American Pern 186- Journal 97(4): 198 (2007) Northward and Range Invasion Expansion of the Fern Invasive Thelypteris dentata (Forssk.) John St. Urban into the Matrix of Three Prefectures Kinki in Japan District, Kentaro Murakami 1 Museum, Natural History Kishiwada City, Japan Matsui Rii; Pacific Consultants Co., Ltd., Japan MORIMOTO YUKIHIRO Graduates School of Global Environmental Studies, Kyoto University, Kyoto, Japan Abstract.— This study investigated the current distribution of an invasive tropical fern, Thelypteris and dentata, its habitat type in three Japanese prefectures (Osaka, Kyoto, and Shiga) in the Kinki The showed District. results that dentata has expanded T. distribution into highly urbanized its areas in Osaka Prefecture and has reached southern Kyoto and Prefecture central Shiga Prefecture. The distribution of T. dentata populations thus seems have expanded northward based to on comparisons with was the distribution determined that the 1980s. Because in the fern's habitat types were mainly the side walls or bottoms of drainage channels, crevices stone walls and in roadsides, the urban matrix has not served as a barrier to the expansion of the range of dentata: T. on may the contrary, be serving it as a type? of heat island corridor that facilitating the spread of is his species. I Words.— Kkv Thelypteris dentata, range expansion, urban heat island, global warming, greenhouse weed downy The maiden fern {Thelypteris dentata (Forssk.) John) St. is considered to be an alien fern that invaded Japan from tropical or subtropical (Yamazumi, countries Honshu 1988; Hotta, 2001). In the the area, first occurrence of dentata was documented Wakayama T. 1951 southern in in Prefecture of Kinki District (Yamazumi, 1988). Although the was fern initially treated as a rare species (Yamazumi, 1988), by the 1970s had been reported it in several additional places, mainly southern Honshu in coastal areas of such Wakayama southern and Mie as prefectures (Manago, The 1986). Japanese distribution of T. dentata was described along with and that of native ferns fern allies by Kurata and Nakaike According (1983). to their survey, the in Waka\ M used map to create this distribution dated mainly from the 1960s to the 1 Corresponding author. Current Museum address: Natural History Kishiwada of City, 6-5 Sakai- maohi, Kishiwada Osaka City, Prefecture, Japan. MURAKAMI ET RANGE EXPANSION OF THELYPTERIS DENTATA 187 AL.: were dentata beginning of the 1980s. Detailed reports of the distribution of T. map Wakayama by Manago Nakajima however, Prefecture (1986). (1998), for expanded dentata might have since predicted that the distribution of T. warm in We map we on focused investigate changes in the national distribution of T. dentata, its trends in 3 prefectures (Osaka, southern Kyoto, and Shiga prefectures) in the Kinki District of western Japan. These study areas were selected because the and growth urban heat island effect (which can facilitate the survival of markabl The term Kida, 2002). temperatures and by rural areas because of the greater retention of heat buildings, concrete, These and thus considered be important asphalt. cities their outskirts are to and which the relationship between increasing temperature the areas in expanding distribution of plant species can be investigated. Methods Materials and Field investigations were carried out by one or two investigators at each site The on 500-m around from 2004 October 2006. study focused a radius July to railway The main reason selecting railway stations as the center of stations. for each study was that has been reported that the primary habitats of the site it human-made such walls (Manago, 1986; Reis 2006) fern are habitats as et a/., common around and that these habitats are features railway stations. In total, we southern Kyoto surveyed 79 railway stations in central Shiga Prefecture, between Prefecture, and Osaka Prefecture; these included 11 stations Yamashina and Omi-Maiko on the Kosei Line of the Japanese Railways (J. between Nagahama and Osaka on Hokuriku Line Tokaido the or 35 stations R.), Line of the 18 stations on the Osaka Loop Line of the R., and 15 stations R., J. J. between