If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Maintenance energy is the energy required to conserve the state of an animal when no work is completed. Dietary energy must be supplied to meet maintenance requirements before milk can be produced. The objectives of the current experiment were to quantify the maintenance energy requirement of Jersey cows when lactating or dry. Energetic measures were collected on 8 Jersey cows and evaluated across 3 physiological phases and nutritional planes: lactation, dry cows fed at maintenance, and fasted dry cows. Through total collection of feces and urine as well as using headbox-style indirect calorimeters, energy balance and heat production data were measured across all phases. Lactation data were collected across four 28-d periods. Data for cows fed at maintenance were collected after 14 d and fasting heat production was measured during the last 24 h of a 96-h fast. Net energy for maintenance (NEM) requirements, and the efficiency of converting metabolizable energy (ME) into net energy were compared between lactating and dry (maintenance or fasting phase) cows. Heat production of dry cows fed at maintenance, which represents ME for maintenance, was 0.146 ± 0.0087 Mcal per unit of metabolic body weight (BW0.75, MBW). Fasting heat production, which represents NEM, was 0.102 ± 0.0071 Mcal/MBW. Energy balance was calculated as tissue energy plus milk energy. When estimated via regressing energy balance on ME intake, NEM was not different between dry and lactating cows (0.120 ± 0.32 vs. 0.103 ± 0.0052 Mcal/MBW). However, the slope of the regression of energy balance on ME intake was greater for dry compared with lactating cows (0.714 ± 0.046 vs. 0.685 ± 0.010) when evaluated with a fixed intercept. This suggests that dry cows were more efficient at converting ME into net energy and that the efficiency of utilizing ME for maintenance may be greater than for lactation. Our measurements of NEM and the slope of ME on energy balance were greater than the value used by the National Research Council (2001), which are 0.080 Mcal/MBW for NEM and approximately 0.64 for the slope. Results of this study suggest that NEM and the efficiency of converting ME into NEM of modern lactating Jersey cows are similar to recent measurements on modern Holstein cows and greater than previous measurements.
Maintenance energy is the energy required to conserve the state of an animal when no work is performed or no products, such as meat or milk, are formed (
). Maintenance energy expenditure can be divided into 3 categories: (1) work and service functions, such as resorption of ion in the kidney, respiration, heart work, nervous tissue functions, and liver functions; (2) membrane transport; and (3) synthesis of cellular components (
) Mcal per unit of metabolic BW (BW0.75; MBW). For a mature Jersey cow weighing 450 kg, the difference between an NEM of 0.073 or 0.124 Mcal/MBW would yield a difference of 5.0 Mcal/d.
suggested that NEM might have increased over time in lactating cows and is associated with increased DMI and milk yield. Changes in the mass of tissue with a high metabolic activity per unit of weight such as the heart, liver, kidney, and gastrointestinal tract can affect maintenance energy expenditure, and increased mass and energy expenditure of these tissues likely occur with increased DMI and milk yield in lactating compared with nonlactating animals (
Effects of diet forage-to-concentrate ratio and intake on energy metabolism in growing beef heifers: Whole body energy and nitrogen balance and visceral heat production.
The point estimate of NEM of a lactating cow is theoretical because it cannot be directly measured. However, NEM can be estimated via 1 of 2 methods: by measuring heat production (HP) at fasting or via regression of energy balance on ME with the y-intercept representing NEM. Published estimates of NEM (Mcal/MBW) of fasted dairy cows are (Mcal/MBW) 0.073 (
Use of indirect calorimetry to evaluate utilization of energy in lactating Jersey dairy cattle consuming diets with increasing inclusion of hydrolyzed feather meal.
The metabolisable energy requirement for maintenance and the efficiency of utilisation of metabolisable energy for lactation by dairy cows offered grass silage-based diets.
observed that the fasting HP of cows immediately following lactation was 9% greater than 31 d after lactating ceased. Because of the abrupt cessation of lactation, this increase in fasting HP may have been a result of mammary involution. To our knowledge, the most recent study where fasting HP was measured in dairy cows was published 20 yr ago (
) and no known measurements have been reported on Jersey cows. In addition, we are unaware of estimates of NEM on the same group of dairy cows during lactation and after drying off while fasted. Therefore, the objective of this work was to quantify maintenance energy requirements of Jersey cows that are lactating or dry. Because of genetic improvements, we hypothesize that our measures of NEM will be greater than what is used by
. Additionally, we expect NEM to be greater when estimated on lactating cows compared with dry cows.
