Conservation Genetics 2: 63–67, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

63

Australian lungfish (Neoceratodus forsteri: Dipnoi) have low genetic variation at allozyme and mitochondrial DNA loci: a conservation alert? Francesca D. Frentiu1 , Jenny R. Ovenden2∗ & Raewyn Street2 1 Department

of Zoology & Entomology, University of Queensland, St. Lucia, Brisbane, Queensland 4072 Australia; 2 Molecular Fisheries Laboratory, Southern Fisheries Centre, Queensland Department of Primary Industries, P.O. Box 76, 13 Beach Road, Deception Bay, Queenland 4508 Australia (∗ author for correspondence: E-mail: [email protected])

Received 9 October 2000; accepted 22 December 2000

Key words: Australian lungfish, allozymes, conservation, mitochondrial DNA, Neoceratodus forsteri, population genetics

Abstract Genetic variation at allozyme and mitochondrial DNA loci was investigated in the Australian lungfish, Neoceratodus forsteri Krefft 1870. Tissue samples for genetic analysis were taken non-lethally from 278 individuals representing two spatially distinct endemic populations (Mary and Burnett rivers), as well as one population thought to be derived from an anthropogenic translocation in the 1890’s (Brisbane river). Two of 24 allozyme loci resolved from muscle tissue were polymorphic. Mitochondrial DNA nucleotide sequence diversity estimated across 2,235 base pairs in each of 40 individuals ranged between 0.000423 and 0.001470 per river. Low genetic variation at allozyme and mitochondrial loci could be attributed to population bottlenecks, possibly induced by Pleistocene aridity. Limited genetic differentiation was detected among rivers using nuclear and mitochondrial markers suggesting that admixture may have occurred between the endemic Mary and Burnett populations during periods of low sea level when the drainages may have converged before reaching the ocean. Genetic data was consistent with the explanation that lungfish were introduced to the Brisbane river from the Mary river. Further research using more variable genetic loci is needed before the conservation status of populations can be determined, particularly as anthropogenic demands on lungfish habitat are increasing. In the interim we recommend a management strategy aimed at conserving existing genetic variation within and between rivers.

Introduction The Australian lungfish Neoceratodus forsteri Krefft 1870 is one of five extant representatives of the ancient and once speciose air-breathing Dipnoan lineage. It is the sole remaining Australian representative of this group and it is also the most morphologically primitive of the extant Dipnoans (Jamieson 1991; Rock et al. 1996). Fossils indicate that the distribution of N. forsteri reached the centre of the Australian continent prior to the Pleistocene (Kemp 1997). However, it is now restricted to the southeast corner of Queensland where it occurs naturally in the Burnett and Mary Rivers. Translocated popu-

lations exist in nearby rivers, some originating in the nineteenth century (Illidge 1892; O’Connor 1895). Although legally protected, the Australian lungfish is potentially threatened as its range coincides with some areas that are extensively utilized for agriculture. Altered flow regimes for irrigation purposes could lead to the destruction of existing population structure, enhancing drift and causing loss of genetic variation. Loss of genetic variation may lead to inbreeding depression, inducing a decline in population size and retarding evolutionary potential, possibly culminating in extinction (Frankel and Soule 1981; O’Brien 1994). This study reports on the extent of genetic diversity within and among three populations of Australian

64

Figure 1. Australian lungfish (Neoceratodus forsteri) sampling sites on Burnett, Mary and Brisbane rivers, south-east Queensland. Australian coastline data is copyright Commonwealth of Australia, provided by AUSLIG.

lungfish at allozyme and mitochondrial DNA loci. Spatial isolation of this saltwater intolerant fish in separate catchments might suggest the presence of at least two evolutionarily significant units (sensu Mortiz 1995) which, if present, would have implications for fisheries management. The genetic consequences of translocation were also measured by surveying genetic variation in the Brisbane River population that may have arisen from a translocation of five Mary River fish in the nineteenth century (Illidge 1892).

