ASSESSING THE EFFICACY OF LEPIDOPTERANS AS POLLINATORS IN A MIDWESTERN POLLINATION NETWORK

Author

Carmen Burkett, Southern Illinois University CarbondaleFollow

Date of Award

8-1-2023

Degree Name

Master of Science

Department

Plant Biology

First Advisor

Sipes, Sedonia

Abstract

Over 200,000 animal species and at least 300,000 plant species are involved in mutualistic plant-pollinator interactions. Understanding pollinator-plant interactions within a community is then of immense importance to both conservation and agriculture as animal pollinators provide critical ecosystem services. Detecting declines in pollinator abundance, pollinator diversity, and pollination rate, however, is complicated by the wide range of natural variability occurring within a system. Studying ecological networks may aid in predicting how extinction waves may impact ecosystemsNot all floral visitors are pollinators, and pollinator quality is dependent on several factors including the variation in the frequency of pollen deposition on compatible stigmas, the suitability and quality of the pollen, and the number of pollen grains deposited. The efficacy of lepidopterans as pollinators is not particularly well known outside of some well-studied moth-plant systems. One way to assess pollinator efficacy is to look at the placement and quantity of pollen carried on the pollen vector’s body. Another approach for assessing an insect’s potential role as a pollinator is to examine pollination and visitation networks. Network metrics such as connectance, species richness, nestedness, complexity, Shannon Diversity of links, functional complementarity, and robustness can provide insight about the community. My research objective of this study is to 1) produce an inventory of diurnally active lepidopteran floral visitors in southern IL using a regional pollinator collection, and 2) examine the role of lepidopterans as pollinators in southern Illinois, using a combination of pollen analyses and network analyses. Specifically, I will address the following questions 1) What lepidopteran floral visitors are found in southern IL and what plants are utilized by lepidopteran floral visitors in southern IL? 2) Which lepidopteran taxa are potentially effective pollinators based on the pattern and amount of pollen carried on their bodies? 3) Which body parts carry the most pollen and do these patterns vary by taxon? 4) How do lepidopterans and plants interact at a community level, and how do non-native species interact? Southern Illinois is an area of high biodiversity and our surveys included 292 sites and 753 collection events across the eleven southernmost counties in Illinois (Williamson, Franklin, Johnson, Jackson, Randolph, Alexander, Gallatin, Pope, Wabash, Hardin, and Union) which alone are home to six of the fourteen natural divisions found in the state: Lower Mississippi River Bottomlands, Ozark, Wabash Border, Southern Till Plain, Coastal Plain, and Shawnee Hills. Most specimens were collected between 2017 and 2019 in the following studies: 1) a 2017 study examining pollinators in agrosystems of Crab Orchard Wildlife Refuge; 2) a 2017-2019 ecologically stratified pollinator inventory study focusing on the eleven southernmost counties of Illinois in the USFS Shawnee national forest and USFWS Crab Orchard Wildlife; and 3) a 2019 inventory of the Illinois National Guard Training Area in Sparta, Randolph County, IL. This study focuses only on the lepidopterans collected during floral visitor surveys of all pollinator taxa. The floral associations recorded with each specimen were used to construct three visitation networks in R for comparison with pollination networks using pollen data—one showing all the lepidopteran and plant interactions within the collection (excluding pan trap specimens), one showing only those from the inventory (natural and semi-natural areas) and National Guard studies, and one showing those from the agricultural studies. There were 3916 lepidopteran specimens collected in total, and these represented 16 families, 108 genera, and 139 species collected from 207 host plants. Themost frequently collected (abundant) species included Cupido comyntas (640 individuals collected), Colias eurytheme (325 collected), Chlosyne nycteis (137 collected), Phyciodes tharos (321 collected), Papilio troilus (164 collected), Hemaris diffinis (161 collected), Epargyreus clarus (178 collected), Euptoieta claudia (324 tcollected), Pieris rapae (71 collected), and Poanes zabulon (77 collected). No threatened or endangered species were collected in the datasets utilized in this project. Insect specimens were examined under a dissecting microscope for pollen presence and amount in nine areas: dorsal head, ventral head, dorsal thorax, ventral thorax, dorsal abdomen, ventral abdomen, dorsal wings, ventral wings, and legs. The pollen was then counted and given a score estimating the number of pollen grains present to the nearest ten grains. The records for pollen carrying individuals were then used to make three more networks more accurately representing potential pollination--one with pollen carrying individuals across all studies, one with pollen carrying individuals from only the Inventory and National Guard studies, and one with pollen carrying individuals and their floral associations from the agricultural study Of the 1637 lepidopteran specimens across 12 families scanned for the presence of pollen, the families Papilionidae, Sphingidae, and Nymphalidae had the most pollen carrying individuals, with 21% or more of individuals carrying pollen. Of the ten most abundant species, only five (Hemaris diffinis, Epargyreus clarus, Euptoieta claudia, Papilio troilus, and Colias eurytheme) carried pollen consistently. Species such as Hylephila phyleus, Pieris rapae, Phoebis sennae, Cirrhophanus triangulifer, Eurytides marcellus, and Ctenucha virginica were more common in the natural and semi-natural areas while species like Cupido comyntas, Colias eurytheme, Battus philenor, Papilio troilus, and Chlosyne nycteis were more common visitors in agricultural areas. Non-native, cultivated, and invasive plant species were present at both types of sites (natural/semi-natural and agricultural). Invasives such as Lonicera japonica and Eleagnus umbellata were recorded as floral associations in the visitation networks, however, they were not reflected in the pollination network floral associations. The majority of the lepidopteran specimens were not found to be carrying any pollen. Of those that were, most carried several types of pollen. The families Asteraceae, Oxalidaceae, Fabaceae, Lamiaceae, Apiaceae, Apocynaceae, Asclepiadaceae, Brassicaceae, and Rubiaceae accounted for the majority of all floral visits with Fabaceae accounting for 37.45%, Asteraceae accounting for 30.24%, Lamiaceae accounting for 5.85%, Apocynaceae accounting for 5.19%, Oxalidaceae accounting for 5.01%, Apiaceae accounting for 2.01%, Rubiaceae accounting for 1.40%, and Brassicaceae accounting for 1.26%. Overall, the semi-natural areas networks exhibited higher biodiversity, connectance, robustness, Shannon diversity of links, nestedness, and functional complementarity than the agricultural field networks. This was true for both the pollination and the visitation networks. The agricultural field networks (both pollination and visitation) had higher web asymmetry than the semi-natural and natural areas visitation and pollination networks. The visitation networks were comparable to other networks produced--bee-plant and syrphid-plant. However, the values were lower than the bee-plant visitation networks but the lepidopteran-plant data set is much smaller than that of the bee-plant networks so this could be a contributing factor. Likewise, the lepidopteran-plant networks show higher robustness, connectance, functional complementarity, and nestedness than the syrphid-plant visitation networks but the syrphid-plant networks included a smaller data set than the lepidopteran-plant networks. In the agricultural visitation network, a few species seem to dominate the network on both the plant and insect side while the semi-natural areas network had a much more even distribution.