Estimates of pollen contamination and selfing in a coastal Douglas-fir seed orchard

Seed orchards are the link from tree breeding to reforestation programs and are theoretically expected to function as closed, perfect populations, ensuring gain and diversity are consistently and predictively delivered as improved seed and seedlings. Seed orchard populations often deviate from panmixia due to fertility variation, reproductive asynchrony, and gene flow, leading to reduced seed crops genetic quality. Here, as a part of multiyear monitoring study, we used DNA fingerprinting (simple sequence repeat markers) to assess a Douglas-fir (Pseudotsuga menziesii) seed orchard's seed crop genetic quality (2009 seed crop). The studied seed crop was produced under ambient temperature (i.e. no reproductive phenology manipulation) and pollination was augmented by pollen from within orchard's pollen donors. DNA fingerprinting of the parental population (66 parents) along with 207 gametophyte (1n) – embryo (2n) pairs of random bulk sample of seed allowed parentage (maternal and paternal) assignment and the direct assessment of pollen contamination (0.18 ± 0.027) and selfing (0.17 ± 0.025) rates as well as parental (pollen, ovule and individual parent) gametic contribution was compared to a previous year's crop (2005). The extended reproductive phenology coupled with variable within-orchard pollen availability has created opportunities for both self and foreign pollen to be successful at various times resulting in the seemingly paradoxical scenario of high selfing and gene flow. These results (2005 and 2009) allowed comparison of seed orchard's crop management practices and are expected to provide scientific foundations to effective seed crops genetic quality improvement.     

Mating System and Pollen Contamination in a Pinus brutia Seed Orchard

Mating system parameters and pollen contamination were estimated in an 11-year-old Pinus brutia Ten. clonal seed orchard, located in Antalya, Turkey, with the aid of isozymes. Isozyme analysis was performed on both maternal (megagametophyte) and embryo tissues of seeds collected from the seed orchard and the surrounding natural stand. Fourteen loci encoding nine enzyme systems were assayed. Based on a multilocus estimator, the proportion of viable seeds originating from outcrossing (tˆ_m) in the seed orchard was found to be 0.947; the remainder (0.053) was due to selfing. We estimated that 85.7% of the orchard seed resulted from pollen contamination by trees in surrounding stands. This may be related to the young age and subsequent low pollen production of the orchard. Only 9% [i.e., 1.0−0.857 (contamination)−0.053 (selfing) = 0.090] of the orchard offspring was due to cross-pollination among clones within the orchard. Due to the high level of pollen contamination, it is estimated that genetic gain in this seed crop is less than 57% of that expected if all offspring resulted from random mating among orchard clones (i.e., no contamination).     

Flowering and Airborne Pollen Occurrence in a Pinus sylvestris Seed Orchard Consisting of Northern Clones

To explain the high background pollination rate detected in earlier studies, this study investigated the flowering phenology of a Scots pine seed orchard with respect to airborne pollen occurrence during 1992 and 1993. The receptive period of the female strobili and the shedding period of the male strobili were observed visually. Airborne pollen was monitored with Rotorod traps located inside and outside the seed orchard. There was no time gap in the onset of flowering between the earliest female and male strobili in 1992. Female flowering began 3 days earlier than male flowering in 1993, and the earliest strobili were probably pollinated by pollen transported long distance from more southerly areas. The almost simultaneous peak in pollen occurrence inside the seed orchard and in surrounding pine stands indicates that the seed orchard clones are not temporally isolated from sources of contamination. However, pollen concentrations inside the seed orchard were higher than those in the surrounding stands. Therefore, flowering phenology cannot sufficiently explain the observed pollen contamination rates in this seed orchard.     

林木种子园花粉污染研究进展

林木种子园花粉污染是影响种子园了遗传品质的重要因子，本文综述了国外林木种子园花粉污染研究的概况、污染水平、影响污染水平的因子、减少污染的措施等。并对我国种子园花粉污染研究和花粉管理提出了建议。     

Mating patterns in an indoor miniature <Emphasis Type="Italic">Cryptomeria japonica</Emphasis> seed orchard as revealed by microsatellite markers

