Avian brood parasites display enhanced annual fecundity compared to other passerine birds. Adult female Brown-headed Cowbirds (Molothrus ater), a brood parasite in the Icterid family, are estimated to lay 40- 50 eggs/year. Some studies suggest annual fecundity in various parasitic species can exceed even this upper estimate. However, boosting reproductive output to this extent may come at the cost of reduced or lost parental care. Steroids, such as testosterone, may act as a pivot point that balances the fitness consequences between reproduction and parental behavior. Moreover, steroids are targets of selection that shape life history strategies and can therefore potentially contribute to the evolution of novel behavioral phenotypes. Here, we examine how reproductive steroids may mediate a possible tradeoff between increased annual fecundity and parental care in female cowbirds. We compare seasonal fluctuations in steroid profiles and follicular development in cowbirds and Red-winged Blackbirds (Agelaius phoeniceus), a related non-parasitic Icterid species. We aim to identify differences in seasonal steroid patterns that correspond to the striking divergence in life history strategies between these species.  We also use GnRH administration in these two species to determine whether species variation in the response to GnRH maps onto the differences in behavioral phenotypes. We identified several mechanisms that would enhance annual fecundity in the cowbird and one mechanism that would do this at the cost of parental care: elevated testosterone. We show that female cowbirds are reproductively ready before female Red-winged Blackbirds, a preferred and abundant host of the cowbird at our study site. We find no significant association between estrogen and follicular size in cowbirds, whereas this association is apparent in Red-winged Blackbirds. However, follicular development is associated with testosterone, and female cowbirds exhibit a pattern of elevated testosterone throughout the breeding season compared to the Red-winged Blackbird. These steroid profiles indicate divergence in the physiological mechanisms that regulate oocyte development in these related species. Moreover, cowbirds produce testosterone significantly quicker and more robustly in response to GnRH administration compared to female Red-winged Blackbirds. Divergence in the steroid seasonal profile and responsivity to GnRH in cowbirds, particularly with respect to testosterone, indicates the HPG axis exhibits consequential modifications in cowbirds that can enhance reproduction while simultaneously inhibiting caregiving behaviors.

Seasonal profiles of reproductive steroids

Female Brown-headed Cowbirds (n = 18) and Red-winged Blackbirds (n = 14) were captured using walk-in traps and mist nets in Ransom and Cass counties of North Dakota during the non-breeding and the peak breeding season in 2021 (May-June). Traps and mist-nets were continuously monitored to ensure the birds were removed within five minutes of capture. Birds were immediately euthanized via rapid decapitation for a separate study. This allowed us to measure the size of the largest follicle using calipers and collect blood to be stored on ice until centrifugation. Collections started May 1 (pre-breeding in North Dakota) and concluded June 25 (peak breeding). We categorized female cowbirds and Red-winged Blackbirds into three categories: pre-breeding, breeding onset, or peak breeding. These categories were defined by observing the breeding activity (or lack thereof) of female Red-winged Blackbirds, the most abundant preferred host available to cowbirds at this location.  The pre-breeding condition was defined as no nest building or egg laying activity present, but males are present, singing, and establishing territories (May 1-14). We captured no female Red-winged Blackbird in this stage because they are either less abundant or significantly more sedentary than males at this time (Wright and Wright, 1944; Holm et al., 1973). The breeding onset condition was marked by the first nest-building activity of a female Red-winged Blackbird. In this category, female Red-winged Blackbirds began building nests, but no eggs were observed in nests yet (May 15-22). The final category is the peak breeding season, which was marked by eggs observed in nests and provisioning behaviors present in female Red-winged Blackbirds (May 23 to June 25). These are subjective categories based on the observations of our field crew; however, these categories do coincide with the timing that female Red-winged Blackbirds transition from breeding to parental behaviors as described by other studies conducted at similar latitudes in North America (Wright and Wright, 1944; Holm et al., 1973). The sample sizes of each species in these categories include cowbirds: pre-breeding (n = 7); onset of breeding (n = 6), and peak breeding (n = 5); red-wings: onset of breeding (n = 6) and breeding (n = 8).  We measured 17-β estradiol (E2), testosterone (T), and follicular sizes across these breeding categories in both species.

