds. Together with the 
Phoenix Islands, they constitute the largest assemblages of 
breeding tropical seabirds in the Pacific, both in species diversity 
and abundance. Christmas Island, with 18 species of breeding 
seabirds. is one of the richest seabird islands in the world 
(Ashmole & Ashmole, 1967; Garnett, 1983), due in part to 
equatorial upw ellings and plankton associated with the seasonal 
movements of the North and South Equatorial Currents and 
Countercurrents (King & Hida, 1957; Ashmole & Ashmole, 
1967). 

Overall, we found greater seabird diversity and density in 
the Line and Phoenix Islands (Tables 1,2,4). Two major groups 
of birds were most abundant in Region I: shearwaters/petrels 
and terns. Of minor importance numerically were boobies and 
storm-petrels (Table 3). Of interest is the high proportion of 
resident breeders and wintering birds (Fig. 3). 



The 1982-1983 anomalous warm waters associated with 
the El Nino Southern Oscillation severely disrupted seabird 
breeding on Christmas Island (Schreiber & Schreiber, 1984). 
Furthermore, feral cats, developmental threats, and increased 
poaching associated with expanding human populations arc 
ongoing problems on Christmas, and these affect the numbers 
of birds seen at-sea in the area (Gilbert and Ellice Islands Gov., 
1974; Garnett, 1983; Teeb*aki, personal communication). 

In the Line Islands, we extended the known range of herald 
petrel and added sightings of Cook's petrel (six) and Stejneger's 
petrel (two), both rarely recorded in the area. 

Region 11 ( 1 ,796 birds. 24 species, 6 families) 

The Phoenix Islands (Fig. 1) form a relatively compact 
group of eight low islands lying from 03 to 05°S. Most are dry 
and waterless. All except Canton are uninhabited and harbor, 
like the Line Islands, some of the richest and largest seabird 
colonies in the world. They lie within the boundaries of the 
South Equatorial Current (04°N to ca. 10°S), a region rich in 
plankton and associated fish schools (King & Hida, 1957; 
Ashmole & Ashmole, 1967). 

At-sea bird observations were dominated by sooty terns 
(96Vr of total), with procellariids next in abundance (Table 3). 
Small numbers of frigatebirds. boobies, and storm-petrels 
occurred, as in Region I (Table 2). Resident breeders, 
nonbreeding visitors and direct migrants were all well- 
represented (Fig. 4). The highest density of any bird on this 
cruise was in this area — sooty terns, at 32.45/10 km 2 . 

In the Phoenix Islands we recorded a range extension of 
the little shearwater, and added three sight records of Cook's 
petrel, and nine of wedge-tailed shearwaters (Table 6). Because 
of the remoteness and unsuitability of the Phoenix Islands for 



Northern 

Mariana 

Islands 



Marshall Islands 






Papua New Guinea 



|»»»| ranrse wm ■uumiumb 

fl — ^ POLrTICM tiiu.c.".. 

[~T"1 CHUUi TB»C«. ««*P<I»« KOWOltv 




J,.s 



Fig. 8. Geographic distribution of feeding flocks (dots). 



245 



human settlement (Howland and Baker are US National Wildlife 

Refuges, and Birnie, McKean. and Phoenix Islands are Kiribati 
Wildlife Refuges), its seabirds appear reasonably safe from 
disturbance for the present. 

Region III (495 birds. 13 species, 5 families) 

The Gilbert Islands (Fig. 1 ) stretch in a compact arc from 
03°17'N to 02°38'S latitude and from 176°49'E to 172°48'E 
longitude. They compose a single archipelago of 1 1 atolls and 
5 reef islands, forming a southerly extension of the Marshall 
Islands. 

All the Gilbert Islands are populated, some heavily. Since 
the I-Kiribati have long utilized seabirds and their eggs for 
food, species sensitive to human disturbance (shearwaters, 
petrels, boobies, tropicbirds) are absent. The primary breeding 
species are tree-nesters such as brown noddies and white terns 
(Table 4). 

Our observations were dominated by larids (Table 3). 
Brown noddies (48% ) and sooty terns (219c) composed 929c of 
(locking birds. With the exception of a few boobies, most other 
birds were nonbreeding visitors (Table 4). Migrants were 
absent (Fig. 3). We also provide additional sightings of such 
rarities in the region as Cook's petrel ( 1 ). Kermadec petrel (2), 
south polar skua (5), and wedge-tailed shearwater (213) 
(Table 6). 

Region IV (799 birds, 18 species, 10 families) 

Our cruise track passed south of the districts of Kosrae and 
Pohnpei, then north of Truk, Yap, and Belau (04°N to 10°N 
latitude, 163°E to 138°E longitude). Micronesia contains 
fewer people on far more islands than the Gilberts. However, 
fishing fleets from several foreign countries are exploiting their 
oceanic waters, and the presence of even a few people on an 
islet deters many seabird species from successful breeding. 

Ourobservations from the Caroline Islands were dominated 
by sooty terns and noddies ( 52% of birds seen), but procellariids 
and boobies were represented in fair numbers. Skuas first 
appeared here, along with the only migrant duck. The wedge- 
tailed shearwater attained a density of 2.93 birds/10 km- 
( Tabic 2). Our observations of brown boobies suggest the 
possible existence of a colony on Magur Islet, Namonuito 
Atoll, in the District of Truk. 

Region V ( 34 birds. 6 species. 4 families) 

The Philippine Sea and Basin (Fig. 2) exhibited the lowest 
biodiversity and species densities (Tables 1-4) of the five 
Pacific regions. This relatively small area, little-known 
ornithological lv and with no islands for hundreds of kilometers. 
stretches west of the Marianas to the Philippines (at 11°N 
latitude, from 136°E to 125°E longitude). The nearest landfalls 
arc the Marianas to the northwest. Belau, New Guinea, and the 
Moluccas (Indonesia) to the south, and the Philippines to the 
w est. Plankton productivity and fish populations arc known to 
be considerably poorer in tropical waters north of 09°N than 
further south (Ashmole & Ashmole, 1967). Our sightings of 
six streaked shearwaters suggest that the southerly migration 
corridor for this species lies entirely cast of the Philippines. 



Region \ I (22 birds. 4 species. 6 families) 

The Bohol (Mindanao) and Sulu Seas stretch from 125°E 
to 1 17°E. Seabird densities were low. Phalaropes. not previously 
encountered on our cruise, accounted for over half the total bird 
count (Table 3). A few larids, skuas, brown noddies, and 
wedge-tailed shearwaters (our most western observations) 
completed the list. The proportions of resident breeders, 
nonbreeding visitors, and direct migrants were equal (Fig. 4). 
Elevated human populations in the Philippines have undoubtedly 
reduced seabird numbers. 

Region VII (65 birds, 8 species. 5 families) 

Our cruise track passed through the South China Sea 
between 07°N to 01°N latitude and 117°Eto 104°E longitude. 
The route passed from the Balabac Strait (southwestern 
Philippine Islands) indirectly to Singapore (Fig. 2). These 
waters, the most heavily polluted of the trip ( see Chapter 2. this 
vol.), are heavily fished commercially, and human population 
densities around their periphery are high. Seabirds were 
sparse. As with Region VI. phalaropes contributed half of the 
total. Migrant stercorariids formed the next most common bird 
grouping; since stercorariid records are few from this area, our 
observations of 1 2 birds of at least 2 species add to their known 
dispersal areas. 

Gulls and terns were in surprisingly few numbers, 
considering that almost 20 species occur in the South 
China Sea. We recorded Caspian, bridled, and crested 
terns (and a few unidentified larids) in very small numbers. 
Resident breeders and nonbreeding visitors numbered only 
one species each, and there were four species of migrants 
(Fig. 3). 

During our indirect. 9-day passage across the South China 
Sea (23-31 October) we encountered approximately 150 land 
birds (including 40 raptors), totaling at least 20 species (Ellis 
et ah, 1990 and Subchapter 3.6, this vol.). The presence of 
owls, nightjars, falcons, a large crake, and small forest birds 
provided a highly interesting replacement for the expected 
terns, gulls, and other seabirds. 

