Hoplitis (Tkalcua) zandeni (Teunissen & van Achterberg 1992 )

Müller, Andreas & Mauss, Volker, 2016, Palaearctic Hoplitis bees of the subgenera Formicapis and Tkalcua (Megachilidae, Osmiini): biology, taxonomy and key to species, Zootaxa 4127 (1), pp. 105-120 : 114-116

Hoplitis (Tkalcua) zandeni (Teunissen & van Achterberg 1992)

Osmia zandeni Teunissen & van Achterberg 1992: 313 . Type material: Holotype Ƌ, “Puerto del Rosario, Fuerteventura, 4.– 16.3.1984 ” ( Spain: Canary Islands), Naturalis Biodiversity Center, Leiden.

Osmia (Microhoplitis) hohmanni Tkalcu 1993: 812 . Type material: Holotype Ƌ, “Fuerteventura, Risco del Paso, 1.5.1988 ” ( Spain: Canary Islands), La Roche Collection. Type species of Microhoplitis Tkalcu. Synonymy in Zanden (1995).

Literature records. SPAIN, Canary Islands: Teunissen & van Achterberg (1992), Hohmann et al. (1993) and Tkalcu (1993) list numerous records of H. zandeni from Fuerteventura and Lanzarote. These records originate mainly from coastal areas not exceeding 150m a.s.l. with the exception of two records at altitudes of 250m and 400m, respectively, and encompass a time period that lasts from the beginning of February to the beginning of May.

New records. SPAIN, Canary Islands: Fuerteventura, Montaña Pelada SW Costa Calma, 29.3.– 5.4.2015 (leg. A. Müller & V. Mauss); Fuerteventura, Istmo de la Pared NW Costa Calma, 4.4.2015 (leg. A. Müller & V. Mauss).

Distribution. Endemic on the two neighbouring Canary Islands Fuerteventura and Lanzarote.

Habitat. Semidesertic areas on sandy ground with scattered shrubs of Salsola divaricata (Amaranthaceae) , Launaea arborescens (Asteraceae) and Zygophyllum fontanesii (Zygophyllaceae) ( Figs. 9, 10 ).

Pollen hosts. Polylectic ( Tab. 1 ). Pollen sources as deduced from field observations and the analysis of both female pollen loads and brood cell contents at the study sites on Fuerteventura were different species of Fabaceae (e.g. Astragalus hamosus , Lotus glinoides , L. lancerottensis , Ononis natrix ), Reseda lancerotae (Resedaceae) , Frankenia laevis (Frankeniaceae) , Heliotropium erosum (Boraginaceae) , Helianthemum canariense (Cistaceae) and Kickxia sagittata (Plantaginaceae) . The females collected pollen from the flowers of Reseda , Frankenia and Helianthemum by rapidly seesawing their metasomal scopa against the anthers, whereas pollen of Heliotropium was removed out of the narrow flower tubes with alternating up and down movements of the forelegs; in contrast to other bee species that collect Heliotropium pollen, the females of H. zandeni are not equipped with specialized bristles to extract pollen from the hidden anthers (see Gotlieb et al. 2014). Interestingly, flowers of Asteraceae do not seem to be exploited for pollen, although several Asteraceae species were flowering at the study sites; thus, the observation of H. zandeni on flowers of the Asteraceae species Launaea arborescens and Phagnalon sp. ( Hohmann et al. 1993) probably refers to individuals that sucked nectar rather than collected pollen.

