Tree Physiology 34, 1376–1387 doi:10.1093/treephys/tpu097
Research paper
Adaptive consequences of human-mediated introgression for indigenous tree species: the case of a relict Pinus pinaster population
1Department
of Forest Ecology and Genetics, INIA-CIFOR, Carretera de La Coruña km 7.5, E-28040 Madrid, Spain; 2Present address: Department of Ecology, Evolution and Behavior, University of Minnesota, 1987 Upper Buford Circle, Saint Paul, MN 55108, USA; 3Corresponding author ([email protected])
Received January 15, 2014; accepted October 15, 2014; published online December 1, 2014; handling Editor João Pereira
Human-induced gene movement via afforestation and restoration programs is a widespread phenomenon throughout the world. However, its effects on the genetic composition of native populations have received relatively little attention, particularly in forest trees. Here, we examine to what extent gene flow from allochthonous plantations of Pinus pinaster Aiton impacts offspring performance in a neighboring relict natural population and discuss the potential consequences for the long-term genetic composition of the latter. Specifically, we conducted a greenhouse experiment involving two contrasting watering treatments to test for differences in a set of functional traits and mortality rates between P. pinaster progenies from three different parental origins: (i) local native parents, (ii) exotic parents and (iii) intercrosses between local mothers and exotic fathers (intraspecific hybrids). Our results showed differences among crosses in cumulative mortality over time: seedlings of exotic parents exhibited the lowest mortality rates and seedlings of local origin the highest, while intraspecific hybrids exhibited an intermediate response. Linear regressions showed that seedlings with higher water-use efficiency (WUE, δ13C) were more likely to survive under drought stress, consistent with previous findings suggesting that WUE has an important role under dry conditions in this species. However, differences in mortality among crosses were only partially explained by WUE. Other non-measured traits and factors such as inbreeding depression in the relict population are more likely to explain the lower performance of native progenies. Overall, our results indicated that intraspecific hybrids and exotic individuals are more likely to survive under stressful conditions than local native individuals, at least during the first year of development. Since summer drought is the most important demographic and selective filter affecting tree establishment in Mediterranean ecosystems, a potential early selective advantage of exotic and hybrid genotypes would enhance initial steps of introgression of non-native genes into the study relict population of P. pinaster. Keywords: conservation genetics, drought avoidance, drought tolerance, gene flow, natural selection, plantations, translocations, water-use efficiency.
Introduction Introduced exotic species are often considered one of the greatest threats to conservation of natural ecosystems (Vitousek et al. 1997, Lee 2002). Their impacts on biodiversity have been demonstrated and the extinction of some
native species has been attributed to processes such as hybridization and competition with exotic species (Ellstrand 1992, McMillan and Wilcove 1994, Rhymer and Simberloff 1996, Vila et al. 2000, Mooney and Cleland 2001, Wolf et al. 2001, Petit 2004).
© The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://treephys.oxfordjournals.org/ at University of Tokyo Library on May 20, 2015
José Alberto Ramírez-Valiente1,2,3 and Juan José Robledo-Arnuncio1
Adaptive consequences of human-mediated introgression for indigenous tree species 1377 this study to what extent gene flow from exotic plantations of maritime pine (Pinus pinaster Aiton) impacts offspring performance in a neighboring natural relict population. Pinus pinaster is a keystone tree species widely distributed in the Mediterranean basin. It shows high intraspecific genetic differentiation at both neutral markers (Vendramin et al. 1998, Salvador et al. 2000, Gonzalez-Martinez et al. 2002, González-Martínez et al. 2004) and functional traits related to growth, reproduction, drought and freezing temperatures (Alía et al. 1995, Correia et al. 2008, Aranda et al. 2010, Corcuera et al. 2011a, 2011b, Santos-del-Blanco et al. 2012, Marguerit et al. 2014). It has been cultivated in many areas of the globe for commercial and ecological purposes (Richardson 1998), both within and out of its natural distribution range. Recent studies have shown that the species has become one of the most widespread invasive alien tree species on Earth (Nuñez and Medley 2011, Richardson and Rejmánek 2011). Within its natural range, maritime pine has a rather fragmented distribution, with several small isolated populations. In Spain, the species represents over 10% of the forest cover, with a total occupied area of 1,324,000 ha, 45% of which are plantations mainly established during the 20th century (De Lucas et al. 2008). Seeds used for maritime pine plantations in Spain have traditionally been collected from a limited