Ecology of the Western Queen Butterfly Danaus gilippus thersippus (Lepidoptera: Nymphalidae) in the Mojave and Sonoran Deserts

The purpose of this study was to assess the ecological knowledge surrounding the western queen butterfly, Danaus gilippus thersippus (H. Bates). Specifically, our objectives were to synthesize existing data and knowledge on the ecology of the queen and use results of this assessment to inform the direction of future research on this understudied species. We identified six core areas for assessment: distribution, the biodiversity of plant resources, western queen and their host plant phenology, chemical ecology, and four key life history traits. We mapped the distribution of D. g. thersippus from museum specimen records, citizen science (e.g., iNaturalist) and image sharing app-based observations, along with other observational data enumerating all current known plant resources and long-range movements. We assembled 14 larval food plants, six pyrrolizidine alkaloids plants and six nectar plants distributed in the western Mojave and Sonoran Desert regions of the United States and Baja California. We report on its phenology and its long-range movement. Butterfly species have declined across the western US, and western monarch populations have declined by 97%. Danaus g. thersippus has received little research attention compared with its famous congener D. plexippus L. Danaus g. thersippus’ desert distribution may be at its temperature limits for the species distribution and for its rare host plant Asclepias nyctaginifolia.


Introduction
Both rare and common butterfly species in California, United States (US), have declined based on a 35-year dataset of butterfly species due to the compounded effects of habitat alteration [1], increased use of neonicotinoid insecticides [2], and climate change [3][4][5]. Aridland butterflies may be even more vulnerable as such species and their host plants are often living at the upper limit of their physiological temperature tolerance [6,7]. Thus, aridland butterfly populations near agricultural areas are at risk from all four unique types of risks.
Plants are essential to the survival of butterflies. Danaus gilippus thersippus require host plants, which support their growth and larvae (such as milkweeds), as well as nectar plants to satisfy their caloric needs as adults. These plant resources have declined due to land development, of soil organisms, primary producers, pollinators, herbivores, predators, parasites and microbiota, with highly variable abiotic resources, including multi-year drought cycles. Specifically, the Ivanpah Solar Electric Generating System (ISEGS) in the northern Mojave Desert is adjacent to a rare population of the milkweed, Asclepias nyctaginifolia A. Gray [17], which is a larval food plant of D. g. thersippus [24]. Several other solar facilities east of Joshua Tree National Park border populations of other milkweed and other nectar species used by D. g. thersippus. Thus, the purpose of this paper is to assess and synthesize the current literature to determine the distribution, phenology, host plant relationships, and movements of this desert butterfly D. g. thersippus, which is interwoven in the complex ecosystems of the Mojave and Sonoran Deserts.

Distribution and Phenology
Locality and phenological records for D. g. thersippus in western United States and Mexico were assembled from databases on the Global Biodiversity Information Facility (GBIF) and Biodiversity Information Serving Our Nation (BISON) (Supplementary Table S1) and Computarización y actualización de la curación de la Colección de Lepidóptera del Museo de Zoología "Alfonso L. Herrera" y su base de datos MARIPOSA, FASE I, which includes museum specimens and citizen science observational records from iNaturalist, and the Lepidopterists' Society Season Summary website for D. g. thersippus hosted by the Florida Museum of Natural History at the University of Florida [25]. Data from "unknown" sources on these sites were redacted and not used. Filtered locality data with duplicates removed for the western US states of Arizona, California, Nevada and Utah were used to produce the D. g. thersippus distribution map ( Figure 1). Phenological data sources include peer-reviewed literature and images of plant associations from collections, iNaturalist, and Flickr after plant identifications were confirmed by James M. Andre, Director of the Granite Mountains Desert Research Center, University of California, Natural Reserve System. We sorted the data by locality and date of collection, removed locality duplicates from the same date, and produced a count of dates in each month when D. gilippus were collected from each unique locality to produce the distribution map [25,26] (Figures 1 and 2). This data is summarized in Table 1. Specimen records in museum databases are assumed to be adults and correlate with a specimen, while observation records are a mix of adults and larval stage and occasionally egg stage data records. However, these are difficult to distinguish in iNaturalist, Flickr, BugGuide and other online databases because the life stage has not been designated and would require viewing each individual record and making a corresponding database of life stage for each record with their identification number. The addition of life stage designation data to online databases would increase the scientific value of citizen science and image sharing app-based Lepidopteran databases, thereby increasing the number of records available for analysis.
We collated the phenological data of D. g. thersippus, and, to avoid collection bias, we counted only unique locality/month/day records for all years. Therefore, if two specimens were collected at one site on the same date, we considered it one record. We visualized the phenological data for Danaus g. thersippus ( Figure 2). We assessed the phenological data for all plant species recorded, based on peer-reviewed literature, databases (Supplementary Table S1), photographic documentation (for example Figure 3, Supplementary Figure S1), that were documented to be used by queens. We predicted that plant species used by conspecifics were used by other Danaus species if they were sympatric with D. g. thersippus and if they contained appropriate larval or nectar chemistry such as pyrrolizidine alkaloids (Tables 2-4). If use was documented, we cited the reference. We then calculated the average plant resource availability per month (Supplementary Figure S2).

