GENETIC DIVERSITY BETWEEN AND WITHIN POPULATIONS OF Handroanthus heptaphyllus (VELL.) MATTOS USING MICROSATELLITE MARKERS

Handroanthus heptaphyllus (Vell.) Mattos, popularly known as ipê-roxo, is a species of the family Bignoneaceae much appreciated for its beauty, excellent quality wood which is used for making medicinal products and also in reforestation programs of degraded areas, as well as landscaping and restoration. The aim of this study was to investigate the genetic diversity between and within populations of H. heptaphyllus using microsatellite markers. The 192 seedlings were produced from seeds collected on 30 trees into the two populations of natural forest fragments in Botucatu region, São Paulo, Brazil. Eight microsatellite loci were analyzed, with allelic polymorphism varying from six alleles for locus TAU22 to 14 alleles for loci TAU12, TAU30, and TAU31, with an expected mean number of alleles per locus ( e A ) of 4.9. The mean expected heterozygosity ( e H ) for the two populations was 0.785, the mean observed heterozygosity ( o H ) was 0.609, and the fi xation index (F) was low between populations, with a mean of 0.222. The gene differentiation between the two populations ( ST G ’) was 0.100. We concluded that the higher genetic diversity is within populations; therefore, as far as germplasm collection programs in Botucatu region are concerned, it is recommended that a larger sampling of individuals should be considered within populations, thereby providing good genetic representativeness. The populations have enough genetic diversity to support genetic improvement and germplasm preservation programs.


Genetic diversity between and within populations... GENETIC DIVERSITY BETWEEN AND WITHIN POPULATIONS OF Handroanthus heptaphyllus (VELL.) MATTOS USING MICROSATELLITE MARKERS 1
Neide Tomita Mori 2 , Mario Luiz Teixeira de Moraes 3 , Caroline Midori Morita 4 , Edson Seizo Mori 5   (received: April 15, 2010; accepted: September 29, 2011)   ABSTRACT: Handroanthus heptaphyllus (Vell.)Mattos, popularly known as ipê-roxo, is a species of the family Bignoneaceae much appreciated for its beauty, excellent quality wood which is used for making medicinal products and also in reforestation programs of degraded areas, as well as landscaping and restoration.The aim of this study was to investigate the genetic diversity between and within populations of H. heptaphyllus using microsatellite markers.The 192 seedlings were produced from seeds collected on 30 trees into the two populations of natural forest fragments in Botucatu region, São Paulo, Brazil.Eight microsatellite loci were analyzed, with allelic polymorphism varying from six alleles for locus TAU22 to 14 alleles for loci TAU12, TAU30, and TAU31, with an expected mean number of alleles per locus ( e A ) of 4.9.The mean expected heterozygosity ( e H ) for the two populations was 0.785, the mean observed heterozygosity ( o H ) was 0.609, and the fi xation index (F) was low between populations, with a mean of 0.222.The gene differentiation between the two populations ( ST G ') was 0.100.We concluded that the higher genetic diversity is within populations; therefore, as far as germplasm collection programs in Botucatu region are concerned, it is recommended that a larger sampling of individuals should be considered within populations, thereby providing good genetic representativeness.The populations have enough genetic diversity to support genetic improvement and germplasm preservation programs.