MATERIALS AND METHODS
Animals and Treatments
The University of Nebraska–Lincoln Animal Care and Use Committee approved animal care and experimental procedures of this study (project # 1894). Eight multiparous Jersey cows sourced from a commercial dairy were used. Cows were housed in individual tiestalls equipped with rubber mats in a temperature-controlled (20°C) barn at the Dairy Metabolism Facility in the Animal Science Complex at the University of Nebraska–Lincoln. All cows were confirmed nonpregnant by blood test for pregnancy-specific protein B (
The experiment consisted of 3 phases: lactation, dry maintenance feeding, and dry fasting (Figure 1). For the lactation phase, data were used from a study conducted in the same 8 cows in which measurements were collected across 4 periods on each cow (n = 32 total; Morris and Kononoff, 2021). A full description of the procedures used to evaluate energy utilization can be found elsewhere (Morris and Kononoff, 2021). For variables that were measured on both dry and lactating cows, methods were identical. According to the objectives of Morris and Kononoff (2021) cows were fed diets that varied in carbohydrates (18.9 to 30.8% starch or 35.9 to 25.5% NDF), fatty acids (2.90 to 6.80% fatty acids), or supplemental Lys (0 to 16 g/d of supplemental digestible Lys).
Figure 1Illustration of time course and samples collected during the experiment.
Before the start of the maintenance feeding phase, cows were dried off. The maintenance feeding phase was 18 d in length. All cows were fed the same TMR (Table 1) to meet maintenance energy requirement, which was assumed to be 0.100 kcal of NEL/MBW (
estimates of dietary NEL. Feed offerings were based upon the most recent BW measure. Dietary ingredients for the base diet (corn silage, alfalfa hay, and concentrate) were placed in a Calan Data Ranger (American Calan Inc.), and subsequently used to mix and deliver TMR once daily at 0930 h. During the final 4 d of the maintenance feeding period, energy balance was determined. Following the maintenance feeding period, cows were fasted for 96 h (Figure 1). Throughout all phases, cows were allowed ad libitum access to water.
Table 1Ingredient and chemical composition of diet fed to Jersey cows during maintenance feeding period (% of DM unless otherwise indicated)
Effects of rumen-protected lysine and histidine on milk production and energy and nitrogen utilization in diets containing hydrolyzed feather meal fed to lactating Jersey cows.
). Briefly, DMI, milk yield (lactation only), as well as fecal and urine output were determined for 4 consecutive days. Catheters were inserted into the bladder were used to separate feces and urine. In urine collection containers, urine was acidified (approximately 400 mL for dry cows and 1,000 mL for lactating cows) to maintain a pH <5.0 with 50% HCl. Daily samples of feed ingredients, feed refusals, milk (lactation only), feces, and urine were collected and composited on a wet-weight basis. All samples were stored at 4°C until further analysis.
Feeds, refusals, and feces were dried at 60°C for 48 h to determine DM and for further analysis. Dry feeds, refusals, and feces were ground to pass a 1-mm screen (Wiley Mill, Arthur A. Thomas Co.), and analyzed by Cumberland Valley Analytical Services Inc. for N (Leco FP-528 N Combustion Analyzer, Leco Corp.), NDF with sodium sulfite (
) by Cumberland Valley Analytical Services Inc. Dry ground samples were analyzed for gross energy (GE) content (Parr 6400 Calorimeter) in the nutrition laboratory of the University of Nebraska–Lincoln. The chemical composition of the maintenance diet is listed in Table 1. Urine energy was determined via bomb calorimetry after drying (60°C) approximately 4 mL of sample in a bomb capsule until dry (4 h). Milk energy was calculated from yield of milk fat, protein, and lactose (
Headbox-style indirect calorimeters were used to measure O2 consumption and CO2 and CH4 production (Figure 2). Total volume of gas flow through the headbox was measured using a mass flow meter (MCW Whisper, Alicat Scientific), and using a data logger (model XR440, Pace Scientific Inc.) flow rate was recorded every minute. From the headbox, continuous samples of incoming and outgoing air were collected into separate bags (44 L, LAM-JAPCON-NSE, Pollution Measurement Corp.) using glass tube rotameters (model 1350E Sho-Rate “50,” Brooks Instruments). Gas bags were analyzed for O2, CO2, and CH4 using a gas analyzer (X-stream, Emerson Process) according to the method of
. To calculate O2 consumption and CO2 and CH4 production, difference between incoming and outgoing gas was multiplied by total air flow. Heat production was calculated using the
Prediction of urinary nitrogen and urinary urea nitrogen excretion by lactating dairy cattle in northwestern Europe and North America: A meta-analysis.