Nanba and Misaki Koen on the Nankai Honsen Line of the Nankai Downy Railways. maiden fern was not found in the additional investigations conducted around Fukui and Tsuruga Stations (which are to the north of Shiga 2005 and August 2006. Prefecture) in July maximum In order to investigate each site at an equivalent intensity, the number was h The individuals each investigation time set at 2 per site. of at The was two (1-10 and 10-100 individuals). classified into categories site was microhabitat type for each fern classified into several categories: crevices between in stone walls, side walls or the bottom of drainage channels, alleys The by buildings, roadside habitats, or in planters. populations identified the autumn of 2005 were investigated in the following spring in order to determine The whether each individual had survived the winter. meteorological data for our study areas were collected from nearby meteorological stations by usin Agency Agency, the Japan Meteorological website (Japan Meteorological 2006). AMERICAN FERN VOLUME NUMBER JOURNAL: 188 97 4 (2007) In order to investigate the increase in temperature each the at site, temperatures were compared between the two study periods (1970-1982 and 1983-2005). Two-tailed were performed with t-tests for statistical analysis, < P significance set at 0.05. Results The shown meteorological data for our study areas are in Table All 1. temperatures were higher 1983 than most temperatures were after before; significantly higher, with the exception of those Torahime and Otsu at was (Table This particularly true the highly urbanized Osaka for cities of 1). and Kyoto, and Hikone for City in central Shiga Prefecture. downy Individuals of the maiden fern were discovered 34 of the 79 at sites The number (43.0%; Fig. Table of individuals in each population 1; 2). shown category recorded for each site is in Table 2; the total count was more than 250 The main newly individuals. habitats of the discovered populations were human-made hard-surfaced habitats such as side walls or the bottom of drainage channels The was or stone walls. frequency of occurrence higher in Osaka southern Prefecture than in the other areas. The northernmost was population discovered Hikone The Shiga in City, in central Prefecture. size of this population and of individual ferns were small, but the population was discovered in the winter of 2004, and had survived two years for at least The until the spring of 2006. distance of this population from the population in Wakayama Prefecture reported by Kurata and Nakaike and Manago (1983) was (1986) approximately 100 km, and the distance from southern Hyogo which was Prefecture, the northernmost location of the fern in Kinki District in map the distribution of Kurata and Nakaike was approximately 60 km. (1983), we assume If that the fern's distribution expanded northward for a total km distance of at least 60 during a period of about 20 years, the expansion km distance averages approximately per 3 year. Discussion — We new Distribution of Thelypteris dentata in Japan. recorded 34 populations of Thelypteris dentata in three prefectures of Japan's Kinki was District. Thelypteris dentata recorded previously and in local flora lists Museum, publications (Nakaike, 1996; Kohata, 1997; Hiratsuka City 2001; Hotta, 2001; Mitsuta, 2002; Matsui et aL, 2003; Murakami et 2003, 2004; a/., Kita-Kawachi Nature Club, and some were 2004), of these reports Kinki in Murakam Matsui (Mitsuta, 2002; aL, 2003; et from 2004 Osaka and Kyoto to in forests in Prefectures, including fragmented and forests, wildlife habitat, reclaimed forest in urban or suburban areas Murakami (Matsui et aL, 2003; et aL, 2003, 2004). In an investigation of the fern's distribution by the Kita-Kawachi Nature Club some populations (2004), of T. dentata were found from central to northern Osaka Prefecture. Mitsuta URAKAMIET AL.: R Table Temperat ure data eight meteorolo^ stations in and near the study area. (The data were obtained from the Japan Meteorological Agency at ;ical 1. ; comp website, 2006.) The locations of each station are shown in Figure The statistical significance was determined by aring the temperatures