Chromosome Numbers and Polyploidy in Life Forms of Asteraceae, Poaceae and Rosaceae in Polish Flora

ACTA BIOLOGICA CRACOVIENSIA Series Botanica 56/1: 7–15, 2014 DOI: 10.2478/abcsb-2014-0001

CHROMOSOME NUMBERS AND POLYPLOIDY IN LIFE FORMS OF ASTERACEAE, POACEAE AND ROSACEAE IN POLISH FLORA

GRZEGORZ GÓRALSKI1*#, MARCIN BAL1, PAULINA GACEK1, TOMASZ MARCIN ORZECHOWSKI2, AND AGNIESZKA KOSECKA-WIERZEJSKA1

1Department of Plant Cytology and Embryology, Jagiellonian University, Gronostajowa 9, 30-387 Cracow, Poland, 2AGH University of Science and Technology, Aleja Adama Mickiewicza 30, 30-059 Cracow, Poland

Received September 30, 2013; revision accepted January 7, 2014

The chromosome numbers and frequency of polyploids were compared in life forms of Asteraceae, Poaceae and Rosaceae. Both parameters were higher in Poaceae and Rosaceae than in Asteraceae. Among the life forms, long- lived plants including perennials and woody plants (shrubs and trees) generally had higher chromosome numbers and consequently polyploid frequencies than short-lived species (annuals and biennials). The families sur- veyed have different frequencies of life forms. Asteraceae and Rosaceae are both dicots, but the life forms in Asteraceae are more similar to Poaceae than to Rosaceae. To separate the influence of life form, in a series of tests we compared life forms from the same families. These results also showed that long-lived forms generally have more chromosomes than short-lived ones. Key words: Chromosome numbers, polyploidy, life forms, plants, evolution, Asteraceae, Rosaceae, Poaceae, threshold method, x method.

INTRODUCTION groups examined. This does not give estimates of polyploid frequency but several of the calculated sta- Polyploidization plays an important role in plant tistical values may give a broader view of the distri- evolution. Estimates of the frequency of polyploids bution of chromosome numbers, which may help in angiosperms range from 30 to 80% (Bennett, identify polyploidization events and to describe 2004) when traditional methods are applied. them using mathematical functions, as was recently However, it is known that the frequency of poly- done for the Italian flora (Bedini et al., 2012b). ploids and chromosome numbers differ between flo- Methods based on molecular data may give ras, geographical regions (e.g., Peruzzi et al., 2011; more precise results but are much more time-con- Peruzzi et al., 2012) and phylogenetic groups suming and expensive. However, they established (Bedini et al., 2012a; Bedini et al., 2012c). that even such small genomes as those of The simplest methods used to estimate poly- Arabidopsis thaliana (n=5) may be of polyploid ori- ploid frequency involve comparing the chromosome gin (Vision et al., 2000; Bowers et al., 2003). In numbers of cytotypes to given chromosome num- recent years, many similar studies have gradually bers, for example the x value for a genus or multi- changed our understanding of the role of polyploids ples of the lowest haploid number in a genus (Wood in plant evolution and have shown that almost all et al., 2009). The so-called threshold methods sim- angiosperms had at least one episode of poly- ply use an arbitrarily chosen n number. They are ploidization in their evolutionary history (Soltis, relatively uncomplicated and are particularly appli- 2005; Soltis et al., 2009). In light of such results it cable to large datasets for which not all x values are may seem pointless to study the frequency of poly- known. The most popular threshold values are n≥11 ploids in given groups of angiosperms. However, if (Goldblatt, 1980) and n≥14 (Grant, 1981). we regard the results obtained by chromosome Indirect information about polyploidy may be number methods as indicating the frequency of obtained by comparing chromosome numbers in the recently formed polyploids rather than the frequen-

