Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Ecology of the invading moss species Orthodontium lineare in Sweden: Spatial distribution and population structure

Ecology of the invading moss species Orthodontium lineare in Sweden: Spatial distribution and... Hedenas, L., Herben. T., Rydin, H. and Soderstrom. L. 1989. Ecology of the invading moss species Orthodontium lineare in Sweden: Spatial distribution and population structure. - Holarct. Ecol. 12: 163-172. The moss species Orthodontium lineare, originally from the southern hemisphere, has been spreading in Europe during this century. We have monitored the distribution patterns of the species in Sweden. The age distribution differed between localities due to variation in the proportion of colonies in old age classes. We attributed this to differences in colony mortality. Although O. lineare had a high spore output in most localities, the colonies showed a clumped pattern, indicating strong neighbourhood effects in the colonization of new spots within the locality. The regional distribution was uneven and markedly southwestern. The distribution was related to regional and local availability of suitable habitats (decaying wood in not-too-dense forests), and to a lesser extent to climatic variables. Spore transport and establishment are suggested to be the limiting factors for colonization of suitable habitats of O. lineare. L. Hedenas, Dept of Botany, Univ. of Stockholm. S-106 91 Stockholm, Sweden. T. Herben, Botanical Inst. of the Czechoslovak Acad. of Sciences. CS-252 43 PrUhonice, Czechoslovakia. H. Rydin, hist, of Ecological Botany. Uppsala Univ., Box 559, S-751 22 Uppsala, Sweden. L. Soderstrom (reprint requests), Dept of Ecological Botany. Univ. of Umed, S-901 87 Umed. Sweden. Introduction Rapid changes with time are general features of many communities (e.g. Connell and Sousa 1983, Sousa 1984. Mooney and Drake 1986), and bryophytc communities are no exception (e.g. Kimmel 1962. During and ter Horst 1987. Herben 1987, Ericson 1977). Until recently the study of species invasions and population biology of invading species has been a greatly neglected field of ecological research, although it can contribute to the knowledge of species regulation in communities. In dynamic communities where dispersal abilities play an important role, the study of invading species, espeeially their population and reproductive biology, increases our understanding of the dispersal process between populations. Two recent examples of successful bryophyte invaders in Europe are Campylopus introflexus (e.g, Frahm 1972, Richards and Smith 197.^) and OrthoAcccpicd 1 October 1988 © HOLAKCTIC ECOLOGY dontium lineare (Ochyra 1982) (nomenclature follows Hallingback and Soderstrom 1987). The ability of competitively inferior species like O. lineare (Hedenas et al. in press) to invade and persist in a dynamic environment is primarily dependent on the relationship between local population size and persistence, dispersal rate and establishment rate (Bazzaz 1986), In this study of O. lineare in Sweden, we deal first with the age structure of populations and discuss this in relation to spore production, dispersal and patch availability. Secondly, we describe the local and regional distribution patterns in relation to climate and availability of suitable habitats. Orthodontium lineare Schwaegr.. a monoicous, acrocarpous moss species (Fig, I), was found only in the Southern Hemisphere until it was detected in England in 1911 (Meijer 1952), Since then it has been spreading in western Europe, and the invasion has been well documented (Oehyra 1982). In Sweden, O. lineare was dis- I- HOLARCTtC ECOLOGY 12:2(1989) addition, we searched for the species during journeys between some of the selected regions. To be able to visit a large number of localities over a large region, we estimated the amount of O. lineare and the number of available sites (at two levels) on rough ordinal scales. At the regional level the density of available forests (regional availability) was estimated using the following categories: 1) 0-1 per 10 km^, 2) 2-5 per 10 km- and 3) >5 per 10 km-. On the basis of earlier knowledge of the habitat preference of O. lineare (Hedenas 1981, Hedenas et al. in press), we considered localities as available when the following criteria were fulfilled: 1) not-too-dense conifer forest with almost no broad-leaved trees and 2) a low cover of herbs and shrubs, with the ground preferably covered by needle litter. From the localities considered available, some were selected for estimation of amount of available substrate at the local level. In areas with few available localities, all were selected, whereas in areas with more localities, four to six were sampled per 25 x 25 km area. Based on studies of the habitat preference of the species (Hedenas et al. in press), we consider "available" or "suitable" patches to be stumps and logs in decay stage 5 or more (sensu Soderstrom 19H7, i.e. Fig. 1. The pattern of growth in O. tineare showing the yearly increments (numbers denote age of branch) and the position of sexual organs from a plant collected in autumn. covered for the first time in the province of Skane in 1969 (Damsholt and Holmen 1971) and has since then spread successfully in southern and westerti Sweden. Today. O. lineare occupies an area in Sweden which is continuous with its distribution elsewhere in western Europe (Ochyra 1982, Hedenas and Soderstrom 1988). Orthodontium tineare grows mostly on temporary substrates (Hedenas et al. in press), limited in time either by decomposition processes or by microsuccession. It is often overgrown by other species on the same substrata but rarely overgrows them (Hedenas et al. in press. Herben 1987). Ecological studies of O. lineare in Europe have focussed on its phytosociologieal affinities and substrate preferences (Hedenas 1981, Herben 1987, von Hubsehmann 1970, Muhle 1970), whereas only Herben (1987) has dealt with aspects of tbe population structure and processes of the invasion. IT Ii •ov 8b vaster/ gotlondpM. -^ Smflland 7/ j X/ 7 1^ ° 7/7 l.Blekinge S L Skdne Methods Regional distribulion of O. lineare and of available localities Within the possible range of O. lineare (judged from earlier reports), we selected 25 x 25 km regions, mostly corresponding to standard Swedish topographic maps (Fig. 2). Within each of these regions we searched for the species on suitable sites along most of the roads. In 164 Fig. 2. Amount of O. tineare found ;u localities visited in southern Sweden. Numbers refer to locality: 1 = SkSne, Brunnby parish. Kullaberg; 2 = SkSne, Stcnestad parish. I km S of Stenestad church (2 plots); 3 = Skanc. Sirtivclstorp parish. Kulltorp (2 plots); 4 ^ Skanc, Tassjo parish, E of igarstorp; * .> = Halland, Voxtorp parish. Yllevadsmossan; 6 = Holland, Harplingc parish, Haverdal, S of Skintan; 7 = Vastergotaland. Grimmarcd parish, Viksliitt; S = Vastergotaland. Vilske-Klcva parish, Mosseberg, by Bergsjon. Filled square (•) = more than 10 small or more than one large spot. Filled circle (•) 4-10 small or one large spot. Plus( + ) = 1-3 small spots. Open circle (O) = absent. HOLARCTIC ECOLOGY 12;2 (WKt) wood soft with small crevices and small pieces lost), except those not lying directly on the ground as they dry out too quickly and too often (cf. Soderstrom 1988), ;ind also bare soil and areas just around tree bases. The density of substrate open to colonization (substrate availability) was estimated, based on the more detailed investigations made earlier, using the following categories: 1) almost no substrate available, 2) some substrate available, and 3) substrate freely available. At each investigated locality we devoted about one bryologist hour to search for O. lineare (3 persons for 20 min, 2 persons for 30 min or 1 person for 60 min). The amount of (). tineare was recorded as 0) absent, 1) only 1-3 small spots (< 1 dm'), 2) 4-10 small spots or one large spot (> 1 dm-), and 3) more than 10 small spots or more than one large spot. From the literature we collected the following data for each locality visited: altitude (from a topographical map), mean annual precipitation and permanence of snow cover (Wallen 1953). mean July temperature and difference between temperatures in January and July (Angstrom 1953a), and length of the growing period (mean day-time temperature ^3°C; Angstrom 1953b). Age structure of populations The population structure of O. lineare was analysed in 10 plots of 10 X 10 m at eight different localities (see Fig. 2 legend for locality names). The plots were selected so as to include at least the central portion of the area where O. tineare occurred. Orihodontium lineare occurs mostly in the form of colonies of about l-IO cm in diameter, and we sampled colonies for studies of their age structure using the following method. Colonies separated from other colonies by more than 10 cm were treated as separate samples. Extensive stands of O. lineare, consisting of several adjacent but still distinguishable colonies, were covered by a randomly placed 10 x 10 cm grid and the moss colonies most close to the intersection points of the grid were collected. When the separation between O. lineare colonies was not clear, samples were taken exactly at the intersection points not representing the whole coUmy. For all sampled colonies, the largest dimension was measured (where possible). Substrate of the colony and presence of protonema was also recorded. Turfs of O. lineare grow primarily at the shoot apices located at the surface. Because