Host Plant Use and Phenology
We searched citizen science and image sharing app-based sites BugGuide, Flickr, and iNaturalist to find temporal observations of host plant use. These data on these sites were then confirmed by cross-checking host plant range data on Calflora and DiscoverLife in addition to botanical authorities.

Distribution and Phenology
We located 720 data points, (172 specimen records, 548 observation records) in the US from 1930 to 2018 and used these to construct the distribution map [27] (Table 1, Figure 1). Thus, 24.89% of the data come from museum specimen records and 76.11% come from observation records via iNaturalist (N = 417), BugGuide (N = 45) and Butterflies and Moths of North America (N = 82) and other sources (N = 4). We located 270 distribution data points (after 71 duplicate data were redacted) based on a total of 166 museum specimen records and 104 observation records in Mexico ( Table 1). The specimens were collected from 1935 through 2019 with a peak of specimen records in 1961 of 39 specimens. These were collated from 26 institutions [28] (SI-1). Danaus g. thersippus are found throughout the Mojave Desert in CA: Imperial Co. Inyo Co., San Bernadino Co., Riverside, Co., San Diego Co.; NV: Clark Co. and they are common throughout the Colorado Desert in California, the Sonoran Desert in AZ [29,30] (Figure 1, Tables 2-4). Their distribution extends south to northwestern Mexico [31,32]. This species occurs in the Chihauhuan desert as well but this was beyond the scope of this study. Records assembled for Figure 1 show a distribution in the Mojave and Sonoran Deserts (ecoregions 14 and 81) and the California coastal range (ecoregion 6) east into the Arizona/New Mexico Mountains (ecoregion 23) and Madrean Archipelago or sky islands (ecoregion 79) and south into Baja California (ecoregion 10.2.3), Sinaloa and the Chihuahuan Desert in Mexico (ecoregions 24, 14.3.2, 13.3.1, and 9.6.1, respectively). Current records show that the California central valley (ecoregion 7) and the Great Basin Desert (ecoregion 13) lack D. g. thersippus, which is likely due to different abiotic and biotic processes. According to Coolidge [33] and Emmel and Emmel [29], adults are found along the Pacific coast to San Diego and in the coastal mountains during fall; however, they are less common in the coastal region during spring and summer and are more typically found in the desert regions [34]. Late summer and fall records from the coast most likely represent adults that have dispersed from the desert [35]. Brown [35] suggests that there may be small stable coastal populations in San Diego County (Co.) or elsewhere where larval host plants are present. Some consider the San Joaquin Desert to possess similar climatic conditions and habit in Kern Co. and Fresno Co. and classify these areas as desert [36], hence we have indicated plant locations in Kern Co. as well (Tables 2-4).
Understanding the complex phenology all life stages of D. g. thersippus is a critical element of the queen's ecology and requires an understanding of its behavior patterns in the western deserts. Currently, we have observations from different studies documenting its presence from January to December in different life stages and a flight period from April to November [29]. Specimen records from Arizona in January as well as in Orange Co. [34] indicate they are active in fall months but also active in summer months ( Figure 2, Supplementary Table S2). Records from the western states of Mexico (Baja, Sonora and Sinaloa) suggest that queens are active in all months but most active in October and November. More adult and larval records are needed to complete a full picture of this species' phenology across its entire geographic range in the western US and Mexico. While the volume of scientific collecting has declined, the observation recording has increased. However, some of these records lack coordinate data, host plant associations and life stage designation in the database and other behavioral data (nectaring, mating, ovipositing, larval feeding, etc.), which would make these observations more useful scientifically. The 1008 data records were sourced from 348 museum specimen and 660 observation records [28,35] highlighting Mojave, Sonoran and Great Basin Desert ecoregions in Arizona, California, Nevada, Utah and Mexico but including all areas of these states (map adapted from Level III US Environmental Protection Agency (EPA) map data [37]). The numbers on the map correspond to the EPA ecoregions of focus where D. g. thersippus occurred. Inset map shows observations near and within the Chihuahuan Desert ecoregion.  Tables S1 and S2). Specimen records are comprised of adult butterfly stage data (100%). Combining museum specimen records with citizen science and image sharing app-based observational data increases numerical strength for analysis; however, observational data from multiple sources often lacked life stage designation.

