The population genetics of the Mute Swan Cygnus olor in Ireland

Transcription

1 The population genetics of the Mute Swan Cygnus olor in Ireland PAUL D. O DONOGHUE, JOHN O HALLORAN, PHILIP J. BACON, PATRICK SMIDDY and THOMAS F. CROSS Mute Swan population genetics were investigated from four localities throughout Ireland using protein electrophoresis. The products o f ten presum ptive lo c i were assayed in red blood cells and one in blood plasma. A ll but the latter (plasm a esterase) were m onom orphic. A lle le frequencies at the EST locus did not differ significantly from previously published values from E n gland and S cotla n d n o r was there a s ig n ific a n t s p a tia l o r te m p o ra l heterogeneity apparent within Ireland. These results do n ot support the hypothesis o f reproductive isolation o f Irish swan populations from England and Scotland o r within Ireland. The previously recorded presence o f lim ited variability in the English and Scottish Mute Swan populations at one locus (Lactate dehydrogenase) com pared to those assayed in Ireland (w here it was m on om orphic) may mean that introduction(s) involved a lim ited num ber o f individuals. The im portance of biochem ical genetics in the understanding of population ecolo gy and dynam ics is w ell described (Lewtonin 1972, Berry 1977). The population genetics of the Mute Swan Cygnus o lo r has been investigated in England and Scotland by Bacon (1980a, 1981), w ho stated that a greater know ledge of Mute Swan genetics m ight be useful in the understanding of population dynam ics (Bacon 1980b). Such studies have since elucidated som e aspects of Mute Swan breeding b iolo gy (Birkhead et al. 1983, Bacon & Andersen- Harild 1987, W alter et al. 1991). In Ireland, the Mute Swan is w idely distributed with population estim ates varying betw een five to seven thousand individuals (M on val & Pirot 1989). T w o recent studies (Collins 1991, Sm iddy & O Halloran 1991) give details of the breeding biolo g y of Irish Mute Swans. M ovem ents betw een Ireland and Britain in general appear rare. Only on e Mute Swan ringed in Ireland has been recovered in W ales (A n glesey) and one ringed in England recovered in Ireland (C ork). Thus it has been suggested b y Sm iddy & O Halloran (1991) that the tw o populations m ay be considered genetically separate. H ow ever, a high resighting rate of Outer Hebridian Mute Swans on the northern coast of Ireland (14 birds colour-ringed in the W estern Isles, Scotland have been resighted on the north coast of Ireland (Hutchinson 1989, Spray 1981)), suggests that mixing of the populations m ay be greater than thought. These sightings seem to be the exception to the ob served patterns of m ovem ent, given the few recoveries from the many other studies of swans in England (Birkhead & Perrins 1986). The Irish Mute Swan appears to be sedentary, though v e ry little is known about swan m ovem ents in Ireland. O Halloran & Collins (1985) in a prelim inary analysis of swan recoveries in Ireland have reported that 90% (n = 61) of ringed birds travelled less than 65 km and no adult travelled this distance. This is similar to m ovements observed in England (O gilvie 1967). Unlike the situation in England, Scotland and W ales (Ticeh u rst 1957), little is known historically about th e Irish Mute Swan, though introduced birds w ere known to be present in Cork in the early 1700s (Smith 1750). Thom pson (1851) reported the species from around Belfast and stated that it was not known in the w ild state, but only in collection s, and that he was uncertain of the date of its 5 W ildfow l 43 (1992 ): 5-11

2 6 M ute Swan Genetics introduction. All authors in the present century follo w ed Thom pson (1851) in regarding the species as an introduction although Collins & W helan (1990) questioned this. T h ey stated that the spread of the Mute Swan in the 19th and 20th centuries could be as a result of the decline in persecution rather than the spread of the species follow ing an introduction. This has happened follow ing protection in Denmark since c. 1930, and Poland and the USSR since In Denmark populations have grown exponentially, doubling in the last 60 years (Bloch 1971, W ieloch 1991). Similar population increases have been observed in introduced swan populations in North America. Protein electrophoresis provides a useful measure of genetic variability and com position (b y the study of allelic distribution of polym orphic loci) (Ferguson 1980). This approach has been used by Bacon (1980a, 1981) in his studies of Mute Swan population genetics in England and Scotland, but not with Irish populations. This study set out to determ ine the level of genetic variability in Irish Mute Swan populations using protein electrophoresis and possibly to provide evidence on the scale of introduction of Mute Swans to Ireland. W e describe the genetic variability of the Irish Mute Swan population as shown in particular for the flock at Cork Lough, and com pare the variability to other sites within Ireland and with sites in Britain (Bacon 1980a, 1981). Material and methods Sample collection Mute Swans w ere captured, under Figure 1. Location of sampling sites for Mute Swans and the pattern of esterase gene frequencies in Irelan d an d Britain; num bers refe r to locations nam ed in T able 2. (D ata fo r Britain recalcu lated from Bacon 1981.)