O. lineare produces sexual organs at the end of the growing season, successive generations of gametangia may serve as age markers. Branches are mostly produced at the beginning of the season (Fig. !) and can be used as an additional age marker. All samples were taken during autumn 1986, when the current year's growth was ahiiost finished and the leaves were green. If fertile, the branches were terminated by sexual organs. The branches of the previous year carried the last generation of capsules (which HOLARCTIC ECOLOGY 12:2 (1989) opened in the summer 1986). Their leaves were mostly green as well. Leaves of older branches were brownish, although still well-preserved. Visible decomposition of the leaves began on four-year old branches. This method yields accurate ageing of the plant for younger colonies. However, for older age categories this method will probably result in an underestimation of the real age because (1) in plants which are not always fertile, it may be difficult to distinguish between two one-year increments because of leaf decay and (2) in some years the yearly increments may be formed not at the apex but at the base exactly at the same position as in the preceding year. Because of these uncertainties the age categories over 4 years were lumped together. The oldest shoot from each colony was used as a measure of colony age. However, in some cases (especially on wood in unfavorable conditions). (). tineare grew with an extensive protonema with very small (< 1 cm'), sparse colonies of 1-3 yr old shoots. In such cases age determination of the colony was not possible, Presence of dead plants indicates that the age of the colony might be greater than the age of the oldest living shoot. Colonies with a targe proportion of protonema and only sparse shoots were excluded from the analyses, as was the whole age data set from the Vikslatt locality, where the majority of colonies were of this type. Capsule number and spore production The number of capsules in each colony was estimated visually in the field. For calibration of these estimates, 20 colonies of different size and capsule densities were collected, their capsule number being first estimated by each investigator and then counted (cf. Pielou 1981). Power curve regressions between estimated and counted numbers were used for calibration (coefficient of determination between estimate and count was greater than 0.9 for all investigators). For an estimation of spore content per capsule, eight just ripe, but unopened, capsules were collected separately. The capsules were opened and spores dissolved in a known amount of water. A fraction of this was put into a sedimentation chamber. After sedimentation the spores were counted in an inverted microscope and the total number was calculated. Pattern at the local scale At Kulltorp in NW SkSne we selected a 50 x 50 m area which was homogeneous with regard to tree layer and ground flora and where O. tineare occurred mainly on woody substrates. The coordinates for all patches of "suitable" (see above) stumps and logs larger than 5 cm in diameter within the square were recorded, together with the degree of decomposition, diameter, length/ height and area of the uncovered wood. In all these patches, total cover in cm' and number of capsules of O. tineare were estimated at all occupied spots. Nine colo- Fig. 3. Amount of available substrate for O. lineare at localities visited in southern Sweden. Filled square ( • ) = substrate freely available. Pius ( + ) = some substrate available. Open circle (O) = almost no substrate available. Hatched area (///) = region with > 5 localities per 10 km-. Dotted area ( : : ; : ) = region with 2-5 localities per 10 km'. Enclosed wiihin dotted line ( •:';• ) = region with ()-l lucality per 1(1 km'. Dotted line (•-.-•• ) = distance with 0-1 locality per tO km^ Tab. 1. Correlation matrix of physical variables and the distribution of O. lineare at a regional scale, OL = amount of O. lineare (ordinal scale). Altit. = altitude (m). Grow. per. = length of growing period (d). Precip. = mean annual precipitation (mm). Snow cover = permanency of snow cover (d). July temp, = mean July temperature ("C), Temp. diff. = difference between mean temperatures in January and July ("C), Substr. avail. = density of available substrate within a localily (ordinal scale). Reg. avail. = density of localities within region. OL Reg. avail. Substr. avail. Alt, Grow. per. Prec. Snow cover Temp. Reg. avail. Substr, avail. Altitude Grow. per. Precip, Snow cover July temp. Temp, diff. *) p<0.05 ,530* .371* -.103 .324* .230* -.323* -.093 -.274* ,307* .019 .163 ,375* -263* -.423* -.197 .008 -.007 .116 .055 -.096 ,134 -.773* -.234 ,525* -.571* .517* -.092 -.820* .506* -.663* .207 -.379* -.114 -.215 .625* -.056 HOLARCTIC ECOLOGY 12:2 (\'im) Tab. 2, Stepwisc regression for the significant predictors of the amount of O. lineare. F = 2 was used as limit for entering variables. Step No, 1 2 3 Variable entered Regional availability Substrate availability Growing period Signif, ,(K)0 .010 .058 r .286 .370 .413 r change ,286 .085 .042 Tab, 3. Cover of O. lineare in the neighbourhood of substrate differing in the presence of O. lineare (OL), - = patches without O. lineare, + = patches with O. lineare. Radius of neighbourhood (m) OL presence Mean cover in neighbourhood (cm') t-test nies of O. lineare were found outside these patehes. and for these colonies we reeorded the position, cover and number of capsules only. We tested whether or not the abundance of O. lineare in the vicinity of a "suitable" substrate pateh influenced the probability that this patch was colonized by O. lineare. The total cover of O. lineare within a given distance from each patch was calculated, and O. lineareeoloni7x'd patches were compared with non-colonized patches. This was done within circular neighbourhoods with radii of 1, 2. 3, 5, 8, 11 and 15 m, excluding spots for which parts of the circle fell outside the 50 x 50 m square. Separate tests were also made for areas around patches that were of different degrees of decomposition. In addition, considering only substrates occupied by O. lineare, we performed a linear regression of O. lineare cover on the substrate against the amount of O. lineare in the neighbourhood defined in the same way. At the same locality, a linear transect of 450 x 5 m, subdivided into 5 x 5 m squares, was studied. In each square, the total cover of O. lineare in cm- and area of available substrate (rotten wood, tree bases, bare soil) was recorded using the following scale: 0) no substrate, I) low substrate availability {0-10 dm-), 2) high substrate availability (> 10 dm-). For this part of the study we used linear neighbourhoods (of dimensions of 1, 2. 3, 5. 10, 15, and 20 squares, representing lengths of 5. 3.8 0.9 11.8 22.0 24.1 51.9 59.5 105,4 161.3 246.3 295.4 355,1 567.2 629.0 1,547 n.s 1.421 n.s 2.300 • 2.435 « 2.643 •* 1.420 n.s H.986 n.s *) 0.05 > p > 0.01; " ) 0.01 > p > 0,001 10.15.25,50, 75 and 100 m) instead of eireular ones (cf. Hill 1973). Separate analyses for the different levels of substrate availability in the squares were made. Results Spatial dislribulion on a regional scale The regional distribution of O. lineare is markedly southern and western in Sweden, with many rich localities in northern Skane and Halland (Fig, 2), With the exception of one report from Stockholm, it has not been reported from northern and eastern Sweden. The substrate availability, both regional and local, showed a clear decrease from the area where spruce forests are usually planted instead of broadlcaved trees (Skane. Halland) to the area with more natural mixed forests in the northeast (Smaland, Vastergtitland; Fig. 3). There were significant correlations between the ocTab, 4. Cover of O. lineare in neighbourhoods (of 5 m radius) of substrate patches with (-I-) and without (-) O. lineare. The area around slightly (stage 5-6) and highly (stage 7-8) decomposed patches was analyzed separately. Decay stage OL presence n Mean cover in neighbourhood (cm-) 62.5 76.1 49.1 127.6 t-test 5-6 7-8 Fig, 4. Spatial distribution of substrates and O. lineare in the 50 X 50 m plot at Kulltorp. HOLARCTIC ECOLOGY 12:2 (1989) 0.483 n.s. 3,005 ' • " ) 0.01 >p>.001 Tab. 5. Difference in average amount of O. lineare (cm^) in neighbourhood of squares with and without O. lineare (from transect data at Kulltorp). Neighbourhood distance (m) High substrate availability Difference t test 1.745 n.s. 3.970 •• 3.380 ** Age structure al selected localities Low substrate availability Difference 65.3 63.9 154.6 117.5 50.6 176.2 354,7 t-test 0.794 0.691 1.039 1.026 0.825 0.554 0.263 n.s. n.s. n.s. n.s. n.s. n.s. n.s. 3.258 •* 2.453 • 1.995 n.s. 2.372 * *)0.05>p>O.OI;'*)0.01>p>O.OOI According to age structure (Fig. 5), the localities can be divided into two groups. Five localities (Kulltorp 1, Kulltorp II. Stenestad I, Kuliaberg and Haverdal) had a high percentage of colonies (> 30%) in the oldest age class (>4 years), and the proportion of colonies in the younger age classes indicated that about 10-20% of colonies regenerated every year. Four localities (Stenestad II, Igarstorp. Mosseberg and Yllevadsmossen) had a low percentage (<20%) of colonies in the oldest age classes and high yearly regeneration. The differences in the age structure between localities from these two groups were generally significant (Tab. 6). There was no apparent difference in substrate availability between these two locality types (Hedenas et al. in press), In some localities the age structure of the colonies