Heredity 64 (1990) 145—159 Received 31 May 1989 The Genetical Society of Great Britain

ThecomplexRobertsonian system of Dichropluspratensis (, ). II.Effectsofthe fusion polymorphisms on chiasma frequency and distribution

ClaudioJ. Bidau*t Laboratoriode Genética, Departmento de Ciencias Biológicas, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Ciudad Universitaria, 1428 Buenos Aires, Repüblica Argentina. The effects of a complex series of Robertsonian polymorphisms on chiasma frequency and distribution were analysed in natural populations of the grasshopper Dichroplus pratensis which has a standard karyotype of 2n =19 (XOd) telocentrics.Populations are usually polymorphic for one to three of seven distinct fusions between the six large (L1—L6) autosomes. The study revealed that: 1. Standard males have an essentially proximal-distal pattern of chiasma distribution. Interstitial chiasmata are less frequent. 2. Fused males have significantly fewer total and proximal chiasmata than standards from the same population but show an increase in interstitial tnd distal chiasmata. 3. The former is due to the fact that centric fusions, either homozygons or heterozygous, produce a decrease in total chiasma frequency of the involved chromosomes and a redistribution of chiasmata resulting in a sharp reduction of proximal chiasmata and an itccse in interstitial and distal ones. In this respect, there are no significant differences between fusion trivalents and corresponding fusion bivalents. The same chromosomes in the basic homozygous state, have a typical standard chiasma pattern. 4. The effects of each fusion are intrachromosomal and although the same chiasma repatterning occurs in all seven fusions, total reduction in chiasma frequency depends on the telocentrics involved since a highly significant positive correlation exists between telocentric length and chiasma frequency. Since all chromosome arms tend to have a minimum of one chiasma, the effects of a fusion are more marked in longer autosomes. 5. Comparisons of populations that differ for number and frequency of fusions, showed a highly significant negative correlation between mean number of different fusions per male (per population) (F) and total chiasma frequency per cell (Xta).Thesame type f correlation exists between F and proximal XtaandF and between-cell variance of totalX. frequency, while a positive correlation was found between F and mean interstitial chiasmata. 6. According to 5, intrachromosomal recombination within populations decreases as number and frequency of fusions increase. This effect is added to the instant reduction of interchromosomal recombination produced by the combination of two linkage groups into one. These results suggest that redistribution of chiasma patterns is due to a direct effect of the rearrangements. Furthermore, the changes in chiasma patterns are discussed in relation to the maintenance of the polymorphisms since trivalents require the elimination of proximal and interstitial chiasmata for balanced orientation and segregation. This was corroborated by the finding of a population in which three fusions were fixed. Here, proximal and interstitial chiasmata in the fused chromosomes are significantly more frequent than expected. Finally, the modifications of recombination potential are discussed relative to the possible adaptive role of these widespread polymorphisms and to a model of chromosomal evolution for the species.

*Memberof the Carrera del Investigador CientIfico y Tec- INTRODUCTION nológico, CONICET. t Present address: Departamento de Genética Facultad de The effects of chromosomal rearrangements on Ciencias Exactas, Qulmicas y Naturales, Universidad Nacional de Misiones, Felix de Azara 174, 3300 Posadas, Misiones, chiasma frequency and localisation have been Repüblica Argentina. extensively studied and reviewed in and 146 C. J. BIDALJ plants. Instances of B chromosomes, interchanges, gent orientation despite strong distal localisation Robertsonian translocations, inversions and super- of chiasmata (Smith, 1966). numerary heterochromatin affecting chiasmata The modifications mentioned above have an and recombination, have been reported. Intra- and additional secondary effect which is the reduction interchromosomal effects have been demon- of intrachromosomal (as well as interchromo- strated (Hewitt, 1979; Parker et a!., 1982; Sperlich somal) recombination and this may be of and Pfriem, 1986; White, 1973). evolutionary significance since it could allow the A distinction must be made however, between maintenance of adaptive supergenes. the effects of spontaneous and polymorphic The grasshopper Dichroplus pratensis Bruner rearrangements since in many cases, the survival offers a unique opportunity to study these effects. and establishment of a stable polymorphic condi- This South American melanopline is widespread tion or conversely, its transience and the con- in Argentina, Uruguay and Southern Brazil sequent elimination or fixation of a given karyo- (Liebermann, 1963) and comprises several morph may depend on a reorganisation of the chromosomal races that differ with respect to num- chiasma patterns in the involved chromosomes. ber and frequency of at least seven different fusions Such is the case of centric fusions. In a small that involve six pairs of large autosomes. With a number of cases of spontaneous fusions studied single exception, all populations already sampled in grasshoppers, no modifications of chiasma pat- are polymorphic for one to three fusions. Karyo- terns were apparent in the resultant trivalents types, geographic distribution of polymorphisms (Colombo, 1987; Kayano and Nakamura, 1960; and meiotic behaviour of heterozygotes have Lopez-Fernandez et a!., 1984; Southern, 1967; already been described (Bidau, 1984, 1988; Bidau Teoh and Yong, 1983). However, in the few cases and Mirol, 1988). of fusion polymorphisms studied in depth, the In this paper the relationship of the fusion rearranged chromosomes showed modifications in polymorphisms of D. pratensis to changes in chiasma frequency and localisation which allow chiasma frequency and localisation are studied in stable convergent orientation and normal segrega- several Argentine populations. tion of trivalents in heterozygotes (Bidau, 1984; Bidau and Hasson, 1984; Bidau and Mirol, 1988; Colombo, 1987, 1988; Hewitt and Schroeter, 1968). MATERIALAND METHODS In some cases however, chromosomes may be pre- adapted to exist in a polymorphic state if they have Thisstudy is based on 297 males of D. pratensis low frequencies and distal localisation of chias- collected by the author and collaborators at the mata as in the Coleopteran Chilocorus (Smith and localities shown in table 1. A much larger number Virkki, 1978). However, this is not always so since of males has been cytologically studied, especially trivalents of hybrids between C. tricyclus and C. at Sierra de la Ventana, being the subject of forth- hexacyclus show high frequencies of non-conver- coming papers; however, some data of these are

