ecological diagnostics for marine mammals : appraisal for … · 2014. 11. 18. · a.k. frie and...
TRANSCRIPT
Ecological Diagnostics for Marine Mammals: Appraisal
of molecular-based methods for dietary and age
estimation
Glenn John Dunshea
BSc (Hons) GCRC
Submitted in fulfillment of the requirements for the degree of
Doctor of Philosophy
Institute of Marine and Antarctic Studies
University of Tasmania
April 2012
“The ballast of factual information, so far from being just about to sink us, is growing less
daily. The factual burden of a science varies inversely with its degree of maturity”
Sir Peter Medawar, OM. “Two conceptions of science”,
Henry Tizard Memorial Lecture, Encounter 143, August 1965
Declarations
Originality;
This thesis contains no material which has been accepted for a degree or diploma by the
University or any other institution, except by way of background information and duly
acknowledged in the thesis, and to the best of the my knowledge and belief no material
previously published or written by another person except where due acknowledgement is
made in the text of the thesis, nor does the thesis contain any material that infringes
copyright.
Signed: Dated:
Glenn J Dunshea
Access;
This thesis may be made available for loan. Copying and communication of any part of this
thesis is prohibited for two years from the date this statement was signed; after that time
limited copying and communication is permitted in accordance with the Copyright Act 1968.
Signed: Dated:
Glenn J Dunshea
Ethics;
The research associated with this thesis abides by the international and Australian codes on
human and animal experimentation, the guidelines by the Australian Government's Office of
the Gene Technology Regulator and the rulings of the Safety, Ethics and Institutional
Biosafety Committees of the University.
This research was approved by the Animal Ethics Committee of the University of Tasmania
(Permit # A0008315).
Signed: Dated:
Glenn J Dunshea
Peer reviewed literature already published from parts of this thesis
Paper 1:
Dunshea, G., N. B. Barros, R. S. Wells, N. J. Gales, M. Hindell, A. and S. N. Jarman (2008)
Pseudogenes and DNA-based diet analyses; a cautionary tale from a relatively well sampled
predator-prey system. Bulletin of Entomological Research 98: 239-248.
Paper 2:
Dunshea, G. (2009). DNA-based diet analysis for any predator. PLoS One 4(4):
e5252.doi:10.1371/journal.pone.0005252.
Paper 3:
Garde E, A.K. Frie, G Dunshea , SH Hansen, KM Kovacs, C Lydersen (2010) Harp seal ageing
techniques-teeth, aspartic acid racemization, and telomere sequence analysis. Journal of
Mammalogy, 91: 1365–1374.
Paper 4:
Dunshea, G., D. Duffield, N. Gales, M. Hindell, R. S. Wells and S. N. Jarman (2011).
Telomeres as age markers in vertebrate molecular ecology. Molecular Ecology Resources, 11:
225-235.
Statement of Co-Authorship
The following people and institutions contributed to the publication of the work undertaken
as part of this thesis:
Paper 1 G. J. Dunshea (70%), N. B. Barros (5%), R. S. Wells (5%) N. J. Gales (5%)
M. A. Hindell (5%), S. N. Jarman (10%)
Paper 3 G. J. Dunshea (20%), Garde E (35%), A.K. Frie (25%), S.H. Hansen (10%),
K.M. Kovacs (5%), C Lydersen (5%)
Paper 4 G. J. Dunshea (65%), D. Duffield (5%), N. Gales (5%), M. Hindell (10%),
R. S. Wells (5%), S. N. Jarman (10%)
Details of the Authors roles:
Paper 1
G. J. Dunshea contributed to the idea of the manuscript, its formalisation and development,
performed the laboratory work, data collection, data analysis and wrote the paper
S. N. Jarman contributed to the idea of the manuscript, its formalisation, development and
manuscript refinement and presentation
N. B. Barros and R. S. Wells assisted with sample collection and manuscript refinement and
presentation
N. J. Gales and M. A. Hindell assisted with manuscript refinement and presentation
Paper 3
G. J. Dunshea, contributed to the idea of the paper, provided harp seal telomere assay data,
data analysis and interpretation, wrote the telomere part of the paper and assisted with
manuscript refinement and presentation
Garde E. contributed to the idea of the paper, its formalisation and development , provided
aspactic acid racemization data, its analysis and interpretation and wrote the bulk of the
paper
A.K. Frie and S.H. Hansen contributed to the idea of the paper, its formalisation and
development , provided teeth growth layer group analysis, sample collection and manuscript
refinement and presentation
K.M. Kovacs and C Lydersen contributed to the idea of the paper, its formalisation and
development, performed data analysis and manuscript refinement and presentation
Paper 4
G. J. Dunshea, contributed to the ideas in the paper, their formalisation and development,
performed most of the laboratory work, data collection, data analysis and wrote the paper
D. Duffield assisted with providing extracted DNA from historically archived samples
N. Gales contributed to the ideas in the paper, their formalisation, development and
manuscript refinement and presentation
M. Hindell contributed to the ideas in the paper, their formalisation, development and
manuscript refinement and presentation
R. S. Wells assisted with sample collection and manuscript refinement and presentation
S. N. Jarman contributed to the ideas in the paper, their formalisation, development and
manuscript refinement and presentation
We the undersigned agree with the above stated “proportion of work undertaken” for each
of the above published (or submitted) peer-reviewed manuscripts contributing to this thesis:
Signed: __________________ ______________________
Mark Hindell Mike Coffin
Supervisor Director
Institute of Marine and Antarctic Studies Institute of Marine and Antarctic
Studies
University of Tasmania University of Tasmania
Date:_____________________
Table of Contents Acknowledgements .......................................................................................................................... 1
Abstract ............................................................................................................................................ 2
- Introduction .................................................................................................................. 7 Chapter 1
1.1 Diet Investigations ................................................................................................................. 7
1.2 Age-based Investigations ..................................................................................................... 10
1.3 Structure of the thesis ......................................................................................................... 13
- A universal DNA-based diet technique for application to potentially any predator .. 15 Chapter 2
2.1 Introduction ......................................................................................................................... 15
2.2 Materials and Methods ........................................................................................................ 16
2.2.1 In silico development .................................................................................................... 16
2.2.2 Development of laboratory protocol ............................................................................ 19
2.2.3 Data analysis ................................................................................................................. 22
2.3 Results .................................................................................................................................. 23
2.3.1 Taxon specific presence/absence of Pac I restriction site in amplicon ......................... 23
2.3.2 ‘Universal’ Complementarity of PCR primers ............................................................... 24
2.3.3 Assessment of target 16SPLSU fragment to identify species from sequence data ...... 27
2.3.4 SSCP scoring of identical clones .................................................................................... 30
2.3.5 Removal of predator amplicons .................................................................................... 30
2.3.6 Prey detection from captive dolphin scat ..................................................................... 31
2.3.7 Prey detection from free ranging Sarasota Bay dolphin scat ....................................... 32
2.4 Discussion ............................................................................................................................. 34
- DNA-based diet analyses to describe cetacean diet; experimental and empirical Chapter 3
evaluations ..................................................................................................................................... 39
3.1 Introduction ......................................................................................................................... 39
3.2 Methods ............................................................................................................................... 40
3.2.1 Initial Tests of Efficacy of Different Protocols for Dolphin faeces DNA Extraction ....... 40
3.2.2 Contamination Control of DNA Extracts and PCR Assays ............................................. 41
3.2.3 Samples Used ................................................................................................................ 41
3.2.4 Universal Assays with SWBD samples ........................................................................... 43
3.2.5 PCR Primer Design and Testing ..................................................................................... 44
3.2.6 Real Time PCR screening for Mugil cephalus DNA in SWBD samples ........................... 45
3.2.7 Universal Assays with Sarasota Bay samples ................................................................ 45
3.2.8 Data Analysis ................................................................................................................. 45
3.3 Results .................................................................................................................................. 48
3.3.1 DNA Extraction Test ...................................................................................................... 48
3.3.2 Contamination Control of DNA Extracts and PCR Assays .............................................. 48
3.3.3 SWBD Universal Assays ................................................................................................. 49
3.3.4 SWBD Mugil cephalus DNA assays ................................................................................ 54
3.4 Sarasota Bay dolphin Universal Assays ............................................................................ 56
3.4.1 Prey Assignment by BLAST threshold and Placement in Neighbour Joining Analysis... 57
3.4.2 Prey Assignment Using the MOTU/Statistical Assignment Package method ............... 59
3.4.3 Comparison of different prey taxonomic assignment techniques ................................ 59
3.4.4 Occurrence and Inferred Relative Importance of Sarasota Bay dolphin prey .............. 64
3.4.5 Comparison Sarasota Bay dolphin diet estimated via DNA and stomach contents
data ........................................................................................................................................ 64
3.5 Discussion ............................................................................................................................. 68
3.5.1 Captive Feeding trials and SWBD prey detection results .............................................. 68
3.5.2 Sarasota bay data .......................................................................................................... 70
3.5.3 Conclusion ..................................................................................................................... 73
- Pseudogenes and DNA-based diet analyses: a cautionary tale from a relatively Chapter 4
well sampled predator-prey system .............................................................................................. 75
4.1 Introduction.......................................................................................................................... 75
4.2 Materials and Methods ........................................................................................................ 77
4.2.1 Sample collection and analysis ...................................................................................... 77
4.2.2 Sequence scoring and analysis ...................................................................................... 77
4.2.3 Confirmation of NUMT origin of spurious sequences ................................................... 78
4.3 Results .................................................................................................................................. 78
4.3.1 pNUMT frequencies, proportions and PCR characteristics ........................................... 78
4.3.2 NUMT sequence characteristics, phylogenetic analysis and substitution pattern ....... 81
4.3.3 Confirmation of NUMT origin of sequences ................................................................. 81
4.4 Discussion ............................................................................................................................. 85
– Telomeres as age markers for marine mammals ....................................................... 89 Chapter 5
5.1 Introduction.......................................................................................................................... 89
5.2 Methods ............................................................................................................................... 91
5.2.1 General Characteristics of Study Populations and Sample Sources ............................. 91
5.2.2 Bottlenose Dolphin samples ......................................................................................... 92
5.2.3 Southern Right Whales ................................................................................................. 95
5.2.4 Harp Seal Samples ......................................................................................................... 95
5.2.5 Data Analysis ................................................................................................................. 97
5.2.6 qPCR Assay Parameter Information .............................................................................. 98
5.3 Results .................................................................................................................................. 99
5.3.1 Bottlenose Dolphins ...................................................................................................... 99
5.3.2 Southern Right Whale qPCR assays ............................................................................ 105
5.3.3 Harp Seals.................................................................................................................... 105
5.4 Discussion ........................................................................................................................... 108
5.4.1 Bottlenose Dolphins and Southern Right Whales ....................................................... 108
5.4.2 Harp Seals.................................................................................................................... 110
5.4.3 Methodological Considerations .................................................................................. 112
– General Discussion ................................................................................................... 113 Chapter 6
6.1 DNA-based methods to study cetacean diet ..................................................................... 113
6.2 Live marine mammal age estimation and telomeres ........................................................ 120
6.2.1 Telomeres for age estimation: Reappraisal of the literature ..................................... 120
6.3 Conclusions ........................................................................................................................ 128
References ................................................................................................................................... 129
List of Figures
Figure 2.1 Proportion of species in alignments provided at
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0005252#s5
(number of species in each alignment provided in parentheses – x-axis) that have priming
regions directly complimentary at the first 8- 14 nucleotide positions of the 3’ ends to the
primers presented in the main manuscript (grey bars). Hatched bars represent the same
analysis when the additional primers (5’-3’): AAGACCCCGTTGAGCTT, AAGACCCTGTCGAGCTT,
AAGACCCTATCGAGCTT, AAGACCCTTTGGAGCTT, AAGACCCTATAAAACTT,
AAGACCCTGTGGAACTT, AAGACCCTATCGAACTT, AAGACCCTATAAATCTT and
AAGACCCTATAGATCTT are included. Exact proportion is indicated above bars. ................... 25
Figure 2.2 Test of universal primers designed for the small fragment of 16s mtDNA targeted
in this study. Lanes: M : 1kb ladder; 1: Mammalia: Arctocephalus pusillus doriferus, 2:
Mammalia: Pagophilus groenlandicus, 3: Mammalia: Tursiops truncatus, 4: Aves:
Aptenodytes patagonicus, 5: Aves: Stercorarius sp. 6: Teleosti: Sardinops sagax, 7: Teleosti:
Trachurus novaezelandiae, 8: Echinodermata: Centrostephanus rodgersii, 9: Echinodermata:
Crinoidea sp. 10: Mollusca: Nototodarus sp. 11: Mollusca: Nototodarus gouldi 12: Mollusca:
Octopoda sp. 13: Mollusca: Pteriomorpha sp. 14: Crustacea: Amphipoda sp. 15: Crustacea:
Penaeidae sp. 16: Crustacea: Thysanoessa macrura, 17 Insecta: Lepidoptera sp. 1, 18:
Insecta: Lepidoptera sp. 2. 19: PCR no template control........................................................ 27
Figure 2.3 Comparison of vertebrate (top graph) and invertebrate (bottom graph) within
species K2P distance (x-axis) and between species K2P distance (y-axis) in taxa from Table
2.3. Diagonal represents equal within and between species K2P distances. Points falling on
or below diagonal line represent species that have equal or higher intra-specific divergence
compared to the pairwise divergence with at least one other species. In such cases specific
identification from sequence data based on distance measures would be ambiguous......... 29
Figure 2.4. Example of SSCP gel used to identify identical clones from a clone library. M =
molecular size marker, NTC = No template control of PCR, E3, F3… = Different clones by well
position in 96 well plate. V1, V2… = Variant clones (i.e. All V1 clones are identical). Clone
identities confirmed by sequencing: V1: Sardinella lemuru/Sardinops sagax, V2: Scomber
australasicus, V3: Trachurus novaezelandiae, V4: Chimera of Trachurus novaezelandiae and
Sardinella lemuru/Sardinops sagax sequences, V5: Tursiops truncatus, V6: Sardinella
lemuru/Sardinops sagax, 2 substitutions difference from V1, V7: Arripis georgianus, V8:
Scomber australasicus 1 substitution difference from V2, V9: Poor sequence – discarded,
V10: Sillago robusta ................................................................................................................ 30
Figure 3.1 Frequency histogram of approximate SWBD scat volumes. The numbers of final
samples used differ from the totals since the original DNA extract may have been cross-
contaminated during DNA extraction procedures and there was no sample remaining to
perform a second DNA extraction .......................................................................................... 42
Figure 3.2 Priming regions of Mugil cephalus species specific primers designed in this study.
Dots represent same nucleotide as Mugil cephalus for that nucleotide position, ~ represents
the gap between priming regions. Note the mismatch between the PCR primers (particularly
at the 3’ end of the primers) for non-target species. ............................................................. 44
Figure 3.3 Bland-Altmann plot of PCR threshold cycle (c(t)) differences between DNA
extraction kits. Bold middle line represents the mean difference between MoBio and
QIAGEN kits (MoBio c(t) – QIAGEN c(t)) and the dotted line the standard deviation of the
mean. ...................................................................................................................................... 48
Figure 3.4 Proportion of predator origin (mtDNA and NuMT) clones from different sample
volume categories of SWBD samples. Sample volume categories and number assayed
indicated on the x-axis. Filled circles are individual data points, the open circles the mean of
each volume category, whiskers bootstrapped 95% binomial confidence interval for the
mean for grouped data. .......................................................................................................... 51
Figure 3.5 Actual Proportion of prey in SWBD diet (Dark Grey horizontal lines) compared to
proportions of amplicons in clone libraries calculated by different means. 1) Prey
proportions averaged across individual scat samples (circles), 2) Prey proportions averaged
across individual dolphins (squares), and 3) absolute total proportion (triangles) by pooling
all prey amplicons in clone libraries. Whiskers represent the bootstrap 2.5 and 97.5
percentiles for 95% confidence intervals (light grey lines) and 95% confidence intervals for
the mean (red lines) due to multinomial sampling error. Symbols represent arithmetic mean
between groups (circles and squares) or total prey proportion (triangle). For comparison,
arithmetic 95% confidence intervals of arithmetic mean proportions of raw data are given
(dotted lines). Prey species as follows: Sa: Scomber australasicus, Ag: Arripis georgianus, Mc:
Mugil cephalus, N: Nototodarus spp., P: Penaeidae spp ........................................................ 53
Figure 3.6 Samples that tested positive for Mugil cephalus DNA per day of the feeding trial
(sequential days from left to right). For the first day, only samples collected after the first
positive sample were included in this figure. Excluding the first test day, numbers above x-
axis are total sample numbers for that day. Data are black dots with actual value to the right
of the data point. Data are joined by lowess smoothed line. ................................................. 55
Figure 3.7 Boxplot with overlaid probability density function estimated with kernel density
estimates of proportion of positive samples from bootstrap replicates (n = 1000) from first
to last prey detection. Filled red line is the actual proportion of positive samples over the
duration of the prey detection period; dotted red lines are the bootstrap 95% confidence
intervals of the proportion of positive samples calculated from the bootstrap variance
estimate. ................................................................................................................................. 66
Figure 3.8 Average taxon (top panel: species; bottom panel: family) amplicons per clone
library. Filled circles represent arithmetic average across individuals, open circles represent
proportion of pooled total prey amplicons. Whiskers are 95% confidence intervals for 1)
population arithmetic mean (dotted lines of filled circles); 2) bootstrapped 95% confidence
intervals for mean due to sampling error in each clone library (red line of filled circles) and 3)
bootstrapped 95% confidence intervals for proportion of pooled total prey amplicons due to
sampling error in each clone library (filled lines of open circles) ........................................... 66
Figure 3.9 Absolute frequency of occurrence (sensu Wright, 2010) of six most common prey
species in Barros & Wells (1998) compared to this study. Whiskers represent exact 95%
binomial confidence intervals. ................................................................................................ 67
Figure 3.10 Average numerical proportion of prey items across all samples (other studies)
and average proportion of prey item clones across all sample (this study). Whiskers
represent truncated 95% confidence intervals of arithmetic mean. Variance was not
available for Berens et al. (2010). ........................................................................................... 67
Figure 4.1 Characteristics of putative NUMTs recovered from all samples (A) Frequency
histogram of the proportion of putative NUMTs in the clone library from each sample (n =
15); first category from 0-0.1 includes libraries with no pNUMTs. (B) Relationship between
the proportion of putative NUMTs in the library (x-axis) and the diversity of prey species
identifiable (filled circles) and putative NUMT haplotypes (open squares) (y-axis). These
relationships were not statistically tested as the variables are not independent. (C)
Relationship between the threshold PCR cycle of the 16S PCR used to amplify prey DNA (x-
axis) and the diversity of the prey species identified (y-axis) (Kendell Tau correlation; tau = -
0.45, z = -2.24, p = 0.03) (D) Relationship between the threshold PCR cycle of the 16S PCR
used to amplify prey DNA (x-axis) and the proportion of putative NUMTs in libraries (y-axis)
(Kendell Tau correlation; tau = 0.44, z = 2.29, p = 0.02) ......................................................... 80
Figure 4.2 Minimum Evolution phylogenetic tree displaying the relationship between
putative NUMT haplotypes and cetacean mtDNA. Other laurasiatherian mtDNA was used for
an outgroup (bottom seven branches). Topology was tested by bootstrapping with 1000
replications and the consensus tree is shown. Only values at nodes with a bootstrap score of
>50% are shown. Note the grouping of all putative NUMT haplotypes within the major
cetacean clade in relation to the outgroup but still distal to and highly diverged from the
majority of cetaceans. GenBank accession numbers are displayed after species name ........ 84
Figure 5.1 Initial terminal restriction fragment experiment on Tursiops truncatus DNA from
multiple individuals. Different individuals indicated by sample identifier at top of image.
Samples with * indicate samples of the same individual collected different years. Lines
indicate problems with sample seepage from bottom of gel well. Dark lines in outer lanes
represent Lambda Hind III size markers (top marker concatenation of 23130bp & 4361bp
fragments (27691bp)) with lines drawn between the same size markers on either side of gel
to account for uneven fragment migration and/or blotting distortion across gel. ................ 99
Figure 5.2 Bal 31 time series digestion of DNA from a single individual to characterize
telomere sequence distribution in bottlenose dolphins. 1. Gel image of experiment. DNA
size in base pairs indicated down left side. Lanes; 1-7 represent 1/6 volume of total Bal 31
treatment - 0 (control), 10, 20, 30, 40, 60 and 120 minutes of Bal 31 digestion. Lanes 8-14
represent remaining 5/6 volume from lanes 1-7 further treated with restriction enzymes as
per a conventional terminal restriction fragment assay. 2. Density plot relative copy number
in each DNA size class from restriction digestion treated DNA from the same experiment.
Digest times represented in top left of graph. Note the rapid digestion of DNA in the ≈23-4
kb portion after 60 minutes of digestion; this fraction represents true telomeric DNA
(Represented by a in both panels). Note also the resistance to Bal 31 digestion of laddering
in smaller size classes and of DNA up to ≈ 8.5kb that overlaps with true telomeric DNA
(represented by b in both panels). ....................................................................................... 100
Figure 5.3 Boxplot of telomere sequence per genome values from bottlenose dolphins.
Bottom whisker is 25th percentile, box is 25th-75th percentile with median dark line, top
whisker above 75th percentile. Outliers (open circles) are defined as over 1.5x the
interquartile range. Note the two extreme outliers, both from historical samples from 1995.
.............................................................................................................................................. 101
Figure 5.4 Telomere sequence per genome values of longitudinal samples from bottlenose
dolphin individuals collected at different ages. Whiskers for data points represent telomere
sequence per genome value standard deviation.................................................................. 102
Figure 5.5 Cross sectional telomere sequence per genome values for Sarasota Bay
bottlenose dolphin samples. Where two data points are from the same individual refer to
the legend on the upper right. There was no significant relationship between telomere
sequence per genome and age with this data. ..................................................................... 103
Figure 5.6 Boxplot of telomere values obtained by densitometric analysis of initial TRF
assays performed in Fig. 5.1. Bottom whisker is 25th percentile, box is 25th-75th percentile
with median dark line, top whisker above 75th percentile. Note the lack of variation across
values. ................................................................................................................................... 104
Figure 5.7 Densitometric profiles from representative lanes from Fig. 5.1. Legend show
representative lane from samples, age of samples in parentheses. FB19 (49 years old) had
the highest telomere metric values and FB157 (An adult of unknown age) had the lowest
telomere metric value. .......................................................................................................... 104
Figure 5.8 Telomere sequence per genome values from 12 samples from 6 cow-calf pairs.
Cows are grey bars, calfs white bars, whiskers are standard error. ..................................... 105
Figure 5.9 Example of Harp Seal TRF autoradiograph. ......................................................... 106
Figure 5.10 Plot of harp seal telomere sequence per genome estimates and tooth age. There
was no significant relationship between the two variables. Whiskers represent standard
deviation of CV-filtered telomere metrics. ........................................................................... 107
Figure 5.11 Plot of harp seal telomere metrics gained from densitometric analysis of TRF
assays and tooth age. There was no significant relationship between the variables with all
data, however if the two points from 1 year old males are excluded there is a significant
relationship between the variables. There was no valid reason to exclude these data points.
............................................................................................................................................... 107
List of Tables
Table 2.1Proportions of species that contain the Pac I restriction site within the amplicon
region out of species represented on GenBank (as of September 2006). .............................. 24
Table 2.2 Forward primers tested for 3’ complementarity with sequences provided in
supplement 2. Primers 1-10 are present in the degenerate forward primer presented in the
main manuscript and used in the study and primer pair empirical trial. Primers 11-19 are not
represented within the degenerate forward primer in the main manuscript; nucleotide
positions where these primers differ from those used in the study are shown in italicised,
bold and larger font. ............................................................................................................... 26
Table 2.3 Summary of comparisons of within species divergence with between species
divergence of species within each familial (or closest higher taxon) group for 16S mtDNA
target fragment. ...................................................................................................................... 28
Table 2.4 Prey and mammal clones identified in clone libraries from 5 captive feeding trial
samples subject to Treatment 2 (Pac I digestion for predator DNA exclusion prior to PCR and
also prior to cloning). Prey species abbreviations are shown below. ). Frequency of prey
occurance across scats is presented in the ‘FOC’ column. ..................................................... 32
Table 2.5 Prey operational taxonomic units (OTUs) identified from scat samples obtained
from free-ranging Sarasota Bay T. truncatus (n = 5 different individuals). Frequency of prey
occurance across scats is presented in the ‘FOC’ column. ..................................................... 33
Table 3.1. Summary of SWBD scat samples available for assays after DNA extraction and
contamination tests. Numbers in parentheses are the actual number of samples collected in
that day. Difference between samples used and numbers in parentheses are due to material
being unavailable for a second round of DNA extraction following contamination tests...... 41
Table 3.2 Summary of SWBD scat samples available for assays after DNA extraction and
contamination tests ................................................................................................................ 43
Table 3.3 Raw data from universal prey detection assays of 12 SWBD samples. Samples with
indicate this sample was assayed as part of the first chapter and the data are included
here. Dashed lines indicate different days of the feeding trial. ............................................. 50
Table 3.4 Nuclear mitochondrial pseudogenes (NuMts) variants isolated from SWBD clone
libraries ................................................................................................................................... 51
Table 3.5 Frequency of Occurrence (FOC) scores weighted by proportion of total prey clones
for SWBD samples. Prey groups arranged in order from most to least important according to
weighted FOC scores ............................................................................................................... 52
Table 3.6 Summary of OTUs identified from Sarasota Bay T. truncatus faeces and gastric
samples using a Threshold/Topological Neighbour Joining Analysis ...................................... 58
Table 3.7 Summary of Sarasota Bay T. truncatus number of clones per individual
taxonomically assigned to prey MOTUs using SAP software and accounting for Taq DNA
polymerase error in clone libraries. If Genus and Family level identifications were > 95%
similar to another identified taxon, they are listed below the most similar (% identity)
species level identification and indented. Blue text indicated a difference in prey assignment
between the ‘% Threshold/Neighbour joining’ method and the MOTU/SAP method. .......... 62
Table 3.8 Species and family Frequency of Occurrence (FOC) and FOC scores for taxa
appearing in more than one sample. Taxa are grouped in descending order of weighted
frequency of occurrence score................................................................................................ 65
Table 4.1 Summary of the occurrence between samples of all recovered putative NUMT
haplotypes and their BLAST closest matches .......................................................................... 79
Table 4.2 Number of pairwise nucleotide differences (bottom diagonal) and pairwise genetic
distances as estimated by the Kimura 2 parameter substitution model (top diagonal)
between suspected NUMT sequences, T. truncatus and the closest BLAST match of the
suspected NUMT’s. Light grey shaded areas are the pairwise comparisons between
suspected NUMTs and true cetacean mtDNA, dark grey shaded areas are the lowest genetic
distance estimate(s) of each suspected NUMT. ...................................................................... 83
Table 4.3 Results of referencing spurious sequences obtained from fecal samples in this
study against Tursiops truncatus draft whole genome shotgun sequences on GenBank by the
BLAST algorithm. Grey shading denotes homologous matches of ≥ 98%. ............................. 85
Table 5.1 Summary of Sarasota Bay samples for telomere assays. ........................................ 92
Table 5.2 PCR assay parameters for q-PCR telomere assays. ................................................. 98
Table 6.1 Summary of methods available to study the diet of live cetaceans.....................119
1
Acknowledgements
I thank my supervisors Nick Gales, Simon Jarman and Mark Hindell. Nick for finding a place
for me in his research program, opening up doors that would usually remain closed to me
and offering advice for developing my still appalling diplomatic skills. Thanks to Simon for
offering jovial advice on many things molecular and science writing, and both Simon and
Nick for facilitating my research through their institutional funding. Thanks to Mark for
providing timely and insightful academic guidance throughout the duration of my
candidature as well as a place in his lab at UTAS (which, importantly, has a fridge with beer
in it most of the time).
Funding for this project was provided by the Australian Antarctic Division, the Winifred
Violet Estate Trust Fund, the ANZ Holsworth Wildlife Research Endowment, The Norwegian
Polar Institute, The University of Tasmania and the Australian Centre for Applied Marine
Mammal Science (now the Australian Marine Mammal Centre). Further support was also
provided by Mote Marine Laboratories and the Chicago Zoological Society. I am extremely
grateful to these organisations for their financial support of my research.