This area is little known ornithologically, at least in 
English publications (Delacour&Mayr. 1946; Delacour. 1947; 
Anon.. 1975; Jing-Xiam & Zi-Yu, 1975; Nelson. 1978; Hails. 
1987; King & Dickinson. 1987). We added three parasitic 
jaegers to records for the South China Sea and a possible range 
extension of the masked booby. Unfortunately, in the last few 
decades numerous large seabird colonies in this area have been 
destroyed by direct human predation (Nelson, 1978) but 
considerable efforts towards conservation of islands, reefs, 
bays, and varied habitats are on-going commitments by all 
countries concerned, assisted by international agencies (UNEP, 
I984a,b; IUCN. 1988a,b). 

This First US-USSR Central Pacific Expedition resulted from 
the efforts of main people in the United States and the Soviet Union. 
On the American side, the primary organization and financial support 
wore from the US Fish & Wildlife Service. We especially appreciate 
the efforts of Harold J. O'Connor, Director, Patuxent Wildlife Research 
Center and Steve Kohl (Office of International Affairs). We thank 
H. Randolph Perry for suggesting and encouraging our participation 
in the expedition, Paul Sykes for willingly assuming the Michigan 



246 



responsibilities of CBK during the long voyage, James Hines, Brett 
Hoover, and Lois Loges for computer programming assistance, and 
Bonnie J. Fancher for help in manuscript preparation. The paper 
benefitted from the comments of D. Ainley. W. King, and P. Gould. 

AKK thanks the other members of the ICBP 1990 Line and 
Phoenix Islands Expedition (M. and A. Garnett, G. Wragg. J. Phillips, 
and M. Linsley ) for use of our collective at-sea data. She also extends 
special thanks to Mr. O'Connor for financial support during preparation 
of the manuscript. 

On the Soviet side, it is a great pleasure to thank Professor 
Alia V. Tsyban, Goskomgidromet, who as leader of the expedition 



provided help, interest, friendship, and outstanding hospitality during 
the voyage. Captain Oleg A. Rostovtsev and his crew (especially the 
navigators) of the R/V Akademik Korolev provided an excellent 
observation platform and abided our frequent imposition for position 
fixes from their satellite and Loran C navigation systems and 
hydrographic charts. Yevgeniy N. Nelepov and Yuri L. Volodkovich 
provided much assistance during the voyage, and Boris Sirenko and 
Boris Alexandrov willingly shared their knowledge of benthic and 
pelagic organisms. Our contacts with our Soviet colleagues would 
have been far less stimulating without the translation skills of 
Valeriya M. Vronskaya and Svetlana V. Petrovskaya. 



3.6 Evidence for a Major Fall Land Bird 

Migration Corridor Across the South China 
Sea from Indo-China to the Greater Sunda 
Islands 

DAVID H. ELLIS. ANGELA K. KEPLER, and CAMERON B. KEPLER 

US Fish & Wildlife Senice, Patu.xent Wildlife Research Center, Laurel, Maryland, USA 



Introduction 

Until 1960, bird migration corridors in eastern Asia were 
poorly known (Wetmore, 1926; Delacour, 1947; McClure, 
1974; Medway & Wells, 1976). In southeast Asia, however, 
the geography of the land masses surrounding the South China 
Sea seems to create natural funnels that should concentrate 
migrant land birds into three primary fall corridors. Important 
flight paths along some of these routes have recently been 
discovered. 

It is known that migrants from Japan and eastern China 
island-hop south through the Philippines (Wetmore. 1926; 
McClure, 1974), with Ng ( 1978) presenting evidence that barn 
swallows (Hirundo rustica) move directly from mainland 
China to the Philippines. McClure ( 1974) asserted that many 
migrants passing through the Philippines to Borneo fly west 
from Palawan, then south to Borneo. Simpson (1983a,b) 
encountered hundreds of migrant birds at the Tembungo offshore 
oil drilling platform near the northeastern tip of Borneo (Fig. 1 ) 
during the fall migration of 1981. Although he reported these 
observations as evidence of a passage directly across the South 
China Sea, his location near Balabac Strait also suggests that 
these migrants could have been moving south from the 
Philippines. 

Geography suggests that many migrant land birds in 
Burma and western Thailand would move south along the 
Malay Peninsula, a pathway known to be important ( McClure, 
1974; Medway & Wells, 1976; Hails, 1987), and thence across 
the narrow Straits of Malacca to Sumatra. However, migrants 
from east Thailand as well as those from China, Laos, and 



Vietnam, moving down the Indo-China Peninsula would 
naturally converge south of the Mekong River Delta on 
Mui Bai Bung. From this tip of the Indo-China Peninsula, 
birds traveling overland must fly northwest into Thailand 
before proceeding south. Those capable of a relatively short 
(ca. 400 km) overwater flight can fly southwest across the Gulf 
of Thailand toward the Malay Peninsula, a route portrayed by 
McClure (1974) and Hails (1987) as a minor pathway for the 
migrants from Indo-China. 

McClure (1968, in McClure, 1974) illustrated a coastal 
migration route from Taiwan to northern Vietnam, thence 
south, crossing the South China Sea to Borneo, another route 
suggested by geography. However, he provides scant evidence 
for such a corridor and no evidence that migrants are 
concentrated at Mui Bai Bung. McClure ( 1 974) discussed a fall 
passage of willow warblers (Phylloscopus sp.) across the South 
China Sea to Sarawak without offering details on their point of 
origin north of the sea. Simpson (Wells, personal 
communication and in prep.) reported a substantial fall 
movement of land birds ( 36 species ) in the Terengganu oil field 
(ca. 05°25'N, 105°13'E, see Fig. 1 ). Although this location is 
only about 200 km east of the Malay Peninsula and west of a 
direct route to Borneo, Simpson's records provide the best 
evidence to date of a direct South China Sea crossing. The birds 
observed by Simpson (1983a) at the Tembungo oil terminal 
could have come from Vietnam, as he suggests, but the source 
of these migrants is clouded by their proximity to the Philippines. 
Although biologists from the Chinese Academy of Science and 
the Beijing Natural History Museum (Anon., 1974) noted 
44 species of land birds during 1974 surveys of islets in the 



247 



South China Sea 




Fig.l. Geography of the southern half of the South China Sea showing bird 
survey locations. Numbered segments are bird location survey 
locations for R/V Akademik Korolev, 23-310ctober 1988. 

northern two-thirds of the South China Sea, demonstrating the 
potential for long-distance (ca. 1,000 km) migration, they did 
not demonstrate that a corridor for land-bird migrants exists 
further south between Vietnam and Borneo. In this paper, we 
present data from the South China Sea that strongly support the 
presence of such a migration route. 

Study Area and Methods 

We encountered migrant land birds during our 
23-31 October 1988 indirect passage (Fig. 1) from Balabac 
Strait to Singapore on the Soviet research vessel Akademik 
Korolev. While in transit, we observed birds during dawn-to- 
dusk seabird surveys from the flying bridge (12 m above sea 
le\ el ). During a 3-day period while the ship was anchored or 
drifting without power to conduct oceanographic research 
(Fig. 1, Station 13;06°01'N, 106°55'E), we conducted periodic 
v. alking inspections of the ship (usually at half-hour intervals) 
and searched the ship each night by flashlight to count roosting 
birds. Five raptors anil several barn swallows were captured by 
hand (primarily at night) and examined for physical condition. 

Results and Discussion 

During our 9-day passage, we encountered about 1 50 land 
buds 1 121 by conservative count, 84 minimum count, Table 1 1 
representing 14 families. Almost all were migrants that winter 
(at least in part) south of the South China Sea. Most of these 
birds (9ft by conservative count ) arrived on the ship during the 
! da) period while we were stationary (Fig. 1, Station 13) 
around 350 km southeast of the southern tip of [ndo-China. 
The presence oi land buds at (Ins location suggests that they 
were in passage across the South China Sea from Indo-Chin.i 
to the ( ireater Sunda Islands. The low bird counts seen before 



arriving at and after leaving this location (Table 2) suggest that 
this spot lies on a rather narrow migratory pathway although, 
alternately, birds may have been reluctant to approach a moving 
vessel. Simpson's 1982 observations (Wells, in prep.), made 
in the Terengganu oil field (Fig. 1 ) very near our cruise track, 
suggested that he was sampling the same corridor we visited; 
if so, the pathway may be somewhat wider than we detected. 