Movement Patterns
Migration is defined as "an adaptation to resources that fluctuate spatiotemporally either seasonally or less predictably" [38]. Dingle and Drake [38] also noted that it often occurs preemptively before resources disappear. Danaus g. thersippus, in the desert in particular, must adapt to ephemeral nectar, larval and pyrrolizidine alkaloid (PA) host plant resources, with brief availability windows, and drought cycles that affect the diversity and abundance of host plants, with preemptive strategies before food sources decline [38,39]. Triggers for seasonal responses such as photoperiod or endogenous rhythms, changes in humidity, or food availability may cue movement patterns [39] of D. g. thersippus. However, D. g. thersippus have not received the intensive research tagging efforts that have advanced the understanding of D. plexippus migration behavior. Close inspection of museum specimen labels and published observations from other web-based sources such as GBIF, BugGuide, DiscoverLife, Scan, and iNaturalist provide some nascent evidence of seasonal movement (Supplementary Table S2 Table S2), suggesting that this species is capable of moving long distances. However, we were only able to locate three tagged individuals during our searches. If these are part of a systematic tagging effort, it would be more informative. Others individuals remain active and appear to remain in desert habitats in winter (Supplementary Table S2). Scott [40] states that D. g. thersippus fly from April to Nov in southern California and southern Nevada. This is an area that merits more research for this species.