). While each cow was inside the headbox, free water intake was measured using a water meter (model DLJSJ75, Daniel L. Jerman Co.).
Figure 2Illustration of headbox-style indirect calorimeter. The top panel show air flow through the system. Air is pulled into the headbox and through the exhaust system by a blower motor. Before entering the exhaust system, air is filtered. A fraction of the exhaust air is recirculated throughout the headbox, which ensures adequate mixing of incoming air and respiratory gases from the animal. The remaining exhaust air exits the headbox, where it is passed through an air flow meter. Positive pressure is created by the air exhaust valve, which allows for diversion of a small fraction of the exhaust air to sample bags. (Illustration by Sara L. Taliaferro, Happy Beetle Studio, Lawrence, KS).
Before the start of the experiment and upon completion, efficiency of gas collection via headboxes was determined by burning 100% ethyl alcohol and measuring O2 consumption and CO2 production relative to expected value calculated from alcohol disappearance. Consumption of O2 and production of CO2 were (average ± SD) 101 ± 0.7 and 99 ± 1.6% of expected.
Gas measurements were collected during each of the 3 phases (Figure 1). During the lactation phase, cows were placed in headboxes for 1 d. During the maintenance feeding phase, each cow was placed in headboxes for 2 consecutive days. During the fasting phase, cows were placed in headboxes 72 h subsequent to the initiation of feed restriction. While cows were in the headboxes, a collection period of 23 h was used to measure O2 consumption and CO2 and CH4 production for cows on all phases. The remaining hour was used for sample processing and to prepare to the subsequent gas collection. All gas data were adjusted to a 24-h period. Cows were adapted to headboxes for a minimum of 3 d before the start of the experiment. For the lactation and maintenance feeding phase, feed was placed in the bottom of the headbox. Cows were allowed ad libitum access to water from a water bowl placed inside the headbox, and free water intake was measured as described above.
Energy Calculations
Respiratory quotient (RQ) was calculated using the ratio of CO2 produced to O2 consumed. Energy loss as CH4 was estimated by multiplying CH4 production by its enthalpy (9.45 kcal/L). Calculations for digested energy (DE) and ME were as follows:
DE (Mcal/d) = GE (Mcal/d) – fecal energy (Mcal/d),
[1]
ME (Mcal/d) = DE (Mcal/d) – urine energy (Mcal/d) – methane energy (Mcal/d).
[2]
Unaccounted energy was assumed to represent tissue energy (TE):
TE (Mcal/d) = ME (Mcal/d) – HP (Mcal/d) – milk energy (Mcal/d).
[3]
To compare NEM and the efficiency of converting ME into NEL (kL) between lactating cows and cows fed at maintenance or fasting (termed dry cows), regression were conducted as described by
Use of indirect calorimetry to evaluate utilization of energy in lactating Jersey dairy cattle consuming diets with increasing inclusion of hydrolyzed feather meal.
), except the term energy balance was used to represent the fact that dry cows were not producing milk. For this, ME was corrected for the ME that was used for tissue retention and energy balance was corrected for heat increment from tissue mobilization. Corrected ME and energy balance were calculated as follows:
corrected ME = ME (kg/d) − |positive TE (Mcal/d)|/kG,
[4]
corrected energy balance = −HP (Mcal/d) for fasting cows or milk energy (Mcal/d) − |negative TE (Mcal/d)| × kT,
[5]
where kT is the efficiency of utilizing body reserve energy for milk production (0.75), kG is the efficiency of utilizing ME intake for tissue gain (0.89;
Corrected energy balance was regressed against the fixed effect of stage (lactating or dry) and the interaction of stage with corrected ME using the “lmer” package in R (v3.6.3; https://www.r-project.org/). Cow was included as a random effect. Differences between regression intercept values (representing NEM) between stage was tested via the effect of stage. Differences between slope values (representing kL) between stage were tested via the interaction of stage with corrected ME.