3 M ute Swan Genetics 7 licence, at four sites (Fig. 1 and Tables 2 and 3) during the w inter of 1986 (Belfast and Cork Lou gh ) and 1990 (C ork Lough, Lim erick and G alway). The data refer to sam ples of swans caught from w intering flocks. Such flocks com prise juveniles, non-breeders and failed breeders from the surrounding area. Family data, i.e. related individuals w h ose gen otypes do not th erefore constitute an independent sample, w ere excluded in determ ining population gen otyp e and allele frequ encies. Similar procedu res w ere follow ed by Bacon (1980a, 1981) and Bacon & Andersen-Harild (1987). Follow ing age and sex determ ination, 3 to 5 ml of b lood w ere drawn from the brachial vein using a 21 gauge needle. T h e b lood was transferred to a 5 ml tube coated w ith dried lithium heparin and returned to the laboratory on ice. Sample preparation and gel running Blood sam ples w ere spun in a centrifuge for 10 minutes at 2000 x g, th e plasma rem oved and the packed red b lood cells w ere suspended and washed in 10 vo l umes of 0.9% of NaCl and shaken, re-centrifuged and the supernatant rem oved. T h e red cells w ere then lysed using an equal volum e of 0.01% Triton X-100 in 0.9% saline solution, re-suspended, vigo r ou sly shaken and re-centrifuged. An aliquot o f the supernatant was then rem oved and stored at -20 C until required for electrophoresis. Starch gels (11% from Sigma Chemical Com pany, St Louis, M O ) w e re prepared using Perspex m oulds m easuring 220 x 180 x 6 mm and tw o buffers (a ) N ew (A yala et al. 1972) and (b ) TCB (T aggart et al. 1981). A sam ple of red cell lysate and/or plasm a was absorbed on to a 2 mm x 6 mm piece of W hatm an No. 3 filter paper for application to gel. T h e gels w ere run at V DC for 4-6 hr at 4 C, w ith additional coolin g using bags of crushed ice. Staining follo w ed Ayala et al. (1972) or Harris & Hopkins (1976) with minor m odifications. Bands w ere record ed by tracing d irectly from the gels. Table 1. Details of enzymes of Mute Swans assayed, showing Enzyme Commission (EC) num ber and putative enzym e loci. O ptim um tissu e and bu ffer com bin ation s and num ber o f sw ans sam pled, a re given fo r each locus. (AAT-1 and M DH -l w ere too weakly expressed to be typed reliably.) Enzyme EC No. Locus Buffer Tissue n Asparate aminotransferase A 4 7-2* New/TCB RBC 20 Esterase EST* TCB Serum 65 Glucosephospate isom erase G P I* TCB RBC 46 H aem oglobin N A H b * N ew RBC 52 Isocitrate dehydrogenase ID H P * New RBC 20 Lactate dehydrogenase LD H -l* New RBC 46 LDH- 2* N ew RBC 46 Malate dehydrogenase M DH- 2* New RBC 52 Phosoglucom utase PGM * TCB RBC 10 Phosogluconate dehydrogenase PG D H New RBC 20 Superoxide dismutase SO D * TCB RBC 20 T able 2. Esterase genotype frequ en cies for Mute Sw ans in Ireland (the present study) a n d England and Scotland (Bacon 1981). P is for X2 tests with ld f for deviation from Hardy-W einberg expectations (see text). Location No.d G en oty p e freq u en cy FF FS SS n P R eferen ce Cork Lougha NS This study Cork Loughb NS This study W est of Ireland 2/ NS This study Belfast NS T h is study Outer Hebrides NS (Bacon 1981)c M idlands NS (B a co n 1981)c O xford NS (B acon 1981)c Salisbury NS (B a co n 1981)c Christchurch NS (B a co n 1981)c W eym outh NS (Bacon 1981)c A b b o tsb u ry NS (B a co n 1981)c»Cork Lough 1990, t>cork Lough 1986, crecalcu lated from Bacon (1981), dsite number in Figure 1.

