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

Learn More →

Microdistribution of butterflies in a mosaic‐like habitat: The role of nectar sources

Microdistribution of butterflies in a mosaic‐like habitat: The role of nectar sources Loertscher, M , Erhardl, A and Zettel, J 1995 Microdistnbution of butterflies in a mosaic-lile habitat The role of nectar sources - Ecography 18 15-26 The microdistnbution of five butterfly species through their flying season was analyzed in a mosaic-like habitat, brought about by secondary succession In order to explain the pattems observed, activity pattems and the use and distribution of nectar sources were determined Emphasis was laid on the changing allocation of visits to flower species and changing abundances of flowers dunng the season The use of nectar sources was basically limited to three flower species, Centaurea scabiosa, C bracteata and Serratuta tmctona As a consequence, niche breadth values were generally low and niche overlaps generally high Some butterflies changed their pattems of flower visits dunng the season and therefore reduced niche overlap with the other butterfly species The microdistnbution oi Melanargia galathea, Lysandra condon, Ochlodes venatus and Lictona achitleae was strongly influenced by the distnbution of their preferred nectar sources as well as by areas generally nch in flowers Changing flower preferences of Melanargia galathea and Lysandra condon males dunng the course of the season were also expressed by changes in the correlations between the distnbution of these butterflies and their nectar planti> The distnbution of nectar sources was not found to be of importance for Coenonympha arcanta, a species which rarely visited flowers M Loenscher and J Zettel, Zoot Inst, Battzerstr. 3, CH-3012 Bern, Switzerland Erhardt, Bot Inst, Schonbetnstr 6, CH-4056 Basel, Switzerland. -A Introduction The microdistnbution of animals in a habitat is influenced by a vanety of abiouc and bioUc factors, such as the distnbution of food, shelter, microchmatically favourable spots or lntra- and interspecific interactions In butterfiies, the distnbution of food and microclimate are thought to be especially important (Douwes 1975, Murphy et al 1984, Boggs 1987). The patchy distribution of particular nectar sources often entails a patchy distribution of butterflies (Douwes 1975, Brown and Ehrlich 1980, Brakefield, 1982). However, if nectar resources are not patchily distnbuted, it appears difficult to find any correlation of microdistribution of butterflies and their adult food sources (Sharp et al 1974), The microdistnbution of adult butterflies can fail to correlate with adult or Accepted 8 November 1994 Copynght © ECOGRAPHY 1995 ISSN 0906-7590 Pnnted in Denmaiic - all nghts reserved ECOGRAPHY 18 1 (1995) larval food sources for other reasons as well. For example in the Satynd Oeneis chryxus, a landmark maUng system IS responsible for a patchy distnbution of the adult butterflies (Daily et al, 1991) In most previous studies there was no quanufication of the relationship between the distnbution of butterflies and their nectar sources In addition, no clear change in the microdistnbution of butterflies due to changes in nectar resource availabihty and the subsequent use of these resources has been previously demonstrated In the present study, we analyzed the microdistnbution of several butterflies on a fine scale in a mosaic-like vegetaUon type on the southem slope of the Swiss Alps In particular, changes in the use of nectar sources over the flying season with respect to changes in abundance and distnbution of nectar sources in the habitat were consid- Fig 1 Location of the study site on Mte S Giorgio The enlargement represents the vegetation mosaic of the investigated study area White = vegetation type "Carex", light grey = vegetation type "Molinia", and dark grey = vegetaUon type "shrubs" (for explanaUon see text) ered m an attempt to explain the observed distribuuon^ patterns of butterflies Material and methods Study site The study site is located on Monte San Giorgio (elevauon 1097 m), about 10 km south of Lugano (Fig.l), The grassland vegetaUon on Monte San Giorgio has under16 gone unhindered secondary succession since the late 195O's, when agricultural use was abandoned (Stampfii et al. 1992) This succession has changed the former haymeadows to a mosaic of three main structural vegetaUon types 1) short-turfed, species-nch vegetation charactenzed by the dominant sedge Carex humtlts (Leysser), the grass Bromus erectus (Huds ) and the herb Hippocrepts comosa (L.) on shallow soils (type "Carex"), 2) highturfed, species-poor vegetaUon with the dominant grass Molinia anmdtnacea (Schrank) and the herbs Potentilla erecta (L.) and Serratula ttnctorta (L.) on deeper soils ECOGRAPHY 18 1 (1995) and in shaded areas (typie "Molinia") and 3) groups of shrubs and trees (typie "shrubs"). These three vegetaUon types are interlocked in a tight, structurally nch vegetation mosaic (Guggisberg 1990, Stampfii et al 1992) An area of 20 x 50 m representauve of the vegetation mosaic and situated at the southfacing slope of the mountain was chosen as study area (Fig 1). The distnbuUon of the three vegetauon typies was analysed by putting a gnd consisting of quadrats of 2 5 x 2 5 m over that area and mapping the three vegetaUon types m each quadrat considered as independent in the cases of M galathea, L Condon and C arcania Only in L achilleae, which is protected by its distastefulness, consecuUve sighungs on the same fiowers may be repeated recordings of the same individuals Statistical analysis Niche breadth and niche overlap measures were calculated using the formulas of Levins (1968) and Hurlben (1978) for breadth, and those of Pianka (1974) and Hurlben (1978) for overlap (Table 2) Resource states in the present study were defined as species of flowers Hurlben's measures of niche breadth and overlap were chosen because they alone incorporate the abundances of resource states in their calculaUon. Only fiower species visited by the butterfiies were included in this analysis For C arcanta no niche breadth or mche overlap was calculated as it was observed imbibing nectar on only five occasions DisuibuUon pattems were analysed for all butterfiy species and their preferred nectar sources by companng the variance and mean of observations per quadrat, and calculaUng an index of dispersion (ID) (Ludwig and Reynolds 1988), an ID greater than 1 implies a clumped distribution (Table 3). If more than 30 sampling units are available, a test staustic d can be computed, which if greater than 1 96 suggests a clumped distnbuuon To assess the importance of the abundance of fiowers for the presence of butterfiies in the quadrats. Spearman rank correlauons between the number of butterfiy observaUons and the number of fiowers per plant species and the total number of fiowers per quadrat were calculated (Table 4) The spaUal parUUoning of use of nectar sources in the habitat was assessed by calculaUng correlation coefficients between the disuibuuon of nectanng acUvlUes of the butterfiy species concerned (elinunaung quadrats where no dnnking acUviUes were detected in both species) A posiUve correlaUon would indicate that the species used the same quadrats for their dnnking acUvites while a negauve correlaUon would indicate the contrary, 1 e a parUUoning of the habiat between the species for nectar acquisiuon. The Bonferroni correcUon for mulUple tests was applied to the significance levels in these analyses To test whether there was a significant shift in the correlaUon with a parucular fiower species between penods, correlaUon coefficients were compared using the method proposed in Zar with correcUon for Ues (1984) Butterfly species Five species, Melanargta galathea (L), Coenonympha arcania (L ), Lysandra cortdon (Poda), Ochlodes venatus (Turati) and Lictorta achtUeae (Esper), accounted for 82% of all observations made dunng the study penod These species were treated further in the analysis The remaimng 18% consisted of observaUons of more than 10 other butterfiy species which were excluded from the analysis Data collection Transect walks were conducted dunng three penods in July and August 1989- 17-21 July (Penod 1), 28 July-1 Aug (Penod 2), 15-19 Aug (Penod 3) Dunng each penod, the transects were earned out on four subsequent days, separated by one day when the available fiowers were recorded by counting the number of fiower heads per species in each quadrat In order to gather the data for the distnbution of adult butterflies in the gnd, the transect walks were conducted along every second row of quadrats From 9 am unul 6 pm two transects per hour were made, each lasting 20-25 minutes Observations of butterfiies