I am grateful to Randy Wells, Nelio Barros (deceased), Dammon and Janet Gannon, Jason
Allen, Aaron Barleycorn, Brian Balmer and all staff and volunteers of the Sarasota Dolphin
Research Program. The work on this thesis on Sarasota dolphins would not have been
possible without these people. I especially thank Randy Wells for being generous in his
support and facilitation of my research. Thank you to Ruth Casper who performed the
dolphin feeding trial that some of the work in this thesis was based upon and who allowed
me to use the samples. Thanks to Sarah Laverick for helping Ruth with this feeding trial and
also providing administrative support. Thanks to C. Scott Baker for help in facilitating the
small Southern Right Whale component in this thesis. Thanks also to Christian Lydersen for
facilitating my participation in the harp seal study.
Lastly thank you to my beautiful girl Krista for putting up with me (for the most part)
throughout this process.
2
Abstract
Traditional molecular ecology has focused on describing the historic processes that lead to
contemporary patterns of diversity within populations, species and higher taxa. Molecular
tools can also identify the origin of biological material and shed light on contemporary
ecological processes of populations and individuals. This thesis is concerned with evaluating
the efficacy of some of these latter nascent applications for diagnosing near real-time
ecological information in marine mammals. The applications under investigation were (i)
DNA-based methods to identify the prey of cetaceans and (ii) using the size of telomeres
within the life of an individual to estimate the age of individuals in the Pinnipedia and
Cetacea.
Diet samples like faeces are complex mixtures of predator, prey and symbiont DNA and as
such they require techniques that can exclusively target prey DNA. Previous DNA-based diet
studies had employed species- or group-specific polymerase chain reaction (PCR) primers to
achieve this and thus were implemented ad hoc or required a priori diet knowledge, which
limited their scope. I developed a prey detection method that employed novel PCR primers
widely complementary (‘quasi-universal’) for most animal 16S mitochondrial DNA (mtDNA)
and a restriction enzyme to selectively exclude predator mtDNA. The method requires no a
priori diet knowledge and can be applied to other predators with a minimum of modification.
Faecal samples were collected from two sources; captive bottlenose dolphins (Tursiops sp.)
fed a known diet and free-ranging bottlenose dolphins from Sarasota Bay, Florida.
Two techniques were applied to detect prey DNA in the captive samples; amplification of a
small mtDNA fragment using a species-specific PCR primer pairs designed to detect a known
prey species and the ‘quasi-universal’ method. Using the species-specific method, a prey
signal was detected within 4-7 hours of feeding the captive dolphins the known diet and
persisted for at least 12-19 hours after the diet ceased. After the first detection, 60 +/- 12%
(mean +/- 95% CI) of captive samples contained a prey DNA signal using species-specific
methods. The ‘quasi-universal’ method was applied to 12 samples from within the time
period with a known diet of 10 prey species from 3 Phyla. Up to six prey species were
detected per sample (range 0-6, mean 3.2 +_ 1.7 (SD) species) and all but one prey species
consisting of 2% wet weight of the total diet were detected across all samples. No prey DNA
was detected from one captive sample using this method. Estimates of prey item amplicon
amounts showed congruence with the total proportion of wet weight of most prey items in
the diet, though variability introduced through sampling amplicon clone libraries puts wide
confidence intervals on these results.
The ‘quasi universal’ method was then applied to 15 faecal and 9 gastric samples from 19
free-ranging Sarasota Bay dolphins. Thirty two prey molecular operational taxonomic units
(MOTUs) were identified across all samples (range 0-9, mean 3.7 +/- 2.2 per individual),
consisting of 28 taxonomic assignments, 18 of which were species level identifications. One
3
sample did not contain prey DNA. The difference in results between samples from the
captive animals and free-ranging animals suggest that factors such as sample collection
methods, sample amount, sample storage duration and whether animals consume live or
dead prey may affect the efficacy of DNA-based techniques, which has ramifications for
interpreting results from captive feeding trials. These results were also congruent with diet
data from this population via traditional hard-parts analysis of stomach contents from
stranded individuals.
An unexpected consequence of using restriction enzymes to exclude predator mtDNA was
the appearance of nuclear mitochondrial pseudogenes (NUMTs) in samples. The appearance
of NUMTs in 15 faecal samples from Sarasota Bay dolphins was further investigated, in
order to understand their impact on DNA-based dietary analysis in a field situation. Nine
unique NUMT paralogs detected in 13 of 15 samples were represented by 1-5 paralogs per
sample and were estimated to be between 5-100% of all amplicons produced per sample.
The diversity of prey DNA and the proportion of NUMT amplicons per sample were related
to real-time PCR cycling characteristics, with lower prey diversity and a higher proportion of
NUMTs recovered with increasing real-time PCR threshold cycle values. This indicated that
low DNA yields from diet samples are more likely to have NUMTs detected and less likely to
contain prey DNA using this technique. This predator-prey system is relatively well sampled,
which facilitated ease of identification of NUMTs, however for many study systems this may
not be the case.
Telomeres are nucleoprotein structures on the end of eukaryote chromosomes that consist
of regions containing ‘telomere-like’ and ‘true’ telomere tandomly repeated DNA sequence
and single stranded telomere sequence overhangs on the 3’end of each anti-parallel DNA
molecule, each with associated proteins. They change size throughout the life of many
animals, suggesting that they be a molecular means to estimate animal age. To examine
whether telomeres would be useful to estimate the age of pinnipeds and cetaceans,
samples were collected from populations of three model species where the age of
individuals was known or could be relatively inferred; harp seals (Pagophilus groenlandicus),
bottlenose dolphins and southern right whales (Eubalaena australis).
In Harp Seals, telomeres were measured using two techniques, (i) de-naturing terminal
restriction fragment analysis and (ii) quantitative PCR (Q-PCR). There was no relationship
between age and telomere length using either telomere measurement method, however
there was a strong correlation between the methods, indicating that they were comparable.
Telomere dynamics in cetaceans were then investigated. Previous studies had shown that
satellite DNA in Mysticete cetaceans contains telomere sequence repeats that may bias
telomere measurement techniques. This was investigated in Odontocete cetaceans by
characterizing interstitial telomere sequence (ITS – that is, telomere repeat DNA sequence
that is not a part of true telomeres) in Bottlenose Dolphins. It was found that substantial ITS
exists in bottlenose dolphins, and given its presence in closely related Mysticete cetaceans,
most likely all cetaceans. The presence of this ITS made denaturing TRF analysis difficult to
4
interpret and so attempts were made to measure Bottlenose Dolphin telomeres using non-
denaturing TRF assays and Q-PCR assays. Attempts at non-denaturing assays were
unsuccessful and further efforts focused on Q-PCR assays. No relationship was found
between Q-PCR telomere metrics and age in bottlenose dolphins. In four instances
longitudinal samples were available from the same bottlenose dolphin individual a number
of years apart and these were compared. While two samples showed the typical pattern of
decline in telomere size with age, one showed no discernable change with age and one
individual displayed an apparent gain in telomere sequence with age.
Given these results a small subsection of 6 cow-calf pair Southern Right Whale samples
were initially analysed. Four adults appeared to contain shorter telomere sequences than
their calves although in two cases, calves contained less telomere sequence then their
presumed mother. In light of these results the use of telomeres to estimate age of individual
marine mammals did not appear a valid technique and could not be recommended.
Overall, this study achieved its aims of appraising the efficacy of these nascent molecular
ecology techniques in marine mammals. In the first instance, DNA-based diet analysis
appears to hold great promise for analysis of cetacean diet. The methods were sensitive,
identified prey to the lowest taxonomic level in many cases, and made use of samples that
cannot be used for traditional diet analyses. They allow high-resolution prey detection of
live animals, a feat that cannot be otherwise achieved in cetaceans failing direct observation
of feeding events. Additionally the ‘quasi-universal’ method can be applied to any predator
where its 16S mtDNA sequence is known or can be gained, and the biases of the technique
can be inferred by using current data from databases such as GenBank. Conversely,
telomeres appear to hold little use for age estimation in marine mammals. All
methodological issues aside, there are many exogenous and endogenous influences other
than chronological time on individual telomere dynamics and these are not well understood.
It is not recommended that telomeres be used for age estimation in any animal group
without considerably more work. The latter outcome is as useful as the former, since any
emerging technique (no matter how promising) must be put through rigorous critical
appraisal in order to understand whether the application is warranted at all, or if so what
the caveats might be.
5
6
7
- Introduction Chapter 1
This thesis investigates new approaches in two fundamental aspects of the natural history,
biology and ecology of any animal species; diet and age estimation. Knowledge of both the
diet and the age of animals have been recognized as fundamental biological factors since
some of the very first scholarly observations of animals (Arisotle ≈350BC). The reasons for
this are easy to understand - in the animal kingdom everything that lives, must eat, and also
inevitably grows old. These are at first, two seemingly disparate areas of a species’ biology.
However, the uniting focus of this thesis is not the inherent biological link between them.
Rather it is the relatively recent conceptual and technological advances in molecular biology,
producing molecular techniques that can be applied to each discipline. The application of
molecular biology to these areas could potentially allow information to be gathered that is
impossible to obtain by other means.
This introductory chapter outlines each of these two areas of study, provides the
background and rationale for the work performed, and ends by outlining and justifying the
structure of the remainder of the thesis.
1.1 Diet Investigations
The struggle for resources is an underlying driver of evolution (Darwin 2003). Food is a
principal key resource, as it is the basis of energy input for animals. Energy is the overriding
currency of all life. Therefore, the flow of energy into and out of systems, whether at the
level of single cell, organism, population, or entire ecosystem, is central to understanding
and integration of all these levels of biological phenomena.
The acquisition of nourishment is strongly influential on the growth, health, reproduction
and survival of individuals. Thus, a large proportion of the activity budgets of most animals
are devoted towards procuring food (Stephens and Krebs 1986). An understanding of the
basic mechanisms of how this is performed, such as what is eaten and how this is gathered,
forms a fundamental basis to quantifying the fitness of individuals, the natural history,
ecology and life history of species and, in due course, the dynamics of ecosystems. These
areas are important in the context of basic research and biological knowledge. They are also
increasingly important in an applied sense, as humans strive to understand and manage the
impact of their activities on natural systems.
The methods used to investigate the diet of animals are diverse, however they can be
broadly divided into two categories: Direct and indirect observation. Direct observation
entails observing what an animal eats in the wild. Indirect observations involve inference of
diet through the use of some sort of indicator or marker, which informs as to what the diet
8
may be. Markers used for indirect studies may be chemical or physical, but in all cases a link
between the marker and potential source(s) of the marker is required. For wild animal
studies both methods have been employed, however indirect methods are most often used.
This is simply because it is rarely possible to observe wild animals feeding in an unbiased
fashion or even at all. For the group of animals that is the focus of this thesis, marine
mammals, this is accentuated further since most are difficult to observe at all.
Except for rare observations of foraging, marine mammal diet has always been investigated
through indirect means. Traditionally, this has involved the quantification of prey physical
remains in the faeces or stomach contents of individuals (Pierce and Boyle 1991). Over the
last three decades, marine mammal diet investigation approaches have expanded to include
chemical markers, such as biological molecules and the stable isotopes of elements. Each of
these techniques has unique characteristics and constraints in terms of the accuracy and
precision of diet information obtainable and the temporal scale over which the information
is relevant. Detection of prey remains in faeces or stomach contents is conceptually simple
and can facilitate a high taxonomic resolution of prey over short timescales (days). Yet it is
biased by the differential survival of the diagnostic remains of different prey species (Pierce
and Boyle 1991). Some biological molecules such as the fatty acids of lipids can offer a
variable taxonomic resolution of prey integrated into the predator over short to medium
(days to weeks) timescales (Budge et al. 2006). However in most cases prey can only be
assigned to broad taxonomic categories qualitatively. Stable isotopes can inform as to the
environmental conditions and diet of a consumer at the time of their integration into tissue,
over timescales of days to years, dependent on the tissue analysed (Newsome et al. 2010).
Yet, in most instances it is difficult to derive any taxonomic information on the prey
consumed with stable isotope techniques, without some sort of in situ comparison.
A relatively recent advance in studying the diet of animals has come from detecting food
DNA in faeces or stomach contents (Kohn and Wayne 1997; Symondson 2002b). This
technique is conceptually similar to detecting prey hard parts, except food DNA is detected
using the tools of molecular biology, instead of visual identification of prey hard parts. DNA-
based diet techniques are appealing because they can be used in situations where the
application of previous techniques is not advantageous, desirable, or even possible
(Symondson 2002a). For example, DNA-based techniques do not suffer from the constraints
of differential hard-parts survival, in that there is no inherent bias towards detecting prey
with more robust hard parts. For animals that are fluid feeders, such as insects, or for other
animals which digest food thoroughly, such as cetaceans and seabirds, it is not possible to
perform faecal hard-parts analysis with any reliability at all.
DNA-based diet techniques have been applied to various marine mammal species with
varying degrees of success. The proof of concept was first demonstrated using faecal
samples from cetaceans (Jarman et al. 2002), however these techniques have since been
predominately applied to pinnipeds (Purcell et al. 2004; Deagle et al. 2005b; Casper et al.
2007a; Casper et al. 2007b; Deagle et al. 2009; Tollit et al. 2009). Experimental feeding trials
9
have demonstrated high rates of detection of known prey seem to vary depending on the
predator species, although detection rates are generally superior to hard parts analysis
(Deagle et al. 2005b; Casper et al. 2007a). As well as facilitating comparatively high rates of
prey detection, these techniques have identified prey taxa unable to be discriminated with
hard parts analysis (Tollit et al. 2009). There also appears to be some quantitative signal of
the mass proportions of prey consumed (Deagle et al. 2005b; Deagle and Tollit 2006; Bowles
et al. 2011), although comparison of these types of studies reveals inconsistencies in these
findings at present (e.g. Casper et al. 2007a; Deagle et al. 2010). Despite this, application of
DNA-based diet techniques to pinniped taxa has already revealed novel trophic insights that
were not possible to detect with other methods (e.g. Tollit et al. 2009).
Compared with pinniped studies, there have been comparatively few studies on the efficacy
of DNA-based diet techniques to investigate cetacean diet. Early studies demonstrated that
prey DNA could be detected in mysticete faeces (Jarman et al. 2002; Jarman et al. 2004).
These studies were important because they showed that there was potentially a technique
that could yield specific prey data from cetacean faecal samples, which was previously not
possible. Most other cetacean diet studies either did not yield specific prey information (e.g.
stable isotope or fatty acids studies), or those that did were performed by visual
examination of the stomach contents of stranded individuals, or those that were by-catch
from commercial fishing operations. There has long been a doubt as to whether data from
stranded or by-caught individuals was representative of the wider population, and few ways
to test the assumption that they were (Barros and Clarke 2002). Furthermore, DNA-based
diet techniques had never been applied to any odontocete cetaceans.
This thesis attempted to address some of the basic outstanding issues when applying DNA-
based diet techniques to cetaceans, using bottlenose dolphins (Tursiops sp.) as a model
species. For example, are these techniques amenable to studying odontocete diet? If so,
does sample quality and/or quantity effect this? How long after ingestion can a prey species
be detected and what is the temporal persistence of the DNA signal from consumed prey?
How well do DNA-based methods describe the diet qualitative and quantitatively, if at all?
Could these techniques be used to test some of the assumptions of traditional approaches
such as hard parts analysis? What are the limitations of these techniques when applying
them in unknown situations, such as studies of wild animals? These were the questions that
the DNA-based diet component of this thesis aimed to address.
Methods were developed to detect diverse prey DNA templates from predators with
unknown diet and the efficacy of these was examined using captive animals fed a known
diet. The techniques were then applied to samples from a wild free-ranging population to
examine their diet and test some assumptions of previous research findings. Finally, some of
the technical complexities of using DNA as a dietary marker were demonstrated and
recommendations on interpretation of results are provided that are applicable to all DNA-
based diet studies.
10
1.2 Age-based Investigations
‘Getting old’ is familiar to us all and is almost a ubiquitous phenomenon in the animal
kingdom. Biologically, ‘getting old’ or ageing, is manifest through the progressive physical,
biochemical , physiological and behavioural changes occurring during the processes of
development, senescence and finally death. These changes are continuous and generally
more pronounced the further apart they are in the lifetime of an individual organism, but
not necessarily in a linear manner. The age of an organism is therefore a time-referenced
record of the state of an organism along its life-schedule, relative to its birth. Thus, an
understanding of the age related changes in an organism, and how long it lives, are crucial
to understanding the life history and biology of the organism itself (Caughley 1977; Stearns
1992). Furthermore, by understanding the changes through time that individuals experience
and how this varies between individuals, it can be ascertained how these dictate population
dynamics processes. Although of fundamental theoretical importance, knowledge of age
related change and population dynamics processes of organisms is critical to detecting,
understanding and managing anthropogenic and natural impacts on individuals and
populations (e.g. Holmes and York 2003).
With the recognition that knowledge of age is fundamental to studying organisms and
populations, there is an obvious need for ways to estimate the age of study subjects. As
with diet methods, age estimation methods can be broken down into the categories of
direct and indirect methods. Direct age estimation methods estimate the age of individuals
through the use of some intrinsic marker which is a record of chronological time. Indirect
methods employ markers that assess the state of development of one individual against the
developmental state of others. Both methods have been widely employed for the purpose
of age estimation, but direct methods are preferred since they are less prone to being
confounded by unpredictable biological variation in the age marker employed.
Age estimation methods are highly diverse. Just about any variable that varies predictably
with age has been used to estimate the age of animals. For example, the diameter of
elephant faeces has been used to estimate the age of the elephant that produced the faeces
(Reilly 2002; Morrison et al. 2005). Similarly, the accumulation of particular biological
molecules with increasing age has been used to estimate the age of invertebrates (Ju et al.
1999). Both of these quite different methods are indirect, since they require the
relationships between the marker and age to be established in known age animals, before
they can be applied to unknown individuals. Examples of direct age estimation methods can
be found amongst studies of vertebrates, where hard tissues are deposited periodically with
respect to chronological time in some taxa. Examples of this include periodic deposition of
lamina or ‘growth rings’ in the otoliths of some fish species (Campana 2001), or of dentine
or cementum in the teeth of mammals (e.g. Hohn et al. 1989; Childerhouse et al. 2004).
There are also some aspects of bone growth and maintenance that vary with chronological
11
time in predictable ways and these have been used to estimate the age of some reptiles and
mammals (Zug et al. 1986; Marmontel et al. 1996).
For many animals there are no reliable techniques to estimate age and most of the
techniques with highest accuracy require either destructive sampling to access the tissue, or
highly invasive capture, restraint and surgical procedures to extract the tissue. These can be
justified in many instances, however where the study species is endangered, or where the
procedures may impact animal welfare significantly or affect the quality of subsequent data,
such procedures are highly questionable and often contentious (e.g. Fiesta-Bianchet et al.
2002 and Nelson 2002; Gales et al. 2005). If they can be applied at all, these techniques
become progressively less desirable the rarer, larger, and more long-lived a taxa is, because
these taxa are less resilient to impacts than short-lived taxa (Western 1979). Ironically, it is
also these taxa where age information is most important for understanding population
dynamics processes, since population processes are highly age structured in long-lived taxa.
Many marine mammals are large, long-lived taxa where gaining age information for
individuals is difficult. Most reliable marine mammal age estimation techniques require
highly invasive or destructive sampling, as they are based on counting periodic increments
in teeth or bone, or for baleen whales, timpanic bullae. There are also techniques which
measure the state of racemisation of asparctic acids enantiomers which require destructive
sampling. Recently, there has been application of bubber fatty acid profiles which vary with
age predictably to estimate the age of humpback and killer whales (Herman et al. 2008;
Herman et al. 2009). This last technique is the only one available that can be applied to live
marine mammals without capture and restraint to remove teeth. However, there is still
much work in understanding the mechanisms behind the relationships and how general
they may be applied within and between taxa (Herman et al. 2008; Herman et al. 2009).
There is a wealth of information that can be gained by ascertaining the in situ age structure
of marine mammal populations and the age of individuals followed in longitudinal studies,
so there is a strong desire amongst marine mammal biologists to develop techniques that
will facilitate this. At the time that the work on this thesis began (2005) there was no
reliable method to estimate the age of live marine mammals of unknown history without
capture to remove teeth, which is not possible for many marine mammal species.
However, there was a suggestion in the literature that telomeres may be used as a genetic
marker to estimate the age of live animals (Haussmann and Vleck 2002; Vleck et al. 2003;
Nakagawa et al. 2004). Telomeres are nucleoprotein structures on the ends of all eukaryote
linear chromosomes. They consist of tandemly repeated DNA sequence ending in a folding
loop structure at the terminus of the chromosome (Greider 1999), with telomere-specific
proteins along their length (Blackburn 1991). Telomeres have many known functions
including facilitating chromosome pairing during cell replication, signally cell cycle arrest and
importantly for the purposes of age estimation, protecting coding DNA from sequence loss
during the process of DNA replication (Blackburn 1991). Sequence loss occurs during DNA
replication because of the unidirectional synthesis of DNA and the biochemical properties of
12
DNA replication molecular machinery. DNA replication is always incomplete on the lagging
strand and DNA is lost during each cell cycle. This is known as the ‘end replication problem’
and was realized theoretically, long before it was demonstrated empirically (Watson 1972;
Olovnikov 1973). One of the principle functions of telomeres is to provide ‘expendable’ DNA
sequence at the end of chromosomes to buffer the loss of coding DNA due to the end
replication problem. When particular telomeres become critically short and unable to
maintain the loop structure on chromosome ends (i.e. are ‘uncapped’; Blackburn 2000),
they signal an arrest of the cell cycle and the cell enters a senescence phase and eventually
apoptosis (Henmann et al. 2001).
The fact that there is DNA lost during each cell cycle is a problem for tissues with high
replicative schedules throughout the life of organisms, such as germ line tissue. In such
tissues there is often an enzyme active called telomerase, a true reverse transcriptase,
which can elongate telomeres de novo (Blackburn 1991). While telomerase activity is high in
germ-line tissues, it is usually inactive or has very low activity in other tissues such as most
somatic tissue. Thus, in many somatic tissues across many taxa, there is a decrease in
telomere length with age, as cellular replication maintains somatic tissue continually with
age. For high maintenance tissues with a large turnover of cells, such as blood and skin,
there is a pronounced reduction of telomere length with age in vivo (Hastie et al. 1990;
Coviello-McLaughlin and Prowse 1997; Haussmann et al. 2003b). Early studies
demonstrated that there was a significant relationship between telomere and age in many
tissues (Harley et al. 1990; Friedrich et al. 2000) and these, in combination with the biology
of cellular and DNA replication, were the basis for the suggestion that telomeres could be
used as an age marker in animals of unknown history (Haussmann and Vleck 2002). The
overriding rationale for use of telomeres as age markers is to assay a tissue with a high rate
of mitotic turnover that is continual throughout life, such that telomere changes with age
will likely be most pronounced and continual.
The age estimation work performed in this thesis examined whether telomeres would be
useful for age estimation of pinnipeds and cetaceans. Skin, and in some cases blood, was
gathered from individuals of known age, or those where age could be relatively inferred.
Three model species were selected, mostly on the basis of availability of supporting data,
which represented three phylogenetically distinct marine mammal taxa: Bottlenose
dolphins, Southern Right Whales (Eubalaena australis) and Harp seals (Pagophilus
groenlandicus). Methodologies were assessed, telomere assays were performed and the
nature of change of these data with age was assessed for each species.
13
1.3 Structure of the thesis
Chapter 2: Since the DNA-based diet studies in this thesis were examining the diet of a
known generalist predator with a diverse diet, a molecular method was developed that
could detect diverse prey DNA from a diet sample. At the time that work started on this
thesis, the prevailing strategy in DNA-based diet studies was to target specific prey species
or taxonomic groups (Symondson 2002a; Jarman et al. 2004). These methods were
insufficient for the scope of work for this thesis as they required a priori knowledge of diet.
Some of the work performed was to test the assumptions of previous diet studies on wild
cetaceans, therefore using a priori diet knowledge to develop methods would have biased
these comparisons unacceptably.
Chapter 3: This chapter applies the method developed in Chapter 2 to faeces samples
collected from captive bottlenose dolphins fed a known diet, to test the efficacy of the
method for describing predator diet qualitatively and quantitatively. A species-specific assay
was also developed and applied to these samples to examine specific prey detection rates
and DNA signal persistence. The technique from Chapter 2 was also applied to diet samples
from free-ranging bottlenose dolphins and these results, the first to describe the diet in
detail of live free-ranging cetaceans, are compared to previous diet studies from this
population based on stomach contents analysis.
Chapter 4: This chapter examines some of the complexity of applying DNA-based diet
methods for detecting diverse prey DNA, such as the one developed in Chapter 2, to
samples from wild animals. It describes some of the challenges in assigning taxonomic
identity to DNA sequences using available databases, where there are few a priori data
available regarding the molecular diversity of predator and prey taxa. The chapter
specifically examines the confounding effects that nuclear mitochondrial pseudogenes may
have on DNA-based diet studies.
Chapter 5: This chapter examines and applies telomere assay methods for telomere-based
age estimation to the model species in this study. Suitable (or the only available) telomere
assays were applied to all study species and the relationships of telomere assay data with
age are examined. Recommendations were produced regarding the use of particular
telomere assay methods in taxa where genomic elements are uncharacterized and the use
of telomeres for age estimation in these taxa.
Chapter 6: This chapter presents a broad overview, general discussion and ramifications of
results from the entire thesis.
14
Chapter Two –
A universal DNA-based diet technique for application to potentially any predator
This chapter has been removed for
copyright or proprietary reasons
Published as :
Dunshea, G. (2009). DNA-based diet analysis for any predator. PLoS One 4(4): e5252.
doi:10.1371/journal.pone.0005252.
38
39
- DNA-based diet analyses to describe cetacean diet; Chapter 3
experimental and empirical evaluations
3.1 Introduction
Knowledge of diet is fundamental to understanding the ecology of any predator
(McLeod et al. 2003), as well as broader integration in trophic ecology and food webs
(Santos et al. 2001; Boyd 2002; Iverson et al. 2004), predator evolution (Gygax 2002; Vitt
and Pianka 2005) and competition processes between species (Heithaus 2001). However,
gathering diet information for marine mammals is particularly challenging due to the
inability to observe feeding while they are submerged or far from land (Barros and Wells
1998). Such constraints necessitate indirect methods for inferring marine mammal diet
such as identification of prey remains parts in stomachs or faeces (hard parts analysis;
Clarke and Macleod 1976; Santos et al. 2001; Flinn et al. 2002), analysis of fatty acid
(Iverson et al. 2004) and/or stable isotope (Newland et al. 2009; Newsome et al. 2010)
signatures in predator tissues and detection of prey DNA in predator faeces (Jarman et
al. 2002; Casper et al. 2007b; Deagle et al. 2009). Each of these methods offers a
different level of detail regarding taxonomic resolution of prey and the timescales over
which prey were ingested. Prey hard parts and detection of prey DNA can offer species
level identification of prey but generally over short (i.e. days) timescales (Deagle et al.
2005b; Casper et al. 2007a). Stable isotope techniques offer little taxonomic detail of
prey but may indicate relative trophic position over considerable time periods (weeks,
months and years). Fatty acid signatures provide a timescale of weeks to months (Budge
et al. 2006) and can identify lower level taxa (Iverson et al. 2004), if considerable
information on prey fatty acid variation and predator metabolism are available (e.g.
Iverson et al. 2004).
Compared to other marine mammal predators, studies of cetacean diet are further
limited by the nature of, and difficulty in obtaining, samples. Stable isotope studies from
biopsies have provided valuable insight into the trophic ecology of cetacean species and
communities (Valenzuela et al. 2009; Riccialdelli et al. 2010), yet lack detail on specific
diet composition. Contemporary investigations aiming to identify cetacean prey taxa
have used hard parts analysis of stomach contents from fisheries by-catch and stranded
animals (e.g. Barros and Odell 1990; Barros and Wells 1998; Evans and Hindell 2004;
Berens McCabe et al. 2010; Jansen et al. 2010). This has yielded information on prey
identity, relative composition and prey size. However, doubt exists whether data gained
from stranding and by-catch samples is representative of the entire population (Santos
et al. 2001; Barros and Clarke 2002). Stomach contents data also have well-recognised,
but unquantified, biases such as under-representation of particular prey with soft bodies
or those lacking robust bone structures and contention regarding prey size
reconstructions (Barros and Clarke 2002). It is possible in some circumstances to collect
faecal material from live cetaceans, however cetacean digestion is thorough and faecal
material generally consists of an amorphous slurry devoid of visually diagnostic predator
remains (Jarman et al. 2002; Jarman 2004; this study).
