There is some debate concerning the degree of availability of organic vs. inorganic forms of trace minerals. The answer depends on the specific mineral, the dietary conditions and the physiologic state of the animal. It is clear that for certain minerals (e.g., selenium, chromium or iron), organic forms are better utilized than inorganic forms (Hallberg and Rossander-Hulthen 1993, Mertz and Roginski 1971). For other minerals (e.g., zinc or copper), it is perhaps less clear. For example, there are as many studies that have failed to show increased bioavailability with organic zinc forms as there are studies demonstrating improved bioavailability. It has been demonstrated in livestock and fish (Paripatananont 1994, Wedekind et al. 1992 and 1994) that growth rate, calcium and phytate levels are factors that significantly affect zinc utilization and thus determine whether the use of organic zinc sources is beneficial. Viewing these data as a whole, results suggest that there is little or no benefit to be gained in using organic zinc in diets low in phytate and calcium (low calcium defined as calcium levels approximating NRC recommendations for respective species). However, as phytate and/or calcium increase, or demand for zinc increases (e.g., fast growth rate), zinc utilization is greater with organic zinc sources (parameters used to assess zinc bioavailability included one or more of the following: bone. zinc, immune response and growth rate). Furthermore, the faster growing the animal, the greater the benefit demonstrated for organic zinc [e.g., fish > chicks (broiler breed > leghorn-type) > pigs]. The efficacy decreased as the animal's age increased, suggesting that there may be less benefit in using organic zinc sources in foods for adult animals. The purpose of this study was to compare the bioavailability of an organic zinc source (zinc propionate) with that of an inorganic zinc source (ZnO) in puppies. These two zinc sources were compared under three dietary conditions as follows: 1) at 10 g Ca/kg diet [the Association of American Feed Control Officials (1997) recommendation for Ca during growth]; 2) 15 g Ca/kg diet: and 3) 15 g Ca/kg diet + 50 g/kg beet pulp. The two Zn sources were compared under varying levels of Ca and fiber to determine whether fiber or excessive Ca had differing effects on the bioavailability of Zn from organic or inorganic Zn. Materials and methods. Forty-two puppies (9 wk mean age at study start) were fed a Zn-deficient egg white/cornstarch/sucrose diet (Table 1) for 2 wk. At the end of this pretest period, puppies were allotted to one of seven dietary treatments. There were three replicates per treatment with two puppies per pen. The puppies were blocked by litter, sex and weight. Within each block, each treatment had a similar mean initial weight and weight distribution. Puppies were housed in stainless steel pens and were allowed free access to food and water. Puppies were fed the experimental diets for 3 wk. However, puppies receiving the control diet were removed from the test if signs of Zn deficiency were severe, and only one of the six puppies receiving that treatment (treatment 1) could continue through d 35 of the study. The experimental protocol was reviewed and approved by the institutional Animal Care and Use Committee. Feed consumption was measured daily and body weights measured weekly. The basal diet (Table 1) was formulated to be adequate in all nutrients except Zn (NRC 1985). Dietary additions of Zn, Ca and beet pulp were made at the expense of cornstarch with Zn added as Zn propionate (ZnP; KemZIN; provided by Kemin Industries, Des Moines, IA) or ZnO and Ca provided as CaC[O.sub.3]. Cellulose was present n the diet to improve stool quality, but has no antagonistic effects on Zn bioavailability and also contributes no phytate to the diet (Wedekind et al. 1995). Treatment 1 was the control with 10 g Ca/kg diet and no added Zn (5.4 mg Zn/kg diet, analyzed); treatments 2 (ZnP) and 3 (ZnO) contained 10 g Ca/kg diet; treatments 4 (ZnP) and 5 (ZnO) contained 15 g Ca/kg diet; and treatments 6 (ZnP) and 7 (ZnO) contained 15 g Ca/kg diet plus 50 g/kg beet pulp. The Zn-supplemented treatments were intended to be equal in zinc concentration (40 mg/kg added Zn); however, actual analyzed Zn concentrations (mg Zn/kg diet) differed between the ZnO and ZnP treatments as shown in Table 2. The basal diet was analyzed for moisture, protein, fat, fiber, ash, Ca, P and Zn; other treatments were analyzed for moisture, Ca and Zn. Blood samples were taken weekly during the pretest period and at the end of the experimental period (i.e., wk 0, 1, 2 and 5). Blood was collected using special-purpose (low Zn) trace element evacuated tubes. At the end of the experiment, dewclaws, one canine deciduous tooth and testes were collected for Zn analysis. The dewclaws were autoclaved to facilitate removal of skin and connective tissue. The bones, teeth and testes were dried at 105 [degrees] C overnight. All samples were dry-ashed overnight in a muffle furnace (650 [degrees] C for bone and teeth samples; 450 [degrees] C for testes) before preparation for Zn analysis. Zn was determined in all tissues by inductively