number of (most frequently allochthonous) populations and individuals, selected for phenotypic traits such as vigor, growth architecture (e.g., apical dominance) and seed cone production and size. Ecologically, maritime pine has a drought avoidance strategy based on the high sensitivity of stomata closure to reductions of water availability (Picon et al. 1996, Fernández et al. 2000). This mechanism allows maritime pine to decrease water loss, increase water-use efficiency (WUE) and, ultimately, survive under severe xeric conditions (de Miguel et al. 2012). Nevertheless, drought tolerance mechanisms such as osmotic adjustment have also been reported to have an important role in adaptation to drought for this species (Nguyen-Queyrens et al. 2002, Nguyen-Queyrens and Bouchet-Lannat 2003). In contrast, the importance of below-ground biomass allocation is less clear for maritime pine. Triboulot et al. (1995) found that water stress stimulated root growth, but only under moderate conditions and for a short-term period (12 days). Aranda et al. (2010) reported differences in root biomass allocation among populations from different climates, but water treatment did not affect this trait. In this study, we set a greenhouse experiment to test for differences in WUE, biomass partitioning, growth rate and mortality under drought conditions between P. pinaster progenies from three parental origins: (i) a small natural relict population growing in ecologically marginal dry conditions, (ii) allochthonous plantations of undocumented origin surrounding the natural stand and (iii) intercrosses between native and allochthonous parents (intraspecific hybrids).
Tree Physiology Online at http://www.treephys.oxfordjournals.org
Downloaded from http://treephys.oxfordjournals.org/ at University of Tokyo Library on May 20, 2015
In plants, the impacts of human-induced gene movement have received considerable attention in recent years in the context of genetically modified crop species (Hoffman 1990, Raybould and Gray 1994, Pilson and Prendeville 2004, Chapman and Burke 2006) and invasive weeds (Small 1984, Van Raamsdonk and Van Der Maesen 1996, Snow et al. 2001). However, population genetic effects of intraspecific translocations have received much less attention, particularly in forest trees (Laikre et al. 2010; but see Steinitz et al. 2012 and references therein). Recent studies have reported that non-local exotic plantations have a significant effect on the genetic composition of the offspring of nearby conspecific populations at neutral loci (Sampson and Byrne 2008, Robledo-Arnuncio et al. 2009, Steinitz et al. 2012). Specifically, Sampson and Byrne (2008) and Robledo-Arnuncio et al. (2009) found that pollen introgression rates from neighboring plantations reached values over 40% among seeds of natural populations. The long-term consequences of intraspecific hybridization for the genetic composition of the local population depend on the level of gene flow and the relative fitness of hybrids compared with pure native genotypes. For example, if F1 hybrids and successive crosses and backcrosses exhibit increased fitness relative to pure local genotypes (heterosis caused by dominant and/or overdominant effects), then hybridization may, throughout generations, lead to the ‘genetic dilution’ and even extinction of the native gene pool (Rhymer and Simberloff 1996, Allendorf et al. 2001, Potts et al. 2003, Barbour et al. 2005). On the contrary, F1 hybrids and later generation backcrosses can suffer reduced fitness as a result of outbreeding depression or maladaptation of the alien gene pool to local environmental conditions. Under the latter scenario, intraspecific hybridization would have a limited long-term impact on the genetic composition of the native population, unless a very high level of gene flow overrides selection, ‘swamping’ the local gene pool (Lenormand 2002). Allochthonous genotypes can also invade native populations in the absence of intraspecific hybridization via seed dispersal from plantations, especially if introduced genotypes have either a numerical or fitness advantage over local genotypes (Hufford and Mazer 2003). Neutral marker studies have contributed greatly to estimate the exposure of natural populations to allochthonous genes from translocated populations (Robledo-Arnuncio et al. 2009, Steinitz et al. 2012, Unger et al. 2014). However, they have limited potential to assess the adaptive consequences of such human-induced gene flow, and in particular the potential risk it poses for the long-term viability of small indigenous populations (Kramer and Havens 2009). Besides knowing the migration rates between populations, a first step (feasible for long-living species) toward understanding such adaptive consequences involves the comparative assessment of fitnessrelated traits of F1 hybrids and pure-bred lines from parental populations (native and exotic). Along this line, we examine in
1378 Ramírez-Valiente and Robledo-Arnuncio potential negative impact of allochthonous gene flow on the genetic composition of the natural stand (Fenster and Galloway 2000, Hufford and Mazer 2003).