4 M ute Swan Genetics Results The nine enzym es and one non-enzymatic protein exam ined w ere presum ed to be determ ined by 11 loci (LDH was coded for by tw o loci) (T able 1). Eight enzym es occurred in red blood cells (RBC) and one in plasma (see Table 1 for the appropriate buffer or tissue com binations used). All proteins listed in Table 1 except the plasma esterase (EST), w ere m onom orphic in Mute Swans at all four locations investigated in Ireland (Fig. 1, Tables 2 and 3). The banding pattern for the esterase locus on a TCB buffer system is shown in Figure 2. Figure 2. Zymogram of the esterase banding pattern of Mute Swans observed on a TCB buffer system. The gen otype frequencies and allele frequencies at the diallelic esterase locus are given in Tables 2 and 3 and Figure 1 which also include English and Scottish allele frequencies recalculated from Bacon (1980a). The genotypic frequencies did not differ significantly from the Hardy-W einberg prediction using a Chi-squared test with one degree of freedom, in cases w here numbers w ere adequate for testing (i.e. expected frequencies >5) (Table 2). The cygnets from a fam ily party from Lim erick w ere not included in the final genotype and allele frequ ency analyses (all other individuals appeared unrelated and w ere thus included). This single fam ily party displayed the follow ing inheritance ratio for the esterase locus: parents, both FS; cygnets, one SS, tw o FS. A number of com parisons of plasma esterase allele frequencies (Table 3) w ere made using Chi-squared test. There was no difference (X ^ j = 0.146, N.S.) between the allele frequencies at the esterase locus for the Cork Lough samples of 1986 and 1990 (T able 3) (all samples from known individuals). Thus the tw o samples w ere pooled (Cork Loughc, Table 3). There was no significant difference in the pattern for the esterase allele frequency between locations in Ireland (X^g = 3.59, N.S.) or England and Scotland (X^g = 3.73, N.S.). The sam ple data from Limerick and Galway w ere p ooled as the expected values at these sites w ere less than five thus invalidating the requirem ents of the X^ test. There was no significant difference (X ^ jq = 13.49, N.S.) when samples from all locations w ere pooled, though the W est of Ireland, Belfast and Abbotsbury show ed most deviation from the average frequencies. Finally there was no significant heterogeneity (X ^ i = 3.46, N.S.) when the pooled data from each country w ere compared. Discussion In the present study w e found that nine enzym e loci and the single non-enzymatic protein locus (haem oglobin ) examined in T a b le 3. Esterase allele frequencies fo r Mute Sw ans in Ireland w ith data from England and Scotland (Bacon 1981) included for comparison. (See Fig. 1.) A lle le Frequ ency Location N o.d F S n R eferen ce Cork Lough« This study Cork Loughb This study Cork LougtV This study Limerick This study Galway This study Belfast This study Outer Hebrides Bacon (1981)c M idlands Bacon (1981)c O xford Bacon (1981)= Salisbury Bacon (1981)c C hristchurch Bacon (1981)c W eym outh Bacon ( 1981)c A b b o tsb u ry Bacon (1 98 1> Superscripts as in Table 2.

