Search

In this thesis, I combined controlled crosses with genetic and demographic data to examine how a main conservation value indicator, population size, was associated with estimates of population viability. I focused on the still common, but decreasing, perennial orchid Gymnadenia conopsea at two spatial scales—locally on the island Öland, in SE Sweden, and regionally across Scandinavia. I aimed to determine whether: (1) population size or density could predict the strength of heterosis and inbreeding depression, (2) local density variation within populations affected self-pollen transfer and relatedness of individuals, (3) genetic diversity increased with population size or density, and genetic differentiation and diversity predicted the strength of heterosis, (4) population growth rate increased with population size and genetic diversity, and (5) large-scale genetic structure indicated several independent colonization events in Scandinavia, with clear genetic groups and genetic diversity hotspots. I found heterosis to decrease and inbreeding depression to increase with local density in Öland populations. The proportion of deposited self-pollen decreased with local density, but established individuals located in sparse patches were not more related than those in dense ones, possibly due to high inbreeding depression at early life stages. Genetic diversity increased with population size, but neither population genetic differentiation or within population genetic diversity was related to the strength of heterosis or inbreeding depression. I found the majority of Öland populations to be declining, and population growth rate to increase with population size. This relationship was driven by higher survival in large populations, and most likely reflects that population size was positively associated with local habitat quality and stability. Population growth rate was not related to the amount of genetic diversity within the population, suggesting declines are not driven by

In this thesis, I combined controlled crosses with genetic and demographic data to examine how a main conservation value indicator, population size, was associated with estimates of population viability. I focused on the still common, but decreasing, perennial orchid Gymnadenia conopsea at two spatial scales—locally on the island Öland, in SE Sweden, and regionally across Scandinavia. I aimed to determine whether: (1) population size or density could predict the strength of heterosis and inbreeding depression, (2) local density variation within populations affected self-pollen transfer and relatedness of individuals, (3) genetic diversity increased with population size or density, and genetic differentiation and diversity predicted the strength of heterosis, (4) population growth rate increased with population size and genetic diversity, and (5) large-scale genetic structure indicated several independent colonization events in Scandinavia, with clear genetic groups and genetic diversity hotspots. I found heterosis to decrease and inbreeding depression to increase with local density in Öland populations. The proportion of deposited self-pollen decreased with local density, but established individuals located in sparse patches were not more related than those in dense ones, possibly due to high inbreeding depression at early life stages. Genetic diversity increased with population size, but neither population genetic differentiation or within population genetic diversity was related to the strength of heterosis or inbreeding depression. I found the majority of Öland populations to be declining, and population growth rate to increase with population size. This relationship was driven by higher survival in large populations, and most likely reflects that population size was positively associated with local habitat quality and stability. Population growth rate was not related to the amount of genetic diversity within the population, suggesting declines are not driven by

Gymnadenia conopsea se koristi kao lekovita biljka u tradicionalnoj, ali i savremenoj medicini istočnih zemalja poput Kine, Koreje i Japana. Uprkos širokoj rasprostranjenosti na teritoriji Evroazije i upotrebi u medicinske svrhe, slabo je istražena u pogledu usvajanja i sadržaja elemenata, kako esencijalnih tako i toksičnih. Shodno tome, u ovom radu je određen sadržaj sledećih elemenata: As, B, Ca, Cd, Co, Cr, Cu, Fe, K, Li, Mg, Mn, Ni, P, Pb, S i Zn, u uzorcima orhideja i zemljištima na kojem su rasle. Distribucija metala u zemljištu i njihova biodsostupnost je određena BCR sekvencijalnom ekstrakcijom. Utvrđeno je da elementi u zemljištu, ali i u biljkama, pokazuju specifičan način raspodele. Većina elemenata se dominatno nalazi u rezidualnoj frakciji zemljišta, koja nije biodostupna. Suprotono, Ca, B, Mn, Co i Pb, su nazastupljeniji u prve tri, biodostupne frakcije. Biljka zadržava većinu elemenata u korenu i tuberima (As, Cr, Pb, Co, Fe, Mn, Ni i Cu), dok su ostali (B, Zn, Ca, K, Mg, P, S i Li) u uporedivim ili višim koncentracijama prisutni u nadzemnim delovima biljke. Analizom biokoncentracionih i translokacionih faktora, utvrđeno je da ova vrsta orhideje akumulira B i Li, dok je u slučaju K primećena hiperakumulacija u svim delovima biljke. Ca se akumulira u korenu i listovima, a Zn i Cu u korenu i cvastima. Translokacija K, Li, P i manjoj meri Ca, se odvija iz korena u listove. Kod K je uočena i translokacija u stablo i cvasti. P se translocira u cvasti, za šta postoje naznake i kod B, Zn i Cu. Analizom glavnih komponenti je potvrđeno postojanje specifičnog profila akumulacije elemenata u biljnim delovima.