Diet
Danaus g. thersippus require three major dietary categories for survival: larval host plants, adult nectar plants for nutrients and energy and nectar plants that contain pyrrolizidine alkaloids (PAs) required for courtship and mating. They use different plant species across their geographic range in the western Mojave and Sonoran Deserts and temporally throughout their annual lifespan (Tables 2-4, Supplementary Figure S2). The nectar of 90% of most species of plants has amino acids [41], 50% has detectable lipids, 60% has phenols [42,43] in addition to vitamins, minerals [44,45], alkaloids which have antimicrobial affects, terpenoids [44,[46][47][48], and phytosteroids [44]. In addition, some nectars contain antimicrobials and antioxidants ( [41] and references within). The availability of nectar can be affected by fluctuations in soil moisture resulting from drought [49], which is an important factor in arid landscapes. Nectar sources contain other constituents including antioxidents [50], alkaloids, microbiota such as bacteria, yeast, and fungi, which may serve functional roles such as pathogen control. Infected insects may employ antimicrobial phytochemicals against their own diseases [51]. Singer et al. [51,52] showed that Apantesis (formerly Grammia) incorrupta (Hy. Edwards) (Lepidoptera: Erebidae: Subfamily Arctiinae) larva ingestion of PA plant toxins improve the survival of parasitized caterpillars by conferring resistance against Exorista mella (Walker) flies in the family Tachinidae, compared to infected caterpillars on feeding on plants without PAs. The alkaloid gelsimine, which is found in the nectar of Gelsemium sempervirens, when fed on by Bombus impatiens, was found to reduce the gut protozoan Crithida bombi [53] in lab experiments. Small amounts of leaf herbivory by D. plexippus caterpillars, consistently resulted in induced increases in foliar cardenolides of A. syriaca [54] and research by Vannette and Hunter [55] suggest that mutualistic partners such as Scutellospora pellucida fungi mediate latex and cardenolide production in some species of Asclepias. Some amino acids in nectar have the potential to modify insect behavior by stimulating insect chemosensory receptors ( [41], review).
The nectar source desert lavender Condea emoryi (Torr.) Harley and J.F.B. Pastore contains secondary compounds such as butelinic acid which have antimicrobial, antitumor, and anti-inflammatory properties [56] and are important resources for D. g. thersippus and other desert butterflies (Table 3, Figure 3A, LSG field observations 2018-2019). Condea shrublands appear to be limited by temperature, as it does not occur > 700 m in the desert mountains and it does not occur very far north into the Mojave Desert [57]. Condea emoryi are found in low washes and tolerate a high degree of flood disturbance that occurs during high precipitation years which occur about every 10 years during winter and monsoon periods in summer. It is a long-lived species that re-sprouts following floods. Populations occur north and east of Joshua Tree Nat. Park (Table 3, Figure 3).
Several sources cite the use of desert palafox Palafoxia arida B.L. Turner as an important nectar source of D. g. thersippus including Pfeiler et al. (Table 5 in [32]). It is an annual which is locally abundant on dunes and sandy washes of bajadas appearing in April through May in Mojave dunes throughout California and northern Mexico [57]. As noted in Tables 2-4, D. g. thersippus caterpillars specialize on plants in the genus Asclepias, and the related genus Funestrum while adult butterflies collect nectar from plants in many families (Asteraceae, Boraginaceae, Fabaceae, Lamiaceae, Apocynaceae). We used data from citizen science and image sharing-based data such iNaturalist, BugGuide, and Flickr to expand the dietary knowledge of D. g. thersippus (Tables 2-4).      Adult female butterflies require nectar with sugars, amino acids and other nutrients for egg production and oviposition (Figure 3). Both males and females feed on plants containing pyrrolizidine alkaloids (PAs). However, males require PAs to produce their courtship attractant to mate with females [see chemical ecology below). All Senecio species tested (186 spp.) contain PAs [58], suggesting that Senecio faccidus var. douglasii and Senecio faccidus var. monoensis ( Table 3) are sources of PAs in the Mojave Desert where they occur. PAs were found in all plant tissues of Senecio; however, the inflorescences contained 90% of the total PAs [103,104]. PAs have been found in the desert species Amsinkia tesellata [84,105] and sixteen species of Eupatorium and one species of Ageratum (Table 3). Amsinkia tessellata, Senecio and Cryptantha are all present in the Ivanpah Valley region of the Mojave where D. g. thersippus occur (Figure 1). Tournefortia bait experiments of ( [61], Table 2) list species found in the western US which attracted male D. gilippus berenice. In Florida, D. gilippus use some of the same food plants as D. plexippus [61], which suggests that other Danaus subspecies may also share host plants with D. plexippus in other regions of the US. Nine species of Cryptantha have been shown to contain PAs [84] and several species of Cryptantha occur broadly in the western deserts and thus may be a source of PAs for D. g. thersippus. However, desert species of Amsinkia, Cryptantha, and Senecio require field experiments and observations to provide confirmation of their use and a deeper understanding of the nexus between D. g. thersippus butterflies, its host plants and local populations across aridlands throughout the western US and Mexico.  Figure S2), which corresponds with D. g. thersippus peak butterfly-larval phenological presence analysis in Figure 2.