Table 1Origin of the populations whose chiasma characteristics are described in table VII. All populations are polymorphic for the involved fusions except C in which three fusions have become fixed. Populations K and SV belong to hybrid zones which result from the overlapping of chromosome races that differ for fusions with monobrachial homologies

Per cent Locality Department Province Collection date Centric fusions standard males

Gaiman (G) Biedma Chubut Jan '83 5/6 930 P. Madryn (PM) Biedma Chubut Jan '83 1/4, 5/6 770 Km 784 (K) Villarino B. Aires Feb '83 1/6, 3/4, 5/6 150 El Condor (EC) A. Alsina Rio Negro Jan '83 2/4, 5/6 40 Tortuguitas (T) G. Sarmiento B. Aires Feb '82 1/6, 3/4 00 Necochea (N) Necochea B. Aires Jan '85 1/6, 3/4 00 Balcarce (B) Balcarce B. Aires Jan '85 1/6, 3/4 00 La LoberIa (LL) A. Alsina RIo Negro Jan '83 2/4, 5/6 0.0 Tandil 1 (Ta 1) Tandil B. Aires Jan '85 1/6, 3/4 0•0 Tandil 2 (Ta 2) Tandil B. Aires Jan '85 1/6, 3/4 00 Sierra de la Ventana (SV) Tornquist B. Aires Feb '83 1/2, 3/4, 5/6, 1/6 00 La Florida (LF) Pringles San Luis Jan '84 1/6, 3/4, 2/5 0.0 Monte Hermoso (MH) C. Dorrego B. Aires Jan '83 1/2, 3/4, 5/6 00 Cerro (C) Tornquist B. Aires Jan '86 1/2, 3/4, 5/6 00 CENTRIC FUSIONS AND CHIASMA FREQUENCY 147 incorported in this paper. All males were processed terns of non-fused individuals of all populations in the field. Testes were fixed in 3: 1 alcohol-acetic studied although variations between populations and squashed in lacto-propionic orcein. probably do occur. Nevertheless, st's from PM and Chiasmata were scored at MI since orientation G were very similar regarding chiasma frequency of bivalents and trivalents at this stage allows easy and localisation (table 2) (see also Bidau, 1984). identification of chiasma position. Ten cells were Fig. 1 shows chiasma patterns of several distinct scored per male and chiasmata classified as fusion karyomorphs. The S bivalents never form proximal (P), interstitial (I) and distal (D) accord- more than one chiasma that may be P, I or D ing to their location in the first (proximal), second although the latter predominante (fig. 1 (a)-(h)). or third portion of the chromosome or chromo- Telocentric L bivalents never form more than two some arm when the latter are divided in equal parts. chiasmata (fig. 1(a)—(c), (f), (h)). Monochiasmate L bivalents usually have a P or a D chiasma (fig. 1(a)—(c), (f), (h)), seldom an I one (fig. 1(b)). RESULTS Bichiasmate telocentric L bivalents are very frequently P/D (fig. 1(a), (b), (h)), sometimes I/D The karyotype and fusion polymorphisms (fig. 1(f)) and only rarely P/I. A summary of of ft pratensis chiasma frequencies and distribution in telocentric TheRobertsonian system of the species has already L bivalents of standard males is given in table 2. been described (Bidau, 1984, 1986, 1988). The Chiasma conditions are radically altered in meta- karyotype comprises in its standard form, six pairs centric fused chromosomes as discussed below (fig. of large telocentric autosomes (L1—L6), three small 1(c)—(h)). There is a highly significant correlation telocentrics (S7—S9) of which S7 is the megameric between chiasma frequency per bivalent and bivalent and S8 carries the single standard NOR, chromosome length (fig. 2; table 3) within the L and one ((3)ortwo (?)medium-sizedtelocentric group. The correlation does not hold for the S X chromosomes. L-L6 chromosomes can be bivalents since there must always be a chiasma unambiguously recognised by distinctive length. present for proper segregation. Only L6 can be identified by heterochromatin Frequencies of total (T), P, I and D chiasmata content since it usually has a C-positive hetero- in 47 Stmalesfrom PM are shown in fig. 3(a)-(d). chromatic block at its telomeric end. The six L These are not significantly different from those of autosomes are involved in fusions and seven the st's from 0. That is, standards have a mainly different fusions have been identified (1/2; 1/4; P/D distribution, I chiasmata being rare (Bidau, 1/6; 2/4; 2/5;3/4;5/6). Some populations are 1984). In PM, 15 males carrying fusions (all of polymorphic for a single fusion, some for two and them heterozygotes) wre found: nine had the 1/4 some for three although not necessarily the same fusion, five the 5/6 fusion and the remaining one ones since the same telocentric is involved in was a double 1/4, 5/6 heterozygote. These males, different fusions. Hybrid zones between chromo- when considered jointly showed significantly fewer somal races have been identified and within them, total chiasmata than the Stmalesfrom the same popultions may be polymorphic for four fusions population (fig. 3(a)). This was mainly due to a with monobrachial homologies (Bidau, 1988). significant decrease in P chiasmata (fig. 3(b)). D and I chiasmata were, on the contrary, higher in heterozygotes (fig. 3(c)— (d)). Chiasma frequencies Chiasmafrequency and localisation in standard and distribution of individual 1-1.4-4 and 5-5.6-6 and fused males trivalents are summarised in table 4. Untilthe present, no completely standard (st) In order to test for possible interchromosomal population has been found in the sampled areas. effects, mean cell chiasma frequencies (X) exclud- Indeed, st individuals are very rare (Bidau, 1988) ing chromosomes 1 and 4 in one case, and 5 and due to the numbers and frequencies of the fusions 6 in the other, were calculated in ten randomly present. Only two populations showed high chosen non-fused males. The same was done with frequencies of Stmales:Puerto Madryn (PM) and nine 1/4 and five 5/6 heterozygotes. The chiasma Gaiman (0) (table 1), both from Chubut and about frequencies thus calculated were not significantly 80 km apart. The first one is polymorphic for different in both cases (non fused, X =786vs. fusions 1/4 and 5/6. In Gaiman, a single 5/6 1/4, X =771;F1189 =156;P>005; non-fused, heterozygote was found among 14 males sampled. X =830vs. 5/6, X=856; F1 149=F92;P> 0.05). The standard individuals from these populations These results suggest that the effect of the fusions were assumed to display the typical chiasma pat- on shiasma frequency are exerted solely upon the 48 C. J. BIDAU