In this thesis, I combined controlled crosses with genetic and demographic data to examine how a main conservation value indicator, population size, was associated with estimates of population viability. I focused on the still common, but decreasing, perennial orchid Gymnadenia conopsea at two spatial scales—locally on the island Öland, in SE Sweden, and regionally across Scandinavia. I aimed to determine whether: (1) population size or density could predict the strength of heterosis and inbreeding depression, (2) local density variation within populations affected self-pollen transfer and relatedness of individuals, (3) genetic diversity increased with population size or density, and genetic differentiation and diversity predicted the strength of heterosis, (4) population growth rate increased with population size and genetic diversity, and (5) large-scale genetic structure indicated several independent colonization events in Scandinavia, with clear genetic groups and genetic diversity hotspots. I found heterosis to decrease and inbreeding depression to increase with local density in Öland populations. The proportion of deposited self-pollen decreased with local density, but established individuals located in sparse patches were not more related than those in dense ones, possibly due to high inbreeding depression at early life stages. Genetic diversity increased with population size, but neither population genetic differentiation or within population genetic diversity was related to the strength of heterosis or inbreeding depression. I found the majority of Öland populations to be declining, and population growth rate to increase with population size. This relationship was driven by higher survival in large populations, and most likely reflects that population size was positively associated with local habitat quality and stability. Population growth rate was not related to the amount of genetic diversity within the population, suggesting declines are not driven by

In this thesis, I combined controlled crosses with genetic and demographic data to examine how a main conservation value indicator, population size, was associated with estimates of population viability. I focused on the still common, but decreasing, perennial orchid Gymnadenia conopsea at two spatial scales—locally on the island Öland, in SE Sweden, and regionally across Scandinavia. I aimed to determine whether: (1) population size or density could predict the strength of heterosis and inbreeding depression, (2) local density variation within populations affected self-pollen transfer and relatedness of individuals, (3) genetic diversity increased with population size or density, and genetic differentiation and diversity predicted the strength of heterosis, (4) population growth rate increased with population size and genetic diversity, and (5) large-scale genetic structure indicated several independent colonization events in Scandinavia, with clear genetic groups and genetic diversity hotspots. I found heterosis to decrease and inbreeding depression to increase with local density in Öland populations. The proportion of deposited self-pollen decreased with local density, but established individuals located in sparse patches were not more related than those in dense ones, possibly due to high inbreeding depression at early life stages. Genetic diversity increased with population size, but neither population genetic differentiation or within population genetic diversity was related to the strength of heterosis or inbreeding depression. I found the majority of Öland populations to be declining, and population growth rate to increase with population size. This relationship was driven by higher survival in large populations, and most likely reflects that population size was positively associated with local habitat quality and stability. Population growth rate was not related to the amount of genetic diversity within the population, suggesting declines are not driven by

A wide majority of orchid populations are decreasing due to habitat fragmentation and to changes in land management. Population size, density and habitat quality are factors that are expected to be positively related to the viability and future growth of a population. We evaluated if population size, density and soil organic matter were good predictors of growth, survival, flowering, recruitment, and growth rate in 18 populations of the long-lived orchid Gymnadenia conopsea. We followed the populations for four years. Recruitment in 2020 increased with population size, and survival in 2018 was higher in denser populations. However, flowering probability and number of flowers both decreased with population size in 2018. Soil organic matter did not significantly influence any vital rate. In total, the studied population factors could explain very little of the variation in demography. The matrix modelling showed that 14 of the 18 populations had a positive stochastic growth rate, even with an increased probability of summer drought (scenario with 50% of the years equal to the dry summer of 2018). In the populations with negative growth rate, the probability of quasi-extinction in the next 50 years varied from 90 to 100%. Declining populations were characterized by low survival following the dry year. In sum, population size, density and soil organic matter did not convincingly explain variation in growth rate of G. conopsea, suggesting that other environmental factors are responsible of governing variation in vitals rates and population dynamics.

Reproductive isolation can be achieved through multiple types of barriers and is essential for speciation. In flowering plants, pre-pollination barriers (e.g. differentiation in pollination niches) are believed to be the most efficient at preventing gene flow across species boundaries. In closely related species that come into secondary contact, such barriers can evolve to prevent competition for pollinator service and/or interspecific pollen transfer, which can have fitness costs. Hence pollination niche differentiation should be stronger in sympatric populations than in allopatric populations (i.e. character displacement). To investigate the differences in pollination niches and to see if it is consistent with a hypothesis of character displacement, I used the two closely related and phenotypically similar orchid species, Gymnadenia conopsea and G. densiflora. I sampled mixed and pure populations of G. conopsea and G. densiflora on Öland during the summer of 2020. In these populations, I used video cameras and pollinator catches to record pollinator activity and characterize the composition of pollinator communities. Estimation of pollinator efficiency was also assessed by analyzing the number of pollinia carried by each pollinator. Contrary to my expectations, I found that both orchids had their visitation peak during the night and that the most frequent and efficient pollinators were Autographa gamma or Deilephila porcellus for both of them. Furthermore, no increased differentiation between the two species was found in mixed compared to pure populations. My results suggest that plant-pollinator interactions do not act as efficient pre-pollination barriers in these two orchid species, and that competition for pollinator service and through interspecific pollen transfer seem to be too weak to drive pollination niche partitioning.