INTRODUCTION
The fragmentation of forests by explorers and the urbanization advent have both had a tremendous impact on biodiversity, upsetting the balance of ecosystems, isolating habitats and changing the behavioral pattern of both fauna and fl ora by disrupting gene fl ow between fragments (KAGEYAMA et al., 2003).As the result, the genetic diversity of forest tree species has been decreasing.
Typically, Handroanthus heptaphyllus (Velloso) Mattos occurs in areas of rain forests.Commonly known as ipê-roxo, the species has become increasingly popular ^^ Mori, N. T. et al.
for several environmental and economic issues.The bark is widely used for making medicinal syrups and the wood is hugely popular for its high economic value.It has also been used in recovery and restoration programs on degraded areas, given the decreasing occurrence of this species in natural habitats (ETTORI et al., 1996), being also used as an ornamental plant in urban public squares and for restoring riparian vegetation, since the species develops well on soils adjacent to watercourses (MOREIRA;SOUZA, 1987).
According to Jankowsky et al. (1990), the species of genus Handroanthus provide hard and heavy wood with low retractibility and density about 1,070 kg.m -3 ; the wood is dark and has pale sapwood (PAULA; ALVES, 1997).Those characteristics have stirred the interest of timber companies and as a consequence the exploration has led to a large decrease in its natural occurrence.However, the species does not reproduce easily in natural environments and seed viability is very low under natural conditions (VIEIRA et al., 2010).
Despite the predatory exploration, the use of the species on a commercial scale is very limited and its seeds are still being collected from natural populations.So far, no genetic improvement programs have been created to offer improved seeds, whether for commercial purposes or even for creation of seed orchards.
In order to use a particular species for reforestation, conservation, or improvement programs, it is important to study its mating system and the genetic structure of populations through molecular markers, to know how genetic diversity is distributed within and between populations.It can help to acquire the necessary knowledge to ensure suitable population management, helping to stablish strategies to support genetic improvement and germplasm preservation programs (BROWN, 1978).
The seasonal semi deciduous forest is one of the forest subtypes that compose the Atlantic forest biome.Today, the total area of remnants of that type of forest in São Paulo State corresponds to less than 5% of the total area by the late nineteenth century (RAMOS et al., 2007) and H. heptaphyllus is a representative species of that vegetation.Therefore, the objective of the study was to investigate the genetic diversity between and within populations of H. heptaphyllus using microsatellite markers, looking to support genetic resource development and germplasm preservation programs for the species in region strongly affected by the impacts of forest fragmentation.

Study site
The research was conducted in the municipality of Botucatu-SP, Brazil, in the region of Médio Tietê Basin.Due to heavy urbanization, the studied region presents many small forest fragments (Figure 1).Two populations were demarcated in those forest fragments: population 1 (Lageado) consisted of 115 seedlings originated from 15 mother trees while population 2 (Indiana) was composed of 77 seedlings originated from 15 mother trees.The idea was to depict the representiveness occurrence of the species in the region, considering locally its low population density.

Microsatellite analysis
For each individual, three leafl ets were collected from adult leaves and then stored at room temperature in a dry environment.For DNA extraction, 150 mg of plant material were used per sample.The protocol proposed by Ferreira and Grattapaglia (1995) was used.The primers were transferred from Handroanthus aureus, as developed by Braga et al. (2007), considering their respective annealing temperatures, through PCR procedures, for replication of DNA fragments (Table 1).The DNA fragments were separated by 6% bis/ acrylamide gels.The gel was developed using nitrate of silver, according to the procedure described by Creste et al. (2001).Genetic diversity between and within populations...

Data analysis
Based on zymograms, the following parameters were estimated: allele frequencies ( ˆi p ), mean number of alleles per locus (A), derived from the arithmetic average of loci; expected number of alleles per locus ( 1 ˆ1 ), according to Weir (1996).The genetic structure between populations was derived from statistics proposed by Nei (1978), where: ˆT H is total gene diversity; ˆST G is coeffi cient of gene diferentiation between populations and ˆS H is gene diversity within populations, using FSTAT software (GOUDET, 2002).The estimation of ˆST G , however, was standardized , ˆST G , according to Hedrick (2005), and presented as follow:

Genetic diversity and fi xation index
Eight loci in two populations of H. heptaphyllus were found to be all polymorphic.A total of 63 alleles were found and the expected number of alleles per locus ( e A ) ranged from 3.4 (TAU 21) to 6.7 (TAU 14) in the population 1, and from 3.1 (TAU 30) to 8.0 (TAU 31) in the population 2, with mean value of 4.9 between populations (Table 2).Estimates of this parameter in other recent studies were found to range from 2.9 (MENDES, 2009) for Cedrela fi ssilis to 10.11 alleles (SILVA, 2010) for Tabebuia aurea.
The fi xation index ( F) followed the same trend with a wider range of variation in population 2, with a minimum of 0.058 (TAU 30) and a maximum of 0.414 (TAU 22), and a mean of 0.222 when both populations were considered, suggesting deviations on the Hardy-Weinberg equilibrium due to homozygous excess, probably due to inbreeding.
Based on the estimated parameters ( e A , o H and e H ) to help identify genetic diversity, and the fi xation index ( F), it can be inferred that population 2 had slight superiority in the mean values of loci and a wider range of variation for those parameters.However, considering the size of those populations and the conditions of fragmentation there are an increase of inbreeding levels than expected over time as a function of likely relatedness of generations, potentially decreasing heterozygous and showing higher structuring within populations.Therefore, effective measures should be taken in order to prevent these populations from remaining isolated and to consequently avoid genetic diversity to decrease for generations.This issue was discussed extensively in studies by Bittencourt and Sebbenn (2007), Gaino et al. (2010), and Sebbenn et al. (2011), on Araucaria angustifolia, Myracroduon urundeuva and Copaifera langsdorffi i respectively.