RESULTS AND DISCUSSION
Maintenance Phase
Respiratory quotient is generally considered to be a gross indicator of whole-body substrate oxidation. Production formation from the oxidation of lipids, protein, and carbohydrates results in an RQ of 0.71, 0.81, and 1.00, whereas lipid synthesis results in an RQ greater than 1 (
). In the current experiment during the maintenance phase, RQ was 0.942 ± 0.014 (Table 2), suggesting that cows were utilizing glucose as an energy source and to a lesser degree some fat, protein, or both. Additionally, RQ was positively correlated with TE (r = 0.80; P = 0.02). This likely occurred due to greater lipid oxidation in cows in negative energy balance and greater lipid synthesis in cows in positive energy balance. Previously, increased RQ has been observed in lactating cows when feeding diets that likely result in increased lipid synthesis (
Energy and nitrogen partitioning in dairy cows at low or high metabolizable protein levels is affected differently by postrumen glucogenic and lipogenic substrates.
When cows were fed at maintenance, average TE was 0.3 ± 1.1 Mcal/d (Table 2), and thus we concluded that NEL supply was near maintenance requirements. Heat production of cows fed at maintenance represents the maintenance energy requirements on an ME basis (MEM;
). In the current study, HP of cows when fed at maintenance was 0.146 ± 0.0087 Mcal/MBW. In a retrospective analysis of 1,038 observations from 284 cows, MEM on average was 0.126 Mcal/MBW (
). However, time was observed to have an effect; observations of MEM increased from 0.121 to 0.177 Mcal/MBW from 1963 to 1973 to 1984 to 1995, respectively (
). Increased MEM from 1963 to 1973 and 1984 to 1995 was suggested to have occurred because of increased milk yield. In an analysis of data from 221 lactating Holsteins, MEM averaged 160 kcal/MBW (
The metabolisable energy requirement for maintenance and the efficiency of utilisation of metabolisable energy for lactation by dairy cows offered grass silage-based diets.
The metabolisable energy requirement for maintenance and the efficiency of utilisation of metabolisable energy for lactation by dairy cows offered grass silage-based diets.
) were much larger than the current experiment (hundred to thousands compared with 8 observations), these databases primarily included observations from lactating cows fed above maintenance. As described by
, energetic parameters are not exclusively independent, and the recursive relationship should be considered. Consequently, examining MEM without considering NEM and the relationship between the two (i.e., kL) can be misleading. For example, increasing dietary forage inclusion is associated with an increase in NEM (
The metabolisable energy requirement for maintenance and the efficiency of utilisation of metabolisable energy for lactation by dairy cows offered grass silage-based diets.
Effects of diet forage-to-concentrate ratio and intake on energy metabolism in growing beef heifers: Whole body energy and nitrogen balance and visceral heat production.
Effects of diet forage:concentrate ratio and metabolizable energy intake on visceral organ growth and in vitro oxidative capacity of gut tissues in sheep.
During fasting, cows lost 6.5 ± 1.7 kg/d of BW (Table 3). This BW loss is because cows were in negative energy balance, and to some extent, loss was due to a decrease in gut fill.
Table 3Descriptive statistics for data set of fasting Jersey cows
Because RQ represents whole-body substrate oxidation, it is an indicator of whether or not a physiological state of fasting has been reached, as an observation of 0.71 is believed to represent a true fasting state (
). In the current experiment, RQ was 0.726 ± 0.011 during the fasting phase (Table 3), which suggests that all animals were at or very near this fasting state.
measured RQ and HP throughout a 120-h fast and reported that although RQ continued to decrease slightly after 72 h of fasting, HP was constant at 0.108 Mcal/MBW. Additionally, in the current experiment, CH4 production during the fasting phase was 18 ± 5 L/d, which suggests minimal rumen fermentation and is much lower than CH4 production for cows fed at maintenance (171 ± 17 L/d; Table 2). Similarly,
), which has likely led to increased NEM requirements. Most historical measurements of fasting HP have been completed on Holsteins. Similarity between our measurements of fasting HP and those reported by
suggests that NEM requirements are similar for modern Jersey and Holstein cows.