A wide majority of orchid populations are decreasing due to habitat fragmentation and to changes in land management. Population size, density and habitat quality are factors that are expected to be positively related to the viability and future growth of a population. We evaluated if population size, density and soil organic matter were good predictors of growth, survival, flowering, recruitment, and growth rate in 18 populations of the long-lived orchid Gymnadenia conopsea. We followed the populations for four years. Recruitment in 2020 increased with population size, and survival in 2018 was higher in denser populations. However, flowering probability and number of flowers both decreased with population size in 2018. Soil organic matter did not significantly influence any vital rate. In total, the studied population factors could explain very little of the variation in demography. The matrix modelling showed that 14 of the 18 populations had a positive stochastic growth rate, even with an increased probability of summer drought (scenario with 50% of the years equal to the dry summer of 2018). In the populations with negative growth rate, the probability of quasi-extinction in the next 50 years varied from 90 to 100%. Declining populations were characterized by low survival following the dry year. In sum, population size, density and soil organic matter did not convincingly explain variation in growth rate of G. conopsea, suggesting that other environmental factors are responsible of governing variation in vitals rates and population dynamics.

Effective population size should be positively related to census size and density, and it is expected to influence the strength of genetic drift, inbreeding and response to selection, and thus the distribution of the genetic load across populations. We examined whether census population size and density predict the strength of inbreeding depression, heterosis and population mean fitness at the seed stage in the terrestrial orchid Gymnadenia conopsea by conducting controlled crosses (self, outcross within and between populations) in 20 populations of varying size (7-30,000 individuals) and density (1-12.8 individuals/m(2)). In the largest population, we also examined how local density affects the occurrence of self-pollination with a pollen staining experiment. The majority of populations expressed strong inbreeding depression at the seed stage (mean delta ID: min-max = 0.26: -0.53 to 0.51), consistent with a mainly outcrossing mating system and substantial genetic load. The effect of between-population crosses varied from strong outbreeding depression to heterosis (mean delta OD: min-max = 0.05: -0.22 to 0.92), indicating varying influence of drift and selection among populations. Census population size did not significantly predict the strength of inbreeding depression, heterosis or population mean fitness. However, inbreeding depression was positively and heterosis negatively correlated with population density. The proportion of self-massulae deposition was three times higher in sparse patches compared to dense ones (41% vs. 14%). Combined effects of density-dependent pollinator behaviour and limited seed dispersal may cause stronger genetic sub-structuring in sparse populations and reduce the strength of the correlation between census and effective population size. The results point to the importance of considering population density in addition to size when evaluating the distribution of recessive deleterious alleles across populations. Synthesis and application

Effective population size should be positively related to census size and density, and it is expected to influence the strength of genetic drift, inbreeding and response to selection, and thus the distribution of the genetic load across populations. We examined whether census population size and density predict the strength of inbreeding depression, heterosis and population mean fitness at the seed stage in the terrestrial orchid Gymnadenia conopsea by conducting controlled crosses (self, outcross within and between populations) in 20 populations of varying size (7-30,000 individuals) and density (1-12.8 individuals/m(2)). In the largest population, we also examined how local density affects the occurrence of self-pollination with a pollen staining experiment. The majority of populations expressed strong inbreeding depression at the seed stage (mean delta ID: min-max = 0.26: -0.53 to 0.51), consistent with a mainly outcrossing mating system and substantial genetic load. The effect of between-population crosses varied from strong outbreeding depression to heterosis (mean delta OD: min-max = 0.05: -0.22 to 0.92), indicating varying influence of drift and selection among populations. Census population size did not significantly predict the strength of inbreeding depression, heterosis or population mean fitness. However, inbreeding depression was positively and heterosis negatively correlated with population density. The proportion of self-massulae deposition was three times higher in sparse patches compared to dense ones (41% vs. 14%). Combined effects of density-dependent pollinator behaviour and limited seed dispersal may cause stronger genetic sub-structuring in sparse populations and reduce the strength of the correlation between census and effective population size. The results point to the importance of considering population density in addition to size when evaluating the distribution of recessive deleterious alleles across populations. Synthesis and application

