Night warming has mixed effects on development of the fall armyworm, Spodoptera frugiperda (Lepidoptera, Noctuidae) in southern China.

1. Introduction

From its’ first detection in China during January 2019, the fall armyworm (FAW), Spodoptera frugiperda (Lepidoptera, Noctuidae), a notorious migratory agricultural pest, spread to 26 provinces (autonomous regions and municipalities) by October 2019 damaging more than 106.5 million hectares of maize fields in that year [1]. FAW effectively completed all stages of the invasion process during 2019; from introduction, colonization, incubation, to spread and causing damage. South of 28°N in China the species can continue to reproduce and could survive winter between 28°-31°N [1].

FAW is potentially a new major pest to Chinese agriculture [2], as it is reputedly polyphagous being recorded on more than 353 plants, of which 180 were agricultural crops [3,4,5], including maize, sorghum, peanut; but see [6]. Two genetically-distinct sub-populations or strains are recognized: “corn-strain” and “rice-strain”, based on their host preference, mating behavior and sex pheromone [7,8]. From a genomic analysis of snips and gene sequences of 318 S. frugiperda moths collected from 13 provinces researchers concluded that the S. frugiperda strain that invaded China was derived from “rice-strain” female parent and “corn-strain” male parent [9,10]. Most field populations have been recorded on corn and other Poaceae, as has been found Africa [11], India [12] and Australia [6]. Some 83% of FAW moths trapped by searchlight on Yongxing island, had fed on C4 plants such as corn and not rice [13]. The S. frugiperda strain in China mainly damages corn, especially fresh corn. Larvae can tunnel and destroy the corn ear during milk maturation stage [14]. FAW has the potential to reduce annual maize production by 21-53% in the absence of pest management [11]. Guangdong, Guangxi and Hainan provinces in southern China are suitable for year-round populations of FAW [15] with about 300,000 hectares of corn [16] at risk.

Temperature influences the life of an ectotherm, determining its body temperature and consequently influencing the rate of physiological functions affecting growth, development, and fitness components [17,18,19]. The global mean surface air temperature has risen substantially during the 20th century [20], with night temperature increases greater than day temperatures [21,22]. The low temperature phase of the thermocycle (daily cycle of temperature) appears to play a major role in determining an insect’s performance [23]. Here our primary goal was to evaluate the effect of increasing night temperature on S. frugiperda growth and development relevant to corn planting area in southern China. Most work on temperature effects on insects use constant temperatures and FAW is no exception [24,25,26,27]. Here we test the effects of day-night temperature regimes relevant to the corn growing area of southern China on S. frugiperda by evaluating various life history traits. Constant temperature studies do not predict the development under our changing temperature. Under some fluctuations FAW develops faster than expected. Our results will better inform attempts at predicting the timing of FAW stages in the field.

2. Materials and Methods

2.1. Insect Rearing

The S. frugiperda used in this study was acquired from Tobacco Research Institute of Beijing Academy of Agricultural Sciences, reared at 27±1℃, humidity 70±5% for more than 20 generations in the insectary. First three (1-3) larval stages were fed fresh maize seedlings (variety: ZD958, height: 30cm) in cages (40cm×30cm×45cm), late larval stages (4-6) were fed with artificial diet individually in 6 well plates [28]. Pupae were kept in Petri dishes until emergence. Adults were kept in mesh cages (40 cm×30 cm×45 cm), for mating and oviposition, and honey water solution (10%) was provided.

Egg masses (n=45) each with at least 30 eggs (≤12h old) were put on fresh corn leaves and placed in 9 climate chambers (5 egg masses/ chamber). Larvae were transferred (twice a day) to fresh corn leaves when eggs hatched, and the time of each transfer noted. Third instar larvae were reared individually on fresh corn leaves in finger tubes (2cm diameter) covered with nylon gauze (120 mesh) until they pupated. Pupae were moved to Petri dishes 30% filled with soil until emergence. We checked the insects twice a day and recorded the instar. Emerging adults were put into cages (40cm×30cm×45cm), with honey water (10%) and the number alive recorded daily until all the moths had died.

2.2. Experimental Design

FAW can survive and develop in the ambient temperature range of 14-32℃ although morality is high at temperatures below 18℃ [29]. Temperatures above 32℃ are less suitable for the species, and largest pupal body weight and highest fecundity were recorded at 20-28℃ suggesting these conditions are optimal [26,29,30,31]. Daily temperature data of five years (2018-2022) were downloaded from a weather website (https://tianqi.2345.com/). Daily maximum temperature and diurnal temperature variation/ range during the corn growing season, October to March, was summarized for southern Provinces in China. Based on the results, we selected three high day temperatures of 27℃, 24℃, 20℃ and three night temperatures regimes that differed by 6℃ to 4℃ and 2℃to give three treatment groups; Group I: 27-21℃, 27-23℃, 27-25℃; Group II: 24-18℃, 24-20℃, 24-22℃; Group III: 20-14℃, 20-16℃, 20-18℃ (see support Figure 1). The temperature cycle was set to 12h day temperature and 12h night. The time needed for temperature to change from one to the other was 30min. The climate chambers temperatures were confirmed with thermometers and checked daily during the study.