Materials and methods Study site We studied a small isolated relict population of maritime pine, located in the southern innerlands of Spain, 1000–1200 m above sea level in the municipality of Fuencaliente (38°25′N, 4°15′W; Figure 1). The entire population consists of ∼200 maritime pine adult trees (Alía et al. 1996). The climate of the area is Mediterranean with continental influence. Average annual precipitation is 833 mm and average annual temperature is 13.4 °C (Spanish Ministry of Agriculture, Food and Environment). Dry season extends over 3 months and summer precipitation reaches one of the lowest values for maritime pine in the Iberian Peninsula (58 mm). Moreover, the native stand grows on a steep crag of quartzite, with poor water retention capacity. The nearest native conspecific populations are located over 100 km east of the study site (at Sierra de SeguraAlcaraz). However, large afforestations were carried out in this area under plantation programs that began in the 1940s (DíazFernández 1994). As a result, the small native population is currently surrounded by widespread maritime pine plantations of undocumented origin (Figure 1). Neutral genetic differentiation between local and planted trees is significant both at nuclear and chloroplast microsatellite loci (Fst = 0.08–0.10), confirming the allochthonous origin of the plantations (Unger et al. 2014). Preliminary results with functional candidate gene SNP markers for biotic and abiotic stress including drought response, wood properties, phenology and growth (markers described in Budde et al. 2014) also suggest some degree of adaptive differentiation between native and planted stands (our unpublished data). Genetic cluster analysis using the same SNP markers indicates that the Fuencaliente native relict is genetically distinct from the plantations and all the studied native Iberian populations of the species, and it also suggests that native populations from southeastern Spain are the most likely origin of the plantations (S.C. González-Martínez,
Figure 1. Location of the study site (left) and distribution of native (light gray) and planted (dark gray) trees in the study area (right).
Tree Physiology Volume 34, 2014
Downloaded from http://treephys.oxfordjournals.org/ at University of Tokyo Library on May 20, 2015
We also tested for differences in two other traits: needle nitrogen (N) concentration and maximum photochemical efficiency of photosystem II (PSII, Fv/Fm). Leaves with high leaf N usually have a high relative growth rate because high leaf N invested in carbon uptake increases the photosynthetic capacity (Antúnez et al. 2001, Ruiz-Robleto and Villar 2005). Leaf N usually shows a positive association with specific leaf area (a measure of leaf area displayed per unit dry mass invested) and a negative association with leaf thickness (Reich et al. 1997, Cunningham et al. 1999, Poorter and Bongers 2006, Schulze et al. 2006, Poorter et al. 2009). Thin leaves are more vulnerable to desiccation, which could lead to reduced survival in arid and nutrient-poor environments where thicker leaves with low N are expected to be beneficial (Reich 2001). In contrast, leaves with high N presenting higher maximum assimilation rates can also allow better growth, more reserves to survive drought periods and less reduction in assimilation due to stomatal closure (Kattge et al. 2009). Maximum photochemical efficiency of PSII of the leaves under dark-adapted conditions provides an estimate of PSII health status, which informs about the state of the whole photosynthetic apparatus, and is commonly estimated by the ratio between light-induced variable fluorescence and maximum fluorescence: Fv/Fm. This metric has been also reported to be a good predictor of growth rate, particularly under stressful conditions (see Cavender-Bares and Bazzaz 2004 and references therein). According to the ecological and functional hypotheses, we expected native progenies to exhibit lower growth rates under well-watered conditions but higher WUE, root investment and survival under water-limiting conditions, as a result of a presumed adaptation to their dry environment. By contrast, we expected individuals carrying allochthonous genes to exhibit higher N and relative growth rates in the absence of water limitation but reduced survival under dry conditions, as a consequence of selection for high-productivity seed sources and presumed (see Materials and methods) lower avoidance of and tolerance to dry conditions. If our initial hypotheses held, it would indicate an early selective advantage of native genotypes under water-limiting conditions, which might reduce the
Adaptive consequences of human-mediated introgression for indigenous tree species 1379