Effective population size should be positively related to census size and density, and it is expected to influence the strength of genetic drift, inbreeding and response to selection, and thus the distribution of the genetic load across populations. We examined whether census population size and density predict the strength of inbreeding depression, heterosis and population mean fitness at the seed stage in the terrestrial orchid Gymnadenia conopsea by conducting controlled crosses (self, outcross within and between populations) in 20 populations of varying size (7-30,000 individuals) and density (1-12.8 individuals/m(2)). In the largest population, we also examined how local density affects the occurrence of self-pollination with a pollen staining experiment. The majority of populations expressed strong inbreeding depression at the seed stage (mean delta ID: min-max = 0.26: -0.53 to 0.51), consistent with a mainly outcrossing mating system and substantial genetic load. The effect of between-population crosses varied from strong outbreeding depression to heterosis (mean delta OD: min-max = 0.05: -0.22 to 0.92), indicating varying influence of drift and selection among populations. Census population size did not significantly predict the strength of inbreeding depression, heterosis or population mean fitness. However, inbreeding depression was positively and heterosis negatively correlated with population density. The proportion of self-massulae deposition was three times higher in sparse patches compared to dense ones (41% vs. 14%). Combined effects of density-dependent pollinator behaviour and limited seed dispersal may cause stronger genetic sub-structuring in sparse populations and reduce the strength of the correlation between census and effective population size. The results point to the importance of considering population density in addition to size when evaluating the distribution of recessive deleterious alleles across populations. Synthesis and application

Effective population size should be positively related to census size and density, and it is expected to influence the strength of genetic drift, inbreeding and response to selection, and thus the distribution of the genetic load across populations. We examined whether census population size and density predict the strength of inbreeding depression, heterosis and population mean fitness at the seed stage in the terrestrial orchid Gymnadenia conopsea by conducting controlled crosses (self, outcross within and between populations) in 20 populations of varying size (7-30,000 individuals) and density (1-12.8 individuals/m(2)). In the largest population, we also examined how local density affects the occurrence of self-pollination with a pollen staining experiment. The majority of populations expressed strong inbreeding depression at the seed stage (mean delta ID: min-max = 0.26: -0.53 to 0.51), consistent with a mainly outcrossing mating system and substantial genetic load. The effect of between-population crosses varied from strong outbreeding depression to heterosis (mean delta OD: min-max = 0.05: -0.22 to 0.92), indicating varying influence of drift and selection among populations. Census population size did not significantly predict the strength of inbreeding depression, heterosis or population mean fitness. However, inbreeding depression was positively and heterosis negatively correlated with population density. The proportion of self-massulae deposition was three times higher in sparse patches compared to dense ones (41% vs. 14%). Combined effects of density-dependent pollinator behaviour and limited seed dispersal may cause stronger genetic sub-structuring in sparse populations and reduce the strength of the correlation between census and effective population size. The results point to the importance of considering population density in addition to size when evaluating the distribution of recessive deleterious alleles across populations. Synthesis and application

Effective population size should be positively related to census size and density, and it is expected to influence the strength of genetic drift, inbreeding and response to selection, and thus the distribution of the genetic load across populations. We examined whether census population size and density predict the strength of inbreeding depression, heterosis and population mean fitness at the seed stage in the terrestrial orchid Gymnadenia conopsea by conducting controlled crosses (self, outcross within and between populations) in 20 populations of varying size (7-30,000 individuals) and density (1-12.8 individuals/m(2)). In the largest population, we also examined how local density affects the occurrence of self-pollination with a pollen staining experiment. The majority of populations expressed strong inbreeding depression at the seed stage (mean delta ID: min-max = 0.26: -0.53 to 0.51), consistent with a mainly outcrossing mating system and substantial genetic load. The effect of between-population crosses varied from strong outbreeding depression to heterosis (mean delta OD: min-max = 0.05: -0.22 to 0.92), indicating varying influence of drift and selection among populations. Census population size did not significantly predict the strength of inbreeding depression, heterosis or population mean fitness. However, inbreeding depression was positively and heterosis negatively correlated with population density. The proportion of self-massulae deposition was three times higher in sparse patches compared to dense ones (41% vs. 14%). Combined effects of density-dependent pollinator behaviour and limited seed dispersal may cause stronger genetic sub-structuring in sparse populations and reduce the strength of the correlation between census and effective population size. The results point to the importance of considering population density in addition to size when evaluating the distribution of recessive deleterious alleles across populations. Synthesis and application