All the studies were conducted simultaneously in 9 climate chambers (Model: RGC-500B, Anhui Youke Instrument and Equipment Company): the variation in temperature is ±1℃ and ±3-7% in humidity. We set the climate chambers on the temperature change mode with the following photoperiod: L:D=14h/10h, humidity:75%, light strength:15000lx, respectively.

2.3. Data Analysis

Mortality rate and development time data were calculated using SPSS (version: 23.0). The effects of temperature treatments on the development time and survival rate were analyzed using General Liner Model (GLM). Survival rate (Binomial) and development time (Poisson) are the dependent variables. Day temperature, night temperature and the interaction of day temperature and night temperature are the independent variables, respectively. Tukey–Kramer HSD test was used for the significant differences.

Linear regression model (Equation 1) was used to describe the relationship between the developmental rate and temperature, R² indicate the degree of fit:

$$V\left(T\right)=\mathrm{a}+\mathrm{b}T$$

(1)

Where a, b are the model parameters, V(T) is the developmental rate of selected stage (1/time in stage) and T is the temperature.

According to the effective accumulated temperature rule, the least squares method is used to calculate the developmental starting temperature and effective accumulated temperature of the armyworm at each developmental stage. The calculation formula (2) (3) is as follows:

$$C=\frac{\Sigma V\sum T-\Sigma V\Sigma T}{n\Sigma V^2-{\left(\Sigma V\right)}^2}$$

(2)

$$K=\frac{n\Sigma VT-\Sigma V\Sigma T}{n\Sigma V^2-{\left(\Sigma V\right)}^2}$$

(3)

In the formula, T is the treatment temperature; C is the developmental threshold temperature; K is the effective accumulated temperature constant; V is the developmental rate (1/developmental days); n is the number of experimental temperature groups, in this study n=9.

We used results from previous studies to compare development among populations (see Table 1).

3. Results

3.1. Development Duration

Development duration of eggs (Figure 2A), larvae (Figure 2B) and pupae (Figure 2C) increased with the decreased daily temperature and night temperature. Moth longevity was reduced with declining daily temperature and night temperature increasing (Figure 2D). Daily temperature treatments levels had significant effects on duration of all stages except for adults (Egg, χ2=9.905; df=2; P=0.000; Larvae, χ2=483.963; df=2; P=0.000; Pupae, χ2=358.403; df=2; P=0.000; Adult, χ2=6.466; df=2; P=0.000; Generation, χ2=1354.096; df=2; P=0.000). Night temperature levels had significant effects on duration of all stages except for adults (Egg, χ2=1.083; df=2; P=0.000; Larvae, χ2=20.248; df=2; P=0.000; Pupae, χ2=58.729; df=2; P=0.000; Adult, χ2=0.125; df=2; P=0.875; Generation, χ2=110.932; df=2; P=0.000). There were significant interaction effects caused by the night temperature increasing in larvae (χ2=9.152; df=4; P=0.000), pupae (χ2=8.885; df=4; P=0.014) and adults (χ2=3.451; df=4; P=0.025).. Generation time was 41.4 d, 35.6 d and 35.5 d in the group of 27℃ day temperature, night temperature 21℃, 23℃ and 25℃, respectively. Generation time was 57.6 d, 55.9 d and 50 d with 24℃ day temperature, night temperature of 18℃, 20℃ and 22℃, respectively. While generation time was 70.3 d, 65.5 d and 60 d at 20℃ day temperature, with night temperatures of 14℃, 16℃ and 18℃, respectively.

3.2. Survival Rate

All eggs survived and hatched and there were no moth deaths over the first three days of adult life. Mortality rate of larvae (Figure 2E) and pupa (Figure 2F) increased with the decreased daily temperature and night temperature range, much more so for pupae (Figure 2F). Significant effects were observed between day temperature regime and mortality in larvae (χ2=0.0978; df=2; P=0.0048) and pupae (χ2=0.5957; df=2; P=0.0003).

3.3. Development of Larvae, Pupae and Adults

Compared to previous studies, estimated lower temperature threshold of larvae and generation in this study are lower, and higher for the pupal stage. The estimated thermal requirement in degree-days of larvae is higher, but the pupae and generation are lower than others (Table 1).

Developmental rate, V(T) increased with increased day and night temperatures. Slope of the linear regression model indicates the development speed of insects. The linear equation for larvae was V(T) =0.0033T-0.0423 (R²=0.8389), for pupae V(T) =0.0113T-0.154 (R²=0.9402) and generation V(T) =0.0016T-0.0123 (R²=0.6582), where T is the average temperature experienced. Compared to results conducted under constant temperature previously [26,29,30,31], observed slope of linear regression model of larvae and adults in this study were similar. The slope for pupae in our study is higher than in previous work.