Comparison of Maintenance and Efficiency in Lactating and Dry Cows
In the current study, comparisons between estimates of maintenance and efficiency of converting ME into NE for lactating and dry cows were made using data from the same 8 cows in the lactation immediately before dry-off. These cows were on average 141 ± 32 DIM, with a DMI of 19.2 ± 2.3 kg/d and ECM of 35.5 ± 4.4 kg/d (Table 4). Additionally, by treatment design, diets fed to lactating cows included a range in starch (minimum to maximum; 18.9 to 30.8% DM), NDF (25.5 to 35.9% DM), and fatty acids (2.90 to 6.80% DM).
Table 4Descriptive statistics for data set of lactating Jersey cows
When regressing corrected energy balance on corrected ME, the derived estimate of NEM was not different between lactating and dry cows (0.120 ± 0.032 vs. 0.103 ± 0.0052 kcal/MBW; P = 0.59; Figure 3A). Standard error for estimating NEM for lactating cow was 6 times that of dry cows (0.032 vs. 0.0052), which was likely a response from the large extrapolation between the minimum ME intake for lactating cows (0.406 Mcal/MBW; Table 3) and the y-intercept. We previously observed a similar large extent of variance (22% of mean) when estimating NEM using the same methods (
Use of indirect calorimetry to evaluate utilization of energy in lactating Jersey dairy cattle consuming diets with increasing inclusion of hydrolyzed feather meal.
). Nevertheless, our data suggest that NEM estimated via regression or as fasting HP do not differ.
Figure 3Regression of energy balance (EB; tissue energy plus milk energy) corrected for the heat increment of tissue mobilization, and ME intake corrected for tissue retention (see Materials and Methods). Both variables are expressed per unit of metabolic body weight (BW0.75, MBW). The intercept, which represents the inverse of NEM, was not different (P = 0.59) between lactating and dry cows (panel A) and was thus removed (panel B). Average NEM was 0.104 ± 0.0051 Mcal/MBW. The slope of each regression represents efficiency of utilizing ME intake for milk or maintenance (kL), which was greater for dry cows compared with lactating cows (0.714 ± 0.046 vs. 0.685 ± 0.010; P < 0.01; panel C). A regression with all data pooled was analyzed (panel C). From this, kL was 0.682 ± 0.0091 and NEM was 0.102 ± 0.0044 Mcal/MBW. Dry = dry cows; Lactating = lactating cows; RMSE = root mean square error.
A second regression was fit with a common intercept for lactating and dry cows because intercepts did not differ. Because a strong negative correlation (r = − 0.61, P < 0.01; data not shown) was observed between intercept and slope coefficients, fitting a common intercept is more appropriate for comparing slope coefficients. Based on previous work the efficiency of utilizing ME for maintenance (0.62) was nearly identical to NEL (0.64) (
assumes a common efficiency for utilizing ME for maintenance and lactation. From the current experiment, the regression slope (i.e., kM or kL) was greater for dry cows compared with lactating cows (0.714 ± 0.046 vs. 0.685 ± 0.010; Figure 3B) when evaluated with a fixed intercept. This suggests that cows may have been more efficient at utilizing ME for maintenance purposes than for lactation. The efficiency between maintenance and fasting (kM) is a measure of the relative efficiency of dietary energy compared with body TE to meet the maintenance needs. This relationship is not similar to the efficiency of utilizing energy above maintenance to generate end products such as meat or milk (
reported that the kM of using a mixture of VFA to meet maintenance energy requirements was 0.68 but was high as 0.82 if only butyrate or propionate were supplied to meet maintenance requirements. In general, these values are similar to the kM we observed. In lactating cows, kL includes heat increment from meeting maintenance requirement plus the efficiency of converting ME into milk fat, protein, and lactose. We recently reported that HP associated with milk protein synthesis is 6.2 Mcal/kg, which translates to a partial efficiency of 0.48 (
, as ME intake increases the overall efficiency approaches the kL due to a dilution of the proportion of ME used for maintenance. This was also observed in the current experiment as kL for the whole data set was similar to kL for lactating cows only (0.682 ± 0.0091 vs. 0.685 ± 0.010).
