Over the past century, the Brazilian Atlantic forest has been reduced to small, isolated fragments of forest. Reproductive isolation theories predict a loss of genetic diversity and increases in inbreeding and spatial genetic structure (SGS) in such populations. We analysed eight microsatellite loci to investigate the pollen and seed dispersal patterns, genetic diversity, inbreeding and SGS of the tropical tree Copaifera langsdorffii in a small (4.8 ha), isolated population. All 112 adult trees and 128 seedlings found in the stand were sampled, mapped and genotyped. Seedlings had significantly lower levels of genetic diversity (A ¼ 16.5 ± 0.45, mean±95% s.e.; H e ¼ 0.838±0.006) than did adult trees (A ¼ 23.2±0.81; H e ¼ 0.893±0.030). Parentage analysis did not indicate any seed immigration (m seeds ¼ 0) and the pollen immigration rate was very low (m pollen ¼ 0.047). The average distance of realized pollen dispersal within the stand was 94 m, with 81% of the pollen travelling o150 m. A significant negative correlation was found between the frequency and distance of pollen dispersal (r ¼ À0.79, Po0.01), indicating that short-distance pollinations were more frequent. A significant SGS for both adults (B50 m) and seedlings (B20 m) was also found, indicating that most of the seeds were dispersed over short distances. The results suggested that the spatial isolation of populations by habitat fragmentation can restrict seed and pollen gene flow, increase SGS and affect the genetic diversity of future generations.

Understanding the coancestry coefficient within openpollinated progenies has long been an area of interest because of the implications of coancestry on estimates of additive genetic variation, variance effective size and the number of seed trees required for seed collection for ex situ conservation, tree breeding and environmental restoration. This study compares three methods to calculate the coancestry coefficient within open-pollinated progenies of the dioecious tree species, Myracrodruon urundeuva, using six microsatellite loci. The methods compared were: i) correlated mating model (CMM) (RITLAND, 1989); ii) TWOGENER method (SMOUSE et al., 2001) to estimate the differentiation among pollen pools (TGM); and iii) HARDY et al.’s 2004 method using the estimate of coancestry from LOISELLE et al. (1995) (HLM) and from RITLAND (1996) (HRM). The data analysis was based on four data sets: two populations were composed of 12 progenies, two of 24 progenies, and all progenies consisted of 15 plants. The coancestry estimated using CMM ranged among populations from 0.145 to 0.158, using TGM it ranged from 0.153 to 0.181, using HLM from 0.153 to 0.162, and HRM from 0.144 to 0.147. To investigate the bias of the estimates of true relatedness within progenies we simulated two half-sib and two full-sib populations. The most accurate method found in the study was CMM because the estimated values presented no bias for true half- or full-sib progenies and these values were very similar to those expected (0.125 and 0.25, respectively). These results have significant implications for breeding and conservation programs because coancestry-within-progenies is a key parameter in assessing the variance effective size.

