## Box 62 Calculation of daydegrees

An outline of a simple method to estimate day-degrees (after Daly et al. 1978) is exemplified by data on the relationship between temperature and development in the yellow-fever mosquito, Aedes aegypti (Diptera: Culi-cidae) (after Bar-Zeev 1958).

1 In the laboratory, establish the average time required for each stage to develop at different constant temperatures. The graph on the left shows the time in hours (H) for newly hatched larvae of Ae. aegypti to reach successive stages of development when incubated at various temperatures.

2 Plot the reciprocal of development time (1/H), the development rate, against temperature to obtain a sig-moid curve with the middle part of the curve approximately linear. The graph on the right shows the linear part of this relationship for the total development of Ae. aegypti from the newly hatched larva to the adult stage. A straight line would not be obtained if extreme development temperatures (e.g. higher than 32°C or lower than 16°C) had been included.

3 Fit a linear regression line to the points and calculate the slope of this line. The slope represents the amount in hours by which development rates are increased for each 1 degree of increased temperature. Hence, the reciprocal of the slope gives the number of hour-degrees, above threshold, required to complete development.

4 To estimate the developmental threshold, the regression line is projected to the x-axis (abscissa) to give the developmental zero, which in the case of Ae. aegypti is 13.3°C. This zero value may differ slightly from the actual developmental threshold determined experimentally, probably because at low (or high) temperatures the temperature-development relationship is rarely linear. For Ae. aegypti, the developmental threshold actually lies between 9 and 10°C.

5 The equation of the regression line is 1/H = k(T° - Tt), where H = development period, T° = temperature, Tt = development threshold temperature, and k = slope of line.

Thus, the physiological time for development is H(T° - Tx) = 1/k hour-degrees, or H(T° - Tt)/24 = 1/k = K day-degrees, with K = thermal constant, or K-value.

By inserting the values of H, T°, and Tt for the data from Ae. aegypti in the equation given above, the value of K can be calculated for each of the experimental temperatures from 14 to 36°C:

K 1008 2211 2834 2921 2866 2755 2861 3415 3882

Thus, the K-value for Ae. aegypti is approximately inde- 34-36°C), and averages about 2740 hour-degrees or pendent of temperature, except at extremes (14 and 114 day-degrees between 16 and 32°C.

night conditions or reduce or increase daytime heat. Thus, predictions of insect life-cycle events based on extrapolation from laboratory to field temperature records may be inaccurate. For these reasons, the laboratory estimates of physiological time should be corroborated by calculating the hour-degrees or day-degrees required for development under more natural conditions, but using the laboratory-estimated developmental threshold, as follows.

1 Place newly laid eggs or newly hatched larvae in their appropriate field habitat and record temperature each hour (or calculate a daily average - a less accurate method).

2 Estimate the time for completion of each instar by discarding all temperature readings below the developmental threshold of the instar and subtracting the developmental threshold from all other readings to determine the effective temperature for each hour (or simply subtract the development threshold temperature from the daily average temperature). Sum the degrees of effective temperature for each hour from the beginning to the end of the stadium. This procedure is called thermal summation.

3 Compare the field-estimated number of hour-degrees (or day-degrees) for each instar with that predicted from the laboratory data. If there are discrepancies, then microhabitat and/or fluctuating temperatures may be influencing insect development or the developmental zero read from the graph may be a poor estimate of the developmental threshold.

Another problem with laboratory estimation of physiological time is that insect populations maintained for lengthy periods under laboratory conditions frequently undergo acclimation to constant conditions or even genetic change in response to the altered environment or as a result of population reductions that produce genetic "bottle-necks". Therefore, insects maintained in rearing cages may exhibit different temperature-development relationships from individuals of the same species in wild populations.

For all of the above reasons any formula or model that purports to predict insect response to environmental conditions must be tested carefully for its fit with natural population responses.

### 6.10.2 Photoperiod

Many insects, perhaps most, do not develop continuously all year round, but avoid some seasonally adverse conditions by a resting period (section 6.5) or migration (section 6.7). Summer dormancy (aestivation) and winter dormancy (hibernation) provide two examples of avoidance of seasonal extremes. The most predictable environmental indicator of changing seasons is photoperiod - the length of the daily light phase or, more simply, day length. Near the equator, although sunrise to sunset of the longest day may be only a few minutes longer than on the shortest day, if the period of twilight is included then total day length shows more marked seasonal change. The photoperiod response is to duration rather than intensity and there is a critical threshold intensity of light below which the insect does not respond; this threshold is often as dim as twilight, but rarely as low as bright moonlight. Many insects appear to measure the duration of the light phase in the 24 h period, and some have been shown experimentally to measure the duration of dark. Others recognize long days by light falling within the "dark" half of the day.

