Matching the green wave: growing season length determines embryonic diapause in roe deer

Abstract

The roe deer (Capreolus capreolus) is Europe's most widespread ungulate, notable for its unique trait of embryonic diapause (delayed blastocyst implantation after mating) and an ongoing debate regarding how climate change affects its parturition timing. Given the relatively constant timing of the rut, roe deer could cope with advancing greening by adjusting its diapause end. Here, we bridge the gap on factors influencing roe deer's diapause by analysing 390 uteri from legally hunted roe deer females in Germany (2017-2020), which we macroscopically examined for the presence of visible embryonic tissue to retrospectively identify the diapause end date. By employing a marginal Cox proportional hazard model, we tested associations between female phenotypic attributes, environmental conditions and the probability of ending embryonic diapause prematurely. Our results confirmed that high-quality, well-conditioned and prime-aged females tend to terminate embryonic diapause earlier. We also demonstrated for the first time that on a population-averaged level, the growing season length in the year of conception significantly influences the diapause timing, even explaining the much-debated shifts in parturition dates in roe deer over the last seven decades. Increased knowledge of mechanisms involved in embryonic diapause may also help decipher embryo-maternal interactions in general, including in vitro fertilization.

Keywords: embryonic diapause; length of season; phenology; phenotypic quality; proportional hazard model; roe deer.

Figure 1.

Reproductive cycle of roe deer with embryonic diapause. Female quality (FQ) influences the timing of rut and parturition; for example, high-quality females mate and give birth earlier. Whereas the timing of the rut is constant, there are contrasting findings on the altered timing of parturition and it is potentially driven by environmental conditions (EC) in the year of conception. Hypotheses depicted on the end of diapause tested in this study investigate, for the first time, the potential effects of environmental conditions in the year of conception and female quality and condition.

Figure 2.

(a) Linear regression between mandible length (cm) and eviscerated weight (kg) (both log-transformed) (Adj. R² = 0.162 given by the linear model; n = 330). (b) QCC of females into (1) better condition and higher quality (n = 81), (2) better condition and lower quality or poorer condition and higher quality (n = 174), and (3) poorer condition and lower quality based on the female’s quality (n = 75) given by the mandible length and the female’s condition given by the residuals from the regression model of (a).

Figure 3.

(a) Date of harvest/death of each female (n = 390) of the study and its estimated date of end of embryonic diapause (if embryonic diapause had ended before demise; n = 80). (b) Boxplot of the estimated date of end of embryonic diapause (if embryonic diapause had ended before demise; n = 80).

Figure 4.

Model predictions (Cox proportional hazard model) for the probability of ending embryonic diapause in relation to the QCC (refer to Model 1, table 1; n = 330 with 67 events), while focusing only on prime-aged females (≥2 to ≤7 years).

Figure 5.

Model predictions (Cox proportional hazard model) for the probability of ending embryonic diapause in relation to the varying LOS in the year of conception (refer to Model 2 table 1; n = 390 with 80 events). We used four different LOS scenarios, ranging from 185 to 240 days, thus mirroring conditions in the 1950s to recent years.