Methods Lungfish were electrofished in 1998 from endemic populations in the Burnett (105 fish from six localities) and Mary (103 fish from five localilties) rivers, and

one population thought to have a translocated origin (Brisbane river, 70 fish from three localities, Figure 1). Allozymes were assayed in all samples. Mitochondrial DNA diversity was assayed in 13 randomly chosen fish from each of the Mary and the Burnett rivers and 14 from the Brisbane river. Muscle tissue was collected non-lethally using biopsy techniques as lungfish are protected in Queensland, and preliminary experiments showed that 24 out of 31 potential allozyme loci could be resolved from muscle. Allozymic variation was screened using horizontal starch gel electrophoresis (Shaklee and Keenan 1986; Frentiu 1998). Twenty four loci were resolved from the enzymes AAT (EC 2.6.1.1; 2 loci), ADH (EC 1.1.1.1), AK (EC 2.7.4.3; 2 loci), CK (EC 2.7.3.2), ENO (EC 4.2.1.11), EST-D (EC 3.1.1.1), FH (EC 4.2.1.2), GAPDH (EC 1.2.1.12), GPI (EC 5.3.1.9), G-3-PDH (EC 1.1.1.8), G-6-PDH

65 (EC 1.1.1.49), IDH (EC 1.1.1.42; 2 loci), LAP (EC 3.4.11.1), LDH (EC 1.1.1.27), MDH (EC 1.1.1.37), ME (EC 1.1.1.40), MPI (EC 5.3.1.8; 2 loci), PGM (EC 2.7.5.1), PK (EC 2.7.1.40), STDH (EC 1.5.1.22), TPI (EC 5.3.1.1). Genomic DNA for use as PCR template was extracted using proteinase K and Chelex 100 resin (Biorad Laboratories P/L) as a chelating agent. The control region and flanking sequence was amplified by primers NEO.12S441.R (TAT AGA CGG TAG TGG CAA GAA GCG, 5’ to 3’) and NEO.CB212.F (TAC TAC GGG TCA TAC CTA TAC AAG, 5’ to 3’) that were designed within existing cytochrome b and the 12S rRNA N. forsteri sequence from GenBank (Acc. Nos. Z21928 and Z21927). Subsequent sequencing allowed the design of three sets of primer pairs that spanned 1350 bp of the control and flanking regions. The pairs were NEO.CR1013.F (CCG CCG CAG ACT AAA ATA GAA CTT) and NEO.12S615.R (GCC AGG ACC AAA CCT TTA TGC TCA), NEO.CR139.F (ACA TTC CTG GCA TTA ACG GCT AGT) and NEO.CR754.R (AAG TTC TAT TTT AGT CTG CGG CGG) and NEO.CR587.F (ACG AGC CAA CAC ATT CCG ACC ATT) and NEO.CR482.R (GTA CTA GCC GTT AAT GCC AGG AAT). Sequence from a protein coding region that included ATPase genes (885 bp) was obtained using primer pairs COIII.2 (GTT AGT GGT CA(GT) GGG CTT GG(AG)) and NEO.ATP343.R (GAT TTA CGC AAC AAC TGA TAC AAC) and NEO.ATP444.R (TGT GAA AGT GTA AGG AAG GAG CCC) and ATP8.2 (AAA GC(AG) T(CT)(AG) GCC TTT TAA GC). Primers COIII.2 and ATP8.2 were designed by E. Bermingham (http://nmg.si.edu/bermlab.html). Sequences were aligned and edited with Sequencher version 3.1.1. Allozyme allele frequencies, including the occurrence of private or rare alleles and heterozygosities were estimated for each sample. Polymorphic allozyme loci were tested for significant deviations from Hardy-Weinberg equilibrium. Allele frequency differences among rivers was tested using Fisher’s exact test (10,000 permutations). We calculated pairwise mtDNA sequence diversity between individuals. The molecular distance among sequences was calculated according to the Kimura 2-parameter method (Kimura 1980). We used Weir and Cockerham’s (1984) framework for analysis of molecular variance (AMOVA) to evaluate the partitioning of genetic variation at nuclear and mtDNA loci among rivers. Signific-