Effective population size should be positively related to census size and density, and it is expected to influence the strength of genetic drift, inbreeding and response to selection, and thus the distribution of the genetic load across populations. We examined whether census population size and density predict the strength of inbreeding depression, heterosis and population mean fitness at the seed stage in the terrestrial orchid Gymnadenia conopsea by conducting controlled crosses (self, outcross within and between populations) in 20 populations of varying size (7-30,000 individuals) and density (1-12.8 individuals/m(2)). In the largest population, we also examined how local density affects the occurrence of self-pollination with a pollen staining experiment. The majority of populations expressed strong inbreeding depression at the seed stage (mean delta ID: min-max = 0.26: -0.53 to 0.51), consistent with a mainly outcrossing mating system and substantial genetic load. The effect of between-population crosses varied from strong outbreeding depression to heterosis (mean delta OD: min-max = 0.05: -0.22 to 0.92), indicating varying influence of drift and selection among populations. Census population size did not significantly predict the strength of inbreeding depression, heterosis or population mean fitness. However, inbreeding depression was positively and heterosis negatively correlated with population density. The proportion of self-massulae deposition was three times higher in sparse patches compared to dense ones (41% vs. 14%). Combined effects of density-dependent pollinator behaviour and limited seed dispersal may cause stronger genetic sub-structuring in sparse populations and reduce the strength of the correlation between census and effective population size. The results point to the importance of considering population density in addition to size when evaluating the distribution of recessive deleterious alleles across populations. Synthesis and application

Why are flowers so diverse? Much of floral evolution is thought to be driven by pollinator-mediated selection. However, the connection between macroevolutionary patterns of floral diversity and microevolutionary processes remains poorly understood. In this thesis, I have used the fragrant orchids Gymnadenia conopsea s.s. and Gymnadenia densiflora to investigate the role of pollinators as agents of selection on floral traits and to test whether they cause spatial variation in selection. I addressed the following questions (1) Is there divergent selection on flowering phenology and floral traits between these two closely related species? (2) What is the contribution of pollinators relative to other selective agents to selection on phenology, visual display, floral scent and spur length? (3) Do diurnal and nocturnal pollinators mediate different selection patterns? (4) Does spatial variation in pollinator communities cause spatial variation in selection? A phenotypic selection study in G. conopsea s.s. and G. densiflora indicated that divergent selection on flowering time contributes to the maintenance of phenological differentiation between the two species. Hand-pollination experiments combined with selection analysis showed that while pollinators were the main selective agent on spur length, their contribution to selection on phenology, visual display and floral scent was more variable and sometimes opposed by non-pollinator mediated selection. Selection analyses combined with a selective exclusion experiment showed that diurnal and nocturnal pollinators exerted different selection patterns on floral traits. Hand-pollination experiments also demonstrated that variation in pollinator-mediated selection largely explained spatial variation in net selection on phenology, visual display and spur length among four populations. A study of floral scent emission of G. conopsea s.s. in the field coupled with a growth-chamber experiment revealed genetically-based variation in f

Why are flowers so diverse? Much of floral evolution is thought to be driven by pollinator-mediated selection. However, the connection between macroevolutionary patterns of floral diversity and microevolutionary processes remains poorly understood. In this thesis, I have used the fragrant orchids Gymnadenia conopsea s.s. and Gymnadenia densiflora to investigate the role of pollinators as agents of selection on floral traits and to test whether they cause spatial variation in selection. I addressed the following questions (1) Is there divergent selection on flowering phenology and floral traits between these two closely related species? (2) What is the contribution of pollinators relative to other selective agents to selection on phenology, visual display, floral scent and spur length? (3) Do diurnal and nocturnal pollinators mediate different selection patterns? (4) Does spatial variation in pollinator communities cause spatial variation in selection? A phenotypic selection study in G. conopsea s.s. and G. densiflora indicated that divergent selection on flowering time contributes to the maintenance of phenological differentiation between the two species. Hand-pollination experiments combined with selection analysis showed that while pollinators were the main selective agent on spur length, their contribution to selection on phenology, visual display and floral scent was more variable and sometimes opposed by non-pollinator mediated selection. Selection analyses combined with a selective exclusion experiment showed that diurnal and nocturnal pollinators exerted different selection patterns on floral traits. Hand-pollination experiments also demonstrated that variation in pollinator-mediated selection largely explained spatial variation in net selection on phenology, visual display and spur length among four populations. A study of floral scent emission of G. conopsea s.s. in the field coupled with a growth-chamber experiment revealed genetically-based variation in f