in the quadrats to the nght and to the left of the path were noted together with their acuviues The following activiUes were disunguished* fiymg, dnnking, resting, basking, courUng and egglaying The observaUons are probably biased towards fiying and dnnking acUviues as these are most prominent Nevertheless the narrowness of the transect area (2.5 meters to each side) reduces this bias A butterfiy was considered dnnking when it introduced its proboscis into the fiower It was sitting on, resung when it was situng on the vegetaUon, on a flower or on the ground without adopUng the species-specific basking posture, and basking when it did adopt that jrosture In one species, L achilleae, no specific basking behavior could be disUnguished Egglaying was noted when a female was observed bending its abdomen and either dropping an egg to the ground or glueing It to a plant surface The method of data coUecUon has the inherent problem that observaUons of successive transects may not be independent. However, data from Scott (1975) and Shreeve (1992), as well as personal observaUons and expenments at another site, mdicate that the observaUons can be 2 ECOGRAPHY 18 1 (1995) Results Total number of observations and activity patterns A total of 2027 observaUons were made for M. galathea, 793 for C. arcania, 673 for L. coridon, of which only 49 17 Table 1 Available flowers and their use by four butterfly species N = number of open flower heads For the three plant species visited most often, data are presented for each penod separately PI, P2, P3 = Penod I, Penod 2 and Penod 3, respectively Available flowers N PI P2 P3 Total PI Centaurea bracteata P2 P3 Total Serratula tinctoria PI P2 P3 Total Total Betonica officinalis Buphthalmum salicifolium Total Scabiosa columbaria Total Leucanthemum vulgaris Total Polygala pedemontana Total Allium spp Total Prunella grandiflora Total Aster amellus Total Hippocrepis comosa Toul Hypochoeris maculata Total lx>tus comiculatus Total Total Knautia drymeia Gymnadema conopsea Total Asperula cvnanchica Total Globularia cordiifolia Total Peucedanum cervana Toul Centaurea scabiosa 50 229 119 398 8 177 29 214 M galathea 100 90 78 268 6 148 45 199 _ 19 40 59 9 7 2 1 _ 2 Flower visits by butterfly species C arcania L condon males 23 30 53 _ 1 3 4 _ 11 146 157 1 — 1 6 3 3 2 — - O venatus 70 53 12 135 2 14 2 18 _ 24 15 39 19 10 1 2 5 2 1 3 1 - L achtlteae 102 40 _ 142 _ _ _ 1 _ 1 6 4 1 2 2 — I 1 _ — - _ _ _ _ _ _ _ 2 2 1 — 1 _ _ I - were ob,servations of females, 401 forO venatus and 270 for L achilleae Because there were so few observations of L. cortdon females all the subsequent analyses were conducted only for the 624 observaUons of males of this species In penod 3 the flight penod of L achilleae had already ended In penod 1 only 29 observauons of L condon were made Therefore in Tables 1 and 3 no numbers appear for this species The five butterfly species displayed the following acuvity patterns In M galathea and L condon, flying predominated, followed by dnnking nectar then resung and basking In O venatus and L achtlleae, dnnking was observed most frequently, followed by flying, resting and basking C. arcanta was mainly encountered in flight, followed by resting, basking and only very rarely visiung flowers The five species differed considerably in the proportion of times they were observed visiUng flowers For O. venatus and L. achilleae about half of the observauons were made while nectanng (48% and 53% resp), whereas in M. galathea (26%) and L condon males (34%) this acuvity was less frequent. Finally just 0 2% of the observations of C. arcanta involved nectanng. In all species, oviposiUon and courtship events were observed only occasionally and the spatial distnbution of these acuviues can at best be used as a hint for explaining the nucrodistribuuon of the respeeuve spiecies In all five butterfly species, most of the basking observaUons were 18 made m the first two observation hours (9 am-11 am) Otherwise no eoncentrauon of particular acuviUes to a specific dayume was observed Nectar resource use and niche measures Out of an array of jxitential nectar sources the three flower species Centaurea scabtosa (L.), C bracteata (Scop ) and S tinctona were visited preferenually (Table 1). They showed different abundance patterns dunng the course of the study (Fig 2a) When visiung these three flowers species, the butterflies exhibited quite different patterns (Fig 2) O venatus showed no clear preferences and allocated their visits about in proporuon to the relaUve abundances of the three fiower species L cortdon males visited mainly C. scabtosa and S ttnctorta, using them in proporuon to their relaUve abundance M galathea paid most visits to the two Centaurea species. In period 2 however, when C bracteata tnpled its abundance relaUve to penod 1, M. galathea allocated nearly ten umes as many visits to this fiower species than in penod 1. In jjenod 3, when the abundance of C bracteata dropped to just 4%, M. galathea sull attnbuted 28% of its visits to that flower species. L achtlleae on the other hand clearly preferred C scabtosa independently of its relaUve abundance. ECOGRAPHY 18 1 (1995) Fig 2a Relative proportion of the three flower species most often visited by the butterflies in the three study penods b-e Proportion of visits by the butterflies attnbuted to these three flower species in each penod a) Preferred flowers b) LachBteae The preferences for the three main nectar sources are also reflected by the low niche breadth values obtained for all species in all penods (Table 2) In conu-ast, the niche overlap measures were high m almost all compansons (Table 2). In penod 1, M. galathea, O venatus and L achtlleae visited the same fiower sf>ecies in similar proportions as shown by the high values obtained with Hurlbert's measure of niche overlap The lower values for Hurlbert's measure of mche overlap in penod 2 between these three species are caused by the different allocaUon of visits to the three mam nectar sources (Fig 2) The lower niche overlap values obtained in the compansons of M galathea and O. venatus with L coridon 2* ECOGRAPHY 18 1 (1995) are mainly caused by the different uulization of C bracteata by both and of S. tmctorta by M. galathea only. Distribution patterns All prefened nectar sources were highly clumped (Table 3) The butterfiies vaned considerably in the degree of aggregaUon of observaUons C arcanta showed moderate clumping while L achilleae marks the other extreme with clearly clumped distnbution pattems (Table 3) There were pronounced differences in the degree of clumping in between penods for L coridon and O vena19 Table 2 Food niche relationships between the four butterfly species, calculated for each penod separately Niche overlap measures following Hurlbert (1978) above diagonal and Pianka (1974) below diagonal For each penod number of flower species visited and niche breadth values following Levins (1968) and Hurlbert (1978) are listed below overlap measures Levins stand = Levins/ Slower species Penod 1 overlap M galathea L. condon O venatus L achilleae Penod 1 breadth ZFlower speeies Levins Levins stand Hurlbert Penod 2 overlap M galathea L condon O venatus L achilleae Penod 2 breadth ZFIower species Levins Levins stand Hurlbert Penod 3 overlap M galathea L. cortdon O venatus L achilleae Penod 3 breadth ZFlower species Levins Levins stand Hurlbert M galathea 0000 0 959 0 995 15 1370 0 090 0 130 M galathea 0 545 0 700 0516 10 0 890 0 090 0210 M galathea L condon 0000 — . 