Genetic structure between and within populations
Results of total gene diversity ( T H = 0.794 or 100%) show higher diversity within populations ( s H = 0.785 or 98.8%) and lower gene diversity ( ST D = 0.09 or 1.2%) between populations.Considering only ST G , low variation was found between loci, the minimum gene diversity of 0.3% (TAU 27) and the maximum of 2.8% (TAU 14) were found.However, when the ' ST G proposed by Hedrick (2005) was used, the variation goes from 2.7% (TAU 27) to 30.3% (TAU 14), which is a considerable estimated value.Being more consistent and accurate parameter, ' ST G was the choice for discussion, indicating a mean variation of 10% between populations, which is something to be considered inasmuch as the presence or absence of alleles is concerned in either population (Table 3).This variation between the populations was becoming evident when studying the genetic diversity and fi xation index parameters, in which a higher range of variation was found in population 2. It is interesting the use of ' ST G statistic for that dispersion measure.ˆ^ˆ^^^ Genetic diversity between and within populations...
Another factor that supports statistic is the study of Hamrick (1983) with wind-dispersed seeds of 38 tree species.Although the author worked with isoenzymes, the estimated gene differentiation between populations ( ST G ) was 0.056 or 5.6%, which is higher than the present result for H. heptaphyllus populations (1.2%).However, when estimating ' ST G statistic, according to Hamrick (1983), the value was 0.093 or 9.3%, close to the estimate for H. heptaphyllus (0.100 or 10%) at the present study.
Populations of Tabebuia aurea were studied by Silva (2010) in Assis-SP, Pedregulho-SP, Selvíria-MS and Três Lagoas-MS, Brazil, who found a ST G mean of 0.025 or 2.5% and a mean ' ST G of 0.444 or 44.4%.Estimates above 10% for ' ST G were also found by Viegás (2009) between two populations of Myracrodruon urundeuva (Aramina-SP and Selvíria-MS, Brazil): 15.9% (pollen) and 23.5% (ovule).This high gene differentiation between populations was expected because of the geographic distance between them, obstructing the gene flow.Isolation and differing historical events in each population, therefore, result in higher gene differentiation, as opposed to geographic proximity which facilitates gene exchange (VIÉGAS, 2009), as is the statement of two populations of H. heptaphyllus in Botucatu region.

Seed collection for genetic resource development
One of the objectives of this study was to create conditions to support germplasm preservation programs for H. heptaphyllus, in Botucatu region, São Paulo, Brazil, area that is strongly affected by the impacts of forest fragmentation.According to Sebbenn (2002Sebbenn ( , 2003Sebbenn ( , 2006)), an estimated number of trees for seed collection intended for forest tree improvement would be 29 to 76, with an average of 45.Therefore, even if collecting from both populations, only 30 trees are found in the study region, which is below the average considered optimal and just one tree above the minimum considered reasonable.It gives an idea on deforestation level of the studied region, which makes seed collection virtually impossible for establishment of seed orchards, since that would require 44 to 114 unrelated trees, with an average of 67 parent trees.
To stablish seed orchard it is necessary a gene enrichment from the other preferably neighbor regions.

CONCLUSIONS
The populations of H. heptaphyllus have shown allelic polymorphism in all eight microsatellite loci.
The species genetic diversity is compatible with results from other tropical tree species.The existing genetic diversity is nonetheless suffi cient to support genetic improvement and germplasm conservation programs for the species.However, there are signs of inbreeding as a result of forest fragmentation in Botucatu region, Brazil.
High genetic variation was found within populations, suggesting that a larger sampling of individuals is required, however the divergence between populations cannot be ignored.

Table 2 -
Expected mean number of alleles per locus ( ), observed heterozygosity ( o H ), expected heterozygosity ( e H ) and fi xation index (F) for eight loci, in two populations of Handroanthus heptaphyllus.