Measurements of efficiency of utilizing ME for maintenance compared with lactation are confounded by many factors and our measures are no exception. For example, maintenance and fasting phases were conducted following the lactation measurements. However, measurements were collected in a climate-controlled environment using identical techniques and personnel; thus, potential confounding effects of sampling period should be minimized. Energy requirements are expressed relative to MBW and dry or fasting cows may have higher values per unit of MBW because of less gut fill and no milk weight. Diets differed between lactating and fasting cows, and diets are well established to affect the efficiency of utilizing ME for lactation (
Use of indirect calorimetry to evaluate utilization of energy in lactating Jersey dairy cattle consuming diets with increasing inclusion of hydrolyzed feather meal.
). Consequently, we concede observed differences in efficiency for lactating and dry cows may be influenced by the diets fed. Specifically dry cows were fed diets with greater forage NDF (39.3 vs. 18.6% DM; data not shown); however, this would be expected to decrease efficiency of utilizing ME (
Effects of diet forage-to-concentrate ratio and intake on energy metabolism in growing beef heifers: Whole body energy and nitrogen balance and visceral heat production.
). Additionally, dry cows were fed a diet with less CP (12.3 vs. 15.7% DM), which should result in a lower heat increment associated with catabolism of excess CP (
). In the field, dry and lactating cows are almost always fed vastly different diets with dry cow diets being higher in forage and lower in CP compared with lactation diets. More research with different diets is needed to confirm our finding that the efficiency of utilizing ME for maintenance is greater than lactation.
reported that kL has increased over time from 0.60 to 0.68, and is positively associated with milk yield, heart rate, and dietary ether extract. Therefore, it is plausible that kL for modern Jersey cows is greater than historic estimates (
). Therefore, both coefficients must be considered simultaneously. In agreement with our data, recent evaluation of historical databases generated similar NEM and kL values as the current experiment (
). Although NEM is likely greater than previous estimates, increased kL will counter this increase in nonproductive energy expenditure. For example, in a 450-kg Jersey cow with a ME intake of 45.8 Mcal, NE will be 2.2 Mcal greater when using kL from the current experiment compared with
and NEM will also be 2.2 Mcal greater. Therefore, energy available for milk and tissue (NEL − NEM) will be computed to be similar across the current estimates for NEM and kL compared with those used by
. Derivation for NEM determined using lactating cow data did not differ from dry cows. The efficiency of utilizing ME in dry cows was greater than the efficiency of utilizing ME in lactating cows (0.721 vs. 0.683), which suggests that ME may be used with a greater efficiency for maintenance than for lactation. Although NEM may be greater for modern lactating dairy cows, increased kL compared with previous research (0.679 vs. 0.63) annul the effect of increased NEM on energy available for milk and tissue production.
ACKNOWLEDGMENTS
The authors thank the University of Nebraska Dairy Metabolism (Lincoln, NE) staff and students for care of the experimental animals and assistance with collections, Kelly Heath (University of Nebraska, Lincoln) for monitoring animal health while fasting, and the American Jersey Cattle Association's AJCC Research Foundation (Reynoldsburg, OH) for partial funding. The authors state no conflicts of interest.
Effects of diet forage:concentrate ratio and metabolizable energy intake on visceral organ growth and in vitro oxidative capacity of gut tissues in sheep.
Use of indirect calorimetry to evaluate utilization of energy in lactating Jersey dairy cattle consuming diets with increasing inclusion of hydrolyzed feather meal.
Effects of rumen-protected lysine and histidine on milk production and energy and nitrogen utilization in diets containing hydrolyzed feather meal fed to lactating Jersey cows.
Energy and nitrogen partitioning in dairy cows at low or high metabolizable protein levels is affected differently by postrumen glucogenic and lipogenic substrates.
Effects of diet forage-to-concentrate ratio and intake on energy metabolism in growing beef heifers: Whole body energy and nitrogen balance and visceral heat production.
Prediction of urinary nitrogen and urinary urea nitrogen excretion by lactating dairy cattle in northwestern Europe and North America: A meta-analysis.
The metabolisable energy requirement for maintenance and the efficiency of utilisation of metabolisable energy for lactation by dairy cows offered grass silage-based diets.
00668-8&id=gr1.jpg){kind=link}
00668-8&id=gr2.jpg){kind=link}
00668-8&id=gr3.jpg){kind=link}