0006 0.000 15 1800 0 120 0.270 L condon 1635 0 979 0 899 10 2 100 0 210 0410 L condon • O venatus 5 053 0000 ——_ 0 970 15 2 630 0 180 0 190 O venatus 2006 2 115 • — . L achilleae 8 610 0000 5 448 15 1380 0090 0 140 L achilleae 1 826 2014 1 896 0 885 10 2410 0 240 0 490 O venatus I 000 0 100 0 300 L achilleae tus Males of L condon showed a clearly more aggregated distnbuuon pattern in penod 3 than in penod 2 The opposite IS the case for O venatus, whose distnbuUon was more aggregated in penods 1 and 2 than in jjenod 3 These differences may be due to the relatively few observations of L condon in penod 2 and of O venatus in penod 3 The frequency of observaUons of each butterfiy species m the different quadrats of the study area is presented in Figs 3-7 In addiuon, proporuons of flower visits are also given ObservaUons of M galathea were concentrated on the northwestern side of the study site (top left in Fig 3), where vegetaUon type "Carex" and type "shrubs" form a small scale mosaic of high structural diversity The projrorUon of flower visits in the distnbuUon pattern does not follow any parucular trend. Males of L. cortdon were most often encountered in the area m the western comer of the gnd dominated by vegetaUon type "Carex" and the quadrats just above that area (Fig. 4). These quadrats also seem to play an important role in the acquisiUon of nectar, as large proporuons of the observations there were 20 fiower visits For C arcanta, no clear concentration of observations was found, even though all quadrats with more than nine observaUons were quadrats bordered by shrubs or trees (Fig 5) The observaUons of O. venatus were concentrated in the northern comer of the gnd, where vegetaUon types "Carex" and "shrubs" are responsible for a sUiicturally highly diverse habitat (Fig 6) However, O venatus was frequently observed in several quadrats in the eastern part, where proporuons of fiower visits were particularly high Finally, L achtlleae did not show a clear preference for a particular area in the study site However, in all quadrats with more than mne observations the proporuon of fiower visits was high (Fig 7) Correlations between butterfly observations and nectar sources and habitat partitioning for nectar acquisition The distnbuUon of the total number of observations of M galathea, L coridon, L achtlleae and O. venatus correECOGRAPHY 18 I (1995) Table 3 Spatial distnbution of the four butterfly species regularly visiUng flowers and of their preferred flower species N = number of open flower heads (Total only) and number of buUerfly observaUons (for each penod separately). Mean = mean number of flower heads or butterfly observations per quadrat, Var = vanance, ID = index of dispersion, d = test statisuc ID and d according to Ludwig and Reynolds (1988) N Mean Var ID d expressed by an increasing correlation value with C bracteata, while the value with C scabiosa remains high This IS again probably due to the correlation m the distnbution of these two plant species C, arcania showed no correlation with any nectar source, as might be expected from the very low projxjrtion of flower visits in the daily activity pattem of this species. Centaurea scabiosa Centaurea bracteata Serratula tinctorta Melanargia galathea Penod 1 Penod 2 Penod 3 Total 54 9 17 9 19,9 86 82 3 46 7 14 6 18 5 15,7 27 9 Discussion The four species regularly visiting flowers linuted their visits to basically three plant species with purple flowers of the family Asteraceae These observations fit well with observations of several authors (e g Jennersten 1984, Steffny et al 1984, SBN 1987, Ebert 1991, Erhardt and Thomas 1991) and the general butterfiy flower syndrome (Faegn and Van der Pijl 1971) The pattems of allocation of visits to these three plant species changed between observation penods With the exception of L achilleae, which clearly preferred C scabtosa, and M galathea which showed a preference for C bracteata when it became abundant, the butterflies visited the three flower species m proportion to their relative abundances The reasons for these differences are largely unknown One possible reason is indicated by the narrow ruche breadths and the changing high niche overlap values which raise the question whether the butterfiy species were comp>etmg for these resources The shift in the visiting pattems exhibited by M galathea and L condon between penod 2 and 3 and the subsequent decrease in niche overlap might lead to this conclusion A possible way to escape competition would be to partition the habitat for acquiring nectar between the species However our analysis indicated no such pattems of habitat partitioning Furthermore analyses of niche relationships are considered lnappropnate for detecting competition (Colwell and Futuyma 1971) It has been argued that preferences for particular nectar sources may represent a major factor for the microdistribution of butterflies in a given habitat (Boggs 1987) Posiuve correlations between the distnbuuon of butterfiies and their preferred nectar sources have been found in the Lycaenid Heodes virgaureae (Douwes 1975), the Satynd Maniola juntna (Brakefield 1982) and the Nymphalid Euphydryas chalcedona (Brown and Ehrlich 1980), Furthermore, Murphy et al (1984) showed that the distnbution of nectar sources lnfiuences oviposition sites in E chalcedona However, only Douwes (1975) quantified the relationship between nectar sources and butterfiies, although he used for his quantification higher systematic groups of plants, such as Ligulifiorae, Tubuliflorae and Dip,sacaceae He found that the distnbuUon of butterflies was strongly lnfiuenced by the distnbution of Ibbulifiorae dunng the whole flying penod of H. vtrgaureae Furthermore he found that the goodness of fit of the regression (R-) of the distribution of the male butterfiies 21 Coenanympha arcania Penod 1 Penod 2 Total Lysandra coridon males Penod 2 Penod 3 Total Ochlodes venatus Penod 1 Penod 2 Total Lictoria achilleae Penod 1 Period 2 Total lated with the distnbution of C scabiosa and the total number of flowers and, with one exception (O venatus), also with C bracteata (Table 3), Although L achilleae and L cortdon males visited the flowers of C. bracteata only occasionally (Table 1), they showed a correlation with this plant This is probably due to the correlation between the distributions of C scabiosa and C. bracteata (Rho = 0,25, p < 0,05) No habitat partitioning for nectar acquisition could be detected. On the other hand, only m one companson (L achtlleae and O venatus) was the correlation of nectanng observations positive When the correlations are examined for each penod separately, it becomes evident that L condon showed correlations with C scabtosa m penods 2 and 3 and with 5 tinctoria in penod 3, In penod 2 the correlation with S tinctorta was distinctly weaker This reflects the changing attnbution of visits to this flower between penod 2 and 3 (Fig, 2) Even though they are not statisically significant, the negative correlaaons with Buphthalmum saltctfolium ( L ) in all penods invites comment (Table 4) This plant grows in shaded areas around trees and shrubs dominated by vegetation typie "Molinia" (Guggisberg 1990), which are usually avoided by L coridon males The changing flower preference of M galathea from C, scabiosa in penod 1 to C bracteata in penod 2 (Fig 2) is ECOGRAPHY 18 I (1995) • 1-6 observations 7-14 observations ^P ^ B 15-24 observations >24 observations (9 O O 3 0% • 1-25% 26-50% 50-75% 75-100% Flower visits Fig 3 Distnbution of the pooled observations (N = 2027) of Melanargia galathea in the study area and the percentage of flower visits for each quadrat 1-2 observations 3-5 observations 6-9 observations >9 observations o o O (5 0% 1-25% 26-50% 51-75% >75% Rower visits Fig 4 Distnbution of the pooled observabons (N = 624) of males oi Lysandra condon in the study area and the percentage of flower visits for each quadrat. ECOGRAPHY 18 I (1995) For symbols see Rg 4 Fig 5 DistnbuUon of the pooled observations (N = 793) ot Coenonsmpha arcania in the study area and the percentage of flower visits for each quadrat with the fiower groups increased between his first and second study penod and remained constantly high between the second and third penod However, since his analysis was conducted with the taxon Tububfiorae, which includes several plant species, seasonal changes in the abundance of different species belonging to this taxon and subsequent changes of the fiower preferences of H virgaureae might have gone unnouced In four of the five species analysed in the present study (M galathea, L cortdon, O venatus and L achtlleae), the distnbution of nectar sources appears to be an important determinant of microdisuibution since they exhibited posiUve correlauons with fiower-nch quadrats and with their respecUve nectar sources In addition, seasonal changes in the allocaUon of visits of the butterfly species to their nectar sources were reflected by increasing or decreasing correlaUon values In C arcania on the other hand, nectar sources do not seem to play a major role m determining microdistnbution Instead the occurrence of isolated shrubs used as basking and perching sites (Ebert 1991, unpubl data) probably accounts for its clustered distnbuUon The quesUon anses whether vanaUon in nectar availabiLty of the different nectar plants could explain the different visitaUon pattems of the observed butterflies Although nectar production was not measured and nectar composiuon not analysed in the present study, it can be stated from numerous lnvestigaUons (e g Baker and Baker 1990 and references therein) that different plant species differ in amount and quality of their nectar In addiUon, microbial decomposiUon of sucrose to glucose and fmctose can significantly reduce the proporUon of sucrose in nectar of ageing flowers This could reduce the attracuveness of older flowers to butterflies, since bunerECOGRAPHY 18 I (1995) flies have a distinct preference for sucrose over glucose