The number of birds we observed (Tables 1,2) is small 
when compared with record counts for well-known migration 
pathways. However, our visit was brief and probably too late 
for detecting the bulk of migrating land birds. Simpson's 
( 1983a) dates for six of nine frequently encountered land birds 
near northeastern Borneo fell before the time of our visit, and. 
just as important, migrating land birds most often aggregate 
where land and water configurations encourage them to collect 
(e.g.. on north or south projecting peninsulas). By contrast, we 
were on the open sea where birds are much less likely to 
concentrate. Considering these factors, it seems likely that 
adequate spatial and temporal sampling will reveal many 
thousands of land birds moving south from Indo-China across 
the South China Sea. 

Although our records and those of Simpson's (Wells, in 
prep.) demonstrate that a sizable migration is probably normal 
across the South China Sea. we should mention an alternate 
hypothesis that may help explain the presence of these birds 
where and when we observed them. First, our passage occurred 
when Typhoon Ruby was ravaging the Philippine Islands 
(Anon.. 1989). Although we did not encounter heavy seas or 
strong winds, some of the birds we observed may have been 
forced out to sea, if nonmigratory. or shunted away from their 
normal migration route, if migratory, by the storm. However, 
most of the birds we observed far from land (Table 3) are 
known to be strong migrants. The four hawks tentatively 
identified as shikras (Accipiter badius) and crested goshaw ks 
(A. trivirgatus), and the dove (Streptopelia sp. ) are the only real 
surprises, although a few others in Table 3 would not be 
expected this far from land. 

Flight direction may give some indication of the likelihood 
of either hypothesis. If the birds were displaced migrants, they 
would probably have been heading southwest (i.e.. away from 
the storm). If in passage from a concentration zone on the Indo- 
China Peninsula to Borneo, they should have been heading 
southeast to encounter our vessel. If. as we observed, the 
raptors (33'f of all land birds) were foraging at sea (Ellis etal., 
1 990 ) rather than migrating, there would likely be no consistent 
trend in their flight direction. In Fig. 2. there is no clear east- 
west trend in arriving or departing flights. However, although 
the data are very few strong southward and westward 
components are e\ idem. In constructing Fig. 2. we eliminated 
directional readings for birds seen on cruise track segments 
1-7 and 20-30 because these segments were near enough to 
land (i.e., within 1 00 km ) that the birds' flight directions could 
have been influenced by sight or sign of nearby land. In 
addition, all Right bearings could have been influenced h\ the 
presence of the ship. 

Physical condition of the birds w e obsen ed mas also be an 
indicator of the regularity with which this migration route is 
used. If a high proportion of the know n o\ erseas migrants were 



248 



TABLE 1 

Land bird totals for R/V Akademik Korolev cruise track segments and stationary watches in the South 

China Sea, October 1988. 

No. Land Birds Observed' 



Date 



23 



24 



25 



Station/Segment 2 


Duration 










All Land 




Length 


ofObs. 


Rapti 


3rs 


Non- 


•raptors 




Birds 


Number 


(km) 


(min.) 


Min. 


Cons. 


Min. 


Cons. 


Min. 


Cons. 


1 


38 


88 




















2 


15 


45 




















3 


6 


17 




















4 


67 


158 




















5 


14 


44 




















6 


19 


48 




















7 


49 


109 








1 


1 


1 


1 


8 


22 


50 




















9 


3 


11 




















10 


30 


67 




















11 


17 


38 




















12 


63 


143 








4 


4 


4 


4 



26-28 13 ca. 912 22 30 32 44 54 74 



28 


14 


ca. 


43 


1 


1 








1 


1 




15 


22 


10 






















16 


26 


78 






















17 


ca. 


195 


1 


4 


4 


8 


5 


12 


29 


18 


ca. 


105 


1 


1 


1 


4 


2 


5 




19 


ca. 


20 






















20 


25 


62 











1 





1 




21 


ca. 


80 




















30 


22 


ca. 


23 


1 


1 


3 


3 


4 


4 




23 


ca. 


IK) 








2 


4 


2 


4 




24 


ca. 


25 











1 





1 




25 


ca. 


95 





1 


3 


3 


3 


4 




26 


ca. 


54 






















27 


ca. 


52 








3 


5 


3 


5 




28 


ca. 


30 


1 


1 








1 


1 


31 


28 


ca. 


15 






















29 


ca. 


15 






















30 


135 


280 


1 


1 


3 


3 


4 


4 


TOTALS 






3.022 


28 


40 


56 


81 


84 


121 



1 Because accurate bird counts were sometimes difficult to obtain for stationary watches (i.e.. some 
birds remained aboard or flew about the ship for extended periods), we report both the minimum 
(min.) number of birds observed (based on subtractive values) and a conservative (cons.) number 
based primarily on new arrivals. The actual number observed is believed to be about 20% higher than 
the conservative count. 

: Cruise track segment locations are illustrated in Fig. 1. 



249 



TABLE 2 

Minimum and conservative land bird counts along cruise track of R/V Akademik Korolev in the 
South China Sea, 23-31 October 1988. ' : 



Species 

Small juv. accipiter (Accipiter sp. ) 

Ad. Japanese sparrow-hawk (Accipiter gularis) 

Ad. shikra (A. badius) 3 

Ad. crested goshawk (A. trivirgatusf 

Eagle/kite (Accipitridae) 

Peregrine falcon (Falco peregrinus) 

Oriental scops owl (Otus siiniaf 

Chinese pond heron (Ardeola bacchus) 

Watercock {Gallicrex cinerea) 

Dove [Streptopelia sp.)' 

Grey nightjar (Caprimulgus indicus) 

Fork-tailed swift (Apus pacificus) 

Swift (Apodidae) 

Dollarbird (Eurystomus orientalis) 

Barn swallow (Hirundo rustica) 

Swallow (Hirundo sp. ) 

Ashy minivet (Pericrocotus divaricatusf 

Lanceolated Warbler (Locustella lanceolata) 4 

Great reed warbler (Acrocephalus arundinaceus i 

Warbler (Acrocephalus sp.) 

Arctic warbler (Phylloscopus boreal is) 

Flycatcher (Ficedula sp.) 

Brown shrike (Lanius cristatus) 

Unidentified passerines or remains 

TOTALS 



Segments 1-7 


Segments 8-19 


Segments 20-30 


(Oct. 


23-24) 


(Oct. 


25-29) 


(Oct 


. 29-3 1 ) 


.Jin. 


Cons. 


Min. 


Cons. 


Min. 


Cons. 








14 


24 


3 


3 








2 


2 














1 


-> 














2 


2 














1 


1 














3 


3 








1) 





2 


2 





1 


1) 











3 


3 








1 


1 














1 


1 














1 


1 


1 


1 








4 


4 














2 


2 








1 


1 




















14 


29 


5 


10 








1 


1 














1 


1 


1 


1 








1 


1 


1 


1 








1 


1 














1 


2 


1 


1 














1 


1 








2 


2 














6 


6 


1 


1 








5 


8 





1 


1 


1 


66 


96 


17 


24 



'Cruise track segments and stations are illustrated in Fig. 1 and described in Table I . 

'Abbreviations in column headings are: Min. (minimum count) and Cons, (conservative count) as explained in Table I. 
Footnote 1 . 

"Because these birds are considered non-migratory, these identifications should be treated as tentative. All are based on 
nearby visual observations aided by lOx binoculars, but without photographic or other substantiation. 

individuals of these species were deposited in the US National Museum: Oriental scops owl. USNM No. 607190; Ashj 
minivet. USNM No. 607193; and Lanceolated warbler, spirit specimen (not assigned numbers at USNM l. 



in good body condition this far from land, it is more tenable to 
suppose that these species regularly use this route. In Table 3, 
our best estimate of physical condition is compared for all 
species that we encountered far from land. We know from 
handling a few captives, and infer from the energetic flight of 
others, that the raptors and the barn swallows at Station 13 
(Fig. I ) were in good physical condition. For the other species, 
too few individuals were present to draw firm conclusions, but 
all birds of most species appeared to be in good condition. 

