FISHERY BULLETIN: VOL. 78, NO. 3 



Table 5. — Manipulations of guppy populations at the end of 

 Phase I (week 36» for test of refuge fence spacing treatment 

 effect during Phase II (weeks 36-58). 



Tank Refuge 



number assignment (mm) Origin 



1 5.0 No. 1 and 2, mixed and reduced' 



2 5.5 No. 1 and 2. mixed and reduced' 



3 5.0 No. 3, reduced 



4 4.5 No 4 and 10, mixed and reduced^ 



5 5.0 — "Long-term control" 



6 5.0 — "Long-term control" 



7 4.5 No, 1 and 2, mixed and reduced' 



8 5.5 No. 4 and 10. mixed and reduced 



9 4.5 No. 9, reduced 



10 5.0 No. 4 and 1 0. mixed and reduced' 



11 5.5 No. 11, reduced 



12 5.0 — "Long-term control' 



'Pluseightfry from tank no. 3. 



Plus five fry and four A5 5 (an adult male size category, refer to Table 1) 

 males from tank no. 11. 



Table 6. — Schedule of daily rations fed to treated guppy pop- 

 ulations during Phase II (weeks 36-58). Total is expressed as a 

 wet weight equivalent. 



Dates 



(1976) 



Dry 

 food 

 Weeks (g) 



Anemia 

 adults 



(g) 



Artemia 

 nauplii 



(g) 



Total 



(g) 



17 Feb. -28 Feb. 



1 (Vlar.-15Mar. 



16IVIar,-18 Apr. 



36-38 

 38-40 

 40-58 



0,15 

 0,18 

 0,20 



1,5 

 1,8 

 2,0 



0.20 

 0.20 

 0.20 



3.2 

 3.8 



4.2 



'To determine, multiply dry food by 10 and add all ration types. 



Since both mixed and single species populations 

 were fed at equal ration levels relative to total 

 population biomass, it seems reasonable to infer a 

 strong competitive advantage to guppy fry in 

 securing food. 



In contrast, the presence of platies apparently 

 had little effect on guppy populations in mixed 

 populations. Tests (Student's t) for differences in 

 mean guppy population numbers and biomass 

 between single species and mixed species popula- 

 tions at week 8 failed to reject, at the a = 0.05 

 level, null hypotheses of equality in population 

 numbers or biomass. After removal of platies, all 

 guppy populations were treated as (equivalent) 

 replicates. 



Weeks 8-.^6 



The 10 replicate guppy populations through 

 week 36 showed a striking contrast between 

 growth in numbers and growth in biomass. Bio- 

 mass steadily increased in all populations and 

 weights attained at each sampling period were 

 nearly equal in all populations. Populations 

 seemed near a common maximum supportable 

 biomass by week 36. In contrast, numerical growth 

 was highly variable and often occurred as discrete 

 pulses of increase. Total population numbers 



always varied greatly among populations and 

 there was no indication of a common approach to 

 an asymptotic or stable numerical population size 

 (Table 7). 



BIOMASS DYNAMICS. — In general, biomass 

 growth in all guppy populations was typified 

 by steady, nearly uniform, biweekly increments 

 through the first 28 wk and by declining incre- 

 ments during weeks 28-36. Removals of individual 

 females at weeks 11, 13, 15, and 17 and restriction 

 of rations to dried food only during weeks 28-29 

 were clearly reflected in depressed biomass incre- 

 ments during these intervals of disturbance (Fig- 

 ure 4). At week 36 mean population biomass was 

 28.906 g and ranged from 26.291 to 30.717 g. 



Using data for weeks 18-28 and 30-36 (periods 

 during which neither removals nor atypical feed- 

 ings occurred, and beyond the time when platies 

 were also present in certain populations), maxi- 

 mum supportable biomass for each population was 

 estimated by assuming a logistic biomass growth 

 model. Application of the logistic model presumed 

 that biomass growth was limited by the final fixed 

 ration level reached at week 22. Maximum bio- 



T 

 O 



T 

 A 



L 



B 

 I 



O 

 M 

 A 

 S 

 S 



g 



r 

 a 

 m 

 s 



12 18 



WEEKS 



Figure 4. — Mean population biomass of guppies during Phase 

 I. Removals of individual females at weeks 11, 13, 15, and 

 17 and restriction of ration to dried food only during weeks 28-29 

 are indicated by arrows. Vertical lines represent standard 

 deviations. 



562 