and fructose (Erhardt 1991, 1992) A novelty effect of fiowers which come into bloom may also attract butterfiies which may then become conditioned to the new nectar plant (e g the disproponionally large fracuon of Melanargia galathea visits to Centaurea bracteata in Penod 2 in the present study) Furthermore, climauc conditions might affect nectar production differently m different plant species An lnvesUgaUon correlating fiower visitaUon pattems of butterfiies with nectar production of preferred versus neglected nectar plants would shed more light to this subject Altemauve explanations for the observed distnbution pattems, such as the distnbuUon of larval foodplants (Brown and Ehrlich 1980) or mate locaung behavior (Rutowski 1991) should also be considered In field studies, such as mark-recapture-studies or transect counts, more males than females are usually observed (e g Erhardt 1985, Daily et al 1991, Loertscher 1991) Differences m the behavior of the sexes, such as a panicular male mate locating behavior, can also lead to differences in microhabitat choice (Brakefield 1982), Males of L Condon showed a clear preference for the vegetaUon tyjie "Carex" and its immediate surroundings TTus could be due not only to a concenu^uon of preferred nectar plants m this vegetauon type, but also to the larval foodplant H comosa, which was found exclusively m vegetaUon type "Carex" (Guggisberg 1990) Male butterfiies should have the highest chance of obtaining a virgin female in the immediate proximity of the female pupaUon sites If these are predictable, for example because larval foodplants are patchily disu-ibuted and larvae pupate near their foodplants, such areas should be preferenually visited by males (Rutowski 1991) These factors may therefore also 23 For symbols see Rg 4 Fig 6 Distnbution of the pooled observations (N = 401) of Ochlodes venatus in the study area and the percenUge of flower visits for edch quadrat determine the microdistnbution of Z, coridon males InL achilleae, which feeds on the same food plant as L condon the clumped distribution may also be caused partly by the same phenomenon descnbed above In conclusion, this study shows that beside the needs for larval resources and specific behaviours such as male mate locating behaviour, species-specific needs for adult resources, the spatial distnbuUon and the temporal varlauon of these resources play a cmcial role for the microdistnbution of butterfiies in the field Acknowledgements - We would like to thank the following persons for the continuous moral and intellectual support dunng this study A Stampfli, F Guggisberg, C Antognoli, A Valsangiacomo and M Zimmermann Constructive cnUcisms on earlier versions of the manuscnpt by M Zimmermann, A Porter and C Wiklund greatly helped to improve it Figure 1 was, drawn by M Kummer and I Scharer The study presented here was part of a multidisciplinary project on unfertilized meadows and pastures of the Canton Tessin managed by O Hegg of the Botanical Inst Univ of Berne and financially supponed by the SWISS N F Project 31-9096 87 and by grants from the governments of Switzerland (BUWAL) and the Cantone Ticino (Dipar- For symbols see Fig 4 Fig 7 DistnbuUon of the pooled observanons (N = 270) of Ltctorta achilleae in the study area and the percentage of flower visits for each quadrat 24 ECOGRAPHY 18 1 (1995) Table 4 Spearman rank correlaUon coefficients between number of butterfly observaUons and number of flower heads per quadrat CorrelaUon coefficients are given for compansons with the most often visited nectar plants and as a contrast with a flower that was only rarely visited (Buphthalmum salicifolium) Correlauons are significant at the Bonferroni-corrected probability level of (0 05/95 = 0 0005) and (0 01/95 = 0 0001) for overall significance levels of 0 05(*) and 0 01(**) over 95 simultaneous tests Spearman rank correlation coeffieienLs (corrected for ties) C scabtosa 0 43** 0 45** 0 34** 051** C bracteata 0 13 0 45** 5 ttnctona 0 11 B salicifolium 01 Flower Tolal 0 47** 0 44** 0 55** 0 55** 0 28* 0 35** 0 46** 0 45** 0 16 0 18 0 15 0 24 0 32** 0 35** 0 33** 0 36** 0 45** 0 25 0 46** Melanargta galathea Penod 1 Penod 2 Penod 3 Total Lysandra cortdon males Penod 1 Penod 2 Penod 3 Total Coenonympha arcanta Penod 1 Penod 2 Penod 3 Total Ochtodes venatus Penod I Penod 2 Penod 3 Total Lictoria achilleae Penod 1 Penod 2 Total 0 48** 0 15 0 05 -0 05 0 15 0 08 -0 -0 -0 -0 16 17 17 18 0 33** 0 28* 0 37** 0 05 0 23 0 07 0 14 0 5** 0 53** 0 22 0 5** 0 44** 031** 048** 0 23 0 II 0 32** 0 27* 0 26 0 07 -0 07 -0 04 0 08 0 03 -01 02 0 06 -0 02 0 16 0 26 -001 0 03 0 24 0 05 0 24 0 29* 0 18 0 17 0 03 -0 15 -0 06 0 18 0 24 -0 07 -0 02 -0 02 phitenor (Lepidoptera, Papilionidae) - Ecol Ent 16 425434 - 1992 Preferences and non-preferences for nectar constituents in Omithoptera prtamus posetdon (Lcpidoptera, Papihonidae) - Oecologia 90 581-585 - and Thomas, J A 1991 Lepidoptera as indicators of change in the seminatural grasslands of lowland and upland Europe References - In Collins, N M and Thomas, J A (eds). The conservation of insects and their habitats Academic Press, London, Baker, H G and Baker, I 1990 The predictive value of nectar pp 213-234 chemistry to the recognition of pollinator types - Israel J Bot 39 157-166 Faegn, K and Van der Pijl, L 1971 The pnnciples of pollination ecology - Pergamon Press. Oxford Boggs, C L 1987 Ecology of nectar and pollen feeding in Lcpidoptera - In Slansky, F and Rodnguez, J G (eds), Guggisberg, F 1990 Verbuschende Wiesen und Weiden auf dem Monte San Giorgio/TI M Sc lhesis, Univ of Berne Nutntional ecology of insects, mites, spiders and related invertebrates John Wiley and Sons, New York, pp 369- Hurlbert, S H 1978 The measurement of niche overlap and some relatives - Ecology 59 dl-ll 391 Brakefield, P M 1982 Ecological studies on the buUerfly Jennersten, O 1984 Flower visitanon and polhnation efficiency of some North European butterflies - Oecologia 63 80-89 Maniolajuntna inBnl&m I Adult behavior, microdistnbution and dispersal - J Anim Ecol 51 713-726 Levins, R 1968 Evolution in changing environments some theoreUcal explorauons - Monographs on population biolBrown, 1 L and Ehrlich, P R 1980 Population biology of the ogy, Pnnceton University Press, New Jersey Checkerspot Butterfly Euphydrtas chalcedona Smicture of Loertscher, M 1991 Population biology of two satynne butterthe Jasper Ridge Colony - Oecologia 47 239-251 flies Erebta meolans (de Prunner, 1798) and Erebia aethiops Col well, R K andFutuyma, D J 1971 On the measurement of (Esper, 1777) (Lepidoptera, Satyndae) - Nola Lep 2 niche breadth and overlap - Ecology 52 567-576 22-31 Daily, G C, Ehrhch, P R and Wheye, D 1991 Deteiroinants of spatial distnbuuon in a populauon of the subalpine butter- Ludwig, J A and Reynolds, J F 1988. StaUsUcal Ecology-A pnmer on methods and computing - John Wiley and Sons, fly Oenets chryxus - Oecologia 88 587-596 New York Douwes, P 1975 Distnbuuon of a population of the butterfly Heodes virgaureae - Oikos 26 332-340 Murphy, D D, Menninger, M S and Ehrlich, P R 1984 Nectar source distnbuuon as a determinant of oviposiUon Ebert, G 1991 Die Schmetterlinge Baden-Wurttembergs Band host species in Euphydnas chatcedona - Oecologia 62 2, Tagfalter 2 - Ulmer, Stuttgart 269-271 Erhardt, A. 1985 Wiesen und Brachen als Lebensraum fUr Pianka, E R 1974 Niche overlap and diffuse competition Schmetterlinge - Birkhkuser, Basel Proc Nat Acad Sci USA 71 2141-2145 - 1991 Nectar sugar and amino-acid preferences of Battus ECOGRAPHY IS 1 (1995) limento dell' Ambiente) as well as from the Lega Svizzera per la protezione della natura (Sezione Ticino) Rutowski, R L 1991 The evoluuon of male mate-locating behavior in butterflies - A m Nat 138 1121-1139 SBN 1987 Schweizenscher Bund fur Naturschutz Tagfaller und lhre Lebensraume - Basel, Scott, J A 1975 Flight patterns among eleven species of diumal Lepidoptera - Ecology 56, 1367-1377 Sharp, M A , Parks, D R and Ehrlich, P R 1974 Plant resources and butterfly habitat selecUon - Ecology 55 870875 Shreeve, T G 1992 Adult behaviour - In Dennis, R L H (ed). The ecology of butterflies in Bntain Oxford University Press, Oxford, pp 22-45 Stampfli, A , Loenscher, M and Guggisberg, F 1992 Artenruckgang in Magerwiesen - Wissenschaftbcher Naturschutz am Monte San Giorgio Gaia 2 105-109 Steffny, H , Kratochwil, A and Wolf, A 1984 Zur Bedeutung verschiedener Rasengesellschaften fur Schmetterlinge (Rhopalocera, Hespenidae, Zygaenidae) und Hummeln (Apidae, Bombus) im Naturschutzgebiet Taubergiessen (Oberrheinebene) - Natur Landsch 59 435-443 Zar, J H (1984) Biostatistical Analysis - PrenUce-Hall, Englewood Cliffs, NJ ECOGRAPHY 18 1 (1995) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ecography Wiley