during ANGE 1. the two study perio ds (1970 to 1982 and 1983 to 2005). EX1 > A minimum maximum Average annual temperature C) verage temperature C) Average temperature C) 7 ( ( ( Significance Significance Significance ISION 1970-1982 1983-2005 1970-1982 1983-2005 of difference 1970-1982 1983-2005 of difference of difference i — -6.3 Torahime 13.2 13.9 n.s. 5.6 n.s. 33.7 34.8 n.s. (a) * * -4.8 -3.7 Hikone 34.3 35.1 14.0 14.7 (b) -3.6 -3.4 Otsu 14.8 14.9 n.s. n.s. 34.7 35.2 n.s. THEL (c) * * -4.2 -3.2 36.4 37.0 Kyoto 15.3 15.9 3 (d) Osaka 16.9 * * -2.5 -1.6 35.9 36.7 16.1 (e) — >TER1 Sakai 15.3 15.8 n.s. -3.9 3.2 n.s. 34.5 36.4 ** (f) S * -2.6 -2.1 Kumatori 14.9 15.6 n.s. 33.1 34.0 * (g) Wakayama * * -2.6 -1.7 35.3 35.7 16.0 16.7 n.s. (h) DENT p< p< Differences betweei periods were determined using a two-tail ed Mest: n.s not significant; *, 0.05; **, 0.01 A i ,, 2 CO CD 6 AMERICAN VOLUME NUMBER FERN JOURNAL: 190 07 4 (2007) 135° E 136 ,--' A 1 dapan Sea *»jy*"*^ f* ^ \r a_. f J H-<* -"-/"' Av' ' 1^.XX J < Lake P Biwa 00 o Q 1 Q \T~ Kyoto Pref. r o \ o o Kyoto urban area I U20 l8 %oo N G 22 21 - x Hyogo *12 Pref. l\ JQ^T* Osaka urban area n- J 30 y Osaka j Osaka Bay <^31^ < Pref, 33^32 bni Mie Pref. ^\ Awaji Island </ ^-T' "-* /-^/3 Nara Pref 34° N //" 40km Wakayama Pref. _j V ,^ Lie. Locations (the gray circles with numbers from to 34) of newly discovered populations of 1. 1 Tholypteris dontata in Osaka. Kyoto, and Shiga prefectures of Japan's Kinki District. The open circles represent the study sites in which T. dentata was nol discovered. Letters a through h represent the meteorological stations Table in 1. (2002) also recorded dentata and Kyoto T. in a list of ferns fern allies in Prefecture, but did not report the locations of collected specimens or their The was distribution. fern not reported in Shiga Prefecture before the present new investigation. Outside Kinki populations have been District, recently km reported from Funabashi which City (Nakaike, 500 1996), lies east of Museum, km Osaka; in Isehara City (Hiratsuka City 2001), which 350 east lies km Osaka; Anjoh which of in City (Hotta, 2001), lies 150 east of Osaka; and in Okayama km which City (Kohata, 1997), lies 140 west of Osaka. Because areas west our study were unknown east or of area not investigated, whether the is it fern's distribution continuous from eastern Shiga Prefecture Kanto is to Prefecture from western Osaka Chugoku or Prefecture southern to the part of Prefecture. Some new researchers noted that dentata species invades T. a that habitats is by escaping from greenhouses and other (Uemura, artificial habitats 2000; Yamazumi, unknown whether 1988), but dentata followed pattern is T. this it However, many we in Japan. likely that of the individuals observed it is MURAKAMI ET RANGE EXPANSION OF THELYPTERIS DENTATA AL.: 191 originated in this manner because T. dentata a greenhouse weed that has is been reported from around the world (Wagner and Smith, 1993; Possley, 2004], and has been observed domestic greenhouses accompanying in or potted it We (Yamazumi, plants in Japan 1988). also found T. dentata in a greenhouse in Kyoto City in January 2006. Therefore, the probability of the fern's dispersal via greenhouses in each region seems be high. to The current results do not clarify whether the Shiga and Kyoto populations Wakayama genetically resemble the Osaka and populations. Thelypteris dentata must have spread from greenhouses or potted plants purchased at each whether floriculture stores in area, but not clear northern individuals it is migrated from the southern Kinki District. However, the most important issue from the perspective of population genetics not the dispersal routes of this is km species, but rather the has dispersed approximately 100 from fact that the it Wakayama which was population, the primary location colonized by this The Moriyama species in the early 1980s (Fig. population identified in City, 2). in central Shiga Prefecture, included more than 20 full-grown individuals, thus will not be difficult for this population to survive in the future. it The distribution of tropical species can usually be explained by the minimum winter temperature