40
DNA-based diet techniques can identify prey in samples without the requirement of
visually recognisable prey parts (Symondson 2002b; King et al. 2008) and have been
proven successful for identifying prey in cetacean faeces (Jarman et al. 2002; Jarman et
al. 2004; first chapter). So far these studies have been restricted in terms of the
molecular approaches used to detecting prey DNA and small sample sizes. Yet questions
remain regarding prey DNA detection rates and passage times, and the ability to
reconstruct diet qualitatively and quantitatively have not been explored. Furthermore,
no studies have addressed the ability of DNA-techniques to compliment other
approaches, and allow appraisal of their biases. This latter point is particularly relevant
since at present there are no other practical means of gaining specific diet information
from free-ranging individuals.
This chapter aims to address some of the gaps in knowledge regarding the use of DNA-
based techniques to study cetacean diet. Specifically, a study of captive bottlenose
dolphins (Tursiops sp.) attempted to determine (i) prey detection rates, signal
persistence in scat samples and variation of these between individual dolphins, and (ii)
how the results were able to reconstruct diet, qualitatively and quantitatively. The
approach was then applied to diet samples (faeces and gastric) from free-ranging
bottlenose dolphins (Tursiops truncatus). The diet of this population is well studied by
stomach contents analysis of stranded animals (Barros and Wells 1998; Berens McCabe
et al. 2010) and experimental evaluations of their hypothesised foraging mode (Gannon
et al. 2005). Finally, these results were compared to stomach contents from the same
population data. I also compared approaches to identifying prey taxa from the DNA
sequence data typical of the results of these types of studies, to illustrate the difficulties
and pitfalls of assigning DNA sequence data to particular taxa in the absence of
complete databases.
3.2 Methods
3.2.1 Initial Tests of Efficacy of Different Protocols for Dolphin faeces DNA
Extraction
Two commercial kits (UltraClean™ Fecal DNA Isolation Kit (MOBIO Laboratories,Inc.) and
QIAamp DNA Stool Mini Kit (QIAGEN Inc.)) were compared for efficacy of DNA extraction
for dolphin scat samples. DNA was extracted from five samples and from at least 2 days
within the ‘Test’ diet period, using both kits following manufacturers protocols (with
modifications for viral DNA extraction for the QIAGEN kit as per Chapter 2). The DNA
extracts from both kits were assayed for presence of Mugil cephalus DNA using Real-
Time PCR with PCR primers BM-1Fwd and BM-2Rv and appropriate reaction and
thermocycling conditions PCR primers (See Below). This prey species was chosen
because of initial robust PCR results from species-specific primer trials, after prey
specific PCR primers when designed and optimised. Reaction profiles were detected on
RotorGene Real Time PCR thermocycler systems (Corbett Life Sciences) and PCR
threshold cycles (c(t) values) were determined for each DNA extract from each sample.
Within each sample consistently lower c(t) values were noted from the QIAGEN DNA
41
extracts indicating a better DNA extraction efficacy with this product (in the case of
dolphin scat samples – See Results). In light of this, the QIAGEN kit was used for all
subsequent DNA extractions.
3.2.2 Contamination Control of DNA Extracts and PCR Assays
For each batch of DNA extractions, a blank DNA extract was included in the batch and in
random order of the pipetting/centrifugation/decanting procedures, to control for
cross-contamination during the DNA extraction procedures. There were up to 23 scat
DNA extractions for each blank extraction, corresponding to the microcentrifuge
capacity available. Blank DNA extracts from each batch were assayed with the universal
PCR primers presented in chapter 2 (conventional PCR for 35 cycles with agarose gel PCR
product check), and in the case of the Sea World Bottlenose Dolphin samples, the Mugil
cephalus species-specific DNA assay (since this detected a smaller DNA fragment – see
below). If any batch blank DNA extraction was found to produce an amplicon signal or
product, this batch of DNA extractions were discard and if more sample was available,
the DNA extractions performed again. Assay PCR runs included 1-4 no template controls
to monitor for cross-contamination at PCR set-up.
3.2.3 Samples Used
3.2.3.1 Seaworld Feeding Trial Experimental Diet Samples
In light of the DNA extraction trial results, the remaining samples from the Seaworld
feeding trial first ‘Test’ (See Chapter 2) period had DNA extracted as described in
Chapter 2 and were used in this chapter. DNA was extracted from samples from 2 days
before the commencement of the ‘Test’ period, through the ‘Test’ period to four days
after the final ‘Test’ period day (Table 3.1). Throughout the text, these samples are
referred to as ‘Sea World Bottlenose Dolphin (SWBD)’ samples. A summary of SWBD
samples analysed in this chapter is presented in Table 3.1.
Table 3.1. Summary of SWBD scat samples available for assays after DNA extraction and contamination tests. Numbers in parentheses are the actual number of samples collected in that day. Difference between samples used and numbers in parentheses are due to material being unavailable for a second round of DNA extraction following contamination tests.
Period Non-Test
Test Non-Test
Day 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Totals
Samples Used
1
(4) 4
(6) 12
(12) 7
(9) 12
(13) 3
(4) 4
(5) 5
(5) 8
(8) 5
(5) 9
(11) 8
(8) 8
(10) 6
(8) 7
(11) 99
(119) Female 1 2 7 3 7 3 4 3 6 4 7 4 5 4 2 62
Male 0 2 5 4 5 0 0 2 2 1 2 4 3 2 5 37
ID Certain?
1 2 9 3 8 3 4 4 6 3 4 4 5 1 6 63
42
Approximate sample volumes were allocated to four categories using container
dimensions and measuring the depth of the scat material settled below the clear
ethanol after samples had sat undisturbed. The categories were: 1) ‘Scraps’ – Less than
1mm in bottom of sample container and only very small amounts in bottom of container
– could see through >50% of the clear bottom of the sample container, 2) <1.26 cm3 –
Less than 1mm in bottom of sample container but >50% of the bottom of the container
covered, 3) 1.26 – 3.77 cm3 – Between 1mm and 3mm in the bottom of sample
container, and 4) >3.77 cm3 – More than 3mm in the bottom of the sample container.
Most SWBD samples were in the <1.26 cm3 category (Fig. 3.1).
Figure 3.1 Frequency histogram of approximate SWBD scat volumes. The numbers of final samples used differ from the totals since the original DNA extract may have been cross-contaminated during DNA extraction procedures and there was no sample remaining to perform a second DNA extraction
3.2.3.2 Sarasota Bay Free-ranging Diet Samples
Scat samples and gastric samples were collected from 19 individual Sarasota Bay
Bottlenose Dolphins (Tursiops sp.) (Table 3.2). Scat samples were collected
opportunistically directly from the animal when they defecated while being handled for
measurements and collection of other samples (Wells et al. 2004). All but three scat
samples were collected on the deck of the research vessel. In 2006, three samples were
hand scooped from the water where the dolphins were being handled. For the samples
collected from the water, a control water sample (20ml of seawater) was also collected
up-current, in the vicinity of where the scat sample was collected.
Gastric samples were collected from 9 animals (2005 n = 5, 2006 n = 4) by running a
10mm plastic hose down the oesophagus of captured animals while they were on deck
of the research vessel. The hose was passed down to the fore-stomach and gastric
juiced were drained off into a 50ml falcon tube, which was subsampled into 5ml falcon
tubes. In five cases, gastric samples were available from the same animal from which a
19
48
26
6
28
54
31
0
10
20
30
40
50
60
Scraps <1.26 cm3 1.26 – 3.77 cm3 >3.77 cm3
Total Samples
Final Samples Used
43
scat was collected, allowing a comparison of the two sample types. Approximate sample
volumes for each sample type were measured as per SWBD samples. Table 3.2
summarizes samples collected from Sarasota Bay dolphins.
Table 3.2 Summary of SWBD scat samples available for assays after DNA extraction and contamination tests
Year IDcode Faeces Gastric Sex Age Length Faecal Volume
(cm3)
2005* SBD146 M 9 247 23.1
2005* SBD155 F 15 243 68.4
2005* SBD182 M 18 263 19.8
2005 SBD195 F Adult 258 65.9
2005 SBD199 F 3? 204 3.9
2005 SBD234 M Adult 221
2005* SBD236 M Adult 242 66
2005* SBD238 M 3? 209 72.6
2006 SBD10 M 25 257
2006 SBD20 M 17 273 6.6
2006 SBD90 F 36 255 36.3
2006 SBD113 F 10 228 14.5
2006 SBD133 F 7 217 25.1
2006 SBD151 F 6 210 23.8
2006 SBD157 F Adult 265 25.1
2006 SBD164 M 17 255 26.4
2006 SBD179 F 4 209 14.5
2006 SBD240 M 2 208 42.9
2006 SBD246 M 2 182
*Data from samples presented in Chapter 2
Samples (faeces and control) were stored for the day at 4°C until they were fixed by
addition of one volume of 100% molecular grade ethanol . Faeces and water samples
were then stored at-20°C until laboratory analysis. Gastric samples were stored for the
day at 4°C and were then stored at -80°C until laboratory analysis. Throughout the text,
these samples are referred to as ‘Sarasota Bay’ samples.
3.2.4 Universal Assays with SWBD samples
Universal diet assays described in Chapter 2 were undertaken on an additional 7
samples from within the test diet period. Additional samples were selected on the basis
of different scat volumes from at least two days within the ‘Test’ period and certainty of
the individual producing the scat. Including results from chapter 2, a total of 12 samples
were assayed with this methodology. Because of the clear effect of sample volume on
this assay (see results Fig. 3.4) and the fact that most samples were small volume
samples (Fig. 3.1), no further samples were analysed with this method.
44
3.2.5 PCR Primer Design and Testing
A species-specific PCR assay was designed to detect Mugil cephalus DNA in the Seaworld
samples. Here 16S sequences from all prey species fed to the dolphins were aligned and
species specific primers were designed using Amplicon Software (Jarman 2004). Primers
were designed to have maximum mismatch between non-target prey species in the 3’
end of both PCR primers and to have otherwise good PCR primer characteristics (i.e.
length, GC content, non-complimentary etc) (Fig. 3.1). Prey sequences were available
from GenBank, or were generated as part of this study as per Chapter 2.
The Mugil cephalus species specific PCR primers were BM-1Fwd (5’-3’):
GCTTAACTGTCTCCTTTTCCCAAC, and BM-2Rv (5’-3’) ACGTGGAACAGGGTTCATTTG, for
amplifying a 162 base pair (bp) length of Mugil cephalus 16S mtDNA (Fig. 3.1). PCR
conditions were initially optimized with AmpliTaq Gold Hot Start Taq DNA polymerase
chemistry using a Taguchi method modified for PCR optimization (Cobb and Clarkson
1994). AmpliTaq Gold chemistry was used for initial amplification and specificity trials
and afterward PCR master mix chemistry (SensiMix – BioLine, Australia) was used. Final
PCR reaction conditions were 1x SensiMix, 0.4µM each primer and 2.5µl scat template
DNA in a 12.5µl reaction volume. For trials of specificity and amplification efficiency
where prey template DNA was extracted direct from prey tissue, 1µl of template DNA
(concentration ≈1-50ng/µl) was used. The final thermal profile used for this assay was
95°C for 7.5 minutes then 40 cycles of 95°C for 15s, 60°C for 20s and 72°C for 20s with a
final extension of 72°C for eight minutes. The melt curve thermal profile was 78°C for
30s followed by sequential 2s holds at subsequent steps of 0.1°C to 95°C.
The specificity of Mugil cephalus species specific PCR primers was examined empirically
by attempting to amplify DNA from 2-5 individuals of each non-target prey species and
examining amplification success with agarose gel electrophoresis initially, then with real-
time amplification and melt curve profiles. Reaction efficiency was calculated from the
slope of c(t) values from a serially diluted plasmid standard containing the Mugil
cephalus LR-J-12887 and LR-N-13398 16S mtDNA fragment.
Figure 3.2 Priming regions of Mugil cephalus species specific primers designed in this study. Dots represent same nucleotide as Mugil cephalus for that nucleotide position, ~ represents the gap between priming regions. Note the mismatch between the PCR primers (particularly at the 3’ end of the primers) for non-target species.
45
3.2.6 Real Time PCR screening for Mugil cephalus DNA in SWBD samples
SWBD samples from two days prior to four days after the first ‘test period (n=99) were
screened for the presence of Mugil cephalus DNA with real-time PCR using Rotorgene
thermocyclers (Corbett Life Sciences) and reactions were set up using the Corbett liquid
handling robotics system. Both c(t) values and melt curve profiles were used to confirm
the amplification of Mugil cephalus 16S mtDNA in samples, as in some cases non-specific
amplification of some other template occurred which resulted in amplicons with
different melt temperature profiles. All samples were run on two separate occasions and
there was no disagreement between runs. The initial intention for the SWBD samples
was to design species specific PCR primers for all ‘Test’ prey species and screen each
sample for them, however after initial results with the Mugil cephalus PCR primers and
the universal prey detection assay these plans were abandoned (See Results and
Discussion).
3.2.7 Universal Assays with Sarasota Bay samples
All additional scat and gastric samples from 2005 and all scat and gastric samples from
2006 were assayed with the universal method (Chapter 2). For these samples up to 19 -
41 clones per individual (19 - 20 per scat; 15-38 per gastric sample) were scored for
sequence variants with single stranded conformational polymorphism analysis prior to
sequencing representative variants (see chapter 2). DNA was extracted from gastric
samples as per the procedure used in Chapter 2 for scat DNA extraction, except the
whole gastric sample (i.e. 5ml falcon tubes) was spun down for 5 minutes at 12000g and
the clear gastric juice decanted. DNA extraction was performed on the solid material
that pelleted at the bottom of the centrifuge tube.
3.2.8 Data Analysis
3.2.8.2 SWBD Data Analysis
Since the SWBD samples were a sub-sample of all defecations within the time period of
the study, bootstrapping was used to estimate the reliability of the estimate of the
proportion of positive samples for Mugil cephalus DNA. There were 68 samples
(inclusive) between the first and last positive sample. The bootstrap procedure re-
sampled these data with replacement 68 times and the number of positives was
recorded. This procedure was repeated 1000 times and bootstrap sample variance used
to calculate confidence intervals for proportions estimates (Efron and Tibshirani 1993).
Where clones were considered of predator or non-predator origin, mean binomial
proportions within groups and confidence intervals (95%) were estimated using the
analysis of overdispersed data (‘aod’) package in R (Lesnoff and Lancelot 2010). The
bootstrap (1000 replicates) estimates of mean binomial proportion and 95% confidence
intervals are presented.
46
Three approaches were used to compare whether prey amplicon proportions from
SWBD universal assays reflected wet weight proportions of diet species; 1) prey
amplicon proportions were calculated for each fecal sample and averaged across all
samples, 2) prey amplicon proportions were calculated for the pooled clone libraries
from each individual dolphin and averaged across individuals, and 3) all clone libraries
were pooled and absolute proportion of prey amplicons compared to actual diet.
Proportions of prey amplicons in clone libraries had two sources of variability that
required description. The first was variability in amplicon proportions within each clone
library due solely to error from sampling a multinomial population. The second was
variation between the groups of samples (i.e. scats and individual dolphins). To estimate
the effect of multinomial sampling error on the mean proportions of prey amplicons
across scats, for individuals and for total proportions, a bootstrapping approach was
implemented. Since the sampling unit of all levels of analysis (i.e. scat, individual and
pooled total) was individual sample clone libraries, the prey amplicons of each clone
library were re-sampled with replacement with a sample equal in size to the original
sample of prey amplicons for that sample. These data were collated by scat, individual
and the total pool and proportions calculated and averaged where applicable (i.e.
between scats and individuals). This procedure was repeated 1000 times and the
bootstrap sample variance used to calculate 95% confidence intervals for the estimate(s)
of the mean or the total proportion. For across group means, the 95% confidence
intervals were also calculated for each bootstrap replicate and the 2.5 and 97.5
percentile value of these used to represent the 95% confidence interval for variability
across groups (i.e. fecal samples and individuals), which takes into account the effect of
multinomial sampling error on across group variation.
3.2.8.3. Sarasota Bay Samples Sequence Data Analysis
Two different species assignment processes were implemented for prey sequences
recovered from Sarasota Bay samples. Firstly, a “percent threshold, neighbour joining
topology” approach identical to the process described in Chapter 2. Secondly, all prey
sequences from all samples were examined and edited concurrently and primer regions
removed. Sequences were assigned into molecular operational taxonomic units (MOTUs)
on the basis of similarity allowing for Taq DNA polymerase error, since cloned amplicons
of the same sequence may differ due to Taq DNA polymerase error. Following Thalmann
et al. (2004), with a Taq DNA polymerase error rate of 7.3 x 10-5(Kobayashi et al. 1999), it
was estimated that a sequence could be the same as a reference if it contained ≤ 2
substitutions over 220bp. However, if the same MOTU containing ≤ 2 substitutions
relative to a reference was present in more than one sample, it was interpreted that the
sequence was a genuine variant since the likelihood of the same random substitution(s)
occurring at the same nucleotide position(s) in the amplicon in different PCRs is
exceedingly low. If MOTUs were not a 100% match to database reference sequences,
they were assigned to the lowest taxonomic unit using Bayesian phylogenetic analysis,
implemented in the Statistical Assignment Package (SAP) (Munch et al. 2008a). The SAP
estimates the posterior probability that a sequence is part of a monophyletic group by
comparing the sequence to databases, gathering significant homologues and associated
47
annotated taxonomic data and performing Bayesian phylogenetic analysis (Munch et al.
2008a). Posterior probabilities for monophyletic groups are estimated using Markov
chain Monte Carlo sampling of all possible Bayesian phylogenetic trees (Munch et al.
2008a) and the lowest level taxon with ≥ 95% prior probability was assigned to the
sequence. For all taxonomic assignments using either method, the assignments were
finally hand-curated to assess if the matching taxon had a distribution that included
Sarasota Bay and surrounding waters using online databases such as FISHBASE
(www.fishbase.org). In some cases further lower level taxonomic assignment, or limited
options for identity were made available at this last step, since there may be only one
genus or few species that are known from the study area.
Because chimeric DNA sequences can be obtained when PCR amplifying mixed,
degraded DNA sources, all sequences unable to be identified to at least Genus were
also analysed with a different more comprehensive chimera detection software package
than that used in Chapter 2; ChimeraSlayer (Haas et al. 2011). Although nuclear
mitochondrial pseudogenes (NUMTs) can confound prey identification using mtDNA
(See Chapter 4), sequences < 98% similar to the closest prey match were all most closely
matching to teleosts and thus were not examined for NUMT origin, given the
conspicuous absence of NUMTs in this chordate group (Bensasson et al. 2001b;
Venkatesh et al. 2006). Sequences identified as predator NUMTs were recovered from a
number of faeces samples from both SWBD and Sarasota bay dolphins.
Proportions of each prey type in the individual clone libraries were estimated based on
variant proportions in the sampled clones. Given the results of the SWBD clone library
analysis, inference regarding the importance of different prey types to Sarasota Bay
dolphins was made using a weighted frequency of occurrence approach and averaging
prey amplicon proportions across individuals. The same bootstrapping procedure was
applied to these data as was used for the SWBD clone library data to represent the
effect of multinomial sampling error on pooled amplicon totals and between individual
means and 95% confidence intervals.
When comparing prey occurrence and amount values between this and past studies,
95% confidence intervals were calculated for inference of difference between studies
following (Wright, 2010). Exact binomial confidence intervals were calculated for
frequency of occurrence values and arithmetic confidence intervals were calculated for
numerical proportions and proportion of prey amplicons across clone libraries.
48
3.3 Results
3.3.1 DNA Extraction Test
The QIAGEN kit yielded lower PCR threshold cycle (c(t) ) values for four out of five
samples positive for Mugil cephalus DNA (Fig 3.2). The MoBio kit had c(t) values on
average 3.59 +/- 2.75 (S.D) cycles higher than the QIAGEN kit (Fig 3.2), indicating a
comparatively lower DNA yield (on average ≈12 times less prey DNA given exponential
DNA amplification at perfect PCR efficiency) and thus the QIAGEN kit was used for
subsequent DNA extractions.
Figure 3.3 Bland-Altmann plot of PCR threshold cycle (c(t)) differences between DNA extraction kits. Bold middle line represents the mean difference between MoBio and QIAGEN kits (MoBio c(t) – QIAGEN c(t)) and the dotted line the standard deviation of the mean.
3.3.2 Contamination Control of DNA Extracts and PCR Assays
DNA extracts for SWBD samples were extracted in 15 batches with between 3-23
samples in each batch. There was no sign of cross-contamination when extraction blanks
from each batch were assayed with the universal PCR primers presented in Chapter 2.
However, when extraction blanks were assayed for Mugil cephalus DNA using species
specific primers, 7 of the 15 batches had late (>34 cycles) amplification signals for Mugil
cephalus DNA. All DNA extracts from these batches were discarded and DNA was
extracted again from these samples with fresh reagents. Of the 48 samples that needed
re-extraction, 19 did not have sufficient material since all material was used in the
49
original DNA extraction (see Table 3.1, Fig 3.1 for details). New DNA extract batch blanks
(n = 2) were assayed for contamination and no amplification was apparent from
universal and Mugil cephalus DNA assays. After this quality control was performed the
total number of SWBD samples available was 100 from 119 with a maximum of 4
samples lost in a single day due to contamination at the initial DNA extraction step (see
Table 3.1).
Sarasota Bay samples were extracted in two batches, one for faeces samples and the
other for gastric samples. Blank extracts from these batches were assayed with the
universal primers from chapter 2 and there was no indication of contamination in either
batch.
3.3.3 SWBD Universal Assays
A total of 470 clones were screened from 12 clone libraries with 29 – 52 clones screened
per clone library. At least one prey species was detected from all but one sample using
the universal method with an average of 3.2 +/- 1.7 (S.D) prey species identified per scat
(range 0-6; Table 3.3). All species of prey except for penaidae were recovered at least
once across all clone libraries (Table 3.3).
3.3.3.1. Clones of predator Origin in SWBD clone libraries
Chapter 4 investigates the NUMTs found in all clone libraries in terms of the sequence
diagnosis of NuMts and ramifications of these for DNA-based diet studies. In this chapter,
the data are presented to summarize the effect of predator origin clones on the efficacy
of this method for SWBD dolphin scats.
There were 10 putative NuMt variants isolated across all clone libraries with 8 appearing
in different samples and 7 in multiple individuals (Table 3.4).The proportion of predator
origin clones (i.e. T. truncatus mtDNA and nuclear mitochondrial pseudogenes – NuMts)
in lower volume samples was greater than in the relatively large volume samples used in
Chapter 2, suggesting sample volume may affect the efficacy of this method for these
dolphin scat samples (Fig. 3.4). The sparseness of this dataset, non-independence of
samples and the variation present in the 1.26 – 3.77 cm3 category meant these data
were unsuitable for modelling with logistic regression, however inference from 95%
confidence intervals of the mean in each volume category supports an effect of sample
volume on the efficacy of this method for these dolphin scat samples (Fig. 3.4).
Table 3.3 Raw data from universal prey detection assays of 12 SWBD samples. Samples with indicate this sample was assayed as part of the first chapter and the data are included here. Dashed lines indicate different days of the feeding trial.
Sample Volume (cm
3 )
Trial Day
Animal Time Sa Ag Mc N P Tn Sr Sl & Ss
Ps Tt Pg Total Clones
Screened
Prey Species
Detected
76 <1.26 5 Gemma 1130 0 0 0 0 0 0 0 0 0 0 29 29 0
77 1.26 – 3.77 5 Moki 1215 5 2 0 9 0 1 1 21 0 2 0 41 6
78 >3.77 5 Nila 1330 11 5 0 2 0 1 0 32 0 0 1 52 5
79 >3.77 5 Gemma 1400 0 1 0 0 0 4 0 16 0 4 9 35 3
81 >3.77 5 Gemma 1535 8 6 0 0 0 4* 1 17 0 1 0 37 5
91 >3.77 7 Gemma 1415 2 13 16 0 0 0 0 2 0 0 5 38 4
97 1.26 – 3.77 9 Stormy 815 1 0 14 0 0 0 0 0 0 0 24 39 2
99 1.26 – 3.77 9 Stormy 935 10 0 0 0 0 1 0 0 1 10 9 31 3
102 >3.77 9 Stormy 1355 5 0 11 0 0 12 0 0 4 6 1 39 4
104 Scraps 9 Buck 1410 0 0 1 0 0 0 0 2 0 6 25 34 2
112 Scraps 10 Stormy 945 2 1 1 0 0 0 0 0 0 19 25 48 3
118 1.26 – 3.77 10 Stormy 1135 1 0 0 0 0 0 0 0 0 17 29 47 1
Combined Totals
45 28 43 11 0 23 2 90 5 65 157 470
FOC* 9/12 6/12 5/12 2/12 0/12 5/12 1/12 6/12 2/12
Sa: Scomber australasicus, Ag: Arripis georgianus, Mc: Mugil cephalus, N: Nototodarus spp., P: Penaeidae spp., Tn: Trachurus novaezelandiae, Sr
Sillago robusta, Sl & Ss: Sardinella lemuru, Sardinops sagax, Ps: Pomatomus saltatrix, Tt: Tursiops truncatus, Pg: Tursiops truncatus pseudogene.* FOC
– frequency of Occurrence
51
Table 3.4 Nuclear mitochondrial pseudogenes (NuMts) variants isolated from SWBD clone libraries
SWBD NuMt Variant
Sample ID Individual 1 2 3 4 5 6 7 8 9 10 Unique Variants
76 Gemma Y Y Y Y 4
77 Moki 0
78 Nila Y 1
79 Gemma Y 1
81 Gemma 0
91* Gemma 1
97 Stormy Y Y Y Y Y 5
99 Stormy Y Y Y Y 4
102 Stormy Y 1
104 Buck Y Y Y Y Y Y Y 7
112 Stormy Y Y Y Y Y Y 6
118 Stormy Y Y Y Y Y Y Y Y Y 9
*Sequence was of poor quality although obviously originating from a Numt. The clone was not re-sequenced and subsequently not classified to any one NuMt variant.
Figure 3.4 Proportion of predator origin (mtDNA and NuMT) clones from different sample volume categories of SWBD samples. Sample volume categories and number assayed indicated on the x-axis. Filled circles are individual data points, the open circles the mean of each volume category, whiskers bootstrapped 95% binomial confidence interval for the mean for grouped data.
52
3.3.3.2 Results of SWBD Universal Assay Prey Detection compared to known diet
For all approaches comparing clone library results to known diet, predator derived
clones were excluded from consideration. Since the actual proportions of each of the
‘non-test’ prey species were unknown, these were pooled together in one category.
Simple frequency of occurrence (FOC) placed Scomber australasicus as the most
important prey item (9/12) followed by ‘non-test’ prey species (8/12), Arripis georgianus
(6/12), Mugil cephalus (5/12) and Nototodarus spp. (2/12) (Table 3.4). This is in contrast
to the actual importance by wet weight of the prey categories of ‘non-test’ species
(43%), equal S.australasicus, A. georgianus and M. cephalus (15% each) and Nototodarus
spp. (10%).
A weighted FOC score was calculated for each prey group where FOC was normalised to
the prevalence of clones of each prey species as a proportion of all prey clones across all
samples. By this scale ‘Non test’ prey were the most important prey group, followed by
S.australasicus, M. cephalus, A. georgianus and Nototodarus spp. (Table 3.5).