Nesting biology. Nesting site: All 19 nests detected near Costa Calma on Fuerteventura from 29 March to 5 April 2015 were in snail shells of Theba geminata (Mousson) . Nests were found in both juvenile and adult shells varying in maximal diameter from 11mm to 15mm. All shells selected by the female bees for nesting lay openly on the ground and were not hidden e.g. among vegetation or under stones. Their position on the ground varied with some shells laying on the umbilicus and others on the apex. As the females were never observed to move nor turn the shells during any phase of the nesting cycle, the free accessibility of the shell opening appears to be an important condition for nest site selection. A second condition pertained to the degree to which the shells were filled with sand. Due to the sandy ground and the permanent wind prevailing at the study sites, almost all shells of T. geminata were more or less filled with sand. Females searching for nesting sites were repeatedly observed to scratch sand out of the shell or even to burrow through sand masses blocking the shell entrance, probably to check whether the inner whorls of the shell are accessible for nesting. Shells with their openings being blocked with hardpacked sand or with their inner whorls being densely filled with sand were never chosen as nesting sites. Nesting cycle: After selection of a suitable snail shell for nesting, the females usually removed sand from the shell entrance to get free access to the inner shell whorls. In one case, this cleaning process lasted 8min. The shells, however, were never completely cleaned from sand with substantial amounts of sand remaining in the entrance area ( Fig. 11 ). Brood cell construction started with the collection of chewed leaf tissue (“leaf pulp”) on both green and yellowed leaves of Helianthemum canariense and possibly also other plant species ( Fig. 17 ). In contrast to other snail shell nesting osmiine bees, e.g. Osmia species of the subgenera Neosmia, Allosmia or Helicosmia ( Müller 2016), H. zandeni did not glue leaf pulp onto the shell surface. Instead, the collected leaf pulp was used i) to tightly fix sand grains still present in the area of the later brood cell to the shell wall and ii) to build - under addition of many sand grains - a narrow and crescent-shaped brink of varying width along the columella of the shell ( Fig. 15 ). The construction of such thresholds, which define the length of the brood cell in advance and are later expanded to the partition that seals the brood cell ( Fig. 16 ), is widespread among osmiine bees (Philipps & Klostermeyer 1978; Frohlich 1983; Müller 1994). In one case, H. zandeni needed 10 flights with nest building material to construct the threshold. The provenance of the numerous sand grains glued together with leaf pulp for the construction of thresholds and cell partitions is not fully clear. Careful examination of photos of females returning from leaf pulp collection flights revealed that sand grains adhered to the leaf pulp packets held by the females in their mandibles. This suggests that the females landed on the ground immediately after having chewed leaf pulp from the Helianthemum leaves, dipped the moist leaf pulp in the sand and transported a mixture of both leaf pulp and sand grains back to their nest. It can not be excluded, however, that part of the sand grains used for nest construction originated from sand remains in the shell entrance area itself. After the threshold had been finalized, the brood cell was provisioned with nectar and pollen. Females returning from provisioning flights, which lasted between 8min and 20min, entered the nest head first to regurgitate nectar before they came out, turned around at the nest entrance ( Fig. 12 ) and entered the nest metasoma first to comb pollen out of the scopa. In larger shells, the female turned around inside the shell rather than at the shell opening. Immediately after egg deposition the threshold was expanded to a cell partition. If a second brood cell was built in front of the first, the cell partition was of moderate and uniform width of 0.75–1.25mm. If, however, no additional cell was constructed, the threshold was expanded to a thick nest plug, which had a maximal width of 2–4.5mm along the shell wall and of 1–2mm in its centre ( Fig. 13 ). The inner side of the nest plug consisted of sand grains irregularly glued together ( Fig. 14 ), whereas the outer side was carefully worked with sand grains and leaf pulp forming a plane wall ( Fig. 16 ). In one case, a female needed 37 flights with nesting material to expand the threshold to a nest plug. After completion of the nest plug, the females left the nest and started searching for another shell to construct additional brood cells. In contrast to other snail shell nesting osmiine bees, e.g. Osmia species of the subgenera Allosmia, Hoplosmia or Neosmia ( Müller 2016), completed nests were neither transported to a safe place nor buried into the ground nor turned in a protected position nor covered with plant material. In one case, a female selected a shell for nesting at 1315pm and left the completed nest containing a single brood cell at 1610pm the following day, suggesting that the construction of a one-celled nest usually lasts about one day under favourable conditions. Nest architecture: Of the 19 nests examined, 10 contained a single brood cell ( Figs. 13, 14 ), 6 contained two cells and 2 had an empty cell in front of the first brood cell containing moderate amounts of sand (the number of brood cells in one nest could not be assessed as the female had just started to provison her first cell). Interestingly, the nest plug in H. zandeni was never built at the shell opening as is the case in all osmiine bee species, which use snail shells as exclusive nesting sites (A. Müller unpublished). Instead, it was built 1/3 to 1 whorl behind the shell opening and thus was usually not visible from the outside. The fact that the eumenine species Leptochilus fortunatus Blüthgen , which also nested in shells of Theba geminata at the study sites, built the nest plug deep inside the shells like H. zandeni suggests that habitat conditions might underlie this unusual behaviour in either species. One possible explanation is that plug construction at the shell opening necessitates the prior removal of sand masses adhering to the shell entrance area, which might be too costly in terms of time, as this sand was often hard packed or tightly adhered to the shell wall. Alternatively, sealing the shell at its opening might not contribute to a better protection of the progeny as the everlasting wind on Fuerteventura is expected to fill the open entrances of the completed nest with sand within a short period of time. Brood parasites: At the study sites, many specimens of Chrysis hohmanni Linsenmaier searched snail shells for putative host nests. As C. hohmanni has the same size and was equally common as H. zandeni at the study sites, H. zandeni is assumed to be its specific host. It can not be excluded, however, that C. hohmanni parasitizes the nests of L. fortunatus , which, however, is larger than the chrysidid wasp and was only rarely encountered.