by The mean or cold index. lowest a temperature Hikone which in City, contained the northernmost population in our survey, was -3.7 C from 1983 to 2005 (Table Although more detailed 1). study is required to understand the ecology of T. dentata, this value provides a good preliminary indication of the threshold temperature for growth and survival of dentata. T. we Because of the limited area covered by our research, the populations that discovered cannot be used to predict the northern limit of populations of this species Kinki More information must be acquired in District. in the future to confirm However, this limit. the climatic conditions in central Shiga may Prefecture be close to the northern distribution limit of the fern because the frequency of occurrence (25% survey remarkably lower of there sites) is than that in southern Osaka Prefecture (91.6% of survey sites). — Habitats of Thelypteris dentata. During our investigation, a large popula- was tion of Thelypteris dentata discovered in the urban areas of Kinki District. Some individuals were also identified in central Osaka City, which highly is The was urban. occurrence ratio lower in central Osaka City (37.5%) than in the northern and southern Osaka areas (50% and 91.6%, respectively). This may have resulted from the lack of crevices the edges of roads and the at fact may that fern not be found in highly developed areas. However, some were individuals nonetheless observed with expanses in locations large of broken pavements and walls with deep from where was in crevices, difficult it remove rhizome. the to fern's Most M (Manago, 1986; Hotta, 2001; side walls of drain channels, crevices in stone walls, roadsides, and alleys Some between were buildings. individuals discovered in pots or planters along may with potted plants such Aloe Cymbidium. These have as or ferns 2 CO number Table Locations of newly discovered populations of Thelypteris dentata and the associated of individuals in each habitat type in Osaka, 2. Kyoto and the Shiga prefectures of Kinki District, Japan. The open circles represent the T. dentata populations that survived from winter 2005 to spring 2006. Number dentata of individuals in each microhabitat type T. Number populations of sites at which dentata survived Sidewalls that T. were recorded/ from winter or bottom Inner Alleys Number between of sites 2005 to (Crevices of) of drainage planted > Name Location surveyed spring 2006 stone walls channels pots Roadside buildings re Shiga Prefecture 8 32 (25.0%) / > 1-10 Kawase N35 13' E136-13' 1 Aduchi N35 08' E136 08' 1-10 1-10 2 re Moriyama N35 03' E135 59' 10-100 3 Z N35 01' E135 57' 1-10 10-100 Kusatsu 4 o N34 59' E135 55' 10-100 5 Seta G N34 0-1 Lshiyama 58' E135°53' 00 6 1 7Z Nishi-Otsu N35 01' E135 51' 1-10 7 > - Otsu N35 00' E135 51' 1-10 re 8 1 1 < Keihanshin area (southern Kyoto Pref. and 6 12 (50.0%) C / re northern Osaka C Pref.) Yamashina N34 59' E135 49' 1-10 9 re N34 58' E135 43' 1-10 1-10 10 Nishi Oji CD Mukoh-machi N34 57' E135 42' 1-10 11 a Yamazaki N34 53' E135 40' 1-10 1 13 Senrioka N34 47' E135 33' 1-10 N34 46' E135 32' 1-10 re 14 Kishibe 4^ n: 3 •^ a Table Continued 2. > Number dentata T. of individuals in each microhabitat type > Number populations of sites at which - dentata that survived Sidewalls T. were recorded/ from winter bottom or Inner Alleys > Number of sites 2005 to (Crevices of) of drainage planted between Name Location surveyed spring 2006 stone walls channels Roadside pots buildings * > Osaka Osaka City (central Prefecture) 9 23 (39.1%) z / 15 HigashiyodogawaN34 40' E135 30' 1-10 O 16 Fukushima N34 41' E135 29' 10-100 1-10 X N34 17 Nocla 41' E135 28' 1-10 1-10 1-10 > 18 Taishn N34 39' E135 29' 10-100 1-10 1-10 z 19 Tsuruhashi N34 40' E135 31' 1-10 10-100 1-10 C/3 c Momodani N34 20 39' E135 31' 1-10 2 N34 21 Terada-cho 38' E135 31' 1-10 1-10 G 22 Tennohji N34 38' E135 30' 1-10 Suminoe N34 23 36' E135 29' 0-1 00 1 I Southern Osaka Prefecture 11 12 (91.6%) / 24 Minato N34 34' E135 27' 1-10 0-1 00 1 25 Hagoromo N34 32' E135 26' 10-100 10-100 26 Takaishi N34 31' El 35 25' 10-100 10-100 CO 27 [zumiotsu N34 30' E135 24' 10-100 10-100 fa 28 Tadaoka N34 29' E135 23' 10-100 29 