V.L. Komarov botanical records from eastern Asia from 1895 to 1897

The datasets were digitised by Hui Kim (data manager), Chin S. Chang was the resource creator and Kae Sun Chang and Chin-Sung Chang were the content providers. With the help of Brahms (Filer 2001), C.S. Chang identified taxonomic changes in Komarov's nomenclature and geo-referenced in the species occurrence data.

The project aims to mobilise and publish the data about V.L. Komarov's collections deposited at LE and other herbaria and primary occurrence data cited in Flora Manshuriae (Komarov 1901, Komarov 1904, Komarov 1907, Komarov 1927a, Komarov 1927b, Komarov 1927c, Komarov 1927d, Komarov 1927e, Komarov 1927f, Komarov 1932a, Komarov 1932b, Komarov 1932c).

This research was supported by Korea National Arboretum (Data-mining project).

Within the LE Herbarium, two groups have been defined for data mobilisation purposes. The first group consists of herbarium specimen records (Fig. 1) of vascular plants collected in the Far East of Russia, northeast China and North Korea by Komarov. V.L. Komarov collected 1,435 specimens in 1895, 1,834 specimens in 1896 and 3,687 specimens in 1897, along with plant records for 51 families that he identified. These data describe a specimen dataset of East Asian plants preserved in the Komarov Botanical Institute (LE) and duplicates distributed to several other institutes (like V, TI, NY, A, Stafleu and Cowan (1979)). The second group includes the primary occurrence data that he observed in the field. Komarov devoted three years to botanical research, meticulously documenting the species he observed in the field. In 1895, he documented 2,270 occurrences from the Amur and Ussuri Rivers in Russia (Fig. 2). The following year, 1896, he extended his observations, documenting another 3,872 occurrences from Primorsky and Amur in Russia and Jilin, Liaoning and Heilongjiang in China (Fig. 3). By 1897, Komarov's efforts had culminated in the recording of another 8,016 occurrences from Hamgyong-bukto, Ryanggang, Chagang-do in North Korea and Jilin, Liaoning in China (Fig. 4). The data from the Komarov's literature were transferred to the draft tables according to the GBIF requirements for occurrence datasets (Robertson et al. 2014).

Figure 1.  

An example specimen label, Komarov1469, registered as Diervilla florida (Bunge) Siebold & Zucc as well as the syntype of Weigela praecox var. tomentosa Nakai. in the Herbarium of the University of Tokyo (TI), was collected from ‘Ostium Fluvii Tumin-gan’, a place name georeferenced to be located in Undok, Hamgyong-bukto, Democratic People's Republic of Korea. It was currently using the accepted name, Weigela florida (Bunge) A.DC.

Figure 2.  

The map of the collecting area for the 1895 V.L. Komarov's botanical expedition. The route is not shown in detail on the map for the sake of clarity. However, the expedition actually started in Ussuri in April 1895. It ended in Pompeyevka in October 1895.

Figure 3.  

The map of the collecting area for the 1896 V.L. Komarov's botanical expedition. The route is not shown in detail on the map for the sake of clarity. However, the expedition actually started in Blagoveshchensk on the Amur in April 1896. It ended in Vladivostok in October 1896.

Figure 4.  

The map of the collecting area for the 1897 V.L. Komarov's botanical expedition. The route is not shown in detail on the map for the sake of clarity. However, the expedition actually started in Vladivostok in April 1897. It ended in Tumin-gan near Langsui-chuinda in November 1897.

Figure 5.  

The portrait of V. L. Komarov (1869-1945) from Wikimedia common website.

Georeferencing: In the case of Komarov's geographic name information, the original Russian-language Flora Manshuriae (Komarov 1901, Komarov 1904, Komarov 1907) transcribed the locality names in Russian Cyrillic, while the later Japanese-translated version (Komarov 1927a, Komarov 1927b, Komarov 1927c, Komarov 1927d, Komarov 1927e, Komarov 1927f, Komarov 1932a, Komarov 1932b, Komarov 1932c) by the South Manchurian Railway Company presents the place names, based on Chinese characters. These two different scribal records, the historically recorded Chinese characters in northern China and on the Korean Peninsula and Komarov's early gazetteer with its corresponding transliterations, are the main sources of information for the georeferencing calculations in this dataset. In many Asian countries, Japanese exonyms are geographic names in the Japanese language that differ from those in the dominant language and these Japanese terms for some names are now a mystery, either because they are quite different from endonyms or because of some other obscure etymology, making it difficult to determine past collection sites (Chang et al. 2021a). Komarov's gazetteer existed before the Japanese altered the place names, and the Japanese version of the gazetteer helped rather than hindered our georeferencing because it relied on locally used Chinese characters rather than Japanese exonyms. Chang et al. (2015) has already published a multilingual botanical gazetteer, which was used in this study to resolve inconsistencies, uncertainties and confusion about the geographic names on the Korean Peninsula used by foreign explorers and Korean plant collectors over the past 120 years. Once the place names have been identified, the next step was to provide an accurate match to a biological collection. We always aimed for accurate georeferencing for place coordinates, but sometimes this was not possible due to insufficient information in the place names. In these situations, we used higher geographic area coordinates, such as counties or cities. To minimise errors, improve data consistency and maintain integrity throughout the georeferencing process, we modified a procedure used by the Chinese Type Collection Project (Lohonya et al. 2020).

