Hosts of Raoiella indica Hirst ( Acari : Tenuipalpidae ) Native to the Brazilian Amazon

The expansion of red palm mite (RPM), Raoiella indica (Acari: Tenuipalpidae) in Brazil could impact negatively the native plant species, especially of the family Arecaceae. To determine which species could be at risk, we investigated the development and reproductive potential of R. indica on 19 plant species including 13 native species to the Brazilian Amazon (12 Arecaceae and one Heliconiaceae), and six exotic species, four Arecaceae, a Musaceae and a Zingiberaceae. Values of the instantaneous rate of increase (ri) were initially estimated at 7, 14, 21 and 28 days after infestation of each species. Higher values of ri (> 0.05) were determined on the Arecaceae Adonidia merrillii, Astrocaryum jauari, Cocos nucifera, Bactris simplicifrons, Mauritia flexuosa, Phoenix dactylifera and Socratea exorrhiza, and on the Heliconiaceae Heliconia psittacorum Sassy; these were classified as “potential primary hosts”. Lower, but still positive values of ri (0-0.05) were determined on the Arecaceae Bactris maraja, Oenocarpus bacaba, Oenocarpus bataua and on the Musaceae Musa × paradisiaca (Prata variety); these were classified as “potential secondary hosts”. Negative values of ri were determined for the remaining plants, i.e., the Arecaceae Astrocaryum aculeatum, Attalea maripa, Bactris gasipaes, Elaeis guineensis, Euterpe oleracea, Euterpe precatoria, and the Zingiberaceae Alpinia rosea; these were considered “non-hosts”. Species with ri < 0.05 were considered not to be threatened by the RPM. Biological parameters of RPM were evaluated on the plant species with positive ri (except B. maraja) and two native species with negative ri (E. oleracea and E. precatoria). Mean developmental time ranged from 14.7 days on C. nucifera to 21.4 days on Musa × paradisiaca, showing a significant influence of the plant substrate. Immature viability, oviposition rate, net reproductive rate (R0) and intrinsic rate of increase (rm) were affected by the plant species.

The objectives of this study were to determine whether selected plants of families Arecaceae, Heliconiaceae, Musaceae and Zingiberaceae commonly found in the Brazilian Amazon could serve as hosts for R. indica, including some species previously reported as hosts (H.psittacorum and the Arecaceae A. maripa, B. gasipaes, E. oleracea, E. precatoria and M. flexuosa), to confirm previous results indicating those plants as potential hosts, given the very common occurrence of these plants in northern Amazon.

RPM Stock Colony
Specimens of RPM were obtained from C. nucifera plants in Boa Vista (02º45′29.2″N60º43′54.7″W).Adults were transferred to leaflets of A. merrillii placed (upside down) onto a piece of foam map moisten daily with distilled water.Mites were weekly transferred to new leaflets.The stock colony was maintained in Entomology Laboratory of Embrapa Roraima at uncontrolled environmental conditions (ca. 26 ºC, 60% RH and natural photophase (ca.12:12 h L: D).Twenty-one day before of the study, 100 adult females were transferred to a new A. merrillii leaflet and removed 24 h later, leaving only the eggs they laid.This new generation was reared to adulthood to be used in the determination of the instantaneous rate of increase (r i ).

Instantaneous Rate of Increase
The experiment was conducted in a screenhouse at Embrapa Roraima.It was conducted under a randomized block design, with five blocks and 19 plant species as treatments, each experimental unit corresponding to a plant.Three recently molted adult females of R. indica were transferred to a rectangle of 1.5 × 3.0 cm delineated with entomological glue onto the lower surface of a randomly selected leaf or leaflet (according to plant species) of each third (basal, median and apical) of each seedling.The number of mites in each developmental stage in each rectangle was counted 7, 14, 21 and 28 days later.The instantaneous rate of increase (r i ) of R. indica on each plant species was estimated using the following equation proposed by Hall (1964): Where, n f is the final number of mites, n o is the initial number of mites, and ∆ t is the elapsed time.Mean values were calculated as follows: Where, r i7 , r i14 , r i21 and r i28 are respectively the instantaneous rate of increase for each species 7, 14, 21 and 28 days after mite confinement.Using a modification of the criterium proposed by Stark, Tanigoshi, Bounfour and Antonelli (1997) and Walthall and Stark (1997), the suitability of plant species as host estimated by the determined values of r i .The plant was classified as "primary or potential primary host" if r i > 0.05, as "potential secondary host" if r i = 0.05-0.00,and as "non-host" if r i < 0.00.