currence of O. lineare and length of growth period, varied depending on substrate (Fig. 6). On non-woody annual precipitation, duration of snow cover, temper- substrates, a large proportion of the colonies usually ature difference between July and January, and sub- belonged to the oldest age class and a low proportion of strate availability on both a regional and local scale the colonies regenerated over the last 4 years. This (Tab. 1). A stepwise regression analysis showed that pattern was most pronounced in the spruce forest localdensity of localities on a regional scale, substrate avail- ities. At Kulltorp 1 and II tbe opposite pattern was ability on a local scale and the length of growing period observed, with the greatest proportion of old colonies accounted for 41% of the variation of O. lineare occur- occurring on wood. rence. No other factor showed significant improvement The relation between colony age and colony size was of the model (Tab. 2). significant (p<0.001). but the predictability of colony size by its age was very low (Fig. 7, after excluding the oldest age category, r^ = 0.15), The slope of the age-size relationship did not differ between substrate types (ANPattern on the local scale COVA. p = 0.38). In the 50 x 50 m square at Kulltorp, we recorded 227 available substrate patches of logs and stumps; 53 of these were colonized by O. lineare {Fig. 4). In addition, we recorded O. lineare growing on tree bases {2 cases), Capsule and spore production by O. lineare on bare soil (1 case) and on rotten logs or branches of As an autoicous moss, O. lineare produces capsules less than 5 cm diam. (6 cases). freely. There was a significant relationship between capThe amount of O. lineare was significantly greater sule number per colony and colony size. Log-trans(t-test) in the vicinity of O. /meare-occupied substrates formed variables gave better fits than untransformed than near non-occupied substrates for the circular ones (logC = 1.03 x log S + 0.39. r-=0.52; C = neighbourhood with a radius of 3. 5 and 8 m, whereas no such relation could be found for the 1, 2, II or 15 m neighbourhoods (Tab. 3, Fig. 4). For the 5 m radius we compared the amount of O. lineare around occupied Kullabwg Kulltorp I Kulltorp tl Havwdal Stansstad I n-e5 Bxd-4 n-24 axd-D n-53 axd-0 n-ig e«d=J n=W BXCI-17 and unoccupied patches separately for neighbourhoods U [-1 I—, around patches in decay stage 5-6 and 7-8. There was significantly more O. lineare around occupied than n » around unoccupied patches of decay stage 7-8. whereas the difference around patches in decay stage 5-6 was 1 I 1 4 >t 1 I ] • *4 oStenastaa II Igarstorp Yllevadsmossen not significant (Tab. 4). .t n-ai BKCKZ7 n-I6 e*d=5 n-'2B oxcl-IO n-12 o)(d-4 -nil Regressions between size of O. lineare colony at a given substrate and amount of O. lineare in neighbourhoods of radius 3. 5 and 8 m were non-significant in all cases. Age of colony (years) In the 450 x 5 m transect, the amount of O. lineare in the neighbourhood differed between squares occupied Fig. 5. Age distribution of selected populations, n = total regeneratand not occupied by O. lineare, but only at high sub- number of colonies, excl. = numberof colonies with cannot be ing protonema and sparse shoots for which age strate availability (Tab. 5). determined. 168 HOLARCTIC ECOLOGY I2 Tab. 6. Chi-square tests of difference in age structure between localities. Line separates localities with a high and low percentage (>3()% and <2n7o, resp,) of the colonies in the oldest age category ( > 4 year). Site A Miissobcrg ( • I) Yllevadsmossen Igarstorp D Stenestad M 6.35 4.66 5.43 12.70* 7.00 10.71* 10.39* 12.29* 5.05 8.61 12.42* 10.7515.07** 17.22** 7.90 4.86 14.15" 7.73 11.14* 7.35 9.78* 10.69* E Haverdal F Kulltorp I (i Kulltorp II 11 Stenestad 1 I Kullaberg 12.05* 10,54 13,81** 2,96 4,08 5,40 8.45 l.M *)0.05>p>0.01; "}0,01>p number of capsules. S - size; Fig. 8). The relationship dill not differ between substriite types or between investigated squares (ANCOVA., p-t).64 for substrate types, and p = 0.53 for investigated squares). Number of spores per eapsule was estimated to be 45MOD (n = 8. SD - 8(KK)). This yielded maximum values of total yearly spore output up to 12 x 10" in a If) X 10 m plot (Stenestad II) and 3.2x10" in a 50 X 50 m plot (Kulltorp; Tab. 7). patches. Orthodontium lineare is known to spread rapidly in the field by means of protonetna (Herben 1987) and will then fill up all available space within reach of the protonema. However, bryophyte establishment consists of two steps, protonema development and gametophore formation. Age obtained by counting annual incretnents on shoots represents age since gametophore formation. We have no information on how this is related to protonema development. In general, a high proportion of the colonies were < 4 years old. This may be due to a still continuing popDiscussion ulation expansion or to a high mortality leading to exDynamics or populations tinction of older colonies. The distinction between these The persistence of O. lineare may be viewed as a dy- alternatives is erucial for interpreting the differences namic equilibrium between establishment and mortality between the two types of age structure observed. The of individual coUinies. Beeause in most of the localities following facts indicate the importance of mortality. (1) (1) the correlation between colony size and colony age In some populations there were direet indications of was very weak and (2) we did not find any eorrelation between Ihe eolony size and the amount of O. lineare in the neighbourhood, we conclude that the time since esIablishtTient was not limiting the size of individual colonies. Thus the size of the individual colonies should be detertnined by the size of the available substrate Kulllofp I f-i t>d-0 Kulltorp II n-ie iicl-O Kullaberg n-26 t>cl-0 Stenostad I n-20 Stenestad II o o o J ^ t i t J J ^ >4 Age of colony (years) Fig. 7, Relation between colony age and size. Bars indicate ± one standard deviation. Age of colony (y«ors) Fig. 6. Age distribution of subpopulations growing on different substrate types in selected populations, n = number of samples, excl. = number of colonies with regenerating protonema and sparse shoots for which age cannot be determined. HOIARCnC ECOLOGY 12: O10' wood non wood Colony size (cm^) Fig. 8. Capstile number per colony in relation to colony size (log scales). colonies and generally bad conditions for the growth of O. lineare. The proportion of these colonies varied between localities (Fig. 5). Thus we interpret the differences between populations with high and low proportion of old colonies as a result of differences in colony mortality, either between different localities or between substrate types. Localities with younger colonies are generally inland localities where the higher mortality may represent some type of population level response to climatic factors. In maritime pine forests, rotten wood is apparently more favourable for survival than in inland spruce forests (cf. Hedenas et al. in press). The same argument can be used for differences between substrate types. The only exceptions to this pattern, i.e. young populations without signs of any mortality (Yllevadsmossen and Igarstorp), probably represent truly recent colonizations. Patterns of occurrence on different scales There was a clear clumping in the occurrence of O. lineare. even in seemingly favourable habitats. With few exceptions, the clumping pattern has been observed on all scales (between regions, between localities and within localities). This indicates that the presence of O. lineare at a given spot is affected by the abundance of the species in the neighbourhood. The pattern on the local scale was investigated within areas of .*5Ox5O and 450x5 m. The area of lowest neighbourhood radii (1-2 m), which did not show any clumping, included only a few substrate patches available for (}. lineare (Fig. 4). For the largest radii (11 :intl 15 m) the clumping was nonsignificant. We doubt that this is a "critical distance effect" (i.e. distance beyond which the presence of O. lineare has no influence) because (I) the data sets for different distances were not independent. (2) lower sample sizes (due to wider border zone) reduced the power of t-tests, and (3) the 450 X 5 m transect demonstrated clumping at larger distances. high mortality. Matiy older colonies were being overgrown by blue-green algae, leprariaceous lichens or other mosses (Hedenas et al. in press). (2) The small differences in the proportion of old-age colonies between populations in localities with dominant old-age classes also indicate a balance between extinction and establishment of colonies. In cases of population expansion we would expect greater differences because of different times since establishment. We would also expect a relationship between the proportion of number of colonies in the age class and its age. In the localities studied (except Stenestad II and Igarstorp) the differences between younger age classes (1-4 year) within a locality, were insignificant, which indicates that gametophore formation did not differ very much between years. (3) The difference in age structure on different substrate types could hardly be explained as a result of on-going expansion. This proportion of younger colonies tends to be increased by excluding from the analysis, colonies with a large proportion of protonema, as these can be reasonably assumed to be older than the individual shoots within them. However, the presence of these colonies indicates mortality of plants within Tab. 7. Capsule and spore production at selected sites. Site Total Total cover of OL (cm-) Number of capsules Calculated total spore output (millions) 150 S30 Kullaberg Slcncstad I Stencstad II Igarstorp Yllevadsmossen Haverdal Vikslatt Mosseberg Kulltorp 1 Kulllorp II Kulllorp 450 X 5 Kulltorp 50 X 50 to o I2(K) 1469 3647 a) 7032 a) sum of values estimated using the power curve for individual colonies. 