Microdistribution of butterflies in a mosaic‐like habitat: The role of nectar sources

Loertscher, Mathias; Erhardt, Andreas; Zettel, Juerg
Ecography , Volume 18 (1) – Mar 1, 1995
Download PDF
13 pages

Loading next page...
 
/lp/wiley/microdistribution-of-butterflies-in-a-mosaic-like-habitat-the-role-of-5X3VpfPQBA

References (23)

Publisher
Wiley
Copyright
Copyright © 1995 Wiley Subscription Services, Inc., A Wiley Company
ISSN
0906-7590
eISSN
1600-0587
DOI
10.1111/j.1600-0587.1995.tb00115.x
Publisher site
See Article on Publisher Site

Abstract

Loertscher, M , Erhardl, A and Zettel, J 1995 Microdistnbution of butterflies in a mosaic-lile habitat The role of nectar sources - Ecography 18 15-26 The microdistnbution of five butterfly species through their flying season was analyzed in a mosaic-like habitat, brought about by secondary succession In order to explain the pattems observed, activity pattems and the use and distribution of nectar sources were determined Emphasis was laid on the changing allocation of visits to flower species and changing abundances of flowers dunng the season The use of nectar sources was basically limited to three flower species, Centaurea scabiosa, C bracteata and Serratuta tmctona As a consequence, niche breadth values were generally low and niche overlaps generally high Some butterflies changed their pattems of flower visits dunng the season and therefore reduced niche overlap with the other butterfly species The microdistnbution oi Melanargia galathea, Lysandra condon, Ochlodes venatus and Lictona achitleae was strongly influenced by the distnbution of their preferred nectar sources as well as by areas generally nch in flowers Changing flower preferences of Melanargia galathea and Lysandra condon males dunng the course of the season were also expressed by changes in the correlations between the distnbution of these butterflies and their nectar planti> The distnbution of nectar sources was not found to be of importance for Coenonympha arcanta, a species which rarely visited flowers M Loenscher and J Zettel, Zoot Inst, Battzerstr. 3, CH-3012 Bern, Switzerland Erhardt, Bot Inst, Schonbetnstr 6, CH-4056 Basel, Switzerland. -A Introduction The microdistnbution of animals in a habitat is influenced by a vanety of abiouc and bioUc factors, such as the distnbution of food, shelter, microchmatically favourable spots or lntra- and interspecific interactions In butterfiies, the distnbution of food and microclimate are thought to be especially important (Douwes 1975, Murphy et al 1984, Boggs 1987). The patchy distribution of particular nectar sources often entails a patchy distribution of butterflies (Douwes 1975, Brown and Ehrlich 1980, Brakefield, 1982). However, if nectar resources are not patchily distnbuted, it appears difficult to find any correlation of microdistribution of butterflies and their adult food sources (Sharp et al 1974), The microdistnbution of adult butterflies can fail to correlate with adult or Accepted 8 November 1994 Copynght © ECOGRAPHY 1995 ISSN 0906-7590 Pnnted in Denmaiic - all nghts reserved ECOGRAPHY 18 1 (1995) larval food sources for other reasons as well. For example in the Satynd Oeneis chryxus, a landmark maUng system IS responsible for a patchy distnbution of the adult butterflies (Daily et al, 1991) In most previous studies there was no quanufication of the relationship between the distnbution of butterflies and their nectar sources In addition, no clear change in the microdistnbution of butterflies due to changes in nectar resource availabihty and the subsequent use of these resources has been previously demonstrated In the present study, we analyzed the microdistnbution of several butterflies on a fine scale in a mosaic-like vegetaUon type on the southem slope of the Swiss Alps In particular, changes in the use of nectar sources over the flying season with respect to changes in abundance and distnbution of nectar sources in the habitat were consid- Fig 1 Location of the study site on Mte S Giorgio The enlargement represents the vegetation mosaic of the investigated study area White = vegetation type "Carex", light grey = vegetation type "Molinia", and dark grey = vegetaUon type "shrubs" (for explanaUon see text) ered m an attempt to explain the observed distribuuon^ patterns of butterflies Material and methods Study site The study site is located on Monte San Giorgio (elevauon 1097 m), about 10 km south of Lugano (Fig.l), The grassland vegetaUon on Monte San Giorgio has under16 gone unhindered secondary succession since the late 195O's, when agricultural use was abandoned (Stampfii et al. 1992) This succession has changed the former haymeadows to a mosaic of three main structural vegetaUon types 1) short-turfed, species-nch vegetation charactenzed by the dominant sedge Carex humtlts (Leysser), the grass Bromus erectus (Huds ) and the herb Hippocrepts comosa (L.) on shallow soils (type "Carex"), 2) highturfed, species-poor vegetaUon with the dominant grass Molinia anmdtnacea (Schrank) and the herbs Potentilla erecta (L.) and Serratula ttnctorta (L.) on deeper soils ECOGRAPHY 18 1 (1995) and in shaded areas (typie "Molinia") and 3) groups of shrubs and trees (typie "shrubs"). These three vegetaUon types are interlocked in a tight, structurally nch vegetation mosaic (Guggisberg 1990, Stampfii et al 1992) An area of 20 x 50 m representauve of the vegetation mosaic and situated at the southfacing slope of the mountain was chosen as study area (Fig 1). The distnbuUon of the three vegetauon typies was analysed by putting a gnd consisting of quadrats of 2 5 x 2 5 m over that area and mapping the three vegetaUon types m each quadrat considered as independent in the cases of M galathea, L Condon and C arcania Only in L achilleae, which is protected by its distastefulness, consecuUve sighungs on the same fiowers may be repeated recordings of the same individuals Statistical analysis Niche breadth and niche overlap measures were calculated using the formulas of Levins (1968) and Hurlben (1978) for breadth, and those of Pianka (1974) and Hurlben (1978) for overlap (Table 2) Resource states in the present study were defined as species of flowers Hurlben's measures of niche breadth and overlap were chosen because they alone incorporate the abundances of resource states in their calculaUon. Only fiower species visited by the butterfiies were included in this analysis For C arcanta no niche breadth or mche overlap was calculated as it was observed imbibing nectar on only five occasions DisuibuUon pattems were analysed for all butterfiy species and their preferred nectar sources by companng the variance and mean of observations per quadrat, and calculaUng an index of dispersion (ID) (Ludwig and Reynolds 1988), an ID greater than 1 implies a clumped distribution (Table 3). If more than 30 sampling units are available, a test staustic d can be computed, which if greater than 1 96 suggests a clumped distnbuuon To assess the importance of the abundance of fiowers for the presence of butterfiies in the quadrats. Spearman rank correlauons between the number of butterfiy observaUons and the number of fiowers per plant species and the total number of fiowers per quadrat were calculated (Table 4) The spaUal parUUoning of use of nectar sources in the habitat was assessed by calculaUng correlation coefficients between the disuibuuon of nectanng acUvlUes of the butterfiy species concerned (elinunaung quadrats where no dnnking acUviUes were detected in both species) A posiUve correlaUon would indicate that the species used the same quadrats for their dnnking acUvites while a negauve correlaUon would indicate the contrary, 1 e a parUUoning of the habiat between the species for nectar acquisiuon. The Bonferroni correcUon for mulUple tests was applied to the significance levels in these analyses To test whether there was a significant shift in the correlaUon with a parucular fiower species between penods, correlaUon coefficients were compared using the method proposed in Zar with correcUon for Ues (1984) Butterfly species Five species, Melanargta galathea (L), Coenonympha arcania (L ), Lysandra cortdon (Poda), Ochlodes venatus (Turati) and Lictorta achtUeae (Esper), accounted for 82% of all observations made dunng the study penod These species were treated further in the analysis The remaimng 18% consisted of observaUons of more than 10 other butterfiy species which were excluded from the analysis Data collection Transect walks were conducted dunng three penods in July and August 1989- 17-21 July (Penod 1), 28 July-1 Aug (Penod 2), 15-19 Aug (Penod 3) Dunng each penod, the transects were earned out on four subsequent days, separated by one day when the available fiowers were recorded by counting the number of fiower heads per species in each quadrat In order to gather the data