5 Mute Swan Genetics 9 the red blood cells of the Mute Swans w ere m onom orphic at all locations sam pled in Ireland. Bacon (1980a) also found most RBC loci to be invariant, but reported one lactate dehydrogenase (LD H ) locus to be variable at low frequency at m ost English and Scottish sites. Exceptions w ere areas of colonial nesting (e.g. Abbotsbury and W eym outh) w here the frequ ency of the h eterozygote LDH phenotype was higher (0.152 and resp ectively) and this is also true in Denmark (Bacon & Andersen- Harild 1987) for colonial Mute Swans. In Ireland the Mute Swan is generally a solitary breeder, but tw o colonial breeding sites have recently been reported (Hutchinson 1989). How ever, these sites w ere not included in the present study. The absence of polym orphism at LDH found in the present study could be due to the fact that it is not present in the Irish population. This could be due to the founder effect, i.e. the Irish population was founded by colonizers w hose genetic com position consisted of a subset of that of the population of origin, which lacked the alternative LDH allele. Interestingly, Bacon (1980a, 1981) failed to detect this polym orphism in another introduced island population, the Outer Hebrides. How ever, the alternate allele m ay be present at such a low level in the Irish population that our sam ple sizes w ere insufficient to detect it. If this is so, then investigation of swans from colonial breeding sites may be fruitful (see above). Plasma esterase was the on ly enzym e shown to be polym orphic in Ireland (Fig. 2). Because we used a different buffer system to Bacon (1981) the appearance of the phenotype was not the same, but factors discussed below confirm that we w ere considering the same system. The gene frequencies at the esterase locus for the Cork Lough swans showed no significant tem poral variation (T able 2). Bacon (1980a) found a similar absence of tem poral variation for his three year study at Oxford. The fact that similar allele frequencies w ere found for Cork Lough swans in 1990 (analysed in Cork) and 1986 (analysed in O xford) permits the intercalibration betw een the results of this study and those of Bacon (1980a, 1981) (also obtained in the Oxford laboratory). Thus, as indicated above, allele frequencies from Bacon (1981) w ere included in Figure 1 and Table 2. Furthermore, segregation ratios from a breeding pair and three cygnets illustrated simple Mendelian inheritance, demonstrating that a single diallelic locus codes for the enzym e, as has been shown previou sly by Bacon (1980a). From a parental cross involving tw o heterozygotes the expected ratio of esterase genotypes, SS:FS:FF, w ould be 1:2:1. The ob served genotype ratio of 1:2:0 from three offspring does not deviate substantially from this pattern of inheritance (see Results.) No significant h eterogeneity was found in allele frequencies at EST between the four Irish locations (though the small number of swans sam pled in areas other than Cork Lough should be noted). One interpretation of this finding is th ere may be sufficient m ovem ents of individuals between the different areas to maintain genetic hom ogeneity (Fig. 1). V ery little is known about swan m ovem ents in Ireland. The main reason is a low level of swan ringing and as a consequence, a low level of recoveries. O Halloran & Collins (1985) in a prelim inary analysis of a small number of Mute Swan ringing recoveries (n = 61) show ed that approxim ately 1% of swans travelled over 100 km. A recent colour ringing study, with several thousands of sightings, has shown that patterns of Mute Swan m ovem ents are along river valleys and to and along the coast, with som e individuals travelling linear distances of 250 km (O Halloran unpublished for Ireland, Ogilvie 1967 for England, Scotland and W ales). Though these latter m ovem ents appear to be less com mon it is conceivable, given that Ireland is a small island, that sufficient mixing of genes exists to prevent the developm ent of population structure. Perrins (1991) working with a population of swans which was almost entirely ringed, achieved a recovery rate of only one in four birds. Considering the much smaller numbers of birds ringed in Ireland, the scarcity of recoveries and hence recorded m ovem ents need not mean that such m ovem ents do not occur. T h e similarities in the esterase gene frequencies both within Ireland and between Britain and Ireland suggests that there is indeed much m ore mixing of the population than recorded by colour-ring resightings or recoveries. It must be em phasised h ow ever, that this result is based on a single locus and may