Why are flowers so diverse? Much of floral evolution is thought to be driven by pollinator-mediated selection. However, the connection between macroevolutionary patterns of floral diversity and microevolutionary processes remains poorly understood. In this thesis, I have used the fragrant orchids Gymnadenia conopsea s.s. and Gymnadenia densiflora to investigate the role of pollinators as agents of selection on floral traits and to test whether they cause spatial variation in selection. I addressed the following questions (1) Is there divergent selection on flowering phenology and floral traits between these two closely related species? (2) What is the contribution of pollinators relative to other selective agents to selection on phenology, visual display, floral scent and spur length? (3) Do diurnal and nocturnal pollinators mediate different selection patterns? (4) Does spatial variation in pollinator communities cause spatial variation in selection? A phenotypic selection study in G. conopsea s.s. and G. densiflora indicated that divergent selection on flowering time contributes to the maintenance of phenological differentiation between the two species. Hand-pollination experiments combined with selection analysis showed that while pollinators were the main selective agent on spur length, their contribution to selection on phenology, visual display and floral scent was more variable and sometimes opposed by non-pollinator mediated selection. Selection analyses combined with a selective exclusion experiment showed that diurnal and nocturnal pollinators exerted different selection patterns on floral traits. Hand-pollination experiments also demonstrated that variation in pollinator-mediated selection largely explained spatial variation in net selection on phenology, visual display and spur length among four populations. A study of floral scent emission of G. conopsea s.s. in the field coupled with a growth-chamber experiment revealed genetically-based variation in f

Why are flowers so diverse? Much of floral evolution is thought to be driven by pollinator-mediated selection. However, the connection between macroevolutionary patterns of floral diversity and microevolutionary processes remains poorly understood. In this thesis, I have used the fragrant orchids Gymnadenia conopsea s.s. and Gymnadenia densiflora to investigate the role of pollinators as agents of selection on floral traits and to test whether they cause spatial variation in selection. I addressed the following questions (1) Is there divergent selection on flowering phenology and floral traits between these two closely related species? (2) What is the contribution of pollinators relative to other selective agents to selection on phenology, visual display, floral scent and spur length? (3) Do diurnal and nocturnal pollinators mediate different selection patterns? (4) Does spatial variation in pollinator communities cause spatial variation in selection? A phenotypic selection study in G. conopsea s.s. and G. densiflora indicated that divergent selection on flowering time contributes to the maintenance of phenological differentiation between the two species. Hand-pollination experiments combined with selection analysis showed that while pollinators were the main selective agent on spur length, their contribution to selection on phenology, visual display and floral scent was more variable and sometimes opposed by non-pollinator mediated selection. Selection analyses combined with a selective exclusion experiment showed that diurnal and nocturnal pollinators exerted different selection patterns on floral traits. Hand-pollination experiments also demonstrated that variation in pollinator-mediated selection largely explained spatial variation in net selection on phenology, visual display and spur length among four populations. A study of floral scent emission of G. conopsea s.s. in the field coupled with a growth-chamber experiment revealed genetically-based variation in f

Why are flowers so diverse? Much of floral evolution is thought to be driven by pollinator-mediated selection. However, the connection between macroevolutionary patterns of floral diversity and microevolutionary processes remains poorly understood. In this thesis, I have used the fragrant orchids Gymnadenia conopsea s.s. and Gymnadenia densiflora to investigate the role of pollinators as agents of selection on floral traits and to test whether they cause spatial variation in selection. I addressed the following questions (1) Is there divergent selection on flowering phenology and floral traits between these two closely related species? (2) What is the contribution of pollinators relative to other selective agents to selection on phenology, visual display, floral scent and spur length? (3) Do diurnal and nocturnal pollinators mediate different selection patterns? (4) Does spatial variation in pollinator communities cause spatial variation in selection? A phenotypic selection study in G. conopsea s.s. and G. densiflora indicated that divergent selection on flowering time contributes to the maintenance of phenological differentiation between the two species. Hand-pollination experiments combined with selection analysis showed that while pollinators were the main selective agent on spur length, their contribution to selection on phenology, visual display and floral scent was more variable and sometimes opposed by non-pollinator mediated selection. Selection analyses combined with a selective exclusion experiment showed that diurnal and nocturnal pollinators exerted different selection patterns on floral traits. Hand-pollination experiments also demonstrated that variation in pollinator-mediated selection largely explained spatial variation in net selection on phenology, visual display and spur length among four populations. A study of floral scent emission of G. conopsea s.s. in the field coupled with a growth-chamber experiment revealed genetically-based variation in f

The link between biotic interaction intensity and strength of selection is of fundamental interest for understanding biotically driven diversification and predicting the consequences of environmental change. The strength of selection resulting from biotic interactions is determined by the strength of the interaction and by the covariance between fitness and the trait under selection. When the relationship between trait and absolute fitness is constant, selection strength should be a direct function of mean population interaction intensity. To test this prediction, we excluded pollinators for intervals of different length to induce five levels of pollination intensity within a single plant population. Pollen limitation (PL) increased from 0 to 0.77 across treatments, accompanied by a fivefold increase in the opportunity for selection. Trait-fitness covariance declined with PL for number of flowers, but varied little for other traits. Pollinator-mediated selection on plant height, corolla size, and spur length increased by 91%, 34%, and 330%, respectively, in the most severely pollen-limited treatment compared to open-pollinated plants. The results indicate that realized biotic selection can be predicted from mean population interaction intensity when variation in trait-fitness covariance is limited, and that declines in pollination intensity will strongly increase selection on traits involved in the interaction.