Controlled pollinations In April 2010, we performed intra- and interpopulation controlled pollinations through tree climbing to obtain ‘pure’ and (intraspecific) ‘hybrid’ progenies. Specifically, (i) we collected pollen from 20 randomly selected adult individuals of the native population and used this mixture to pollinate several female strobili at each of other 20 randomly selected individuals from the native population (mother trees were different from pollen donors) and (ii) we collected pollen from 20 randomly selected adult individuals of the plantations and used the mixture to pollinate a different set of strobili from the same 20 mother trees of the native population. Female strobili were isolated with paper and polyethylene bags before maturation to prevent open pollination. Pollinations were carried out when female strobili were receptive (as indicated by the separation of the scales and their position at almost right angle to the conelet axis), by pollen injection with a syringe twice within 72 h. Bags were removed some weeks later when the scales of female strobili were closed. Intrapopulation crosses yielded ‘pure’ progenies with a local native genetic background (L), whereas interpopulation crosses provided intraspecific ‘hybrid’ (native × plantation) progenies (H), mimicking introgression via pollen. In October 2011, we collected all mature cones from the mother trees. Female strobili inside bags left unpollinated during the receptive period aborted without producing seeds, confirming effective isolation from open pollination. We also collected cones from 20 open-pollinated mother trees from the plantation to have ‘pure’ exotic plantation progenies (E). These 20 trees were different from the 20 pollen donors previously sampled in the plantations, and were randomly selected from the core of planted stands, keeping a minimum separation distance of
100 m from each other to avoid paternal correlations among families. Given the very small size of the native stand relative to the plantations, we assumed a negligible probability of effective pollen flow from the former to the latter. In particular, pollen immigration rates are expected to be proportional to sourceto-sink population size ratios (Holsinger 1991), so knowing that the pollen flow rate from the plantations (comprising hundreds of thousands of trees) to the small relict (consisting of only 200 trees) is ∼8% (Unger et al. 2014), the reciprocal rate is probably
Research paper
Adaptive consequences of human-mediated introgression for indigenous tree species: the case of a relict Pinus pinaster population
1Department
of Forest Ecology and Genetics, INIA-CIFOR, Carretera de La Coruña km 7.5, E-28040 Madrid, Spain; 2Present address: Department of Ecology, Evolution and Behavior, University of Minnesota, 1987 Upper Buford Circle, Saint Paul, MN 55108, USA; 3Corresponding author ([email protected])
Received January 15, 2014; accepted October 15, 2014; published online December 1, 2014; handling Editor João Pereira
Human-induced gene movement via afforestation and restoration programs is a widespread phenomenon throughout the world. However, its effects on the genetic composition of native populations have received relatively little attention, particularly in forest trees. Here, we examine to what extent gene flow from allochthonous plantations of Pinus pinaster Aiton impacts offspring performance in a neighboring relict natural population and discuss the potential consequences for the long-term genetic composition of the latter. Specifically, we conducted a greenhouse experiment involving two contrasting watering treatments to test for differences in a set of functional traits and mortality rates between P. pinaster progenies from three different parental origins: (i) local native parents, (ii) exotic parents and (iii) intercrosses between local mothers and exotic fathers (intraspecific hybrids). Our results showed differences among crosses in cumulative mortality over time: seedlings of exotic parents exhibited the lowest mortality rates and seedlings of local origin the highest, while intraspecific hybrids exhibited an intermediate response. Linear regressions showed that seedlings with higher water-use efficiency (WUE, δ13C) were more likely to survive under drought stress, consistent with previous findings suggesting that WUE has an important role under dry conditions in this species. However, differences in mortality among crosses were only partially explained by WUE. Other non-measured traits and factors such as inbreeding depression in the relict population are more likely to explain the lower performance of native progenies. Overall, our results indicated that intraspecific hybrids and exotic individuals are more likely to survive under stressful conditions than local native individuals, at least during the first year of development. Since summer drought is the most important demographic and selective filter affecting tree establishment in Mediterranean ecosystems, a potential early selective advantage of exotic and hybrid genotypes would enhance initial steps of introgression of non-native genes into the study relict population of P. pinaster. Keywords: conservation genetics, drought avoidance, drought tolerance, gene flow, natural selection, plantations, translocations, water-use efficiency.