170 HOLARCTIC ECOLOGY On a regional scale, O. lineare shows a marked southwestern distribution, occurring in many of the available localities in northwestern Skane and southern Halland and only in a few localities in Vast ergot land. It is absent in the eastern and central parts of southern Sweden. As the prevailing winds in southern Sweden are southwesterly, one would expect O. lineare to also have spread to areas further north and east. The regional pattern is instead correlated with the density of available forests, whereas climatic factors seem to be of less importance. Because the main substrate for O. lineare, deeaying wood, is a temporal substrate and O. lineare is a poor competitor (Hedenas et al. In press), O. lineare relies on the continuous availability of new substrate patches. The proportion of patches occupied is then determined by the ratio between establishment rate and colony mortality rate. If the establishment rate is high compared with the extinction rate, the species will occur in most of the available localities, i.e. be a core species (cf. Hanski 1982). Colonization rate is regulated by production, transport and establishment of diaspores. Spore production did nt)t appear to be limiting for O. lineare as there were enormous numbers of spores produced on well established localities (up to 12 million per m'; Tab. 7). However, 0. lineare release the spores very close to the ground which make them less likely to be caught by the wind. Lane and eoworkers (quoted in Wyatt 1982) found a strongly leptocurtic relationship, with 97% of all spores from Airicham angusfaiuin trapped within 2 m from the colony and yet 1% was found beyond 15 m from the colony. Miles and Longton (1987) found that although most of the spores of Xm'c/jum undulatum were transported more tban 2 m, the spore deposition {per tmit of area) deereased rapidly and was very low at distances > 15 cm from the colony edge. Soderstrom and Jon.s.son (in press) found that 43% of the spores from Piilidium pulcherrimum, an epixyllc forest hepatic, were deposited witbin 2.5 m from the source. Furthermore, the germination of moss spores in the field is rarely observed (Longton and Miles 1982, Longton and Schuster 1983, Miles and Longton 1987), even if they germinate well under experimental conditions, as is the case for O. lineare (Herben unpubl. data). Dispersal, including both spore transport and establishment, is then probably restricting the colonization rate of O. lineare. The same process occurring at the local scale may operate at the regional seale (Tab. 3). Wbere the density of suitable loealities/patches is high, localities/patches where O. lineare has disappeared may be efficiently colonized from nearby localities/patches. At a regional scale O. lineare will occur as a core species. This is the case in northern SkSne and southern Halland. where much of the natural broad-leaved forests are cut down and replaced by coniferous forests, mostly of spruee. which are suitable for O. lineare. In Sm^land and Vastergotland, eoniferous forests are l i O l ARCTIC ECOLOGY 12:2 (1989) more natural and much of the broad-leaved forests remain unreplaeed. In these forests, the man-made disturbances are not so large and available localities are more scattered. Therefore, the chance for colonization is much less, and O. lineare will occur only on a few of the loealities, i.e. as a satellite speeies (ef. Hanski 1982). This will make these forests less subject to invasion by a foreign species (cf. Orians 1986). Within tbe occupied localities, it may, however, build up large local populations if substrate is abundant, as on Mosseberg in Vastergotland. Patterns like this have also been observed for epixylie speeies in northern Sweden (Soderstrom 1987). Acknowledgements - We ihank H. During. T. Ebenhard. L. Ericson and H. Hytleborn for comments on the manuscript. T. Hallingback and S. Franscn for help with localiiies. K. Lindahl for help with spore counting, A. Nordgren for drawing Fig. 4. and N. Rollison for correcting the English. Financial support was received from the Karin and Axel Btnning's Foundation. References Angstrom, A. 1953a. Temperatur, humiditet. - In: Svenska Sallskapet for Antropologi & Geografi (eds). Atlas over Sverige. Stockholm, sheet 25-26. (In Swedish and English). - 1953b. Maximi- och mininiitomperaturer, Srslider. vegetationsperioden, temp.-ktimatets forandringar. - In: Svenska Sallskapet for Antropologi & Geografi (eds). Atlas over Sverige. Stockholm, sheet 27-28. (In Swedish and English). Bazzaz. F. A. 1986. Life history of colonizing plants: some demographic, genetic, and physiological features. - In: Mooney, H, A. and Drake. J, A. (eds), Ecology of biological invasions of North America and Hawaii. Eeol. Stud. 58, pp. 9fr-ilO. Connell, J. H. and Sousa, W. P. 1983. On the evidence needed to judge ecological stability or persistence. - A m . Nat, 121: 789-824. Damsholt. K. and Holmen, K. 1971. Orlhodontiurn lineare Schwaegr. fundet I Sverige. - Lindbergia 1: 115.(In Danish with English summary). During. H. and ter Horst. B. 1987. Diversity and dynamics in bryophyte communities on earth banks in a Dutch forest. Symp. BioL Hung. 35: 447-155. Ericson. L. 1977, The influence of voles and lemmings on the vegetation in a coniferous forest during a 4-year period in northern Sweden. - Wahlenbergia 4: 1-114, Frahm, J.-P. 1972. Die Ausbreitung von Campylopus introflexus (Hedw.) Brid. in Mitteteuropa. - Herzogia 2: 317330. Hallingback, T. and Soderstrom. L. 1987. Sveriges mossor och deras svenska namn. En kommenterad checklista. - Svensk Bot. Tidskr. 81: 357-388. (In Swedish with English summary). Hanski, I. 1982. Dynamics of regional distribution: the core and satellite species hypothesis. - Oikos 38: 210-221. Hedenas, L. 1981. Orthodonfium lineare ~ en mos.sa pa frammarsch. - Svensk Bot, Tidskr. 75: 157-161. (In Swedish with English summary). - and Soderstrom. L. 1988. Kapmossan. Orihodantium lineare, iSverige.-SvenskBot.Tidskr. 81:217-220. (In Swedish with English summary). - , Herben, T., Rydin. H. and Soderstrom, L. Ecology of the invading moss species Orthodontium lineare in Sweden: substrate preference and interactions with other species. J. Bryol. (in press). Herben, T. 1987. The ecology of the invasion of Orthodontium lineare Schwaegr. in central Europe. - Symp, Biol. Hung. 35: 323-333. Hill. M. O. 1973. The intensity of spatial pattern in plant communities. - J. Ecol. 61: 225-235. Hiibschmann, A. von. 1970. Ueber die verbreitung einiger seltener Laubmoose in Nordwesldeutschen Pflanzengesellschaften. - Herzogia 2: 63-75. Kimmel, U. 1962. Entwicklung einiger Moose und Flechten auf Dauer-Untersuchungsflachen. -Oberh. Ges. Natur- u. Heilkunde Ber. N.F. Nalurw. Abt. 32: 131-160. Longion. R. E. and Miles. C. J. 1982. Studies on the reproductive biology of mosses. - J. Hattori Bot. Lab. 52: 219-240. - and Schuster, R. M. 1983. Reproductive biology. - In: Schuster. R. M. (ed.). New manual of bryology. Hattori Bot. Lab,. Nichinan, pp. 386-462. Mcijer, W. 1952. The genus Orthodonlium. - Acta Bot. Necrl. 1: 1-80. Miles. C. J. and Longton. R. E. 1987. Reproductive biology of Atrichum undulatiim. - Symp. Biol. Hung. 35: 193-207. Mooncy, H. A. and Drake, J. A. (eds) 1986. Ecology of biological invasions of North America and Hawaii. - Ecol. Stud. 5S. Springer. Muhle, H. 1970. Zur Ausbreitung vom Orlhodondum lineare Schwaegr.: Orthodontium in Schwiirzwald. - Herzogia 2: 107-112. Ochyra, R. 1982. Orthodonlium lineare Schwaegr. - a new Species and genus in the moss flora of Poland. - Bryologische Beitrage 1: 23-36. Orians, G. H. 1986. Site characteristics favouring invasions. In: Mooney, H. A. and Drake. J. A. (eds). Ecology of biological invasions of North America and Hawaii. Ecol. Stud. 5S. pp. 133-148. Pielou, E. C. 1981. Rapid estimation of standing crop of intertidai fucoids on an exposed shore. - J . Environ. Manage. 13: 85-98. Richards, P. W. and Smith, A. J. E. 1975. A progress report on Campytopus inlroflexus iWuiis.) Brid- and C- polvtrichoides De Not. in Britain and Ireland. - J. Bryol. 8: 293-298. Soderstrom. L. 1987. Dispersal as a limiting factor for distribution among epixylic bryophytes. - Symp. Biol. Hung. 35: 475-484. - 1988, Cryptogam species sequence in relation to substrate variables of decaying coniferous wood in northern Sweden. - Nord. J. Bot. 8: 89-97. - and Jonsson, B. G, Spatial pattern and dispersal in the leafy hepatic Ptitidium pukherrimum. - J. Bryol. (in press). Sousa, W. P. 1984, The role of disturbance in natural communities. - Ann. Rev. Ecol. Syst, 15: 353-391, Wall^n, C. C. 1953. Sno, hagel, aska. Arsncderbord. arets medeltemperatur. - ln: Svenska Siillskapct for Antropologi & Geografi (eds). Attas over Sverige. Stockholm, sheet 31-32, (In Swedish and English). Wyatt. R, 1982. Populationecoiogy of Bryophytes.-J. Hattori Bot. Lab. 52: 179-198, HOLARCTIC ECOLOGY 12:2 (1989) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ecography Wiley

Ecology of the invading moss species Orthodontium lineare in Sweden: Spatial distribution and population structure

Download PDF
11 pages

Loading next page...