*e-mail: [email protected] #Góralski contributed 80% to this paper

PL ISSN 0001-5296 © Polish Academy of Sciences and Jagiellonian University, Cracow 2014 8 Góralski et al. cy of plants that had at least one episode of poly- al., 2007). Poaceae (grasses) is one of the largest and ploidization in their history, they become a useful most widespread families and is very important in tool in the search for relations between ploidy and agriculture. Other interesting features of grasses are different factors or features of plants (Góralski et al., their wide range of x (2–18) and 2n (4 to more than 2014). 260) values achieved via such processes as chromo- Studies of polyploidy very often focus on the fea- some doubling, aneuploid reduction and hybridiza- tures of polyploids that might explain their evolution (Hilu, 2004). Rosaceae is also an economically tionary success by genetic and phenotypic superior- important group of angiosperms that contains about ity (e.g., Ronfort, 1999; Otto and Whitton, 2000; 100 genera and 3,000 species representing perenni- Soltis and Soltis, 2000; Levin, 2002; Wang et al., al, tree and shrub life forms. Phenomena that influ- 2006; Otto, 2007; te Beest et al., 2012) and features ence evolution, such as hybridization, polyploidiza- more directly related to competition and invasive- tion and apomixis, are frequently observed in these ness (Lowry and Lester, 2006; Pandit, 2006; Hull- families (Hummer and Janick, 2009). They are also Sanders et al., 2009; Treier et al., 2009; Mráz et al., the most data-rich families in our database 2011a,b; Pandit et al., 2011; te Beest et al., 2012; (Góralski et al., 2009). Góralski et al., 2014). For decades there has also The Polish flora is relatively young because in been an ongoing search for the relationship between the Pleistocene almost the entire area of the country chromosome number and life form, which may also was covered by a glacier (Szafer and Zarzycki, be related to evolutionary success. In the second 1977). This makes it an interesting area for observ- quarter of the 20th century, Stebbins (1938) com- ing many evolutionary phenomena. In recent pared the x values of dicots in temperate and boreal decades the chromosome numbers of ~40% of the zones and found that the average x value was 9 for Polish angiosperms have been examined (Gacek et herbaceous plants (perennials, annuals and biennial., 2011) and their 2n values have been collected in als) but 14.2 in woody plants, and that therefore the a database accessible online (Góralski et al., 2009). frequency of polyploids was lower in woody than in herbaceous plants, in which, according to Stebbins, the tendency toward polyploidy was manifested MATERIALS AND METHODS chiefly in perennial herbs. Similar observations have been reported for the DATA AND DATA SOURCES flora of Pakistan (Baquar, 1976; Khatoon, 1991; Khatoon and Ali, 2006). The general view in the first Chromosome numbers were obtained from the two of those cited papers is that polyploidy is most Chromosome Number Database (Góralski et al., common in perennials and less common in annuals, 2009), which contains the chromosome numbers followed by shrubs, and finally least common in known for Polish angiosperms. The life forms of the trees. Khatoon and Ali (2006) also found the lowest examined species were found in different sources frequencies of polyploids in trees and higher ones in (Szafer et al., 1953; Rutkowski, 2004; USDA, 2007; shrubs, although the life form they found most Snowarski, 2012). The x values were estimated abundant in polyploids was annual plants, which using data on somatic chromosome numbers and prevailed over perennial herbs. Another sequence of consulted with various syntheses and taxon-specific polyploid percentages was given by Haskell (1952) literature (previously described in Gacek et al., for the British flora: lowest in annuals, higher in 2011). trees and shrubs, and highest in perennial herbs. Some other studies found no such differences GROUPS TESTED between life forms: for example, Petit and Thompson (1999) for the flora of the Pyrenees, and For statistical analyses and comparisons at different Vamosi and Dickinson (2006) for Rosaceae. levels we created 19 groups containing the data col- Differences between the flora studied and the lected for three families: Asteraceae, Poaceae and methods used by the various authors make it diffi- Rosaceae. The first group contained all of the data cult to sort out their results and come to general (called "all tested"). The next three groups contained conclusions. Although studies in this area have been data for these three families separately done for at least 75 years, new research is still need- ("Asteraceae," "Poaceae," "Rosaceae"). Five groups ed if we are to understand the connections between contained chromosome numbers for six life forms ploidy and life form in plants. represented by at least ten cytotypes with self- We studied representatives of three angiosperm explanatory names – "perennial," "annual," "bienni- families in the Polish flora: Asteraceae, Poaceae and al," "shrubs" and "annual or biennial" (containing Rosaceae. The first is the largest family of species that can be annual or biennial). In addition, angiosperms, comprising over 1,600 genera and two groups containing two similar life forms were 23,000 species, and rich in life forms (Anderberg et created – "annual and biennial" and "trees and Chromosome numbers and polyploidy in Asteraceae, Poaceae and Rosaceae 9