A final hypothesis may explain the presence of some of the 
raptors. Many were opportunistically foraging at sea. During 
our 3-day layover dig. I, Station 13), we recorded raptors 
perching for extended periods, roosting nightly on the ship, and 



engaging in at least 21 hunting forays (Ellis et ai. 1990). Of 
14 forays for which the outcome was known, 13 (93%) were 
successful. Some accipiters even used the ship's deck lights to 
forage at night. We gathered prey remains, totaling at least 
20 kills. Two species, barn swallow and brown shrike (Lcmius 
cristatus), suffered heavy mortality from predation. Of 14 barn 
swallows (minimum count) observed from 25-29 October, at 
least 7 turned up as prey. Even more significant, five of six 
(minimum count) brown shrikes seen during the same 5-day 
period were observed as prey. Simpson's (1983a; Wells, in 
prep. ) observations of raptor behavior at both oil fields led him 
to conclude that Japanese sparrow-hawks were hunting and 
"commuting between nearby rigs." Our observations confirm 



250 



TABLE 3 

Physical condition and migratory status of land birds arriving on R/V Akademik Korolev, 25-29 October 1988, South China Sea. 1 



Physical 
Condition" 



Mobility Classes 2 



Taxon [Number] 



Known Known Known 

Known Over-water Colonizer Straggler 
Migrant Migrant 5 of Islands to Islands Comments 



Good 



Good 



Japanese sparrow-hawk + 

(Accipiter gularis) [2] 

Shikra (A. badius) [2] + 



Common migrant 



Western population is highly migratory; 
eastern population migratory in Malaysia 



Good Crested goshawk 

{A. trivirgatus) [2] 

Good Peregrine falcon 

{Falco peregrinus) [3] 



Non-migrant throughout range 



Fatigued 


Watercock 




(Gallicrex cinerea) [ 1 ] 


Good 


Dove (Streptopelia sp.) [ 1 


Emaciated 


Oriental scops owl 




(Otus sunia) [1] 


Good 


Gray nightjar 




(Caprimulgus indicus) [ 1 



Winters in Greater Sunda Islands and Celebes; 
very few records as straggler 



Strongly migratory, scatters across Malaysia in 
winter 



Good Fork-tailed swift 

(Apus pacificus) [4] 

Good Barn swallow 

(Hirundo rustica) [29] 

Fatigued Ashy minivet (Pericrocotus 

divaricatus) [1] 

Fatigued Lanceolated warbler 

{Locustella lanceolata) [1 ] 



A few migratory stragglers recorded as far east 
as Marshall Islands 

Winters throughout region and tropics 
worldwide 

Winters on larger islands of Indonesia, but not 
on islands separated by large bodies of water 

Winters in Greater Sunda Islands 



Good Great reed warbler (Ac ro- 

cephalus arundinaceus) [ 1 ] 

Good Brown shrike 

[Lanius cristatus) [6] 



Common migrant in Indonesia 



Common migrant to Greater Sunda Islands; 
recorded in Palau 



! Data are included only for that portion of the cruise track (stations 8-19) where the ship was far (>100 km) from land. 

"Symbols in these columns: + = yes.- = no. Assignment to mobility class (i.e., regular migrant over land and over large bodies of water 
>500 km. colonizer of distant land masses and islands as a breeding bird, straggler either on migration or as a resident) is at best tentative 
for some species, but was derived from information in Brown & Amadon (1968), Clements (1978), King & Dickinson (1975), Medway & 
Wells (1976), and Pratt et al. (1987). 

'Physical condition was reported "Good" if bird flew well and was adept at avoiding capture by hand, "Fatigued" if readily captured by 
hand, and "Emaciated" if sternum was sharply protruding upon capture. 



251 



N 

i 




A. ACCIPITERS IN =17) 




B PASSERINES and SWIFTS i N = 7 I 



o 



Fig. 



C. BARN SWALLOWS IN=4I 

Arriving and departing flight directions for birds seen on cruise track 
segments 8-19 of R/V Akademik Korolev, 26-29 Oclober 1988. 



that the raptors were opportunistically using our stationary ship 
for perching, roosting, hunting, and eating. When the ship was 
moving under power, however, none of the raptors perched for 
any extended period, and none roosted on the ship. 

From all available evidence, it seems most likely that the 
birds we encountered were a small part of what must be a 
si/able wave of fall migrants on their way across the South 
China Sea. The configuration of the land masses suggests that 
the point of departure for these birds was the southern tip of 
Indo-China; however, further land-based research is needed to 
substantiate the point of origin and destination of birds crossing 
the South China Sea to Borneo. Additional work at sea will also 
be helpful in determining the timing and magnitude of the 
migration, as well as corridor w idth. Work on islands in the 
South China Sea or stationary platforms may substitute in part 
for the at-sea studies, but it is also important to determine body 
condition of birds arriving in the Greater Sunda Islands. 
Intensive banding operations in Vietnam, at sea. and in Borneo 
could reveal much about survival rates and all other aspects of 
this little-known migration route. 

This project was part of the First US-USSR Joint Pacific 
Expedition to the Pacific Ocean and South China Sea. We express 
appreciation to the agents of both governments who made our 
participation possible. We especially thank Professor Alia V. Tsyban 
(Chief Scientist). Harold J. O'Connor (Director. Patuxent Wildlife 
Research Center and US organizer of the expedition ). Captain Oleg A. 
Rostovtsev, and our many Soviet friends whoenhancedourenjo) ment 
of the trip. Joe Marshall and Ralph Browning assisted in specimen 
identification at the Smithsonian Institution. Linda J. Miller. Cath) 
Ellis, and Bonnie Fancher assisted greatly in data handling and 
manuscript preparation. 

The following people commented on an earh version of this 
manuscript and thereby compensated for our inexperience w ith the 
birds of the South China Sea region: H. Elliott McClure. Duncan 
Parish. Philip D. Round. Lucia Liu Severinghaus. and David R. Wells. 
We appreciate additional editorial comments from Mark R. Fuller. 
George F. Gee. and Gary H. Heinz. 



252 



Chapter 3 References 



Ainley, D. G. & Boekelheide, R. J. ( 1983). An Ecological 

Comparison of Oceanic Seabird Communities of the Souith 

Pacific Ocean. Stud. Avian Biol. 8. 2-23. 
Amerson. A. B., Jr. ( 1969). Ornithology of the Marshall and 

Gilbert Islands. Atoll Res. Bull. No. 127. 1-348. 
Anonymous ( 1974). Report of the zoological survey of the 

islands of the South China Sea. ACTA Zoologica Sin. 20(2), 

113-130. 
Anonymous (1989). Marine weather review. Mariners 

Weather Log 33(2), 40-50. 
AOU. (1983). Check-list of North American Birds. 6th Ed. 

American Ornithologists' Union, Washington, D.C. 
AOU. (1985). Thirty-fifth supplement to the American 

Ornithologists' Union Check-list of North American Birds. 

Auk 102. 680-686. 
Arashkevich, E. G. (1972). Vertical distribution of various 

trophic groups of copepods in boreal and tropical regions of 

the Pacific Ocean. Oceanology 12(2). 315. (in Russian) 
Ashmole, N. P. & Ashmole, M. J. (1967). Comparative 

feeding ecology of seabirds of a tropical oceanic island. 

PeabodyMus. Nat. Hist. Yale Univ. Bull. 24. 1-131. 
Bailey, S. F., Pyle, P. & Spear, L. B. (1989). Dark Pterodroma 

petrels in the North Pacific: Identification, status, and North 

American occurrence. American Birds 43. 400-415. 
Baker, R. H. (1951). The avifauna of Micronesia: Its origin, 

evolution, and distribution. Univ. Kans. Publ. Mus. Nat. 

Hist. 3(1). 1-359. 
Beklemishev. K. B. (1973). Ranges of tropical far-neritic 

species. In Tropical Zone of the World Ocean and the 

Global Processes Associated with It, pp. 292-299. Nauka 

P. jlishers. Moscow, (in Russian) 
Bloxham, A. ( 1925). Diary of Andrew Bloxham. naturalist of 

the "Blonde" on her trip from the Hawaiian Islands from 

England, 1824-1825. B. P. Bishop Mus. Spec. Publ. 10. 
Boaden. P. J. S. & Seed, R. (1985). An Introduction to Coastal 

Ecology. Blackie and Son Ltd., Glasgow and London. 218 

pp. 
Briggs. K. T„ Tyler, W. B.. Lewis, D. B. & Carlson, D. R. 

