New records of fruit trees as host for Neosilba species (Diptera, Lonchacidae) in southeast Brazil

Fruits of thirty-five cultivated native plant species (19 orders and 12 families) were sampled in farms of fruit production from two municipalities of São Paulo state, Brazil (January 2010 to March 2012) to evaluate species diversity of Neosilba flies. Thirty-one species of plants were the host for Neosilba species while four were not infested. Some aspects of the biology and patterns of species diversity, abundance, infestation rates, puparias viability and the interactions among species of frugivorous flies and their host plants were quantified. Seven species of Neosilba were reared: Neosilba bella Strikis & Prado (4 hosts), Neosilba certa (Walker) (4 hosts), Neosilba glaberrima (Wiedemann) (5 hosts), Neosilba inesperata Strikis & Prado (6 hosts) Neosilba pendula (Bezzi) (15 hosts), Neosilba pradoi Strikis & Lerena (8 hosts) and Neosilba zadolicha McAlpine (26 hosts). The association between the lance flies and the host fruit species is discussed.


Introduction
The Neotropical genus Neosilba McAlpine (Diptera: Lonchaeidae) is comprised of fly species whose larvae feeds on many species of commercially important fruit species (Araújo & Zucchi 2002, Strikis & Prado 2005, Bittencourt et al. 2006).The genus is restricted to the Neotropical region, being known from Caribe, Mexico, and Colombia to Brazil.Forty species were described so far and at least more 60 species are waiting for descriptions (McAlpine & Steyskal 1982, Strikis 2011;Galeano-Olaya & Canal 2012).
Although of the economic importance of some species of Neosilba that occur as pests on fruit and vegetables in several countries, such as Colombia (Steyskal 1978, Peñaranda et al. 1986), Peru (Korytkowski & Ojeda 1971), Costa Rica (Sánchez et al. 1991), and Brazil, the knowledge of these dipterans is still very scarce (Uchoa et al. 2002).
Studies on the biology and ecology of economically important species of fruit flies have contributed for the management and control of agricultural pests (Carey 1993, Vargas et al. 2001, Papadopoulos et al. 2002).Integrated pest management has been more difficult by the lack of basic studies on taxonomy, biology, and ecology.In Brazil, records on the genera of Lonchaeidae associated with fruits are scarce.Regional surveys are very important because they can provide basic information for managing insect pest populations and their natural enemies (Uchoa et al. 2002).
Our hypothesis is that there are new associations between Neosilba species and native host fruit species in the southeast of Brazil.
The aim of this study was to provide original information on fruit infestation rates, puparia's viability and interactions with species of host plants by Neosilba species in fruits sampled in the southeast of Brazil.