 
/lp/wiley/ecology-of-the-invading-moss-species-orthodontium-lineare-in-sweden-gAVad5oS50

References (30)

Publisher
Wiley
Copyright
Copyright © 1989 Wiley Subscription Services, Inc., A Wiley Company
ISSN
0906-7590
eISSN
1600-0587
DOI
10.1111/j.1600-0587.1989.tb00835.x
Publisher site
See Article on Publisher Site

Abstract

Hedenas, L., Herben. T., Rydin, H. and Soderstrom. L. 1989. Ecology of the invading moss species Orthodontium lineare in Sweden: Spatial distribution and population structure. - Holarct. Ecol. 12: 163-172. The moss species Orthodontium lineare, originally from the southern hemisphere, has been spreading in Europe during this century. We have monitored the distribution patterns of the species in Sweden. The age distribution differed between localities due to variation in the proportion of colonies in old age classes. We attributed this to differences in colony mortality. Although O. lineare had a high spore output in most localities, the colonies showed a clumped pattern, indicating strong neighbourhood effects in the colonization of new spots within the locality. The regional distribution was uneven and markedly southwestern. The distribution was related to regional and local availability of suitable habitats (decaying wood in not-too-dense forests), and to a lesser extent to climatic variables. Spore transport and establishment are suggested to be the limiting factors for colonization of suitable habitats of O. lineare. L. Hedenas, Dept of Botany, Univ. of Stockholm. S-106 91 Stockholm, Sweden. T. Herben, Botanical Inst. of the Czechoslovak Acad. of Sciences. CS-252 43 PrUhonice, Czechoslovakia. H. Rydin, hist, of Ecological Botany. Uppsala Univ., Box 559, S-751 22 Uppsala, Sweden. L. Soderstrom (reprint requests), Dept of Ecological Botany. Univ. of Umed, S-901 87 Umed. Sweden. Introduction Rapid changes with time are general features of many communities (e.g. Connell and Sousa 1983, Sousa 1984. Mooney and Drake 1986), and bryophytc communities are no exception (e.g. Kimmel 1962. During and ter Horst 1987. Herben 1987, Ericson 1977). Until recently the study of species invasions and population biology of invading species has been a greatly neglected field of ecological research, although it can contribute to the knowledge of species regulation in communities. In dynamic communities where dispersal abilities play an important role, the study of invading species, espeeially their population and reproductive biology, increases our understanding of the dispersal process between populations. Two recent examples of successful bryophyte invaders in Europe are Campylopus introflexus (e.g, Frahm 1972, Richards and Smith 197.^) and OrthoAcccpicd 1 October 1988 © HOLAKCTIC ECOLOGY dontium lineare (Ochyra 1982) (nomenclature follows Hallingback and Soderstrom 1987). The ability of competitively inferior species like O. lineare (Hedenas et al. in press) to invade and persist in a dynamic environment is primarily dependent on the relationship between local population size and persistence, dispersal rate and establishment rate (Bazzaz 1986), In this study of O. lineare in Sweden, we deal first with the age structure of populations and discuss this in relation to spore production, dispersal and patch availability. Secondly, we describe the local and regional distribution patterns in relation to climate and availability of suitable habitats. Orthodontium lineare Schwaegr.. a monoicous, acrocarpous moss species (Fig, I), was found only in the Southern Hemisphere until it was detected in England in 1911 (Meijer 1952), Since then it has been spreading in western Europe, and the invasion has been well documented (Oehyra 1982). In Sweden, O. lineare was dis- I- HOLARCTtC ECOLOGY 12:2(1989) addition, we searched for the species during journeys between some of the selected regions. To be able to visit a large number of localities over a large region, we estimated the amount of O. lineare and the number of available sites (at two levels) on rough ordinal scales. At the regional level the density of available forests (regional availability) was estimated using the following categories: 1) 0-1 per 10 km^, 2) 2-5 per 10 km- and 3) >5 per 10 km-. On the basis of earlier knowledge of the habitat preference of O. lineare (Hedenas 1981, Hedenas et al. in press), we considered localities as available when the following criteria were fulfilled: 1) not-too-dense conifer forest with almost no broad-leaved trees and 2) a low cover of herbs and shrubs, with the ground preferably covered by needle litter. From the localities considered available, some were selected for estimation of amount of available substrate at the local level. In areas with few available localities, all were selected, whereas in areas with more localities, four to six were sampled per 25 x 25 km area. Based on studies of the habitat preference of the species (Hedenas et al. in press), we consider "available" or "suitable" patches to be stumps and logs in decay stage 5 or more (sensu Soderstrom 19H7, i.e. Fig. 1. The pattern of growth in O. tineare showing the yearly increments (numbers denote age of branch) and the position of sexual organs from a plant collected in autumn. covered for the first time in the province of Skane in 1969 (Damsholt and Holmen 1971) and has since then spread successfully in southern and westerti Sweden. Today. O. lineare occupies an area in Sweden which is continuous with its distribution elsewhere in western Europe (Ochyra 1982, Hedenas and Soderstrom 1988). Orthodontium tineare grows mostly on temporary substrates (Hedenas et al. in press), limited in time either by decomposition processes or by microsuccession. It is often overgrown by other species on the same substrata but rarely overgrows them (Hedenas et al. in press. Herben 1987). Ecological studies of O. lineare in Europe have focussed on its phytosociologieal affinities and substrate preferences (Hedenas 1981, Herben 1987, von Hubsehmann 1970, Muhle 1970), whereas only Herben (1987) has dealt with aspects of tbe population structure and processes of the invasion. IT Ii •ov 8b vaster/ gotlondpM. -^ Smflland 7/ j X/ 7 1^ ° 7/7 l.Blekinge S L Skdne Methods Regional distribulion of O. lineare and of available localities Within the possible range of O. lineare (judged from earlier reports), we selected 25 x 25 km regions, mostly corresponding to standard Swedish topographic maps (Fig. 2). Within each of these regions we searched for the species on suitable sites along most of the roads. In 164 Fig. 2. Amount of O. tineare found ;u localities visited in southern Sweden. Numbers refer to locality: 1 = SkSne, Brunnby parish. Kullaberg; 2 = SkSne, Stcnestad parish. I km S of Stenestad church (2 plots); 3 = Skanc. Sirtivclstorp parish. Kulltorp (2 plots); 4 ^ Skanc, Tassjo parish, E of igarstorp; * .> = Halland, Voxtorp parish. Yllevadsmossan; 6 = Holland, Harplingc parish, Haverdal, S of Skintan; 7 = Vastergotaland. Grimmarcd parish, Viksliitt; S = Vastergotaland. Vilske-Klcva parish, Mosseberg, by Bergsjon. Filled square (•) = more than 10 small or more than one large spot. Filled circle (•) 4-10 small or one large spot. Plus( + ) = 1-3 small spots. Open circle (O) = absent. HOLARCTIC ECOLOGY 12;2 (WKt) wood soft with small crevices and small pieces lost), except those not lying directly on the ground as they dry out too quickly and too often (cf. Soderstrom 1988), ;ind also bare soil and areas just around tree bases. The density of substrate open to colonization (substrate availability) was estimated, based on the more detailed investigations made earlier, using the following categories: 1) almost no substrate available, 2) some substrate available, and 3) substrate freely available. At each investigated locality we devoted about one bryologist hour to search for O. lineare (3 persons for 20 min, 2 persons for 30 min or 1 person for 60 min). The amount of (). tineare was recorded as 0) absent, 1) only 1-3 small spots (< 1 dm'), 2) 4-10 small spots or one large spot (> 1 dm-), and 3) more than 10 small spots or more than one large spot. From the literature we collected the following data for each locality visited: altitude (from a topographical map), mean annual precipitation and permanence of snow cover (Wallen 1953). mean July temperature and difference between temperatures in January and July (Angstrom 1953a), and length of the growing period (mean day-time temperature ^3°C; Angstrom 1953b). Age structure of populations The population structure of O. lineare was analysed in 10 plots of 10 X 10 m at eight different localities (see Fig. 2 legend for locality names). The plots were selected so as to include at least the central portion of the area where O. tineare occurred. Orihodontium lineare occurs mostly in the form of colonies of about l-IO cm in diameter, and we sampled colonies for studies of their age structure using the following method. Colonies separated from other colonies by more than 10 cm were treated as separate samples. Extensive stands of O. lineare, consisting of several adjacent but still distinguishable colonies, were covered by a randomly placed 10 x 10 cm grid and the moss colonies most close to the intersection points of the grid were collected. When the separation between O. lineare colonies was not clear, samples were taken exactly at the intersection points not representing the whole coUmy. For all sampled colonies, the largest dimension was measured (where possible). Substrate of the colony and presence of protonema was also recorded. Turfs of