Table 1. Australian lungfish (Neoceratodus forsteri) allele frequencies and within locus heterozygosities (H) for two allozyme loci (GPI and PGM) for three catchments in south-east Queensland. Significant differences between Hexpected and Hobserved are denoted by an asterisk, with signficance values in brackets. Locus

Brisbane (N = 70)

Burnett (N = 105)

Mary (N = 103)

GPI A Hexpected Hobserved

0.725 0.489 0.429

0.586 0.401 0.385

PGM A Hexpected Hobserved

0.847 0.374 0.319

0.754 0.261 0.180∗ (p = 0.005)

0.737 0.390 0.290∗ (p = 0.036) 0.777 0.349 0.255∗ (p = 0.020)

ance was determined using 10,000 permutations with Arlequin 1.1 (Schneider et al. 1997).

Results and discussion The Australian lungfish displayed low allelic diversity at allozyme and mtDNA loci and minimal genetic differentiation among spatially distinct catchments. In a sub-sample of 42 individuals, two of 24 allozyme loci (GPI and PGM) were polymorphic, yielding an average heterozygosity across all loci of 0.030. Across all samples, two alleles were found at each polymorphic locus (Table 1). Fifteen polymorphic positions were found among 2,235 base pairs from the mitochondrial genome of 40 lungfish revealing eight haplotypes (Table 2). Sequence diversity was 0.000423 (±0.000356) in the Burnett, 0.001470 (±0.000931) in the Mary and 0.000515 (±0.000397) in the Brisbane samples. Five mtDNA haplotypes were identified in the samples from the Burnett river and five from the Mary river. Two only were found in the samples from the Brisbane river. A low level of overall nuclear and mitochondrial genetic variation was somewhat unexpected given its current abundance (Brooks and Kind 2000), geographic separation of drainages and the long evolutionary history of the species in eastern Australia. Severe or prolonged reductions in population size may account for this finding. Genetic drift is known to rapidly deplete genetic variation during population bottlenecks (for example, Ovenden and White 1990).

66 Table 2. Character state table for Neoceratodus forsteri mtDNA partial sequence for 2,235 nucleotides of the control (1350 bp consisting of the 5’ part of cytochrome b, tRNAthr and tRNApro , 927 bp of the control region and 3’ part of tRNAphe , GenBank AF344663) and ATPase (885 bp consisting of 5’ end of tRNAlys , ATPase8 and 3’ part of ATPase6, GenBank AF344662) regions. Each character consists of a single nucleotide substitution or indel (359, 861) that were polymorphic among the 40 N. forsteri sampled. H’type