Introduction Introduced exotic species are often considered one of the greatest threats to conservation of natural ecosystems (Vitousek et al. 1997, Lee 2002). Their impacts on biodiversity have been demonstrated and the extinction of some
native species has been attributed to processes such as hybridization and competition with exotic species (Ellstrand 1992, McMillan and Wilcove 1994, Rhymer and Simberloff 1996, Vila et al. 2000, Mooney and Cleland 2001, Wolf et al. 2001, Petit 2004).
© The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
Downloaded from http://treephys.oxfordjournals.org/ at University of Tokyo Library on May 20, 2015
José Alberto Ramírez-Valiente1,2,3 and Juan José Robledo-Arnuncio1
Adaptive consequences of human-mediated introgression for indigenous tree species 1377 this study to what extent gene flow from exotic plantations of maritime pine (Pinus pinaster Aiton) impacts offspring performance in a neighboring natural relict population. Pinus pinaster is a keystone tree species widely distributed in the Mediterranean basin. It shows high intraspecific genetic differentiation at both neutral markers (Vendramin et al. 1998, Salvador et al. 2000, Gonzalez-Martinez et al. 2002, González-Martínez et al. 2004) and functional traits related to growth, reproduction, drought and freezing temperatures (Alía et al. 1995, Correia et al. 2008, Aranda et al. 2010, Corcuera et al. 2011a, 2011b, Santos-del-Blanco et al. 2012, Marguerit et al. 2014). It has been cultivated in many areas of the globe for commercial and ecological purposes (Richardson 1998), both within and out of its natural distribution range. Recent studies have shown that the species has become one of the most widespread invasive alien tree species on Earth (Nuñez and Medley 2011, Richardson and Rejmánek 2011). Within its natural range, maritime pine has a rather fragmented distribution, with several small isolated populations. In Spain, the species represents over 10% of the forest cover, with a total occupied area of 1,324,000 ha, 45% of which are plantations mainly established during the 20th century (De Lucas et al. 2008). Seeds used for maritime pine plantations in Spain have traditionally been collected from a limited number of (most frequently allochthonous) populations and individuals, selected for phenotypic traits such as vigor, growth architecture (e.g., apical dominance) and seed cone production and size. Ecologically, maritime pine has a drought avoidance strategy based on the high sensitivity of stomata closure to reductions of water availability (Picon et al. 1996, Fernández et al. 2000). This mechanism allows maritime pine to decrease water loss, increase water-use efficiency (WUE) and, ultimately, survive under severe xeric conditions (de Miguel et al. 2012). Nevertheless, drought tolerance mechanisms such as osmotic adjustment have also been reported to have an important role in adaptation to drought for this species (Nguyen-Queyrens et al. 2002, Nguyen-Queyrens and Bouchet-Lannat 2003). In contrast, the importance of below-ground biomass allocation is less clear for maritime pine. Triboulot et al. (1995) found that water stress stimulated root growth, but only under moderate conditions and for a short-term period (12 days). Aranda et al. (2010) reported differences in root biomass allocation among populations from different climates, but water treatment did not affect this trait. In this study, we set a greenhouse experiment to test for differences in WUE, biomass partitioning, growth rate and mortality under drought conditions between P. pinaster progenies from three parental origins: (i) a small natural relict population growing in ecologically marginal dry conditions, (ii) allochthonous plantations of undocumented origin surrounding the natural stand and (iii) intercrosses between native and allochthonous parents (intraspecific hybrids).