• U PD PD• I Dli •PD 4D$ DDr a b ix I. a D D D% D(_JD 4 D'D D DtID 3 C DJD 144 4 pa 4 x o:I e f

0

a I. I. 7 4 a.PD

9 %x h CENTRIC FUSIONS AND CHIASMA FREQUENCY 149

Table 2Asummaryof chiasma frequencies and distribution in the L bivalents of standard (2n =19)males of Dichropluspratensis. Data ofeach L bivalent (L1-L6) are given for the G (upper row) and PM (lowerrow) populations. X, Xi,, X and X4 are meannumbers oftotal, proximal,interstitial anddistal chiasmata per bivalent,respectively. Percentagesare giveninparentheses. Theright part of the table represents the distribution of bivalents with different localisation of 1 and 2 chiasmata

Bivalents with n chiasmata

Chiasnia frequencies 1 2 Total X, X,, X1 X P 1 D I P/D l/D P/I I II's

1 15538 05538 00308 09692 2 2 54 58 70 2 0 72 130 (1'54) (154) (4154) (53'84) (l'54) 16063 06271 00646 09146 30 ii 148 189 271 20 0 291 480 (630) (230) (30.80) (56•50) (420) 2 1•4835 O'5462 00923 08000 14 12 47 73 57 0 0 57 130 (l080) (9.20) (36'20) (4380) 14667 0'6646 01188 06833 117 32 107 256 199 22 3 224 480 (24'40) (6.70) (22.30) (41.50) (460) (060) 3 1'3692 0'6154 00846 06462 35 8 39 82 45 3 0 48 130 (26.90) (6'IO) (30'OO) (3460)(2'30) 12833 0'6063 0'1458 0'5313 170 52 122 344 118 15 3 136 480 (3540) (1080) (2S40) (24.60) (3.10) (060) 4 12615 0'6077 01385 05154 49 14 33 96 30 4 0 34 130 (3770) (1080) (2540) (23'lO (310) l'1833 05125 01333 05396 169 48 175 392 73 11 4 88 480 (3520) (1000) (3650) (15.20) (2'30) (0.80)

5 1'1615 06538 00692 04385 65 8 36 109 20 1 0 21 130 (5000) (620) (27'70) (15.40) (0.80) 11521 0'5167 0'0333 06021 176 15 216 407 72 1 0 73 480 (36.70) (310) (45.00) (15.00) (020)

6 11154 0'7308 0'0308 0'3538 80 4 31 115 15 0 0 15 130 (61'SO) (3.10) (2390) (11.50) 1'0917 0'6000 00313 0'4604 244 15 177 436 44 0 0 44 480 (50.80) (310) (36'90) (9.20) involved chromosomes. The same type of corn- with their fusion counterparts as in fig. 4(a), (b) it parison was made for P, I and D chiasmata and is obvious that in both cases, chiasma frequency again, differences were not statistically significant. is reduced in the fused chromosomes though the The other fusions studied produce similar effects effect is less marked in the 5/6 fusion as expected; on chiasma frequency and distribution (see for (c) when 1/4 and 5/6 heterozygotes were corn- example, table 6). pared with respect to total chiasma frequency, the Furthermore, these results indicate that the two difference was highly significant (F1,139 =1826; fusions affect chiasma frequency differently. This P <0.01) but the reduction in proximal chiasmata is expected since in stmales,chromosomes 1 and is similar (F,,139=0'31; P>0'05). 4 have on average more chiasmata than 5 and 6 It was considered relevant to compare fusion (table 3; fig. 2). This is corroborated by the follow- heterozygotes and homozygotes regarding chiasma ing: (a) while mean cell chiasma frequency of st's frequency. Since males differing karyotypically is10'75, that of 1/4 heterozygotes is 983 (F1,559 = only for the fusion in question are difficult to find 42'30; P<0'Ol) and that for 5/6 heterozygotes, in sufficiently large numbers owing to the nature 10'52 (F1,519= 1.46; P>0.05); (b) when unfused of the polymorphisms, comparisons were made chromosomes 1 and 4 or 5 and 6 are compared solely between the trivalent and the corresponding

FIgure 1 Chiasma distribution at MI in Dichroplusprazensis males.(a), (b). Two cells from a standard male (2n 19).(c) Horn 1/6,Het 3/4(2n= 16).(d) Triple 2/5, 3/4, 1/6 Het; the 1/6 trivalent is linearly oriented; note distal localisation of chiasmata (2n=16).(e)Het 1/6, Horn 2/5, 3/4 (2n=14).(1) Het 1/6, Horn 3/4(2n=16).(g) Double 1/6, 3/4 Met, 2/5 Hom(2n 15); note exceptionalproximalchiasma in bivalent 2/5 (left). Het 1/6, Horn 3/4. References: P (proximal), I (interstitial), D (distal). Bar =10m. 150 C. J. BIDAU

x .6-

I 5

I4-

3.