Table 3.5 Frequency of Occurrence (FOC) scores weighted by proportion of total prey clones for SWBD samples. Prey groups arranged in order from most to least important according to weighted FOC scores
Prey Group FOC Total Clones (%)
Weighted FOC Score
Ratio to 'Non-Test' Weighted FOC
Ratio to 'Non-test' by wet weight
‘Non-test' 8/12 48.6 0.324 - -
S.australasicus 9/12 18.2 0.137 1:2.4 1:2.86
M. cephalus 5/12 17.4 0.073 1:4.5 1:2.86
A. georgianus 6/12 11.3 0.057 1:5.7 1:2.86
Nototodarus sp. 2/12 4.5 0.007 1:43.6 1:4.3
Penaidae 0/12 0 0.000 - -
Mean prey proportions between all scat samples indicated the same relative importance
of prey items as the weighted frequency of occurrence approach (Fig. 3.5). Arithmetic
confidence intervals for mean prey proportions between scats incorporated the actual
percentage of prey in the diet for all of the fish prey, but not for Nototodarus spp. or
Penaidae (Fig.3.5). When considering averages across individual dolphins the relative
importance of prey items was Non-Test species, followed by M. cephalus, S.australasicus,
A. georgianus and Nototodarus spp. although the differences between M. cephalus and
S.australasicus averages were small (Fig. 3.5). Arithmetic confidence intervals for mean
prey proportions between individuals incorporated the actual percentage of the prey in
the diet for all prey items detected (i.e. all but Penaidae, Fig. 3.5). For between scat and
between individual averages, bootstrapped estimates of 95% confidence intervals of the
mean and of the 2.5 and 97.5 percentiles of 95% confidence intervals were very wide,
indicating that the process of sampling individual sample clone libraries adds
considerable variability to these estimates (Fig 3.5).
When all clone libraries were pooled and absolute prey proportions estimated, the
relative importance of prey matched the weighted FOC approach, although the
53
differences between S.australasicus, M. cephalus and A. georgianus were the smallest of
any of the above approaches (Fig. 3.5). Confidence intervals for pooled sample absolute
proportions incorporated the actual percentage of the prey in the diet for
S.australasicus, M. cephalus and A. georgianus but not for other prey groups (Fig.3.5).
Figure 3.5 Actual Proportion of prey in SWBD diet (Dark Grey horizontal lines) compared to proportions of amplicons in clone libraries calculated by different means. 1) Prey proportions averaged across individual scat samples (circles), 2) Prey proportions averaged across individual dolphins (squares), and 3) absolute total proportion (triangles) by pooling all prey amplicons in clone libraries. Whiskers represent the bootstrap 2.5 and 97.5 percentiles for 95% confidence intervals (light grey lines) and 95% confidence intervals for the mean (red lines) due to multinomial sampling error. Symbols represent arithmetic mean between groups (circles and squares) or total prey proportion (triangle). For comparison, arithmetic 95% confidence intervals of arithmetic mean proportions of raw data are given (dotted lines). Prey species as follows: Sa: Scomber australasicus, Ag: Arripis georgianus, Mc: Mugil cephalus, N: Nototodarus spp., P: Penaeidae spp
54
3.3.4 SWBD Mugil cephalus DNA assays
3.3.4.1 First and last detection of Mugil cephalus DNA
There were two independent positive detections of Mugil cephalus DNA 2.83 and 5.08
hours after the potential first ingestion of this prey ≈9am. All other scats produced on
the first day of the ‘Test’ period were of ambiguous origin.
The last independent detections of the prey signal were at 11:15am and 2:10pm the day
after the test diet ceased. Since the last possible feeding time for this prey was≈ 4pm
the day before, this suggests that prey signals can persist for a minimum of 19.25 –
22.16 hours after prey ingestion. Assuming the final signal was from prey consumed the
day before and given the first feeding time ≈9am these final prey detections could also
represent prey ingested up to 26.25 – 29.16 prior. Similar to minimum time to detection,
it is not possible to estimate the maximum prey signal persistence time, since positive
scats may have been produced, but not collected, after the last positive scat was
collected.
3.3.4.2. Positive samples from first to last prey detection
After the first prey detection between 33 – 80% of samples per day tested positive until
the last detection of the prey item (fig. 3.6). Of the 68 samples between the first and last
positive sample (inclusive) 60% were positive for the prey item. The bootstrap estimate
of the variance of the proportion positive was 6% (i.e. 60 +/- 11.9% (95% confidence
intervals); Fig. 3.7).
3.3.4.3. Effects of sex and sample volume on Mugil cephalus DNA detections
There were 42 samples collected between the first and last prey detection where the
animal producing the scat was certain. In an attempt to account for the serial
dependence of samples from the same individual, a mixed-effects logistic model was fit
to these data to examine the effects of sex and sample volume on Mugil cephalus assay
success (response: presence/absence of Mugil cephalus DNA) with individual as a
random effect. With an average of 5 +/- 3 (SD; range 1-10) samples per individual (n = 10)
data were too sparse to model (over dispersion). Ignoring serial dependence from
individual dolphins, pooling results from first to last prey detection in volume categories
resulted in data too sparse for robust contingency table tests, with expected cell counts
falling below 5 (Zar, 1999). Pooling data within sex, there was no significant difference in
Mugil cephalus DNA detection between sexes (X2=1.8, P = >0.1).
55
Figure 3.6 Samples that tested positive for Mugil cephalus DNA per day of the feeding trial (sequential days from left to right). For the first day, only samples collected after the first positive sample were included in this figure. Excluding the first test day, numbers above x-axis are total sample numbers for that day. Data are black dots with actual value to the right of the data point. Data are joined by lowess smoothed line.
Figure 3.7 Boxplot with overlaid probability density function estimated with kernel density estimates of proportion of positive samples from bootstrap replicates (n = 1000) from first to last prey detection. Filled red line is the actual proportion of positive samples over the duration of the prey detection period; dotted red lines are the bootstrap 95% confidence intervals of the proportion of positive samples calculated from the bootstrap variance estimate.
1 4 7 7 12 3 4 5 8 5 9 8 8 6 7 0 0
0.71
0.57
0.50
0.33
0.50
0.80 0.75
0.60 0.56
0.63
0 0 0 0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
NT NT Test Test Test Test Test Test Test Test Test NT NT NT NT
Pro
po
rtio
n p
osi
tive
Day of Trial from left to right
56
3.3.4.4. Evidence for prey pulse signals from Universal and Q-PCR assays
For samples assayed with the universal technique, there were three days where
sequential scats were collected from the same individual (Table 3.3). The first of these
consisted of three scats from a female (Gemma) where 0, 3 and 5 prey species were
detected from sequential scats 76, 79 and 81 respectively (Table 3.3).The second scat
was passed 2.5 hours after the first and the third 1.58 hours after the second with all of
the prey in the second scat detected in the third. The second day there were three scats
(97, 99, 102) from a male (Stormy) were passed with 2, 3 and 4 prey species detected
sequentially. The second scat was passed 1.3 hours after the first and the third 4.3 hours
after the second with one species from the first scat not detected in the second and all
species from the second detected in the third. The third day there were two scats (112,
118) from a male (Stormy) were passed with 3 and 1 prey species detected sequentially.
The second scat was passed 1.8 hours after the first and the single species detected in
the second scat was also detected in the first. In summary, there was substantial
variation in prey detection between sequential scats from the same individual indicating
prey DNA pulse signals: Both the appearance and disappearance of prey signals can
occur in sequential scats within 1.58 and 1.3 hours respectively.
For the Q-PCR results, there were 4 scenarios to consider the pulse signal of Mugil
cephalus DNA in sequential scats from the same individual: two no transition (i.e. no
detection -> no detection, or detection -> detection) and two transition (i.e. no
detection -> detection, or detection -> no detection). Three individuals showed no
transition passages (Gemma, four no detection -> no detection passages between 17m –
5.3 hours, Moki and Nila, one detection -> detection passage each, 1.58 and 3.7 hours
respectively) and one individual displayed a no detection -> detection transition passage
of 4.3 hours.
3.4 Sarasota Bay dolphin Universal Assays
A total of 485 clones were screened from 24 clone libraries, with 15 – 38 clones
screened per clone library. At least one prey species was detected from all samples
(Table 3.6), except one fecal sample which had a substantially lower volume than the
others (Table 3.2). The different bioinformatics routines used for prey identification in
this chapter yielded, in some cases, substantially different results compared to each
other and Chapter 2. For sequence quality control, in the case where no chimeras were
detected using CCode software, two chimeras were identified with ChimeraSlayer
software. As chimeras are not recognised with CCode software, there was an artificial
inflation of diet diversity estimates for the individual sample used in chapter 2. In both
samples, the chimeras consisted of different amplicon ends if two different prey species
identified from the sample, which may confound the chimera detection strategy used
CCode software. The sequences of predator origin identified appear to be NUMTs and
are the subject of Chapter 4, so they are not discussed in detail in this chapter.
57
3.4.1 Prey Assignment by BLAST threshold and Placement in Neighbour Joining
Analysis
There were 25 operational taxonomic units (OTUs) identified using the ‘percent
threshold/neighbour joining analysis’, with 20 OTUs assigned to species level
identifications and 13 of these displaying at least one OTU at a 100% match to GenBank
nucleotide records for that species (Table 3.6). There were an average of 3.2 +/- 1.7 (S.D.;
range 0-7) prey OTU’s identified per individual, with an average of 3.1 +/- 1.4 (S.D.;
range 0-7) OTU’s identified per scat sample and an average of 2.7 +/- 0.4 (S.D.; range 2-3)
per gastric sample (Table 3.6).
Table 3.6 Summary of OTUs identified from Sarasota Bay T. truncatus faeces and gastric samples using a Threshold/Topological Neighbour Joining Analysis
Prey OTU Identified (# of unique sequences)
FOC SBD
10* SBD 113
+
SBD 133
SBD 146
SBD 151
+
SBD 155
+
SBD 157
+
SBD 164
+
SBD 179
+
SBD 182^
SBD 195*
SBD 199
+
SBD 20
+
SBD 234*
SBD 236^
SBD 238
+
SBD 240
+
SBD 246*
SBD 90
Archosargus probatocephalus 1/19 10
Brevoortia patronus 2/19 1* 5 Centropomus undecimalis 1/19 7* Chaetodipterus faber 1/19 3 Cynoscion nebulosus 2/19 1 6
+*
Cynoscion sp. 1/19 2*
Elops saurus 4/19 4 4 2 11+
Epinephelus itajara 1/19 1
Gerreidae sp. 1/19 1 Hemiramphus brasiliensis 1/19 1
+
Lagodon rhomboids 12/19 18 31+* 25
+* 9 1 6 16 29
+* 2
+ 6 10 9
+*
Leiostomus xanthurus 9/19 10+* 1 1 4 4
+* 11 17
+* 3 3
+
Menticirrhus littoralis 1/19 7
Microgobius gulosus 1/19 2 Mugil cephalus 2/19 2
+ 1
Mugil gyrans/curema 3/19 2 2 2
Mugilidae spp** 1/19 1
Opisthonema oglinum 1/19 1 Opsanus beta 6/19 11 1
+ 2 9 5 9
+
Opsanus spp** 2/19 1 1+
Orthopristis chrysoptera 1/19 3 Paralichthys albigutta 1/19 3
+*
Prionotus sp. 2/19 1 2 Sardinella aurita 1/19 1 Sphoeroides parvus 4/19 1 6
+ 1 3
Predator Origin Clones 0 3 3 1 4 1 9 12 4 1 26 19 15 1 1 13 11 0 6
Total Clones Screened 20 19 40 41 19 20 20 20 20 37 38 19 20 20 39 19 19 20 35
OTU identifications 3 3 2 6 4 7 3 2 1 3 3 0 3 3 5 3 2 4 5
Indicates 100% match between at least one prey sequence and GenBank record; + Fecal Sample only; * Gastric Sample Only;
+* Both Fecal and Gastric
Sample; ^ Single clone variant in library chimera as identified with ChimeraSlayer.
59
3.4.2 Prey Assignment Using the MOTU/Statistical Assignment Package method
There were 53 unique sequences identified across all prey sequences which sorted to
28 taxonomic assignments consisting of 32 different MOTU’s using the statistical
assignment package (SAP): 13 MOTU’s were assigned to species due to 100% agreement
between the MOTU and database records and 5 MOTU’s were assigned to species level
by SAP (Table 3.7). Different MOTU’s were maintained on the basis of similarity to each
other allowing for Taq DNA polymerase error (See Table 3.7, Cynoscion sp., Scianidae,
Elops sp. and Opsanus sp.), and if below the ≥ 2 substitutions Taq error threshold, on the
basis of identical MOTUs in different samples (See Table 3.7, Scianidae 2 and Scianidae
3). There were an average of 3.7 +/- 2.2 (S.D.; range 0-9) prey MOTU’s identified per
individual, with an average of 3.3 +/- 1.8 (S.D.; range 0-9) MOTU’s identified per scat
sample and an average of 2.8 +/- 1.2 (S.D.; range 2-3) per gastric sample.
For some taxonomic groups there were few options for potential prey with distributions
that range into the Sarasota Bay area. For example, samples assigned to Elops sp. could
be only 1 of 2 species that are known from the Sarasota region; Elops saurus and Elops
smithi. The latter of this species is indistinguishable from the former without performing
counts of vertebrae. For the Lagodon rhomboides MOTU, some of the unique sequences
were assigned to Lagodon sp. in SAP, yet only Lagodon rhomboides occurs in the study
region. For the Opsanus sp. MOTU, although this sequence was substantially different
from Opsanus beta sequence records, Opsanus beta is the only Opsanus species known
to inhabit Gulf of Mexico waters. Considering the Scianidae assignments, there are 21
Scianidae species known to inhabit the Sarasota Bay region and 13 of these have
nucleotide records available on GenBank. It may therefore be the case that the
Scianidae MOTU’s represent one or more of the eight Scianidae sp. not represented on
GenBank. There is some difficultly in interpreting the Cynoscion sp. MOTU, since all
Cynoscion spp. known from the study area have multiple sequence records represented
on GenBank. It is noteworthy that for both the Opsanus sp and the Cynoscion sp.
MOTU’s which both have assignment interpretation difficulty, there are sequences of a
closely related MOTU, confidently identified to species within the same sample. It is
possible that these MOTUs are therefore experimental artefacts of the PCR process
outside of the normal bounds of Taq DNA polymerase error (e.g. prey mtDNA
hetroplasmy).
3.4.3 Comparison of different prey taxonomic assignment techniques
The difference in results between the two taxonomic assignment methods used on
recovered prey sequences originates largely from considering all sequences together
and accounting for Taq DNA polymerase error, as opposed to applying blanket cut-off
thresholds for each sequence when comparing it to database (i.e. GenBank) records. A
case in point is the different Scianidae MOTUs present in the SAP assignments, most
similar to Leiostomus xanthurus (Table 3.7). Using the threshold/neighbour joining
assignment method, all of these MOTUs grouped with Leiostomus xanthurus, since they
60
were all within 98% identity to this prey item (Table 3.6). By contrast using SAP and
accounting for Taq DNA polymerase error and multiple records from different samples,
these same sequences grouped into four MOTU’s which are probably more
representative of diet, given that 44% of the possible prey species in the family these
MOTUs could represent are not present in databases and intra-specific sampling in
databases is also low (i.e. < 3 separate records for all except Cynoscion spp.). Regardless
of whether Genus or familial level MOTU’s represent different prey species or intra-
specific variation, by representing them as MOTU’s they allow for the possibility they
could be different prey species or at least present a biological component of diet
diversity that can be compared between samples as a prey species might be. For this
reason further summaries of prey data from Sarasota Bay samples are presented with
the MOTU/SAP assignment method results.
61
...This page is intentionally blank...
Table 3.7 Summary of Sarasota Bay T. truncatus number of clones per individual taxonomically assigned to prey MOTUs using SAP software and accounting for Taq DNA polymerase error in clone libraries. If Genus and Family level identifications were > 95% similar to another identified taxon, they are listed below the most similar (% identity) species level identification and indented. Blue text indicated a difference in prey assignment between the ‘% Threshold/Neighbour joining’ method and the MOTU/SAP method.
SBD Number
Prey MOTU Taxonomic Assignment (Number of unique sequences if >1)
FOC 10* 20+ 90 113
+ 133 146 151
+ 155
+ 157
+ 164
+ 179
+ 182^ 195* 199
+ 234* 236^ 238
+ 240
+ 246*
Archosargus probatocephalus 1/19 10
Brevoortia patronus – (2) 2/19 1* 5
Centropomus undecimalis 1/19 7*
Chaetodipterus faber 1/19 3
Cynoscion nebulosus -- 2/19 1 6+*
Cynoscion sp. 1 1/19 1*
Scianidae 1 1/19 1*
Elops saurus (5) -- 4/19 1 3 4 1+
Elops sp. 1 1/19 1
Elops sp. 2 1/19 10+
Elops sp. 3 1/19 1
Epinephelus itajara 1/19 1
Eucinostomus sp. %
1/19 1
Haemulidae 1/19 3
Hemiramphus brasiliensis 1/19 1+
Lagodon rhomboides (9) -- 12/19 18 9+* 31
+* 25
+* 9 1 6 16 29
+* 2
+ 6 10
Leiostomus xanthurus (4) 6/19 3+ 7
+* 4 2
+ 12* 2
Scianidae 2 4/19 3* 1 11 1
Scianidae 3 (2) 3/19 1 2* 4+
Scianidae 4 1/19 1*
Menticirrhus littoralis 1/19 7
Microgobius sp. 1/19 2
Table 3.7 continued. SBD Number
Prey MOTU Taxonomic Assignment (Number of unique sequences if >1)
FOC 10* 20+ 90 113
+ 133 146 151
+ 155
+ 157
+ 164
+ 179
+ 182^ 195* 199
+ 234* 236^ 238
+ 240
+ 246*
Mugil sp. %
1/19 1
Mugil cephalus 2/19 2+ 1
Mugil gyrans/crema 3/19 2 2 2
Opisthonema sp. 1/19 1
Opsanus beta (4) -- 6/19 9+ 11 1
+ 2 9 5
Opsanus 1 %
2/19 1+ 1
Paralichthys albigutta 1/19 3+*
Prionotus sp. (2) 2/19 1 2
Sardinella aurita 1/19 1
Sphoeroides parvus 4/19 1 6+ 1 3
Total Prey MOTUs Identified 3 4 5 4 2 7 4 7 3 2 1 4 3 0 3 9 4 2 4
Total Predator 0 15 6 3 3 1 4 1 9 12 4 1 26 19 1 1 13 11 0
Total Clones screened 20 20 35 19 40 41 19 20 20 20 20 37 38 19 20 39 19 19 20
+ Fecal Sample only; * Gastric Sample Only;
+* Both Fecal and Gastric Sample;
^ Single clone variant in library chimera as identified with ChimeraSlayer
64
3.4.4 Occurrence and Inferred Relative Importance of Sarasota Bay dolphin prey
Simple frequency of occurrence (FOC) placed Lagodon rhomboides as the most
important prey item (12/19) followed by Opsanus beta and Leiostomus xanthurus (6/19),
Sphoeroides parvus, Scianidae 2 and Elops saurus/smithi (4/19), Scianidae 3 and Mugil
gyrans/crema (3/19), Prionotus sp., Mugil cephalus, Cynoscion nebulosus and Brevoortia
patronus (2/19) with all other prey items occurring in only one individual (Table 3.7). At
the familial level FOC placed Sparidae as the most important prey family (12/19)
followed by Scianidae (10/19), Batrachoididae (6/19), Mugilidae (5/19), Clupeidae,
Elopidae and Tetraodontidae (4/19) and Triglidae 2/19 with all other families occurring
in only one individual.
A weighted FOC score was calculated for each prey group where FOC was normalised to
the prevalence of clones of each prey species as a proportion of all prey clones across all
samples. By this scale all species level FOC assignments remained similar as basic FOC
rankings except for relative differences in each FOC ranking category (Table 3.8).
Weighed FOC of prey families remained similar to basic FOC, except Elopidae ranked
above Muligidae, and as with species, there were relative differences in each FOC
ranking category (Table 3.8).
Averaged prey amplicon proportions indicate that Lagodon rhomboides amplicons were
considerably more prevalent in clone libraries compared to any other species (Fig. 3.8).
Other prey species amplicon proportions show the same trend of relative prey
prevalence as weighted FOC scores, however there was considerable overlap in the 95%
confidence intervals of mean amplicon proportions for all prey species other than
Lagodon rhomboides (Fig. 3.8). Similarly, at the familial level, Sparidae amplicons were
considerably more prevalent in clone libraries compared to any other prey family.
Scianidae and Batrachoididae amplicons were the next most prevalent prey families at
an average of 20% and 11.2% of amplicons across individuals, respectively. All other
prey families were represented by on average <5.8% of prey amplicons with
considerable overlap in 95% confidence intervals.
3.4.5 Comparison Sarasota Bay dolphin diet estimated via DNA and stomach
contents data
Comparing data from this study to the six most frequenctly occurring prey taxa in Barros
& Wells (1998), FOC values were similar between the two studies with all but one
(Orthopristis crysoptera) within 10% of each other (Fig 3.9). Confidence intervals are
wide in both studies due to the low number of total samples analysed and show
considerable overlap in all species compared (Fig. 3.9). Averages of numerical
proportions of total prey in each stomach in the 12 most numerically dominant prey
from Berens et al. (2010) were also compared with data from Barros & Wells (1998) and
average total proportion of clones from these prey in this study (Fig. 3.10).
65
Table 3.8 Species and family Frequency of Occurrence (FOC) and FOC scores for taxa appearing in more than one sample. Taxa are grouped in descending order of weighted frequency of occurrence score.
Prey Group FOC Total Clones (%) Weighted FOC Score
Species
Lagodon rhomboides 12/19 43.3 0.273
Opsanus beta 6/19 9.9 0.031
Leiostomus xanthurus 6/19 8.0 0.025
Elops saurus/smithi 4/19 5.9 0.012
Scianidae 2 4/19 4.3 0.009
Sphoeroides parvus 4/19 2.9 0.006
Scianidae 3 3/19 1.9 0.003
Mugil gyrans/crema 3/19 1.6 0.003
Cynoscion nebulosus 2/19 1.9 0.002
Brevoortia patronus 2/19 1.6 0.002
Prionotus sp 2/19 0.8 0.001
Mugil cephalus 2/19 0.8 0.001
Family
Sparidae 12/19 48.5 0.306
Scianidae 10/19 19.7 0.104
Batrachoididae 6/19 11.0 0.035
Elopidae 4/19 5.9 0.012
Mugilidae 5/19 2.8 0.007
Tetraodontidae 4/19 3.1 0.007
Clupeidae 4/19 2.3 0.005
Triglidae 2/19 0.8 0.001
Of the 12 most numerically dominant prey from Berens et al. (2010), at least 9 and up to
11 taxa were also identified as prey in this study (range given since only inclusive higher
taxa were identified for some taxa in either study) and there were 8 common items with
Barros & Wells (1998) (Fig 3.10). Measures of variation for mean proportions were not
given in Berens et al. (2010), but were calculated from data provided in Barros and Wells
(1998) and in this study via the arithmetic mean of prey clone proportions across all
samples. In 4 cases the 95% confidence intervals for estimates of means from this study
encompassed mean estimates provided in Berens et al. (2010). In all instances
confidence intervals overlapped between Barros & Wells (1998) data and this study. The
mean diversity of prey items per individual with prey present was higher in this study
compared to Barros & Wells (1998): 3.9 +/- 2 (+/- S.D; range 1-9) and 2.4 +/- 1.4 (+/- S.D;
range 1-6), respectively, although estimates from this study could be overinflated if
MOTUs represent intra-specific variation. This information was not provided in Berens
et al. (2010). In terms of all taxa identified, Berens et al (2010) identified 22 species from
13 families, Barros & Wells (1998) identified 15 species from 10 families (and one
elasmobranch family) and this study identified at least 23 different species from 16
families. Of the species level identifications from this study, 5 species (Chaetodipterus
faber, Epinephelus itajara, Hemiramphus brasiliensis, Paralichthys albigutta and
Sphoeroides parvus) have not been found in the above prey identification studies for this
population.
66
Figure 3.8 Average taxon (top panel: species; bottom panel: family) amplicons per clone library. Filled circles represent arithmetic average across individuals, open circles represent proportion of pooled total prey amplicons. Whiskers are 95% confidence intervals for 1) population arithmetic mean (dotted lines of filled circles); 2) bootstrapped 95% confidence intervals for mean due to sampling error in each clone library (red line of filled circles) and 3) bootstrapped 95% confidence intervals for proportion of pooled total prey amplicons due to sampling error in each clone library (filled lines of open circles)
67
Figure 3.9 Absolute frequency of occurrence (sensu Wright, 2010) of six most common prey species in Barros & Wells (1998) compared to this study. Whiskers represent exact 95% binomial confidence intervals.
Figure 3.10 Average numerical proportion of prey items across all samples (other studies) and average proportion of prey item clones across all sample (this study). Whiskers represent truncated 95% confidence intervals of arithmetic mean. Variance was not available for Berens et al. (2010).
0
10
20
30
40
50
60
70
80
90
100
Lagodonrhomboides
Leiostomusxanthurus
Opsanus beta Mugil spp. Orthopristiscrysoptera
Elops spp
Ab
solu
te F
req
ue
ncy
of
Occ
ura
nce
(%
)
Barros & Wells (1998), n = 16
This Study, n=19
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Ave
rage
Nu
me
rica
l Pro
po
rtio
n o
f to
tal P
rey/
Clo
ne
s
Berens et al. (2010) n = 15
This Study
Barro & Wells (1998) n = 16
68
3.5 Discussion
3.5.1 Captive Feeding trials and SWBD prey detection results
Field based trophic ecology studies generally aim to determine population level diet
characteristics in the context of biologically relevant variation, such as spatial, seasonal
or ontogenetic factors. In such studies it is assumed that one or few scats are collected
from many individuals, in contrast to feeding trial experiments, where many scats are
collected from few individuals (Casper et al. 2007). This inevitably limits vertebrate
captive feeding trials, such as the feeding trial in this study. It is difficult to justify
treatment of each scat/DNA extraction from the same animal in feeding trials as
independent samples (e.g. Deagle et al. 2005; Deagle and Tollit 2006; Bowles et al. 2011;
OEHM et al. 2011), since previous studies have shown meal specific pulse signals for
particular prey items (Deagle et al. 2005; Casper et al. 2007). This implies time-series
correlation structure of scat derived DNA-diet data, which invalidates many traditional
statistical methodologies that require independence of samples (Zuur et al. 2009).
Although the nature of serial dependence is also of interest in terms of DNA gut passage
times and signal persistence, the constraints this places on inferences about results are
compounded by potential differences in digestive physiology between captive and real
world environments (Casper et al. 2007). These factors in combination with the under-
representation of inter-individual digestive variability (i.e. small n of different trial
subjects), regardless of potential environmental biases, and potential differences in pre-
ingestion DNA quality between live (i.e. predated) and dead (i.e. most items fed in
captive trial studies) prey, must temper the context of interpretation of feeding trial
data for DNA-based diet methods.
The SWBD feeding trial suffered from some of the constraints mentioned above, but
also had some unique limitations. Although the number of subjects was higher than
most feeding trials, the ability to identify them was poor: only ≈ 62% of samples
between first and last detection of Mugil cephalus DNA could be confidently assigned to
one individual. This resulted in small sample sizes which were not suitable for statistical
analysis. Other constraints were 1) that not all scats produced could be collected; 2)
judging by the fecal volumes collected from wild animals, only a small proportion of any
single scat was collected, and; 3) The times of ingestion of any particular prey item were
not recorded but were variable throughout the trial. Points 1 & 2 are in contrast to most
other marine mammal feeding trials, which were able to collect the majority of every
scat produced (e.g. Deagle et al. 2005; Casper et al. 2007; Bowles et al. 2011) and in
some cases, know the identity of the animals producing the scat (e.g. Deagle et al. 2005;
Casper et al. 2007). Indeed, the differences between the nature of the scats and the
time spent out of water between pinnipeds and cetaceans mean that comparisons
between this work and pinniped feeding trials may be misguided. The results of time
series scats from this study (3.3.4.1) also indicate prey (or meal) specific DNA signal
pulses. Taken together, the implication of these points for this study is that initial
detection and signal persistence cannot be ascertained from this experimental design. It
69
is also not possible to ascertain whether failure to detect prey items was due to lack
sensitivity of the molecular method (i.e. a false negative) or whether prey DNA was not
present to be detected (either in the entire bowel movement or in the portion of scat
collected). Treating the samples collected as a sample of all scats produced in the study
period, the detection rate of this study (60 +/- 12 %) was considerably lower than Deagle
et al (2005) and Parsons et al. (2005) and similar to the median value from three prey
species in Casper et al. (2005). Most samples collected in this feeding trial were small
volume, and sample volume appeared to be related to the proportion of prey amplicons
in clone libraries (Fig. 3.4). Deagle et al. (2005) found that prey DNA detection was less
consistent in small parts of individual pinniped scats as opposed to homogenized whole
scats and Oehm et al. (2011) also found that DNA extracted from relatively larger
volumes of avian scat had higher rates of prey detection than DNA extractions
performed on smaller homogenised samples. Another consideration is that these
samples were analysed upwards of 2.5 years after they were collected, which may not
be ideal for samples of marginal DNA quality to begin with.