Mating Behavior
Danaus g. thersippus eclose shortly after sunrise (30 min. past) [62]. Male D. g. thersippus patrol all day [62]. Males release an attraction pheromone during courtship that functions as a premating isolating mechanism [40,105]. The duration of copulation is 100 min to 12 h [29,40,106] or 4.5 h and longer according to Pliske and Eisner [107]. Danaus gilippus fly in copula [30] and the male usually carries the female D. gilippus: [40,108]. Females can mate up to 10 times [40]. Unreceptive females simply fly away [40]. Danaus gilippus males use PAs in courtship as honest advertisements of nuptial gifts of protective PAs which they transfer to females upon mating; females then transfer these PAs to the eggs [22]. Thus, plants containing PAs are as essential to the survival of Danaus butterflies as are milkweed plants with defensive cardenolides.

Chemical Ecology
The variability of cardenolide chemistry in host plants may affect the survivorship of larvae and adult butterflies. The choice of host plant oviposition site by females is one the most importance choices to ensure offspring success after mate choice. Due to the co-evolution of Danaus and the chemical ecology of its host plants this is a fertile subject of research with a growing body of literature.
Danaus g. thersippus in the western deserts rely on three categories of plants resources without which they cannot survive. The specialized host plants that contain cardenolides in the genera Asclepias and Funastrum (Apocynaceae) and other secondary compounds are required by the larva. The plants used by males to collect the precursor of their sex attractant are essential for attracting females, courtship and mating. In addition, both females and males require nectar sources for their entire adult lifespan which contain an appropriate suite of nutrients and energy resources. If one of these categories of host plant resources is absent, D. g. thersippus and Danaus populations in general will not be sustainable. Danaus g. thersippus are mobile flyers so they may mate in one location and oviposit their eggs in a different location; however, nectar resources are required for fueling travel and searching behavior.
Plants containing pyrrolizidine alkaloids possess two types of volatile compounds that attract male butterflies. Substances on the plant surface act as phagostimulants when a butterfly lands on a plant. Male D. gilippus sequester the precursor for the production of dihydropyrrolizines from plants ( Table 3) that they release via their hairpencils and that contain pyrrolizidine alkaloids [107,109] review. This attractant called danaidone attracts females to males during courtship [110]. Plants in the genus Senecio, (Figure S1) Amsinkia, Ageratina and Cryptantha, contain pyrrolizidine alkaloids sequestered by adult males from plants, which is transferred to females during mating, and then transferred by the female to the eggs [22]. It has been suggested that the eggs are protected from predators and parasites as a result but this requires further testing.
Milkweed plants in the genus Asclepias produce up to 200 structurally different cardenolides, which appear to all have the same inhibition of the Na + /K + ATPase in animal cells [69]. These cardiac glycosides block a key transmembrane carrier in animals [111]. Cardenolides are steroids which occur as glycosides with one or more sugars attached in plant or insect tissue [112]. For a detailed discussion of the chemical structure of cardiac glycosides, see Malcolm [112], and, for a thorough review of all aspects of the chemical ecology of cardenolides and danaids, see Agrawal et al. [21]. The polarity of cardenolides determines its absorption rate in the herbivore [21]. Cardenolides can be found in all Asclepias species [21,109], and all plant tissues, including latex, and nectar and cardenolide expression is tissue specific [21,112,113]. In addition, the concentration varies in time and in type of cardenolide, which vary in polarity [112,114,115], and in the diversity of compounds [116]. The binding of cardenolides to Na + /K + ATPase is temperature dependent [21]. Latex has higher concentration of cardenolides in some species such as A. eriocarpa [21,115,117]. Moranz and Brower [118] also found that concentrations of cardenolides, which varied temporally in Danaus gilippus, were mediated by their host plants.