I 2 I I'

I 0. 09- i=iI11 08 0.7

// 9 0 II '2 113 '4 l'5 l'6 '7 I'SRL Figure 2 Relationship between relative L chromosome length (RL) and chiasma frequency per chromosome or chromosome arm in ten standard males (*),tenrandomly chosen males from a polymorphic population for fusions 1/2, 3/4 and 5/6(0)and ten males from a population with the same fusions fixed (•). Regression equations and their significance are given in table 3. fusion bivalent. The fact that no interchrornosomal are combined into one thus reducing the number effects have been detected (see above) was con- of elements that are randomly transmitted to the sidered. The comparisons were made using gametes. Chiasma frequency reduction is added individuals from the SV, LF and EC populations. to this effect producing a generalised decrease in As can be seen in tables 5 and 6, no significant recombination. differences between Hets. and Horns, are apparent To test this prediction, the mean number of (interstitial chiasmata are however, more frequent different fusions per individual (F) was calculated in fusion bivalents than in trivalents). for each population (table 7). Mean cell total, These results suggest that the recombination proximal, interstitial and distal chiasma frequency capacity of a given population (hence its potential and inter cell variance of total chiasma frequency, ability for the liberation of genetic variability) were plotted against F. The results of the regression depends on the quality, number and frequency of analyses (figs. 5 and 6) show that T, P and I chiasma the fusions present. frequency and inter-cell variance depend sig- nificantly upon F. A further confirmation of this effect was obtained when populations SV, PM and C were Inter-populationcomparisons compared for chiasma frequency on each L Itis worth noting that interchromosomal recombi- chromosome or chromosome arm. Ten males from nation is instantly reduced as a consequence of a each sample were considered: ten st'sfromPM, centric fusion. Two independent linkage groups ten triple homozygotes (1/2, 3/4, 5/6) from C and

Table 3 Chiasma frequency per L chromosome or chromosome arm based on ten randomly chosen standard males (a) from PM population, ten from a polymorphic population (b) (SV population polymorphic for 1/2, 3/4 and 5/6 fusions) and ten from a monomorphic population (c) (C population; fusions 1/2, 3/4 and 5/6 fixed) L chromosome or chromosome arm

1 2 3 4 5 6

Mean relative length (RL) 1785 1467 1368 12-20 1093 972

Mean chiasma frequency per (a) 149±005140±005125±004120±0-04115±004105±002 L chromosome or chromosome (b) 133±005126±0-04 112±003103±002118±004106±002 arm (±SE) (c) 1-08±003 102±002104±002098±003095±0030-94±003

(a) RL vs. .Y=054+0-055X;F(1,4)=9370; P<001. (b) RL vs. X. Y=074±O032X; F(1,4)=6-96; P<001. (c) RL vs. X. Y=076+0018X; F(1,4)=78-86; P<001. CENTRIC FUSIONS AND CHIASMA FREQUENCY 151

a b

1075 N

3 j77

C d

N

6,2 5

nD 0•1 23 •4 ni Figure 3 Distribution of the number of (a) total (T), (b) proximal (P), (c) distal (D) and (d) interstitial (I) chiasmata in standard (open histograms) and fused (shaded histograms) males of the PM population. Means are indicated. Results of the respective analyses of variance are: (a) F1168 =325;P <0001; (b) F1168 =1346;P<0001; (c) F1168 =123;P <0001; (d) F1168 =54; P < 000 1. ten randomly chosen males (including individuals profound effects on intra- and/or interchromo- with 1, 2 and 3 fusions) from the highly polymor- somal recombination. Examples can be found in phic SV population. From fig. 2 it can be seen that different organisms and include inversions in the relationship between chromosome length and Drosophila and some grasshoppers (Confalonieri, chiasma frequency varies according to the number 1988; Confalonieri and Colombo, 1989; Goñi et of fusions present (table 3). a!., 1985; Hewitt, 1979; Sperlich and Pfriem, 1986; Weissman, 1976; White and Morley, 1955), super- numerary segments in many (Hewitt, DISCUSSION 1979; John, 1983; Riva et a!., 1984), B chromo- somes (Jones and Rees, 1982), etc. The mechanism Thepresence of chromosomal polymorphisms in by which recombination is modified depends on natural populations is frequently associated with the rearrangement involved and may imply van- BIDAU + J.

C. c 6666 4rs + —

a c- — , 6 Sa ri 0 000 O r-1 0 n o 0 0 en — en 000

— — a- 0000 en a- — 6 o r — ——0 — 0 '7

.2a •_oo

0 F- a- en -

a a 8 a 0V a I j .0 .0 1- a0. 0 U < V •a o .2 . V aE 0 vU V U U— V o V V . — >, 0 . 0 .0 a ao U V.2 0 0 a a V U '-0 8 ii a LHV _9 'ii' .

a b 18, (0) (a) do the the do the —13xt (It) both. poly- PM. df= muta- 1987). 1=310, genetic by - or natural fused different chromo- — as or from 7-48, Y=2-1195— rearrange- restrictions the recombina- in F frequency 12 certainly between t P<0-70. modifications (S) r=0-74, in well and and males is reduced as heterozygotes S 0-87, Fused: same (Colombo, in Y=0-4498+0-1661X; involved 090< rearrangements 1988). chromosome chromosomal chiasma localisation - r that unfused the a original II changes effect the total +01954X; df=4, them chromosomal of I the severely interaction properties a Mirol, of of Unfused: the the to themselves

exactly chromosomes 0-2391 that and between (b) overall crossing-over, (=0-30, frequency,that rise (b) 0-01

7A5_6 Figure ation It tion not produced tions Furthermore, result mechanical of factors selection. ment individuals morphism not meiotic 152 XL4 CENTRIC FUSIONS AND CHIASMA FREQUENCY 153

Table 5A summary of chiasma frequency and distribution of different fusions of Dichroplus pratensis in the homozygous, heterozygous and standard states. Only those chiasma configurations that were observed at least once, are included. The configurations for the fusions in the standard form result from the combination of the chiasma patterns of both telocentric bivalents from the same cell; the same applies for standard chiasma frequencies