With the exception of penaeid prey, the diet composition of SWBD samples was
accurately qualitatively defined using the universal technique. Penaeid DNA may not
have been detected due to a number of factors, including more severe DNA degradation
in dead prey pre-ingestion than other species, differential survival of prey species DNA
in the digestive tract, the fact that the item was not favoured by subjects and sometimes
found after feeding events in the water column (R. Casper pers. comm) or simply due to
variability in prey detection due to prey signal pulses coupled with the small overall
dietary proportion (2% wet weight) of the item. At the individual level it seems this
technique cannot inform as to all prey items ingested over the previous 1-2 day period.
However, at the population level where sample results are combined the technique
performed adequately for describing the major dietary components.
Like other studies, this work found a reasonable agreement between the weight of
proportions of prey ingested and the proportions of DNA signals from prey detected in
scats (Deagle et al. 2005; Deagle and Tollit 2006; Nejstgaard et al. 2008; Bowles et al.
2011). Where samples were analysed grouped by individuals or total pooled clone
libraries, the average estimated proportion of each prey group was within 11% of the
actual dietary proportion and in the case of grouping by individual dolphins, the 95%
confidence intervals for the arithmetic means overlapped prey wet weight proportions
for detected items. In comparison to grouping by scats, grouping data by individuals or
pooling is likely to be more representative of a field situation, where few scats are
collected from many individuals. Furthermore, individually grouped clone proportion
confidence intervals overlapped for items fed at similar proportions (10-15% wet
weight), but not for ‘non test’ items fed at 43% of wet weight, indicating that reliable
relative information can be gained by comparisons of prey metrics and confidence
intervals. The frequency of occurrence (FOC) and weighted FOC values did not compare
well in this regard and only appeared to be useful for comparison with amplicon
proportions estimates. The boot-strapping approach used to represent variance in
estimated prey proportions and their confidence intervals due to clone library sampling
error demonstrated the large variance introduced by sampling individual scat amplicon
70
pools. An inevitable constraint of cloning techniques is the time and expense of
sequencing individual clones and the consequences of this have long been recognised
for quantifying diversity in mixed (i.e. diet, environmental) samples (Hughes et al. 2001;
Dunbar et al. 2002). With the advent of next generation sequencing techniques such as
454 sequencing, there is no longer a requirement to clone samples to identify and
quantify amplicon diversity (e.g. Deagle et al. 2009). In addition to comparative ease of
this procedure, sample sizes of amplicon pools number in the thousand to hundreds of
thousands. As well as increasing the likelihood of detecting rare items, such large
samples of amplicon pools will also decrease variance in amplicon proportion estimates
due to sampling error. Despite this, it is still recommended that re-sampling or diversity
statistical summary procedures (e.g. Chao and Lee 1992) are implemented to
understand the impact of sampling error on study metrics and conclusions.
The feeding regime employed by this feeding trial is unlikely to represent any
experienced by wild animals, since all subjects were fed the same proportions of prey
each day. For extrapolation of results to field studies, it may be more appropriate for
captive studies to vary individual meals randomly each day and have the overall dietary
composition matching pre-defined proportions in feeding trials. In field studies the
individual scat is the sampling unit and it is the sample of these that is summarised per
group, rather than comparisons of individual scats themselves (Wright 2010). It may
therefore be more realistic to vary feeding trial diets per meal per individual, as this is
certainly more likely to be a realistic representation of the diet of wild animals. In any
case, it is also clear that cetacean feeding trials have their own unique set of practical
limitations, in that only small proportions of any scat can be collected and it is not
possible to collect all scats. Despite the caveats of this experimental design and the
inability to rule out false negatives, most prey species were detected. The ramification
of these results for interpreting data from wild animals are that although false negatives
cannot be ruled out, it is certain that prey species detected were consumed and, with
the universal technique used in this study, there is quantitative signal in the data - albeit
with wide confidence intervals. This is a significant advance for gathering trophic
information from these types of samples, as this was not previously possible at all.
3.5.2 Sarasota bay data
The application of the universal technique to Sarasota Bay samples yielded species level
prey information from all but one sample, but the bioinformatics routines implemented
to identify prey from DNA sequences yielded quite different results. An important issue
for using DNA sequence data to identify organisms is the effect of incomplete databases
on identifications (Nielsen and Matz 2006). While statistical procedures can be applied
to existing datasets to assign probabilities of inclusion to represented taxa (Nielsen and
Matz 2006; Munch et al. 2008b; Munch et al. 2008c), it is not possible to assign DNA
sequences to the correct taxa when they are not represented in databases (Munch et al.
2008a). Applying blanket similarity threshold cut-offs to existing data is appealing due to
the ease of implementation; however it is not representative of biological reality.
Implicit in a threshold value approach for taxonomic identity is that the cut-off value
71
incorporates both experimental error, such as Taq DNA polymerase error (in the case of
cloning), and intra-specific diversity. While the likelihood of experimental error fitting
into pre-defined dissimilarity thresholds may be appropriate, it is not likely these
adequately cope with variation in intra-specific diversity (Will and Rubinoff 2004).
Dissimilarity thresholds alone will therefore limit false positive results, but not
misidentifications (Ross et al. 2008), furthermore they cannot assign probabilities to
identifications. In essence a threshold approach was employed within the SAP
framework, since Taq DNA polymerase error was accounted for using an amplicon
length specific dissimilarity cut-off. Just as important to the final results were retaining
sequences that were identical between samples but which were under the threshold
cut-off value. These sequences likely represent common aspects of prey mtDNA
diversity. Prey mtDNA heteroplasmy may be responsible for these sequences,
particularly where a highly similar prey sequence was positively identified to species
level in the same sample. It is difficult to estimate the likelihood that this is the case
given the paucity of data on heteroplasmic mtDNA copy number variation in wild
populations. The SAP approach is appealing because it assigns a robust probability of
taxonomic inclusion based on Bayesian phylogenetic analysis. It therefore uses existing
data in a phylogenetic framework, instead of speculating on appropriate similarity
thresholds where there is inadequate sampling within and between species, as is likely
to be the case for most DNA-diet studies in non-model systems that use online
databases of reference sequences.
Failing bioinformatic assignment errors, there is the possibility of false positive results
from DNA sequences amplified from scat samples which merit consideration. False
positives may arise from, 1) environmental DNA contamination or cross-contamination
during the DNA extraction process, 2) Chimeric DNA formation in vitro, 3) mtDNA
heteroplasmy (as discussed above), 4) Nuclear mitochondrial pseudogenes (NUMTs -see
chapter 4) and 5) Where prey consumed by other prey and not the predator itself (i.e.
‘secondary predation’) are detected. Cross-contamination during DNA extraction,
chimeric DNA formation in vitro and NUMTs were all explicitly considered with
experimental controls and analysis methods. Environmental contamination may have
occurred for the scat samples, but seems unlikely for the gastric samples given that
these pass straight from a gastric collection tube to a sample vial. All but three scat
samples were collected aboard the research vessel, in most cases straight from the anus
of the individual and thus were unlikely to be contaminated. The three samples
collected in the water column yielded results qualitatively similar to vessel-collected
samples in terms of prey species detected, so it is not considered that they were greatly
affected by environmental contamination. Given the lack of detection of extraneous
DNA in the captive samples analysed (either environmental or secondary predation) it is
considered unlikely that this significantly affected results of the Sarasota Bay samples.
This position is congruent with the strong similarities between the DNA data and
previous stomach contents data from this population. However, given the different
circumstances of the captive and wild sample sets, false positives cannot be definitively
ruled out from the Sarasota bay data. Thus novel results presented, such as the
72
identification of new prey species or rare prey only occurring in small proportions in one
sample, should be interpreted with caution.
The adequacy with which the universal method described SWBD diet diversity does
suggest that the Sarasota data are qualitatively representative of actual Sarasota
dolphin diet. The diversity of diet items identified in this study is high compared with
past studies of dolphin diet based on hard-parts analysis (Barros and Wells 1998; Berens
McCabe et al. 2010). New dietary items were identified which may represent rare prey
or potentially those with hard parts unable to survive the digestive environment of
dolphin alimentary tracts for long. The SWBD data also suggest that averages of clone
library proportions should be broadly indicative of the relative mass of prey consumed
and sample diet proportions. These, in combination with FOC data, indicate that
Lagodon rhomboides (the pinfish) constitutes an important prey species for this
population. This is in strong agreement with a past study on the foraging ecology of this
population (Barros and Wells 1998).
In fact, there is a striking degree of similarity between data from this study and all past
works on the foraging ecology of this population. This is the first time a population scale
species level diet description from live cetaceans has been produced unconstrained by
the biases of other diet analysis methods, allowing comparison of diet gathered directly
from live free-ranging animals and stranded individuals from the same population. Not
only were frequencies of occurrence similar between this study and Barros and Wells
(1998), but mean proportions of prey clones and mean numerical abundances were
similar between this study and data from Berens McCabe et al. (2010). The DNA data
was intermediate between the two past studies for five of the 12 prey species where
these metrics were compared. There was also greater overlap between prey species
composition from Berens McCabe et al. (2010) and this study compared to both the past
stomach contents studies. Mean proportions of prey clones and mean numerical
abundances may not be directly comparable. However, considered together with prey
species composition they indicate that the DNA data are intermediate between the two
past stomach contents studies, suggesting that stomach contents diet data accurately
reflects diet composition and variation of live individuals in this population.
Bottlenose dolphins exhibit regionalized foraging behaviours in response to the
distribution, abundance and behaviour of their prey (e.g. Smokler et al. 1997; Connor et
al. 2000; Torres and Read 2009; Allen et al. 2011). It is hypothesized that Sarasota Bay
dolphins intercept sound from soniferous prey and use this to locate prey(the 'passive
listening' hypothesis; Barros 1993). The DNA data supports this hypothesis, both through
broad agreement of important prey taxa with the past studies, and the prevalence of
obligate soniferous taxa in four of the top 8 prey taxa from this study, despite their low
relative abundance in this ecosystem (data from Berens McCabe et al. 2010).
73
3.5.3 Conclusion
DNA-based diet analysis techniques were able to detect prey DNA in faecal samples
from live captive and free-ranging bottlenose dolphins. Using a specifically designed
universal prey detection assay, all known prey items except one (comprising only 2% of
the wet weight of the diet) were detected in 12 faecal samples from 5 different captive
individuals. The proportions of prey DNA were broadly congruent with the wet weight
proportion of prey items fed to the captive animals, however there was substantial
variation in prey DNA proportion estimates introduced by sampling error within the
molecular methodology. The small volume of most samples influenced the efficacy of
prey DNA detection using this method. The first detection of prey was 2.8 hours after
initial feeding and the signal persisted for a minimum of 19.3 hours after feeding ceased.
Between the first and last positive sample, approximately 60% of samples were positive
for the prey item. The detection rates from the feeding trial differed substantially from
past studies, highlighting some of the practical difficulties of performing robust feeding
trial experiments with cetaceans. Prey detected from samples obtained from 19 wild
free-ranging individuals consisted of up to 32 taxa across at least 13 families of teleosts.
The nature of the bioinformatics analysis applied to the recovered prey sequences
substantially influenced results and a Bayesian phylogenetic technique yielding
statistical confidence metrics for taxonomic inclusion of prey sequences was preferred.
The applicability of DNA-based diet techniques in situations where gaining specific diet
via other means information is impossible was demonstrated by the applying the
method to free-ranging dolphins.There was remarkable agreement between genetic
assessment of diet n free-ranging dolphins and past studies based on stomach contents
analysis of stranded animals in the same population. These techniques appear to hold
great promise for the analysis of cetacean diet, if samples can be obtained.
74
Chapter Four –
Pseudogenes and DNA-based diet analyses: a cautionary tale from a relatively well sampled predator-prey system
&
Chapter 5 –
Telomeres as age markers for marine mammals
These chapters have been removed for
copyright or proprietary reasons
Published as :
Dunshea, G., N. B. Barros, R. S. Wells, N. J. Gales, M. Hindell, A. and S. N. Jarman (2008) Pseudogenes and DNA-based diet analyses; a cautionary tale from a relatively well sampled predator-prey system. Bulletin of Entomological Research 98: 239-248. Garde E, A.K. Frie, G Dunshea , SH Hansen, KM Kovacs, C Lydersen (2010) Harp seal ageing techniques-teeth, aspartic acid racemization, and telomere sequence analysis. Journal of Mammalogy, 91: 1365–1374. Dunshea, G., D. Duffield, N. Gales, M. Hindell, R. S. Wells and S. N. Jarman (2011). Telomeres as age markers in vertebrate molecular ecology. Molecular Ecology Resources, 11: 225-235.
113
– General Discussion Chapter 6
Marine mammals, and particularly those without terrestrial phases in their life histories,
are challenging to study. Primary questions for both basic and applied ecology, such as
the role of marine mammals in the structuring of ecosystems and their dynamics, for the
most part remain contentious (e.g. Wade et al. 2007; Springer et al. 2008). Similarly,
habitat requirements, use, and key resources are ill defined for many marine mammal
species and particularly cetaceans. This all stems from the fact that these animals spend
the vast majority of their time, underwater and far away from land. In addition to the
scientific value of such information, this paucity of knowledge affects our ability to
understand the nature and scale of marine mammal interactions with anthropogenic
activities, and the impacts of these. This thesis has been concerned with developing and
evaluating molecular genetic techniques to indirectly examine specific aspects of marine
mammal ecology. The main aim was to find ways to maximise the information that can
be gained from samples, when the rare opportunity arises to sample these animals.
6.1 DNA-based methods to study cetacean diet
The notion that an animal's diet can be determined by examining its faeces is not at all
new (e.g. Storr 1961), but the methods to do so have changed. Although procedurally
complex, the conceptual simplicity of the DNA-based diet method is that it requires only
DNA to partially survive the digestive process. This is also one of the advantages of this
technique compared to examining faeces visually, where recognisable diet remains are
required, as is the expertise to identify them. As opposed to the specialized skills,
experience and reference collections required for visual faecal analysis, workers with
molecular biological skills and appropriate facilities are practically a ubiquitous feature
of contemporary research institutions (Bowles et al. 2011). For many different animals,
some or all dietary items are digested beyond any visual recognition (Symondson 2002b).
Cetacean faeces fall into this category, and without genetic techniques contain little
dietary information. Other uses for cetacean faeces includes individual hormone profiles
(Rolland et al. 2005), parasitic prevalence (Aznar et al. 2011) and as a source of DNA for
the animal itself (Parsons et al. 1999; Green et al. 2007) . Although DNA-based diet
methods had been applied to cetacean faeces previously (Jarman et al. 2002; Jarman et
al. 2004), this is the first study to examine their efficacy systematically and apply them
to a wild population to address an outstanding question for cetacean stomach contents
analysis.
The results from the captive trials highlighted some of the technical as well as practical
difficulties of captive feeding trials for determining cetacean diet. Many of these
difficulties constrained interpretation of the feeding trial data, but they paradoxically
may also represent the ‘best case’ scenario for the majority of studies sampling wild
cetacean scats. It is rare that cetaceans are captured at all, and rarer still that faeces are
collected during this, as with the Sarasota Bay research program. It is more likely that
114
use of these techniques for studies of wild cetaceans will generally require sampling
from the water column. Sampling from the water column needs careful consideration of
appropriate control samples to test for the possibility of microscopic organism
contamination (King et al. 2008). While the aquarium water is filtered and UV sterilized,
making it unlikely that fish larval or environmental DNA contamination occurred, it is
possible that contamination from cellular debris from prey fed would contaminate
samples.
Although the issue of microscopic organism contamination has been raised previously
(King et al. 2008), it is not clear what constitutes appropriate control samples. Simply
collecting a water sample with the faecal sample has inherent assumptions about the
distribution of zooplankton, and it is not clear whether the presence or absence of an
item in a control sample should preclude the presence or absence of the same item in
the faecal sample. For example, how many water samples should be collected and in
what vicinity to the faecal sample? It is known that zooplankton distribution is random
and varies over small temporal and spatial scales (e.g. Jenkins 1988) and therefore it is
currently not clear what form appropriate control samples should take. Additionally, if
the predator under study feeds on zooplankton it is likely that their prey would be
present in the water column where samples are collected. Therefore control samples
would not be useful for differentiating prey from environmental contamination. It is
assumed in this study that the sequences and templates detected in faecal samples
were from ingested prey with further discussion of these results.
The captive feeding trial demonstrated that prey DNA fragments up to 270 bp can be
detected in dolphin faecal samples. The composition of these fragments within
individual samples did not represent the breadth of diet over the previous 1 – 2 day
period; however across all samples 9 of 10 prey items were detected. That
undetermined, but small, fractions of individual bowel movements were collected
probably strongly influenced prey detection rates within samples, given the pulse nature
of the DNA signal from individual meal items shown in this and other studies. Sample
volume did affect the efficacy of the universal method for detecting any prey DNA, it
therefore follows that it affected detecting all prey DNA, in addition to the pulse nature
of prey DNA signals. The proportions of prey DNA in samples estimated with the
universal technique and summarized across samples reflected proportions of some prey
items (particularly fish) ingested, albeit with large error due to the hierarchical nature of
the sampling processes within the analysis of samples. That is, sampling of bowel
movements during the day, sampling within bowel movements (i.e. material actually
collected) and sampling of amplicon pools produced from PCRs via sampling clone
libraries. Quantitative relationships of prey DNA proportions reflecting the mass of prey
ingested have been reasonably inconsistent across different controlled feeding studies,
but different molecular methods and summaries of data across samples have been
applied to each (Deagle et al. 2005b; Deagle and Tollit 2006; Nejstgaard et al. 2008;
Bowles et al. 2011). Averaging or pooling across samples likely reduces the effects of
prey DNA pulses in individual samples and individual predator heterogeneity in digestive
processes, in addition to decreasing chances of a type II statistical errors. Squid
proportion in the diet was consistently underestimated by any data summary method in
115
this study, which could be related to differences in mtDNA density/copy number
between fish and cephalopods. It is likely that more distantly related taxa are more
likely to have increasingly different mtDNA densities for any given mass of tissue, which
could explain the similarity of fish DNA proportions, as opposed to squid, to each other
and the mass ingested. This could be accounted for by developing correction factors for
tissue mass specific copy number differences between prey, as has been attempted
elsewhere (Bowles et al. 2011). For future studies on wild animals where correction
factors are not available, interpretation of prey DNA proportions seems problematic
without some a priori information, both on the molecular technique being employed
and on prey inter-specific copy number and detection rate variation. As a starting point,
comparisons may be best reserved by examining proportions within constrained
phylogenetic groups, although even this may be problematic. In their feeding trial,
Bowles et al. (2011) found that one fish in mixed diets of four fish species was
consistently underestimated despite application of correction factors. A reasonable
approach may be that workers should use prey DNA proportions as ‘working hypotheses’
and treat them with wide confidence intervals, until further information can be obtained
by molecular experimentation or other diet analysis techniques.
It may be reasonable to assume that the individual level data may not represent all prey
items consumed within the previous 1-2 days, but the summaries across all samples
adequately reflect the vast majority of dietary items consumed by weight in these
individuals. Although beyond the scope of this study, some simple modelling could
examine this under a variety of scenarios. For example, simulations of feeding trial data
can inform how many samples are required to detect specific prey in a multi-species diet
with certain probabilities of false negatives (e.g. Casper et al. 2007). This framework
could also be extended to determine minimum sample sizes required to detect, with a
given statistical power, diet differences between populations (see Trites and Joy 2005).
Since far greater volumes of faecal material were collected from Sarasota Bay animals, it
may be that individual results are more indicative of all prey consumed than for the
captive feeding trial samples. Regardless, these results are significant for two reasons: 1)
Excluding observational reports, they are the first description of the specific dietary
composition of a population of live free-ranging cetaceans, and, 2) They are the first
data to allow direct comparison of diet from stomach contents of dead individuals with
the diet of live free-ranging individual cetaceans from the same population. This
comparison has therefore provided a separate validation of the conclusions of past
studies regarding the diet and prey selection preferences of this population, and in
doing so has provided further support for the proposed mode of foraging hypothesized
for this population (i.e. the ‘passive listening hypothesis’).
While the universal diet technique developed and employed throughout this thesis was
powerful for detecting prey DNA for known and unknown situations, it had
complications. The first was that the technique could not exclude unknown predator
nuclear mitochondrial pseudogenes (numts). This was more of an issue for samples
from the feeding trial than for Sarasota Bay samples: numt’s were sequenced in 10 out
of 12 faecal samples from the captive feeding trial and 13 of the 15 Sarasota Bay faecal
116
samples. The origin of these sequences needed careful consideration in light of potential
prey and mammalian 16S mtDNA sequence diversity to be properly assigned as numts.
For cetaceans this type of analysis is likely to be more straightforward then for lesser
known predatory taxa due to the availability of extant cetacean mtDNA sequence data
and mammalian mtDNA sequence data. But the ramifications of numts on DNA-based
diet analyses will depend on the strategy employed to detect prey DNA. Where DNA
sequences are the end results of prey detection methods, spurious matches to database
sequences should be considered with the utmost caution and should be examined for
potential numt origin as a matter of course. Where prey DNA sequences are not the end
result of prey detection methods, predator or prey numt amplification and scoring can
confound these results and this requires careful consideration in the design and
validation of these methods.
Despite repeated pleas over the years for laboratory and analytical strategies that can
detect or avoid numts when using mtDNA markers in ecology and evolutionary biology
(Sorenson and Quinn 1998; Bensasson et al. 2001b; Thalmann et al. 2004b; Thalmann et
al. 2005; Triant and Dewoody 2007; Song et al. 2008), many studies employing these
markers are still being published with no mention of these (Morin et al. 2010). This is
also the case in the DNA-based diet literature (e.g. Bohmann et al. 2010) despite
recognition of the ramifications of errors in mtDNA-based taxonomic identifications as a
result of mtDNA heteroplasmy or numts (Rubinoff et al. 2006). Clearly this is an issue
that working molecular ecologists, publication authors, journal editors and article
reviewers should all be aware of to rigorously perform their roles, although it is
currently not clear that this is the case.
This work has provided guidelines and proof of the efficacy of these nascent trophic
ecology methods for studying cetacean diet and foraging ecology. Other methods
applicable to live cetaceans have particular advantages and disadvantages, dependent
on the timescale of interest and prey taxonomic resolution required. Some of the
characteristics, relative advantages and disadvantages of other diet analysis techniques
for live animals are given in Table 6.1. Stable isotope (SI) analysis can yield a variety of
ecological information on live animals such as diet and trophic level information, habitat
use and migratory patterns, pathways of food web contaminants and physiological
processes (Newsome et al. 2010). Generally, only skin and blubber can be practically
sampled from most live cetaceans, however some exceptions exist (Blood or teeth;
Wells et al. 2004; Barros et al. 2010 respectively). Although low in taxonomic resolution,
SI techniques offer a temporal window of days to years for samples from live animals,
dependent on the tissue used (Table 6.1). This wide range in timescales has obvious
advantages for long/large scale questions such as ontogentic dietary shifts, variation in
foraging strategies and the effects of inter-annual variation in foraging success on
growth and breeding. The complementary ecological information gained from SI analysis
is also appealing as it is often directly related to the foraging ecology of the focal species.
A recent derivative of SI analysis called compound-specific stable isotope analysis
characterises stable isotopes at the level of individual organic molecules (Newsome et al.
2010). This new technique has the potential to indicate dietary sources at the base of
food webs, allowing disentanglement of spatial and trophic/physiological processes in
117
isotopic signatures; an important issue for migratory marine mammals (Newsome et al.
2010).
Fatty acids signature analysis (FASA) and its derivatives (‘quantitative fatty acids
signature analysis’ or QFASA; Iverson et al. 2004) characterise the fatty acids present in
the lipids of animals. The basis for FASA is that for mono-gastric predators, fatty acids
consumed by predators are deposited without modification, allowing inference of their
dietary intake over time (Budge et al. 2006). However, this very basis is not without
controversy, with some authors believing that deposition of prey fatty acids in predator
tissue occurs with modification and is unpredictable (Grahl-Nielsen et al. 2004). Prior
information on modification and synthesis of particular FA can alleviate this problem,
but is difficult to obtain in many instances (Grahl-Nielsen et al. 2004). In principle,
QFASA provides quantitative estimates of specific prey items consumed, however this
requires extensive a priori characterisation of variation in prey fatty acids and predator
metabolism (Budge et al. 2006). This has rarely been performed as it requires extensive
experimentation, development and validation of complex statistical models (Iverson et
al. 2004). More typically, FASA has been used to qualitatively describe intra-specific
variation in diet (e.g. Bradshaw et al. 2003; Newland et al. 2009), prey-base
separation/overlap in sympatric predators (e.g. Iverson et al. 2007) and spatial and/or
temporal patterns in diet (e.g. Wang et al. 2009). The timescale over which FASA
represents diet, days to months, is dependent on the tissue analysed and the ecology of
the predator, although few estimates of this are available for specific predators (Budge
et al. 2006). Like stable isotopes, these techniques therefore yield longer term dietary
data (Table 6.1), however data interpretation can be constrained by a lack of prior
information on prey fatty acid variation, predator metabolism and the appropriate
timescale for consideration.
In comparison, DNA-based dietary analysis facilitates high resolution taxonomic
identification of prey, but over comparatively very short timescales (Table 6.1). Thus,
aside from technique specific constraints, the trade-off when considering the best
technique revolves around the timescale of the study question and the taxonomic
resolution of prey required. Clearly, if questions are centred around specific trophic
interactions then specific techniques (i.e. DNA or in some instances FASA) are required.
However, if questions centre on long term foraging strategies and their effects on other
variables, a longer time-frame, but less taxonomically informative indicator of trophic
processes may be more suitable. For questions of resource partitioning, competition or
spatial and temporal segregation of prey sources within and among predatory taxa, each
technique can be informative over the appropriate temporal window. Coupling DNA
techniques with those with a wider temporal range is potentially a powerful
methodological system to understand the dynamics of the specific diet of predators. If
the timescales of integration of dietary fatty acids and/or stable isotopes can be
estimated, and longitudinal DNA-diet samples can be collected, there is potential to
match the lag in dietary integration of prey biomarkers with specific prey data collected
from around the lag time.
118
Of primary importance in these methodological considerations is the life history of the
study species. For example, most mysticete cetaceans undertake extensive seasonal
migrations from high latitude feeding grounds segregated from low latitude
breeding/calving grounds (Clapham et al. 1999). It is generally assumed that feeding
during migration and in breeding/calving grounds is minimal, especially for lactating
mothers, although there are few data to verify this (Clapham et al. 1999; Oftedal 2000).
At any rate, it is assumed that the most important nutritional input for these animals
occurs during the time spent on the high latitude feeding grounds (e.g. Leaper et al.
2006). Longer time-frame diet analysis techniques are thus more appropriate for
samples collected from the breeding/calving grounds of these taxa. For other taxa that
have less pronounced habitat segregation, shorter range dietary techniques such as
DNA-based methods are as relevant as longer term methods. Given that sampling of
mysticete cetaceans on feeding grounds is rare and DNA-based diet methods would be
prone to environmental contamination of potential food sources, it seems more likely
that DNA-based diet techniques will find more use in studies of odontocete cetaceans
and other marine mammal species.
Table 6.1 Summary of methods available to study the diet of live cetaceans
Method Timescale Prey Taxonomic Resolution (Qualitative/Quantitative diet signatures?)