Biological Parameters
Developmental time, oviposition and daily survivorship rates of R. indica were determined on the following plant species on which r i was positive: A. The study was initiated with recently laid eggs.For such, three R. indica adult females were transferred from the stock colony to rectangles (2.0 × 3.0 cm) delimited with entomological glue on the undersurface of undetached leaflets (one on each of seven leaflets of composite leaves) or leaves (seven rectangles on a single non-composite leaf) at taken from the median stratum of the plant canopy.Twenty-four hours later, the females and part of the eggs laid were removed, leaving one to five eggs per rectangle.The units were examined once a day for counting the number of mites in each stage.Newly emerged females were transferred to new similar experimental units, adding an adult male taken from the stock colony to each unit.Dead males were replaced by new ones until the female died.Sex ratio was estimated based on offspring obtained in the first 10 days of oviposition (Schulten, 1985).
All evaluations for the determination of r i and for the study of the biology of R. indica were carried out using 30× or 60× hand lens (LED HG498).The environmental conditions were recorded every 10 minutes using a Hobo Data Loggers ® .

Statistical Analysis
Values of r i , viability of immature forms, developmental time, oviposition period, sex ratio, longevity and daily oviposition rates were subjected to an analysis of variance (PROC GML), using the statistical package SAS version 9.0 (SAS Institute, 2003, 2008), after testing the data for normality and homoscedasticity (Kolmogorov and Bartlett tests, respectively).Data that could not be normalized were subjected to Kruskal-Wallis non-parametric test (Proc NPAR1WAY), SAS version 9.0 (SAS Institute, 2003, 2008).Fertility life table parameters [net reproductive rate (Ro), generation time (T), and intrinsic rate of increase (r m )] were estimated using the Jackknife procedure, as described by Maia, Luiz and Campanhola (2000), using SAS/Enterprise Guide 4.3 (SAS Institute, 2003, 2008).Pearson correlation between r i and r m was performed (SAS/Enterprise Guide 4.3, SAS Institute, 2003Institute, , 2008)).
Net reproductive rate (R 0 ) and intrinsic rate of increase (r m ) were also affected by plant species (p < 0.05) (Table 4).The highest R 0 was on P. dactylifera (14.7) and the lowest on O. bataua (0.9), while the highest r m was on A. merrilliii, C. nucifera and S. exorrrhiza (0.094-0.103) and the lowest on O. bataua (-0.002).Mean generation time (T) was also affected by host (p < 0.05), being shorter on A. merrillii (20.3 days) and longer on O. bataua (36.1 days) (Table 4).A significant and positive correlation was observed between r i and r m (n = 33; r = 0.62; p < 0.0001).