6 10 M ute Swan Genetics not reflect the actual situation in the genom e as a whole. On the other hand, if isolation of these populations is real, sufficient genetic differences may not have arisen by selection or drift to be identified by these techniques. Based upon the proteins examined in this study, there appears to be no significant genetic difference between Mute Swan flocks within Ireland or between those in Britain and Ireland. There is an observable difference in that an LDH locus which is weakly variable in English populations, was invariant in the Irish populations w e investigated. How ever, such a small difference could not be tested statistically. The genetic similarity of the EST allele frequencies of Mute Swans in England and Scotland suggest that the Irish Mute Swan probably arrived from that country, though whether by anthropogenic introduction or natural colonisation is of course im possible to say. N or can the size of the introduction be estimated. The lack of LDH polym orphism in Ireland dem onstrated by this study suggests that the introductions w ere small. On the other hand the presence of the esterase polymorphism in Ireland with similar allele frequencies to Britain suggests fairly substantial introductions unless balancing selection is operating. Since the variation in allozym es is so low it is suggested that mini-satellite DNA analysis (Burke 1989) for example, might show m ore variation and thus would be m ore useful in addressing problem s of Mute Swan population structure. We are grateful to P. Reilly for assistance in the laboratory and I. Forsyth, T. Carruthers and T. Whilde for help with ñeldwork and facilities provided. To all the swan<atchers in Britain and Ireland for their help with ñeldwork. References Ayala, F.J., Powell, J.R., Tracey, M.L., Maurao, C.A. & Perez, S.S Enzyme variability in the D rosophila w illistoni group. IV. Genic variation in natural populations of D. willistoni. Genetics 70: Bacon, P.J. 1980a. Population genetics of the Mute Swan Cygnus University of Oxford. D32193/80 (BLLD F). olor. D. Phil. Thesis, Bacon, P.J. 1980b. Status and dynamics of a Mute Swan population near Oxford between Wildfowl 31: Bacon, PJ Population genetics of Cygnus olor. In: G.V.T. Matthews & M. Smart (Eds.) Proc. Second International Swan Symposium, Sapporo, Japan. IWRB, Slimbridge, Glos. UK. Bacon, P.J. & Andersen-Harild, P Colonial breeding in Mute Swans associated with an allozym e of lactate dehydrogenase. Biol. J. Linn. Soc. 30: Berry, R.J Inheritance and Natural History. Collins, London. Birkhead, M.E., Bacon, P.J. & Walter, P Factors affecting the breeding success of the Mute Swan ( Cygnus olor). J. Anim. Ecol. 52: Birkhead, M.E. & Perrins, C.M The Mute Swan. Croom Helm, London. Bloch, D Ynglebastanden of Knopsvane ( Cygnus o lo r) i Danmark i Dansk Vildtundersogelser 16:1-47. (Vildbilogisk Station 1971.) Burke, T DNA Fingerprinting and Other M ethods for the Study of Mating Success. TREE 4: Collins, R Breeding perform ance of an Irish Mute Swan Population. In: J. Sears & P.J. Bacon (Eds.) Proc. 3rd Int. Swan Symposium, O xford Wildfowl Supplement No. I. pp Collins, R. & Whelan, J The Mute Swan in Dublin. Irish Birds 4: Ferguson, A Biochem ical Systematics and evolution. Blackie, London. Harris, H. & Hopkins, D.A Handbook o f enzyme electrophoresis in human genetics. North-Holland, Oxford.

7 Mute Swan Genetics 11 Hutchinson, C.D Birds in Ireland. Poyser, Calton. Lewtonin, R.C The genetic basis o f evolutionary change. Columbia U niversity Press, New York. Monval, J-Y. & Pirot, J-Y Results of the IWRB International W aterfow l Census IWRB Special Publication No. 8. O Halloran, J. & Collins, R Prelim inary results of ringing Mute Swans in Ireland. Irish Birds 3: O gilvie, M.A Population changes and m ortality of the M ute Swan in Britain. Wildfowl Trust Ann. Rep. 18: Perrins, C.M Survival rates of young Mute Swans Cygnus olor. In: J. Sears & P.J. Bacon (Eds.) Proc. 3rd Int. Swan Symposium, Oxford Wildfowl Supplement No. I. pp Smiddy, P. & O Halloran, J The breeding b iolo gy of Mute Swans Cygnus o lo r in southeast Cork, Ireland. Wildfowl 42: Smith, C The ancient and present state o f the county and city o f Cork. P rivately published, Dublin. Spray, C An isolated population of Cygnus o lo r in Scotland. In: G.V.T. Matthews & M. Smart (Eds.) Proc. 2nd Int. Swan Symposium, Sapporo, Taggart, J., Ferguson, A. & Mason, F.M Genetic variation in Irish populations of brow n trout (Salm o trutta): E lectroph oretic analysis of allozym es. Comp. Biochem. Physiol. 69B: Thom pson, W The Natural History o f Ireland, Vol. 3. Reeve & Benham. London. Ticehurst, N.F The Mute Swan in England. Cleaver Hume, London. Walter, P., Bacon, P. & Sears, J An analysis of Mute Swan Cygnus o lo r breeding data. In: J. Sears & P.J. Bacon (Eds.) Proc. 3rd IWRB International Swan Symposium, Oxford Wildfowl Supplement No. 1. pp W ieloch, M Population trends of the Mute Swan Cygnus o lo r in the Palearctic. In: J. Sears & P.J. Bacon (Eds.) Proc 3rd IWRB International Swan Symposium, Oxford Wildfowl Supplement No. 1. pp Paul D. O Donoghue, John O Halloran and Thomas F. Cross, Department of Zoology, University College, Cork, Ireland. Philip J. Bacon, Institute of Terrestrial Ecology, Banchory Research Station, Glassel, Banchory, Kincardinshire, AB31 4BY, Scotland. Patrick Smiddy, Ballykeneally, Ballymacoda, Co. Cork, Ireland. Correspondence to Dr John O Halloran, Department of Zoology, University College, Cork, Ireland.