TABLE 1. Comparisons and statistical significance of tests. Signs indicate the statistical significance of the difference (- P>0.01, + P≤0.01) 1st – chromosome number (Mann-Whitney test), 2nd – polyploid frequency at n≥11 threshold (χ2 test), 3rd – polyploid frequency at n≥14 threshold (χ2 test), 4th – polyploid frequency estimated by x value (χ2 test)

shrubs." When a given life form was represented Core Team, 2012). Two types of tests were applied only by members of one family, the name of the fam- for comparisons of groups: comparisons of chromo- ily in the tables and charts is given in parentheses some numbers and comparisons of the frequency of before the name of the life form and this life form is polyploids. Several values were calculated for the not displayed separately, when the life forms of a chromosome numbers and are given in the tables: given family are specified. Two groups were created mean and SD, mode, median, minimum and maxi- from life forms shared by two families: "perennial mum values. Differences in chromosome number (Asteraceae & Poaceae)" and "annual (Asteraceae & between groups were tested using the nonparamet- Poaceae)". The next groups contained data on the ric Mann-Whitney U (Wilcoxon) test (Wilcox.test chromosome numbers of life forms represented by a function from stats library). large enough number of cytotypes in each family: Additionally, the frequency of polyploids was "Asteraceae – annual," "Asteraceae – perennial," estimated using two threshold methods, n≥11 "Poaceae – annual," "Poaceae – perennial" and (Goldblatt 1980) and n≥14 (Grant 1981), and the "Rosaceae – perennial." As explained above, the x value. The frequency of polyploids in the studied groups "Asteraceae – biennial," "Rosaceae – shrubs" groups that had at least ten cytotypes was compared and "Rosaceae – trees & shrubs" are the same as using Pearson's χ2 test (chisq.test from stats "(Asteraceae) biennial", "(Rosaceae) shrubs" and library). This test was also used when the propor- "(Rosaceae) trees and shrubs," respectively, and are tions of life forms were calculated. not specified separately in the tables and charts. The Because multiple comparisons were used for all last two groups included two life forms with similar of the tests in our study (32 comparisons for every characteristics in two families ("Asteraceae – annual test) we took 0.01 as the significance level. and biennial", "Rosaceae – trees and shrubs"). All comparisons and the statistical significance of the results are shown in Table 1. RESULTS

We tested 1,645 cytotypes belonging to three fami- STATISTICAL TOOLS AND PROCEDURES lies (Asteraceae, Poaceae, Rosaceae). The chromo- All of the statistical tests and graphs employed the R some numbers (2n) of the tested cytotypes ranged environment for statistical computing (R Development from 6 to 96. The mean for the cytotypes was slight- 10 Góralski et al.

TABLE 2. Chromosome number parameters calculated for the tested groups

TABLE 3. Diploid and polyploid numbers and frequencies as estimated by n≥11 and n≥14 threshold methods and by x value; (%) Chromosome numbers and polyploidy in Asteraceae, Poaceae and Rosaceae 11

Fig. 1. Chromosome numbers in the tested groups. Horizontal bar – median; notched boxes – lower and upper quartiles; notches – +/-1.58 IQR divided by the Fig. 2. Basic chromosome numbers (x values) for the test- square root of the number of data elements (IRQ – the dif- ed groups. Horizontal bar – median; notched boxes – ference between upper and lower quartiles); whiskers – lower and upper quartiles; notches – +/-1.58 IQR divided the range of the data, excluding outliers; black dots – out- by the square root of the number of data elements (IRQ – liers (values that are farther than 1.5 IQR from the box). the difference between upper and lower quartiles); whiskers – the range of the data, excluding outliers; black dots – outliers (values that are farther than 1.5 IQR from the box). ly higher than 31, and the median and mode equalled 28 (Tab. 2, Fig. 1). Depending on the method adopted, the frequency of polyploids was 70% (threshold Asteraceae and Poaceae for the n≥14 threshold n≥11), 62% (n≥14) and 64% (x method) (Tab. 3). (Tab. 3). Among the tested families Asteraceae had the Perennials, which shared the highest mode (28) lowest mean (~30) and median (25.5). The medians and median (28) with shrubs and the group "trees for the other two families were the same (28) but the and shrubs" had the highest mean (~32.5). Biennial mode in Rosaceae (28) was twice that in Poaceae plants were the group with the lowest mode (12) of (14). The Mann-Whitney test did not indicate statis- chromosome number, but the lowest mean (22.5) tical significance (P>0.01). and median (19) were for "annual or biennial" The frequency of polyploids was highest for plants. This paper will not consider comparisons of Rosaceae for both thresholds (84.5%) and for the x groups containing many life forms versus the mem- method (84%). Asteraceae had the lowest share of bers of these groups, or comparisons of similar polyploids for both n≥11 (64%) and n≥14 (46%) groups (e.g., biennial vs. "annual and biennial"). and for the x method (52%). Asteraceae had a high- In the chromosome number comparisons of er median (9) and a much wider spread of x values groups of life forms, 3 of 17 comparisons gave than the other two families, for which the median results at P≤0.01: "annual or biennial" versus was 7. The differences in polyploid frequency were shrubs, "annual or biennial" versus "trees and significant (P≤0.01) between Rosaceae and the shrubs," and "annual and biennial" versus peren- other families for all three methods, and between nial. 12 Góralski et al.

