Morphological characteristics as a key attribute for a successful determination of selected Cotoneaster species

© Slovak University of Agriculture in Nitra Faculty of Agrobiology and Food Resources

The classification of taxa included in Cotoneaster genus is relatively problematic. Morphological features, such as colour of fruit and bark or shape and size of leaves, are not always reliable attributes. A pome is a type of fruit specific to subfamily formerly known as Maloideae. In Cotoneaster, the ovary develops into the pyrene (Rohrer, 1991). The total amount of pyrenes in a single pome is variable. The number of pyrenes in pome and the number of pomes in infructescence are the most important morphological features for determination of previously mentioned Cotoneaster species. Other features include the presence of trichomes on leaves and pomes, and the colour of pomes (Baranec, 1992). Since we believe, that there is a lack of information and available data from contemporary articles, we decided to prove, whether these morphological characteristics are reliable for identification of selected Cotoneaster species. We also wanted to prove, whether it is possible to rely only on these selected characteristics. And last but not least, it should be mentioned, that such a large-scale study in the Western Carpathians has been lacking so far.

Material and methods
The majority of collected individuals grew in dry habitat with a permanent sunlight, mainly on grassy steppes or cliff edges. Some of the individuals grew in forest in the community of other shrubs and larger trees. Each population was relatively small and it included 1-5 individuals, rarely 6-10(-15). Pomes and infructescences were collected from various localities in four countries situated in the Western Carpathians -the Czech Republic, Hungary, Poland and Slovakia ( Figure  1, 2). Infructescences were calculated in the field and from herbarium specimens stored in the herbarium NI (Herbarium Collection of Slovak University of Agriculture in Nitra). The list of localities selected for counting of pyrenes in pome and for counting of pomes in infructescence is listed in Appendix 1 and 2, respectively. Individuals were assigned to different species by the colour of pomes and the presence of trichomes according to determination key of Baranec (1992).
Samples were divided in two groups, one designed for counting of pyrenes in pome and the other for counting Localities within the Western Carpathians selected for counting of pyrenes in pome: red circles -Cotoneaster integerrimus, blue diamonds -C. melanocarpus agg., yellow squares -C. tomentosus. Background terrain layer is from Stamen Design with data by OpenStreetMap.
of pomes in infructescence. Overall, we collected 3-4 twigs per individual. Herbarium specimens from NI herbarium (Herbarium Collection of Slovak University of Agriculture in Nitra) were also used to increase the total amount of infructescences. The collection of samples, designed for counting of pyrenes in pome, included 2,353 pomes of >130 individuals. The collection of samples, designed for counting of pomes in infructescence, included 1019 infructescences of 141 individuals. The data were collected and further statistically analysed by Microsoft® Excel 2010 and STATISTICA, version 10 (StatSoft, Inc., 2011) data analysis software system (Tukey's HSD test). Values were calculated as a mean per species.
Herbarium abbreviations are according to Thiers (2019). Nomenclature of flowering plants follows Marhold and Hindák (1998). The map was created by the software QGIS, version 3.2 (QGIS Development Team, 2018) with QuickMapServices plug-in and terrain background layer from Stamen Design with data by OpenStreetMap.

Figure 2
Localities within the Western Carpathians selected for counting of pomes in infructescence: red circles -Cotoneaster integerrimus, blue diamonds -C. melanocarpus agg., yellow squares -C. tomentosus. Background terrain layer is from Stamen Design with data by OpenStreetMap.
We also compared results from this paper, obtained by counting of pomes in infructescence, with other published data. These results are listed in Table 4.
The vast majority of C. integerrimus pomes grew one by one and did not form infructescences. Single pome represented 86.96 % of the collection. Infructescences with 2 pomes represented 11.88 % and infructescences with 3 pomes represented only 1.16 % of the collection. Number of pomes in infructescences of C. melanocarpus agg. ranged mainly from 1 to 2. Single pome represented 57.42 % of the collection and infructescences with 2 pomes represented 33.09 % of the collection. Infructescences with 3 pomes (8.05 %), 4 pomes (1.25 %) and 5 pomes (0.18 %) were also present. Number of pomes in infructescences of C. tomentosus ranged mainly from 1 to 2. Single pome represented 61.74 % of the collection. Infructescences with 2 pomes represented 28.70 % of the collection. Infructescences with 3 pomes (6.96 %) and 4 pomes (2.61 %) were also present.
Nowadays, modern methods, such as flow cytometry method, are very important in taxonomy of plants (Hajrudinović et al., 2015;Macková et al., 2017;Žabka et al., 2018), which is a big step forward in systematic botany. These methods in combination with classic morphological and population biology research (Rohrer et al., 1991;Ďurišová, et al., 2016;Ďurišová and Baranec, 2016) give us the opportunity to solve many different problems in complicated groups, like Rosaceae.
The purpose of these morphological analyses was to determine a difference between two closely related tetraploid species C. integerrimus and C. melanocarpus agg.) (Rothleutner et al., 2016), since C. tomentosus is considered to be pentaploid (Macková et al., 2018). Flow cytometry method, used in the last study, revealed no significant difference in the genome size of tetraploid C. integerrimus and C. melanocarpus agg., including C. matrensis (Kšiňan et al., 2019). Therefore, we decided to prove, if these morphological characteristics are reliable for identification of selected tetraploid Cotoneaster species.

Conclusion
Results from this paper proved, that even in the era of modern cytological and molecular methods, classic morphological methods are still relevant in the field of botany. However, we do not recommend relying only on morphological characteristics, because in some cases a determination of problematic closely related species can be more complicated. We suggest a combination of available methods, from classic morphological research to modern methods, or at least a combination of several morphological characteristics.