Costs of reproduction are expected to vary with environmental conditions thus influencing selection on life-history traits. Yet, the effects of habitat conditions and climate on trade-offs among fitness components remain poorly understood. For 2-5years, we quantified costs of experimentally increased reproduction in two populations (coastal long-season vs. inland short-season) of two long-lived orchids that differ in natural reproductive effort (RE; 30 vs. 75% fruit set). In both species, survival costs were found only at the short-season site, whereas growth and fecundity costs were evident at both sites, and both survival and fecundity costs declined with increasing growing season length and/or summer temperature. The results suggest that the expression of costs of reproduction depend on the local climate, and that climate warming could result in selection favouring increased RE in both study species.

Mutualists and antagonists are known to respond to similar floral cues, and may thus cause opposing selection on floral traits. However, we lack a quantitative understanding of their independent and interactive effects. In a population of the orchid Gymnadenia conopsea, we manipulated the intensity of pollination and herbivory in a factorial design to examine whether both interactions influence selection on flowering phenology, floral display, and morphology. Supplemental hand-pollination increased female fitness by 31% and one-quarter of all plants were damaged by herbivores. Both interactions contributed to selection. Pollinators mediated selection for later flowering and herbivores for earlier flowering, while both selected for longer spurs. The strength of selection was similar for both agents, and their effects were additive. As a consequence, there was no net selection on phenology, whereas selection on spur length was strong. The experimental results demonstrate that both pollinators and herbivores can markedly influence the strength of selection on flowering phenology and floral morphology, and cause both conflicting and reinforcing selection. They also indicate that the direction of selection on phenology will vary with the relative intensity of the mutualistic and antagonistic interaction, potentially resulting in both temporal and among-population variation in optimal flowering time.

Spatial variation in pollinator-mediated selection (Delta beta(poll)) is a major driver of floral diversification, but we lack a quantitative understanding of its link to pollen limitation (PL) and net selection on floral traits. For 2-5 years, we quantified Delta beta(poll) on floral traits in two populations each of two orchid species differing in PL. In both species, spatiotemporal variation in Delta beta(poll) explained much of the variation in net selection. Selection was consistently stronger and the proportion that was pollinator-mediated was higher in the severely pollen-limited deceptive species than in the rewarding species. Within species, variation in PL could not explain variation in Delta beta(poll) for any trait, indicating that factors influencing the functional relationship between trait variation and pollination success govern a major part of the observed variation in Delta beta(poll). Separating the effects of variation in mean interaction intensity and in the functional significance of traits will be necessary to understand spatiotemporal variation in selection exerted by the biotic environment.

Most plants attract multiple flower visitors that may vary widely in their effectiveness as pollinators. Floral evolution is expected to reflect interactions with the most important pollinators, but few studies have quantified the contribution of different pollinators to current selection on floral traits. To compare selection mediated by diurnal and nocturnal pollinators on floral display and spur length in the rewarding orchid Gymnadenia conopsea, we manipulated the environment by conducting supplemental hand-pollinations and selective pollinator exclusions in two populations in central Norway. In both populations, the exclusion of diurnal pollinators significantly reduced seed production compared to open pollination, whereas the exclusion of nocturnal pollinators did not. There was significant selection on traits expected to influence pollinator attraction and pollination efficiency in both the diurnal and nocturnal pollination treatment. The relative strength of selection among plants exposed to diurnal and nocturnal visitors varied among traits and populations, but the direction of selection was consistent. The results suggest that diurnal pollinators are more important than nocturnal pollinators for seed production in the study populations, but that both categories contribute to selection on floral morphology. The study illustrates how experimental manipulations can link specific categories of pollinators to observed selection on floral traits, and thus improve our understanding of how species interactions shape patterns of selection.