Environmental variability can have significant impacts on foliar traits, such as soil moisture content [49], and seasonal temperature variation, which will have an impact on the concentration of toxins experienced by herbivores [113]. Secondary metabolism is influenced by water stress, which is relevant in desert systems and CO 2 level; a 20%-30% decline in cardenolides was recorded from water stress (86% reduction in leaf H 2 O potential). Cardenolide content increased with elevated CO 2 (1000 ppm) [21,119]. Herbivore damage can induce cardenolide expression as well [72,120,121]. The genus Asclepias contains cardenolides and other alkaloids ( Table 2) [21] and variation within a species is a heritable trait [80]. Paired comparison field experiments of plants in full sun or deep shading resulted in decreased concentration and diversity of cardenolides [21]. Cardenolides may function defensively against bacteria, fungi, protozoa and viruses [122][123][124][125]; however, research has yielded varied results. Vannette and Hunter [55] found a root infection by fungal species Scutellospora pellucida which induced increased leaf cardenolides and plant growth in A. syriaca but found no effect with fungus Glomus etunicatum.
The ability to store cardenolides is different for different species of milkweed species depending on the polarity of the cardenolide glycosides. Less polar cardenolides are more able to cross membranes [121,126,127]. In addition, female herbivores store a higher rate of toxins than males when reared on plants of equal concentration [126]. In addition, cardenolides such as calotropin and calactin can be found at different ratios in host plants and selectively concentrated differently in D. g. thersippus and D. plexippus [128]. Danaus gilippus stored calotropin, 10-12 times less than D. plexippus but not calactin when both were fed on Asclepias curassavica [87]. These two compounds, which are present in A. curassavica and other species are very effective in inducing animal emesis [88]. In a study by Oyeyele and Zalucki [129], sister species D. plexippus laid 70% of their eggs on low foliar cardenolide plants, concurring with another study with Asclepias curassavica in which butterflies preferred plants with lower cardenolides as well [90]. Species having higher concentrations of total cardenolides tend to have fewer polar compounds on average, thus having a smaller number of polar cardenolides overall [113].
Researchers have hypothesized that more toxic species not only have high concentrations of cardenolides, but also more apolar forms, which are more easily absorbed in the insect hemolymph [21,113]. Both D. g. thersippus and D. plexippus are cardiac glycoside insensitive; however, D. gilippus have a different molecular modification at position 122 than D. plexippus [130].
Agrawal et al. [131] documented latex exuded in response to leaf damage in 53 species of Asclepias, including some found in the western deserts. Latex response was high in Asclepias erosa compared with A. subulata, A. albicans, A. nyctaginifolia, A. fascularis, A. asperula, A. speciosa, A. vestita, A. latifolia, A. linaria, respectively (16. [131]. This same pattern was not repeated in cardenolide content [92]. Latex-transporting canals are pressurized so when a leaf or vein is cut, latex flows and coagulates rapidly, drying into a glue-like substance. Hence, it is considered a plant defense [93]. Young D. gilippus caterpillars vein cut to block the flow of latex to feeding sites to counteract this defensive secretion of milkweeds [132]. "Self-medication" has been attributed to the behavior of parasite-infected D. plexippus, which lay their eggs on antiparasitic milkweed to protect their larval offspring from parasitic growth when they hatch [124].