Material and Methods
Characterization of the study area.These orchards are located in an important fruit producing area in the state of São Paulo, Brazil.According to Setzer (1976), the regions are inserted in a climatic transition between very humid subtropical with marked dry seasons (Mu-Cw), with mean temperatures around 24ºC and minimum temperatures around 16ºC in the summer.Mean annual precipitation is about 1,300 mm/year and the predominant soil type in the studied areas is Latosol with good aeration, permeability and drainage.
Collecting of the Host Fruits.Fruits from 35 species (Table 1) were sampled from January 2010 to March 2012, in fruit farms from the municipalities of Campina do Monte Alegre (23º 53' 37" S, 48º 51' 06" W, 612m) (site 1), and Paraibuna (23º 27' 94" S, 45º 42' 88" W, 647m) (site 2).On a monthly basis fruits were collected from each studied area.The mature fruits were collected directly from the plant and the amount of  fruits collected in each sample varies depending on the availability of fruit at the time of sampling.Were collected about 50 fruits per specie plant on average.The samples (fruits individualized) were kept at room temperature and humidity, stored in plastic boxes (25 × 50 × 10 cm) sealed with nylon organza and lined with moist, autoclaved fine sand (± 2 cm layer) as a substrate for larval pupation.After the larvae had pupated 15 to 25 d after the material was brought in from the field, the substrate was sieved to collect the puparia, which were counted and transferred to emergence boxes kept under the same environmental conditions.Daily over a period of 40 d, substrate humidity was checked and the emergence of flies and parasitoids was monitored.The emerged parasitoids were fed with honey and water for 3 d, to fix the coloring that would allow their correct identification.They were then killed and preserved in 85% ethanol in labeled flasks for subsequent counting and species identification (Uchoa & Zucchi 1999).Fly Identification.The adults were identified in the Departamento de Biologia Animal, Instituto de Biologia, Universidade Estadual de Campinas (Unicamp), Campinas-SP, by the first author.Neosilba species were identified using keys and original descriptions (Korytkowski & Ojeda 1971, McAlpine & Steyskal 1982, Strikis 2011, Galeano-Olaya & Canal 2012).Only males were used since the group taxonomy is based on analysis of male genitalia (McAlpine & Steyskal 1982).Plant species were identified by botanists at the Departamento de Botânica, Universidade Estadual de Campinas (Unicamp), Campinas, SP, Brazil.Voucher specimens of the insects (stored in 85% alcohol) were deposited at Coleção Zoológica (ZUEC), Universidade Estadual de Campinas (UNICAMP), Campinas, SP, Brazil.
Quantitative variables, infestation indexes, and viability of the puparia.Two quantitative variables were evaluated: the number of puparia and of emerged flies.The fruit infestation levels were evaluated by two indexes: number of puparia/fruit, and the number of puparia/mass (g) of fruit.The viability of the puparia was calculated by the equation: %V = No. of ЕА/ PUP) X 100, where: %V = Percent of viability), No. of EA = number of emerged adults, and PUP =A total number of puparia; and the quotient was multiplied by 100.
Faunistic analysis.The faunistic analysis of Neosilba species was conducted according to Silveira-Neto et al. (1976).The following parameters were estimated: frequency, dominance, abundance, and constancy, using the Anafau software developed by the Departamento de Entomologia, Fitopatologia e Zoologia Agrícola, ESALQ/USP (Moraes et al. 2003): Frequency: Number of individuals of one species divided by the total number of individuals in the sample.
Dominance: The ratio given by the number of individuals of a given species divided by the number of individuals of all collected species.A species is considered dominant when its frequency is higher than 1/S.(S = total number of species in the community).Species classification according to dominance: Super-dominant (sd): number of individuals is higher than the upper limit of the 5% confidence interval (CI).Dominant (d): number of individuals is within the range of the 5% CI.Non-dominant (nd): number of individuals is lower than the lower limit of the 5% CI.
Abundance: Refers to the number of individuals of a given taxonomical category per unit of surface or volume.Species abundance was classified into five categories, as follows: Super-abundant (sa): number of individuals is higher than the upper limit of the 1% CI; Very abundant (va): number of individuals is situated between the upper limits of the 5% and 1% confidence intervals; Common (c): number of individuals is within the Biota Neotrop., 17(1): e20160213, 2017 http://www.scielo.br/bnhttp://dx.doi.org/10.1590/1676-0611-BN-2016-02135% CI range; Dispersed (d): number of individuals is situated between the lower limits of the 5% and 1% confidence intervals, and Rare (r): number of individuals is lower than the lower limit of the 1% CI.Constancy: Percentage of samples in which a given species is present.Species constancy was classified into three categories: Constant (w): when the species was present in more than 50% of collections; Accessory (y): when the species was present in 25% to 50% of collections; Accidental (z): when the species was present in less than 25% of the collections.Malpighiae marginata was the host fruit species with the highest Neosilba diversity: six species, being all of them found in this work, with the exception of N. certa.Strikis et al. (2011) also found no association record between N. certa and Malpighia sp. in his survey about frugivorous flies in the Amazon Rain Forest.P.cattleianum was the host with the largest number of Neosilba specimens, from which emerged 398 individuals (males + females).

Host
Infestation Indexes and Puparia's Viability.The mean of infestation was: 0.46 (puparia / total fruit) and 0.05 (puparia / fruit weight in grams).The mean of the viability of the puparia was 67.23%.C. brasiliensis was the plant species that showed the highest infestation indexes based on the number of fruit (2.30 puparia / fruit) and, B. crassifolia was the plant species presented the highest infestation indexes based on the mass of fruit (0.23 puparium/ mass in grams).
Faunistic analysis.The faunistica analysis is shown in Table 2.The Shannon diversity index (H), Equitability index (E) and Margalef (α) indexes were similar between the two areas.Neosilba zadolicha was the most frequent, most dominant and the most abudant specie.Neosilba bella was the less frequent specie.Faunistic indices classified all species as "accidental".(Table 2).