O. lineare grow primarily at the shoot apices located at the surface. Because O. lineare produces sexual organs at the end of the growing season, successive generations of gametangia may serve as age markers. Branches are mostly produced at the beginning of the season (Fig. !) and can be used as an additional age marker. All samples were taken during autumn 1986, when the current year's growth was ahiiost finished and the leaves were green. If fertile, the branches were terminated by sexual organs. The branches of the previous year carried the last generation of capsules (which HOLARCTIC ECOLOGY 12:2 (1989) opened in the summer 1986). Their leaves were mostly green as well. Leaves of older branches were brownish, although still well-preserved. Visible decomposition of the leaves began on four-year old branches. This method yields accurate ageing of the plant for younger colonies. However, for older age categories this method will probably result in an underestimation of the real age because (1) in plants which are not always fertile, it may be difficult to distinguish between two one-year increments because of leaf decay and (2) in some years the yearly increments may be formed not at the apex but at the base exactly at the same position as in the preceding year. Because of these uncertainties the age categories over 4 years were lumped together. The oldest shoot from each colony was used as a measure of colony age. However, in some cases (especially on wood in unfavorable conditions). (). tineare grew with an extensive protonema with very small (< 1 cm'), sparse colonies of 1-3 yr old shoots. In such cases age determination of the colony was not possible, Presence of dead plants indicates that the age of the colony might be greater than the age of the oldest living shoot. Colonies with a targe proportion of protonema and only sparse shoots were excluded from the analyses, as was the whole age data set from the Vikslatt locality, where the majority of colonies were of this type. Capsule number and spore production The number of capsules in each colony was estimated visually in the field. For calibration of these estimates, 20 colonies of different size and capsule densities were collected, their capsule number being first estimated by each investigator and then counted (cf. Pielou 1981). Power curve regressions between estimated and counted numbers were used for calibration (coefficient of determination between estimate and count was greater than 0.9 for all investigators). For an estimation of spore content per capsule, eight just ripe, but unopened, capsules were collected separately. The capsules were opened and spores dissolved in a known amount of water. A fraction of this was put into a sedimentation chamber. After sedimentation the spores were counted in an inverted microscope and the total number was calculated. Pattern at the local scale At Kulltorp in NW SkSne we selected a 50 x 50 m area which was homogeneous with regard to tree layer and ground flora and where O. tineare occurred mainly on woody substrates. The coordinates for all patches of "suitable" (see above) stumps and logs larger than 5 cm in diameter within the square were recorded, together with the degree of decomposition, diameter, length/ height and area of the uncovered wood. In all these patches, total cover in cm' and number of capsules of O. tineare were estimated at all occupied spots. Nine colo- Fig. 3. Amount of available substrate for O. lineare at localities visited in southern Sweden. Filled square ( • ) = substrate freely available. Pius ( + ) = some substrate available. Open circle (O) = almost no substrate available. Hatched area (///) = region with > 5 localities per 10 km-. Dotted area ( : : ; : ) = region with 2-5 localities per 10 km'. Enclosed wiihin dotted line ( •:';• ) = region with ()-l lucality per 1(1 km'. Dotted line (•-.-•• ) = distance with 0-1 locality per tO km^ Tab. 1. Correlation matrix of physical variables and the distribution of O. lineare at a regional scale, OL = amount of O. lineare (ordinal scale). Altit. = altitude (m). Grow. per. = length of growing period (d). Precip. = mean annual precipitation (mm). Snow cover = permanency of snow cover (d). July temp, = mean July temperature ("C), Temp. diff. = difference between mean temperatures in January and July ("C), Substr. avail. = density of available substrate within a localily (ordinal scale). Reg. avail. = density of localities within region. OL Reg. avail. Substr. avail. Alt, Grow. per. Prec. Snow cover Temp. Reg. avail. Substr, avail. Altitude Grow. per. Precip, Snow cover July temp. Temp, diff. *) p<0.05 ,530* .371* -.103 .324* .230* -.323* -.093 -.274* ,307* .019 .163 ,375* -263* -.423* -.197 .008 -.007 .116 .055 -.096 ,134 -.773* -.234 ,525* -.571* .517* -.092 -.820* .506* -.663* .207 -.379* -.114 -.215 .625* -.056 HOLARCTIC ECOLOGY 12:2 (\'im) Tab. 2, Stepwisc regression for the significant predictors of the amount of O. lineare. F = 2 was used as limit for entering variables. Step No, 1 2 3 Variable entered Regional availability Substrate availability Growing period Signif, ,(K)0 .010 .058 r .286 .370 .413 r change ,286 .085 .042 Tab, 3. Cover of O. lineare in the neighbourhood of substrate differing in the presence of O. lineare (OL), - = patches without O. lineare, + = patches with O. lineare. Radius of neighbourhood (m) OL presence Mean cover in neighbourhood (cm') t-test nies of O. lineare were found outside these patehes. and for these colonies we reeorded the position, cover and number of capsules only. We tested whether or not the abundance of O. lineare in the vicinity of a "suitable" substrate pateh influenced the probability that this patch was colonized by O. lineare. The total cover of O. lineare within a given distance from each patch was calculated, and O. lineareeoloni7x'd patches were compared with non-colonized patches. This was done within circular neighbourhoods with radii of 1, 2. 3, 5, 8, 11 and 15 m, excluding spots for which parts of the circle fell outside the 50 x 50 m square. Separate tests were also made for areas around patches that were of different degrees of decomposition. In addition, considering only substrates occupied by O. lineare, we performed a linear regression of O. lineare cover on the substrate against the amount of O. lineare in the neighbourhood defined in the same way. At the same locality, a linear transect of 450 x 5 m, subdivided into 5 x 5 m squares, was studied. In each square, the total cover of O. lineare in cm- and area of available substrate (rotten wood, tree bases, bare soil) was recorded using the following scale: 0) no substrate, I) low substrate availability {0-10 dm-), 2) high substrate availability (> 10 dm-). For this part of the study we used linear neighbourhoods (of dimensions of 1, 2. 3, 5. 10, 15, and 20 squares, representing lengths of 5. 3.8 0.9 11.8 22.0 24.1 51.9 59.5 105,4 161.3 246.3 295.4 355,1 567.2 629.0 1,547 n.s 1.421 n.s 2.300 • 2.435 « 2.643 •* 1.420 n.s H.986 n.s *) 0.05 > p > 0.01; " ) 0.01 > p > 0,001 10.15.25,50, 75 and 100 m) instead of eireular ones (cf. Hill 1973). Separate analyses for the different levels of substrate availability in the squares were made. Results Spatial dislribulion on a regional scale The regional distribution of O. lineare is markedly southern and western in Sweden, with many rich localities in northern Skane and Halland (Fig, 2), With the exception of one report from Stockholm, it has not been reported from northern and eastern Sweden. The substrate availability, both regional and local, showed a clear decrease from the area where spruce forests are usually planted instead of broadlcaved trees (Skane. Halland) to the area with more natural mixed forests in the northeast (Smaland, Vastergtitland; Fig. 3). There were significant correlations between the ocTab, 4. Cover of O. lineare in neighbourhoods (of 5 m radius) of substrate patches with (-I-) and without (-) O. lineare. The area around slightly (stage 5-6) and highly (stage 7-8) decomposed patches was analyzed separately. Decay stage OL presence n Mean cover in neighbourhood (cm-) 62.5 76.1 49.1 127.6 t-test 5-6 7-8 Fig, 4. Spatial distribution of substrates and O. lineare in the 50 X 50 m plot at Kulltorp. HOLARCTIC ECOLOGY 12:2 (1989) 0.483 n.s. 3,005 ' • " ) 0.01 >p>.001 Tab. 5. Difference in average amount of O. lineare (cm^) in neighbourhood of squares with and without O. lineare (from transect data at Kulltorp). Neighbourhood distance (m) High substrate availability Difference t test 1.745 n.s. 3.970 •• 3.380 ** Age structure al selected localities Low substrate availability Difference 65.3 63.9 154.6 117.5 50.6 176.2 354,7 t-test 0.794 0.691 1.039 1.026 0.825 0.554 0.263 n.s. n.s. n.s. n.s. n.s. n.s. n.s. 3.258 •* 2.453 • 1.995 n.s. 2.372 * *)0.05>p>O.OI;'*)0.01>p>O.OOI According to age structure (Fig. 5), the localities can be divided into two groups. Five localities (Kulltorp 1, Kulltorp II. Stenestad I, Kuliaberg and Haverdal) had a high percentage of colonies (> 30%) in the oldest age class (>4 years), and the proportion of colonies in the younger age classes indicated that about 10-20% of colonies regenerated every year. Four localities (Stenestad II, Igarstorp. Mosseberg and Yllevadsmossen) had a low percentage (<20%) of colonies in the oldest age classes and high yearly regeneration. The differences in the age structure between localities from these two groups were generally significant (Tab. 6). There was no apparent difference in substrate availability between these