(1987). Bird communities at sea off California: 1975 to 

1983. Stud. Avian Biol. 11. 74 pp. 
Brown, L. & Amadon, D. ( 1968). Eagles, Hawks and Falcons 

of the World. Country Life Books, London. 
Chislenko. L. L. (1968). Nomograms for Determining the 

Weight of Aquatic Organisms from Body Size and Shape. 

Nauka Publishers. Leningrad, (in Russian) 
Clapp. R. B.( 1967). Line Islands. Unpublished report. 32 pp. 

USNM, Smithsonian Institution, Washington. D.C. 
Clapp. R. B. & Sibley. F. C. ( 1 967 ). New records of birds from 

the Phoenix and Line Islands. Ibis 109. 122-125. 
Clapp. R. B. & Sibley. F. C. (1968). Additional new records 

of birds from the Phoenix and Line Islands. Ibis 110. 

573-575. 
Clapp. R. B. & Sibley. F. C. (1971a). Notes on the vascular 

flora and terrestrial vertebrates of Caroline Atoll. Southern 

Line Islands. Atoll Res. Bull. 145. 1-16. 



Clapp. R. B.& Sibley. F. C. (1971b). The vascular flora and 

terrestrial vertebrates of Vostok Island, south-central Pacific. 

Atoll Res. Bull. 144, 1-9. 
Clements, J.F.I 1978). Birds of the world: A check list. Two 

Continents Publishing Group, New York. 
Crossin, R. S. (1974). The storm-petrels (Hydrobatidae). In 

Pelagic Studies of Seabirds in the Central and Eastern 

Pacific Ocean (W. B. King, ed. ), pp. 154-205. 
Day, R. A., Wehle, D. H. S. & Coleman, F. C. ( 1984). Ingestion 

of plastic pollutants by marine birds. In Proceedings of 

the Workshop of the Fate and Impact of Marine Debris, 

26-29 November 1984, Honolulu, Hawaii, pp. 344-386. 

NOAA Technical Memorandum SWEC-54, NMFS, 

NOAA-TA, US Department of Commerce. 
Delacour, J. (1947). Birds of Malaysia, xvi + 382 pp. 

Macmillan Co., New York. 
Delacour, J. & Mayr, E. (1946). Birds of the Philippines, 

xv + 309 pp. Macmillan Company, New York. 
Ellis, D. H., Kepler, A. K. & Kepler. C. B. (1990). Evidence 

for a fall raptor migration pathway across the South China 

Sea. J. Raptor Res. 24(1-2), 12-18. 
Engbring, J. (1983). Avifauna of the southwest islands of 

Palau. Atoll Res. Bull. 267, 1-22. 
Garnett, M. C. (1983). A Management Plan for Nature 

Consen'ation in the Line and Phoenix Islands. Pt. I: 

Description, 318 pp.; Pts. II, III: Policy and Recom- 
mendations. 131 pp. 
Garnett, M. C. (1984). Conservation of seabirds in the South 

Pacific region: A review. ICBP Tech. Pub. 2, 547-557. 
Geinrikh, A. K. (1960). The near-surface plankton of the 

central World Ocean. Oceanology 11. (in Russian) 
Geinrikh, A. K. (1964). On the surface plankton of the 

northeastern Pacific. Trudy Institute of Oceanology AN 

SSSR 65, 77-94. (in Russian) 
Geinrikh. A. K. (1968). Quantitative distribution of certain 

plankton animals in the western Pacific Ocean. In Plankton 

of the Pacific Ocean. Gidrometeoizdat Publishers, Moscow. 

(in Russian) 
Gilbert and Ellice Islands Government. (1974). Line Islands 

Expedition report, 89 pp. Unpublished manuscript. 
Gill, R. E. (1990). Ecology of curlews in French Polynesia 

during spring. In Summary of the Proceedings from the 

Workshop on Bristle-thighed Curlews. Anchorage. Alaska 

(R. E. Gill & C. M. Handel, eds. ). 22 pp. 
Gorshkov. S. G. (1974). Atlas of the Oceans. The Pacific 

Ocean. Head Department of Navigation and Oceanography, 

Ministry of Defense of the USSR. 
Gould, P. J. (1974a). Introduction, pp. 1-5. In Pelagic 

Studies of Seabirds in the Central and Eastern Pacific 

Ocean (W. B. King, ed.), 277 pp. Smithson. Contrib. 

Zool. 158. 
Gould. P. J. (1974b). Sooty tern (Stemafuscata), pp. 6-52. In 

Pelagic Studies of Seabirds in the Central and Eastern 

Pacific Ocean ( W. B. King. ed. ). 277 pp. Smithson. Contrib. 

Zool. 158. 



253 



Gould, P. J. (1983). Seabirds between Alaska and Hawaii. 

Condor 85, 286-291. 
Gould, P. J. & Forsell, D. J. ( 1989). Techniques for shipboard 

surveys of marine birds. US Fish Wildl. Serv. Tech. Report 

25. 1-22. 
Gould, P. J. & King. W. B. ( 1967). Records of four species 

of Pterodroma from the central Pacific Ocean. Auk 84. 

591-594. 
Gould, P. J., King, W. B. & Sanger. G. A. (1974). Red-tailed 

tropicbird {Phaethon rubricauda), pp. 206-231. In 

Pelagic Studies of Seabirds in the Central and Eastern 

Pacific Ocean ( W. B. King. ed. ). 277 pp. Smithson. Contrib. 

Zool. 158. 
Grossman, H. & Grossman, H. ( 1974). Preliminary Report 

(Birds). In Line Islands Expedition Report, August-October 

1974, 5 pp. Gilbert and Ellice Islands Government. 

Unpublished manuscript. 
Haile, N. S. (1964). SabakSoc.J. 11. 135-137. 
Hails. C.( 1987). Birds of Singapore, 168 pp. Times Editions, 

Singapore. 
Haney. J. C. (1985). Counting seabirds at sea from ships: 

comments on interstudy comparisons and methodological 

standardization. Auk 102(4). 897-898. 
Harrison. P. (1985). Seabirds: An Identification Guide, 

448 pp. Houghton Mifflin Company. Boston. 
Herring, J. L. (1961). The genus Halobates (Hemiptera, 

Gerridae). Pac. Insects 3(2-3). 223-305. 
Huber. L. N. (1971). Notes on migration of Wilson's storm- 
petrel {Oceanites oceanicus) near Eniwetok Atoll, western 

Pacific Ocean. Notornis 18( 1 ). 38-42. 
Humphrey. P. S. (1965). An ecological survey of the Central 

Pacific. Smithson. Year 1%5. 24-30. 
Hunt, G. L., Eppley. Z., Burgeson. B. & Squibb. R. (1981). 

Reproductive ecology, foods and foraging areas of seabirds 

nesting on the Pribilof Islands, 1975-1979. In Final Reports 

of Principal Investigators, Vol. 12. pp. 1-258. NOAA. 

Boulder. Colorado. 
ICBP. ( 1990). Unpublished field notes. International Council 

for Bird Protection (ICBP) 1990 Line and Phoenix Islands 

Expedition. In author's possession. 
IUCN. (1988a). Coral Reefs of the World. Vol. 2. United 

Nations Environmental Program. Intemation Union for 

Conservation of Nature and Natural Resources. Compiled 

by C. Sheppard & S. M. Wells. 389 pp. 
IUCN. ( 1988b). Coral Reefs of the World. Vol. 3. United 

Nations Environmental Program. Intemation Union for 

Conservation of Nature and Natural Resources. Compiled 

by C. Sheppard & S. M. Wells. 329 pp. 
Jenkins, 1. A. F. ( 1979). Observations on the wedge-tailed 

shearwater {Puffinus pacificus) in the south-west Pacific. 

Notornis 26. 331-348. 
Jing-Xian, L. & Zi- Yu. W. ( 1 975 ). The red-footed booby in the 

\isha i Paracel I Islands. (Translated from Chinese bj [aian 

C. On'. ) Zool. Mag. (Dongwuxue Zazhi) 3. 40. 
Kepler. A. K. (1490). ICBP 1990 Line and Phoenix Islands 

1 \pedition. 20 February- 31 May 1990, 12pp. Unpublished 

report, Cambridge. England. 