GnRH administration

Breeding female Brown-headed Cowbirds (n = 10) and Red-winged Blackbirds (n = 8) were captured using walk-in and drop-down traps during the breeding season of 2022 at the Marine Nature Study Area and Lido Beach Nature Area of NY (red-wings) and the Balcones wildlife refuge in Marble Falls, Texas (cowbirds). Blood samples were obtained from each female via the branchial vein within five minutes of capture. This initial blood sample was used to obtain baseline hormone measurements. Immediately following blood collection, females were injected with 1.5 µg of gonadotropin-releasing hormone (GnRH; Bachem catalog # 4030773) intramuscularly in a 50 μl volume dissolved in phosphate-buffered saline (Jawor et al., 2006; Burns et al., 2014; Rosvall et al., 2013; 2014). Treated birds were placed in a cage approximately 18"L x 14"W x 24"H so that additional blood samples could be collected at 30 min (timepoint 2), 1 hour (timepoint 3), and 2 hours (timepoint 4) after GnRH treatment. Red-wing females were released with a color band in case of repeated capture, whereas female cowbirds were euthanized. Blood samples were frozen until centrifugation. Plasma was used to measure testosterone so that we could identify the immediate steroid response from the gonads, which initially produces testosterone before it is aromatized into estrogen. We then measured E2 with subjects that had remaining plasma: Brown-headed Cowbirds (n = 7); Red-winged Blackbirds (n =7).

All procedures described here were conducted in accordance with Hofstra IACUC procedures as well as federal (MB96705A-0), NY state (1181), Texas state (SPR-0521-069), and North Dakota state (GNF05428921) scientific collecting permits.

 Steroid hormone assays

Circulating T and E2 concentrations were quantified as described in Lynch et al., 2018; Pellicano, 2019; Lynch and Wilczynski, 2005; 2006; 2008; Lynch et al., 2006. Briefly, steroids were extracted from plasma using 3 ml of diethyl ether. Extracted steroids were resuspended in assay buffer for T and E2 assays using ELISA kits from Arbor Assays (17-β estradiol: cat # KB30-H1; Ann Arbor, Michigan) and Cayman Chemical (testosterone; cat # 582701; Ann Arbor, Michigan). These kits have been validated for use in birds in our previous studies (Lynch et al., 2018; Pellicano, 2019). Both steroids were measured using a single standard curve, thereby precluding an inter-assay variation measurement. Estradiol EIA kits have 100% cross reactivity with 17β-Estradiol, and all other reported cross reactivities were less than 1%. Testosterone kits have 27.4% cross reactivity with 5a-dihydrotestosterone, 18.9% cross reactivity with 5b-dihydrotestosterone, 4.7% cross reactivity with methyl testosterone, 3.7% cross reactivity with androstenedione, and 2.2% cross reactivity with 11-keto testosterone. All other cross-reactivities were less than 1%. The detection limit is 3.9 pg/ml for testosterone and 3.75 pg/ml for estradiol.

We used a two-way ANOVA for independent samples to determine whether there was an interaction between the two species across breeding stages. This analysis was used for steroid hormone data as well as follicular size. However, because we collected cowbirds in three breeding stages, whereas Red-winged Blackbirds could only be collected in two of these stages, we did an additional analysis to determine if pre-breeding hormone concentration or follicular size in cowbirds was different from the early breeding state in cowbirds and Red-winged Blackbirds. We used a t-test for independent samples and selected tests for unequal variances where data exhibited heteroscedasticity.  Steroid hormone concentrations were logarithmically transformed to achieve a normal distribution. Alpha values were set at 0.05.

We also used regression to examine the relationship between follicular size and steroid concentration in both species. These correlational analyses will allow us to identify species differences in the relationship between follicular growth and steroid concentration. These variables should be correlated if a typical associative breeding pattern is present.

Finally, we examined species comparisons in T and E₂ across four timepoints in the GnRH study using a 2x4 two-way ANOVA with one factor as a repeated measure. Hormone concentrations were logarithmically transformed to achieve a normal distribution. Raw data is represented in all graphs for simplicity.