For all three polyploid estimation methods the ploids was highest in Rosaceae (77–80%); in the frequency of polyploids was lowest in the short-lived other two groups the frequencies were at least 9% plants (30–55%), higher in perennials (57–70%) and lower; the χ2 test indicated significant differences at highest in woody plants (87–93%). The differences P≤0.01 for Asteraceae versus Poaceae (n≥14 thresh- in polyploid frequency were significant between old) and for Rosaceae versus Asteraceae (n≥14 shrubs and the other groups, and between "trees threshold and x method). and shrubs" and the other groups, short-lived plants Annuals were shared between Asteraceae and and perennials differed significantly in some cases Poaceae. In this case the mean chromosome num- for the n≥11 threshold and the x method but not for bers were similar (~26) but there were greater dif- the n≥14 threshold. ferences in the modes (Asteraceae 18, Poaceae 14) We compared annuals with perennials in two of and medians (Asteraceae 24, Poaceae 28). For fre- the three families, Asteraceae and Poaceae. quency of polyploids neither the Mann-Whitney test Perennials had a higher mean (~32 vs. 26), median nor the χ2 test showed significant differences (28 vs. 24) and mode (28 vs. 18) but the differences between families. were not significant. Next we compared life forms within families. Three life forms (annual, biennial, perennial) had at DISCUSSION least 10 cytotypes in Asteraceae. A common group (Asteraceae – annual and biennial) for short-lived Surprisingly, Rosaceae and Asteraceae (both dicots) plants that included annual, biennial and "annual or seem to differ more from each other than they differ biennial" was also created. Perennials had the high- from Poaceae (monocots). The mean and median est 2n mean (~30) and median (27), and "annual chromosome numbers were higher in Poaceae and and biennial" plants the lowest (~25.5 mean, 20 Rosaceae than in Asteraceae. This order is similar to median). The mode of chromosome numbers was that given recently for the Italian flora (Bedini et al., 12 for biennials and 18 for the rest of the forms in 2012c); also worth noting is that for Asteraceae and Asteraceae. In Asteraceae the differences between Poaceae but not for Rosaceae the mean chromosome "annual and biennial" and perennial were significant numbers for the Polish flora are higher than the (P≤0.01). Perennials had the highest polyploid fre- means from that Italian study. quency under all estimation methods (57–68%); in The order from polyploid-richest to polyploid- all the short-lived groups it ranged from 26 to 54%. poorest in our study is Rosaceae→Poaceae→ These differences were significant in some cases: for Asteraceae, but the differences between Asteraceae "annual and biennial" versus perennial for the n≥11 and Poaceae are weaker than between those two threshold and the x method, and for biennial versus families and Rosaceae. Thus, for the three families perennial for the x method. that were examined, the phylogenetic similarities Poaceae were represented by two groups with seem less responsible than other factors for the ten or more cytotypes (annuals, perennials). Of chromosome number distributions; life form appar- these, perennials had higher mean and mode chro- ently may be the factor chiefly responsible. mosome numbers but the differences were not sig- The examined life forms generally did not show nificant by the Mann-Whitney test or tests for poly- statistically significant differences in chromosome ploid frequency. number or polyploid frequency between the annual, Perennials and shrubs were two life forms hav- biennial and "annual or biennial" groups, so a com- ing more than 10 cytotypes in Rosaceae, also common group of short-lived plants was formed for mon group ("trees and shrubs") was created for all them: "annual and biennial" (short-lived plants). For woody plants: shrubs, trees and species that can be similar reasons the common group "trees and trees or shrubs. shrubs" (woody plants) was created. The perennials had higher mean (~34 vs. ~29) When the life forms were compared for chro- and median (31.5 vs. 28) chromosome numbers mosome numbers, perennials appeared to have than shrubs; the modes were the same (28). The higher chromosome numbers than "trees and polyploid frequencies tended to be higher in peren- shrubs," followed by "annuals and biennials." nials than in shrubs and all woody plants but these Polyploid frequency gives a different series: "trees differences were not significant. and shrubs" followed by perennials and then "annu- The next series of comparisons was between al and biennial" plants. groups of life form types from different families. The Our results are generally in line with those most common life form was perennial, so it was described for the British flora by Haskell (1952) and compared among all three families. Generally, all similar to earlier results from Stebbins (1938). The means, medians and modes for chromosome num- order that we arrived at, perennials→"trees and ber were similar and the Mann-Whitney test did not shrubs"→"annual and biennial" (with an occasional show significant differences. The frequency of poly- switch between perennials and woody plants), sug- Chromosome numbers and polyploidy in Asteraceae, Poaceae and Rosaceae 13 gests that the most obvious factor in increase of smaller box and the