Kepler. C. B., Kepler. A. K. & Ellis. D. H. ( 1992). Ecological 

studies on Caroline Atoll. Republic of Kiribati, south-central 

Pacific Ocean. Part II. Seabirds, other tenestrial animals. 

and conservation. (Subchapter 1. 2, this volume). 
King. B. F. & Dickinson. E. C. (1975). A Field Guide to the 

Birds of South-East Asia, 480 pp. Houghton Mifflin Co.. 

Boston. 
King. J. E. & Hida. T. S. ( 1957 ). Zooplankton abundance in the 

Central Pacific. Pt II. US Fish Wildl. Serv. Fish. Bull. 

57(118), 365-395. 
King. W. B. ( 1967). Seabirds of the Tropical Pacific Ocean. 

126 pp. Smithsonian Institution, Washington. D.C. 
King, W. B. (1970). The Trade Wind Zone Oceanography 

Pilot Study. Part VII: Observations of Sea Birds, March 

1964 to June 1965, 136 pp. US Fish Wildl. Sen: Spec. Sci. 

Rep. Fish. 586. 
King, W. B. ( 1973). Conservation status of birds of central 

Pacific islands. Wilson Bull. 85. 89-103. 
King. W. B. (ed.) (1974a). Pelagic Studies of Seabirds in the 

Central and Eastern Pacific Ocean. Smithson. Contrib. 

Zool. 158. 
King. W. B. (1974b). Wedge-tailed shearwater {Puffinus 

pacificus), pp. 53-95. In Pelagic Studies of Seabirds in the 

Central and Eastern Pacific Ocean, (W. B. King. ed). 

277 pp. Smithson. Contrib. Zool. 158. 
Korshenko, A. N. (1988). Ecological Conditions of the Eastern 

Equatorial Region of the Northern Pacific Ocean. 

Gidrometeoizdat Publishers, Moscow, (in Russian) 
MacDonald. J. D. & Lawford, P. A. ( 1954). Sight records of 

birds in the Pacific compiled from the bird log kept during the 

recent cruises of the H. M. S. Challenger. Emu 54( 1 ). 7-28. 
Marchal. E. ( 1 959 ). Analyse de quelques contrenus stomacaus 

de Neothunnus albacora (Lowe). Bull. Inst. Afrique Noire 

B(a). 1123-1136. (in French) 
Marks. J. S., Redmond. R. L., Hendricks. P.. Clapp. R. B. & 

Gill. R. E. (1990). Notes on longevity and flightlessness in 

bristle-thighed curlews. Auk 107(4). 779-781. 
Mayr. E. ( 1 945 ). Birds of the Southwest Pacific: A Field Guide 

to the Birds of the Area Between Samoa, NewCaledonia, and 

Micronesia. 316 pp. Macmillan Company. New York. 
McClure. H. E. ( 1974). Migration and Survival of the Birds of 

Asia. US Army Medical Component, SEATO Medical 

Project. Bangkok. Thailand. 
Medway, L. & Wells. D. R. (1976). Birds of the Malax 

Peninsula, Vol. V. Witherby, London. 
MorzerBruyns. W. F.J. ( 1965). Birds seen during west to east 

trans-Pacific crossing along Equatorial Counter-current 

around latitude 7°N, in the autumn of 1 960. Sea Swallow 17. 

57-66. 
Murphy, R. C. ( 1936). Oceanic Birds of South America. Am. 

Mus. Nat. Hist.. New York, pp. 1-1245. 
Murphy. R. C. ( 195 1 ). The populations of the wedge-tailed 

shearwater {Puffinus pacificus). Am. Mus. Novit. 1512. 

1-21. 
Murphy. R. C. & Snyder. J. P. ( 1 952 ). The "Pealea" phenomena 

and other notes on storm-petrels. Amer. Mus. Novit. 1596. 

1-16. 



254 



Nelson. J. B. (1975). Breeding biology of frigatebirds — a 

comparative review. The Living Bird 14, 1 13-156. 
Nelson. J. B. (1978). The Sulidae: Gannets and Boobies. 

Oxford University Press, Oxford, England. 1. 012 pp. 
Ng. S. C. ( 1978). Migratory barn swallows (Hirundo rustica) 

in the South China Sea. Malay. Nat. J. 32, 67-73. 
Nicholson, E. M. & Douglas, G. ( 1969). Draft checklist of 

Pacific oceanic islands. Micronesica 5(2), 327-463. 
NID. ( 1 945 ). Pacific Islands. Vol. IV. Western Pacific, 526 pp. 

Great Britain Naval Intelligence Division, B. R. 519C 

Geographical Handbook Series, Butler & Tanner, Ltd., 

London. 
Niering, W. A. & Sachet, M. -H. (1961). Observations on 

Puluwat and Gaferut. Caroline Islands, with historical and 

climatic information on Gaferut Island. Atoll Res. Bull. 76, 

1-15. 
Osipov. N. N. & Charykov, A. K. (1987). Concentration of 

heavy metals present in sea water by films of petrolic origin. 

In Geographic and Economic Problems Involved in the 

Investigation and Harnessing of the Southern Seas of the 

USSR. Abstracts: the Third Ail-Union Conference on Ocean 

Geography and Cartography (Rostoc-na-Donu, May 1987), 

Leningrad, (in Russian) 
Owen, R. P. ( 1 97 1 ). The birds of the Palau Islands including 

Helen Reef, and Palau bird name list, 10 pp. Trust Territory 

of the Pacific Islands. Office of the Chief Conservationist, 

Biology Laboratory, Koror, Palau, Caroline Islands. 

Unpublished report. 
Owen, R. P. (1977a). New bird records for Micronesia and 

major island groups in Micronesia. Micronesica 13( 1 ). 

57-63. 
Owen. R. P. (1977b). A checklist of the birds of Micronesia. 

Micronesica 13( 1 ), 65-81. 
Perry. R. (1980). Wildlife conservation in the Line Islands, 

Republic of Kiribati (Gilbert Islands). Envir. Cons. 7, 

311-318. 
Pratt. D. H., Bruner. P. L. & Berrett, D. G. (1987). The Birds 

of Hawaii and the Tropical Pacific. Princeton University 

Press. Princeton. New Jersey. 
RNZAF. (1986). Aerial photos of Maiden Island, SN 657 1 . 

Royal New Zealand Air Force, Dept. of Survey & Land 

Information, Wellington, New Zealand. 
Romanenko, V. I. & Kuznetsov, S. I. (1974). Ecology of 

Freshwater Microorganisms. A Laboratory Handbook. 

Nauka Publishers. Leningrad. 194 pp. (in Russian) 
Savilov. A. I. ( 1969). Pleuston of the Pacific Ocean. In Tikhii- 

Pacific Ocean. Part II. Deep-water Benthic Fauna. Pleuston, 

pp. 264-353. Nauka Publishers, Moscow, (in Russian) 
Schreiber. R. W. & Ashmole, N. P. (1970). Seabird breeding 

seasons on Christmas Island, Pacific Ocean. Ibis 112, 

363-394. 
Schreiber. R. W. & Schreiber. E. A. ( 1984). Central Pacific 

seabirds and the El Nino Southern Oscillation: 1982 to 1983 

perspectives. Science 225. 7 1 3-7 1 6. 
Sherman. K. (1963) Pontellidcopepod distribution in relation 

to surface water types in the central North Pacific. Limnol. 

Oceanogr. 8(2). 214-227. 



Sibley, F. C. & Clapp, R. B. ( 1 967). Distribution and dispersal 

of central Pacific lesser frigatebirds Fregata ariel. Ibis 109. 

328-337. 
Simpson, D. M. ( 1983a). Autumn migration of land birds off 

North Borneo in 1981. Sea Swallow 32, 48-54. 
Simpson, D. M. ( 1 983b). Birds seen at the Tembungo gas flare. 

North Borneo during the development of Typhoon Clara. 

Sea Swallow 32, 82-83. 
Sorokin. Yu. I. (1971a). Assessment of the role of 

bacterioplankton in the productivity of Pacific Ocean waters. 