Behaviour. Flight activity: H. zandeni was active also during windy, rather cool and cloudy weather when most other bee species had ceased their activities. H. zandeni thus seems to be well adapted to the windy conditions and often moderate temperatures prevailing on Fuerteventura during the species’ main flight period from February to April. One possible adaptation underlying this moderate sensitivity against unfavourable weather conditions might be the special flight behaviour of H. zandeni : both sexes used to fly very close to the sandy ground and often landed on the ground for shorter or longer resting periods (see below), thus probably benefitting from the heat emission of the sand. Male mating behaviour: On their search for females, males patrolled along more or less fixed routes close to the ground, which were regularly interrupted by short resting periods on the ground ( Fig. 18 ). Mating attempts were observed both on the ground and on flowers. In one case, a male constantly patrolled flowers of Frankenia and Reseda along a fixed circular route of 1.5m diameter around a low shrub and landed once per passage at exactly the same same spot to rest for a few seconds. Females also regularly interrupted their provisioning flights by short resting periods on the ground. As the frequency and length of these resting periods in both sexes increased with decreasing temperature and increasing wind velocity, this behaviour probably served to heat up the body on the warm sandy ground. Sleeping places: Females passed the night and bad weather in their own nest, while males were observed to sleep singly in snail shells.

Figures

FIGURE 9 – 14. Hoplitis zandeni. 9, 10: Semidesertic habitat on sandy ground with scattered shrubs of Salsola divaricata (Amaranthaceae), Launaea arborescens (Asteraceae) and Zygophyllum fontanesii (Zygophyllaceae) (Canary Islands, Fuerteventura, Costa Calma). 11: Nest in snail shell of Theba geminata, where the female had to burrow through sand masses that blocked the shell entrance. 12: Pollen loaden female at nest entrance. 13: Opened nest with one brood cell and nest plug built inside the shell. 14: Opened nest with inner side of nest plug consisting of sand grains irregularly glued together with leaf pulp.

FIGURE 15 – 18. Hoplitis zandeni. 15: Opened nest with threshold made of a mixture of leaf pulp and sand grains. 16. Opened nest with outer side of nest plug carefully worked with sand grains and leaf pulp to form a plane wall. 17: Yellowed leaf of Helianthemum canariense with chewing marks along apical and left margin. 18: Male resting on the ground.