two locality types (Hedenas et al. in press), In some localities the age structure of the colonies currence of O. lineare and length of growth period, varied depending on substrate (Fig. 6). On non-woody annual precipitation, duration of snow cover, temper- substrates, a large proportion of the colonies usually ature difference between July and January, and sub- belonged to the oldest age class and a low proportion of strate availability on both a regional and local scale the colonies regenerated over the last 4 years. This (Tab. 1). A stepwise regression analysis showed that pattern was most pronounced in the spruce forest localdensity of localities on a regional scale, substrate avail- ities. At Kulltorp 1 and II tbe opposite pattern was ability on a local scale and the length of growing period observed, with the greatest proportion of old colonies accounted for 41% of the variation of O. lineare occur- occurring on wood. rence. No other factor showed significant improvement The relation between colony age and colony size was of the model (Tab. 2). significant (p<0.001). but the predictability of colony size by its age was very low (Fig. 7, after excluding the oldest age category, r^ = 0.15), The slope of the age-size relationship did not differ between substrate types (ANPattern on the local scale COVA. p = 0.38). In the 50 x 50 m square at Kulltorp, we recorded 227 available substrate patches of logs and stumps; 53 of these were colonized by O. lineare {Fig. 4). In addition, we recorded O. lineare growing on tree bases {2 cases), Capsule and spore production by O. lineare on bare soil (1 case) and on rotten logs or branches of As an autoicous moss, O. lineare produces capsules less than 5 cm diam. (6 cases). freely. There was a significant relationship between capThe amount of O. lineare was significantly greater sule number per colony and colony size. Log-trans(t-test) in the vicinity of O. /meare-occupied substrates formed variables gave better fits than untransformed than near non-occupied substrates for the circular ones (logC = 1.03 x log S + 0.39. r-=0.52; C = neighbourhood with a radius of 3. 5 and 8 m, whereas no such relation could be found for the 1, 2, II or 15 m neighbourhoods (Tab. 3, Fig. 4). For the 5 m radius we compared the amount of O. lineare around occupied Kullabwg Kulltorp I Kulltorp tl Havwdal Stansstad I n-e5 Bxd-4 n-24 axd-D n-53 axd-0 n-ig e«d=J n=W BXCI-17 and unoccupied patches separately for neighbourhoods U [-1 I—, around patches in decay stage 5-6 and 7-8. There was significantly more O. lineare around occupied than n » around unoccupied patches of decay stage 7-8. whereas the difference around patches in decay stage 5-6 was 1 I 1 4 >t 1 I ] • *4 oStenastaa II Igarstorp Yllevadsmossen not significant (Tab. 4). .t n-ai BKCKZ7 n-I6 e*d=5 n-'2B oxcl-IO n-12 o)(d-4 -nil Regressions between size of O. lineare colony at a given substrate and amount of O. lineare in neighbourhoods of radius 3. 5 and 8 m were non-significant in all cases. Age of colony (years) In the 450 x 5 m transect, the amount of O. lineare in the neighbourhood differed between squares occupied Fig. 5. Age distribution of selected populations, n = total regeneratand not occupied by O. lineare, but only at high sub- number of colonies, excl. = numberof colonies with cannot be ing protonema and sparse shoots for which age strate availability (Tab. 5). determined. 168 HOLARCTIC ECOLOGY I2 Tab. 6. Chi-square tests of difference in age structure between localities. Line separates localities with a high and low percentage (>3()% and <2n7o, resp,) of the colonies in the oldest age category ( > 4 year). Site A Miissobcrg ( • I) Yllevadsmossen Igarstorp D Stenestad M 6.35 4.66 5.43 12.70* 7.00 10.71* 10.39* 12.29* 5.05 8.61 12.42* 10.7515.07** 17.22** 7.90 4.86 14.15" 7.73 11.14* 7.35 9.78* 10.69* E Haverdal F Kulltorp I (i Kulltorp II 11 Stenestad 1 I Kullaberg 12.05* 10,54 13,81** 2,96 4,08 5,40 8.45 l.M *)0.05>p>0.01; "}0,01>p number of capsules. S - size; Fig. 8). The relationship dill not differ between substriite types or between investigated squares (ANCOVA., p-t).64 for substrate types, and p = 0.53 for investigated squares). Number of spores per eapsule was estimated to be 45MOD (n = 8. SD - 8(KK)). This yielded maximum values of total yearly spore output up to 12 x 10" in a If) X 10 m plot (Stenestad II) and 3.2x10" in a 50 X 50 m plot (Kulltorp; Tab. 7). patches. Orthodontium lineare is known to spread rapidly in the field by means of protonetna (Herben 1987) and will then fill up all available space within reach of the protonema. However, bryophyte establishment consists of two steps, protonema development and gametophore formation. Age obtained by counting annual incretnents on shoots represents age since gametophore formation. We have no information on how this is related to protonema development. In general, a high proportion of the colonies were < 4 years old. This may be due to a still continuing popDiscussion ulation expansion or to a high mortality leading to exDynamics or populations tinction of older colonies. The distinction between these The persistence of O. lineare may be viewed as a dy- alternatives is erucial for interpreting the differences namic equilibrium between establishment and mortality between the two types of age structure observed. The of individual coUinies. Beeause in most of the localities following facts indicate the importance of mortality. (1) (1) the correlation between colony size and colony age In some populations there were direet indications of was very weak and (2) we did not find any eorrelation between Ihe eolony size and the amount of O. lineare in the neighbourhood, we conclude that the time since esIablishtTient was not limiting the size of individual colonies. Thus the size of the individual colonies should be detertnined by the size of the available substrate Kulllofp I f-i t>d-0 Kulltorp II n-ie iicl-O Kullaberg n-26 t>cl-0 Stenostad I n-20 Stenestad II o o o J ^ t i t J J ^ >4 Age of colony (years) Fig. 7, Relation between colony age and size. Bars indicate ± one standard deviation. Age of colony (y«ors) Fig. 6. Age distribution of subpopulations growing on different substrate types in selected populations, n = number of samples, excl. = number of colonies with regenerating protonema and sparse shoots for which age cannot be determined. HOIARCnC ECOLOGY 12: O10' wood non wood Colony size (cm^) Fig. 8. Capstile number per colony in relation to colony size (log scales). colonies and generally bad conditions for the growth of O. lineare. The proportion of these colonies varied between localities (Fig. 5). Thus we interpret the differences between populations with high and low proportion of old colonies as a result of differences in colony mortality, either between different localities or between substrate types. Localities with younger colonies are generally inland localities where the higher mortality may represent some type of population level response to climatic factors. In maritime pine forests, rotten wood is apparently more favourable for survival than in inland spruce forests (cf. Hedenas et al. in press). The same argument can be used for differences between substrate types. The only exceptions to this pattern, i.e. young populations without signs of any mortality (Yllevadsmossen and Igarstorp), probably represent truly recent colonizations. Patterns of occurrence on different scales There was a clear clumping in the occurrence of O. lineare. even in seemingly favourable habitats. With few exceptions, the clumping pattern has been observed on all scales (between regions, between localities and within localities). This indicates that the presence of O. lineare at a given spot is affected by the abundance of the species in the neighbourhood. The pattern on the local scale was investigated within areas of .*5Ox5O and 450x5 m. The area of lowest neighbourhood radii (1-2 m), which did not show any clumping, included only a few substrate patches available for (}. lineare (Fig. 4). For the largest radii (11 :intl 15 m) the clumping was nonsignificant. We doubt that this is a "critical distance effect" (i.e. distance beyond which the presence of O. lineare has no influence) because (I) the data sets for different distances were not independent. (2) lower sample sizes (due to wider border zone) reduced the power of t-tests, and (3) the 450 X 5 m transect demonstrated clumping at larger distances. high mortality. Matiy older colonies were being overgrown by blue-green algae, leprariaceous lichens or other mosses (Hedenas et al. in press). (2) The small differences in the proportion of old-age colonies between populations in localities with dominant old-age classes also indicate a balance between extinction and establishment of colonies. In cases of population expansion we would expect greater differences because of different times since establishment. We would also expect a relationship between the proportion of number of colonies in the age class and its age. In the localities studied (except Stenestad II and Igarstorp) the differences between younger age classes (1-4 year) within a locality, were insignificant, which indicates that gametophore formation did not differ very much between years. (3) The difference in age structure on different substrate types could hardly be explained as a result of on-going expansion. This proportion of younger colonies tends to be increased by excluding from the analysis, colonies with a large proportion of protonema, as these can be reasonably assumed to be older than the individual shoots within them. However, the presence of these colonies indicates mortality of plants within Tab. 7. Capsule and spore production at selected sites. Site Total Total cover of OL (cm-) Number of capsules Calculated total spore output (millions) 150 S30 Kullaberg Slcncstad I Stencstad II Igarstorp Yllevadsmossen Haverdal Vikslatt Mosseberg Kulltorp 1 Kulllorp II Kulllorp 450 X 5 Kulltorp 50 X 50 to o I2(K) 1469 3647 a) 7032 a) sum of values estimated using the power curve for individual colonies. 