The fragrant orchid Gymnadenia conopsea s.l. is a controversial taxon with two commonly distinguished species, G. conopsea s.str. and G. densiflora. Despite morphological similarity, differentiation between the taxa has been reported for several characters; however, character variation within taxa has obviated a clear consensus. We assessed ITS sequences, microsatellite variation and chromosome numbers on the European scale (1,420 samples) and conducted morphological analyses for 626 samples from Germany. ITS analysis revealed a 2% nucleotide divergence between the taxa, similar to the divergence between other Gymnadenia species. The ITS sequences of G. densiflora form a well-supported monophyletic group sharing a most recent common ancestor with G. nigra and G. austriaca. Thus, G. conopsea and G. densiflora are not sister species, and a species rank is supported for G. densiflora (Wahlenb.) Dietrich and G. conopsea (L.) R.Br. s.str. This was confirmed by the microsatellite analysis, which revealed a strong genetic differentiation between the taxa because of largely non-overlapping sets of alleles. Chromosome numbers showed that G. conopsea was either diploid or tetraploid, whereas G. densiflora was diploid throughout. Morphologically, the taxa differed significantly in the mean value of a number of diagnostic characters. However, a discriminant analysis showed that the morphological variability is substantial, and on the individual level an unequivocal assignment is not possible as 96% of G. conopsea, but only 77% of G. densiflora could be assigned correctly. Further studies are needed on character variation within and among species and ploidy levels to allow for a better identification of the genetically differentiated but morphologically similar taxa.

The fungal community associated with the terrestrial photosynthetic orchid Gymnadenia conopsea was characterized through PCR-amplification directly from root extracted DNA and cloning of the PCR products. Six populations in two geographically distinct regions in Germany were investigated. New ITS-primers amplifying a wide taxonomic range including Basidiomycetes and Ascomycetes revealed a high taxonomic and ecological diversity of fungal associates, including typical orchid mycorrhizas of the Tulasnellaceae and Ceratobasidiaceae as well as several ectomycorrhizal taxa of the Pezizales. The wide spectrum of potential mycorrhizal partners may contribute to this orchid's ability to colonize different habitat types with their characteristic microbial communities. The fungal community of G. conopsea showed a clear spatial structure. With 43 % shared taxa the species composition of the two regions showed only little overlap. Regardless of regions, populations were highly variable concerning taxon richness, varying between 5 and 14 taxa per population. The spatial structure and the continuous presence of mycorrhizal taxa on the one hand and the low specificity towards certain fungal taxa on the other hand suggest that the fungal community associated with G. conopsea is determined by multiple factors. In this context, germination as well as pronounced morphological and genetic differentiation within G. conopsea deserve attention as potential factors affecting the composition of the fungal community.

Assisted gene flow can restore genetic diversity when genetic drift has driven deleterious alleles to high frequencies in small, isolated populations. Previous crosses among 20 populations of Gymnadenia conopsea documented the strongest heterosis and the weakest inbreeding depression in sparse and small populations, consistent with fixation of mildly deleterious alleles by genetic drift. We genotyped the populations used for crosses for 1,728 SNPs and tested the following predictions: 1) heterosis increases with genetic differentiation (FST) to donor populations and decreases with genetic diversity in the recipient population, 2) inbreeding depression increases with genetic diversity, and 3) genetic diversity increases, and mean FST to other populations decreases, with population size and density. Pairwise FST ranged from very low to moderate (0.005-0.20) and genetic diversity varied moderately among populations (proportion of polymorphic loci=0.52-0.75). However, neither FST between populations, nor genetic diversity in the recipient population, was related to the strength of heterosis. There was also no association between genetic diversity and the strength of inbreeding depression. Genetic diversity increased and mean FST decreased with population size, consistent with reduced diversity and increased differentiation of small populations by genetic drift. The results indicate that the loci conferring heterosis are not mirrored by overall population differentiation, and limited additional information on potential source populations for genetic rescue is gained by the genetic data. Instead, controlled crosses can directly reveal positive effects of introducing new genetic material, and is a simple method with high potential in conservation.

Population size is a main indicator of conservation value, thought to indicate both current and long-term population viability. However, few studies have directly examined the links between the size and the genetic and demographic properties of populations using metrics that integrate effects across the whole life cycle. In this study, we combine six years of demographic data with SNP-based estimates of genetic diversity from 18 Swedish populations of the orchid Gymnadenia conopsea, a species that is becoming increasingly rare in Europe. We investigated how population growth rate varied among populations and years, and used LTRE analysis to evaluate how underlying vital rates contribute to the variation. We examined if stochastic growth rate increased with population size and genetic diversity. For each population, we estimated the probability of quasi-extinction and of a 90% decrease in population size, within the next 30 years. Deterministic growth rate varied among populations and years, and the LTRE decomposition suggested that variation in both fecundity, and survival and growth of adults, contributed to this variation. Estimates of stochastic growth rate indicated that ten populations are declining, seven increasing, and one population approximately stable. Stochastic growth rate increased with population size, mainly due to higher survival in larger populations, but was not related to genetic diversity. One third of the populations were predicted to go extinct and eight populations to undergo a 90% decrease in population size in the coming 30 years. The results suggest that ongoing declines are not due to genetic erosion. Low survival in small populations most likely reflects a positive association between local environmental conditions and population size. More generally, the results suggest that demographic metrics can be more informative than genetic metrics, regarding conservation priority.