part below the median is very chromosome number is plant longevity. The differ- compact, which indicates a high concentration of the ences were seen mainly between short-lived plants same 2n values. Indeed, almost half of the chromo- and those that live many years, such as perennials some numbers are at 28, the value of the mode and and woody plants. The differences between the latter median. Shrubs are mainly responsible for the con- two groups were not so marked in our study and centration of values at 2n=28, found in over 62% of both are regarded as long-lived forms. the cytotypes (data not shown). The span of 2n val- Our results seem to support the idea that the ues is wider in the group of all woody species in differences between life forms are much more Rosaceae but still narrower than for the other important determinants of differences in chromo- groups. A similar concentration of chromosome some number than phylogenetic relations (though it numbers was also found more generally for the should be mentioned that Asteraceae and Rosaceae order Rosales in the Italian flora (Bedini et al., are not very closely related). A closer look at the data 2012c). shows that the situation is much more complicated. This concentration of chromosome numbers The important fact is that the distribution of life strongly influences the results of the tests for poly- forms between families is not correlated with phylo- ploid frequency. The threshold for the n≥14 method genetic relations. The families that are more similar is 2n=28; this means that most of the cytotypes are to each other (Asteraceae and Poaceae) in chromo- qualified as polyploids for both threshold methods, some number and polyploid frequency are also sim- resulting in a high representation of polyploids in ilar in their life forms. Both groups share annuals, woody plants and consequently in Rosaceae. This "annual or biennial" and perennials. Only the last explains why "trees and shrubs" are more abundant category, perennials, is shared by Rosaceae, which in polyploids than the perennials in our compar- is abundant in woody plants not present in the other isons. Such a strong influence of the exact 2n num- families. Thus, a comparison of woody plants versus ber on polyploid frequency as estimated by the short-lived plants is almost entirely a comparison of threshold method suggests that this method is over- part of Rosaceae versus part of Asteraceae and sensitive in such situations. The result would have Poaceae (two Rosaceae cytotypes are annuals). been dramatically different if the threshold had been In this situation, it is not a simple exercise to only one or two chromosomes more. It is doubtful distinguish the effects of life form and phylogeny on that such a small difference can really reveal the chromosome number or polyploidy. For example, it polyploid frequency. may be that Rosaceae has more chromosomes The third method, based on x values, should because it is abundant in long-lived plants, but it give more reliable results. Trees are generally may also be that the ancestors of that family had a regarded as having a higher basic chromosome high number of chromosomes and that this is why number than herbaceous plants (Stebbins, 1938; perennials and woody plants in that family have high Levin and Wilson, 1976) but our dataset does not chromosome numbers. The first of those explana- show that (Fig. 2). The median x values for short- tions may be supported by the higher chromosome lived plants were higher (7.5–9) than for woody numbers in perennials of Rosaceae generally than in plants (Rosaceae) and for the perennials of Rosaceae the other families examined. and Poaceae (7), and they had a much wider spread, To separate the influence of life form from phy- mainly at the high values. The 2n values for the logenetic factors we performed the series of com- groups within Rosaceae were rather high, so conse- parisons of life forms from the same families. The quently the polyploid frequency as estimated by the results indicated that the differences between the life x method was also high. However, for short-lived forms within the tested families are comparatively plants (especially in Asteraceae), high x values and weak, but there were some differences between low 2n numbers resulted in comparatively low poly- short-lived and long-lived species. The weaker dif- ploid frequencies. ferences between life forms from the same family Although the differences in chromosome num- may suggest that taxonomy may yet be relevant here, ber distributions between Rosaceae and the other but our results suggest that the relation between two families obscure the general gradation of chro- chromosome number and the phylogenetics of the mosome number from perennials to woody plants to studied families is not a straightforward one. short-lived "annual and biennial" plants, this pattern Figure 1 may shed some light on this problem. is easily seen in Figure 1. The distributions of the chromosome numbers of Perennials were also found to be richest in Asteraceae and Poaceae are wide, and although the polyploids in the flora of Pakistan (Baquar, 1976; medians are different, the ranges of values overlap Khatoon, 1991), but other studies suggest that significantly and therefore the statistical tests do not annual plants are the most polyploid-rich result in significant differences between these (Khatoon and Ali, 2006). Such a difference proba- groups. By contrast, Rosaceae is represented by a bly is connected with the different histories of the 14 Góralski et al. floras of Pakistan and Poland; the latter is thought numbers and higher polyploid frequencies to have formed mainly during the Holocene after than short-lived species (annuals and bienni- the last glaciation (Szafer and Zarzycki, 1977). als). These results are similar to those for the Also, alien plants in the Polish flora generally have British flora but differ from some results from lower chromosome numbers and lower polyploid Pakistan. frequency than native plants (Góralski et al., 6. Threshold methods may be oversensitive in sit- 2014), and alien plants may be more common in uations in which, as in Rosaceae, the most fre- annuals (and biennials) than in perennial or quent 2n value is 28, the threshold for the n≥14 woody plants; that hypothesis needs to be con- method. firmed in future studies. The higher polyploid frequency in perennial REFERENCES herbs than in short-lived herbs may be explained by connecting polyploidization with hybridization – ANDERBERG AA, BALDWIN BG, BAYER RG, BREITWIESER J, hybrids frequently are not fertile, but fertility may be JEFFREY C, DILLON MO, ELDENÄS P, FUNK V, GARCIA-JACAS N, reestablished by polyploidization, which is much and HIND DJN. 2007. Compositae. In: Kadereitt JW, more probable in long-lived perennials that can Jeffrey C [ed.], Families and Genera of Vascular Plants, reproduce vegetatively than in herbs that live for only vol. 8, Flowering Plants· Eudicots, 61–588. Springer one or two years (Stebbins, 1938). Under such Berlin Heidelberg. assumptions, "trees and shrubs" should have a high BAQUAR SR. 1976. Polyploidy in the flora of Pakistan in rela- Taxon chromosome number, and indeed they seem to have tion to latitude, life form, and taxonomic groups. 25: 621–627. more chromosomes than "annual and biennial" plants BEDINI G, GARBARI F, and PERUZZI L. 2012a. Karyological but fewer than perennials. Stebbins explained that knowledge of the Italian vascular flora as inferred by the polyploidy disturbs the size relationships of the cells analysis of "Chrobase. it". Plant Biosystems 146: so that wood fibers cannot be produced satisfactorily 889–899. by cambium, and he also noted that trees are cytolog- BEDINI G, GARBARI F, and PERUZZI L. 2012b. Chromosome ically stable, but the relations between ploidy level number variation of the Italian endemic vascular flora. and anatomical and histological features have not yet State-of-the-art, gaps in knowledge and evidence for an been thoroughly examined and understood. exponential relationship among even ploidy levels. The relationships between chromosome number Comparative Cytogenetics 6: 191–211. or polyploidy and life forms are not yet clear. Our BEDINI G, GARBARI F, and PERUZZI L. 2012c. Does chromosome results add some valuable data to the discussion. number count? Mapping karyological knowledge on Analyses and comparisons of more families and the Italian flora in a phylogenetic framework. Plant Systematics and Evolution life forms within them are needed. The differences in 298: 739–750. chromosome numbers between life forms may be cor- BENNETT MD. 2004. Perspectives on polyploidy in plants – ancient and neo. Biological Journal of the Linnean related with other features such as the ability to colo- Society 82: 411–423. nize new areas or invasiveness, as recently found for BOWERS JE, CHAPMAN BA, RONG J, and PATERSON AH. 2003. the Polish flora (Góralski et al., 2014). Unravelling angiosperm genome evolution by phylogenetic analysis of chromosomal duplication events. Nature 422: 433–438. CONCLUSIONS GACEK P, GÓRALSKI G, and JOACHIMIAK AJ. 2011. Chromosome numbers and polyploidy in Polish angiosperms. Acta 1. Chromosome numbers and polyploid frequen- Biologica Cracoviensia Series Botanica 53/2: 37–49. cies were described for three families GOLDBLATT P. 1980. Polyploidy in angiosperms: monocotyle- (Asteraceae, Poaceae, Rosaceae) in the Polish dons. In: Lewis WH [ed.], Polyploidy: Biological flora and for their life forms. Relevance. Plenum Press, New York, London. 2. Asteraceae and Rosaceae are both dicots but GÓRALSKI G, JUDASZ A, GACEK P, GRABOWSKA-JOACHIMIAK A, Asteraceae is more similar in life forms to and JOACHIMIAK AJ. 2014. Polyploidy, alien species and Poaceae than to Rosaceae. invasiveness in Polish angiosperms. Plant Systematics and Evolution 300: 225–238. 3. Poaceae and Rosaceae generally have higher GÓRALSKI G, LUBCZYŃSKA P, and JOACHIMIAK AJ. 2009. chromosome numbers and higher polyploid fre- Chromosome Number Database. http://chromosomes. quencies than Asteraceae. binoz.uj.edu.pl/, accessed 2012. 4. The distribution of chromosome numbers is GRANT V. 1981. Plant Speciation. Columbia University Press, comparatively narrow in Rosaceae (in most New York. cases 2n=28) and for its woody plants, espe- HASKELL G. 1952. Polyploidy, ecology and the British flora. cially shrubs. Wider distributions were found in The Journal of Ecology 40: 265–282. Asteraceae and Poaceae and their life forms. HILU KW. 2004. Phylogenetics and chromosomal evolution in 5. Long-lived plants (perennials, trees and the Poaceae (grasses). Australian Journal of Botany 52: shrubs) generally have higher chromosome 13–22. Chromosome numbers and polyploidy in Asteraceae, Poaceae and Rosaceae 15