coupled plasma-emission spectrophotometry. Bioavailability of Zn was determined for the Zn propionate and ZnO treatments via multiple regression slope ratio analysis. Plasma Zn was regressed on supplemental Zn intake by comparing the change in plasma Zn between d 35 and 14 (end of Zn-deficient pretest period) for treatments 2-7. The model included block and the amount (milligrams) of Zn fed from each of the Zn sources such that six regression lines with a common intercept were determined. There were six observations per treatments 2-7; treatment one was not used), yielding six data points for the construction of a line of best fit. The slopes were compared with the standard [arbitrarily set as treatment 2, (ZnP at 10 g Ca/kg diet)]. Analyzed Zn values from treatments 2-7 were used in the regression analysis; thus discrepancies in dietary Zn levels between treatments were corrected for. Figure 1 shows a graphical representation of the data. The intercept was calculated by assuming that block effects summed to zero. The slopes of treatments 2-7 were statistically compared using a heterogeneity of slopes procedure to determine if significant differences in slopes existed. Differences between individual slopes were then tested using t tests among all pairs only if F-tests for differences among slopes were significant (P < 0.05). All ANOVA and regression analyses were performed using the General Linear Modes procedure of SAS (SAS Institute, Cary, NC). Probability values of P < 0.05 between treatments were considered significant. Results. Plasma Zn decreased over time during the Zn-deficient pretest period from an average of 11.32 [+ or -] 0.40 (SEM) to 4.44 [+ or -] 0.18 [micro]mol/L. Signs of Zn deficiency occurred as early as 5 d into the pretest period. The symptoms were suggestive of parakeratosis (e.g., crusted plaques with erosions on face and feet). Other symptoms included lethargy and anorexia. No differences in weight gain were seen between the Zn-supplemented treatments, but puppies fed treatment 1 (no added Zn) lost weight over the 3-wk experimental period. Because of small sample size (i.e., < 0.1 g average), the analytical variation was high for Zn concentrations in dewclaws, teeth and testes (Table 2). Other studies have shown bone to be a sensitive measure of Zn status (Wedekind et al. 1992), but sample size in those studies exceeded 1 g. Despite the higher Zn intakes for puppies fed the ZnO treatments, the plasma Zn concentrations tended to be higher for puppies fed the Zn propionate treatment under each of the three dietary regimens (10 g/kg Ca, 15 g/kg Ca and 15 g/kg Ca + 50 g/kg beet pulp). With increasing dietary Ca, Zn concentrations in plasma, teeth and testes tended to decline. This tendency was observed for bot h Zn sources. When plasma Zn was regressed on supplemental Zn intake, thus correcting for differences in dietary Zn intake, significant differences (P < 0.05) in Zn bioavailability were observed between Zn propionate and ZnO (Table 3). When dietary Ca was at 10 g/kg, the bioavailability of Zn in Zn propionate was 80% greater (i.e., 0.01041/0.00575 = 1.8) than the Zn in ZnO. The bioavailability of both Zn sources decreased when dietary Ca was increased to 15 g/kg, but an advantage was seen for Zn propionate vs. ZnO regardless of dietary Ca level and/or presence of beet pulp. Zn utilization was not significantly affected by beet pulp addition. Discussion. The results of our study showed that the bioavailability of Zn in Zn propionate is 60-80% greater than that of the Zn in ZnO. However, it is clear from other studies (Wedekind et al. 1992 and 1994) that the Zn in ZnS[O.sub.4] is approximately twofold more available than that in ZnO. Thus, it would appear from this study that the efficacy of Zn propionate, an organic Zn source, is no greater than that of certain other inorganic Zn sources. Other studies have been conducted in dogs comparing Zn bioavailability between organic and inorganic Zn sources. Lowe et al. (1994) found that organic Zn sources resulted in greater Zn retention, higher concentrations of Zn in hair and greater hair growth compared with ZnO. It should be noted that the diet in that study did contain phytate (soybean meal was in the diet at 16%) and Ca levels were high (3.2%). Slope-ratio analysis was not used in that study; however, if one compares the hair Zn concentrations between the Zn sources as a ratio (i.e., 21.09/10.83), then a 95% advantage in Zn bioavailability is yielded for the zinc amino chelate vs. ZnO at 1.2% dietary Ca. This is similar to the 60-80% advantage we saw in this study between organic and inorganic Zn sources. When the zinc amino acid chelate is compared with ZnO at 3.2% dietary Ca, that advantage increased to 290% (21.15/7.28). The difference in Ca levels and phytate content between that study and this one may explain the greater bioavailability estimate obtained (at 3.2% dietary Ca) from the earlier study. Thus, in conclusion, organic Zn sources are efficacious only under certain dietary conditions and may not yield significant improvements in pet food diets that are low in Ca and/or phytate. KEY WORDS: * zinc * bioavailability * puppies * calcium * organic zinc