Most insects can be described as "long-day" species, with growth and reproduction in summer and with dormancy commencing with decreasing day length. Others show the reverse pattern, with "short-day" (often fall and spring) activity and summer aestivation. In some species the life-history stage in which photo-period is assessed is in advance of the stage that reacts, as is the case when the photoperiodic response of the maternal generation of silkworms affects the eggs of the next generation.

The ability of insects to recognize seasonal photo-period and other environmental cues requires some means of measuring time between the cue and the subsequent onset or cessation of diapause. This is achieved through a "biological clock" (Box 4.4), which may be driven by internal (endogenous) or external (exogenous) daily cycles, called circadian rhythms. Interactions between the short time periodicity of circa-dian rhythms and longer-term seasonal rhythms, such as photoperiod recognition, are complex and diverse, and have probably evolved many times within the insects.

### 6.10.3 Humidity

The high surface area : volume ratio of insects means that loss of body water is a serious hazard in a terrestrial environment, especially a dry one. Low moisture content of the air can affect the physiology and thus the development, longevity, and oviposition of many insects. Air holds more water vapor at high than at low temperatures. The relative humidity (RH) at a particular temperature is the ratio of actual water vapor present to that necessary for saturation of the air at that temperature. At low relative humidities, development may be retarded, for example in many pests of stored products; but at high relative humidities or in saturated air (100% RH), insects or their eggs may drown or be infected more readily by pathogens. The fact that stadia may be greatly lengthened by unfavorable humidity has serious implications for estimates of development times, whether calendar or physiological time is used. The complicating effects of low, and sometimes even high, air moisture levels should be taken into account when gathering such data.

### 6.10.4 Mutagens and toxins

Stressful conditions induced by toxic or mutagenic chemicals may affect insect growth and form to varying degrees, ranging from death at one extreme to slight phenotypic modifications at the other end of the spectrum. Some life-history stages may be more sensitive to mutagens or toxins than others, and sometimes the phenotypic effects may not be easily measured by crude estimates of stress, such as percentage survival. One sensitive and efficient measure of the amount of genetic or environmental stress experienced by insects during development is the incidence of fluctuating asymmetry, or the quantitative differences between the left and right sides of each individual in a sample of the population. Insects are usually bilaterally symmetrical if grown under ideal conditions, so the left and right halves of their bodies are mirror images (except for obvious differences in structures such as the genitalia of some male insects). If grown under stressful conditions, however, the degree of asymmetry tends to increase.

The measurement of fluctuating asymmetry has many potential uses in theoretical and economic entomology and in assessment of environmental quality. For example, it can be used as an indicator of developmental stability to determine the effect on non-target organisms of exposure to insecticides or vermicides, such as avermectins. Bush flies (Musca vetustissima) breeding in the dung of cattle treated for nematode control with Avermectin B1 are significantly more asymmetric for two morphometric wing characters than flies breeding in the dung of untreated cattle. Fluctuating asymmetry has been used as a measure of environmental quality. For example, water quality has been assessed by comparing the amount of asymmetry in aquatic insects reared in polluted and clean water. In industrially polluted waters, particular bloodworms (larvae of chironomid midges) may survive but often exhibit gross developmental abnormalities. However, at lower levels of pollutants, more subtle effects may be detected as deviations from symmetry compared with clean-water controls. In addition, measures of developmental effects on non-target insects have been used to assess the specificity of biocides prior to marketing. The technique is not completely reliable, with doubts having been raised about interpretation (variation in response between different organ systems measured) and concerning the underlying mechanism causing any responses measured.

### 6.10.5 Biotic effects

In most insect orders, adult size has a strong genetic component and growth is strongly determinate. In many Lepidoptera, for example, final adult size is relatively constant within a species; reduction in food quality or availability delays caterpillar growth rather than causing reduced final adult size, although there are exceptions. In contrast, in flies that have limited or ephemeral larval resources, such as a dung pat or temporary pool, cessation of larval growth would result in death as the habitat shrinks. Thus larval crowding and/or limitation of food supply tend to shorten development time and reduce final adult size. In some mosquitoes and midges, success in short-lived pool habitats is attained by a small proportion of the larval population developing with extreme rapidity relative to their slower siblings. In pedogenetic gall midges (section 5.10.1), crowding with reduced food supply terminates larva-only reproductive cycles and induces the production of adults, allowing dispersal to more favorable habitats.

Food quality appears important in all these cases, but there may be related effects, for example as a result of crowding. Clearly, it can be difficult to segregate out food effects from other potentially limiting factors. In the California red scale, Aonidiella aurantii (Hemiptera: Diaspididae), development and reproduction on orange trees is fastest on fruit, intermediate on twigs, and slowest on leaves. Although these differences may reflect

0 0