-Mating system in Myracrodruon urundeuva (Anarcadiaceae): implications for conservation genetics. The aims of this study were to investigate the mating system of a fragmented population of the dioecious tropical tree Myracrodruon urundeuva Allemão, using five microsatellite loci and the mixed mating and correlated mating models. The study was conducted in the Estação Ecológica de Paulo de Farias (436 ha), where the population occupies about 142 ha. The mating system was estimated using 514 open-pollinated offspring, collected from 30 seed-trees. Estimates of the multilocus outcrossing rate confirm that the species is dioecious (t m = 1.0). Low levels of mating among relatives were detected in the population (1 -t s = 0.020). The estimate of paternity correlation (r p(m) ) indicated that offsprings were composed of mixtures of half-sibs and full-sibs, with the latter occurring at a low frequency (average of 0.148). The estimated coancestry coefficient within families (Θ = 0.147) was larger and the effective population size (N e(v) ) was lower (N e(v) = 2.98) than expected in progenies from panmictic populations (Θ = 0.125, N e(v) = 4, respectively). In terms of conservation, the results indicate that to retain an effective population size of 150, is necessary to collect seeds from at least 50 seed-trees.Key words -correlated matings, fragmentation, microsatellite, variance effective population size RESUMO -Sistema de reprodução em Myracrodruon urundeuva (Anarcadiaceae): implicações para a conservação genética. O objetivo deste estudo foi investigar o sistema de reprodução de uma população fragmentada da espécie arbórea dióica, Myracrodruon urundeuva Allemão, utilizando cinco locos microssatélites e os modelos mistos de reprodução e de cruzamentos correlacionados. O estudo foi conduzido na Estação Ecológica de Paulo de Faria (436 ha), onde a população ocupa uma área aproximada de 142 ha. Foram amostradas 514 progênies de polinização aberta coletadas de 30 árvores matrizes da população. A estimativa da taxa de cruzamento multilocus confirma que a espécie é dióica (t m = 1,0). Baixos níveis de cruzamentos entre parentes foram detectados na população (1 -t s = 0,020). A estimativa da correlação de paternidade (r p(m) ) indicou que as progênies são compostas por misturas de meios-irmãos e irmãos-completos, embora este último ocorra em menor proporção (média de 0,148). O coeficiente de coancestria dentro das progênies (Θ = 0,147) foi maior e o tamanho efetivo de variância (N e(v) = 2,98) foi menor do que o esperado em progênies de populações panmíticas (Θ = 0,125, N e(v) = 4, respectivamente). Em termos de conservação genética, os resultados indicam que para se obterem lotes de sementes com tamanho efetivo de referência de 150, é necessário coletar sementes de pelo menos 50 árvores matrizes.Palavras-chave -cruzamentos correlacionados, fragmentação, microssatélites, tamanho efetivo populacional 1.Parte da tese de doutorado da primeira autora,

RESUMO.A mancha areolada de Thanatephorus é uma das doenças mais importantes da seringueira (Hevea brasiliensis) na Amazônia brasileira. Além da seringueira, o fungo T. cucumeris (fase anamórfica Rhizoctonia solani) causa doenças foliares em outras espécies nativas ou cultivadas na região. Baseando-se na ausência de informações sobre quais os grupos de anastomose (AG) de R. solani estão associados à seringueira e a outros hospedeiros na Amazônia, testou-se a hipótese de que estes isolados pertencem a AG distintos. Também não há informação sobre a patogenicidade cruzada, à seringueira, de isolados provenientes de outros hospedeiros. A determinação do AG foi efetuada com base em análise filogenética da região ITS do rDNA. Concluiu-se que um novo subgrupo de R. solani (o AG-2-2 Hb) é o principal agente causal da mancha areolada em seringueira. Entretanto, isolados de outros AGs (AG-1 IB e ID, AG-4 HGI) e de Ceratobasidium sp. (Rhizoctonia sp. binucleada AG-R), oriundos de hospedeiros distintos, também infectam seringueira. Isto implica diversidade de fonte de inóculo para a sobrevivência e disseminação dos patógenos. De maneira inédita se relata a ocorrência de um novo AG de R. solani infectando citros no Acre, Brasil, distinto dos demais relatados no mundo (provavelmente AG-14).Palavras-chave: Rhizoctonia solani, grupamento de anastomose, patogenicidade cruzada, seringueira, citrus.ABSTRACT. Phylogenetic inference reveals the complex etiology of the target and leaf spot diseases on rubber tree and other species cultivated in the Amazon. Thanatephorus target spot is one of the most important diseases of the rubber tree (Hevea brasiliensis) in the Amazon region of Brazil. In addition to rubber tree, the fungus Thanatephorus cucumeris (anamorphase = Rhizoctonia solani) causes several foliar diseases on other agricultural crops or native species in the region. Based on the lack of information on which anastomosis groups (AG) of R. solani are associated with rubber tree and other hosts in the Amazon, we tested the hypothesis that these isolates belong to distinct AGs. There is no information about the cross pathogenicity (to rubber tree) of isolates from other hosts. The AG was determined based on phylogenetic analyses of the ITS region of the rDNA. We concluded that a new subgroup of R. solani (the AG-2-2 Hb) is the main pathogen causing the rubber tree leaf spot. However, isolates from other AGs (AG-1 IB and ID, AG-4 HGI) and from Ceratobasidium sp. (binucleate Rhizoctonia sp. AG-R) from distinct hosts can also infect the rubber tree. This implicates in diversity of inoculum sources for the survival and spread of the leaf spot pathogens. For the very first time, we report the occurrence of a new AG of R. solani infecting citrus in Acre, Brazil, which is distinct from all the others already described worldwide (probably AG-14).