HULL-SANDERS HM, JOHNSON RH, OWEN HA, and MEYER GA. RONFORT J. 1999. The mutation load under tetrasomic inher- 2009. Effects of polyploidy on secondary chemistry, itance and its consequences for the evolution of the self- physiology, and performance of native and invasive geno- ing rate in autotetraploid species. Genetical Research types of Solidago gigantea (Asteraceae). American 74: 31–42. Journal of Botany 96: 762–770. RUTKOWSKI L. 2004. Klucz do Oznaczania Roślin HUMMER KE, and JANICK J. 2009. Rosaceae: taxonomy, eco- Naczyniowych Polski Niżowej. Wydawnictwo Naukowe nomic importance, genomics. In: Folta KM, Gardiner SE PWN, Warszawa. [eds.], Genetics and Genomics of Rosaceae. Springer, SNOWARSKI M. 2012. Atlas of Vascular Plants of Poland. New York. http://www.atlas-roslin.pl/ KHATOON S. 1991. Polyploidy in the flora of Pakistan and ana- SOLTIS PS, and SOLTIS DE. 2000. The role of genetic and lytical study. Ph.D. dissertation, University of Karachi. genomic attributes in the success of polyploids. KHATOON S, and ALI SI. 2006. Chromosome numbers and Proceedings of the National Academy of Sciences 97: polyploidy in the legumes of Pakistan. Pakistan Journal 7051–7057. of Botany 38: 935–945. SOLTIS DE, ALBERT VA, LEEBENS-MACK J, BELL CD, PATERSON LEVIN DA. 2002. The Role of Chromosomal Change in Plant AH, ZHENG CF, SANKOFF D, DEPAMPHILIS CW, WALL PK, Evolution. Oxford University Press, Oxford. and SOLTIS PS. 2009. Polyploidy and angiosperm diver- LEVIN DA, and WILSON AC. 1976. Rates of evolution in seed sification. American Journal of Botany 96: 336–348. plants: Net increase in diversity of chromosome numbers SOLTIS PS, SOLTIS DA, ENDRESS PK, and CHASE MW. 2005, and species numbers through time. Proceedings of the Phylogeny and Evolution of Angiosperms, Sinauer National Academy of Sciences 73: 2086–2090. Associates, Inc., Sunderland. LOWRY E, and LESTER SE. 2006. The biogeography of plant STEBBINS GL. JR. 1938. Cytological characteristics associated reproduction: potential determinants of species' range with the different growth habits in the dicotyledons. sizes. Journal of Biogeography 33: 1975–1982. American Journal of Botany 25: 189–198. MRÁZ P, BOURCHIER RS, TREIER UA, SCHAFFNER U, and MÜLLER- SZAFER W, KULCZYŃSKI S, and PAWŁOWSKI B. 1953. Rośliny SCHÄRER H. 2011a. Polyploidy in phenotypic space and Polskie. Państwowe Wydawnictwo Naukowe, Warszawa. invasion context: a morphometric study of Centaurea SZAFER W, and ZARZYCKI K. 1977. Szata Roślinna Polski. T. I, stoebe sl. International Journal of Plant Sciences 172: Państwowe Wydawnictwo Naukowe,Warszawa. 