Kishiwada N34 27' E135 22' 10-100 10-100 30 Kaiduka N34 26' E135 21' 1-10 2 31 Izumisano N34 24' E135 19' 1-10 1-10 N34 32 Tarui 22' E135 15' 1-10 10-100 33 Ozaki N34 21' E135 14' 1-10 1-10 34 Misaki-koen N34 19' E135 09' 1-10 Total 34 11 16 6 13 8 CO VOLUME NUMBER AMERICAN FERN JOURNAL: 104 97 4 (2007) Fie. 2. Change in the distribution of Thalyptcris dent at a from 1083 lo 2006. The open circles represent the distribution in 1083 (Kurata and Nakaike, 1083), and the gray circles represenl the -4 additional records of 2004-2006. The dashed line represents the C isothermal line of the minimum average temperature (1000-2003) based on the data from the Japan Meteorological Agency website. MURAKAMI RANGE EXPANSION OF THELYPTEHIS DENTATA KY 195 AI,: may purchased and originated as spores in the potting soil of plants, the ferns have survived due supply of water by humans. However, some ferns that to a may removed owners invaded planters have been the did not like the if and invaders enough retain them. Therefore, data from planters gardens to when should not be used rigorously evaluating habitat suitability for the downy we maiden have information Table Nonetheless, retained this in 2 fern. Even garden the sake of completeness. the absence of a fern in a planter or for if does not provide evidence of absence in a region, the presence of a fern does its provide evidence of presence in that region. its — Can expansion heat island corridors explain of the fern's distribution?. Mormon Northward expansion of the range of the great butterfly [Papilio km memnon occurred a rate of about 400 over 22 years, from 1981 to L.) at 2003 (Yoshio, 2004; Yoshio and 2004). The annual average expansion Ishii, thus equals about 18 km. In contrast, the results of the present study rate speed Thelypteris dentata indicate that the dispersal of the fern is km km approximately 60 to 100 over 20 years, for a rate of 3 to 5 per year. Although the fern has expanded distribution more slowly than the butterfly, its this likely because the butterfly can fly freely, whereas the fern sporophytes is downy maiden grows urban are sessile. Because the fern in artificial habitats rather than in a natural environment such as that of forest-floor ferns, the city may matrix not function as a barrier to northward expansion of the distribution of this species. Our study focused on areas along railway lines, and almost all were urban suburban these the matrix almost sites in or areas. In areas, city is continuous and there are few geomorphic obstacles between the study sites. If warm this spatial structure represents a continuous corridor from Shiga to Wakayama Wakayama, grow tropical or subtropical species that often in the may invade northern region be capable of dispersal along these corridors to fragmented surrounded by hard-surfaced areas In contrast, forests a (Fig. 2). from expanding urban matrix generally prevent species of the forest interior two even an urban heat island corridor connects patches of forest. easily, if more Urban weeds can expand their distribution rapidly because the city warming matrix functions as a source of habitat, not as a barrier. global If expand and continues, the distribution of such species will further the species common become wider range urban will across a of areas. — The Crinum Line and heat islands. In Japan, the Crinum Line and effect its known on phytogeography well (Koshimizu, 1938; Nakanishi, 1980). This is line represents a border formed by the line representing an average annual temperature approximately 15°C and an annual lowest temperature of of -3.5°C (Koshimizu, 1938), and regarded as the standard distribution limit is subtropical or tropical species (Nakanishi, 1980). This line used as for is still Crinum japonicum, the distribution borderline for not only asiaticum L. var. and name, the species that gives the line but also for various subtropical its hamabo tropical species, including Hibiscus Sieb. et Zucc, Canavalia lineata DC, (Thunb.) Chloranthus glaber (Thunb.) Makino, Debregeasia edulis many (Siebold & Zucc.) Wedd. and more (Taniguchi, 1956; Murata, 1968; The and Horikawa, 1976; Nakanishi, 1980). annual average temperature