Page 8 - Mantovani1998
P. 8
satDNA Structure in Unisexual Stick Insects 1295
Table 1 Table 2
Distance Mean Values and Standard Errors (SES) for Student’s t-Test on Bacillus p-Distance Values
Bacillus Samples
df t P
N Mean Ϯ SE 1. GG vs. GB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 3.01 **
2. GM vs. GB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.52 **
a. B. grandii . . . . . . . . . . . . . . 136 0.126 Ϯ 0.003 3. GG vs. GM . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.16 NS
b. GG . . . . . . . . . . . . . . . . . . . . 10 0.102 Ϯ 0.008 4. GG, GB, GM vs. GG-GB-GM. . . . . . . . . . . . . 134 13.06 **
c. GB . . . . . . . . . . . . . . . . . . . . 36 0.079 Ϯ 0.003 5. AAT vs. ACY . . . . . . . . . . . . . . . . . . . . . . . . . . 7 3.18 NS
d. GM . . . . . . . . . . . . . . . . . . . 3 0.121 Ϯ 0.014 6. ACA vs. ACY . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.56 NS
e. GG, GB, GM . . . . . . . . . . . 49 0.086 Ϯ 0.003 7. AAT vs. ACA . . . . . . . . . . . . . . . . . . . . . . . . . . 7 0.06 NS
f. GG-GB-GM. . . . . . . . . . . . . 87 0.148 Ϯ 0.003 8. AAT, ACY, ACA vs. AAT-ACY-ACA . . . . . . 43 0.50 NS
g. B. atticus . . . . . . . . . . . . . . . 45 0.115 Ϯ 0.003 9. GG vs. AAT. . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 0.46 NS
h. AAT. . . . . . . . . . . . . . . . . . . 6 0.107 Ϯ 0.004 10. GG, GB, GM vs. AAT, ACY, ACA. . . . . . . . . 59 3.53 ***
i. ACY . . . . . . . . . . . . . . . . . . . 3 0.131 Ϯ 0.006 11. GG-GB-GM vs. AAT-ACY-ACA. . . . . . . . . . . 118 6.53 ***
l. ACA . . . . . . . . . . . . . . . . . . . 3 0.106 Ϯ 0.014 12. B. grandii vs. B. atticus . . . . . . . . . . . . . . . . . . 179 1.78 NS
m. AAT, ACY, ACA . . . . . . . 12 0.112 Ϯ 0.005 13. BLYgrandii vs. BLYatticus . . . . . . . . . . . . . . . 274 24.60 **
n. AAT-ACY-ACA . . . . . . . . . 33 0.116 Ϯ 0.003 14. BLYgrandii vs. GG. . . . . . . . . . . . . . . . . . . . . . 179 7.11 ***
o. BLYgrandii. . . . . . . . . . . . . 0.061 Ϯ 0.001 15. BLYatticus vs. AAT . . . . . . . . . . . . . . . . . . . . . 109 1.55 NS
p. BLYatticus . . . . . . . . . . . . . 171 0.122 Ϯ 0.002
105
NOTE.—Mean values of p-distances obtained between Bag320 monomers of: NOTE.—Abbreviations are as in table 1. NS, not significant; * 0.01 Ͻ P Ͻ
B. grandii grandii, GG; B. grandii benazzii, GB; B. grandii maretimi, GM; B. 0.05; ** 0.001 Ͻ P Ͻ 0.01; *** P Ͻ 0.001.
atticus atticus, AAT; B. atticus cyprius, ACY; B. atticus carius, ACA; B. lyn-
ceorum grandii grandii–like sequences, BLYgrandii; B. lynceorum atticus–like showing a significantly lower mean p-distance value. It Downloaded from http://mbe.oxfordjournals.org/ by guest on April 9, 2016
sequences, BLYatticus. GG, GB, GM and AAT, ACY, ACA make up the b, c, might seem, therefore, that the same mechanisms and
d and h, i, l samples, respectively. GG-GB-GM and AAT-ACY-ACA values are rates of genomic turnover are at work in both reproduc-
obtained from comparisons between clones pertaining to the different B. grandii tive systems, allowing comparable levels of variability
and B. atticus subspecies, respectively. Bacillus grandii and B. atticus samples as measured by nucleotide substitutions in the majority
comprise both intrasubspecific (GG, GB, GM or AAT, ACY, ACA) and inter- of the samples. It should be noted that intra- and even
subspecific (GG-GB-GM or AAT-ACY-ACA) p-distances obtained for the two interchromosomal exchanges are possible in an autom-
taxa. N ϭ number of p-distances obtained from pairwise sequence comparisons. ictic system. On the other hand, very different trends
emerge when intersubspecific p-distances are consid-
grandii grandii–like sequences within the majority of ered. For B. grandii, a value significantly higher than
populations, the only exception being a Mascalucia sam- that observed for intrasubspecific comparisons is found,
ple with all clones pertaining to the same cluster (the thus indicating that sequence homogeneity is higher
grandii grandii one). At present, this situation does not within than between B. grandii subspecies. Bacillus at-
appear to invoke any special explanation, possibly being ticus shows comparable p-distance values (and therefore
due to a sampling error that occurred in the choice of sequence similarity) at both intra- and intersubspecific
clones to be sequenced. In addition, three out of four levels. All this could be explained by bisexuality acting
Melilli clones cluster all in one group (the atticus one). strongly on sequence variant homogenization/fixation in
It must also be pointed out that within each cluster, B. the same genomic pool and by parthenogenetic repro-
lynceorum sequences do not follow any geographical duction leaving random levels of individual variability
trend. This behavior has already been noticed (and is owing to the absence of chromosomal reassortment in
here observed again) for B. atticus monomers (Manto- the progeny.
vani et al. 1997). Therefore, a basic character of parthe-
nogenetic systems seems to be the absence of intrapo- In the hybrid B. lynceorum, the Bag320 satellite
pulation or intrasubspecific higher sequence similarity, monomers differ qualitatively and depart significantly in
which is possibly linked to unisexual reproduction: the terms of p-distance values. The nucleotide composition
absence of chromosome exchange among individuals at the diagnostic loci of the clones Mel12, Not14, and
hampers sequence variant fixation (sensu Dover 1986) Rag22 suggests the possibility that gene conversion has
in the population. occurred between an atticus- and a grandii grandii–like
monomer. On the other hand, no clones were found that
The dendrogram also shows quite directly that could be the products of crossing over between the two
while all B. atticus clones are completely intermingled kinds of repeats. On the whole, this supports the pos-
with the related B. lynceorum sequences, the grandii sibility of interchromosomal exchange occurring even in
grandii cluster is further split in two subgroups, one an apomictic, as has also been demonstrated for ribo-
comprising only clones of the parental species. The somal DNA repeats in obligately parthenogenetic clones
higher sequence similarity between the grandii grandii– of Daphnia pulex and in the hybrid lizards of the Het-
like sequences of B. lynceorum and GG/Not4 and 10 eronotia binoei complex (Crease and Lynch 1991; Hillis
clones suggests that only a few lineages participated in et al. 1991).
the hybridization events leading to the triploid hybrid.
On the other hand, the coexistence of two distinct
The statistical analysis on sequence similarity per- kinds of sequences in the same genomic complement
formed here, although on limited samples in some in- and the so far limited homogenization between them can
stances, indicates that at the population level, the bisex- be explained taking into account the apomictic game-
ual B. grandii and the unisexual B. atticus may behave
comparably, with only the B. grandii benazzii sample