Predators, Parasites and Pathogens
Tachinid flies Lespesia archippivora Riley [Diptera: Tachinidae) [133] attack the larva of D. g. thersippus [134]. Danaus gilippus is parasitized by two chalcid wasps, B. ovata Say [134], Brachymeria annulata Fabricus [Hymenoptera Chalcididae) [135]. Ophryocystis elektroscirrha is a protozoan parasite that was first recovered from Danaus plexippus and Danaus gilippus butterflies in Florida in 1966 [136]. New infections occur when larvae ingest parasite spores as they feed on contaminated egg shells or milkweed leaves. Most spores are transmitted from infected adults to their offspring (vertical transmission), although horizontal transmission may also occur. Following ingestion, spores lyse in larval guts. Heavily infected adults have difficulty emerging from their pupal cases and expanding their wings, although adults with low parasite loads appear normal [136]. McLaughlin and Myers [136] found an Ophryocytis infection rate of 22% and 50% on the scales of adult Danaus g. berenice butterflies in Florida. A major cause of mortality in reared D. gilippus is a highly contagious polyhedral virus [136]. Birds are well documented predators of D. plexippus [137] and likely prey on Danaus g. thersippus as well.

Distribution
We assembled the spatial distribution of D. g. thersippus in the western USA and Mexico based on current literature, digital specimen records, observations, and image records (Figure 1). This distribution encompasses Arizona, California, Nevada, New Mexico, Utah, and Baja California to focus on the Mojave and Sonoran Desert distribution of D. g. thersippus. This analysis documents that D. g. thersippus uses both desert and coastal regions. It also suggests that the Great Basin Desert is not a preferred habitat perhaps due to climatic and other biotic limitations or this may be the result collecting-observation bias. We did not include the Chihuahuan Desert in this inquiry due to the current geographic focus of our research and field surveys; however, this species does extend into the Chihuahuan Desert as shown by museum and observation records [25,28]. Of note, is the absence of D. g. thersippus in the Central Valley likely due to anthropogenic activities including habitat alteration, resulting in the loss of all three categories of required host plants [8], and increased use of insecticides and herbicides [10,13,14]. We have also documented two populations of D. g. thersippus adjacent to protected areas: one in the interspace between the Ivanpah Solar Electric Generating System (ISEGS) and the Mojave National Preserve and a second near Desert Sunlight Solar and Joshua Tree National Park.
Analysis of the available distribution data reveals that only 24.89% of the data come from museum specimen records, revealing the shortage of this type of information for this species in the desert ecoregion. Observation records for adult and larval queen butterflies are easily identified; however, host plant associations would be best documented with specimen-based records so they can be verified by specialists.

Movement
Three tagging records document three specimens of D. g. thersippus long-distant movement events up to 1404 km from east to west. These appear to be in a manner similar to migratory movement [38] which its congener D. plexippus undertakes annually. All three movement records occurred in the fall (two in September and one in October), hence the time of year is also suggestive. The data provide some early hint that this population may move from east to west seasonally. However, the movement patterns and ecology of D. g. thersippus needs to be greatly expanded to understand the movement biology of this species. These three records all came from Lepidopterists' Society Season Summary [25] website which highlights the benefit of exploring multiple forms of data and organizations involved in data aggregation and more structured, and coordinated research for this species.

Phenology
We assembled the phenology of D. g. thersippus across its entire western range in the US covering Arizona, California, Nevada, New Mexico, Utah, and Baja California. Our analysis of phenological data from museum records and citizen-based observations (Figure 2, Figure S2) show that June, July, August, September and October are the peak activity periods for Danaus g. thersippus butterflies and larvae across their range in western range. Detailed site-specific research is required across their range to provide a more accurate phenology that can separate larval from adult phenology and to reveal if this species has migratory patterns. Plant record databases and literature show that April through Aug are the highest average periods for host plant (larval, nectar and pyrrolizidine alkaloid plant) resource availability across the Mojave and Sonoran Deserts. However, more detailed local site-specific phenology for each plant species' emergence and bloom period is needed to make this information biologically meaningful. Phenology can also change significantly with changes in precipitation and temperature annually, particularly in arid ecosystems where species are living at the upper limit of their physiological temperature tolerance [6,7].

Diet
The use of online photographic databases, such as BugGuide, iNaturalist, Flickr and those associated with DiscoverLife, provided documentation of host plant use, including the locality of host plant, and phenological data, including flowering or vegetation availability for the larval stage. Most images used were focused on the adult butterfly stage, due to the ease of identification. Plant species identification was sometimes challenging. Additional photographs with detailed close-ups of flower structures and leaf shapes in focus would improve the research quality of these images. Observations on all these digital databases that show multiple taxa would benefit by including latitude and longitude coordinates.
A great deal more research is needed on both the larval and adult diets of D. g. thersippus across the desert ranges. Researchers require details about which specific species of nectar plants are used to sustain this long-lived butterfly, temporally beginning with its emergence from its chrysalis, through courtship, mating, searching for oviposition sites, and spatially as plant resources change in different localities and over time as the seasons progress in the extreme temperatures of desert climates inter and intra-annually. More research is needed to understand how other organisms, including other herbivores [138] and humans, affect the resources required to sustain populations of these and other desert butterflies and pollinators of their host plants, which are mostly bees. Loss of larval and adult host PAs and nectar plants resources is one of the major factors in the 97% population decline of D. plexippus butterflies in the western United States along with insecticide use [20], Pelton). Danaus gilippus thersippus is equally vulnerable and of interest scientifically due to its behavior, chemical ecology and hypothesized long-distance movement, which is similar to the monarch butterfly. In addition, D. g. thersippus' host plant species may be at their temperature limits due to its distribution in desert ecoregions, hence alterations in the regional hydrology or climate might adversely affect its ecological network as well.