3.4. Pupation and Emergence Uniformity

Higher uniformity of S. frugiperda pupation and emergence are observed in high daily temperature and night temperature. The start of pupation and emergence in the treatment of 27-25℃ were 20d and 27d while 40d and 60d in the treatment of 20-14℃.

Figure 3. Cumulative proportion pupation (A) and adult emergence (B) for each temperature regime plotted against degree-days experienced.

Figure 3. Cumulative proportion pupation (A) and adult emergence (B) for each temperature regime plotted against degree-days experienced.

4. Discussion

In ectotherms body temperature more or less tracks changes in ambient temperature, although behavior can greatly modify the temperature experienced —e.g., [32,33,34,35]. The temperature experienced in the organism’s physical environment will strongly influence development, survival, potential and realized fecundity, and migration; indeed, virtually all aspects of its physiological ecology [36]. Invariably when an insect species invades a new area there are a flood of papers reporting these temperature effects, and Spodoptera frugiperda is no exception; e.g. China [24,25,26,30,37,38], India [31,32,33,34,35,36,37,38,39] and Kenya [40]. These papers are locally based and essentially reinvent a wheel, as they are generally based on constant temperatures. The premise and indeed promise is that using these data we can predict development based on some fitted model to the laboratory results [41]. Surprisingly constant temperature results and even simple degree day or linear models and rate summing ‘work’, although they can be right for the wrong reasons. Much of the ability to predict depends on what temperatures are experienced and in practice real temperatures fluctuate and various studies have reported such effects [42,43,44,45]. Here we tested simple changes in temperature experienced based on day-night temperature changes relevant to the invaded range in tropical parts of southern China. In these regions the shape of the temperature curve can indeed be more of less flat, between small day-night changes.

As expected, significant effects on development were observed in all temperature treatments – generally the higher the temperature the faster the development, but there were significant interactions effects of temperature changes. The greatest effect of fluctuation was to slow down development at intermediate temperatures (24℃) relative to the simple expectations of a degree-day model. Fluctuations increase rate of development at low temperature treatments (20 ℃), consistently so in this treatment, the greater the fluctuation the faster the development. Similar fluctuating temperature effects, namely increasing rates of development, have been observed in other studies — e.g., [46]. There was little to no effect at near optimal temperatures (27℃). Linear degree day models are meant to apply well to the intermediate temperature range in which our experimental treatments fell. However, predictions are likely to be out by potentially up to 5 days for larvae. The timing of management decisions may need to be adjusted for such effects of day-night temperature changes.

Other studies too have shown that larval growth rate, survival rate, adult longevity and fecundity are increased with the moderate increase of night temperature [47] and for the red scale population density in the Covina Valley is inversely correlated with the number of nights with the temperature dropped [48]. Whitney and Johnson (2005) showed that increasing night temperature increased the intrinsic rate of increase in Pieris rapae [49]. The effects of night temperatures on adult lifespan of the Finnish Glanville fritillary butterfly (Melitaea cinxia) [50] were similar with this study. The larval and pupal survival rate was only significantly affected by the daily highest temperature. Low nighttime temperatures had little effect on the survival and the total amount of maturation feeding of Monochamus alternatus adults [51].

Geographical ranges, population dynamics and phenologies of many organisms are altered by climate change. For ectotherms, increased ambient temperatures frequently have direct consequences for metabolic rates, activity patterns and developmental rates. Consequently, in many insect species, earlier and prolonged seasonal duration occur with recent global warming. However, from an ecological and evolutionary perspective, the voltinism and investment into each generation may be even more important than seasonality, since an additional generation per unit time may accelerate population growth and the potential for adaptation [52,53,54]. Pupation and emergence of S. frugiperda increased along with the increased daily temperature and night temperature. The uniformity of pupation and emergence is higher in the warmer night in this study indicating that the growth, development, population and voltinism will increase. We suggest management of S. frugiperda will become more costly to the economy and the environment.

Chinese government has announced the rules on S. frugiperda management, for Hainan, Guangdong, Guangxi and Yunnan provinces which are the winter reproducing areas of S. frugiperda. Monitoring the invaded population and decreasing the number of pest sources which may migrate to northern China from this area are two critical objectives [55,56,57]. However, as mentioned earlier, with the huge demand of fresh corn consumption in China, winter corn planting area has increased rapidly in Guangxi, Guangdong and Hainan provinces. The population of S. frugiperda in this area may provide a source population for annual migration and invasion of northern China, but transgenic varieties and pesticide cannot be used as sweet corn is consumed by people directly. It will be a big challenge to policy makers to balance the in’s and out’s of S. frugiperda management. In conclusion, the results of this study could help the local authorities to better time biological and physical control interventions.