Advantages Disadvantages
Stable Isotopes Variable dependent on tissue used: Blood Plasma – Hours to Days
[1]
Epidermis – One to 3-4 months[2]
Muscle or whole blood – A Few to Several Months
[1] Bone Collagen or teeth –
Several Months to Years[1]
Baleen - Years
[1,3]
Very Low – Can only indicate relative trophic position of prey (This may change with compound-specific analyses) (N/Y)
- Variable, but relatively long timescales for inferring processes through time - Unlikely to be confounded by secondary predation
- Low taxonomic resolution - Potential confounding of diet/trophic processes with movement or age/physiological/growth processes - Isotopic enrichment needs characterisation for robust interpretation of data
Fatty Acid Analysis
Variable dependent on species ecology and tissue used
[4]
Hours- Months[4]
Intermediate – Can sometimes discriminate between prey higher taxa (Y/N)
- Variable, but relatively long timescales for inferring processes through time - Large multivariate datasets powerful for ordination techniques
- Timescales of dietary integration are not well characterised for most tissues in most species
[4]
- Sampling process can be inherently subjective
[5]
- Complex data sets and interpretation - Potentially confounded by secondary predation
[4]
Quantitative Fatty Acid Signature Analysis
Variable dependent on species ecology and tissue used
[4]
Hours - Months[4]
Very High – Species Level (Y/Y– in some instances where prior information is available)
- Variable, but relatively long timescales for inferring processes through time - Large multivariate datasets powerful for ordination techniques
- Timescales of dietary integration are not well characterised for most tissues in most species
[4]
- Substantial prior knowledge is required through experimentation and characterisation of prey variability
[4, 5]
- Potentially confounded by secondary predation
[4]
Observation of Feeding
Present moment Variable – Can be species, usually at least Phyla level (Y/N)
- Insights into hunting and prey capture strategies can be gained
- Inevitably biased towards surface events
DNA-based Analysis
Previous 3-6 to 19 - ≈48 hours Very High – Species level (Y/potentially – in some instances where prior information is available)
- High taxonomic resolution - Targeted sampling of potential unknown prey (phylogenetic information in recovered DNA sequences)
- Database of prey DNA sequences is required for specific identification - Very short timescale of information - Potentially confounded by secondary predation or environmental contamination
[1] Rubenstein
and Hobson (2004);
[2] Based on estimates of odontocete epidermal turnover cited in Ruiz-Cooley et al. (2004);
[3] Best and Schell (1996);
[4]
Budge et al. (2006); [5]
Grahl-Nielsen et al. (2004)
120
6.2 Live marine mammal age estimation and telomeres
It would appear that age estimation for live marine mammals currently remains elusive,
however some recent work has provided potential new directions. Herman et al. (2008;
2009) have demonstrated strong predictive relationships between blubber fatty acid
metrics and age in free ranging killer whales and humpback whales. Although these
relationships are empirical and the underlying biological processes uncertain, their
accuracy and precision is relatively high. They also appeared to be applicable
independent of diet, sex or ecotype in killer whales but were probably less general in
application to humpback whales (Herman et al. 2008; Herman et al. 2009). There have
also been promising examples of using transcriptional profiles to estimate the age of
mosquitos (Cook et al. 2006; Cook et al. 2007). Despite these promising initial results,
until an understanding of the mechanistic biological processes responsible for these
results occurs, it remains difficult to validate their applicability for unknown individuals
in field situations. Many of the issues raised in this discussion regarding the influence,
interactions of, and variation in endogenous and exogenous variables on telomere-age
relationships are applicable to any biomarker approach.
This study highlighted some difficulties in telomere measurement methodologies and
conceptual problems in the notion of using telomeres for age estimation. It is clear that
there are issues in applying denaturing terminal restriction fragment (TRF) telomere
measurement methodologies to non-model species. I recommend that, genome-wide
telomere sequence should be characterised with Bal 31 time series experiments, or
other telomere sequence characterisation procedures such as cytogenetic procedures.
Without this, it is not clear that what is being detected and measured in denaturing TRF
assays is actually telomere sequence. If not, there is potential for whatever genomic
element that is being measured to co-vary with the processes under study (i.e. sexual,
age or individual or population specific differences) and provide non-random results that
have nothing to do with telomere length variation and telomere dynamics processes.
There are high profile telomere studies reporting novel evolutionary processes, where
there are clear (and potentially non-telomeric) banding patterns throughout the size
range of the ‘telomere’ fragments measured (Olsson et al. 2010; Olsson et al. 2011).
In light of Chapter 4 and the many other recent telomere studies, a reappraisal and
critical review of the notion of using telomeres for age estimation was undertaken. Here
the studies advocating age estimation via telomeres were assessed against contradictory
work. This involved considering processes influencing telomere dynamics as well as
experimental design issues. It was concluded that despite earlier work, the weight of
evidence now indicates that telomeres alone should not be used for age estimation
without considerable further work.
6.2.1 Telomeres for age estimation: Reappraisal of the literature
The best examples of where telomere length (specifically TRF measurements) may be
used to estimate age of wild animals come from birds. Birds are ideal candidates for
telomere length analysis: samples of blood can be relatively easily obtained, cells can be
121
effectively buffered immediately to prevent DNA degradation and as little as 50–500µL
of blood can yield upwards of 10µg of DNA due to avian red blood cells being nucleated.
Thus, there is optimal collection, storage and DNA yield conditions for samples of bird
blood. These studies (Haussmann and Vleck 2002; Haussmann et al. 2003b) were able to
differentiate age into three lifespan stages in all measured individuals in all species
studied except zebra finches (Vleck et al. 2003), although a few statistical details are
given on this.
Attempts to improve the accuracy of age estimates gained from TRF analysis
(Haussmann and Mauck 2008a) involve selectively analysing size classes of telomeric
DNA that increase the correlation coefficients of relationships between telomere length
and known age samples. These methods were able to improve correlation coefficients (r 2 values) from <0.2 to >0.8. Yet, at the best correlation that could be found between
telomere length and age, there are data points where telomere metrics overlap
between 6-month and 2.5-year-old individuals. Zebra finches have a maximum lifespan
of more than 5 years (Haussmann et al. 2003b); thus, for individuals, this age resolution
is unable to distinguish an individual approximately one-tenth through its life from one
halfway through its life. This level of predictive accuracy for the ‘best’ possible situation
for individual telomere-based age estimation has obvious consequences for the
downstream use of the data. This also raises the issue of the use of the telomeres on a
‘population age structure’ scale rather than an ‘individual age estimate’ scale, which is
addressed later.
The logic employed by Haussmann & Mauck (2008) is also questionable. They used
samples of known age birds and numerically determined the size classes of telomeres to
analyse that yield the tightest relationship between age and the telomere metric for that
sample. As the premise of the relationships is to use them to predict age in unknown
individuals, this tactic assumes that the estimates of the best size class of telomeres to
analyse are independent of the characteristics of the individuals used in the analysis. It is
not clear that this is the case and it cannot be assessed from the experimental design. If
this assumption is not satisfied, the predictive relationships should only be used on
animals from that particular sample set, which are already known age. The authors
provide an explanation for why focusing the telomere size range selected numerically in
their analyses should be appropriate for unknown samples; specifically, that accruing of
short telomeres is a marker of older age. However, if these telomere size classes are
most illustrative of age and this is constant (an assumption when applying these
relationships to unknown individuals), there should be no need to calibrate this on a
species, and indeed sample-specific basis.
Another recent study has found that telomere data are useful for age estimation of
martens (Martes spp.; Pauli et al. 2011). This study is a tremendous conceptual advance
to this field because it acknowledges the effect of influential exogenous covariates on
telomere dynamics and attempts to include these in a Bayesian network statistical
analysis. Despite this, the study suffers from a lack of reporting of technical details of
qPCR assays, however it is not alone in this regard (see Horn et al. 2010). There are also
aspects of the data analysis that are unclear and problems with data interpretation. The
122
authors used muscle blocks from tooth-aged, harvested animals as a tissue source to
generate telomere data, developing models for age class estimation with covariate
datasets available during ‘non-invasive’ sampling (i.e. collection of skin or plucked hair
samples) or ‘live caught’ sampling, presumably where non-destructive sampling is
desirable. Unfortunately telomere length and dynamics are not homogeneous among
tissues (Friedrich et al. 2000; Nakamura et al. 2002), therefore it is unlikely that the
application of their models to telomere length from non-muscle tissue is appropriate.
These authors also argue that interstitial telomere sequence is unlikely to affect their
results because it is correlated with terminal restriction fragment data in other species.
As stated previously, this will be dependent on the genomic characteristics of ITS, which
can vary over small phylogenetic distances (Meyne et al. 1990). Additionally, Pauli et al.
(2011) used covariates that were likely to be strongly influential on the age structure of
their sample as priors in the Bayesian modelling framework, rather than those likely to
affect telomere dynamics. Given that there are no mechanistic links between their
covariates with telomere dynamics processes, it is not clear whether the nature of the
relationships between variables used for model parameterisation are representative of
the effects of more important variables for telomere dynamics of these species. These
authors also used the qPCR technique and generated relative telomere metrics in
relation to an arbitrarily chosen animal. Therefore, all future application of these models
rely on comparisons with this same arbitrarily chosen animal, since the telomere metrics
used to parameterise the predictive model were all created with reference to this
individual. Lastly, although sensitivity analysis was performed on model outputs, this did
not incorporate uncertainty in telomere metrics determination, which was a
(presumably, mean) co-efficient of variation of 14.6% from a single run of duplicates of
each loci for sample. There was no consideration in variation of metrics among runs, as
was noted in this study, in Pauli et al. (2011). These drawbacks highlight some of the
complexities involved in the notion of using telomeres as age markers and the need for
interdisciplinary approaches. However, despite these drawbacks, this style of study and
data handling will likely be most useful for telomere based age estimation studies. This
will of course be limited to study taxa where information on appropriate covariates will
be available, which is unlikely for many marine mammal taxa.
Other studies have concluded that telomere metrics are not useful for age estimation
(Hall et al. 2004; Juola et al. 2006; Karlsson et al. 2008; Garde et al. 2010) and found no
relationship between telomere length and age (Hatase et al. 2008; Bize et al. 2009).
These, and other studies, have shown large variation in telomere length within age
classes and telomere dynamics between individuals at similar developmental stages and
ages in a wide range of vertebrate taxa and using a wide range of methods (Hall et al.
2004; Juola et al. 2006; Pauliny et al. 2006; Aviv et al. 2008; Hatase et al. 2008;
Haussmann and Mauck 2008b; Bize et al. 2009; Nordfjall et al. 2009; Salomons et al.
2009), highlighting the importance of both genetic and environmental factors in
determining initial telomere length and rate of attrition. This betrays the context-
specific nature of relationships between telomere length and age. For example,
favourable relationships between human telomere length and age were reported for
forensic age estimation (Tsuji et al. 2002). However, similar relationships derived with
123
similar methods but different study subjects show very different models for human
telomere–age relationships (Unryn et al. 2005; Valdes et al. 2005). This suggests that the
relationships are, at best, useable for the populations they were developed from, but
not for different populations. While it was initially thought telomeres may provide an
age marker for forensic use (Tsuji et al. 2002; Takasaki et al. 2003; Nakagawa et al. 2004),
this has been rejected because of issues of reliability and error of age estimates
(Schmeling et al. 2007; Karlsson et al. 2008). These data suggest that, not only are there
influences other than chronology on telomere dynamics, but that these also vary
between individuals and subject groups. For wild animals, population- and cohort-level
variations are also expected.
There a number of endogenous and exogenous factors that affect telomere dynamics
(see Monaghan and Haussmann 2006; Aviv 2008). The level of telomerase activity in
tissues obviously has a profound and tissue-specific influence. Here, I focus on the
factors that vary at the individual level, which may influence initial telomere length,
telomere attrition rate, telomerase activity or all of these. Initial telomere length, while
being greatly variable, has a heritable component (Slagboom et al. 1994; Rufer et al.
1999; Nordfjall et al. 2005; Andrew et al. 2006) and in human offspring is positively
linked to paternal age (Unryn et al. 2005), although the extent and nature of heritability
is uncertain (e.g. Bischoff et al. 2005). There also appears to be heritable aspects to
regulation of telomere dynamics in vivo (Graakjaer et al. 2004). Telomere attrition rate is
dependent on average telomere length, with individuals with relatively long average
telomere length displaying more telomere attrition over time (Aviv et al. 2008; Bize et al.
2009; Nordfjall et al. 2009), although care is needed interpreting these data to avoid
regression to the mean artifacts (Kelly and Price 2005). It also appears that this relatively
higher rate of attrition acts preferentially on the longer telomere size classes (Salomons
et al. 2009). This is intriguing considering that longer telomeres are generally considered
to be a proxy for better individual quality or condition (e.g. Pauliny et al. 2006); it
suggests that the presence of longer telomeres at old ages is indicative of relatively
increased telomere maintenance mechanisms as much, or more than, decreased loss
through processes that effect attrition. The above longitudinal studies and others (Hall
et al. 2004; Ujvari and Madsen 2009; Chapter 5) also demonstrate that individuals can
exhibit stable or increased average telomere lengths over considerable periods of their
lifespan. Apparent telomere lengthening has appeared in a number of different
longitudinal studies on a number of different taxa with a variety of methods, suggesting
it may be a real phenomenon and not measurement error (see Salomons et al. 2009).
Telomere lengthening has considerable importance in relation to interpretation of cross-
sectional data and the models fit to it.
There are also exogenous factors that influence telomere dynamics, albeit through their
effects on endogenous processes. Telomerase activity and telomere dynamics have
been linked to stress levels in humans and mice, with greater stress associated with
shorter telomeres and decreased telomerase activity in otherwise healthy women (Epel
et al. 2004) and shorter telomeres in mice (Kotrschal et al. 2007). There is also a link
between socio-economic status (Cherkas et al. 2006), obesity and related variables
(Gardner et al. 2005; Valdes et al. 2005; Cherkas et al. 2008; Nettleton et al. 2008),
124
smoking (Valdes et al. 2005), vitamin D levels (Richards et al. 2007) and age-adjusted
telomere length in very large samples of humans. There appears to be a great diversity
of influential exogenous factors, although in most cases a link is proposed between the
exogenous factor and oxidative stress [and⁄or suppression of anti-oxidative capacity;
(Serra et al. 2003)] or processes that increase cellular turnover, such as tissue
inflammation, thereby providing a mechanism for effecting telomere dynamics.
Telomere length and dynamics appear to chronicle heredity, cumulative somatic cell
growth and replicative history, exposure to oxidative stress and associated influences of
exogenous processes and agents, over and above chronological age. This has led many
authors to propose that telomeres are indicative of ‘biological age’ if not chronological
age. Here, I define biological age as variation in longevity (Nakagawa et al. 2004) and
age-relative physiological condition among individuals and species. A number of studies
on both wild animals (Haussmann et al. 2005; Pauliny et al. 2006; Bize et al. 2009;
Salomons et al. 2009) and humans (Cawthon et al. 2003) have demonstrated that
telomere length and rates of telomere loss are predictive of life expectancy (but
seeMartin-Ruiz et al. 2005; Bischoff et al. 2006 for humans). While the notion of
biological age is interesting in relation to life history strategies and trade-offs in wild
animals (Monaghan and Haussmann 2006), the complex telomere⁄genetic⁄environment
associations can rarely differentiate between cause and covariation or consequence
(Aviv 2004), particularly in an ecological setting. Additionally, reconciling the biological
age concept with chorological age and its practical use in ecological investigations of
wild animals may be difficult. Many uses of age estimates are for age structure
information, terms of elapsed time, indicators of sexual maturity status or ‘life
experience’. All of which, by definition, cannot be accurately indicated by measures of
biological age.
Determining telomere–age relationships has predominately involved evaluating cross-
sectional relationships in populations. These studies are population specific and
generally constrained to a single time point. The relationships derived are inevitably
dependent on the coverage and characteristics of the starting age distribution and so
are influenced by sampling error and⁄or bias. The method advocated for age estimation
via telomere analysis was a development of a telomere length on age ‘calibration’ for a
particular species or more recently, population (Vleck et al. 2003; Haussmann and
Mauck 2008a). Unfortunately, there have been few statistics offered in many of these
studies that allow appraisal of the error of the technique for individual age estimates -
the coefficient of determination (r2 value), while being related to predictive confidence,
in itself does not allow this (Dapson 1980). Assuming that a representative sample is
initially collected, this rationale also assumes that the mechanisms acting to produce the
telomere length relationships are homogeneous among individuals in the species or
population and temporally. This is often not the case, particularly if applying a
‘calibration’ from captive known history populations to a wild population (e.g. Bjorndal
and Zug 1995). In a wild population setting, homogeneity in the host of variables
affecting telomere dynamics, other than age, is implausible. Deriving a significant cross-
sectional relationship between telomere length and age should therefore be regarded as
a starting point, not the end point, of investigating this age estimation technique. A
125
more appropriate end point would be an understanding of what influences this
relationship spatially and temporally, providing an understanding of the error of the
technique when applying it to unknown samples. These are intriguing ecological
questions in themselves, given the links between telomere dynamics and individual
condition, survival and fitness that have been identified.
Investigations into Leach’s storm petrel (Oceanodroma leucorhoa) telomere dynamics
provide a good illustration of the dangers of interpreting cross-sectional data in an
individual framework. The telomere rate of change in this relatively long-lived species
showed a positive relationship between telomere length and age, adding 75 ± 10 (SE)
base pairs per year (Haussmann et al. 2003b). This suggested that telomeres
‘lengthened’ considerably over time in this species (Vleck et al. 2003; Nakagawa et al.
2004; Juola et al. 2006 and others). However, Haussmann and Mauck (2008b) examined
variation in initial telomere length in a very large sample of newly hatched chicks and
compared that to old adults. They compared mutually exclusive mechanistic hypotheses
of the observed telomere ‘lengthening’ and concluded that long-lived individuals start
with long telomeres and variation in telomere length decreases with age as those with
short telomeres are selected against for survival to old age (see Fig. 2). Similarly,
Halaschek-Wiener et al. (2008) have demonstrated reduced variation in telomere length
in very old humans, and Salomons et al. (2009) have demonstrated selective
disappearance of individuals with short telomeres. These intriguing observations are
nonetheless not immediately reconcilable with other observations at the individual level.
For example, individuals can actually increase telomere length over time (Aviv et al.
2008; Bize et al. 2009; Hall et al. 2004; Nordfjall et al. 2009; Ujvari & Madsen 2009;
Chapter 5) and those with the longest telomeres also appear to show the highest rates
of telomere attrition (Hall et al. 2004; Aviv et al. 2008; Bize et al. 2009; Nordfjall et al.
2009). There are two points to take away from these studies: (i) it is impossible to imply
individual level processes from small cross-sectional samples, and (ii) using the original
cross-sectional relationship as the ‘calibration’ to use for age estimates of unknown
individuals would produce erroneous individual age estimates given the proposed
processes at work (Fig. 2).
Species suitable for telomere-based chronological age estimation are those that meet
two criteria; a large change in telomere length over their lifetime and low variation in
initial telomere length (Nakagawa et al. 2004). Given that the rate of telomere change
decreases with increasing lifespan (Haussmann et al. 2003b), it is probable that species
that meet the first criterion will only be those with relatively short lifespans. However,
the second criterion may prove to be more problematic. There appears to be
considerable variation in telomere length for any one age. For example, telomere
metrics from hatchling Leach’s storm petrels vary between 3 and 4 kb (dependent of
telomere size classes measured): the same order of variation shown across the entire
age distribution in earlier cross-sectional studies in this species (Haussmann et al. 2003b).
The level of variation in initial telomere length and dynamics is likely to be species, and
given parental processes, cohort specific. Indications from some short-lived mammals
(Coviello-McLaughlin and Prowse 1997), teleosts (Ying 2005; Hatakeyama et al. 2008)
and bird species (Haussmann et al. 2003b) indicate generally high intra-age class
126
telomere variation, which is to be expected given the host of factors influencing initial
telomere length and subsequent dynamics.
In this way, the hypotheses that telomere length is a marker of chronological age and of
‘biological age’ are conflicting. If telomere length reflects ‘biological age’, the well-
documented effects of early life processes (e.g. parental condition, provisioning and
resource acquisition) on growth, condition and survival would be reflected in telomere
dynamics. Evidence is mounting that this is the case (see above) and if so chronological
age estimates from telomere length alone are, by definition of biological age,
confounded by condition.
There appear to be a few successful applications of using telomere length to estimate
individual age. However, a taxon may display correlations between telomere length and
age, particularly when telomere metrics are obtained from TRF style assays. It is clear
that, given the host of factors affecting telomere dynamics, telomeres cannot simply be
viewed as a ‘molecular clock’. There is large intra-individual variation in telomere length
at any given age and therefore, in taxa where telomere length declines with age, it is not
apparent whether an individual with comparatively short telomeres is biologically, or
chronologically, old.
Telomere length may be useful at a population scale for comparison of relative age
structures. Many endogenous and exogenous variables can affect telomere length, but
these may average out at larger scales. However, for this approach to be valid it must be
assumed that the sum of the effects of endogenous and exogenous influences is not
significantly different between the units of comparison and over time. There does not
appear to be any reason to assume a priori that this will be the case without further
evidence, given the many population-based genetic and ecological differences in species.
It would also need to be assumed that the data used to parameterise models is
representative of the spectrum of variation of influential endogenous and exogenous
factors over time. If telomere length distributions between populations are used as a
supporting basis for such studies, the influences of ecological⁄genetic processes and⁄or
differences must be accounted for. Despite recognition of the large variation in telomere
length and dynamics in individuals, the most influential factors on this variation are still
barely understood for the most intensively studied group of animals: humans. It
therefore seems unlikely that these influential factors can be accounted for robustly in
wild animal populations at present.
Current evidence suggests that telomere length alone should not be used to predict
chronological age. More promising approaches are likely to involve identification and
inclusion of influential covariates to telomere dynamics in predictive models. However,
more basic information is needed to expand and properly evaluate this. I recommend
three broad areas for future work. The first is the nature of variability in telomere length
at any given age and the processes affecting this. In the context of age estimation, this
will require assessing variation in telomere length within age classes from large samples
and across cohorts. It will also require a more sound understanding of both genetic and
environmental factors which contribute to telomere length and dynamics, best informed
127
by experimental manipulations in addition to following individuals over time exposed to
different resource acquisition regimes and other relevant environmental conditions.
The second is whether the age resolution obtained is sufficient to be used for
downstream applications. This will depend on the relationships derived, their associated
errors, an understanding of what affects them and the scale of application of the data.
However, even if a resolution of only approximately one-third of a lifespan can be
obtained (and even this with individual outliers; see above) (Takasaki et al. 2003; Vleck
et al. 2003), this is clearly insufficient for calculating terms of elapsed time, judging
sexual maturity status, individual experience and quantitative applications such as age
structure information for population dynamics assessment. If, however, one wishes to
differentiate between ‘mature’ and ‘very old’, it may be that telomere length could
provide this information if the nature of variability and processes affecting it are
understood.
Of course, if telomere length yields a sufficiently strong robust relationship for age and
its variability is well established and understood, other applications are feasible.
Currently, studies have not provided enough information to evaluate this, most that
have concluded telomeres are unfavourable for most age data applications, and there is
mounting evidence to suggest that a single telomere length age relationship should not
be applied to unknown individuals. Finally, regardless of the complexity of the models
fitted, they must be validated using known age individuals not used to parameterise
them, like other age estimation methods. The one telomere-based age estimation study
that has performed this analysis [for humans; (Takasaki et al. 2003)] had a standard
error for age estimates of 7.5 years (and therefore a 95% C. I. = 14.7 years). This study
found that 20% of estimates from a small blind sample (3⁄15) were outside of the
standard error of the technique and 13% (2⁄15) were outside of the 95% confidence
limits.
Correlations exist between telomere length and age in many taxa. Examining telomere
dynamics may lead to profound ecological and evolutionary insights, but these must be
examined and interpreted with rigour and in the right ecological context. Some basic
mechanisms of telomere dynamics are understood; however, the influences,
interactions and variation in influential endogenous and exogenous factors on these
mechanisms are currently less clear, particularly in wild animal populations. There is a
limitation on using cross-sectional sampling methods alone to shed light on these
relationships. For age estimation techniques to be successful, there is a need to have an
understanding of the error of age estimates derived from the marker under study and
how this may vary when applied to unknown individuals. This has not been made
apparent through studies undertaken thus far that promote the use of telomeres for age
estimation. The appropriateness of methods also needs to be assessed and standardized
given the nature of the samples collected and their potential genomic characteristics.
Given the conflicting nature of the biological and chronological age concepts and the
fact that these remain disparate for most practical uses of age estimates in population
ecology, telomeres are currently of limited use for determining accurate and precise
chronological age in most situations. However, the application of complex statistical
128
modelling frameworks that can incorporate influential covariates and prior information,
such as Bayesian approaches, will likely be more fruitful for future studies than the
regression based techniques used previously. Such models will require validation and
assessment as to the scale (i.e. population or individual) of their applicability and should
not be assumed to be valid for prediction because they have been made. Telomeres may
yet prove useful as age markers in some short-lived species; even so, more work needs
to be performed to understand the nature of variability in initial telomere length and
telomere dynamics. The best information about this will probably come from integrated
cross-sectional and longitudinal studies from populations under different environmental
conditions.
6.3 Conclusions
Molecular genetic techniques have revolutionized ecological and evolutionary studies
over the past three decades. Early studies predominately focused on genetic diversity
and inference of historical processes that influenced contemporary diversity within and
among species (e.g. phylogenetic and phylogeography studies). It was soon realised that
these same molecular tools could be applied over shorter timescales to illuminate
previously intractable aspects of a species’ ecology. For example, the genetic
identification of individuals facilitated unprecedented insight into parentage, kin groups,
breeding success, movement patterns and population size estimation of wild animals, to
name a few applications. This study has developed and appraised molecular genetic
tools for investigating diet and age in live marine mammals: some of the most basic of
aspects of the ecology of any species. Although results within this study were mixed
regarding specific technique efficacy, these results provide a basis for further work. Just
as this study relied upon previous conceptual advances in the understanding the genetic
patterns in species boundaries and the genetic mechanisms underlying development, it
was facilitated by technological advances in the manipulation and analysis of DNA. With
the rapid growth of molecular genetic technologies and accompanying evolving genetic
techniques, promising existing ecological genetics approaches, such as some of those in
this thesis, will be strengthened and developed further. The application of next
generation sequencing to DNA-based diet studies is already allowing larger scale studies
to be undertaken than was previously possible (e.g. Deagle et al. 2009). In addition to
strengthening previous approaches, they will also undoubtedly deliver new
opportunities. The number of genome enabled species has risen dramatically since the
human genome was published, which could eventually facilitate development of
chromosome specific telomere assays in non-model species. This would allow
unparalleled insight into telomere biology and ecology potentially aiding applications of
telomeres for age estimation, among other things. It certainly is an exciting time to be
involved in molecular ecology research.
129
References Allen, S. J., L. Bejder and M. Krutzen (2011). "Why do indo-pacific bottlenose dolphins (Tursiops sp.) carry conch shells (Tubinella sp.) in Shark Bay, Western Australia." Marine Mammal Science 27(2): 449-454. Altschul, S. F., W. Gish, W. Miller, E. W. Myers and D. J. Lipman (1997). "Basic local alignment search tool." Journal of Molecular Biology 215: 403-410. Andrew, T., A. Aviv, M. Falchi, G. L. Surdulescu, J. P. Gardner, X. Lu, M. Kimura, B. S. Kato, A. M. Valdes and T. D. Spector (2006). "Mapping genetic loci that determine leukocyte telomere length in a large sample of unselected female sibling pairs." The American Journal of Human Genetics 78: 480-486. Arisotle (≈350BC). History of Animals (Books I & II).Translated by D'arcy Wentworth Thompson.Classics.mit.edu/Aristole/history_anim.html. Arnason, U., P. W. Allderdice, J. Lien and B. Widegren (1988). "Highly Repetitive DNA in the Baleen Whale Genera Balaenoptera and Megaptera." Journal of Molecular Evolution 27: 217-221. Aviv, A. (2004). "Telomeres and Human Aging: Facts and Fibs." Science of Aging Knowledge Environment 51: pe43. Aviv, A. (2008). "The epidemiology of human telomeres: faults and promises." The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences 63: 979-983. Aviv, A., W. Chen, J. P. Gardner, M. Kimura, M. Brimacombe, X. Cao, S. R. Srinivasan and G. S. Berenson (2008). "Leukocyte Telomere Dynamics: Longitudinal Findings Among Young Adults in the Bogalusa Heart Study." Amercian Journal of Epidemiology 169(3): 323-329. Aznar, F. J., J. Hernández-Orts, A. A. Suárez, G. García-Varela, J. A. Raga and H. I. Cappozzo (2011). "Assessing host–parasite specificity through coprological analysis: a case study with species of Corynosoma (Acanthocephala: Polymorphidae) from marine mammals." Journal of Helminthology: DOI: 10.1017/S0022149X11000149. Baker, C. S., A. Perry, J. L. Bannister, M. T. Weinrich, R. B. Abernethy, J. Calambokidis, J. Lien, R. H. Lambertsen, J. U. Ramirez and O. Vasquez (1993). "Abundant mitochondrial DNA variation and world-wide population structure in humpback whales." Proceedings of the National Academy of Sciences of the United States Of America 90(17): 8239-8243. Barros, N. B. (1993) Feeding ecology and foraging strategies of bottlenose dolphins on the central east coast of Florida. PhD Dissertation. University of Miami. Coral Gables, Florida. 328 pp Barros, N. B. and M. R. Clarke (2002). Diet. Encyclopedia of Marine Mammals. W. F. Perrin, Wursig, B., Thewissen, J. G. M. San Diego, Academic Press: 323-327.