170 HOLARCTIC ECOLOGY On a regional scale, O. lineare shows a marked southwestern distribution, occurring in many of the available localities in northwestern Skane and southern Halland and only in a few localities in Vast ergot land. It is absent in the eastern and central parts of southern Sweden. As the prevailing winds in southern Sweden are southwesterly, one would expect O. lineare to also have spread to areas further north and east. The regional pattern is instead correlated with the density of available forests, whereas climatic factors seem to be of less importance. Because the main substrate for O. lineare, deeaying wood, is a temporal substrate and O. lineare is a poor competitor (Hedenas et al. In press), O. lineare relies on the continuous availability of new substrate patches. The proportion of patches occupied is then determined by the ratio between establishment rate and colony mortality rate. If the establishment rate is high compared with the extinction rate, the species will occur in most of the available localities, i.e. be a core species (cf. Hanski 1982). Colonization rate is regulated by production, transport and establishment of diaspores. Spore production did nt)t appear to be limiting for O. lineare as there were enormous numbers of spores produced on well established localities (up to 12 million per m'; Tab. 7). However, 0. lineare release the spores very close to the ground which make them less likely to be caught by the wind. Lane and eoworkers (quoted in Wyatt 1982) found a strongly leptocurtic relationship, with 97% of all spores from Airicham angusfaiuin trapped within 2 m from the colony and yet 1% was found beyond 15 m from the colony. Miles and Longton (1987) found that although most of the spores of Xm'c/jum undulatum were transported more tban 2 m, the spore deposition {per tmit of area) deereased rapidly and was very low at distances > 15 cm from the colony edge. Soderstrom and Jon.s.son (in press) found that 43% of the spores from Piilidium pulcherrimum, an epixyllc forest hepatic, were deposited witbin 2.5 m from the source. Furthermore, the germination of moss spores in the field is rarely observed (Longton and Miles 1982, Longton and Schuster 1983, Miles and Longton 1987), even if they germinate well under experimental conditions, as is the case for O. lineare (Herben unpubl. data). Dispersal, including both spore transport and establishment, is then probably restricting the colonization rate of O. lineare. The same process occurring at the local scale may operate at the regional seale (Tab. 3). Wbere the density of suitable loealities/patches is high, localities/patches where O. lineare has disappeared may be efficiently colonized from nearby localities/patches. At a regional scale O. lineare will occur as a core species. This is the case in northern SkSne and southern Halland. where much of the natural broad-leaved forests are cut down and replaced by coniferous forests, mostly of spruee. which are suitable for O. lineare. In Sm^land and Vastergotland, eoniferous forests are l i O l ARCTIC ECOLOGY 12:2 (1989) more natural and much of the broad-leaved forests remain unreplaeed. In these forests, the man-made disturbances are not so large and available localities are more scattered. Therefore, the chance for colonization is much less, and O. lineare will occur only on a few of the loealities, i.e. as a satellite speeies (ef. Hanski 1982). This will make these forests less subject to invasion by a foreign species (cf. Orians 1986). Within tbe occupied localities, it may, however, build up large local populations if substrate is abundant, as on Mosseberg in Vastergotland. Patterns like this have also been observed for epixylie speeies in northern Sweden (Soderstrom 1987). Acknowledgements - We ihank H. During. T. Ebenhard. L. Ericson and H. Hytleborn for comments on the manuscript. T. Hallingback and S. Franscn for help with localiiies. K. Lindahl for help with spore counting, A. Nordgren for drawing Fig. 4. and N. Rollison for correcting the English. Financial support was received from the Karin and Axel Btnning's Foundation. References Angstrom, A. 1953a. Temperatur, humiditet. - In: Svenska Sallskapet for Antropologi & Geografi (eds). Atlas over Sverige. Stockholm, sheet 25-26. (In Swedish and English). - 1953b. Maximi- och mininiitomperaturer, Srslider. vegetationsperioden, temp.-ktimatets forandringar. - In: Svenska Sallskapet for Antropologi & Geografi (eds). Atlas over Sverige. Stockholm, sheet 27-28. (In Swedish and English). Bazzaz. F. A. 1986. Life history of colonizing plants: some demographic, genetic, and physiological features. - In: Mooney, H, A. and Drake. J, A. (eds), Ecology of biological invasions of North America and Hawaii. Eeol. Stud. 58, pp. 9fr-ilO. Connell, J. H. and Sousa, W. P. 1983. On the evidence needed to judge ecological stability or persistence. - A m . Nat, 121: 789-824. Damsholt. K. and Holmen, K. 1971. Orlhodontiurn lineare Schwaegr. fundet I Sverige. - Lindbergia 1: 115.(In Danish with English summary). During. H. and ter Horst. B. 1987. Diversity and dynamics in bryophyte communities on earth banks in a Dutch forest. Symp. BioL Hung. 35: 447-155. Ericson. L. 1977, The influence of voles and lemmings on the vegetation in a coniferous forest during a 4-year period in northern Sweden. - Wahlenbergia 4: 1-114, Frahm, J.-P. 1972. Die Ausbreitung von Campylopus introflexus (Hedw.) Brid. in Mitteteuropa. - Herzogia 2: 317330. Hallingback, T. and Soderstrom. L. 1987. Sveriges mossor och deras svenska namn. En kommenterad checklista. - Svensk Bot. Tidskr. 81: 357-388. (In Swedish with English summary). Hanski, I. 1982. Dynamics of regional distribution: the core and satellite species hypothesis. - Oikos 38: 210-221. Hedenas, L. 1981. Orthodonfium lineare ~ en mos.sa pa frammarsch. - Svensk Bot, Tidskr. 75: 157-161. (In Swedish with English summary). - and Soderstrom. L. 1988. Kapmossan. Orihodantium lineare, iSverige.-SvenskBot.Tidskr. 81:217-220. (In Swedish with English summary). - , Herben, T., Rydin. H. and Soderstrom, L. Ecology of the invading moss species Orthodontium lineare in Sweden: substrate preference and interactions with other species. J. Bryol. (in press). Herben, T. 1987. The ecology of the invasion of Orthodontium lineare Schwaegr. in central Europe. - Symp, Biol. Hung. 35: 323-333. Hill. M. O. 1973. The intensity of spatial pattern in plant communities. - J. Ecol. 61: 225-235. Hiibschmann, A. von. 1970. Ueber die verbreitung einiger seltener Laubmoose in Nordwesldeutschen Pflanzengesellschaften. - Herzogia 2: 63-75. Kimmel, U. 1962. Entwicklung einiger Moose und Flechten auf Dauer-Untersuchungsflachen. -Oberh. Ges. Natur- u. Heilkunde Ber. N.F. Nalurw. Abt. 32: 131-160. Longion. R. E. and Miles. C. J. 1982. Studies on the reproductive biology of mosses. - J. Hattori Bot. Lab. 52: 219-240. - and Schuster, R. M. 1983. Reproductive biology. - In: Schuster. R. M. (ed.). New manual of bryology. Hattori Bot. Lab,. Nichinan, pp. 386-462. Mcijer, W. 1952. The genus Orthodonlium. - Acta Bot. Necrl. 1: 1-80. Miles. C. J. and Longton. R. E. 1987. Reproductive biology of Atrichum undulatiim. - Symp. Biol. Hung. 35: 193-207. Mooncy, H. A. and Drake, J. A. (eds) 1986. Ecology of biological invasions of North America and Hawaii. - Ecol. Stud. 5S. Springer. Muhle, H. 1970. Zur Ausbreitung vom Orlhodondum lineare Schwaegr.: Orthodontium in Schwiirzwald. - Herzogia 2: 107-112. Ochyra, R. 1982. Orthodonlium lineare Schwaegr. - a new Species and genus in the moss flora of Poland. - Bryologische Beitrage 1: 23-36. Orians, G. H. 1986. Site characteristics favouring invasions. In: Mooney, H. A. and Drake. J. A. (eds). Ecology of biological invasions of North America and Hawaii. Ecol. Stud. 5S. pp. 133-148. Pielou, E. C. 1981. Rapid estimation of standing crop of intertidai fucoids on an exposed shore. - J . Environ. Manage. 13: 85-98. Richards, P. W. and Smith, A. J. E. 1975. A progress report on Campytopus inlroflexus iWuiis.) Brid- and C- polvtrichoides De Not. in Britain and Ireland. - J. Bryol. 8: 293-298. Soderstrom. L. 1987. Dispersal as a limiting factor for distribution among epixylic bryophytes. - Symp. Biol. Hung. 35: 475-484. - 1988, Cryptogam species sequence in relation to substrate variables of decaying coniferous wood in northern Sweden. - Nord. J. Bot. 8: 89-97. - and Jonsson, B. G, Spatial pattern and dispersal in the leafy hepatic Ptitidium pukherrimum. - J. Bryol. (in press). Sousa, W. P. 1984, The role of disturbance in natural communities. - Ann. Rev. Ecol. Syst, 15: 353-391, Wall^n, C. C. 1953. Sno, hagel, aska. Arsncderbord. arets medeltemperatur. - ln: Svenska Siillskapct for Antropologi & Geografi (eds). Attas over Sverige. Stockholm, sheet 31-32, (In Swedish and English). Wyatt. R, 1982. Populationecoiogy of Bryophytes.-J. Hattori Bot. Lab. 52: 179-198, HOLARCTIC ECOLOGY 12:2 (1989)

Journal

Ecography – Wiley

Published: Jun 1, 1989

There are no references for this article.