strength of current are of great significance in nursery feeding, With a stock density 

 of 128 trout per liter (1.7 fish per square centimeter) and 6^ trout per liter (0.85 

 trout per square centimeter) in two identical long-stream boxes (160 x 30 x 13 centi- 

 meters water space), the denser stock showed about 15 percent better grov/th in two 

 months, than the weaker stock in three months, A corresponding picture was given by 

 experiments in variou : sized boxes vath the same number of fish. Of course, the "outflow 

 factor" shows its influence here, A box virith /iOO square centimeters of surface, in con- 

 trast to one with l,Zi53 square centimeters of surface, both stocked vdth 200 fish, showed 

 about 26 percent poorer growth after 29 days. Thus, the follcnvfing rule may be stated: 

 In the early development stage the density of stock must not be too weak, with larger 

 growth a lower density of stock is more favorable. Accordingly, in case of longer term 

 "nursery feeding", the brood should be sorted and distributed in larger space. In every 

 case, the "nursery boxes" should have ample surfaces, not less than 1500 square centi- 

 meters. 



Experiments with a through current of 12 liters per minute (in a box of 80 x 30 x 13 

 centimeters water space) showed a superior growth in t he fry as compared with fry, kept in 

 boxes with a through current of only 6 liters per minute. The fish had grown by ll per- 

 cent better within 29 days. The current must not be strong enough to press the brood 

 against the outflow sieve. 



The influence of light upon the groTfth of the fry is also noticeable. T7e experimented 

 with fr:/ in nursery bcoces kept in the open and in houses, exposed to light and to darkness. 

 Best results were shown by boxes kept in the open. Equally good results were noted with 

 bcDces (kept in houses) that were painted white on the inside. The results were less good 

 by 24. percent with boxes, painted black on the inside and by 34 percent less good with 

 boxes kept in the open but completely darkened. 



In addition to the action of light, the better visibility of food — in light exposed 

 boxes — plays undoubtedly a great role. 



rTithin 2 to 3 weeks (at a temperature of 12 degrees centigrade) a differentiation 

 in growth becomes noticeable. If the "nursing" is continued, a sorting out of the brood- 

 lings becomffl necessary. Sorting apparatus or nets of different sized meshes are used. 

 Individuals, large or small, may be selected by hand. 



The "nursing" of the brood has its decided advantages as shown by numerous experiments. 

 In a hatchery under intensive culture, the losses among unnursed fry amounted to 62 percent 

 as against 48 percent in the nursed stock (under otherwise like conditions). 



In experiments at the Forest Academy in Eberswalde with fry from brook trout, kept in 

 natural ponds, I had losses of only 40 percent with nursed fry, as against losses of 80 

 percent of unnursed fry. Losses among some 1,000 "nursed" fry (in various natural ponds) 

 amounted — in the fall of 1931 — to 33 percent among the non-sorted stock, to 52 percent 

 among especially small selected stock and to 32 percent among especially large selected 

 stock. 



There now remain for brief discussion, both of the named methods of fingerling 

 rearing in ponds. In the first method, the natural rearing, about 2 individual brood- 

 lings per square meter of water surface are set out. Besides larger natural ponds, carp 

 nursing ponds are advantageously stocked with rainbow trout brood along with carps. In 

 case of long "nursing", only half the stock is set in. If the pond is too fished-out 

 after three to four months, then three broodlings per square meter are set in. These 

 figures are average values. Naturally, fertility and local conditions must be considered. 

 In the Trout Fishery at Fuerstenberg in Vfestphalla, it is customary' (according to 

 Schaeperclaus) to first stock the ponds with brook trout brood, then to fish these out 

 as 4-6 cm. length "June fingerlings", and then from June to September to grow similar 

 rainbow trout fingerlings in them. No "nursery feeding" is done, and the average loss 

 is 60 to 70 percent, according to Schaeperclaus, 



In the second method, the "nursed" or feed competent brood is set in small ponds 

 (Fi^. 37) of 0,75 to 1.00 meter depth, not too strong flawing if there is danjer of 

 gyrodactjlus, at a density of about 100 fry per square meter. Artificial feeding, mostly 



122 