386–402. TE BEEST M, LE ROUX JJ, RICHARDSON DM, BRYSTING AK, SUDA J, MRÁZ P, GARCIA-JACAS N, GEX-FABRY E, SUSANNA A, BARRES L, KUBESOVA M, and PYSEK P. 2012. The more the better? and MÜLLER-SCHÄRER H. 2011b. Allopolyploid origin The role of polyploidy in facilitating plant invasions. of highly invasive Centaurea stoebe s.l. (Asteraceae). Annals of Botany 109: 19–45. Molecular Phylogenetics and Evolution 62: 612–623. TREIER UA, BROENNIMANN O, NORMAND S, GUISAN A, SCHAFFNER U, OTTO SP. 2007. The evolutionary consequences of polyploidy. STEINGER T, and MÜLLER-SCHÄRER H. 2009. Shift in cyto- Cell 131: 452–462. type frequency and niche space in the invasive plant OTTO SP, and WHITTON J. 2000. Polyploid incidence and evo- Centaurea maculosa. Ecology 90: 1366–1377. lution. Annual Review of Genetics 34: 401–437. USDA, NRCS. 2013. The Plants Database. http://plants. PANDIT MK. 2006. Continuing the search for pattern among usda.gov. National Plant Data Team, Greensboro, NC rare plants: are diploid species more likely to be rare? 27401-4901 USA. Evolutionary Ecology Research 8: 543–552. VAMOSI JC, and DICKINSON TA. 2006. Polyploidy and diversifi- PANDIT MK, POCOCK MJO, and KUNIN WE. 2011. Ploidy influ- cation: a phylogenetic investigation in Rosaceae. ences rarity and invasiveness in plants. Journal of International Journal of Plant Sciences 167: 349–358. Ecology 99: 1108–1115. VISION TJ, BROWN DG, and TANKSLEY SD. 2000. The origins of PERUZZI L, GÓRALSKI G, JOACHIMIAK AJ, and BEDINI G. 2012. genomic duplications in Arabidopsis. Science 290: Does actually mean chromosome number increase with 2114–2117. latitude in vascular plants? An answer from the compar- WANG J, TIAN L, LEE HS, WEI NE, JIANG H, WATSON B, ison of Italian, Slovak and Polish floras. Comparative MADLUNG A, OSBORN TC, DOERGE RW, and COMAI L. Cytogenetics 6: 371–377. 2006. Genomewide nonadditive gene regulation in PERUZZI L, DAWSON MI, and BEDINI G. 2011. Chromosome Arabidopsis allotetraploids. Genetics 172: 507–517. number variation in two antipodean floras. AoB PLANTS WOOD TE, TAKEBAYASHI N, BARKER MS, MAYROSE I, GREENSPOON 2011 plr020 doi:10.1093/aobpla/plr020. PB, and RIESEBERG LH. 2009. The frequency of polyploid PETIT C, and THOMPSON JD. 1999. Species diversity and eco- speciation in vascular plants. Proceedings of the logical range in relation to ploidy level in the flora of the National Academy of Science of the United States of Pyrenees. Evolutionary Ecology 13: 45–65. America 106: 13875–13879.