130
Barros, N. B. and D. K. Odell (1990). Food habits of bottlenose dolphins in the southeastern United States.In: The Bottlenose Dolphin.S. Leatherwood and R. R. Reeves Academic Press.San Diego. Barros, N. B., P. H. Ostrom, C. A. Stricker and R. S. Wells (2010). "Stable isotopes differentiate bottlenose dolphins off west-central Florida." Marine Mammal Science 26(2): 324-336. Barros, N. B. and R. S. Wells (1998). "Prey and feeding patterns of resident bottlenose dolphins (Tursiops truncatus) in Sarasota Bay, Florida." Journal of Mammalogy 79(3): 1045-1059. Bensasson, D., D. Zhang, D. L. Hartl and G. M. Hewitt (2001a). "Mitochondrial pseudogenes: evolutions's misplaced witnesses." Trends in ecology and evolution 16(6): 314-321. Bensasson, D., D. X. Zhang, D. L. Hartl and G. M. Hewitt (2001b). "Mitochondrial pseudogenes: evolution's misplaced witnesses." Trends In Ecology & Evolution 16(6): 314-321. Berens McCabe, E. B., D. P. Gannon, N. B. Barros and R. S. Wells (2010). "Prey selection by resident common bottlenose dolphins (Tursiops truncatus) in Sarasota Bay, Florida." Marine Biology 157: 931-942. Best, P. B. and D. M. Schell (1996). "Stable isotopes in southern right whale (Eubalaena australis) baleen as indicators of seasonal movements, feeding and growth." Marine Biology 124: 483-494. Bischoff, C., J. Graakjaer, H. C. Peterson, J. V. B. Hjelmborg, J. W. Vaupel, V. A. Bohr, S. Kølvraa and K. Christensen (2005). "The heritability of Telomere length among the elderly and oldest-old." Twin Research and Human Genetics 8(5): 433-439. Bischoff, C., H. C. Peterson, J. Graakjaer, K. Andersen-Ranberg, J. W. Vaupel, V. A. Bohr, S. Kolvraa and K. Christensen (2006). "No association between telomere length and survival among the elderly and oldest old." Epidemiology 17(2): 190-194. Bize, P., F. Criscuolo, N. B. Metcalfe, L. Nasir and P. Monaghan (2009). "Telomere dynamics rather than age predict life expectancy in the wild." Proceedings of the Royal Society B-Biological Sciences: doi: 10.1098/rspb.2008.1817. Bjorndal, K. A. and G. R. Zug (1995). Growth and age of sea turtles. Biology and conservation of sea turtles, revised edn. K. A. Bjorndal. Washington, Smithsonian Institution Press. Blackburn, E. H. (1991). "Structure and function of telomeres." Nature 350: 569-573. Blackburn, E. H. (2000). "Telomere states and cell fates." Nature 408: 53-56. Blackburn, E. H. and P. B. Challoner (1984). "Identification of a telomeric DNA sequence in Trypanosoma brucei." Cell 36(2): 447-457.
131
Blankenship, L. E. and A. A. Yayanos (2005). "Universal primers and PCR of gut contents to study marine invertebrate diets." Molecular Ecology 14(3): 891-899. Bohmann, K., A. Monadjem, C. L. Noer, M. Rasmussen, M. R. K. Zeale, E. Clare, G. Jones, E. Willerslev and M. T. P. Gilbert (2010). "Molecular Diet Analysis of Two African Free-Tailed Bats (Molossidae) Using High Throughput Sequencing." PLoS One 6(6): e21441. doi:10.1371/journal.pone.0021441. Bowles, E., P. M. Schulte, D. J. Tollit, B. E. Deagle and A. W. Trites (2011). "Proportion of prey consumed can be determined from faecal DNA using real-time PCR." Molecular Ecology Resources 11: 530-540. Boyd, I. L. (2002). "Estimating food consumption of marine predators: Antarctic fur seals and macaroni penguins." Journal of Applied Ecology 39: 103-119. Bradshaw, C. J. A., M. Hindell, A., N. J. Best, K. L. Phillips, G. Wilson and P. D. Nichols (2003). "You are what you eat: describing the foraging ecology of southern elephant seals (Mirounga leonina) using blubber fatty acids." Proceedings of the Royal Society of London. B Series 270: 1283-1292. Broccoli, D. (2004). "Function, replication and structure if the mammalian telomere." Cytotechnology 45: 3-12. Budge, S. M., S. J. Iverson and H. N. Koopman (2006). "Studying trophic ecology in marine ecosystems using fatty acids: A primer on analysis and interpretation." Marine Mammal Science 22(4): 759-801. Burk, A., J. P. Douzery and M. S. Springer (2002). "The Secondary Structure of Mammalian Mitochondrial16S rRNA Molecules: Refinements Based on a Comparative Phylogenetic Approach." Journal of Mammalian Evolution 9(3): 225-252. Campana, S. E. (2001). "Accuracy, precision and quality control in age determination, including a review of the use and abuse of age validation methods." Journal of Fish Biology 59: 197-242. Campbell, R. A., N. J. Gales, G. M. Lento and C. S. Baker (2008). "Islands in the sea: extreme female natal site fidelity in the Australian sea lion, Neophoca cinerea." Biology Letters 4(1): 139-142. Casper, R. M., S. N. Jarman, B. E. Deagle, N. J. Gales and M. Hindell, A. (2007a). "Detecting prey from DNA in predator scats: A comparison with morphological analysis, using Arctocephalus seals fed a known diet." Journal Of Experimental Marine Biology And Ecology: doi:10.1016/j.jembe.2007.04.002. Casper, R. M., S. N. Jarman, N. J. Gales and M. A. Hindell (2007b). "Combining DNA and morphological analyses of faecal samples improves insight into trophic interactions: a case study using a generalist predator." Marine Biology 152(4): 815-825. Caughley, G. (1977). Analysis of Vertebrate Populations John Wiley & Sons.London. Cawthon, R. M. (2002). "Telomere measurement by quantitative PCR." Nucleic Acids Research 30(10): e47.
132
Cawthon, R. M., K. R. Smith, E. O'Brien, A. Sivatchenko and R. A. Kerber (2003). "Association between telomere length in blood and mortality in people aged 60 years and older." Lancet 361: 393-395. Chao, A. and S. M. Lee (1992). "Estimating the number of classes via sampling coverage." Journal of the American Statistical Association 87: 210-217. Charlesworth, B., P. Sniegowski and W. Stephan (1994). "The evolutionary dynamics of repetitive DNA in eukaryotes." Nature 317(6494): 215-220. Cherkas, L. F., A. Aviv, A. M. Valdes, J. L. Hunkin, J. P. Gardner, G. L. Surdulescu, M. Kimura and T. D. Spector (2006). "The effects of social status on biological aging as measured by white-blood-cell telomere length." Aging Cell 5: 361-365. Cherkas, L. F., J. L. Hunkin, B. S. Kato, J. B. Richards, J. P. Gardner, G. L. Surdulescu, M. Kimura, X. Lu, T. D. Spector and A. Aviv (2008). "The Association Between Physical Activity in Leisure Time and Leukocyte Telomere Length." Archives of Internal Medicine 168(2): 154-158. Childerhouse, S., G. Dickie and H. G (2004). "Ageing live New Zealand sea lions (Phocarctos hookeri) using the first post-canine tooth." Wildlife Research 31(2): 177-181. Clapham, P. J., S. B. Young and R. L. Brownell Jr (1999). "Baleen whales: conservation issues and the status of the most endangered populations." Mammal Review 29(1): 35-60. Clarke, M. R. and N. Macleod (1976). "Cephalopod remains fomr sperm whales caught off Iceland." Journal of the marine biology association of the United Kingdom 56: 733-749. Cobb, B. D. and J. M. Clarkson (1994). "A simple procedure for optimising the polymerase chain reaction (PCR) using modified Taguchi methods." Nucleic Acids Research 22(18): 3801-3805. Collura, R. V. and C. Stewart (1995). "Insertions and duplications of mtDNA in the nuclear genomes of Old World monkeys and hominoids." Nature 378: 485-489. Connor, R. C., M. R. Heithaus, P. Berggren and J. L. Miksis (2000). "'Kerplunking': Surface fluke-splashes during shallow water bottom foraging by bottlenose dolphins." Marine Mammal Science 16: 646-653. Cook, P. E., L. E. Hugo, I. Iturbe-Omaetxe, C. R. Williams, S. F. Chenoweth, S. A. Ritchie, P. A. Ryan, B. H. Kay, M. W. Blows and J. P. O'Neill (2006). "The use of transcriptional profiles to predict adult mosquito age under field conditions." Proceedings of the National Academy of Sciences of the United States of America 103(48): 18060 - 18065. Cook, P. E., L. E. Hugo, I. Iturbe-Omaetxe, C. R. Williams, S. F. Chenoweth, S. A. Ritchie, P. A. Ryan, B. H. Kay, M. W. Blows and S. L. O'Neill (2007). "Predicting the age of mosquitoes using transcriptional profiles." Natuer Protocols 2(11): 2796 - 2806.
133
Coviello-McLaughlin, G. M. and K. R. Prowse (1997). "Telomere length regulation during postnatal development and ageing in Mus spretus." Nucleic Acids Research 25: 3051-3058. Criscuolo, F., P. Bize, L. Nasir, N. B. Metcalfe, C. G. Foote, K. Griffiths, E. A. Gault and P. Monaghan (2009). "Real-time quantitative PCR assay for measurement of avian telomeres." Journal of Avian Biology 40: 342-347. Dapson, R. W. (1980). "Guidelines for statistical usage in age-estimation technics." Journal of Wildlife Managment 44(3): 541-548. Darwin, C. (2003). The Origin of Species.Signet Classic.New York. de la Mare, W. K. (1987). "Fitting population models to combinations of CPUE series and absolute abundance estimates." Report to the International Whaling Commission 37: 379-381. Deagle, B. E., A. Chiaradia, J. McInnes and S. N. Jarman (2010). "Pyrosequencing faecal DNA to determine diet of little penguins: is what goes in what comes out?" Conservation Genetics 11: 2039-2048. Deagle, B. E., J. P. Eveson and S. N. Jarman (2006). "Quantification of damage in DNA recovered from highly degraded samples – a case study on DNA in faeces." Frontiers in Zoology 3(11). Deagle, B. E., N. J. Gales, K. Evans, S. N. Jarman, S. M. Robinson, R. Trebilco and M. Hindell, A. (2007). "Studying seabird diet through genetic analysis of faeces: A case study on Macaroni Penguins (Eudyptes chrysolophus)." PLoS One 2(9): e831. doi:10.1371/journal.pone.0000831. Deagle, B. E., S. N. Jarman, D. Pemberton and N. J. Gales (2005a). "Genetic screening for prey in the gut contents from a giant squid (Architeuthis sp.)." Journal Of Heredity 96(4): 417-423. Deagle, B. E., R. Kirkwood and S. N. Jarman (2009). "Analysis of Australian fur seal diet by pyrosequencing prey DNA in faeces." Molecular Ecology. Deagle, B. E., R. Kirkwood and S. N. Jarman (In Press). "Analysis of Australian fur seal diet by pyrosequencing prey DNA in faeces." Molecular Ecology. Deagle, B. E. and D. J. Tollit (2006). "Quantitative analysis of prey DNA in pinniped faeces: potential to estimate diet composition?" Conservation Genetics: DOI 10.1007/s10592-00609197-7. Deagle, B. E., D. J. Tollit, S. N. Jarman, M. A. Hindell, A. W. Trites and N. J. Gales (2005b). "Molecular scatology as a tool to study diet: analysis of prey DNA in scats from captive Steller sea lions." Molecular Ecology 14(6): 1831-1842. Delany, M. E., A. B. Krupkin and M. M. Miller (2000). "Organization of telomere sequences in birds: Evidence for arrays of extreme length and for in vivo shortening." Cytogenetics and Cell Genetics 90: 139-145.
134
Dmitriev, P. V. and Y. S. Vassetzky (2009). "Analysis of telomeric DNA: Current approaches and methods." Russian Journal of Developmental Biology 40: 125-144. Dunbar, J., S. M. Barns, L. M. Ticknor and C. R. Kuske (2002). "Empirical and Theoretical Bacterial Diversity in Four Arizona Soils." Applied And Environmental Microbiology 68(6): 3035-3045. Edgar, R. C. (2004). "MUSCLE: multiple sequence alignment with high accuracy and high throughput." Nucleic Acids Research 32(5): 1792-1797. Efron, B. and R. J. Tibshirani (1993). An introduction to the Bootstrap CRC Press LLC.Boca Raton, Florida, USA. Epel, E. S., E. H. Blackburn, J. Lin, F. S. Dhabhar, N. E. Adler, J. D. Morrow and R. M. Cawthon (2004). "Accelerated telomere shortening in response to life stress." Proceedings of the National Academy of Sciences of the United States of America 101(49): 17312-17315. Evans, K. and M. Hindell, A. (2004). "The diet of sperm whales (Physeter macrocephalus) in southern Australian waters." ICES Journal of Marine Science 61(8): 1313-1329. Fiesta-Bianchet, M., P. Blanchard, J. M. Gaillard and J. M. Hewison (2002). "Tooth extraction is not an acceptable technique to age live ungulates." Wildlife Society Bulletin 30(1): 282-283. Flinn, R., A. W. Trites, E. J. Gregr and R. I. Perry (2002). "Diets of fin, sei, and sperm whales in British Columbia: an analysis of commercial whaling records, 1963 - 1967." Marine Mammal Science 18: 663-679. Frie, A. K., V. A. Potelov, M. C. S. Kingsley and T. Haug (2003). "Trends in age-at-maturity and growth parameters of female Northeast Atlantic harp seals, Pagophilus groenlandicus (Erxleben, 1777)." Ices Journal Of Marine Science 60(5): 1018-1032. Friedrich, U., E. Griese, M. Schwab, P. Fritz, K. Thon and U. Klotz (2000). "Telomere length in different tissues of elderly patients." Mechanisms of Ageing and Development 119: 89-99. Gales, N. J., T. Kasuya, P. J. Clapham and R. L. Brownell Jr (2005). "Japan's whaling plan under scrutiny." Nature 435: 883-884. Gannon, D. P., N. B. Barros, D. P. Nowacek, A. J. Read, D. M. Waples and R. S. Wells (2005). "Prey detection by bottlenose dolphins, Tusiops truncatus: an experimental test of the passive listening hypothesis." Animal Behaviour 69: 709-720. Garde, E., A. K. Frie, G. J. Dunshea, S. H. Hansen, K. M. Kovacs and C. Lydersen (2010). "Harp seal ageing techniques-teeth, aspartic acid racemization, and telomere sequence analysis." Journal of Mammalogy 91: 1365-1374. Garde, E., M. P. Heide-Jørgensen, S. H. Hansen, G. Nachman and M. C. Forchhammer (2007). "Age-specific growth and remarkable longevity in narwhals (Monodon monoceros) from West Greenland as estimated by aspartic acids racemization." Journal of Mammology 88(1): 49-58.
135
Gardner, J. P., S. Li, S. R. Srinivasan, W. Chen, M. Kimura, X. Lu, G. S. Berenson and A. Aviv (2005). "Rise in Insulin Resistance Is Associated With Escalated Telomere Attrition." Circulation 111: 2171-2177. George, J. C., J. Bada, J. Zeh, L. Scott, S. E. Brown, T. O'Hara and R. Suydam (1999). "Age and growth estimates of bowhead whales (Balaena mysticetus) via aspartic acid racemization." Canadian Journal of Zoology 77: 571-580. Gonzalez, J. M., J. Zimmermann and C. Saiz-Jimenez (2004). "Evaluating putative chimeric sequences from PCR amplified products." Bioinformatics 21(3): 333-337. Graakjaer, J., L. Pascoe, H. Der-Sarkissian, G. Thomas, S. Kolvraa, K. Christensen and J. Londoño-Vallejo (2004). "The relative lengths of individual telomeres are defined in the zygote and strictly maintained during life." Aging Cell 3: 97-102. Grahl-Nielsen, O., M. Andersen, A. E. Derocher, C. Lydersen, O. Wiig and K. M. Kovacs (2004). "Reply to Comment on Grahl-Nielsen et al. (2003): sampling, data treatment and predictions in investigations on fatty acids in marine mammals." Marine Ecology Progress Series 281: 303-306. Green, M. L., D. L. Herzing and J. D. Baldwin (2007). "Noninvasive methodology for the sampling and extraction of DNA from free-ranging Atlantic spotted dolphins (Stenella frontalis)." Molecular Ecology Notes 7(6): 1287-1292. Green, S. J. and D. Minz (2005). "Suicide polymerase endonuclease restriction, a novel technique for enhancing PCR amplification of minor DNA templates." Applied And Environmental Microbiology 71(8): 4721-4727. Greider, C. W. (1999). "Telomeres Do D-Loop-T-Loop." Cell 97: 419-422. Gygax, L. (2002). "Evolution of group size in the superfamily Delphinoidea (Delphinidae, Phocoenidae and Monodontidae): a quantitative comparative analysis." Mammal Review 32(4): 295-314. Haas, B. J., D. Gevers, A. M. Earl, M. Feldgarden, D. V. Ward, G. Giannoukos, D. Ciulla, D. Tabbaa, S. K. Highlander, E. Sodergren, B. Methe, T. Z. DeSantis, J. F. Petrosino, R. Knight and B. W. Birren (2011). "Chimeric 16S rRNA sequence formation and detection in Sanger and 454-pyrosequenced PCR amplicons." Genome Research 21(3): 494-504. Hajibabaei, M., D. H. Janzen, J. M. Burns, W. Hallwachs and P. D. N. Hebert (2006). "DNA barcodes distinguish species of tropical Lepidoptera." Proceedings Of The National Academy Of Sciences Of The United States Of America 103(4): 968-971. Hall, M. E., L. Nasir, F. Daunt, E. A. Gault, J. P. Croxall, S. Wanless and P. Monaghan (2004). "Telomere loss in relation to age and early environment in long-lived birds." Proceedings of the Royal Society of London. B Series 271: 1571-1576. Hall, T. A. (1999). "BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT." Nucleic Acids Symposium Series 41: 95-98.
136
Harcourt, R. G., J. J. Kingston, M. Cameron, J. R. Waas and M. Hindell, A. (2007). "Paternity analysis shows experience, not age, enhances mating success in an aquatically mating pinniped, the Weddell seal (Leptonychotes weddellii)." Behavioral Ecology and Sociobiology 61: 643-652. Harley, C. B., A. B. Futcher and C. W. Greider (1990). "Telomeres shorten during ageing of human fibroblasts." Nature 345: 458-460. Harper, G. L., R. A. King, C. S. Dodd, J. D. Harwood, D. M. Glen, M. W. Bruford and W. O. C. Symondson (2005). "Rapid screening of invertebrate predators for multiple prey DNA targets." Molecular Ecology 14(3): 819-827. Hastie, N. D., M. Dempster, M. G. Dunlop, A. M. Thompson, D. K. Green and R. C. Allshire (1990). "Telomere reduction in human colorectal carcinoma and with ageing." Nature 346: 866-868. Hatakeyama, H., K. I. Nakamura, N. Izumiyama-Shimorriura, A. Ishii, S. Tsuchida, K. Takubo and N. Ishikawa (2008). "The teleost Oryzias latipes shows telomere shortening with age despite considerable telomerase activity throughout life (vol 129, pg 550, 2008)." Mechanisms Of Ageing And Development 129(11): 692-692. Hatase, H., R. Sudo, K. K. Watanabe, T. Kasugai, T. Saito, H. Okamoto, I. Uchida and K. Tsukamoto (2008). "Shorter telomere length with age in the loggerhead turtle: a new hope for live sea turtle age estimation." Genes and Genetic Systems 83: 423-426. Haussmann, M. F. and R. A. Mauck (2008a). "New strategies for telomere-based age estimation." Molecular Ecology Notes 8(2): 264-274. Haussmann, M. F. and R. A. Mauck (2008b). "Telomeres and longevity: Testing an evolutionary hypothesis." Molecular Biology and Evolution 25: 220-228. Haussmann, M. F. and C. M. Vleck (2002). "Telomere length provides a new technique for aging animals." Oecologia 130: 325-328. Haussmann, M. F., C. M. Vleck and I. C. T. Nisbet (2003a). "Calibrating the telomere clock in common terns, Sterna hirundo." Experimental Gerontology 38: 787-789. Haussmann, M. F., D. W. Winkler, K. M. O'Reilly, C. E. Huntington, I. C. T. Nisbet and C. M. Vleck (2003b). "Telomeres shorten more slowly in long-lived birds and mammals than in short-lived ones." Proceedings of the Royal Society of London. B Series 270(1522): 1387-1392. Haussmann, M. F., D. W. Winkler and C. M. Vleck (2005). "Longer telomeres associated with higher survival in birds." Biology Letters 1: 212-214. Heithaus, M. R. (2001). "Predator-prey and competitive interactions between sharks (order Selachii) and dolphins (suborder Odontoceti): a review." Journal of Zoology 253: 53-68. Henmann, M. T., M. A. Strong, L. Y. Hao and C. W. Grieder (2001). "The shortest telomere, not average telomere length, is critical for cell viability and chromosome stability." Cell 107: 67-77.
137
Herman, D. P., C. O. Matkin, G. M. Ylitalo, J. W. Durban, M. Bradley Hanson, M. E. Dahlheim, J. M. Straley, P. R. Wade, K. L. Tilbury, R. H. Boyer, R. W. Pearce and M. M. Krahn (2008). "Assessing age distributions of killer whale Orcinus orca populations from the composition of endogenous fatty acids in their outer blubber layers." Marine Ecology Progress Series 372(289-302). Herman, D. P., G. M. Ylitalo, R. J., J. M. Straley, C. M. Gabriele, P. J. Clapham, R. H. Boyer, K. L. Tilbury, R. W. Pearce and M. M. Krahn (2009). "Age determination of humpback whales Megaptera novaeangliae through blubber fatty acid compositions of biopsy samples." Marine Ecology Progress Series 392: 277-293. Hoelzel, A. R., J. Hey, M. E. Dahlheim, C. Nicholson, V. Burkanov and N. Black (2007). "Evolution of Population Structure in a Highly Social Top Predator, the Killer Whale." Molecular Biology and Evolution 24(6): 1407-1415. Hoffman, J. I., C. W. Matson, W. Amos, T. R. Loughlin and J. W. Bickham (2006). "Deep genetic subdivision within a continuously distributed and highly vagile marine mammal, the Steller’s sea lion (Eumetopias jubatus)." Molecular Ecology 15: 2821-2832. Hogg, C. J., T. L. Rogers, A. Shorter, K. Barton, P. J. O. Miller and D. Nowacek (2009). "Determination of steroid hormones in whale blow: It is possible." Marine Mammal Science 25(3): 605-618. Hohn, A. A., M. D. Scott, R. S. Wells, J. C. Sweeny and A. B. Irvine (1989). "Growth layers in teeth from known-age, free-ranging bottlenose dolphins." Marine Mammal Science 5(4): 315-342. Holmes, E. E. and A. E. York (2003). "Using Age Structure to Detect Impacts on Threatened Populations: a Case Study with Stellar Sea Lions." Conservation Biology 17(6): 1794-1806. Horn, T., B. C. Robertson and N. J. Gemmell (2010). "The use of telomere length in ecology and evolutionary biology." Heredity 105: 497-506. Huber, J. A., D. A. Butterfield and J. A. Baross (2002). "Temporal changes in archaeal diversity and chemistry in a mid-ocean ridge subseafloor habitat." Applied And Environmental Microbiology 68(4): 1585-1594. Hughes, J. B., J. J. Hellmann, T. H. Ricketts and B. J. M. Bohannan (2001). "Counting the Uncountable: Statisical Approaches to Estimating Microbial Diversity." Applied And Environmental Microbiology 67(10): 4399-4406. Iverson, S. J., C. Feild, W. D. Bowen and W. Blanchard (2004). "Quantitative fatty acid signature analysis: a new method of estimating predator diets." Ecological Monographs 74: 211-235. Iverson, S. J., A. M. Springer and A. S. Kitaysky (2007). "Seabirds as indicators of food web structure and ecosystem variability: qualitative and quantitative diet analyses using fatty acids." Marine Ecology Progress Series 352: 235-244.
138
Jansen, O. E., M. F. Leopold, E. H. W. G. Meesters and C. Smeenk (2010). "Are white-beaked dolphins Lagenorhynchus albirostris food specialists? Their diet in the southern North Sea." Journal of the Marine Biological Association of the United Kingdom 90(8): 1501-1508. Jarman, S. N. (2004). "Amplicon: software for designing PCR primers on aligned DNA sequences." Bioinformatics 20(10): 1644-1645. Jarman, S. N. (2006). "Cleaver: software for identifying taxon specific restriction endonuclease recognition sites." Bioinformatics 22: 2160-2161. Jarman, S. N., B. E. Deagle and N. J. Gales (2004). "Group-specific polymerase chain reaction for DNA-based analysis of species diversity and identity in dietary samples." Molecular Ecology 13: 1313-1322. Jarman, S. N., N. J. Gales, M. Tierney, P. C. Gill and N. G. Elliot (2002). "A DNA-based method for identification of krill species and its application to analysing the diet of marine vertebrate predators." Molecular Ecology 11: 2679-2690. Jarman, S. N., K. S. Redd and N. J. Gales (2006). "Group-specific primers for amplifying DNA sequences that identify Amphipoda, Cephalopoda, Echinodermata, Gastropoda, Isopoda, Ostracoda and Thoracica." Molecular Ecology Notes 6(1): 268-271. Jenkins, G. P. (1988). "Micro- and fine-scale distribution of microplankton in the feeding environment of larval flounder." Marine Ecology Progress Series 43: 233-244. Ju, S. J., D. H. Secor and H. R. Harvey (1999). "Growth rate variability and lipofuscin accumulation rates in the blue crab Callinectes sapidus." Marine Ecology Progress Series 224: 197-205. Juen, A. and M. Traugott (2005). "Detecting predation and scavenging by DNA gut-content analysis: a case study using a soil insect predator-prey system." Oecologia 142(3): 344-352. Juola, F. A., M. F. Haussmann, D. C. Dearborn and C. M. Vleck (2006). "Telomere shortening in a long-lived marine bird: cross-sectional analysis and test of an aging tool." The Auk 123(3): 775-783. Karlsson, A. O., A. Svensson, A. Marklund and G. Holmlund (2008). "Estimating human age in forensic samples by analysis of telomere repeats." Forensic Science International: Genetics Supplement Series 1: 569-571. Kelly, C. and T. D. Price (2005). "Correcting for regression to the mean in behavior and ecology." American Naturalist 166: 700-707. Kimura, M. and A. Aviv (2011). "Measurement of telomere DNA content by dot blot analysis." Nucleic Acids Research: doi:10.1093/nar/gkr235. Kimura, M., R. C. Stone, S. C. Hunt, J. Skurnick, X. Lu, X. Cao, C. B. Harley and A. Aviv (2010). "Measurement of telomere length by southern blot analysis of the terminal restriction fragment lengths." Nature Protocols 5: 1596-1607.
