Date: 16 September 1998
From: Deborah A Bidwell <dbidwell@hopper.unh.edu>
To: AQUA-L@killick.ifmt.nf.ca
QUESTION:
What is the proper way to calculate degree days when the rearing temperature falls below zero on some occasions?
Deborah A. Bidwell
Zoology Department
University of New Hampshire
(603) 862-4153
dbidwell@hopper.unh.edu
*************
COMMENTS 1:
"Degree days" is a unit that has had application for expressing the
incubation period, based on i) the average temperature (in degrees Celsius) over the period of egg development to hatching and ii) the number of days to hatching. It is the product of the value of the average temperature multiplied by the number of days.
References to "degree days" are not hard to find in general texts on
aquaculture and fisheries biology, and in more specialist papers. Thus,
for example, Landau (Introduction to Aquaculture, 1992, John Wiley & Sons, pages 230-231) briefly discussed the unit.
As will be obvious the unit "degree days" is really nothing more then a
short-cut form of expression. It has mainly been used to provide for
convenient comparison of incubation periods within single species at
different temperatures, as well as for inter-species comparisons of egg
development. In these contexts, it should be noted, as pointed out by
Landau, that while the hatching time may vary over a wide range of degree days for eggs in the same fertilization batch, the average number of degree days is relatively constant for each particular species - being 400 to 460 for the brown trout and 290 to 330 for the rainbow trout. For the aquaculturist, the main inference that can be drawn from these values is that eggs of rainbows develop faster to hatching than eggs of browns.
With regard to the question concerning how do you calculate degree days if the temperature is zero or temperatures fall below zero, we need to note how the unit is defined and then to evaluate how and when it should be used. As indicated above it is the average temperature over the whole period to hatching that is the factor used in the multiplication, so, unless the zero or sub-zero temperatures predominate throughout the hatching period, there should be no problem. If temperatures during the incubation period are mainly zero or sub-zero (and, as I asked previously, is this really physiologically possible for any fish eggs during development to hatching under normal conditions?) the calculation of degree days as presently defined would not be possible. The information for such a case would need to be described by other means than use of this short-cut unit. The manner of description devised would presumably be more meaningful for the eggs of the particular species and for the conditions under which they were being studied.
Dr Norman E. Milward,
Hon. Research Associate, Department of Zoology,
James Cook University, Townsville, Queensland 4811, Australia.
Ph: (07) 4781 4193 Fax: (07) 4725 1570 :
e-mail: norman.milward@jcu.edu.au
***************
COMMENTS 2:
With abalone (H. discus hannai) larvae, development is also calculated on degree days. However, the threshold temperature for these critters is taken as ( I think) 7-8C - being the "biological zero" or the temperature under which no development occurs. This seems to me to be a valid reason for shifting the threshold level away from zero . . .
Tom Lewis
e-mail: Tom.Lewis@utas.edu.au
***************
COMMENTS 3:
In the application most familiar to aquaculture, salmonid egg incubation, Celcius degrees are used and the threshold is zero (freezing) which causes it to be unstated or even not realised.
Though the most familiar use is in the prediction of egg hatching the concept was not invented in salmonid hatcheries. It is widely used in climatological studies to compare seasonal data for a variety of purposes, for example, agricultural crop production, applying temperature thresholds often well above freezing, as appropriate. I have used it in a similar way with oceanographic data to compare seasonal potentials for oyster growth and reproduction, applying temperature thresholds appropriate to each.
In summary, to calculate "day-degrees", you add together mean
daily temperatures minus the threshold you choose. For complete clarity the temperature scale and threshold used should be stated. e.g. Degree-days above 4 degrees Celcius.
Roy
e-mail: rdrinnan@netcom.ca
***************
COMMENTS 4:
In salmonid work the calculation is based on average degrees Centigrade and, hence, as you also point out, could be said to have a
threshold of zero (freezing).
No doubt, the average temperature above some other threshold could be adopted but clearly for every different base threshold used a different degree days value would be obtained. This would not appear to be of much benefit for enabling easy comparison of development rates within or between species, even if the temperature scale and threshold are always stated.
The period of egg development from fertilization to hatch is usually quoted in the literature as the number of days at some particular temperature in degrees Centigrade. Thus, for example, Embody (1934) gave values of 97 days at 4.5 degrees C. for brown trout and 75 days at 4.5 degrees C. Hence, the respective degree days values on Embody's data would be 436.5 and 337.5 respectively for the two species. As postulated by you it would be possible to calculate different degree days values on Embody's data, such as 145.5 and 112.5 using a threshold temperature of 3 degrees C., but I don't see
what this would achieve.
It may be worth noting that the constancy assumed by Apstein (1909) for the product of time in days to reach a certain stage of development and the prevailing temperature in degrees C., i.e. degree days, is not strictly correct, since over the full range of temperatures at which development may occur the relationship tends to decrease at lower temperatures and vice versa. However, for hatchery purposes the incubation time can usually be calculated with sufficient accuracy for the prevailing temperatures (in degrees C.) from the known degree days values.
Norman Milward
e-mail: norman.milward@jcu.edu.au