for the distnbution of adult butterflies in the gnd, the transect walks were conducted along every second row of quadrats From 9 am unul 6 pm two transects per hour were made, each lasting 20-25 minutes Observations of butterfiies in the quadrats to the nght and to the left of the path were noted together with their acuviues The following activiUes were disunguished* fiymg, dnnking, resting, basking, courUng and egglaying The observaUons are probably biased towards fiying and dnnking acUviues as these are most prominent Nevertheless the narrowness of the transect area (2.5 meters to each side) reduces this bias A butterfiy was considered dnnking when it introduced its proboscis into the fiower It was sitting on, resung when it was situng on the vegetaUon, on a flower or on the ground without adopUng the species-specific basking posture, and basking when it did adopt that jrosture In one species, L achilleae, no specific basking behavior could be disUnguished Egglaying was noted when a female was observed bending its abdomen and either dropping an egg to the ground or glueing It to a plant surface The method of data coUecUon has the inherent problem that observaUons of successive transects may not be independent. However, data from Scott (1975) and Shreeve (1992), as well as personal observaUons and expenments at another site, mdicate that the observaUons can be 2 ECOGRAPHY 18 1 (1995) Results Total number of observations and activity patterns A total of 2027 observaUons were made for M. galathea, 793 for C. arcania, 673 for L. coridon, of which only 49 17 Table 1 Available flowers and their use by four butterfly species N = number of open flower heads For the three plant species visited most often, data are presented for each penod separately PI, P2, P3 = Penod I, Penod 2 and Penod 3, respectively Available flowers N PI P2 P3 Total PI Centaurea bracteata P2 P3 Total Serratula tinctoria PI P2 P3 Total Total Betonica officinalis Buphthalmum salicifolium Total Scabiosa columbaria Total Leucanthemum vulgaris Total Polygala pedemontana Total Allium spp Total Prunella grandiflora Total Aster amellus Total Hippocrepis comosa Toul Hypochoeris maculata Total lx>tus comiculatus Total Total Knautia drymeia Gymnadema conopsea Total Asperula cvnanchica Total Globularia cordiifolia Total Peucedanum cervana Toul Centaurea scabiosa 50 229 119 398 8 177 29 214 M galathea 100 90 78 268 6 148 45 199 _ 19 40 59 9 7 2 1 _ 2 Flower visits by butterfly species C arcania L condon males 23 30 53 _ 1 3 4 _ 11 146 157 1 — 1 6 3 3 2 — - O venatus 70 53 12 135 2 14 2 18 _ 24 15 39 19 10 1 2 5 2 1 3 1 - L achtlteae 102 40 _ 142 _ _ _ 1 _ 1 6 4 1 2 2 — I 1 _ — - _ _ _ _ _ _ _ 2 2 1 — 1 _ _ I - were ob,servations of females, 401 forO venatus and 270 for L achilleae Because there were so few observations of L. cortdon females all the subsequent analyses were conducted only for the 624 observaUons of males of this species In penod 3 the flight penod of L achilleae had already ended In penod 1 only 29 observauons of L condon were made Therefore in Tables 1 and 3 no numbers appear for this species The five butterfly species displayed the following acuvity patterns In M galathea and L condon, flying predominated, followed by dnnking nectar then resung and basking In O venatus and L achtlleae, dnnking was observed most frequently, followed by flying, resting and basking C. arcanta was mainly encountered in flight, followed by resting, basking and only very rarely visiung flowers The five species differed considerably in the proportion of times they were observed visiUng flowers For O. venatus and L. achilleae about half of the observauons were made while nectanng (48% and 53% resp), whereas in M. galathea (26%) and L condon males (34%) this acuvity was less frequent. Finally just 0 2% of the observations of C. arcanta involved nectanng. In all species, oviposiUon and courtship events were observed only occasionally and the spatial distnbution of these acuviues can at best be used as a hint for explaining the nucrodistribuuon of the respeeuve spiecies In all five butterfly species, most of the basking observaUons were 18 made m the first two observation hours (9 am-11 am) Otherwise no eoncentrauon of particular acuviUes to a specific dayume was observed Nectar resource use and niche measures Out of an array of jxitential nectar sources the three flower species Centaurea scabtosa (L.), C bracteata (Scop ) and S tinctona were visited preferenually (Table 1). They showed different abundance patterns dunng the course of the study (Fig 2a) When visiung these three flowers species, the butterflies exhibited quite different patterns (Fig 2) O venatus showed no clear preferences and allocated their visits about in proporuon to the relaUve abundances of the three fiower species L cortdon males visited mainly C. scabtosa and S ttnctorta, using them in proporuon to their relaUve abundance M galathea paid most visits to the two Centaurea species. In period 2 however, when C bracteata tnpled its abundance relaUve to penod 1, M. galathea allocated nearly ten umes as many visits to this fiower species than in penod 1. In jjenod 3, when the abundance of C bracteata dropped to just 4%, M. galathea sull attnbuted 28% of its visits to that flower species. L achtlleae on the other hand clearly preferred C scabtosa independently of its relaUve abundance. ECOGRAPHY 18 1 (1995) Fig 2a Relative proportion of the three flower species most often visited by the butterflies in the three study penods b-e Proportion of visits by the butterflies attnbuted to these three flower species in each penod a) Preferred flowers b) LachBteae The preferences for the three main nectar sources are also reflected by the low niche breadth values obtained for all species in all penods (Table 2) In conu-ast, the niche overlap measures were high m almost all compansons (Table 2). In penod 1, M. galathea, O venatus and L achtlleae visited the same fiower sf>ecies in similar proportions as shown by the high values obtained with Hurlbert's measure of niche overlap The lower values for Hurlbert's measure of mche overlap in penod 2 between these three species are caused by the different allocaUon of visits to the three mam nectar sources (Fig 2) The lower niche overlap values obtained in the compansons of M galathea and O. venatus with L coridon 2* ECOGRAPHY 18 1 (1995) are mainly caused by the different uulization of C bracteata by both and of S. tmctorta by M. galathea only. Distribution patterns All prefened nectar sources were highly clumped (Table 3) The butterfiies vaned considerably in the degree of aggregaUon of observaUons C arcanta showed moderate clumping while L achilleae marks the other extreme with clearly clumped distnbution pattems (Table 3) There were pronounced differences in the degree of clumping in between penods for L coridon and O vena19 Table 2 Food niche relationships between the four butterfly species, calculated for each penod separately Niche overlap measures following Hurlbert (1978) above diagonal and Pianka (1974) below diagonal For each penod number of flower species visited and niche breadth values following Levins (1968) and Hurlbert (1978) are listed below overlap measures Levins stand = Levins/ Slower species Penod 1 overlap M galathea L. condon O venatus L achilleae Penod 1 breadth ZFlower speeies Levins Levins stand Hurlbert Penod 2 overlap M galathea L condon O venatus L achilleae Penod 2 breadth ZFIower species Levins Levins stand Hurlbert Penod 3 overlap M galathea L. cortdon O venatus L achilleae Penod 3 breadth ZFlower species Levins Levins stand Hurlbert M galathea 0000 0 959 0 995 15 1370 0 090 0 130 M galathea 0 545 0 700 0516 10 0 890 0 090 0210 M galathea L condon 0000 — . 