In The Functioning of Pelagic Communities in Tropical 

Ocean Waters, pp. 92-122. Nauka Publishers, Moscow, (in 

Russian) 
Sorokin, Yu. I. ( 1 97 1 b). Quantitative assessment of the role of 

bacterio-plankton in the biological productivity of the tropical 

Pacific. In The Functioning of Pelagic Communities in 

Tropical Ocean Waters. Nauka Publishers. Moscow, (in 

Russian) 
Sorokin. Yu. I. ( 1973). Productivity of bacterioplankton in the 

western Pacific. Oceanology 13( 1 ), 95-106. (in Russian) 
Sorokin, Yu. I. ( 1976). Primary production in seas and oceans. 

In Science and Technology Findings. General Ecology. 

Biocenology. Hydrobiology. Vol. 1. VINITI. Moscow, (in 

Russian) 
Sorokin. Yu. I. ( 1978). Description of primary production and 

heterotrophic microplankton in the Peruvian upwelling 

region. Oceanology 18(1 ). 97-1 10. (in Russian) 
Sorokin, Yu. I. (1982). The role of microheterotrophic 

organisms in the functioning of marine ecosystems. Usp. 

Sovrem. Biol. 93(2), 236-252. (in Russian) 
Sorokin, Yu. I., Korsak, M. N„ Mamaeva, T. I. & Kogel'shchitu, 

D. D. ( 1983). Primary production and activity of microflora 

in Peruvian upwelling areas. In Bioproductivity of Upwelling 

Ecosystems, pp. 74-87. Moscow, (in Russian) 
SPREP. ( 1 989). The Northern Marshall Islands natural diversity 

and protected areas survey, 7-24 September 1988, 133 pp. 

South Pacific Regional Environment Program. Noumea. 

New Caledonia, and East-West Center. Honolulu, Hawaii. 
Stepan'yants, S. D. (1977). Siphonophores of the central 

Pacific Ocean. In Marine Plankton (Systematics and 

Faunistics), pp. 54-81. Leningrad, (in Russian) 
Stoddart, D. R. (MS). Scientific importance and conservation 

of Central Pacific Islands, 28 pp. Dept. of Geography, 

Cambridge University, England. Unpublished typescript 

manuscript. 
Tasker. M. L.. Jones. P. H„ Dixon, T. & Blake, B. F. ( 1984). 

Counting seabirds at sea from ships: A review of methods 

employed and a suggestion for a standardized approach. Auk 

101(31,567-577. 
Tsyban, A. V. (ed. ) (1980). Handbook of Methods for the 

Biological Analysis of Sea Water and Benthic Deposits 

(1980). Gidrometeoizdat. Leningrad. 191pp. (in Russian) 
Tsyban, A. V.. Panov, G. V.. Subbotin, V. G.. Timoshenkova, 

N. P. & Chernyak, S. M. (eds.). (1988). Methodological 

Foundations of Integrated Ecological Monitoring of the 

Ocean. Gidrometeoizdat Publishers, Moscow. 287 pp. (in 

Russian) 



255 



UNEP. (1984a). Marine Environmental Centres: East Asian 

Seas. United Nations Environment Programme, Regional 

Seas Directories and Bibliographies, Rome, Italy. 139 pp. 
UNEP. (1984b). Marine Environmental Centres: South Asian 

Seas. United Nations Environment Programme, Regional 

Seas Directories and Bibliographies. Rome, Italy. 31 pp. 
Vinogradov, M. E. ( 1978). Physical and biological features of 

equatorial upwellings. Bulletin MOIP, otd. bioL 83( 1 ). 

5-15. (in Russian) 
Vinogradov, M. E. & Shushkina, E. I. ( 1987). Functioning of 

the plankton communities of the ocean's epipelagial. Nauka 

Publishers, Moscow, (in Russian) 
Vinogradov, M. E. & Voronina, N. M. ( 1963). Distribution of 

plankton in waters of equatorial currents of the Pacific 

Ocean. Oceanology 21, 22. (in Russian) 



Voronina, N. M. ( 1964). Distribution of surface zooplankton 
in equatorial- current waters of the Pacific Ocean (focusing 
on copepods of the Pontellidae family). Trudy Institute of 
Oceanology AN SSSR 65, 95-106. (in Russian) 

Wells, D. R. (in prep. ). Malayan bird report: 1982 and 1983. 
Malay. Nat. J. 43. 

Wetmore, A. (1926). The Migrations of Birds. Harvard 
University Press, Cambridge, Massachusetts. 

Worcester, D. C. (191 1). Hybridism among boobies. Philipp. 
J.Sci. Sect. D6, 179-182. 

Zaitsev, Yu. P. ( 1971 ). Marine Neustonology. Published for 
NMFS. NOAA, US Department of Commerce and National 
Science Foundation, Washington, D.C. 207 pp. (Translated 
from the Russian) 



256 



General Conclusions 



The BERPAC expedition of 1988 instituted a series of 
research projects that greatly amplified the scope of the original 
program's overall research objectives. The cruise included an 
8-day stay on Caroline Atoll, an uninhabited necklace of islets 
enclosing a pristine lagoon. Chapter 1 deals at length with the 
history, geology, reef structure, botany, and ornithology of this 
little-studied ecosystem. Notable features of the atoll included 
a nearly-continuous reef surmounted by 39 islets in various 
stages of plant succession, many of them covered in virgin 
forest (Subchapter 1.1); a community of 1 1 species of breeding 
seabirds. numbering in excess of 1,000,000 individuals 
(Subchapter 1.2); and a remarkable Acropora -Tridacna reef 
containing the world's densest known colony of Tridacna 
clams (Subchapter 1 .4). The inner reef system has developed 
within a lagoon that is perched several inches above sea level 
at low tide. New species of plants, lizards, land birds 
(Subchapter 1.3), and seabirds were discovered for the atoll 
during the 1988 expedition. 

Caroline also provided an opportunity to compare basic 
oceanographic parameters between the open ocean, an enclosed 
lagoon, and the confined waters of the South China Sea. 

In Chapter 2, one of the more recent aspects of marine 
pollution — plastic contaminants — was examined in the central 
Pacific Ocean and South China Sea. Surface and subsurface 
water was sampled and the quantity and distribution of plastic 
debris determined. Plastics collected from the water were 
extracted and analyzed for organic pollutants to assess the 
potential hazard of the transfer of pollutants to marine organisms 
that ingest plastics. Results of sampling revealed that plastic 
debris, and specifically raw polyethylene pellets used in 
manufacturing, is widespread in the Pacific Ocean and South 
China Sea. Plastics occurred more frequently than did tar balls. 



even in the South China Sea. Organic contaminants were not 
associated with plastics at any detectable levels. However, 
subsequent studies demonstrated that plastics can adsorb certain 
contaminants and that this could represent a potential hazard to 
marine life. 

In Chapter 3, primary productivity was compared between 
waters close to and distant from Caroline Atoll. Waters close 
to Caroline showed less productivity than waters nearer to 
the equator (Subchapters 3.1 and 3.2). Not surprisingly, 
mesozooplankton (Subchapter 3.3) and neuston 
(Subchapter 3.4) diversity and biotnass also increased toward 
the equator, relative to the waters near Caroline, due in part 
to the hydrodynamics of the water columns near the equator. 
The equatorial parts of the Pacific were also much richer in 
seabird diversity and density than waters at higher latitudes 
(Subchapter 3.5). One surprise of the expedition was 
the discovery of a major land bird migration corridor over 
the South China Sea from Vietnam to Borneo 
(Subchapter 3.6). 

The second leg of the 1988 BERPAC expedition was the 
first of a planned series of similar cruises in the tropical Pacific 
and, as such, should be viewed as a forerunner of expeditions 
to come. Many disciplines relevant to the unique biology of the 
central Pacific were not represented in 1988, particularly those 
concerned with the functioning of coral reef ecosystems. Steps 
have been taken to issue a broader call to marine biologists to 
participate in future expeditions in order to further our 
understanding of the World Ocean and the marine-derived 
shallow waters and terrestrial ecosystems of which it is 
composed. If the publication of this volume inspires other 
biologists to follow, it has fulfilled part of its goal in presenting 
the varied findings of a successful initial expedition. 