139
King, R. A., D. S. Read, M. Traugott and W. O. C. Symondson (2008). "Molecular analysis of predation: a review of best practice for DNA-based approaches." Molecular Ecology 17(4): 947-963. Kipling, D. and H. J. Cooke (1990). "Hypervariable ultra-long telomeres in mice." Nature 347: 400-402. Kirby, L. T. (1992). DNA Fingerprinting W. H. Freeman and Company.New York. Klevezal, G. A. and S. E. Kleinenberg (1969). Age determination of mammals from annual layers in teeth and bones.The U.S. Department of the Interior and The National Science Foundation.T. U. S. D. o. t. I. a. T. N. S. Foundation.Washington D. C. Kobayashi, N., K. Tamura and T. Aotsuka (1999). "PCR error and molecular population genetics." Biochemical Genetics 37: 317-321. Kohn, M. H. and R. K. Wayne (1997). "Facts from feces revisited." Trends in ecology and evolution 12: 223-227. Kotrschal, A., P. IImonen and D. J. Penn (2007). "Stress impacts telomere dynamics." Biology Letters 3(128-130). Krafft, B. A., K. M. Kovacs, A. K. Frie, T. Haug and C. Lydersen (2006). "Growth and population parameters of ringed seals (Pusa hispida) from Svalbard, Norway, 2002–2004." Ices Journal Of Marine Science 63(6): 1136-1144. Kumar, S., K. Tamura and M. Nei (2004). "MEGA3: Integrated software for Molecular Evolutionary Genetics Analysis and sequence alignment." Briefings in Bioinformatics 5: 150-163. Kvitrud, M. A., S. D. Riemer, R. F. Brown, M. R. Bellinger and M. A. Banks (2005). "Pacific harbor seals (Phoca vitulina) and salmon: genetics presents hard numbers for elucidating predator-prey dynamics." Marine Biology 147(6): 1459-1466. Lasley, B. L. and J. F. Kirkpatrick (1991). "Monitoring of ovarian function in captive and free-ranging wildlife by means of urinary and fecal steroids." Journal of Zoo and Wildlife Medicine 22(1): 23-31. Leaper, R., J. Cooke, P. Trethan, K. Reid, V. Rowntree and R. Payne (2006). "Global climate drives southern right whale (Eubalaena australis) population dynamics." Biology Letters 2: 289-292. Lesnoff, M. and R. Lancelot (2010). "aod: Analysis of Overdispersed Data." R Package version 1.2, URL http://cran.r-project.org/package=aod. Li, T., C. Kim, C. M. Rubin and C. W. Schmid (2000). "K562 cells implicate increased chromatin accessibility in Alu transcriptional activation." Nucleic Acids Research 28(16): 3031-3039. Lockyer, C. (1984). "Age determination by means of the ear plug in baleen whales." Report to the International Whaling Commission 34: 692-696.
140
Lopez, J. V., N. Yuhki, R. Masuda, W. Modi and S. J. O'brien (1994). "Numt, a recent transfer and tandem amplification of mitochondrial DNA to the nuclear genome of the domestic cat." Journal of Molecular Evolution 39(2): 174-190. Marmontel, M., T. J. Nhea, H. I. Kochman and S. R. Humphrey (1996). "Age determination in manatees using growth-layer-group counts in bone." Marine Mammal Science 12(1): 54-88. Martin-Ruiz, C. M., J. Gussekloo, D. van Heemst, T. von Zglinicki and R. G. J. Westendorp (2005). "Telomere length in white blood cells is not associated with morbidity or mortality in the oldest old: a population-based study." Aging Cell 4: 287-290. Martin, D. L., R. M. Ross, L. B. Quetin and A. E. Murray (2006). "Molecular approach (PCR-DGGE) to diet analysis in young Antarctic krill Euphausia superba." Marine Ecology Progress Series 319: 155-165. McLeod, C. D., M. B. Santos and G. J. Pierce (2003). "Review of data on diets of beaked whales: evidence of niche separation and geographic segregation." Journal of the Marine Biological Association of the United Kingdom 83: 651-665. Messier, T. L., J. P. O'Neill, S. Hou, J. A. Nicklas and B. A. Finette (2003). "In vivo transposition mediated by V(D)J recombinase in human T lymphocytes." The EMBO Journal 22: 1381-1388. Meyne, J., R. J. Baker, H. H. Hobart, T. C. Hsu, O. A. Ryder, O. G. Ward, J. E. Wiley, D. H. Wurster-Hill, T. L. Yates and R. K. Moyzis (1990). "Distribution of non-telomeric sites of the (TTAGGG)n telomeric sequence in vertebrate chromosomes." Chromosoma 99: 3-10. Mirol, P. M., S. Mascheretti and J. B. Searle (2000). "Multiple nuclear pseudogenes of mitochondrial cytochrome b in Ctenomys (Caviomorpha, Rodentia) with either great similarity to or high divergence from the true mitochondrial sequence." Heredity 84: 538-547. Monaghan, P. and M. F. Haussmann (2006). "Do telomere dynamics link lifestyle and lifespan?" Trends in ecology and evolution 21(1): 47-53. Morin, P. A., N. M. Hedrick, K. M. Robertson and C. A. Leduc (2006). "Comparative mitochondrial and nuclear quantitative PCR of historical marine mammal tissue, bone, baleen, and tooth samples." Molecular Ecology Notes 7(3): 404-411. Morin, P. A., K. K. Martien, F. I. Archer, F. Cipriano, D. Steel, J. Jackson and B. L. Taylor (2010). "Applied Conservation Genetics and the Need for Quality Control and Reporting of Genetic Data Used in Fisheries and Wildlife Management." Journal of Heredity 101(1): 1-10. Morrison, T. A., P. I. Chiyo, C. J. Moss and S. C. Alberts (2005). "Measures of dung bolus size for known-age African elephants (Loxodonta africana): implications for age estimation." Journal of Zoology 266: 89-94. Moyzis, R. K., J. M. Buckingham, L. S. Cram, M. Dani, L. L. Deaven, M. D. Jones, J. Meyne, R. L. Ratliff and J. Wu (1988). "A highly conserved repetitive DNA sequence, (TTAGGG)n,
141
present at the telomeres of human chromosomes." Proceedings of the National Academy of Sciences of the United States of America 85: 6622-6626. Munch, K., W. Boomsma, J. P. Huelsenbeck, E. Willerslev and N. Neilsen (2008a). "Statistical Assignment of DNA Sequences using Bayesian Phylogenetics." Systematic Biology 57(5): 750-757. Munch, K., W. Boomsma, E. Willerslev and N. Neilsen (2008b). "Fast phylogenetic DNA barcoding." Philosophical Transactions Of The Royal Society B-Biological Sciences. Munch, K., W. Boomsma, E. Willerslev and N. Neilsen (2008c). "Statistical Assignment of DNA Sequences using Bayesian Phylogenetics." Systematic Biology 57(5): 750-757. Nakagawa, S., N. J. Gemmell and T. Burke (2004). "Measuring vertebrate telomeres: applications and limitations." Molecular Ecology 13: 2523-2533. Nakamura, K., N. Izumiyama-Shimomura, M. Sawabe, T. Arai, Y. Aoyagi, M. Fujiwara, E. Tsuchiya, Y. Kobayashi, M. Kato, M. Oshimura, K. Sasajima, K. Nakachi and K. Takubo (2002). "Comparative Analysis of Telomere Lengths and Erosion with Age in Human Epidermis and Lingual Epithelium." Journal of Investigative Dermatology 119: 1014-1019. Nejstgaard, J. C., M. E. Frischer, P. Simonelli, C. Troedsson, M. Brakel, F. Adiyaman, A. F. Sazhin and L. F. Artigas (2008). "Quantitative PCR to estimate copepod feeding." Marine Biology 153: 565-577. Nelson, M. E. (2002). "The Science, Ethics and Philosophy of Tooth Extractions from Live-Captured White-Tailed Deer: A Response to Fiesta-Bianchet et al. (2002)." Wildlife Society Bulletin 30(1): 284-288. Nettleton, J. A., A. Diez-Roux, N. S. Jenny, A. L. Fitzpatrick and D. R. Jacobs Jr (2008). "Dietary patterns, food groups, and telomere length in the Multi-Ethnic Study of Atherosclerosis(MESA)." American Journal of Clinical Nutrition 88: 1405-1412. Newland, C., I. C. Feild, P. D. Nichols, C. J. A. Bradshaw and M. Hindell, A. (2009). "Blubber fatty acid profiles indicate dietary resource partitioning between adult and juvenile southern elephant seals." Marine Ecology Progress Series 384: 303-312. Newsome, S. D., M. R. Clementz and P. L. Koch (2010). "Using stable isotope biochemistry to study marine mammal ecology." Marine Mammal Science 26(3): 509-572. Nielsen, R. and M. Matz (2006). "Statistical approaches for DNA barcoding." Systematic Biology 55(1): 162-169. Nordfjall, K., A. Larefalk, P. Lindgren, D. Holmberg and G. Roos (2005). "Telomere length and heredity: Indications of paternal inheritance." Proceedings of the National Academy of Sciences of the United States of America 102(45): 16374-16378. Nordfjall, K., U. Svenson, K. Norrback, R. Adolfsson, P. Lenner and G. Roos (2009). "The Individual Blood Cell Telomere Attrition Rate is Telomere Length Dependent." PLoS Genetics 5(2): e1000375.doi:10.1371/journal.pgen.1000375.
142
O'Callaghan, N. J., V. S. Dhillon, P. Thomas and M. Fenech (2008). "A quantitative real-time PCR method for absolute telomere length." Biotechniques 44: 807-809. O'Corry-Crowe, G. M., R. S. Suydam, A. Rosenberg, K. J. Frost and A. E. Dizon (1997). "Phylogeography, population structure and dispersal patterns of the beluga whale Delphinapterus leucas in the western Nearctic revealed by mitochondrial DNA." Molecular Ecology 6(10): 955-970. OEHM, J., A. Juen, K. Nagiller, S. Neuhauser and M. Traugott (2011). "Molecular scatology: how to improve prey DNA detection success in avain faeces." Molecular Ecology Resources: doi:10.1111/j.1755-0998.2011.03001.x. Oftedal, O. T. (2000). "Use of maternal resreves as a lactation stragey in large mammals." Proceedings of the Nutrition Society 59: 99-106. OKazaki, S., K. Tsuchida, H. Maekawa, H. Ishikawa and H. Fujiwara (1993). "Identification of a Pentanucleotide Telomere Sequence, (TTAGG)n, in the Silkworm Bombyx mori and in Other Insects." Molecular and Cellular Biology 13(3): 1424-1432. Olovnikov, A. M. (1973). "A theory of marginotomy." Journal of Theoretical Biology 41: 181-190. Olsen, L. E. and A. D. Yoder (2002). "Using Secondary Structure to Identify Ribosomal Numts: Cautionary Examples from the Human Genome." Molecular Biology and Evolution 19(1): 93-100. Olsson, M., A. Pauliny, E. Wapstra, T. Uller, T. Schwartz and D. Blomqvist (2010). "Sex Differences in Sand Lizard Telomere Inheritance: Paternal Epigenetic Effects Increases Telomere Heritability and Offspring Survival." PLoS One 6(4): e17473. doi:10.1371/journal.pone.0017473. Olsson, M., A. Pauliny, E. Wapstra, T. Uller, T. Schwartz, E. Miller and D. Blomqvist (2011). "Sexual differences in telomere selection in the wild." Molecular Ecology 20(10): 2085-2099. Palomeque, T. and P. Lorite (2008). "Satellite DNA in insects: a review." Heredity 100: 564-573. Parsons, K. M., J. F. Dallas, D. E. Claridge, J. W. Durban, K. C. Balcomb, P. M. Thompson and L. R. Noble (1999). "Amplifying dolphin mitochondrial DNA from faecal plumes." Molecular Ecology 8(10): 1766-1768. Parsons, K. M., S. B. Piertney, S. J. Middlemas, P. S. Hammond and J. D. Armstrong (2005). "DNA-based identification of salmonid prey species in seal faeces." Journal Of Zoology 266: 275-281. Patenaude, N. J. and C. S. Baker (2001). "Population status and habitat use of southern right whales in the sub-Antarctic Auckland Islands of New Zealand." Journal of Cetacean Research & Management Special Issue.(Special Issue 2): 111-116.
143
Pauli, J. N., J. P. Whiteman, B. G. Marcot, T. M. McClean and M. Ben-David (2011). "DNA-based approach to aging martens (Martes americana and M. caurina)." Journal of Mammology 92(3): 500-510. Pauliny, A., R. H. Wagner, J. Augustin, T. Szep and D. Blomqvist (2006). "Age-independent telomere length predicts fitness in two bird species." Molecular Ecology 15: 1681-1687. Pereira, S. L. and A. J. Baker (2004). "Low number of mitochondrial pseudogenes in the chicken (Gallus gallus) nuclear genome: implications for molecular inference of population history and phylogenetics." BMC Evolutionary Biology 4(17): doi:10.1186/1471-2148-4-17. Pierce, G. J. and P. R. Boyle (1991). "A review of methods for diet analysis in picivorous marine mammals." Oceanography and Marine Biology: An Annual Review 29: 409-486. Polz, M. F. and C. M. Cavanaugh (1998). "Bias in template-to-product ratios in multitemplate PCR." Applied And Environmental Microbiology 64(10): 3724-3730. Poulakakis, N., P. Lymberakis, K. Paragamian and M. Mylonas (2005). "Isolation and amplification of shrew DNA from barn owl pellets." Biological Journal Of The Linnean Society 85(3): 331-340. Punt, A. E. (1999). "A full description of the standard BALEEN II model and some varients thereof." Journal of Cetacean Research & Management 1: 267-276. Purcell, M., G. Mackey, E. LaHood, H. Huber and L. Park (2004). "Molecular methods for the genetic identification of salmonid prey from Pacific harbor seal (Phoca vitulina richardsi) scat." Fishery Bulletin 102(1): 213-220. Reilly, J. (2002). "Growth in the Sumatran elephant (Elephas maximus sumatranus) and age estimation based on dung diameter." Journal of Zoology 258(2): 205-213. Riccialdelli, L., S. D. Newsome, M. L. Fogel and R. N. P. Goodall (2010). "Isotopic assessment of prey and habitat preferences of a cetacean community in the southwestern South Atlantic Ocean." Marine Ecology Progress Series 418: 235-248. Richards, J. B., A. M. Valdes, J. P. Gardner, D. Paximadas, M. Kimura, A. Nessa, X. Lu, G. L. Surdulescu, R. Swaminathan, T. D. Spector and A. Aviv (2007). "Higher serum vitamin D concentrations are associated with longer leukocyte telomere length in women." American Journal of Clinical Nutrition 86: 1420-1425. Richly, E. and D. Leister (2004). "NUMTs in Sequenced Eukaryotic Genomes." Molecular Biology and Evolution 21(6): 1081-1084. Ridgeway, S. H., R. F. Green and J. C. Sweeny (1975). "Mandibular anesthesia and tooth extraction in the bottlenosed dolphin." Journal of Wildlife Diseases 11: 415-418. Rodrigue, K. L., B. P. May, T. R. Famula and M. E. Delany (2005). "Meiotic instability of chicken ultra-long telomeres and mappings of a 2.8 megabase arrays to the W-sex chromosome." Chromosome Research 13(6): 581-591.
144
Rolland, R. M., K. E. Hunt, S. D. Kraus and S. K. Wasser (2005). "Assessing reproductive status of right whales (Eubalaena glacialis) using fecal hormone metabolites." General and Comparative Endocrinology 142(3): 308-317. Rosel, P. E., L. Hansen and A. A. Hohn (2009). "Restricted dispersal in a continuously distributed marine species: common bottlenose dolphins Tursiops truncatus in coastal waters of the western North Atlantic." Molecular Ecology 18(24): 5030-5045. Ross, R. A., S. Murugan and W. L. S. Li (2008). "Testing the reliability of genetic methods of species identification via simulation." Systematic Biology 57(2): 216-230. Rubenstein, D. R. and K. A. Hobson (2004). "From birds to butterflies: animal movement patterns and stable isotopes." Trends In Ecology & Evolution 19(5): 256-263. Rubinoff, D., S. Cameron and K. Will (2006). "A genomic perspective on the shrtcomings of mitochondrial DNA for "barcoding" identification." Journal Of Heredity 97(6): 581-594. Rufer, N., T. H. Brümmendorf, S. Kolvraa, C. Bischoff, K. Christensen, L. Wadsworth, M. Schulzer and P. M. Lansdorp (1999). "Telomere Fluorescence Measurements in Granulocytes and T Lymphocyte Subsets Point to a High Turnover of Hematopoietic Stem Cells and Memory T Cells in Early Childhood." Journal of Experimental Medicine 190(2): 157-167. Ruiz-Cooley, R. I., D. Gendron, S. Aguiniga, S. Mesnick and J. D. Carriquiry (2004). "Trophic relationships between sperm whales and jumbo squid using stable isotopes of C and N." Marine Ecology Progress Series 277: 275-283. Salomons, H. M., G. A. Mulder, L. van de Zande, M. F. Haussmann, M. H. K. Linskens and S. Verhulst (2009). "Telomere shortening and survival in free-living corvids." Proceedings of the Royal Society B-Biological Sciences 276: 3157-3165. Sambrook, J., E. F. Fritsch and T. Maniatis (1989). Molecular Cloning: A laboratory Manual Cold Spring Harbor Laboratory Press.Cold Spring Harbor, NY. Santos, M. B., M. R. Clarke and G. J. Pierce (2001). "Assessing the importance of cephalopods in the diets of marine mammals and other top predators: problems and solutions." Fisheries Research 52: 121-139. Schmeling, A., G. Geserick, W. Reisinger and A. Olze (2007). "Age estimation." Forensic Science International 165: 178-181. Scott, N. M., M. F. Haussmann, R. M. Elsey, P. L. Trosclair III and C. M. Vleck (2006). "Telomere length shortens with body length in Alligator mississippiensis." Southeastern Naturalist 5(4): 685-692. Serra, V., T. von Zglinicki, M. Lorenz and G. Saretzki (2003). "Extracellular superoxide dismutase is a major antioxidant in human fibroblasts and slows telomere shortening." Journal of Biological Chemistry 278: 6824-6830. Sheppard, S. K. and J. D. Harwood (2005). "Advances in molecular ecology: tracking trophic links through predator-prey food-webs." Functional Ecology 19(5): 751-762.
145
Simon, C., F. Frati, B. A and e. al. (1994). "Evolution, weighing, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers." Annals of the Entomological Society of America 87: 651-701. Slagboom, P. E., S. Droog and D. I. Boomsma (1994). "Genetic determination of telomere size in humans: a twin study of three age groups." American Journal of Human Genetics 55: 876-882. Smokler, R., A. Richards, R. C. Connor, J. Mann and P. Berggren (1997). "Sponge carrying by dolphins (Delphinidae, Tursiops sp.): A foraging specialization involving tool use?" Ethology 103: 454-465. Song, H., J. E. Buhay, M. F. Whiting and K. A. Crandall (2008). "Many species in one: DNA barcoding overestimates the number of species when nuclear mitochondrial pseudogenes are coamplified." Proceedings of the National Academy of Sciences of the United States of America 105(36): 13486-13491. Sorenson, M. D. and R. C. Fleischer (1996). "Multiple independent trans-positions of mitochondrial DNA control region sequences to the nucleus." Proceedings Of The National Academy Of Sciences Of The United States Of America 93: 15239-15243. Sorenson, M. D. and T. W. Quinn (1998). "Numts: A challenge for avian systematics and population biology." The Auk 115(1): 214-221. Springer, A., J. Estes, G. van Vliet, T. Williams, D. Doak, E. Danner and B. Pfister (2008). "Mammal-eating killer whales, industrial whaling, and the sequential megafaunal collapse in the North Pacific Ocean: A reply to critics of Springer et al. 2003." Marine Mammal Science 24(2): 414-442. Stearns, S. C. (1992). The Evolution of Life Histories Oxford University Press.Oxford. Steinke, D., M. Vences, W. Salzburger and A. Meyer (2005). "TaxI: a software tool for DNA barcoding using distance methods." Philosophical Transactions Of The Royal Society B-Biological Sciences 360(1462): 1975-1980. Stephens, D. W. and J. R. Krebs (1986). Foraging Theory Princeton University Press.Princeton. Storr, G. M. (1961). "Microscopic analysis of faeces, a technique for ascertaining the diet of herbivorous mammals." Australian Journal of Biological Sciences 14(1): 157-164. Sunnucks, P., A. C. C. Wilson, L. B. Beheregaray, K. Zenger, J. French and A. C. Taylor (2000). "SSCP is not so difficult: the application and utility of single-stranded conformational polymorphism in evolutionary biology and molecular ecology." Molecular Ecology 9: 1699-1710. Sutherland, R. M. (2000) Molecular analysis of avian diets. PhD Thesis. University of Oxford. Oxford. Symondson, W. O. C. (2002a). Diagnostic techniques for determining carabid diets. Agroecology Of Carabid Beetles. Andover, Intercept Ltd Scientific, Technical & Medical Publishers: 137-164.
146
Symondson, W. O. C. (2002b). "Molecular identification of prey in predator diets." Molecular Ecology 11(4): 627-641. Takasaki, T., A. Tsuji, N. Ikeda and M. Ohishi (2003). "Age estimation in dental pulp DNA based on human telomere shortening." International Journal of Legal Medicine 117: 232-234. Team, R. D. C. (2010). R: A language and environment for statistical computing R. F. f. S. Computing.Vienna, Austria. Thalmann, O., J. Hebler, H. N. Poinar, S. Paabo and L. Vigilant (2004a). "Unreliable mtDNA data due to nuclear insertions: a cautionary tale from analysis of humans and other great apes." Molecular Ecology 13: 321-335. Thalmann, O., J. Hebler, H. N. Poinar, S. Paabo and L. Vigilant (2004b). "Unreliable mtDNA data due to nuclear insertions: a cautionary tale from analysis of humans and other great apes." Molecular Ecology 13(2): 321-335. Thalmann, O., D. Serre, M. Hofreiter, D. Lukas, J. Eriksson and L. Vigilant (2005). "Nuclear insertions help and hinder inference of the evolutionary history of gorilla mtDNA." Molecular Ecology 14(1): 179-188. Tollit, D. J., A. D. Schulze, A. W. Trites, P. F. Olesiuk, S. J. Crockford, T. S. Gelatt, R. R. Ream and K. M. Miller (2009). "Development and application of DNA techniques for validating and improving pinniped diet estimates." Ecological Applications 19(4): 889-905. Torres, L. G. and A. J. Read (2009). "Where to catch a fish? The influence of foraging tactics on the ecology of bottlenose dolphins (Tursiops truncatus) in Florida Bay, Florida." Marine Mammal Science 25(4): 797-815. Triant, D. A. and J. A. Dewoody (2007). "The occurrence, detection, and avoidance of mitochondrial DNA translocations in mammalian systematics and phylogeography." Journal Of Mammalogy 88(4): 908-920. Trites, A. W. and R. Joy (2005). "Dietary analysis from fecal samples: How many scats are enough?" Journal of Mammalogy 86(4): 074-712. Tsuji, A., A. Ishiko, T. Takasaki and N. Ikeda (2002). "Estimating age of humans based on telomere shortening." Forensic Science International 126: 197-199. Ujvari, B. and T. Madsen (2009). "Short telomeres in Hatchling Snakes: Erythrocyte Telomere Dynamics and Longevity in Tropical Pythons." PLoS One 4(10): e7493. doi:10.1371/journal.pone.0007493. Unryn, B. M., L. S. Cook and K. T. Riabowol (2005). "Paternal age is positively linked to telomere length of children." Aging Cell 4: 97-101. Valdes, A. M., T. Andrew, J. P. Gardner, M. Kimura, E. Oelsner, L. F. Cherkas, A. Aviv and T. D. Spector (2005). "Obesity, cigarette smoking, and telomere length in women." Lancet 366: 662-64.
147
Valenzuela, L. O., M. Sironi, V. J. Rowntree and J. Seger (2009). "Isotopic and genetic evidence for culturally inherited site fidelity to feeding grounds in southern right whales (Eubalaena australis)." Molecular Ecology 18(5): 782-791. Vences, M., M. Thomas, A. van der Meijden, Y. Chiari and D. R. Vieites (2005). "Comparative performance of the 16S rDNA gene in DNA barcoding of amphibians." Frontiers in Zoology 2(5): doi:10.1186/1742-9994-2-5. Venkatesh, B., N. Dandona and S. Brenner (2006). "Fugu genome does not contain mitochondrial pseudogenes." Genomics 87(2): 307-310. Vestheim, H. and S. N. Jarman (2008). "Blocking primers to enhance PCR amplification of rare sequences in mixed samples - a case study on prey DNA in Antarctic krill stomachs." Frontiers in Zoology 5(12): doi:10.1186/1742-9994-5-12. Vitt, L. J. and E. R. Pianka (2005). "Deep history impacts present-day ecology and biodiversity." Proceedings of the National Academy of Sciences of the United States of America 102(22): 7877-7881. Vleck, C. M., M. F. Haussmann and D. Vleck (2003). "The natural history of telomeres: tools for aging animals and exploring the aging process." Experimental Gerontology 38: 791-795. von Wintzingerode, F., O. Landt, A. Ehrlich and U. B. Gobel (2000). "Peptide nucleic acid-mediated PCR clamping as a useful supplement in the determination of microbial diversity." Applied And Environmental Microbiology 66(2): 549-557. Wade, P. R., V. N. Burkanov, M. E. Dahlheim, N. A. Friday, L. W. Fritz, T. R. Loughlin, S. A. Mizroch, M. M. Muto, D. W. Rice, L. G. Barrett-Lennard, N. A. Black, A. M. Burdin, J. Calambokidis, S. Cerchio, J. K. B. Ford, J. K. Jacobsen, C. O. Matkin, D. R. Matkin, A. V. Mehta, R. J. Small, J. M. Straley, S. M. McCluskey, G. R. VanBlaricom and P. Clapham (2007). "Killer whales and marine mammal trends in the North Pacific - A re-examination of evidence for sequential megafauna collapse and the prey-switching hypothesis." Marine Mammal Science 23(4): 766-802. Wang, S. W., S. J. Iverson, A. M. Springer and S. A. Hatch (2009). "Spatial and temporal diet segregation in northern fulmars (Fulmarus glacialis) breeding in Alaska: insights from fatty acid signatures." Marine Ecology Progress Series 377: 299-307. Watson, J. D. (1972). "Origin of concatameric T7 DNA." Nature New Biology 239: 197-201. Wells, R. S., H. L. Rhinehart, L. J. Hansen, J. C. Sweeny, F. I. Townsend, R. Stone, D. R. Casper, M. D. Scott, A. A. Hohn and T. K. Rowles (2004). "Bottlenose Dolphins as Marine Ecosystem Sentinels: Developing a Health Monitoring System." EcoHealth 1: 246-254. Western, D. (1979). "Size, life history and ecology in mammals." African Journal of Ecology 147: 185-204. Will, K. W. and D. Rubinoff (2004). "Myth of the molecule: DNA barcoding for species cannot replace morphology for identifications and classification." Cladistics 20: 47-55.
148
Woischnick, M. and C. T. Moraes (2002). "Pattern of Organization of Human Mitochondrial Pseudogenes in the Nuclear Genome." Genome Research 12(885-893). Wright, B. E. (2010). "Use of chi-square tests to analyse scat-derived diet composition data." Marine Mammal Science 26(2): 395-401. Ying, T. C. (2005) Evaluation of telomere length as an age marker in marine teleosts. MSc Thesis. University of Hong Kong. Hong Kong. Zaidi, R. H., Z. Jaal, N. J. Hawkes, J. Hemingway and W. O. C. Symondson (1999). "Can multiple-copy sequences of prey DNA be detected amongst the gut contents of invertebrate predators?" Molecular Ecology 8: 2081-2087. Zhang, W., C. Yi, W. Bao, B. Lui, J. Cui, H. Yu, X. Cao, M. Gu, M. Lui and Z. Cheng (2005). "The Transcribed 165-bp CentO Satellite Is the Major Functional Centromeric Element in the Wild Rice Species Oryza punctata." Plant Physiology 139: 306-315. Zischler, H. (2000). "Nulcear integrations of mitochondrial DNA in primates: Inferences of associated mutational events." Electrophoresis 21: 531-536. Zug, G. R., A. H. Wynn and C. Ruckdeschel (1986). Age Determination of Loggerhead Sea Turtles, Caretta caretta, by Incremental Growth Marks in the Skeleton.Smithsonian Contributions to Zoology. Number 427.Smithsonian Institution Press.City of Washington. Zuur, A. F., E. N. Ieno, N. J. Walker, A. A. Saveliev and G. M. Smith (2009). Mixed Effects Models and Extensions in Ecology with R.Statistics for Biology and Helth.M. Gail, K. Krickeberg, J. M. Samet, A. Tsiatis and W. Wong Springer.New York.