Common A B C D E F G

River

Control region

ATPase region

Brisbane

Burnett

Mary

131

264

355

359

859

861

899

209

339

404

413

440

632

675

744

13 0 0 0 0 0 0 1

8 2 1 0 0 1 1 0

5 3 0 1 2 0 0 2

G G G G G A A G

T T T T T T T C

C C C C C C C T

— — — — — A A —

T T T T T T T A

G G G G G G G —

A A A A A A A G

G A G A G G G G

G G G G G A G G

G G G G G G G A

A A A A A A A G

G G G G G G G A

A A A A A A A G

C C T C C C C C

C C C T T C C C

Furthermore, two alleles were observed at each of the two polymorphic nuclear loci, possibly indicating that if they ever existed, low frequency alleles may have been lost presumably via enhanced genetic drift during the bottleneck event. We also observed a somewhat truncated distribution of mtDNA haplotypes, with one haplotype (G) being up to nine character state changes removed from the common haplotype with no intermediaries (Table 2). This may indicate that ‘pruning’ of mtDNA lineages occurred during past bottleneck events. Kemp (1991) regards Pleistocene aridity as the primary cause of lungfish range contractions. Episodic, or prolonged, drought both ancient or recent may have been responsible for reductions in population size in the Mary and Burnett rivers. Post-bottleneck recovery may have been slow due to long generation times (Brooks and Kind 2000) and high predator vulnerability of juveniles (Bancroft 1928). Tests for population subdivision among rivers indicated most genetic variation to be within, not between, populations. Pairwise tests showed no significant allozyme allele frequency differences between the Mary and the Burnett or between the Mary and the Brisbane populations, but they did occur between the Burnett and the Brisbane populations (p = 0.035). No private alleles were found. A small but significant FST (0.0188, p = 0.006) was found following AMOVA on nuclear loci, possibly due to the differences between the Burnett and Brisbane populations. No pair-wise differentiation was found between populations at mtDNA markers and the overall Fst of 0.044 was not significant.

The minimal genetic differentiation found among lungfish populations was surprising given the current geographical separation of the catchments. However, confluence of the lower reaches of the Burnett and the Mary rivers in the Hervey Bay region may have occurred at the height of the Pleistocene when sea levels were lower than present (Chappell 1987). Coastal bathymetric data to test for past confluence of drainage channels are not available. During the period 15,000 to 8,000 years before present, admixture of lungfish populations in the Mary and Burnett rivers may have been possible. Confluence and admixture of drainages in this area of Queensland is suggested by a similar pattern found in another freshwater species, Nannoperca oxleyana (Hughes et al. 1999). Together, the overall low genetic variation and lack of subdivision between the Mary and Burnett rivers suggests that the severe population reductions which left lungfish populations genetically depauperate occurred prior to, or during, the proposed confluence of the Mary and Burnett river systems. The forces of mutation, drift and selection have not led to significant regional population differentiation since lungfish became geographically isolated in adjacent catchments by rising sea levels. Significant departures from Hardy Weinberg equilibrium, suggesting a deficiency of heterozygotes, were found at the two nuclear loci in the Mary river population and for the PGM locus in the Burnett river sample (Table 1). This may be indicative of inbreeding, which can be associated with inbreeding depression (for example, Coltman et al. 1998; Eldredge et al. 1999). However further infor-

67 mation on population sizes and life history strategies in lungfish is necessary to assess this proposal. An alternative, but less likely, explanation is the presence of a Wahlund effect, but a more polymorphic set of genetic markers is needed before this can be tested. The low level of polymorphism among the genetic markers used in this study could not confirm a translocated origin for the Brisbane river population. However, the Brisbane population shared a rare haplotype (haplotype G, Table 2) with the Mary river population, a finding most consistent with a translocation scenario. This is indirect evidence that the Mary and the Burnett populations were the only lungfish populations in existence at the time of European settlement (for example, Illidge 1892; O’Connor 1895). The observed low levels of genetic variation at two classes of neutral, independent markers would suggest a paucity of variability across the whole lungfish genome, presumably including loci directly linked to fitness such as those implicated in disease resistance. Further studies using more powerful genetic tools such as microsatellites or AFLPs (Mueller and Wolfenbarger 1999) are recommended particularly as low genetic variation is implicated in population declines (Frankel and Soulé 1981), and anthropogenic effects on lungfish habitat is escalating. Priority should be given to ranking the populations according to their conservation significance and an evaluation of the genetic implications of population enhancement including the establishment of gene flow among populations and further translocations.

Acknowledgements The authors thank Steven Brooks, Peter Kind and Myles Waller for assistance with the collection of genetic material; the Queensland Department of Natural Resources for financial and logistic support; Dr. Anne Kemp and Professor Gordon Grigg for advice and discussions and Kate Yeomans for mapping skills. We thank Lisa Pope, Gordon Grigg, Jean Joss, Michael Hutchison, Mike Dredge and two anonomyous reviewers for their assistance with this manuscript.