Chemical Ecology
The chemical ecology of D. g. thersippus closely resembles the chemical ecology of D. plexippus, which has been thoroughly and eloquently reviewed by Agrawal et al. [21]. Of note is the paired comparison field experiments of plants in full sun or deep shading, which resulted in a decreased concentration and diversity of cardenolides [21]. The impact on plants containing cardenolides merits further assessment where the distribution of Asclepias species, Danaus species and ground-mounted utility-scale photovoltaic and concentrated solar energy (USSE) development intersect [139]. The shading of soil and host plants used by Danaus species may have ecological consequences on this complex network across both spatial and temporal scales. The cardenolide and pyrrolizidine alkaloid concentration of Danaus host plants is critical to the future of this and related species.

Predators, Parasites and Pathogens
Danaus g. thersippus is parasitized by tachinid flies Lespesia archippivora [133], and by two chalcid wasps, B. ovata [133] and Brachymeria annulata [135]. Ophryocystis elektroscirrha is a protozoan parasite that infects D. g. thersippus. Ophryocytis infection rate was 10% and 50% on the scales of adult D. gilippus from Florida in experiments [140,141]. A major cause of mortality in reared i. gilippus is a highly contagious polyhedral virus [136]. In addition, birds are well documented predators of D. plexippus [137] and likely prey on D. g. thersippus as well.

Suggestions for Future Studies
The purpose of this paper is to assemble the current information available on Danaus g. thersippus in the western Mojave and Sonoran Deserts of the US. This is particularly critical in light of the decline in numbers of all butterfly species due to rapid large-scale development in the western deserts from urban expansion, utility-scale ground-mounted solar development in and near protected landscapes, and the increased use of systemic insecticides. Our intent is to stimulate further research on D. g. thersippus, due to its similarities to and differences from its close congener D. plexippus, by giving a preliminary roadmap to future work. In addition, the intersection of D. g. thersippus, their chemical ecology, complex relationship with their host plants, their movement ecology and vulnerability due to their distribution in Mojave and Sonoran Deserts and coastal California merits research attention.

Conclusions
We assembled the distribution of the queen butterfly Danaus gilippus thersippus in the western U.S. states of Arizona, California, Nevada and Utah and western Mexico based on 1008 data points, (182 specimen records, 556 observation records) in the US and 166 museum specimen records and 104 observation records in Mexico to construct its distribution map [27] (Figure 1).
We assembled the phenology of D. g. thersippus in this ecoregion using data based on a combination of museum specimen records (24.8%, N = 184) and observation records (75.1%, N = 556).
We assembled a referenced table of D. g. thersippus' dietary resources in the western deserts, including 14 larval host plant species records, six pyrrolizidine alkaloid host plant records, and six nectar host plant records with focus on the Mojave Desert and Sonora Desert ecoregions in the western US and Mexico. This is a starter list to stimulate much needed additional research.
We report on the long-distance movement behavior of D. g. thersippus moving from east to west, data assembled from the Lepidopterists' Society Season Summary [25].
We assembled a phenology of larval and adult host plant resource availability by month in the Mojave and Sonoran Deserts ( Figure S2).
We summarized the mating behavior, chemical ecology, predator, parasites and pathogens attacking queen butterflies focusing on D. g. thersippus in the western United States.