To evaluate the practical application of an indoor seed orchard, we compared the seed production, final germination rates and mating patterns of potted Cryptomeria japonica D. Don clones in two miniature seed orchards: one in a glasshouse and the other outdoors. There was no statistically significant difference in total seed production between the orchards (P = 0.275, ANOVA). However, the final germination rate of seeds produced in the indoor orchard was significantly lower than that of seeds produced in the outdoor orchard (P < 0.01, Wilcoxon test). The average self-fertilization rate was higher in the indoor orchard (27.2%) than in the outdoor orchard (5.6%), and this parameter was strongly associated with the reduced final germination rates of the seeds harvested in the glasshouse. Pollen contamination was considerably lower in the indoor seed orchard (48.9% in the outdoor seed orchard and 4.4% in the indoor seed orchard). χ² tests indicated that the paternal contributions of each constituent clone differed significantly in both orchards (P < 0.001 in both cases). The strong variations in paternal contributions among indoor orchard clones was significantly correlated with their respective pollen production capacities (Spearman ρ = 0.693, P < 0.001). To reduce the high self-fertilization rates, we recommend the use of orchard clones with similar pollen production capacity in indoor seed orchards.     

Effect of patch size and isolation on mating patterns and seed production in an urban population of Chinese pine (Pinus tabulaeformis Carr.)

Fragmentation is a critical issue for tree populations because the creation of small patches can reduce local population size and increase isolation, both of which can promote inbreeding and its negative consequences, as well as loss of genetic diversity. To test the hypothesis that patch size and isolation influence mating patterns or seed production in forest trees, we utilize the spatial array of trees of an urban population of Chinese Pine (Pinus tabulaeformis Carr.) planted in patches around Beijing. Our design includes 28 urban patches, with patch size ranging from 1 to 2000 adult trees and isolation (edge distance index) ranging from 37.5 m to 245.8 m. We examined the average number of seeds per cone and percentage of viable seed per cone for each patch as measures of seed production. By utilizing seven paternally inherited chloroplast microsatellite loci, we estimated the mating pattern parameters for each patch, including the level of selfing, the amount of immigrant pollen and the effective number of pollen sources (N_ep). Using a general linear model selection procedure based on AIC value, we found patch size was the best predictor of the selfing and immigration rate; smaller patches had a higher selfing and immigration rate. Small patches with one adult had relatively high N_ep which indicates connectivity among urban patches. However, due to the reduced amount of immigrant pollen and limited diversity of local pollen, intermediate sized patches (with 5–10 adults) had the lowest N_ep among the study patches. For patches with more than 10 adults, N_ep was increased with patch size. The percentage of viable seeds per cone significantly decreased with patch size, indicating a possible negative consequence of inbreeding. The effect of patch size on mating patterns and seed production suggests that the patches of trees experience less connectivity than trees within continuous forest. These findings indicate that forest management practices should emphasize the maintenance of an optimal patch size because, despite the fact that tree species show the potential for long distance pollen movement, the number of local trees strongly influences the mating patterns.     

Low rates of pollen contamination in a Scots pine seed orchard in Sweden: the exception or the norm?

We investigated mating structure and gene flow in a clonal seed orchard of Scots pine (Pinus sylvestris L.) over three consecutive pollination seasons (2010–2012) with nine nuclear microsatellite markers. The paternity of 1991 offspring from four maternal parents was assigned to 28 candidate fathers using an exclusion procedure and a likelihood-based method implemented in the program CERVUS. Relative reproductive success was highly variable among pollen parents but consistent across years and ranged from 0.1% to 18.3%. Consequently, the seed crops’ effective number of fathers was reduced to 52.9%, 48.8%, and 45.7% of the census in the three seasons, respectively. Self-fertilization fluctuated around the orchard's expected value of 5.1%, reaching 4.05%, 7.71%, and 6.61%, respectively. Pollen contamination was estimated to be 5.64%, 7.29%, and 4.89%, respectively, after correction for cryptic gene flow. CERVUS provided similar results as the exclusion method, but estimates greatly varied depending on the input parameters, mainly the proportion of fathers sampled. These results indicate the studied seed orchard is a well-functioning production population with only minor negative effects of self-fertilization and pollen contamination on the quality of seed crops. Genotyping issues associated with microsatellites as a potential source of false paternity assignment and exclusion are discussed.     

Background pollination in Pinus sylvestris seed orchards

The proportion of background pollen grains in the seed crops of two Pinus sylvestris seed orchards in Central Finland was estimated with the aid of multilocus allozyme markers. The orchards studied were more than 30 years old and in full pollen production. For the bulked seed crops of the seed orchard with southern clones the estimated average of background pollination over four years was 26%. There were statistically significant differences between years. No significant heterogeneity in the degree of background pollination between clones was found. Among single ramets there was significant heterogeneity in the estimated contamination rates, but the variation was not related to position in the orchard. For the seed orchard with northern clones the bulked seed crop was studied only for one year and the level of background pollination was found to be 33 %. These estimates are fairly high, but lower than for many other orchards. Background pollination at this level will cause losses in expected genetic gains. Part of the seeds from northern orchards will not be adapted to the intended area of use.     