References Bancroft TL (1928) On the life history of Ceratodus. Proc. Linn. Soc. NSW, 53, 315–317.

Brooks S, Kind P (2000) Lungfish and General Fisheries Surveys in the Burnett River – Annual Progress Report. March, 2000 Department of Primary Industries, Southern Fisheries Centre, Queensland, Australia. Chappell J (1987) Late Quaternary sea-level changes in the Australian region. In: Sea Level Changes (eds. Tooley MJ, Shennan I), pp. 297–331. Blackwell, Oxford, UK. Coltman DW, Bowen WD, Wright JM (1998) Birth weight and neonatal survival of harbour seal pups are positively correlated with genetic variation measured by microsatellites. Phil. Trans. R. Soc. Lond. B, 265, 803–809. Eldredge MDB, King JM, Loupis AK, Spencer BSP, Taylor AC, Pope LC, Hall GP (1999) Unprecedented low levels of variation of the black footed rock wallaby. Conserv. Biol., 13, 531–541. Frankel OH, Soulé ME (1981) Conservation and Evolution. Cambridge University Press, Cambridge. Frentiu, FD (1998) Patterns of Genetic Diversity in the Australian Lungfish. Honours Thesis, Zoology Department, University of Queensland. Hughes J, Ponniah M, Hurwood D, Chenoweth S, Arthington A (1999) Strong genetic structuring in a habitat specialist, the Oxleyan Pygmy Perch Nannoperca oxleyana. Heredity, 83, 5–14. Illidge T (1892/ 1984) Report to the Royal Society of Queensland. Proc. Roy. Soc. Qld., X, 41–44. Jamieson BG (1991) Fish Evolution and Systematics: Evidence from Spermatozoa. Cambridge University Press, Cambridge. Kemp A (1991) Australian Mesozoic and Cainozoic Lungfish. In: Vertebrate Palaeontology of Australasia (eds. Vickers-Rich P, Monaghan JM, Baird RF, Rich TH), pp. 465–497. Pioneer Design Studio Pty. Ltd., Victoria. Kemp A (1997) A revision of Australian Mesozoic and Cenozoic Lungfish of the Family Neoceratodidae (Osteichthyes: Dipnoi) with a description of four new species. J. Paleontol., 71, 713– 733. Kimura M (1980) A simple method for estimating the evolutionary rate of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol., 16, 111–120. Mortiz C (1995) Uses of molecular phylogenies for conservation. Phil. Trans. R. Soc. Lond. B, 349, 113–118. Mueller UG, Wolfenbarger LL (1999) AFLP genotyping and fingerprinting. Trends in Ecol. & Evol., 14, 389–394. O’Brien SJ (1994) A role for molecular genetics in biological conservation. Proc. Natl. Acad. Sci. (USA), 91, 5748–5755. O’Connnor D (1895) Letter to the Royal Society of Queensland. Proc. Roy. Soc. Qld., xi, xxi. Ovenden J, White R (1990) Mitochondrial and allozyme genetics of incipient speciation in a landlocked population of Galaxias truttaceus (Pisces: Galaxiidae). Genetics, 124, 701–716. Rock J, Eldridge M, Champion A, Johnston P, Joss J (1996) Karyotype and nuclear DNA content of the Australian lungfish Neoceratodus forsteri (Neoceratodidae: Dipnoi). Cytogenet. Cell Genetics, 73, 187–189. Schneider S, Kueffer JM, Roessli D, Excoffier L (1997) Arlequin ver. 1.1: A Software for Population Genetic Data Analysis. Genetics and Biometry Laboratory, University of Geneva, Switzerland. Shaklee JB, Keenan CP (1986) A Practical Laboratory Guide to the Techniques and Methodology for Electrophoresis and its Application to Fish Fillet Identification. CSIRO Marine Laboratories Report 177. Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358–1370.