Discussion
Estimated instantaneous rates of increase (r i ) showed that RPM can colonize several of the evaluated plant species, both native and introduced, suggesting the ability to maintain itself in the Amazon region on wild and cultivated plants.Obviously this does not mean that the mite actually threatens those plants in nature, given that prevailing climatic factors as well as natural enemies and other biotic factors may restrain RPM from reaching high populations.
The trend for the reduction of r i values along the duration of the experiment in several of the evaluated plant species might be spurious, being conceivably due to the progressive reduction of the suitability of the plant substrate with the increasing mite population in a restricted space (Reddall, Sadras, Wilson, & Gregg, 2004), which could lead to increasing intraspecific competition for food and space (Elkinton, 2009).Partial support to this hypothesis is the fact that for plants with lower r i value, this parameter remained about constant (ex.O. bataua) or tended to increase (on M. paradisiaca and O. bacaba) during the experiment.In about half of the plants tested, RPM population decreased through time (A. aculeatum, A. maripa, A. rosea, B. gasipae, B. maraja, E. guineensis, E. oleraceae, E. precatoria and O. bataua), leading to the conclusion that these species are not threatened by the expanding geographic distribution of RPM in Brazil.Note. a Number of mite specimens.Within a column, means followed by the same letter are not significantly different (Kruskal-Wallis test: p > 0.05).
Table 4. Net reproductive rate (R 0 ), intrinsic rate of increase (r m ) and mean generation time (T) (mean ± CI) of Raoiella indica on different plant species.Boa Vista, Roraima, Brazil.T = 30±4,61 ºC, RH = 69±15,04%, natural light: darkness (ca.12:12 h) The reasons for those different levels of susceptibility are not known, but could be due to plant defense factors (Painter, 1958;Price, 1986;Stout, 2013).According to Isman and Grieneisen (2014) plants developed complex chemical arsenals (terpenoids, flavonoids, tannins and alkaloids) with phytochemical activities that can act as deterrents to feeding or oviposition, reducing the performance of herbivores (Novaes, Molinillo, Varela, & Macías, 2013;Pavela, 2010;Rattan, 2010).Variability of leaf structures can also provide mechanical or physiological characteristics that can interfere with the ability of a particular mite species to use a plant as host.
In palm trees, such structures can be scales of the abaxial leaf surface, presence of epicuticular wax, as well as the presence of trichomes in different densities (Beard et al., 2012;Horn, Fisher, Tomlinson, Lewis, & Laubengayer, 2009;Valverde, Fornoni, & Núñez-Farfán, 2001).Another important aspect concerning RPM refers to the characteristics of the stomata.Beard et al. (2012) mentioned that this mite feeds by piercing cells inside the stomata.These authors showed that feeding by RPM is greatly influenced by the behavior of the stomata, and that Raoiella species cannot feed on hosts with closed or covered stomatal openings.
Results of the present study suggest that native Brazilian plant species (A.jauari, B. simplicifrons, M. flexuosa, O. bacaba, S. exorrhiza and H. psittacorum) could be at risk with the expansion of the distribution of RPM, given that the values of r i for those species are similar to the values obtained for well known exotic hosts, as A. merrillii and C. nucifera (Carrillo et al., 2012;Vásquez et al., 2015).However, quite different life table parameters were found in this and previous studies.The value of r m was much higher on A. merrillii in this study than reported by Vásquez et al. (2015) at a slightly lower temperature (29 o C; r m = 0.042, against 0.1 in this study).This difference could be due to the fact that in the present study the rearing units were established on undetached leaves, which would supposedly be more adequate for not promoting the closure of the stomata (Beard et al., 2012).However, quite the opposite was found on C. nucifera, for which those authors reported r m of 0.166 (against 0.097 in this study).Thus, detachment of leaves may disturb RPM performance, but obviously other factors might have important bearing as well.In any case, in both studies C. nucifera and A. merrillii were the hosts on which R. indica best performed.
The intrinsic rate of natural increase (r m ) value is used for revealing the impact of a parameter (e.g.temperature, host plant) on the demographic potential of an organism (Birch, 1948;Dent, 1997).Under optimum conditions, r m will attain its maximum possible value (Dent, 1997).Meanwhile, the instantaneous rate of increase (r i ) measures the ability of a population to increase over time (Hall, 1964).Walthall and Stark (1997) considered that r i provides information on short-term and real populations are evaluated in shorter time and at less expense than the traditional life table method.These authors obtained high correlation between these two demographic parameters (r 2 = 0.91, n = 21, p < 0.01) for populations of pea aphid, Acyrthosiphon pisum (Harris) (Hemiptera: Aphididae), used as the model species and exposed to the nicotinergic insecticide imidacloprid, as xenobiotic sprayed on potted broad bean, Vicia faba L. (variety Banner), suggesting that r i can be used as a substitute measure for r m .The r i has been used in the assessment of xenobiotics, especially concentration-response laboratory bioassays (Stark et al., 1997;Tuelher, Venzon, Guedes, & Pallini, 2014;Walthall & Stark, 1997).
All native plant species classified as potential primary hosts of RPM in this study (A. jauari, B. simplicifrons, M. flexuosa, O. bacaba, S. exorrhiza and H. psittacorum) have great relevance to indigenous people, for extractivism and/ or maintenance of wildlife of the Amazon biome, providing important elements as food, construction items, shelter, medicine, handcraft raw material, fibers etc. (Mesa & Galeano, 2013;Paniagua-Zambrana et al., 2007;Sosnowska & Balslev, 2009).Some of the palm species, as S. exorrhiza, are among the 20 most abundant plants in the Amazon biome (Steege et al., 2013).Many of these palms are found in conservation areas, where mite control is difficult to apply.The conceivable significant role of prevailing natural enemies on RPM population on those plants remains to be demonstrated.These could eventually prevent RPM from causing significant damage to those hosts.The study of these and related aspects is warranted.
In conclusion, the red palm mite has the potential to cause severe damage to plants native to the Brazilian Amazon.Eight species were concluded in this study to be potential primary hosts, of which five are native to that biome, namely the Arecaceae A. jauari, B. simplicifrons, M. flexuosa and S. exorrhiza, and the Heliconiaceae H. psittacorum.This seems relevant, because all of these species are important for extractivism by indigenous people and/or for maintenance of wildlife.
on two native species on which r i was negative (E. precatoria and E. oleracea); the latter two plants were included because of their economic and ecological importance in the Amazon region.A completely randomized design was adopted, using five seedlings of A. merrillii, E. precatoria, E. oleraceae, M. flexuosa, O. bacaba, O. bataua, P. dactylifera and Musa × paradisiaca and 8-10 seedlings of A. jauari, B. simplicifrons, C. nucifera, S. exorrhiza and H. psittacorum (numbers varying according to the architecture/structure of each plant).
days)Note.a Number of mite specimens.Within each column, means followed by the same letter are not significantly different (T test: P > 0.05).