Levels of outcrossing and contamination in two Pinus sylvestris L. seed orchards in Northern Sweden

The joint levels of outcrossing and contamination in two young northern Swedish Scots pine (Pinus sylvestris L.) seed orchards were studied with the aid of allozyme markers. High levels of contamination (21–36%) were estimated and were attributed to the observed low within orchard pollen production. Outcrossing rate estimates were high (0.957–0.961), though they departed significantly from complete outcrossing (t= 1.0) in spite of the observed high contamination levels. These results confirmed reported relationships between outcrossing and contamination levels in open‐pollinated seed orchards. It was concluded that a crop management option relying on supplemental mass‐pollination would be effective in reducing both contamination and inbreeding levels.     

Density-dependent selfing and its effects on seed performance in a tropical canopy tree species, Shorea acuminata (Dipterocarpaceae)

In the conservation and management practices of natural forests, sound reproduction and regeneration form the basis of the maintenance and viability of the tree populations. To obtain and serve biological information for sustainable forest management, we investigated reproductive biology and inbreeding depression in seeds of an important dipterocarp tree species, Shorea acuminata (Dipterocarpaceae), by both field and laboratory experiments. Results of parental analysis of immature and mature seeds showed that selfing rates varied greatly, from 7.6 to 88.4% among eight mother trees, and the mean overall selfing rate was 38.3%. Observed outcrossing events within a 40-ha study plot were predominantly (76.5%) short-distance events with a mating distance (md) ≤ 100 m. Since the selfing rate sharply decreased with increase in the number of flowering conspecifics (i.e., individuals of the same species) within a 100-m radius from the mother trees, the local density of flowering conspecifics appears to be the key factor determining the outcrossing rate in S. acuminata. However, the extremely high selfing rate (88.4%) observed for one tree could not be simply explained by the low local density of flowering conspecifics. Instead, differences in its flowering phenology (its flowering peaked ca. a week earlier than most of the other examined individuals) may have severely limited its receipt of pollen from other conspecifics, and thus promoted selfing. Since there were no significant differences in the proportion of selfed progeny between immature and mature seed stages, there was no evidence of selective abortion of selfed seeds during seed development. However, the seed mass of outcrossed progeny was heavier than that of selfed progeny, and heavier seeds showed higher success rates at germination and seedling establishment. These results suggest that inbreeding depression resulted in reductions in seed mass and may reduce the fitness of selfed seeds in S. acuminata. In addition, the outcrossing rate of S. acuminata was more sensitive to low local conspecific flowering-tree densities than that of a sympatric bee-pollinated dipterocarp species with greater pollination distances. These results suggest that the management of local adult-tree densities is important for avoiding selfing and inbreeding depression in future generations, especially in a species like S. acuminata with predominantly short-distance pollination.     

宝兴百合花柱偏转促进延迟自交试验研究

2011年7~8月对云南香格里拉高山植物园的宝兴百合(Lilium duchartrei Franch.)种群进行了传粉生物学研究实验,以期找到这种花朵开放初期雌雄异位的植物传粉生物学特点。结果表明,记录到的178次昆虫对宝兴百合的访问中能触碰到柱头的仅有蝴蝶类1次,有效传粉者传粉频率严重不足;单花期5.5~6.5 d,花开放的前3天,柱头远离花药,从花开放的第二天起花柱缓慢向上偏转,第四天将柱头从远离花药位置送到刚好可以接受自花花粉的位置,利于早期没有昆虫传粉的花朵延迟自交,自花授粉发生在花期末期,自交之前仍然保持异交传粉机制;自然状态(92.3%)及完全套袋处理(90%)的结实率接近,说明宝兴百合传粉存在花柱偏转促进的延迟自交,但2种处理的每果实种子数自然状态(121)显著高于完全套袋处理(78),又说明由昆虫传粉的异交更利于宝兴百合结实。延迟自交避免了自交与异交竞争造成的花粉或者种子的折损,为宝兴百合在有效访花昆虫不足导致异交失败时提供繁殖保障。     

Southern pollen sired more seeds than northern pollen in southern seed orchards established with northern clones of Pinus sylvestris

Abstract

Contamination by southern pollen is a considerable problem in Scots pine (Pinus sylvestris L.) seed orchards established with northern clones. This study investigated whether the contamination was due to the competitive superiority of southern pollen by carrying out competition trials using mixtures of pollen from northern and southern populations of Scots pine. Trials were performed in a southerly seed orchard established with clones originating from northern populations. Seed paternity (siring) was determined through the analysis of allozyme variation. Southern genotypes sired significantly more seeds (76%) than their northern competitors and across all mixed-pollen crosses. Maternal genotype had no effect on seed siring success. The mean flower abortion rate was lower in southern pure-pollen crosses and mixed-pollen crosses than in northern pure-pollen crosses. The results show that local pollen may induce high levels of background pollination in southern seed orchards composed of northern genotype grafts. These results must be taken into account when aiming to produce suitable reforestation material for northern areas.     