1. The data mobilisation phase includes the extraction of data from the three volumes of Komarov's Flora Mandshuiae, the filling of the initial datasets and the georeferencing. Quality control is carried out at all stages and includes a final cross-check of the data. In particular, the master files and initial datasets will be archived in the BRAHMS database. We used the taxon module provided by BRAHMS to manage scientific names and we used the verbatim scientific names given by Komarov to reflect the most recent nomenclature (POWO 2024), together with the regional checklist (Chang et al. 2020).

2. The botanical records are taken from the three volumes of Flora Manshuriae, the lists of available material (gazetteers and data) being synchronised and updated according to the latest taxonomy. When only literature was available for species occurrence records, the scientific names recorded in Komarov's flora were transcribed into the current scientific names, with typographical errors removed. The rank of the original scientific name and the corresponding taxonomic status were retained. Where specimens were available, the earliest scientific name recorded on the label was checked against literature and the earliest scientific name recorded was chosen as the default scientific name. For types of new taxa by Komarov as well as other authors, the scientific name of the corresponding type is given in the "typeStatus", for example, "Isosyntype of Acer tschonoskii var. rubripes Kom.". Where the accepted name differed, we presented it in a column called 'acceptedNameUsage'.

3. A multilingual gazetteer has been produced to resolve the confusion of place names in northeast China, Far Eastern Russia and North Korea. The gazetteer thesaurus was created as a tool to correlate the often radically different names given to a single place or feature and to provide geographic coordinates for each.

4. The geographical information on the labels of this period included not only the name of the place where the specimen was collected, but also other information, such as a topographic description of the river or its location on a particular mountain. To know the distribution of a plant species, it is first necessary to georeference the collections so that they can be plotted on a map.

Komarov journeyed through Amur Oblast, Jewish Autonomous Oblast and Khabarovsk Krai regions of Russia in 1895. The following year, in 1896, he explored the Primorsky Krai, Heilongjiang and Jilin regions of China. Subsequently, in 1897, he revisited the Primorsky Krai in Russia. His travels continued to lead him to Rason, Hamgyongbuk-do, Ryangang and Chagang in North Korea, as well as Liaoning and Jilin in China.

40.88 and 49.46 Latitude; 122.87 and 133.43 Longitude.

The majority of specimen records and primary occurrence data belong to class Magnoliopsida (4,669 specimens + 12,228 occurrence data from literature) and Liliopsida (1,605 specimens +1,337), followed by Polypodiopsida (445 specimens + 442), Lycopodiopsida (54 specimens + 43) and Coniferophyta (183 specimens + 108). Our dataset represents 123 families (Fig. 6), of which 13.9% of the data belong to the monocot families Poaceae (3.3%), Cyperaceae (2.5%), Orchidaceae (1.8%), Asparagaceae (1.3%) and Liliaceae (1.2%). The dicot comprises 80.0% of the total, followed by families Asteraceae (11.5%), Rosaceae (7.3%) and Ranunculaceae (6.0%), respectively, followed by Fabaceae (4.2%), Lamiaceae (3.2%), Apiaceae (2.9%), Caryophyllaceae (2.5%), Campanulaceae (2.2%), Betulaceae (2.1%), Caprifoliaceae (2.0%) and Salicaceae (1.9%). It further includes 604 genera, with the significant ones being Artemisia (416), Acer (396), Carex (334), Viola (297), Polygonum (271), Salix (255), Potentilla (247), Betula (243), Clematis (229), Aster (218), Lespedeza (209), Aconitum (207), Vicia (198), Saussurea (195), Thalictrum (193), Ulmus (184), Angelica (183), Galium (181), Adenophora (178), Euonymus (172), Geranium (165), Prunus (165), Lilium (162), Cacalia (158), Quercus (154), Corylus (149), Lonicera (144), Lactuca (141), Veronica (139), Stellaria (135), Sedum (132), Tilia (132), Spiraea (131), Populus (130), Patrinia (125), Rosa (125), Athyrium (119), Chrysosplenium (119) and Iris (118).

Figure 6.  

Taxonomic distribution of occurrences amongst vascular plant genera and families in the dataset. The figure was prepared with the “treemap” package in R (Tennekes 2010).

This work is licensed under a Creative Commons Attribution (CC-BY) 4.0 License.