1 2

Fusion D-O P-O 1-0 D-1 D-P I-P P/D-OI/D-O O-D 0-P 0-I D-D 1-D P-D 1-I P-I P-P O-P/D0-l/D

1/2 SV Horn — — — 700 175 — 2'5 — — — — Het — — — 540 21'O — 40 — — — St — — — 4.4 — 14'4 — 1.1 1'I — — 3/4 SV Horn 0'7 — 0'3 724 103 O'3 31 — — — 0•3 Het — — — 580 273 07 1'3 — — — — 3/4 LF Horn — — — 867 83 — — — 0'S — — Het — — — 762 200 12 — — — 12 St — — — 300 -- .— 100 —.— 5/6 I/C Horn 50 — 10 712 170 20 30 — 10 — — I-let — — — 74-6 200 36 18 — — — — St — — — 2-5 5-0 12-5 5-0 7-5 300 — — Horn — — — 57-5 20(1 25 — — — — — 5/6 SV Het — — — 68-2 20-9 0-9 2-7 1-8 — — — St — — — 15-S 1-7 27-5 08 42 4-2 — — 1/6 LF Horn 71 1-4 2-9 32-9 32-9 14 43 14 — 14 — Het 23 — — 68-5 14-5 0-8 46 — — — —- St — — — —- —--- 20-0 — — 200 — --- 1/6 SV Horn — — — 93.4 3.3 Het — — — 70-0 18-2 09 09 — — — —-- St — — — 22-5 2-5 25-0 — — 2-5 — —-- 2/5 I/F Horn 10 — — 60-0 23-0 — 20 3-0 — 1-0 -—- Het — — — 82-0 11-4 St — — — 2-5 2-5 50 — 5-0 400 — — 2/4 I/C Horn 6-0 2-0 — 60-0 280 — 4-0 -—- — Het — — — 52-0 41-0 — 7-0 — — — -—- St — — — —-- 2-9 114 29 5-6 22-9 — -

Fusion P/D-P P11)-I P/ D- P 1/ft-P l/D—I 1/D—D P/I--P P—I'/D I-P/I) D—P/D P-I/i)I—I/I) D—I/D P—P/I

1/2 SV Horn 25 5-0 —— Het — 30 —— 10 170 St 100 4.4 33.3 4.4 1-1 1-1 3/4 SV Horn 03 0-3 0-3 — 03 11-0 — Het 13 11-3 3/4 LFHorn — 08 2-5 Het — 1-2 — 10-0 200 5/6 EC HornSt — Het St 225 7.5 7.5 5/6 SVHorn — 2-5 — 25 125 — Het — — 18 36 -— St 150 5-8 150 08 — 3.3 — 1/6 I/C Horn — —- 29 114 Het — 0-8 - 23 62 — St 100 30-0 1/6 SV Horn 3.3 Het — — 09 8-2 St 225 7.5 50 25 2-5 5-0 2/5 LF Horn 1-0 1-0 — 1°0 60 1-0 Het — 8-6 St 250 2-5 7.5 2/4 EC Horn —- Het — 2-0 St 314 7-1 8-6 154 C. J. BIDAU

Table5 continued

4 Fusion P/D-I/D P/I-P/D I/D-I/D I/D-P/D P/D-P/D P/D-P/I N Xt Xp Xi Xd 1/2 SV Horn 2-5 — — — 40 213 003 033 177 Het — — — — 100 2-21 0-03 0-48 170 St — 7-8 144 22 90 3-03 1-23 022 1-58 3/4 5V Horn — — — — 290 2-12 0-02 029 1-81 Het — — — — 150 2-13 002 0-41 1-70 3/4LF Horn — — — — 120 203 0-02 013 188 Het — — — — 80 2-01 001 023 1-77 St — — — — 10 2-30 0-80 0-10 1-40 5/6EC Horn — — — — 100 194 0-04 0-24 1-66 Het — — — — 110 2-00 0-04 0-24 172 St — — — — 40 2-38 1-40 0-38 060 5/6 SV Horn 2-5 — — — 40 2-23 0-05 043 1-75 Het — — — — 110 205 003 0-35 1-67 St — 0-8 7-5 — 120 253 1-02 018 133 1/6 LF Horn — — — — 70 2-03 006 0-63 134 Het — — — — 130 207 0-02 0-35 1-70 St — 10-0 10-0 — 10 280 1-40 0-10 1-30 1/6SV Horn — — — — 30 2-03 0-00 0-07 196 Het — 09 — — 110 2-11 0-02 031 1-79 St — — 2-5 — 40 2-50 1-03 020 127 2/5 LF Horn — — — — 100 2-09 007 0-35 167 Het — — — — 70 2-09 0-00 020 189 St — 2-5 7-5 — 40 2-55 168 013 0-74 2/4EC Horn — — — — 50 1-92 0-02 0-36 1-54 Het — — — — 100 2-02 0-02 0-57 1-43 St — — 43 — 70 2-56 1-53 0-24 0-79 somes or else, that the loss in fertility is sufficiently Df the rearrangement or by selection against compensated by some selective advantage that the proximal chiasma formation (unless preadaption heterozygote might possess. in the form of strong distal localisation in In the case of fusions, one condition that telo/acrocentrics occurs). ensures proper orientation and segregation of When considering the multiple fusion poly- trivalents is the absence of proximal chiasmata morphisms of D.pratensistwo points are then which, when present, frequently determine linear worthy of discussion; first, the initial establishment orientation. For the maintenance of a fusion poly- of the polymorphisms and second, the possible morphism it is then necessary that proximal chias- adaptive significance of the fusions themselves and mata are eliminated either through a direct effect the modifications on recombination. Both points are, however, intimtely related. Table 6 Comparison between chiasrna frequencies of different In the three thoroughly studied cases of fusion L chromosome cornbinations in the heterozygous (Het) polymorphism in Acrididae, those of and homozygous (Horn) states enigma (Hewitt and Schroeter, 1968), Leptysma ? ? argentina (Bidau and Hasson, 1984; Colombo, FusionHet Horn F d P 1987, 1988) and D. pratensis (Bidau, 1984, 1986, 1988; Bidau and Mirol, 1988; Tosto et a!., 1987; 1/2 n = 100 n = 40 3-971, 138 0-05—0-01 2-25±0.04 2-10±0-05 this paper), a marked reduction in proximal chias- 3/4 n=100 n=100 0-871,198 >0-05 mata accompanying a general redistribution of 2-12±0-03 2-19±0-04 chiasmata in the involved chromosomes of 5/6 n = 100 n = 50 2-951, 148 >0-05 heterozygotes was observed. In all three cases, 2-08±0-03 2-18±0-07 trivalents show fairly regular orientation and segre- 1/6 n 110 n =30 0-001,138 >0-05 2-10±0•03 2-10±0-06 gation. Indeed, in D. pratensis non-convergently oriented trivalents have on average more proximal CENTRIC FUSIONS AND CHIASMA FREQUENCY 155