Pollination dynamics in a Douglas-fir seed orchard as revealed by pedigree reconstruction

• ? Pollination dynamics was studied in a Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) seed orchard using 8 nuclear microsatellite markers and pedigree reconstruction.
• ? The seed orchard consisted of 49 parents (clones). Cone-crop management included bloom delay and supplemental mass pollination (SMP) using 12 internal and 4 external pollen donors.
• ? A random sample of 801 bulk seeds was genotyped for both haploid megagametophyte and corresponding diploid embryo.
• ? Using the parental population’s multilocus genotypes, full pedigree reconstruction generated all the information needed to estimate the maternal, paternal, and parental reproductive success, selfing, pollen contamination, and pollination success of the 4 external pollen donors.
• ? Maternal, paternal, and parental reproductive success varied with 80% of gametes being produced by 23, 45, and 37% of the orchard’s parents, respectively, resulting in a drastically reduced effective population size as compared to the census number (14 vs. 53).
• ? Selfing, pollen contamination, and aggregate SMP success (internal and external) were estimated to be 15.2, 10.4, and 15.0%, respectively.
• ? Full pedigree reconstruction was effective in unraveling the orchard’s pollination dynamics and both female and male reproductive success.

     

Pollen dispersal,mating patterns and pollen contamination in an insect-pollinated seed orchard of <Emphasis Type="Italic">Schima superba</Emphasis> Gardn. et Champ.

Pollen flow and mating patterns are the most important factors influencing the genetic structure of insect-pollinated forest trees and are essential parts of genetic management in seed orchards. We investigated pollen flow, the mating system and the level of pollen contamination in a clonal seed orchard of Schima superba Gardn. et Champ. In total, 328 open-pollination progenies coming from 11 mother trees were identified using 13 polymorphic microsatellite loci. A total of 203 full-sib families were identified and were nested among the 11 studied seed donors. The male reproductive success rate from 0.49 to 7.77% for most male parents, with an average of 2.44%. More than 80% of the crosses were found within a distance of 60 m, and the most frequent pollination distance between female parent and male parents was approximately 20 m. Mating system analysis showed that the outcrossing level was very high (t _m  = 1.000, outcrossing rate = 98.5%) in the seed orchard and that there was an average of 2.3 effective pollen donors (N _ep ) per female parent. In addition, the mating success of individual males within neighborhoods was moderately influenced by their fecundity and the direction of their location relative to mother trees. The pollen contamination from outside the seed orchard was high (7.01%). Our findings are valuable for the assessment of seed orchards, and it may be worthwhile to use pollen management strategies to decrease pollen contamination and increase the genetic quality of the seeds produced.     

Annual Variation in Pollen Contamination and Outcrossing in a Picea abies Seed Orchard

A mature Norway spruce [Picea abies (L.) Karst.] seed orchard, established in southern Finland with 67 clones from northern Finland, was analysed in three different years in order to estimate the pollen contamination ratios. Allozyme-based paternity analysis revealed that the contamination rate was high, 69-71%, and did not differ between the years studied. It appears that, in areas where spruce is the dominant tree species, the contamination rate will be very high even in mature seed orchards. However, the contamination rate in the thinned parts of the orchard was significantly lower than that in the unthinned parts in two of the three years studied. The outcrossing rate was also high, 96-100% in all years, even though the ramets of each clone were planted using a clonal-row design, and there were no significant differences in the outcrossing rates between the different parts of the orchard.     