Tree Physiology Online at http://www.treephys.oxfordjournals.org
Downloaded from http://treephys.oxfordjournals.org/ at University of Tokyo Library on May 20, 2015
In plants, the impacts of human-induced gene movement have received considerable attention in recent years in the context of genetically modified crop species (Hoffman 1990, Raybould and Gray 1994, Pilson and Prendeville 2004, Chapman and Burke 2006) and invasive weeds (Small 1984, Van Raamsdonk and Van Der Maesen 1996, Snow et al. 2001). However, population genetic effects of intraspecific translocations have received much less attention, particularly in forest trees (Laikre et al. 2010; but see Steinitz et al. 2012 and references therein). Recent studies have reported that non-local exotic plantations have a significant effect on the genetic composition of the offspring of nearby conspecific populations at neutral loci (Sampson and Byrne 2008, Robledo-Arnuncio et al. 2009, Steinitz et al. 2012). Specifically, Sampson and Byrne (2008) and Robledo-Arnuncio et al. (2009) found that pollen introgression rates from neighboring plantations reached values over 40% among seeds of natural populations. The long-term consequences of intraspecific hybridization for the genetic composition of the local population depend on the level of gene flow and the relative fitness of hybrids compared with pure native genotypes. For example, if F1 hybrids and successive crosses and backcrosses exhibit increased fitness relative to pure local genotypes (heterosis caused by dominant and/or overdominant effects), then hybridization may, throughout generations, lead to the ‘genetic dilution’ and even extinction of the native gene pool (Rhymer and Simberloff 1996, Allendorf et al. 2001, Potts et al. 2003, Barbour et al. 2005). On the contrary, F1 hybrids and later generation backcrosses can suffer reduced fitness as a result of outbreeding depression or maladaptation of the alien gene pool to local environmental conditions. Under the latter scenario, intraspecific hybridization would have a limited long-term impact on the genetic composition of the native population, unless a very high level of gene flow overrides selection, ‘swamping’ the local gene pool (Lenormand 2002). Allochthonous genotypes can also invade native populations in the absence of intraspecific hybridization via seed dispersal from plantations, especially if introduced genotypes have either a numerical or fitness advantage over local genotypes (Hufford and Mazer 2003). Neutral marker studies have contributed greatly to estimate the exposure of natural populations to allochthonous genes from translocated populations (Robledo-Arnuncio et al. 2009, Steinitz et al. 2012, Unger et al. 2014). However, they have limited potential to assess the adaptive consequences of such human-induced gene flow, and in particular the potential risk it poses for the long-term viability of small indigenous populations (Kramer and Havens 2009). Besides knowing the migration rates between populations, a first step (feasible for long-living species) toward understanding such adaptive consequences involves the comparative assessment of fitnessrelated traits of F1 hybrids and pure-bred lines from parental populations (native and exotic). Along this line, we examine in
1378 Ramírez-Valiente and Robledo-Arnuncio potential negative impact of allochthonous gene flow on the genetic composition of the natural stand (Fenster and Galloway 2000, Hufford and Mazer 2003).
Materials and methods Study site We studied a small isolated relict population of maritime pine, located in the southern innerlands of Spain, 1000–1200 m above sea level in the municipality of Fuencaliente (38°25′N, 4°15′W; Figure 1). The entire population consists of ∼200 maritime pine adult trees (Alía et al. 1996). The climate of the area is Mediterranean with continental influence. Average annual precipitation is 833 mm and average annual temperature is 13.4 °C (Spanish Ministry of Agriculture, Food and Environment). Dry season extends over 3 months and summer precipitation reaches one of the lowest values for maritime pine in the Iberian Peninsula (58 mm). Moreover, the native stand grows on a steep crag of quartzite, with poor water retention capacity. The nearest native conspecific populations are located over 100 km east of the study site (at Sierra de SeguraAlcaraz). However, large afforestations were carried out in this area under plantation programs that began in the 1940s (DíazFernández 1994). As a result, the small native population is currently surrounded by widespread maritime pine plantations of undocumented origin (Figure 1). Neutral genetic differentiation between local and planted trees is significant both at nuclear and chloroplast microsatellite loci (Fst = 0.08–0.10), confirming the allochthonous origin of the plantations (Unger et al. 2014). Preliminary results with functional candidate gene SNP markers for biotic and abiotic stress including drought response, wood properties, phenology and growth (markers described in Budde et al. 2014) also suggest some degree of adaptive differentiation between native and planted stands (our unpublished data). Genetic cluster analysis using the same SNP markers indicates that the Fuencaliente native relict is genetically distinct from the plantations and all the studied native Iberian populations of the species, and it also suggests that native populations from southeastern Spain are the most likely origin of the plantations (S.C. González-Martínez,
Figure 1. Location of the study site (left) and distribution of native (light gray) and planted (dark gray) trees in the study area (right).