Table 7Chiasma characteristics of 14 natural populations of D. pratensis(forreferences see table 1)

Population N F X, X X, Xd S G 14 007±007 1076±011 357±013 077±007 647±011 1674 PM 62 0•26±006 1059±005 345±006 077±003 637±006 1555 K 20 120±017 984±007 226±009 092±007 667±009 0939 EC 24 154±012 992±005 342±009 1'09±006 541±0'lO 0706 T 13 1•57±014 955±006 2•01±012 0•94±008 6•61±014 0529 N 5 180±020 954±010 220±015 106±013 628±020 0539 B 10 180±013 943±0•06 234±011 139±013 568±014 0389 LL 6 183±017 1003±009 263±013 117±012 623±017 0440 Ta2 15 187±009 950±007 201±010 152±010 597±012 0681 Tal 18 200±000 956±007 163±007 123±009 670±009 0837 SV 46 2'33±015 991±004 133±005 131±005 727±006 0780 LF 21 271±010 9•29±005 091±006 116±007 721±009 0454 MH 6 300±000 903±0'06 073±013 162±013 668±0'lS 0236 C 16 3•00±000 904±006 058±006 196±012 654±013 0489

(1) P vs. Xi. Y= 1065 —053X; F(1, 12) =5194;P <001 .r=—09014; 1(12) =721;P<0001 (2) Pvs. .p. Y=390—102X; F(l,12)=5588; P<001 .r=—09071;i(12)=717; P<0001 (3) F vs. Xi Y=064+032X; F(l, 12)=2584; P<001 .r=08257;i(12)= 507; P<0001 (4) Pvs. Xd. Y=613+017X; F(1, 12)=113; P>005. r=02931; t(12)=106; 03

a b

3

Io S

0 0 xi C Xd d

0 7. 2 0 S. •0 6 S

5.

I I —, 2 0 2 F

Figure 5 Relahonship betweenmeannumber of different fusions per male (F) and mean (a) total (.i),(b)proximal (gp), (c) interstital (Xanddistal (Xd)chiasmataper cell in 14 populations of D.pratensis. Regressionequations and their sigriificances are givrn flle7. pattern in heterozygous mutants. Extensive tightly linked with the metacentric fusion chromo- chiasma modifications certainly do occur in some. The latter is improbable since in D. pratensis mutants hetero7vgous for interchanges or fissions all seven fusions behave equally: in all cases (Arana et a1. SO;Hewitt, 1967; Parker, 1987; telocentrics have high proximal chiasrna frequen- Parker et aL, 2) and these are obviously a result cies (see the results section) and besides, control of the rearn ment itself. Second, comparisons of chiasma frequency is probably polygenic (Shaw, have been between different chromosomes 1972). which do n'.'t necessarily respond equally to the It is then more plausible that the fusions exert rearrangement.Ideally, comparisons should some direct restrictions on proximal chiasma for- involve the same fusion, polymorphic in one popu- mation at the heterozygous and homozygous con- lation and spontaneous in another. ditions through for example, modifications of Perhaps more relevant is the fact that it is synapsis. A possible model is the following: if difficult to envisage a selective process by which chiasmata are formed preferentially in regions proximal chiasmaa are eiiminated in the fusion which include or are close to zones of pairing chromosomes without a concomitant elimination initiation which are paired early in prophase I and in the corresponding telocentrics of standards, have the most protracted period of synapsis, the unless genes responsible for chiasma control are chance of chiasma formation in these regions may CENTRIC FUSIONS AND CHIASMA FREQUENCY 157 r2 2 a b

