Aphis craccivora, Koch

Nafría, Juan Manuel Nieto, Moreno-González, Víctor, Ortego, Jaime, Favret, Colin & Durante, M. Pilar Mier, 2022, Aphis mirabilis sp. n. (Hemiptera, Aphididae, Aphidina), a South American species resembling A. craccivora, and further data on A. tropaeoli, Zootaxa 5183 (1), pp. 22-48 : 35-37

3, i. Diversity of A. craccivora , especially in relation to populations living on Robinia pseudoacacia

Two nominal taxa included in the genus Aphis have been described from material collected on R. pseudoacaciae : A. robiniae Macchiati, 1885 and A. craccivora pseudoacaciae Takahashi, 1966 .

Aphis robiniae was described as a species with similarities to Myzus cerasi (Fabricius) and to A. laburni Kaltenbach ( Macchiati, 1885) . Myzus cerasi is a valid species and A. laburni is synonymous with A. cytisorum Hartig. Macchiati's description agrees with that of A. craccivora and also, but less clearly, with that of A. cytisorum , because the mention of M. cerasi reveals the bright black color in life and lack of waxy powder that are characteristic of this species. The types of A. robiniae are not known to be preserved. The name A. robiniae was used sporadically during the first two thirds of the 20th Century, for example Börner & Heinze (1957) (with the name Pergandeida (Doralis) robiniae ) considered it valid over A. leguminosae Theobald, 1915 , A. papilionacearum van der Goot, 1918 and A. cistiella Theobald, 1923 , and also the appropriate name for multiple previous citations of A. craccivora . They wrote that A. robiniae was present in Central Europe, but only as a vagrant from the Mediterranean region, and that it occurs in the subtropics and tropics of the Old World on numerous legumes. These authors thus did not consider R. pseudoacacia to be the only host of A. robiniae .

Host transfer experiments document the viability of A. craccivora on a broad range of hosts, including from R. pseudoacacia to M. sativa and vice versa ( Nevsky, 1929; Falk, 1958). Nevsky (1929) further suggested that A. robiniae may be synonymous with A. craccivora , a decision officially taken by Eastop & Hille Ris Lambers (1976). However, Zhang & Zhong (1981) considered A. robiniae valid, and even added to it the subspecies A. robiniae canavaliae Zhang , collected in Beijing and Henan ( China) on Canavalia ensiformis , Dolichos lablab (now Lablab purpureus ) and M. sativa . Remaudière & Remaudière (1997) considered both A. robiniae and its subspecies as synonyms of A. craccivora .

Aphis craccivora pseudoacaciae was described by Takahashi (1966) from specimens collected in Japan on R. pseudoacacia . He wrote that it is "similar to A. craccivora Koch , but differs in the longer setae on body and its appendages", from which it can be inferred that he saw no differences in either the dorsal sclerotization or the length of the processus terminalis in relation to the base of the antennal segment VI. This feature of setal length is the only character mentioned by Blackman & Eastop (2020) to separate the two subspecies from each other.

The recorded host plants of A. craccivora pseudoacaciae in the Palearctic Region are R. pseudoacacia , one species of Astragalus and three species of Vicia ( Holman 2009) . Subsequently , Stekolshchikov & Novgorodova (2015) cited this subspecies in the Republic of Altai (Russian Federation) on plants of four genera of Fabaceae , none of them Robinia . Apterae from Altai have long setae (in accordance with the original description of the subspecies), a wide and complete discal plate and a processus terminalis more than two times the length of the base of antennal segment VI (Stekolshchikov, email pers. com.). The setal lengths of apterous viviparae from our samples collected on R. pseudoacacia and on M. sativa (samples 5-12, Table 1 ) are variable and only partly overlap with those established by Takahashi (1966) for his subspecies. Based on these data, it is possible to think that A. craccivora pseudoacaciae could be synonymous with A. craccivora , but to ensure this, it will be necessary to examine the syntypes of the subspecies and study the variability of the length of the setae of apterae collected in Japan on various plants. In any case, based on the existing data, A. craccivora pseudacaciae is not restricted to R. pseudoacacia .