Tree Physiology Volume 34, 2014
Downloaded from http://treephys.oxfordjournals.org/ at University of Tokyo Library on May 20, 2015
We also tested for differences in two other traits: needle nitrogen (N) concentration and maximum photochemical efficiency of photosystem II (PSII, Fv/Fm). Leaves with high leaf N usually have a high relative growth rate because high leaf N invested in carbon uptake increases the photosynthetic capacity (Antúnez et al. 2001, Ruiz-Robleto and Villar 2005). Leaf N usually shows a positive association with specific leaf area (a measure of leaf area displayed per unit dry mass invested) and a negative association with leaf thickness (Reich et al. 1997, Cunningham et al. 1999, Poorter and Bongers 2006, Schulze et al. 2006, Poorter et al. 2009). Thin leaves are more vulnerable to desiccation, which could lead to reduced survival in arid and nutrient-poor environments where thicker leaves with low N are expected to be beneficial (Reich 2001). In contrast, leaves with high N presenting higher maximum assimilation rates can also allow better growth, more reserves to survive drought periods and less reduction in assimilation due to stomatal closure (Kattge et al. 2009). Maximum photochemical efficiency of PSII of the leaves under dark-adapted conditions provides an estimate of PSII health status, which informs about the state of the whole photosynthetic apparatus, and is commonly estimated by the ratio between light-induced variable fluorescence and maximum fluorescence: Fv/Fm. This metric has been also reported to be a good predictor of growth rate, particularly under stressful conditions (see Cavender-Bares and Bazzaz 2004 and references therein). According to the ecological and functional hypotheses, we expected native progenies to exhibit lower growth rates under well-watered conditions but higher WUE, root investment and survival under water-limiting conditions, as a result of a presumed adaptation to their dry environment. By contrast, we expected individuals carrying allochthonous genes to exhibit higher N and relative growth rates in the absence of water limitation but reduced survival under dry conditions, as a consequence of selection for high-productivity seed sources and presumed (see Materials and methods) lower avoidance of and tolerance to dry conditions. If our initial hypotheses held, it would indicate an early selective advantage of native genotypes under water-limiting conditions, which might reduce the
Adaptive consequences of human-mediated introgression for indigenous tree species 1379
Controlled pollinations In April 2010, we performed intra- and interpopulation controlled pollinations through tree climbing to obtain ‘pure’ and (intraspecific) ‘hybrid’ progenies. Specifically, (i) we collected pollen from 20 randomly selected adult individuals of the native population and used this mixture to pollinate several female strobili at each of other 20 randomly selected individuals from the native population (mother trees were different from pollen donors) and (ii) we collected pollen from 20 randomly selected adult individuals of the plantations and used the mixture to pollinate a different set of strobili from the same 20 mother trees of the native population. Female strobili were isolated with paper and polyethylene bags before maturation to prevent open pollination. Pollinations were carried out when female strobili were receptive (as indicated by the separation of the scales and their position at almost right angle to the conelet axis), by pollen injection with a syringe twice within 72 h. Bags were removed some weeks later when the scales of female strobili were closed. Intrapopulation crosses yielded ‘pure’ progenies with a local native genetic background (L), whereas interpopulation crosses provided intraspecific ‘hybrid’ (native × plantation) progenies (H), mimicking introgression via pollen. In October 2011, we collected all mature cones from the mother trees. Female strobili inside bags left unpollinated during the receptive period aborted without producing seeds, confirming effective isolation from open pollination. We also collected cones from 20 open-pollinated mother trees from the plantation to have ‘pure’ exotic plantation progenies (E). These 20 trees were different from the 20 pollen donors previously sampled in the plantations, and were randomly selected from the core of planted stands, keeping a minimum separation distance of
100 m from each other to avoid paternal correlations among families. Given the very small size of the native stand relative to the plantations, we assumed a negligible probability of effective pollen flow from the former to the latter. In particular, pollen immigration rates are expected to be proportional to sourceto-sink population size ratios (Holsinger 1991), so knowing that the pollen flow rate from the plantations (comprising hundreds of thousands of trees) to the small relict (consisting of only 200 trees) is ∼8% (Unger et al. 2014), the reciprocal rate is probably