0006 0.000 15 1800 0 120 0.270 L condon 1635 0 979 0 899 10 2 100 0 210 0410 L condon • O venatus 5 053 0000 ——_ 0 970 15 2 630 0 180 0 190 O venatus 2006 2 115 • — . L achilleae 8 610 0000 5 448 15 1380 0090 0 140 L achilleae 1 826 2014 1 896 0 885 10 2410 0 240 0 490 O venatus I 000 0 100 0 300 L achilleae tus Males of L condon showed a clearly more aggregated distnbuuon pattern in penod 3 than in penod 2 The opposite IS the case for O venatus, whose distnbuUon was more aggregated in penods 1 and 2 than in jjenod 3 These differences may be due to the relatively few observations of L condon in penod 2 and of O venatus in penod 3 The frequency of observaUons of each butterfiy species m the different quadrats of the study area is presented in Figs 3-7 In addiuon, proporuons of flower visits are also given ObservaUons of M galathea were concentrated on the northwestern side of the study site (top left in Fig 3), where vegetaUon type "Carex" and type "shrubs" form a small scale mosaic of high structural diversity The projrorUon of flower visits in the distnbuUon pattern does not follow any parucular trend. Males of L. cortdon were most often encountered in the area m the western comer of the gnd dominated by vegetaUon type "Carex" and the quadrats just above that area (Fig. 4). These quadrats also seem to play an important role in the acquisiUon of nectar, as large proporuons of the observations there were 20 fiower visits For C arcanta, no clear concentration of observations was found, even though all quadrats with more than nine observaUons were quadrats bordered by shrubs or trees (Fig 5) The observaUons of O. venatus were concentrated in the northern comer of the gnd, where vegetaUon types "Carex" and "shrubs" are responsible for a sUiicturally highly diverse habitat (Fig 6) However, O venatus was frequently observed in several quadrats in the eastern part, where proporuons of fiower visits were particularly high Finally, L achtlleae did not show a clear preference for a particular area in the study site However, in all quadrats with more than mne observations the proporuon of fiower visits was high (Fig 7) Correlations between butterfly observations and nectar sources and habitat partitioning for nectar acquisition The distnbuUon of the total number of observations of M galathea, L coridon, L achtlleae and O. venatus correECOGRAPHY 18 I (1995) Table 3 Spatial distnbution of the four butterfly species regularly visiUng flowers and of their preferred flower species N = number of open flower heads (Total only) and number of buUerfly observaUons (for each penod separately). Mean = mean number of flower heads or butterfly observations per quadrat, Var = vanance, ID = index of dispersion, d = test statisuc ID and d according to Ludwig and Reynolds (1988) N Mean Var ID d expressed by an increasing correlation value with C bracteata, while the value with C scabiosa remains high This IS again probably due to the correlation m the distnbution of these two plant species C, arcania showed no correlation with any nectar source, as might be expected from the very low projxjrtion of flower visits in the daily activity pattem of this species. Centaurea scabiosa Centaurea bracteata Serratula tinctorta Melanargia galathea Penod 1 Penod 2 Penod 3 Total 54 9 17 9 19,9 86 82 3 46 7 14 6 18 5 15,7 27 9 Discussion The four species regularly visiting flowers linuted their visits to basically three plant species with purple flowers of the family Asteraceae These observations fit well with observations of several authors (e g Jennersten 1984, Steffny et al 1984, SBN 1987, Ebert 1991, Erhardt and Thomas 1991) and the general butterfiy flower syndrome (Faegn and Van der Pijl 1971) The pattems of allocation of visits to these three plant species changed between observation penods With the exception of L achilleae, which clearly preferred C scabtosa, and M galathea which showed a preference for C bracteata when it became abundant, the butterflies visited the three flower species m proportion to their relative abundances The reasons for these differences are largely unknown One possible reason is indicated by the narrow ruche breadths and the changing high niche overlap values which raise the question whether the butterfiy species were comp>etmg for these resources The shift in the visiting pattems exhibited by M galathea and L condon between penod 2 and 3 and the subsequent decrease in niche overlap might lead to this conclusion A possible way to escape competition would be to partition the habitat for acquiring nectar between the species However our analysis indicated no such pattems of habitat partitioning Furthermore analyses of niche relationships are considered lnappropnate for detecting competition (Colwell and Futuyma 1971) It has been argued that preferences for particular nectar sources may represent a major factor for the microdistribution of butterflies in a given habitat (Boggs 1987) Posiuve correlations between the distnbuuon of butterfiies and their preferred nectar sources have been found in the Lycaenid Heodes virgaureae (Douwes 1975), the Satynd Maniola juntna (Brakefield 1982) and the Nymphalid Euphydryas chalcedona (Brown and Ehrlich 1980), Furthermore, Murphy et al (1984) showed that the distnbution of nectar sources lnfiuences oviposition sites in E chalcedona However, only Douwes (1975) quantified the relationship between nectar sources and butterfiies, although he used for his quantification higher systematic groups of plants, such as Ligulifiorae, Tubuliflorae and Dip,sacaceae He found that the distnbuUon of butterflies was strongly lnfiuenced by the distnbution of Ibbulifiorae dunng the whole flying penod of H. vtrgaureae Furthermore he found that the goodness of fit of the regression (R-) of the distribution of the male butterfiies 21 Coenanympha arcania Penod 1 Penod 2 Total Lysandra coridon males Penod 2 Penod 3 Total Ochlodes venatus Penod 1 Penod 2 Total Lictoria achilleae Penod 1 Period 2 Total lated with the distnbution of C scabiosa and the total number of flowers and, with one exception (O venatus), also with C bracteata (Table 3), Although L achilleae and L cortdon males visited the flowers of C. bracteata only occasionally (Table 1), they showed a correlation with this plant This is probably due to the correlation between the distributions of C scabiosa and C. bracteata (Rho = 0,25, p < 0,05) No habitat partitioning for nectar acquisition could be detected. On the other hand, only m one companson (L achtlleae and O venatus) was the correlation of nectanng observations positive When the correlations are examined for each penod separately, it becomes evident that L condon showed correlations with C scabtosa m penods 2 and 3 and with 5 tinctoria in penod 3, In penod 2 the correlation with S tinctorta was distinctly weaker This reflects the changing attnbution of visits to this flower between penod 2 and 3 (Fig, 2) Even though they are not statisically significant, the negative correlaaons with Buphthalmum saltctfolium ( L ) in all penods invites comment (Table 4) This plant grows in shaded areas around trees and shrubs dominated by vegetation typie "Molinia" (Guggisberg 1990), which are usually avoided by L coridon males The changing flower preference of M galathea from C, scabiosa in penod 1 to C bracteata in penod 2 (Fig 2) is ECOGRAPHY 18 I (1995) • 1-6 observations 7-14 observations ^P ^ B 15-24 observations >24 observations (9 O O 3 0% • 1-25% 26-50% 50-75% 75-100% Flower visits Fig 3 Distnbution of the pooled observations (N = 2027) of Melanargia galathea in the study area and the percentage of flower visits for each quadrat 1-2 observations 3-5 observations 6-9 observations >9 observations o o O (5 0% 1-25% 26-50% 51-75% >75% Rower visits Fig 4 Distnbution of the pooled observabons (N = 624) of males oi Lysandra condon in the study area and the percentage of flower visits for each quadrat. ECOGRAPHY 18 I (1995) For symbols see Rg 4 Fig 5 DistnbuUon of the pooled observations (N = 793) ot Coenonsmpha arcania in the study area and the percentage of flower visits for each quadrat with the fiower groups increased between his first and second study penod and remained constantly high between the second and third penod However, since his analysis was conducted with the taxon Tububfiorae, which includes several plant species, seasonal changes in the abundance of different species belonging to this taxon and subsequent changes of the fiower preferences of H virgaureae might have gone unnouced In four of the five species analysed in the present study (M galathea, L cortdon, O venatus and L achtlleae), the distnbution of nectar sources appears to be an important determinant of microdisuibution since they exhibited posiUve correlauons with fiower-nch quadrats and with their respecUve nectar sources In addition, seasonal changes in the allocaUon of visits of the butterfly species to their nectar sources were reflected by increasing or decreasing correlaUon values In C arcania on the other hand, nectar sources do not seem to play a major role m determining microdistnbution Instead the occurrence of isolated shrubs used as basking and perching sites (Ebert 1991, unpubl data) probably accounts for its clustered distnbuUon The quesUon anses whether vanaUon in nectar availabiLty of the different nectar plants could explain the different visitaUon pattems of the observed butterflies Although nectar production was not measured and nectar composiuon not analysed in the present study, it can be stated from numerous lnvestigaUons (e g Baker and Baker 1990 and references therein) that different plant species differ in amount and quality of their nectar In addiUon, microbial decomposiUon of sucrose to glucose and fmctose can significantly reduce the proporUon of sucrose in nectar of ageing flowers This could reduce the attracuveness of older flowers to butterflies, since bunerECOGRAPHY 18 I (1995) flies have a distinct preference for sucrose over glucose and fructose (Erhardt 1991, 1992) A novelty effect of fiowers which come into bloom may also attract butterfiies which may then become conditioned to the new nectar plant (e g the disproponionally large fracuon of Melanargia galathea visits to Centaurea bracteata in Penod 2 in the present study) Furthermore, climauc conditions might affect nectar production differently m different plant species An lnvesUgaUon correlating fiower visitaUon pattems of butterfiies with nectar production of preferred versus neglected nectar plants would shed more light to this subject Altemauve explanations for the observed distnbution pattems, such as the distnbuUon of larval foodplants (Brown and Ehrlich 1980) or mate locaung behavior (Rutowski 1991) should also be considered In field studies, such as mark-recapture-studies or transect counts, more males than females are usually observed (e g Erhardt 1985, Daily et al 1991, Loertscher 1991) Differences m the behavior of the sexes, such as a panicular male mate locating behavior, can also lead to differences in microhabitat choice (Brakefield 1982), Males of L Condon showed a clear preference for the vegetaUon tyjie "Carex" and its immediate surroundings TTus could be due not only to a concenu^uon of preferred nectar plants m this vegetauon type, but also to the larval foodplant H comosa, which was found exclusively m vegetaUon type "Carex" (Guggisberg 1990) Male butterfiies should have the highest chance of obtaining a virgin female in the immediate proximity of the female pupaUon sites If these are predictable, for example because larval foodplants are patchily disu-ibuted and larvae pupate near their foodplants, such areas should be preferenually visited by males (Rutowski 1991) These factors may therefore also 23 For symbols see Rg 4 Fig 6 Distnbution of the pooled observations (N = 401) of Ochlodes venatus in the study area and the percenUge of flower visits for edch quadrat determine the microdistnbution of Z, coridon males InL achilleae, which feeds on the same food plant as L condon the clumped distribution may also be caused partly by the same phenomenon descnbed above In conclusion, this study shows that beside the needs for larval resources and specific behaviours such as male mate locating behaviour, species-specific needs for adult resources, the spatial distnbuUon and the temporal varlauon of these resources play a cmcial role for the microdistnbution of butterfiies in the field Acknowledgements - We would like to thank the following persons for the continuous moral and intellectual support dunng this study A Stampfli, F Guggisberg, C Antognoli, A Valsangiacomo and M Zimmermann Constructive cnUcisms on earlier versions of the manuscnpt by M Zimmermann, A Porter and C Wiklund greatly helped to improve it Figure 1 was, drawn by M Kummer and I Scharer The study presented here was part of a multidisciplinary project on unfertilized meadows and pastures of the Canton Tessin managed by O Hegg of the Botanical Inst Univ of Berne and financially supponed by the SWISS N F Project 31-9096 87 and by grants from the governments of Switzerland (BUWAL) and the Cantone Ticino (Dipar- For symbols see Fig 4 Fig 7 DistnbuUon of the pooled observanons (N = 270) of Ltctorta achilleae in the study area and the percentage of flower visits for each quadrat 24 ECOGRAPHY 18 1 (1995) Table 4 Spearman rank correlaUon coefficients between number of butterfly observaUons and number of flower heads per quadrat CorrelaUon coefficients are given for compansons with the most often visited nectar plants and as a contrast with a flower that was only rarely visited (Buphthalmum salicifolium) Correlauons are significant at the Bonferroni-corrected probability level of (0 05/95 = 0 0005) and (0 01/95 = 0 0001) for overall significance levels of 0 05(*) and 0 01(**) over 95 simultaneous tests Spearman rank correlation coeffieienLs (corrected for ties) C scabtosa 0 43** 0 45** 0 34** 051** C bracteata 0 13 0 45** 5 ttnctona 0 11 B salicifolium 01 Flower Tolal 0 47** 0 44** 0 55** 0 55** 0 28* 0 35** 0 46** 0 45** 0 16 0 18 0 15 0 24 0 32** 0 35** 0 33** 0 36** 0 45** 0 25 0 46** Melanargta galathea Penod 1 Penod 2 Penod 3 Total Lysandra cortdon males Penod 1 Penod 2 Penod 3 Total Coenonympha arcanta Penod 1 Penod 2 Penod 3 Total Ochtodes venatus Penod I Penod 2 Penod 3 Total Lictoria achilleae Penod 1 Penod 2 Total 0 48** 0 15 0 05 -0 05 0 15 0 08 -0 -0 -0 -0 16 17 17 18 0 33** 0 28* 0 37** 0 05 0 23 0 07 0 14 0 5** 0 53** 0 22 0 5** 0 44** 031** 048** 0 23 0 II 0 32** 0 27* 0 26 0 07 -0 07 -0 04 0 08 0 03 -01 02 0 06 -0 02 0 16 0 26 -001 0 03 0 24 0 05 0 24 0 29* 0 18 0 17 0 03 -0 15 -0 06 0 18 0 24 -0 07 -0 02 -0 02 phitenor (Lepidoptera, Papilionidae) - Ecol Ent 16 425434 - 1992 Preferences and non-preferences for nectar constituents in Omithoptera prtamus posetdon (Lcpidoptera, Papihonidae) - Oecologia 90 581-585 - and Thomas, J A 1991 Lepidoptera as indicators of change in the seminatural grasslands of lowland and upland Europe References - In Collins, N M and Thomas, J A (eds). The conservation of insects and their habitats Academic Press, London, Baker, H G and Baker, I 1990 The predictive value of nectar pp 213-234 chemistry to the recognition of pollinator types - Israel J Bot 39 157-166 Faegn, K and Van der Pijl, L 1971 The pnnciples of pollination ecology - Pergamon Press. Oxford Boggs, C L 1987 Ecology of nectar and pollen feeding in Lcpidoptera - In Slansky, F and Rodnguez, J G (eds), Guggisberg, F 1990 Verbuschende Wiesen und Weiden auf dem Monte San Giorgio/TI M Sc lhesis, Univ of Berne Nutntional ecology of insects, mites, spiders and related invertebrates John Wiley and Sons, New York, pp 369- Hurlbert, S H 1978 The measurement of niche overlap and some relatives - Ecology 59 dl-ll 391 Brakefield, P M 1982 Ecological studies on the buUerfly Jennersten, O 1984 Flower visitanon and polhnation efficiency of some North European butterflies - Oecologia 63 80-89 Maniolajuntna inBnl&m I Adult behavior, microdistnbution and dispersal - J Anim Ecol 51 713-726 Levins, R 1968 Evolution in changing environments some theoreUcal explorauons - Monographs on population biolBrown, 1 L and Ehrlich, P R 1980 Population biology of the ogy, Pnnceton University Press, New Jersey Checkerspot Butterfly Euphydrtas chalcedona Smicture of Loertscher, M 1991 Population biology of two satynne butterthe Jasper Ridge Colony - Oecologia 47 239-251 flies Erebta meolans (de Prunner, 1798) and Erebia aethiops Col well, R K andFutuyma, D J 1971 On the measurement of (Esper, 1777) (Lepidoptera, Satyndae) - Nola Lep 2 niche breadth and overlap - Ecology 52 567-576 22-31 Daily, G C, Ehrhch, P R and Wheye, D 1991 Deteiroinants of spatial distnbuuon in a populauon of the subalpine butter- Ludwig, J A and Reynolds, J F 1988. StaUsUcal Ecology-A pnmer on methods and computing - John Wiley and Sons, fly Oenets chryxus - Oecologia 88 587-596 New York Douwes, P 1975 Distnbuuon of a population of the butterfly Heodes virgaureae - Oikos 26 332-340 Murphy, D D, Menninger, M S and Ehrlich, P R 1984 Nectar source distnbuuon as a determinant of oviposiUon Ebert, G 1991 Die Schmetterlinge Baden-Wurttembergs Band host species in Euphydnas chatcedona - Oecologia 62 2, Tagfalter 2 - Ulmer, Stuttgart 269-271 Erhardt, A. 1985 Wiesen und Brachen als Lebensraum fUr Pianka, E R 1974 Niche overlap and diffuse competition Schmetterlinge - Birkhkuser, Basel Proc Nat Acad Sci USA 71 2141-2145 - 1991 Nectar sugar and amino-acid preferences of Battus ECOGRAPHY IS 1 (1995) limento dell' Ambiente) as well as from the Lega Svizzera per la protezione della natura (Sezione Ticino) Rutowski, R L 1991 The evoluuon of male mate-locating behavior in butterflies - A m Nat 138 1121-1139 SBN 1987 Schweizenscher Bund fur Naturschutz Tagfaller und lhre Lebensraume - Basel, Scott, J A 1975 Flight patterns among eleven species of diumal Lepidoptera - Ecology 56, 1367-1377 Sharp, M A , Parks, D R and Ehrlich, P R 1974 Plant resources and butterfly habitat selecUon - Ecology 55 870875 Shreeve, T G 1992 Adult behaviour - In Dennis, R L H (ed). The ecology of butterflies in Bntain Oxford University Press, Oxford, pp 22-45 Stampfli, A , Loenscher, M and Guggisberg, F 1992 Artenruckgang in Magerwiesen - Wissenschaftbcher Naturschutz am Monte San Giorgio Gaia 2 105-109 Steffny, H , Kratochwil, A and Wolf, A 1984 Zur Bedeutung verschiedener Rasengesellschaften fur Schmetterlinge (Rhopalocera, Hespenidae, Zygaenidae) und Hummeln (Apidae, Bombus) im Naturschutzgebiet Taubergiessen (Oberrheinebene) - Natur Landsch 59 435-443 Zar, J H (1984) Biostatistical Analysis - PrenUce-Hall, Englewood Cliffs, NJ ECOGRAPHY 18 1 (1995)

Journal

Ecography – Wiley

Published: Mar 1, 1995

There are no references for this article.