Wang et al. (2011) recognized population-based COI sequence differences in Chinese samples of A. craccivora , with some affinities for certain fabaceous hosts. Their results suggest the presence of host-adapted races in China. Specifically, 13 of the 14 samples collected on R. pseudoacacia were grouped on a single branch, mixed with 16 samples collected on other plant hosts. They did not report on the morphology of their specimens.

Wongsa et al. (2017) found that aphids identified as A. craccivora collected in Thailand are clustered on four separate branches based on their COI sequences, but it is important to note that among those examined, none were collected on R. pseudoacacia , and that there is no information on possible morphological differences between each cluster.

Mehrparvar et al. (2012) reported that A. craccivora specimens collected in Iran showed differences among themselves (the most informative for configuring groups are the lengths of the siphunculi and the ultimate rostral segment) that they considered to be related to the host plant. According to them, the best-defined group is the one formed by the specimens collected on R. pseudoacacia . However, Borowiak-Sobkowiak et al. (2017) found no differences in the estimated fertility in Polish populations of A. craccivora reared on R. pseudoacacia , Vicia faba and Vigna sesquipedalis , and they considered that the morphometric differences found by Mehrparvar et al. (2012) were the result of environmental conditions, especially regarding the nature of the host plant. It should be noted that in the study by Mehrparvar et al. (2012), several metric characteristics were used, but not setal length. In 19 apterous viviparous specimens of those collected on R. pseudoacacia : (a) discal plate is complete and wide; (b) siphunculi are long and cylindrical; (c) ratio of two parts of antennal segment VI is 1.62-2.33, usually more than 1.9; and (d) setae on abdominal segment 8 are approximately 1.3-1.6 times longer than the subarticular diameter of antennal segment III (Mehrparvar, email pers. com.). These metric data partially overlap with the data from Argentine A. craccivora collected on R. pseudoacacia , which are indistinguishable from those collected on M. sativa (see principal component analysis).

Brady & White (2013) reported that samples of A. craccivora collected on R. pseudoacacia and on M. sativa throughout central Kentucky ( USA) differ from each other by the presence of different facultative symbionts: Hamiltonella was found only in specimens from M. sativa , Arsenophonus only in specimens from R. pseudoacaciae . In a similar study, Chen et al. (2009) did not find either of those facultative symbionts in A. craccivora aphids on Vig. sesquipedalis in Taiwan.

We have not found any publication that records any peculiarity of populations of A. craccivora on R.pseudoacacia in North America, the native range of this tree.

The details described above confirm the wide variability of A. craccivora with regards to morphological characters, gene sequences and biological characteristics (host plants and endosymbions), and even suggests that a group of species may be hidden under that name. However, these details also confirm that there is no unequivocal relationship between the several variants or entities associated with R. pseudoacacia .

3, ii. General features of sexuals of A. craccivora

As is well known and already mentioned above, A. craccivora is usually anholocyclic. Sexuals have been described or recorded by: (1) Falk (1958), from specimens collected on R. pseudoacacia in the north of the former German Democratic Republic; (2) Basu et al. (1968), from specimens collected on Tinospora cordifolia (Menispermaceae) in West Bengal ( India); (3) La Rossa et al. (1993), from specimens collected on Gleditsia triacanthos (Fabaceae) and R. pseudoacacia in the province of San Luis ( Argentina); (4) Borowiak-Sobkowiak et al. (2017) in Poland without mention of host plant; and (5) Descamps et al. (2017) from specimens collected on Medicago sativa in the southeast of the province of Buenos Aires ( Argentina).

Oviparous females of A. craccivora have the usual general aspect of oviparae of other Aphis species, and males are winged. In contrast, males of the problem entity are wingless, and in addition, present differences in other qualitative and quantitative features relative to those of A. craccivora . One of the differential features is the ratio “processus terminalis / base of antennal segment VI”, which in males and oviparae of the problem entity is always much less than 2; in viviparae, only exceptionally is it equal to or greater than 2 times. Most of the native South American species of Aphis also have a processus terminalis that is less than 2 times the base of the antennal segment VI; exceptions are, for example, A. alstroemeriae , A. gaultheriae , A. luzuriagae or A. maulensis (Essig, 1953; Mier Durante et al., 2016; López Ciruelos et al., 2018).