257 



Appendix A 



Participants of the First Joint US-USSR 
Pacific Expedition, Fall 1988. 



Tsyban. A. V. USSR Chief Scientist 

Volodkovich, Y. L. USSR Assistant Chief Scientist 

Smith. G. J. US Chief Scientist 

Rostovtsev, O. A. Captain of the R/V Akademik Korolev 

Nelepov, Y. N. Assistant to the Captain on Scientific Affairs 

Vaytekaya, Y. I. Scientific Secretary 

Petrovskaya, S. V. Interpreter 

USSR Participants: 

Alexandrov, Boris G. (Microbiologist — Neuston) 
Junior Scientist, 
Institute of Biology of the South Seas & 

Academy of Sciences of Ukranian SSR. Odessa Branch 
37 Pushkinskaya Street, 27001 1 Odessa, UkSSR 

Barinova, Svetlana (Microbiologist — Microorganism indicators) 
Leading Engineer, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Belyaeva, Olga L. (Hydrobiologist — Polyaromatic hydrocarbons) 
Junior Scientist, 

Institute of Global Climate and Ecology 
USSR State Committee for Hydrometeorology & 
Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Chernyak, Sergei M. (Chemist — Chlorinated hydrocarbons) 
Senior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow. USSR 

Drakov, Sergei N. (Hydrobiologist — Vertical profiles) 
Leading Designer, 
Institute of Physics & 
Academy of Sciences, Beleorousskoi SSR 
Leninsky Prospekt. 70 
Minsk, BSSR 220602 



259 



Irha, Natalya I. (Organic Chemist — Photochemical oxidation of PAH's) 
Scientist, 

Institute of Chemistry & 
Academy of Sciences, Estonia SSR 
Akademia Teye Street, 1 5 
Tallin. ESSR 200108 

Kolobova, Tatiana P. (Analytical Chemist — Trace metals) 
Senior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Koltun, Vladimir M. (Hydrobiologist — Benthos) 
Department Head and Leading Scientist, 
Zoological Institute & 

USSR Academy of Sciences 
Universitetskaya Street, I 
Leningrad, USSR 199034 

Korsak, Mikhail N. (Hydrobiologist — Primary production) 
Senior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Korzhikov, Igor A. (Hydrobiologist — Primary production) 
Junior Scientist, 

Far Eastern Regional Research Institute of 
Goskomgidromet USSR 
24 Dzerjinskiy Stret, 690600 
Vladivostok, USSR 

Kudryatsev, Vassiliy (Microbiologist — Bacterial production) 
Senior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Kulikov, Audrey S. (Hydrobiologist — Mesozooplankton) 
Junior Scientist. 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Kumeisha, Alexander A. (Hydrooptician — Hydrooptics) 
Senior Scientist. 
Institute of Physics & 

Academy of Sciences, Beleorousskoi SSR 
Leninsky Prospekt. 70 
Minsk, BSSR 220602 

260 



Levina, Olga N. (Hydrobiologist — Microzooplankton) 

Engineer, 

Southern Division of the Oceanographic Institute & 

Academy of Sciences 
Gelendzhik 7,353470 
Oceanologiya, USSR 

Lukin, Alexander E. (Hydrooptician — Hydrooptics) 
Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 
Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Mamaev. Vladimir O. (Microbiologist — Numbers and biomass of microorganisms) 
Post Graduate Student, 
Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Mamaeva, Nila V. (Protozoologist — Microzooplankton) 
Senior Scientist, 
Southern Division of the Oceanographic Institute of the USSR & 

Academy of Sciences 
Gelendzhik 7. 353470 
Oceanologiya, USSR 

Marchenko, Alexanders. (Hydrobiologist — Technician) 
Senior Engineer, 

Institute of Biology of the South Seas & 
Academy of Sciences of Ukranian SSR. Odessa Branch 
37 Pushkinskaya Street. 27001 I 
Odessa, UkSSR 

Medinets, Vladimir I. (Hydrobiologist — Biosedimentation) 
Engineer, 

Odessa Department of the State Oceanographic Institute 
Goskomgidromet, USSR 
Proletarski Boulevard, 89 
Odessa, USSR 15, 270015 

Nelepov, Yeugeniy N. (Assistant to the Captain on scientific affairs) 
Far Eastern Regional Research Institute of 
Goskomgidromet of USSR 
24 Dzerjinskiy Street. 690600 
Vladivostok. USSR 

Panov, Gennadiy V. (Microbiologist — Bacterial degradation of pollutants) 
Senior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 



261 



Pershina, Irina V. (Analytical Chemist — Dissolved organic matter) 
Junior Scientist. 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Petrovskaya. Svetlana (Expedition Interpreter) 
Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street. 107258 
Moscow, USSR 

Polishchuk, Leonid N. (Hydrobiologist — Biogeography of zooneuston) 
Senior Scientist, 
Institute of Biology of South Seas & 

Academy of Sciences of Ukranian SSR. Odessa Branch 
37 Pushkinskaya Street. 27001 1 
Odessa, UkSSR 

Rostovtsev, Oleg A. (Captain of the R/V Akademik Korolev) 
Far Eastern Regional Research Institute of 
Goskomgidromet of USSR 
24 Dzerjinskiy Street, 690600 
Vladivostok. USSR 

Shigaev. Viktor V. (Oceanographer — Conductivity/temperature/depth) 
Senior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Sirenko, Boris I. (Hydrobiologist — Macros-benthos) 
Senior Scientist. 
Zoological Institute & 

USSR Academy of Sciences 
Universitetskaya Street. I 
Leningrad, USSR 199034 

Tsyban, Alia V. (USSR Chief Scientist) 

Professor of Ecology and Deputy Director. 
Institute of Global Climate and Ecology. 
USSR State Committee for Hydrometeorology & 
Academy of Sciences 
12. Morozo\ Street. 123376 
Moscow, USSR 

I Irbas, Eha R. (Organic Chemist — Photochemical oxidation of PAH's) 
Scientist. 

Institute of Chemistry & 
Academy of Sciences. Estonia SSR 
Akademia Teye Street. 15 
l.illin. ESSR 200I0S 



262 



Vaytekaya, Yanina I. (Hydrochemist — Scientific Secretary) 
Hydrometeorological Observatory of Klaipeda 
Board of Hydrometeorology of Lithuania 
Goskomgidromet of USSR 
Taikos Street. 26, 235800 
Klaipeda, USSR 

Ventsel, Mikhail V. (Hydrobiologist — Phytoplankton) 
Junior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 
Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Volodkovich, Yuriy L. (Hydrobiologist — USSR Assistant Chief Scientist) 
Senior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 

Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

Vronskaya, Valeriya M. (Hydrobiologist — Photochemical oxidation of PAH's) 
Junior Scientist, 

Institute of Global Climate and Ecology, 
USSR State Committee for Hydrometeorology & 
Academy of Sciences 
Glebovskaya Street, 107258 
Moscow, USSR 

US Participants: 

Ellis, David H. (Research Biologist, Ph.D. — Atoll ecology and oceanic seabird surveys) 
US Fish and Wildlife Service 
Patuxent Wildlife Research Center 
Laurel, Maryland 20708 

Kepler, Angela K. (Biologist, Ph.D. — Atoll ecology and oceanic seabird surveys) 
400 Snapfinger Drive 
Athens, Georgia 30605 

Kepler, Cameron K. (Research Biologist, Ph.D. — Atoll ecology and oceanic seabird surveys) 
US Fish and Wildlife Service 
Patuxent Wildlife Research Center 
Southeast Research Group 
University of Georgia 
Athens, Georgia 30602 

Sibley. Thomas H. (Aquatic Toxicologist, Ph.D. — Bioaccumulation of contaminants) 
Fisheries Research Institute. WH-10 
University of Washington 
Seattle, Washington 98195 



263 



Smith, Gregory J. (Biologist, Ph.D. — Methods of collection of plastics) 
Wildlife International, Ltd. 
305 Commerce Drive 
Euston, Maryland 21601 

Stafford. Charles J. (Analytical Chemist — Methods of analysis) 
US Environmental Protection Agency 
Analytical Chemistry Laboratory 
Beltsville, Maryland 20705 



264