Congruence between theory and practice: reduced contamination rate following phenotypic pre-selection within the Breeding without Breeding framework

Wind-pollinated seed orchards are often subjected to pollen intrusion from outside pollen sources. When wind-pollinated seed orchards seed are used to establish progeny trials, the offspring is expected to harbour varying proportion sired by outside sources. Theoretically the magnitude of selection differential between the orchard's population and the contaminant pollen sources will affect the proportion of offspring resulting from outside pollen sources matings. If phenotypic pre-selection is implemented through selecting the top phenotypically ranked individuals from these progeny trials, then it is expected that the proportion of individuals sired by outside pollen sources will be lower within the pre-selected vs. unselected individuals. Here, we present empirical data from two Scots pine progeny trials supporting these theoretical expectations. The observed reduced contamination rate among fingerprinted elite offspring was 3.9 and 4.2%, suggesting a significant reduction in comparison to reported contamination rates between 21 and 70% in Scots pine. Results provide support to the proposed phenotypic pre-selection during the implementation of Breeding without Breeding scheme.     

Seed cone abortion and full seed production in an indoor seed orchard with potted grafts of Picea abies

Abortion of seed cones and production of full seeds have been recorded in an indoor‐potted Norway spruce (Picea abies (L.) Karst.) seed orchard at Biri nursery in Norway. Seed cone buds aborted their normal development at much higher frequencies (22–45%) inside isolation bags than in an open‐air (4–12%) greenhouse environment. The grafts had been induced to flower the year prior to seed cone development. Treatment with gibberellin A_4/7 in conjunction with heat reduced the proportion of aborted, isolated flowers (13%) compared to treatment with heat alone (22%) in a clonally balanced material. The production of full seeds per cone increased from 3 in 1987 to 50 in 1991. The full seed production is discussed in relation to methods and timing of pollen forcing. It is anticipated that 1 kg of seeds could be produced annually in a 1250 m² greenhouse.     

Are Phenological Observations Sufficient to Estimate the Quality of Seed Crops from a <Emphasis Type="Italic">Eucalyptus grandis</Emphasis> Open-pollinated Seed Orchard? Consequences for Seed Collections

Reproductive phenology was observed over three years in a Eucalyptus grandis seed orchard in Madagascar to determine the impact of geographic differences of parental selections on the expected genetic composition of seed crops. Pollen flow (observed pollen cloud) was deduced from a paternity assignment based on sampling from one year’s seed production. The two approaches were used to verify whether knowledge of phenology is sufficient to predict the genetic quality of the seeds collected. Despite the high flowering level, with a constant cycle over three years, the results demonstrated wide reproductive phenological differences associated with the parents’ origin, suggesting putative pollination disequilibrium. From both observed and expected pollen clouds, the results showed preferential mating among different provenances, which has consequences for seed crop composition. Phenological observations, which can be made inexpensively, can be used to promote methods of effective seed orchard management in order to improve the genetic quality of seeds.     

Jackknife resampling for precision measurement of direct gene flow estimates

Gene flow from outside pollen sources reduces the expected genetic gain of seed orchards, and has been quantified by genotyping seed samples from crop years to determine the proportion of seed sired by outside sources. Random bulk seed sampling and microsatellite genotyping allows for fast, unbiased identification of orchard contamination rates. Bulk seed sampling has been employed in multiple studies of seed orchard gene flow, but a method of determining precision of gene flow estimates from this sampling method has yet to be described. In this study, we used jackknife resampling to generate random samples of migrant and within-population genotypes obtained from an empirical sample of Douglas-fir (Pseudotsuga menziesii [Mirb.] Franco) genotypes of five microsatellite loci. We generated samples of 50, 100, 150, 200, and 250 genotyped individuals from a sample of 400 seeds, with genotypes corresponding to low (6%), moderate (14%), or high (56%) pollen contamination levels. The mean gene flow estimate and confidence interval were calculated for each sample size and gene flow scenario to determine the precision of gene flow estimates obtained from each sample size in each contaminant gene flow scenario. Genotyping 150 individuals was sufficient to produce gene flow estimates with calculated confidence interval values below ± 1%, indicating that 150 genotyped individuals were sufficient for precise gene flow estimation in this population. This resampling method will enable precise and unbiased estimation of gene flow into seed orchards for future monitoring of gene flow, to allow better management of seed orchard genetic resources.     

蓝桉树木园异型杂交及花粉污染的微卫星分子标记研究

本文应用4对微卫星分子多态引物,对利用优株嫁接建立的蓝桉(E.globulus)树木园中的118个自由授粉子代的DNA微卫星指纹进行了分析.结果表明:选用的引物具有高度多态性,各个位点扩增得到的平均等位基因数为14.25,树木园中母株的异型杂交率为15%(95%,整体异型杂交率为47.9%,外源花粉污染率为17.6%.由于该树木园的异型杂交率低,其自由授粉的种子应用价值不高,因此应采用大规模人工授粉来制种或选用树木园中异型杂交率高的母株采种.