0

3

2 0 S • 0 S

— 0 a 2 a 10 15 20

Figure6 Relationship between (a) F and mean betweenceli variance of total chiasmata (S) in 14 populations of 1). prafnth and (b) mean frequencyof interstitial(Xi)anoproximal (Xp) cniasmasa in the same populations. Regression equations and their significance are given in table 7. well be increased. Then, if unfused telocentrics could be initially successful because of favourable have both ends attached to the nuclear membrane, position effects. The need for stabilising segrega- they can initiate pairing from either or both ends, tiort in heterozygotes could then, by restricting which could explain the P/D chiasma distribution, chiasma formation, allow the surging of super- In fused chromosomes however, four original ends genes. That is, the decrease in recombination being are replaced by two and the number of early pairing a byproduct of the establishment of the poly- regions is consequently reduced and shifted to morphism would not be the initial cause of its distal ends only. This would explain both the maintenance. In the second case one could reduction in total chiasrna frequency and the envisage a situation in which some coadaptation reduction in proximal chiasmata in fusion already exists and any fusion that involves such heterozygones and iiornozygotes. gene arrangements should be favoured since it Turning to the possible adaptive significance would protect them from recombination. of the D.pratensis's polymorphismsone is tempted A model proposed for the chromosomal evol- to relate their maintanence to the restrictions ution of D.pratensis(Bidau, 1984, 1988) can be imposed upon recombination. Fusions are not as expanded when the effects of the fusions on re- effective suppressors of recombination as inver- combination are considered. It proposes that range sions bum a fusion instantly reduces interchromo- expansion of the species occurred mainly in the somal recombination by bringing together two for- monomorphic standard form since these popula- merly independent linkage groups, usually accom- tions have more recombination potential and thus, panied by a suppression of chiasmata in proximal a greater capacity for liberating genetic variability portions of chromosomes as discussed above. which allows the production of more gene arrange- These facts could allow the conservation of ments to be selected for in order that the population coadapted supergenes partially or completely pro- becomes adapted to a new environment. Once a tected from recombination. Thus the maintenance new colony adapts successfully to its new habitat, of the polymorphisms would become adaptive the fortuituous appearance of one or more fusions since fixation could, by eliminating the segregation that include the adaptive gene combinations would restrictions of heterozygotes, result in an increase generate the polymorphic condition to preserve in crossing-over in regions previously protected them as explained above. (as in the C population described in Results) and It is possible to relate this model to the central- the disruption of supergenes. marginal model of evolutionary biology as it has Two hypothese relate polymorphic rearrange- been applied for example to Drosophila (Brussard, ments and supergenes, the postadaptation and the 1984; Sperlich and Pfriem, 1986). The model preadaptation hypotheses (Sperlich and Pfriem, assumes that populations belonging or near to the 1986). In the first case, the newly arisen fusion centre of a species' range, have certain characteris- 158 C. J. BIDAU tics such as high density, high levels of genetic and tion higher in marginal populations; on the other, phenotypic variation, etc. In contrast, peripheral phenotypic variation shows no significant differen- populations are isolated, sparse, have low levels ces. The latter could be related to the fact that of chromosomal polymorphism, etc. A number of allozyme variability is not correlated with chromo- hypotheses concerning the inversion polymorph- somal variability (Brussard, 1984). It is relevant isms in Drosophila as related to this model have however to note that although central populations been put forward (Brussard, 1984). may be chromosomally more polymorphic, the When considering the distribution of fusion type of polymorphism conditions the propositions polymorphisms in D. pratensis it was shown that, of the central-marginal model: central populations at least in the sampled area, no clear pattern is may be highly polymorphic for B chromosomes evident (Bidau, 1988). Certainly, some fusions are which raise chiasma frequency, or inversions or widespread and others restricted to particular fusions that decrease it. Hence, the chromosomal localities, and populations are generally very poly- situationispossibly related not only to morphic. Exceptions however, do exist; two modifications of recombination but to other effects Patagonian populations have exceptionally high due to the rearrangements themselves. frequencies of standard individuals (PM and G) The hypothesis that the PM and G populations and are consequently the less polymorphic ones are recent invaders is appealing but by no means and those with the highest recombination indexes. definitive. Despite the wide ecological tolerance displayed by the species, these populations are the best candi- Acknowledgements I am fully indebted to Dr Cristina dates for being considered marginal (in the eco- D'Aiutolo and my sons, Claudio Jr and Pablo, for help in the logical sense) since environmental conditions in collection and processing of the grasshoppers. lam also indeb- these areas are certainly more strenuous (see ted to my collaborators, Miss Daniela Tosto and Miss Patricia Bidau, 1984). Mirol for their patient work, useful comments and good com- pany. Thanks are also due to my colleague and friend Dr These populations, having more recombination Pablo Colombo for the critical review of the manuscript and potential, could liberate more genetic variability many stimulating discussions. I also wish to thank an anony- (if available) that would be of advantage in a new, mous reviewer whose corrections and suggestions improved unexploited environment. The PM and G popula- the manuscript substantially. tions would then agree with Carson's (1959) model This work was supported by grants from CONICET and in having low polymorphism and high recombina- Universidad de Buenos Aires. tion. It is however relevant to discuss some morphological data of D. pratensis (Bidau and Tosto, 1986). Although the putative marginal REFERENCES populations seem to be phenotypically more vari- able than those presumed to be central (or optimal ARANA, P., SANTOS, J.L. ANDGIRALDEZ,R. 1980. Chiasma interference and centromere co-orientation in a spon- in he ecological sense), no statistically significant taneous translocation heterozygote of Euchorthippus pul- correlation was found between degree of poly- vinatus gallicus (Acrididae: Orthoptera). Chromosoma, 78, morphism and morphological variability. This is 327—340. in contrast with the report of Westerman (1983) BIDAU, c. i. 1984. Ph.D. Thesis, Universidad de Buenos Aires. on Phaulacridium in which a negative correlation BIDAU, c. i. 1986. Geographic distribution of chromosome polymorphisms in Dichroplus pratensis Bruner (Melano- between recombination potential and phenotypic plinae, Acrididae) in Argentina. Proc. IV Triennial Meet., variability was demonstrated. He produced a very Pan Amer. Acridol. Soc., Saskatoon, 165-166. apt explanation: those populations with high cross- BIDAU,C. .j.1988.Zonas hIbridas en ortópteros: el ejemplo de over frequencies liberate much genetic variability Dichroplus pratensis (Acrididae). Revista Soc. Entomol. Argentina (In press). which is subjected to selection very rapidly and BIDAU, C. J. AND HASSON, E. R. 1984. Population cytology of consequently, rapidly lost. Populations with a Leptysma argentina Bruner (Leptysminae, Acrididae). lower recombination potential would preserve the Genetica, 62, 161-175. extant genetic variability. In Westerman's (1983) BIDAU,C.J. AND MIROL, i'. 1988. Orientation and segregation case however, low recombination populations of Robertsonian trivalentsinDichroplus pro tensis (Acrididae). Genome, 30,947—955. were the peripheral ones. A further case of a BIDAU, C. J. AND TOSTO, D. 1986. Análisis de la variabilidad central-marginal distribution of a fusion poly- morfológica de razas cromosómicas de Dichroplus pratensis morphism in a grasshopper was reported by en relación a la capacidad potencial de recombinacjón. Colombo (1988). XVII Congr. Arg. Genet., RIo Cuarlo, 65. BRUSSARD,i.F.1984.Geographic patterns and environmental The case of D. pratensis is puzzling. On the gradients: the central marginal model in Drosophila one side, polymorphism is lower and recombina- revisited. Ann. Rev. Ecol. Syst., 15,25—64. CENTRIC FUSIONS AND CHIASMA FREQUENCY 159