Discussion
Seven Neosilba species were recorded (Table 1), and this species richness is within the range reported in other inventories carried out in the state of São Paulo.In a previous study carried out in the municipality of Monte Alegre do Sul, Souza-Filho et al. (2009) reported eight species during a sampling period of one year.Silva et al. (2006) carried out a two-year study in two locations in the Southern Brazil, and reported five Neosilba species.In the south in the east-west direction across the state of Mato Grosso do Sul, nine species were registered (Nicácio & Uchoa 2011).
Neosilba zadolicha was the most predominant, frequent, dominant and abundant species indicating its importance in the region (Table 2).This species is a very common in citrus orchards reaching high rates of infestation (Uchoa et al. 2002, Raga et al. 2006, Raga et al. 2011).State of São Paulo has approximately 600,000 hectares of sweet oranges [Citrus sinensis (L.) Osbeck], with different varieties fruiting all year long.This fact may contribute to the abundance, dominance and constancy of N. zadolicha (Raga et al. 2015).
In our study, all species captured presented low constancy being considered accidental.This result suggests that adults of the low constancy species were not resident on the orchards, but they came from other hosts nearby the farm and/or the surrounding forest area.The sampled areas are surrounded by one of the few and largest remnants of the highly endangered mature coastal rainforest in Brazil (Faria et al. 2006).The Brazilian Atlantic rainforest is considered one of the richest biomes on earth, and southwest São Paulo harbors high species richness, high levels of endemism and local sites of diversity of trees in families that comprise species which are known hosts of Neosilba, such as Fabaceae, Malpighiaceae, Myrtaceae, Rutaceae, and Sapotaceae (Thomas et al. 1998, Faria et al. 2006, Martini et al. 2007).Thus, the forest areas surrounding the orchards can provide an important reservoir for lonchaeid populations that probably migrate to the orchards.The movement of fruit flies from the adjacent native vegetation, particularly forest fragments, into orchards was demonstrated by Vargas et al. (2001), and Kovaleski et al. (1999).Uchoa et al. (2002) found Neosilba species infesting fruits of C. sessiflora and M. jaboticaba in the cerrado of Mato Grosso do Sul, Brazil, and Strikis et al. (2011) found N. glaberrima and N. zadolicha in P. macrophylla fruits, but in this paper, we did not record these associations.Many different biotic and abiotic stimuli can account for the presence of the lesser abundant fruit flies species in environments that do not provide optimal host plants, such as commercial orchards (Aluja et al 1996).The authors suggest that the odor of ripening fruit, shelter conditions of perennial trees, and emission of volatiles by certain tree species that are similar to those found in the sexual pheromones of fruit flies could draw adult fruit flies into the orchard.Adaime et al. (2012), recorded N. bella in B. crassifolia in the Amazon region.Bittencourt et al. (2013) reared N. bella from E. stipitata fruits in northeastern Brazil.In previous surveys, N. bella have been found in twelve plant species from nine plant families (Table 2).
According to Bittencourt et al. (2013), N. bella has a wide geographical distribution in Brazil, ranging from Atlantic Forest, Amazon Rain Forest, and Cerrado.Its plasticity in occupying such different biomes, and attacking different host plants, makes this species a candidate in becoming an important pest, once it is found in environments occupied by crops plantation, especially coffee crops.However, N. bella was one of the species with the lowest number of host fruit recorded in this research.
McAlpine & Steyskal (1982) also found N. certa in I. vera, and Souza-Filho et al. (2009) recorded this species in orchards of P. guajava.N. certa has already been registered in fifty host fruits, belonging to seventeen plant families ( N. inesperata had already been reported in twenty-eight host fruits in sixteen plant families (Table 2).This species was previously reported by Nicácio & Uchoa (2011) in P. guajava.Raga et al. (2015) also found this species in M. emarginata.
In the literature N. pendula had already been reported infesting fifty-five host fruit species from twenty-five plant families.Seventeen plant species from nine families have been reported previously as hosts of N. pradoi (Table 1).N. zadolicha has already been recorded in ninety-four host species from thirdy-two plant families (Table 1).
Malpighia emarginata was the host fruit species with the highest Neosilba diversity: six species, being all of them found in this work, with the exception of N. certa.Strikis et al. (2011) also found no association record between N. certa and Malpighia sp. in his survey about frugivorous flies in the Amazon Rain Forest.P. cattleianum was the host with the largest number of Neosilba specimens, from which emerged 398 individuals (males + females).Adaime et al. (2012) found lower infestation indexes to that found in this work: 0.06 puparia/fruit to N. bella hosted in B. crassifolia.Aguiar-Menezes et al. (2004) reported to N. zadolicha hosted in P. alata infestation index (2.1 puparia/fruit) higher than all indexes found herein.Souza et al. (2012) reported infestation indexes to Neosilba species hosted in P. guajava (0.03 puparia/ g of fruit) similar than mass fruit infestation that was found here (0.01 puparia/ g of fruit).However to Neosilba species hosted in Z. joazeiro, Souza et al. (2012) found higher infestation indexes (3.28 puparia/ g of fruit) than the herein obtained (0.09 puparia/ g of fruit).
c Number of puparia/weight of fruits (in grams).

Plant Taxa Neosilba male species Indices/Rate Collecting sites Nº of fruits Mass of fruits (g) Pupae (n) Emerged adults N. bella N. certa N. glaberrima N. inesperata N. pendula N. pradoi N. zadolicha Total fruit infestation b Fruit infestation c Puparia viability (%)
a Site 1: Campina do Monte Alegre; site 2: Paraibúna (São Paulo, Brazil); b Number of puparia/number of fruit; c Number of puparia/weight of fruits (in grams).
N. certa (90 individuals) also had four species of host fruits, occurring in three plant family: Spondias mombin L. (Anacardiaceae), Inga vera N