CARSON, H. L. 1959.Genetic conditions that promote or retard PARKER, i. S. 1987.Increased chiasma frequency as a result of the formation of species. Cold Spring Harbor Symp. Quant. chromosome rearrangement. Heredity, 58, 87-94. BioL, 24, 87-103. PARKER,J. S., PALMER, R. W., WHITEHORN, M. A. F. AND COLOMBO,P. C. 1987.Effects of centric fusions on chiasma EDGAR, L.A. 1982.Chiasma frequency effects of structural frequency and position in Leptysma argentina. 1. Spon- chromosome change. Chromosoma, 85, 673-686. taneous and stable polymorphic centric fusions. Genetica, RIVA,E., FOX, D. P., GIRALDEZ, R. AND SANTOS, J. L. 1984. 72, 171—179. Chiasma frequency and distribution in the presence and COLOMBO,P. C. 1988.Polimorfismos cromosómicos en absence of supernumerary chromosome segments in the Orthoptera. Revista Soc. Entomol. Argentina (In press). grasshopper, Euchorthippus pulvinatus gallicus. Heredity, CONFALONIERI,V. A. 1988.Effects of chromosome polymorph- 53, 101-106. isms on chiasma conditions in Trimerotropis pallidipennis SHAW,D. D. 1972.Genetic and environmental components of (Oedipodinae: Acrididae), Genetica 76, 171—179. chiasma control. II. The response to selection in Schis- CONFALONIERI,V. A. AND COLOMBO, P. C. 1989.Inversion tocerca.Chromosoma, 37,297—308. polymorphism in Trimerotropis pallidipennis (Orthoptera): SMITH,S. G. 1966.Natural hybridization in the coccinellid clinal variation along an altitudinal gradient. Heredity, 62, genus Chilocorus. Chromosoma, 18, 380-406. 107—I 12. SMITH,S. G. AND VIRKKI. N. 1978.Coleoptera. In John, B. GONI,B., DE VAIO, E. S., BELTRAMI, M., LEIRA, M. S., CRIVEL, (Ed) Cytogenetics 3. Insecta 5. Gebruder Born- M., PANZERA, F., CASTELLANOS. P. AND BASSO, A. 1985. traeger, Berlin-Stuttgart. Geographicpatterns of chromosomal variation in popula- SOUTHERN,D. 1. 1967.Spontaneous chromosome mutations tions of the grasshopper Trimerotropis pallidipennis from in Truxaline grasshoppers. Chromosoma,22,241-257. Southern Argentina. Can. J. Genet. Cytol., 27,259—271. SPERLICH,D. AND PFRIEM, P. 1986.Chromosomal polymorph- HEwITT,0. M. 1967.An interchange which raises chiasma isms in natural and experimental populations. In The frequency. Chromosoma, 21, 285-295. Genetics and Biology of Drosophila, vol. 3e, chapter 37, HEWITT,G. M. 1979.Orthoptera. Grasshoppers and crickets. Academic Press, London, pp. 257-309. In Animal cytogenetics. 3. Insecta 1. (B. John, ed). Gebruder TEOH,S. B. AND YONG, M.s. 1983. A spontaneous fusion Born traeger, Berlin-Stuttgart. heterozygote in the tropical grasshopper, Valanga nigri- HEWITT,G. M. AND SCHROETER, 0. L. 1968.Population cornis (Burmeister). Caryologia, 36, 165-173. cytology of Oedaleonotus. 1. The karyotypic facies of TOSTO, D., MIROL, P. AND BIDAU, C. J. 1987. Aspectos de las (Scudder). Chromosoma, 25, 121- relaciones entre las fusiones céntricas y Ia distribución de 140. los quiasmas en Dichroplus pratensis (Melanoplinae, JOHN,B. Therole of chromosome change in the evolution of Acrididae). XVIII Congr. Arg. Genet., Buenos Aires, 21. Orthopteroid . In Sharma, A. K. and A. S. (Eds) WEISSMAN, D. B. 1976. Geographical variability in the pen- Chromosomes in the Evolution of Eukaryotes. Vol. 1, CRC centric inversion system of the grasshopper Trimerotropis Press, Boca Raton, pp. 1-110. pseudofasciata. Chromosoma, 55, 325-347. JONES,R. N. AND REES, H. 1982.B chromosomes. Academic WESTERMAN,M.1983. Chiasma frequency and variability Press, London-New York of morphological characters in populations of two grass- KAYANO,H. AND NAKAMURA, K. 1960.Chiasma studies in hopper species. Heredity, 51, 501—506. structural hybrids. V. Heterozygotes for a centric fusion WRITE,M. J.D. 1973. Animal cytology and evolution. 3rd Edit. and a translocation in Acrida lata. Cytologia, 25,476-480. Cambridge Univ. Press, London-New York LIEBERMANN,.j. 1963.Contribucibn al conocimiento de WHITE,M. J.D.AND MORLEY, F. H.W.1955.Effects of pen- Dichroplus pratensis (Orth., Acrid.) IDIA, 191, 23—28. centric rearrangements on recombination in grasshopper LOPEZ-FERNANDEZ.C., RUFAS, J. S., GARCIA DE LA VEGA, C. chromosomes. Genetics, 40, 604-619. AND GOSALVEZ, j. 1984.Cytogenetic studies on Chort- hippus jucundus (Fisch.) (Orthoptera). III. The meiotic consequences of a spontaneous centric fusion. Genetica, 63, 3-7.